WO1997040170A2 - Functional expression of heterologous adenylyl cyclase - Google Patents

Abstract

Heterologous adenylyl cyclases are functionally expressed in host cells. The subject host cells may be used to screen for inhibitors or activators of the adenylyl cyclase, or of a regulator of adenylyl cyclase which is functionally co-expressed in the yeast cell. Mutants of adenylyl cyclases are also disclosed.

Description

FUNCTIONAL EXPRESSION OF HETEROLOGOUS ADENYLYL CYCLASE

Background of the Invention .

The invention relates inter alia, to expression of a heterologous adenylyl cyclase in a host cell, the transformed host cells, and their use, e.g., in identifying potential inhibitors or activators of the heterologous adenylyl cyclase, or of other proteins which are natively or artificially coupled to the heterologous adenylyl cyclase in the engineered host cell.

In some instances, for a drug to cure a disease or alleviate its symptoms, the drug must be delivered to the appropriate cells, and trigger the proper "switches." The cellular switches are known as "receptors." Hormones, growth factors, neiirotrartsrnitters and many other biomolecules normally act through interaction with specific cellular receptors. Drugs may activate or block particular receptors to achieve a desired pharmaceutical effect Cell surface receptors mediate the transduction of an

"external" signal (the binding of a ligand to the receptor) into an "internal" signal (the modulation of a pathway in the cytoplasm or nucleus involved in the growth, metabolism or apotosis of the cell). In many cases, transduction is accomplished by the following signaling cascade:

• An agonist (the ligand) binds to a specific protein (the receptor) on the cell surface.

• As a result of the ligand binding, the receptor undergoes an allosteric change which activates a transducing protein in the cell membrane. • The transducing protein activates, within the cell, production of so-called

"second messenger molecules."

• The second messenger molecules activate certain regulatory proteins within the cell that have the potential to "switch on" or "off* specific genes or alter some metabolic process. This series of events is coupled in a specific fashion for each possible cellular response. The response to a specific ligand may depend upon which receptor a cell expresses. For instance, the response to adrenalin in cells expressing α-adrcnergic receptors may be the opposite of the response in cells expressing β-adrenergic receptors. The above "cascade" is idealized, and variations on this theme occur. For example, a receptor may act as its own transducing protein, or a transducing protein may act directly on an intracellular target without mediation by a "second messenger". Signals initiated by a variety of mammalian hormones and neurotransmitters are received by seven transmembrane domain receptors in the plasma membrane of cells and are transduced to intracellular effectors via heterotrimeric G proteins. Many different G proteins are known to interact with receptors. G protein signaling systems include three components: the receptor itself, a GTP-binding protein (G protein), and an intracellular target usually a protein.

The cell membrane acts as a switchboard. Messages arriving through different receptors can produce a single effect ifthe receptors act on the same type of G protein. On the other hand, signals activating a single receptor can produce more than one effect ifthe receptor acts on different kinds of G proteins, or ifthe G proteins can act on different effectors.

The heterotrimeric G protein is composed of a guanine nucleotide-binding α subunit together with a tight complex of β and γ subunits. In their resting state, the G proteins, which consist of alpha (α), beta (β) and gamma (γ) subunits, are complexed with the nucleotide guanosine diphosphate (GDP) and are in contact with receptors. When a hormone or other first messenger binds to receptor, the receptor changes conformation and this alters its interaction with the G protein. This spurs the α subunit to release GDP, and the more abundant nucleotide guanosine tri-phosphate (GTP), replaces it, activating the G protein. The G protein then dissociates to separate the α subunit from the still complexed beta and gamma subunits. The free Gα and the Gβγ subunits both may be capable of influencing the activity of specific effector molecules (e.g., the enzymes adenylyl cyclase, cyclic GMP phosphodiesterase (PDE), phospholipase C, phospholipase A2, and selected ion channels). The effector (which is often an enzyme) in turn converts an inactive precursor molecule into an active "second messenger," which may diffuse through the cytoplasm, triggering a metabolic cascade.

After a few seconds, G protein signaling is terminated with the hydrolysis of GTP to GDP through the intrinsic GTPase activity of the Gα subunit and the subsequent reassociation of Gα-GDP with Gβγ to form the inactive heterotrimer. This reassociation is driven by the high affinity of GDP-bound Gα for Gβγ. Hundreds, if not thousands, of receptors convey messages through heterotrimeric

G proteins, of which at least 17 distinct forms have been isolated. Although the greatest variability has been seen in the α subunit, several different β and γ structures have been reported. There are, additionally, several different G protein-dependent effectors.

The study of microorganisms indicates that the development of G protein signal transduction pathways arose early in the evolution of eukaryotic cells. G protein regulatory function is intrinsic to the response to mating pheromones in yeast (Whiteway et al. 1989) and the development of the cellular slime mold Dictvostelium discoideum is controlled by G protein-mediated responses to cAMP (Devreotes 1989).

Adenylyl cyclase is among the best studied of the effector molecules which function in mammalian cells in response to activated G proteins. Activation of adenylyl 5 cyclase occurs when signals transduced from specific cellular receptors result in the release of GTP-bound Gas. Gas ("s" denotes stimulatory) was originally identified as a regulator of adenylyl cyclase activity in mutant S49 cells which lacked adenylyl cyclase activity. Gαs-GTP stimulated adenylyl cyclase activity in those eye- cells (Northup et al. (1980) Proc. Natl. Acad. Sci. USA 77, 6516-6520). The production of cAMP can be l o stimulated by pure GTP-γS-bound Gas (GTP-γS is a non-hydroi yzable foιτn of iii nucleotide). Activation of cyclase by GTP-bound Gas is reversed by excess 0$_!% inhibition is assumed to occur as an inactive G protein heterotrimer re-forms.

Molecules which signal through receptors that interact with another class of G proteins, Gαi (including Gail, Gai2, and Gαi3), mediate inhibition of adenylyl cyclase. is Upon agonist binding to Gi-coupled receptors, bom activated Gαi protein and the released Gβγ complex appear to be capable of inhibiting the activity of adenylyl cyclase [Taussig et al. (1993) Science 261, 218-221]. The Gβγ complex may inhibit the enzyme's activity by reforming a heterotrimer with free Gas, thereby sequestering that stimulatory molecule (Gilman (1984) Cell 36, 577-579). In addition, the Gαi subunit o may directly inhibit adenylyl cyclase activity (Taussig et al. ( 1993) Science 261, 218- 221.) A third mechanism for the negative regulation of adenylyl cyclase involves direct inhibition by the Gβγ complex. Purified type 1 adenylyl cyclase has been shown to be directiy inhibited by βγ subunits (Taussig et al. (1993) J. Biol. Chem. 268, 9-12). Cyclic nucleotides play an important role in the regulation of a multitude of 5 cellular activities. The synthesis of adenosine 3', 5'-cyclic phosphate (cyclic adenosine monophosphate or cAMP) is catalyzed by adenylyl cyclase, an enzyme which, in mammalian cells, is an integral membrane protein. Cyclic AMP is a second messenger which acts in response to cellular signals through a specific protein kinase (cAMP- dependent protein kinase or protein kinase A) to phosphorylate target molecules, e.g., o other protein kinases or proteins involved in transport or cellular morphology. Through stimulation of the kinase, intracellular cAMP mediates many of the effects of hoπtωnes in the regulation of cellular metabolism and cell growth. Cyclic AMP is hydrolyzed by several phosphodiesterases (PDE) and can be actively secreted from some cell types, presumably via a specific transporter, or sequestered from the cytoplasm via transporters 5 present in the membranes of intracellular organelles.

In vertebrate cells, adenylyl cyclase is regulated by heterotrimeric G proteins [Gilman (1984) Cell 36, 577-579] while in yeast, RAS proteins regulate adenylyl cyclase [Toda et al. (1985) Cell 40, 27-36; Broek et al. (1985) Cell 41, 763-769]. In turn, the activity of both the heterotrimeric G proteins and RAS proteins are controlled by the forms of guanine nucleotides to which they are bound. While most adenylyl cyclases are found associated with the plasma membrane, certain forms of the enzyme expressed in bacteria are cytosolic, as is a mammalian enzyme found in testis. Peripheral membrane adenylyl cyclases are expressed in E. coli (Aiba et al. 1984) and in S. cerevisiae (Kataoka et al. 1985). The adenylyl cyclase encoded by the ACG gene of Dictyostelium appears to have a single transmembrane domain (Pitt et al. 1992). A second adenylyl cyclase gene from Dictyostelium (ACA) (Pitt et al. 1992), the Drosophila rutabaga gene (Levin et al. 1992), and the six full- length cDNAs encoding mammalian adenylyl cyclases that have been cloned to date code for integral membrane proteins.

Haploid yeast cells are able not only to grow vegetatively, but also to mate to form a diploid cell. The two mating types ("sexes") of haploid cells are designated a and α. The a cells produce the dodecapeptide a-factor, and the α cells, the tridecapeptide α- factor. Because a-factor and α-factor elicit a mating response in the yeast cell of the opposite "sex", they are called "pheromones". These pheromones, as well as other proteins specifically involved in the production or transport of, or response to, pheromones, are considered "pheromone system proteins".

The gene encoding a-factor pheromone, like the α-factor receptor gene, is an a cell-specific gene; a cell-specific genes are only expressed in a cells. The gene encoding α-factor pheromone, like the a-factor receptor gene, is an α cell-specific gene; α cell- specific genes are only expressed in α cells. Other yeast genes belong to a haploid- specific gene set and are expressed in haploid cells (a cells or α cells) but not in diploid (a/α) cells. In addition, there exists a diploid cell-specific gene set, including those genes involved in sporulation.

In eukaryotic cells, RNA polymerase II promoters contain a specific sequence (the TATA box) to which the transcription factor TFIID (TATA binding protein or TBP) binds. An active transcription initiation complex includes TFIID, accessory initiation proteins, and RNA Pol II. As in higher eukaryotic cells, the TATA box is an essential control sequence in yeast promoters. Yeast TATA-box-binding protein (TBP) was identified by its ability to substitute in function for mammalian TFIID [Buratowski et al., Nature 334, 37 (1988); Cavallini et al., Nature 334, 77 (1988)]. With only a few apparent exceptions [transcription of some glycolytic enzyme genes, see Struhl, Mol. Cell. Biol. 6, 3847 (1986) and Ogden et al., Mol. Cell Biol. 6, 4335 (1986)] transcription of yeast genes requires the proximal TATA box element and TFIID binding for initiation of transcription. Also required for efficient transcription are gene-specific activator proteins; the precise mechanism whereby these gene-specific regulatory proteins influence transcription has not been completely elucidated. 5 MCM 1 p (encoded in the MCM 1 gene) is a non-cell-type-specific transcription factor in yeast. MCMlp acts alone or in concert with other regulatory proteins to control expression of a- and α- cell specific genes. Yeast mating type loci encode the regulatory proteins that contribute to the control of cell type-specific expression. These proteins are Matalp (encoded by the MATa gene) and Matαlp and Matα2p (encoded by the MATα

I o locus). MCM 1 p activates transcription of a-specifϊc genes by binding to an upstream activation sequence (UAS) located in the control region of a- specific genes. Matαlp and MCMlp interact to enhance each other's binding to specific UAS binding sites to activate α-cell-specific gene transcription in α-cells. Matα2p associates with MCMlp to repress a-specific gene transcription in α-cells. In diploid (a/α) cells, Matαlp and is Matα2p associate to repress the transcription of haploid-specific genes. The Matα Ip/Matα2p regulatory entity is found only in diploid cells.

Yeast contain two genes encoding the α-factor pheromone, MFα 1 and MFα2. Analysis of yeast bearing mutations in these sequences indicates that MFαl gives rise to the majority of α-factor produced by cells. Expression occurs at a higher level from MF

20 α 1 than from MFo2 (Kurjan, Mol. Cell. Biol. 5, 787 ( 1985).

The MFαl gene of yeast encodes a 165 aa precursor protein containing an 85 aa leader sequence at the N-terminus. The leader includes a 19 aa signal sequence and a 66 aa sequence which contains sites for the addition of three oligosaccharide side chains (Kurjan and Herskowitz, Cell 39, 933 (1982); Singh et al. Nuc. Acids Res. 1 1.4049

25 (1983); Julius et al. Cell 36, 309 ( 1984). Four tandem copies of the 13 aa α-factor are present in the C-terminal portion of the precursor; 6-8 aa spacer peptides precede the α- factor sequences (see Figure 2).

After translocation of the nascent α-factor polypeptide to the ER, the signal sequence is cleaved from the precursor protein to yield pro-α-factor (Waters et al. J_* o Biol. Chem. 263, 6209 ( 1988). The core N-linked carbohydrate is added to three sites in the N-terminus of pro-α-factor (Emter et al. Biochem. Biophys. Res. Commun. 116, 822 (1983); Julius et al. Cell 36, 309 (1984); Julius et al. Cell 37, 1075 (1984). Additional glycosylation occurs in the Golgi prior to cleavage of pro-α-factor by the KEX2 endopeptidase. This enzyme cleaves within each of the spacer repeats leaving a Lys-Arg 5 sequence attached to the C-terminus of α-factor peptide (Julius et al. Cell 37, 1075 (1984). The Lys-Arg sequence is removed by the action of the KEX-1 protease (Dmochowska et al. Cell 50, 573 (1987). The additional spacer residues present at the N-terminus of α-factor peptide are removed by the dipeptidyl aminopeptidase encoded by STE13 (Julius et al. Cell 32, 839 (1983). Four α-factor peptides are released from each precursor protein via the proteolytic processing outlined above and the mature α- factor is secreted from the cell.

Precursors of the 12 aa mature a-factor peptide are encoded in the MFal and MFa2 genes and are 36 aa and 38 aa residues, respectively (for schematic of MFal gene see Figure 5). The precursors contain one copy of a-factor and the products of the two genes differ in sequence at one amino acid. The two forms of a-factor are produced in equal amounts by a cells (Manney et al. in Sexual interactions in eukaryotic microbes, p21, Academic Press, New York (1981).

Processing of a-factor entails a process that differs in every detail from that of α- factor. The processing of a-factor begins in the cytosol and involves the farnesylation of the C-terminal cysteine residue near the carboxyl terminus (-CVIA) by a farnesyl transferase (Schafer et al. Science 245, 379 ( 1989); Schafer et al. Science 249, 1133 (1990). The α and β subunits of the farnesyl transferase are encoded by the RAM2 and RAM1 genes, respectively (He et al. Proc. Natl. Acad. Sci. 88, 11373 (1991). Subsequent to farnesylation is the proteolytic removal of the three amino acids that are C-terminal to the modified cysteine by a membrane-bound endoprotease. Next, the carboxy-terminal famesylated cysteine residue is modified further: the carboxyl group is methylated by the product of the STE14 gene. STE14p is a membrane-bound S- farnesyl-cysteine carboxyl methyl transferase (Hrycyna et al. EMBO. J. 10, 1699 (1991). The mechanisms of the N-terminal processing of a-factor have not been elucidated. After processing of the precursors is complete, mature a-factor is transported to the extracellular space by the product of the STE6 gene (Kuchler et. al. EMBO J. 8, 3973 (1989), an ATP-binding cassette (ABC) transporter.

In normal S. cerevisiae (budding yeast) a cells, the α- factor binds the G protein- coupled membrane receptor STE2. The G protein dissociates into the Gα and Gβγ subunits, and the Gβγ binds an unidentified effector, which in turn activates a number of genes. STE20, a kinase, activates STE5, a protein of unknown function. STE5 activates STE11 kinase, which stimulates STE7 kinase, which induces the KSS1 and/or FUS3 kinases. These switch on expression of the transcription factor STE 12. STE12 stimulates expression of a wide variety of genes involved in mating, including FUS1 (cell fusion), FAR1 (cell-cycle arrest), STE2 (the receptor), MFA1 (the pheromone), SST2 (recovery), KAR3 (nuclear fusion) and STE6 (pheromone secretion). Other genes activated by the pathway are CHS1, AGαl, and KAR3. The multiply tandem sequence TGAAACA has been recognized as a "pheromone response element" found in the 5'- flanking regions of many of the genes of this pathway.

One of the responses to mating pheromone is the transient arrest of the yeast cell in the GI phase of the cell cycle. This requires that all three GI cyclins (CLN1, CLN2, CLN3) be inactivated. It is believed that FUS3 inactivates CLN3, and FAR1 inbiWts CLN2. (The product responsible for inactivating CLN1 is unknown).

The growth arrest is terminated by a number of different mechanisms. First, the α-factor receptor is internalized following binding of the pheromone, resulting in a transient decrease in the number of pheromone binding sites. Second, the C-terminal tail of the receptor is phosphorylated consequent to ligand binding, resulting in uncoupling of the receptor from the transducing G proteins. Third, pheromone-induced increases in expression of GPAlp (the Gα-subunit of the heterotrimeric G protein) increase the level of the α subunit relative to the Gβ and Gγ subunits, resulting in reduction in tiie level of free Gβγ and consequent inactivation of the pheromone response pathway. Additional mechanisms include induction of the expression of SST2 and BAR1 and phosphorylation of the α subunit (perhaps by SVG1).

Signaling is inhibited by expression of a number of genes, including CDC36, CDC39, CDC72, CDC 73, and SRM1. Inactivation of these genes leads to activation of the signaling pathway. A similar pheromone signaling pathway may be discerned in α cells, but tie nomenclature is different in some cases (e.g., STE3 instead of STE2).

Other yeast also have G protein-mediated mating factor response pathways. For example, in the fission yeast S. pombe, the M factor binds the MAP3 receptor, or the P- factor the MAM2 receptor. The dissociation of the G protein activates a kinase cascade (BYR2, BYR1, SPK1), which in turn stimulates a transcrip-tion factor (STE11).

However, in S. pombe, the Gα subunit transmits the signal, and there are of course other differences in detail.

A wide variety of foreign proteins have been produced in S. cerevisiae, that remain in the yeast cytoplasm or are directed through the yeast secretory pathway (Kingsman et al. ΗBTECH 5, 53 ( 1987). These proteins include, as examples, insulin- like growth factor receptor (Steube et al. Eur. J. Biochem. 198, 651 (1991), influenza virus hemagglutinin (Jabbar et al. Proc. Natl. Acad. Sci. 82, 2019 (1985), rat liver cytochrome P-450 (Oeda et al. DNA 4, 203 (1985) and functional mammalian antibodies (Wood et al. Nature 314, 446 (1985). Use of the yeast secretory pathway is preferred since it increases the likelihood of achieving faithful folding, glycosylation and stability of the foreign protein. Thus, expression of heterologous proteins in yeast often involves fusion of the signal sequences encoded in the genes of yeast secretory proteins (e.g., α- factor pheromone or the SUC2 [invertase] gene) to the coding region of foreign protein genes.

A number of yeast expression vectors have been designed to permit the 5 constitutive or regulated expression of foreign proteins. Constitutive promoters are derived from highly expressed genes such as those encoding glycolytic enzymes like phosphoglycerate kinase (PGK1) or alcohol dehydrogenase I (ADH1) and regulatable promoters have been derived from a number of genes including the galactokinase (GAL1) gene. In addition, supersecreting yeast mutants can be derived; these strains o secrete mammalian proteins more efficiently and are used as "production" strains to generate large quantities of biologically active mammalian proteins in yeast (Moir and Davidow, Meth. in Enzymol. 194, 491 (1991).

Heterologous G protein-coupled receptors have been functionally expressed in S. cerevisiae. Marsh and Hershkowitz, (Cold Spring Harbor Symp., Quant. Biol., 53: 557- 5 65 (1988)) replaced the S. cerevisiae STE2 with its homologue from S. Kluyven. More dramatically, a mammalian beta-adrenergic receptor and Gα subunit have been expressed in yeast and found to control the yeast mating signal pathway. King, et al., Science, 250: 121-123 (1990).

Duke University, WO92/05244 (April 2, 1992) describes a transformed yeast cell o which is incapable of producing a yeast G protein α subunit, but which has been engineered to produce both a mammalian G protein α subunit and a mammalian receptor which is "coupled to" (i.e., interacts with) the aforementioned mammalian G protein α subunit. Specifically, Duke reports expression of the human beta-2 adrenergic receptor (hβAR), a seven transmembrane receptor (STR), in yeast, under control of the GAL1 5 promoter, with the hβAR gene modified by replacing the first 63 base pairs of coding sequence with 11 base pairs of noncoding and 42 base pairs of coding sequence from the STE2 gene. (STE2 encodes the yeast α-factor receptor). Duke co-expressed a rat G protein α subunit in the same cells, yeast strain 8C, which lack the cognate yeast protein. Duke found that the modified hβAR was functionally integrated into the membrane, as o shown by studies of the ability of isolated membranes to interact properly with various known agonists and antagonists of hβAR. The ligand binding affinity for yeast- expressed hβAR was said to be nearly identical to that observed for naturally produced h βAR. Ligand binding resulted in G protein-mediated signal transduction. Duke did not co-express a mammalian adenylyl cyclase in these cells. ₅ African trypanosomes are protozoan parasites which are able to evade host immune defenses by altering their surface glycoproteins. The variable antigenicity is accomplished by sequential expression of genes encoding coat proteins. The variable surface glycoprotein genes (VSG) are transposed from silent regions to active, telαmere- linked expression sites. Additional open reading frames (ORFs) termed the Expression Site Associated Genes (ES AGs), are found at these expression sites. ESAG4, cloned from Tiypanosoma brucei, contains a sequence which is homologous to S. cerevisiae adenylyl cyclase [Pays et al. (1989) Cell 57, 835-845]. In addition, an ESAG from Trypanosoma equiperdum (eESAG4c), which is homologous to the ESAG4 of T. brucei, has been shown to encode an adenylyl cyclase which will complement an adenylyl cyclase (cyr-1) deletion mutant of S. cerevisiae [Ross et al. (1991) EMBO J. 10,2047- 2053].

The eESAG4c ORF contains sequence with homology to both S. cerevisiae and S. pombe adenylyl cyclases [Kataoka et al. (1985) Cell, 43, 493-505; Yamawaki- Kataoka et al. (1989) PNAS 86, 5693-5697; Young et al. (1989) PNAS 86, 7989-7993]. The region that is conserved between trypanosomes and yeast is within the yeast adenylyl cyclase catalytic domain and exhibits a sequence identity on the order of 50%. The eESAG4c sequence is approximately 40% identical to that of bovine brain adenylyl cyclase type 1 (Krupinski et al. (1990) Science 244, 1558-1562). The protein predicted by the eESAG4c sequence bears an N-terminal sequence that encodes a putative transmembrane domain flanking the sequence that is homologous to the adenylyl cyclase catalytic domain.

Also identified within the ESAG of Trypanosoma equiperdum are sequences which bear homology to a "leucine-rich repeat" gene family (Takahashi et al. (1985) PNAS 82, 1906-1910; Lopez et al. (1988) PNAS 85, 2135-2139; Hashimoto etal. (1988) Cell 52, 269-279). Proteins encoded by members of this family are involved in diverse functions, however, the repeat sequences are believed to be involved in membrane association and in protein-protein interactions. In S. cerevisiae the repeat domain of adenylyl cyclase is required for regulation of the enzyme by RAS proteins and for the association of the enzyme with the plasma membrane [Colicelli et aL (1990) Mol. Cell. Biol. 10, 2539-2543; Field et al. (1990) Science 247, 464-467; Mitts etal, (1990) Mol. Cell. Biol. 10, 3873-3883.] Also within the ESAG are sequences witii limited homology to nucleotide binding domains [Florent et al. (1991) Mol. Cell. Biol. 11, 2180-2188] that have been hypothesized to have a regulatory function in trypanosomes analogous to that of Ras in yeast. Neither the leucine-rich repeat region nor the nucleotide binding domain were included in the sequences that complemented the yeast cyr deletion mutants [Ross et al. (1991) EMBO J. 10, 2047-2053]. Ross et al. (1991) speculated that the lack of these potential regulatory sequences would account for the much greater adenylyl cyclase activity exhibited by cyr-1 deletion mutants expressing eESAGc than was seen in yeast expressing the endogenous CYR gene from plasmids.

In addition to ESAG4, cloned from the VSG region of Trypanosoma brucei, at least three other genes cloned from T. brucei, GRESAG 4.1 , 4.2 and 4.3, bear sequence homology to eukaryotic adenylate and guanylate cyclases (Alexandre et al. (1990) Mol. Biochem. Parasitol. 43, 279-288). ["GRESAG" indicates Genes Related to Expression Site Associated Genes.] It has been demonstrated that both ESAG 4 and GRESAG 4.1 can complement a S. cerevisiae adenylyl cyclase deletion mutant, cyrl . The trypanosome cyclases associate with the yeast membrane fraction, differ in their response to Ca2+, and do not appear to be properly regulated in yeast [Paindavoine et al. (1992) Mol. Cell. Biol. 12, 1218-1225].

Thus, the heterogenous adenylyl cyclases that have been shown to exhibit activity, although unregulated, in yeast are derived from trypanosome species. The trypanosome cyclase genes lie in regions near sequences encoding leucine-rich motifs with homology to a regulatory domain of yeast adenylyl cyclase. This suggests that proteins which derive from the two different trypanosome sequences may interact to form a regulatory complex. This could be analogous to the situation in Saccharomyces cerevisiae where activity of adenylyl cyclase is controlled through the interaction of the enzyme with regulatory RAS proteins. The homologies of sequence and regulation between the yeast and trypanosome enzymes appear to have favored the complementation of yeast deleted for adenylyl cyclase with sequences encoding the trypanosome enzyme.

Attempts to express mammalian adenylyl cyclases in yeast

Previous attempts by other laboratories to express mammalian adenylyl cyclase in yeast were unsuccessful. Ronald Taussig, working in the laboratory of Alfred Gilman at the University of Texas Southwestern Medical Center, attempted to express mammalian type 1 adenylyl cyclase in Saccharomyces cerevisiae (personal communication). The protocol used by Taussig involved rescue of eye (cyclase minus) cells by transformation with mammalian type 1 adenylyl cyclase; the metric of cyclase activity was growth of the test cells on forskolin-containing medium. Forskolin is known to bind directly to and to stimulate adenylyl cyclase types 1-6 in mammalian cells. Taussig was unable to detect enzyme activity in eye cells transformed with the mammalian enzyme, i.e., he was unable to detect growth of transformed cells on forskolin-containing medium. The mammalian type 2 cyclase has been functionally expressed, by means not publicly disclosed, in the primitive eukaryote Dictyostelium discoideum [personal communication from P. Devreotes cited in Iyengar (1993)]. The structure of one of the two adenylyl cyclase genes that have been isolated from Dictyostelium, ACA, iβ 5 predicted to be structurally analogous to the mammalian cyclases in that it is also an integral membrane protein [Pitt et al. 1992]. In addition, Dictyostelium can express eight Gα subunits, each bearing approximately 45% sequence homology to mammalian Gα proteins [Hadwiger et al. 1991 ; Wu and Devreotes 1991 ]. The lack of success in Gilman's laboratory at expressing a functional mammalian type 1 adenylyl cyciase in

I o yeast, and the successful expression of the mammalian enzyme in Dictyostelium, indicate that differences in the transduction of signal to this enzyme exist between yeast and the higher eukaryotes. Furthermore, those differences must be taken into consideration in any attempt to recapitulate a signal transduction pathway witii mammalian adenylyl cyclase in yeast. is Peptide libraries are systems which simultaneously display, in a form which permits interaction with a target, a highly diverse and numerous collection of peptides. These peptides may be presented in solution (Houghten 1991), or on beads (Lam 1991), chips (Fodor 1991), bacteria (Ladner USP 5,223,409), spores (Ladner USP 409), plasmids (Cull 1992) or on phage (Scott, Devlin, Cwirla, Felici, Ladner *409). Many of o these systems are limited in terms of the maximum length of the peptide or the composition of the peptide (e.g., Cys excluded). Steric factors, such as the proximity of a support, may interfere with binding. Usually, the screening is for binding in vitro to an artificially presented target, not for activation or inhibition of a cellular signal transduction pathway in a living cell. While a cell surface receptor may be used as a 5 target, the screening will not reveal whether the binding of the peptide caused an allosteric change in the conformation of the receptor.

Ladner, USP 5,096,815 describes a method of identifying novel proteins &t polypeptides with a desired DNA binding activity. Semi-random ("variegated") DNA encoding a large number of different potential binding proteins is introduced, in o expressible form, into suitable host cells. The target DNA sequence is incorporated into a genetically engineered operon such that the binding of the protein or polypeptide will prevent expression of a gene product that is deleterious to the cell under selective conditions. Cells which survive the selective conditions are thus cells which express a protein which binds the target DNA. While it is taught that yeast cells may be used for s testing, bacterial cells are preferred. The interactions between the protein and tiie target DNA occur only in the cell, not in the periplasm, and the target is a nucleic acid, not a protein.

Substitution of random peptide sequences for functional domains in cellular proteins permits some determination of the specific sequence requirements for the accomplishment of function. Though the details of the recognition phenomena which operate in the localization of proteins within cells remain largely unknown, the constraints on sequence variation of mitochondrial targeting sequences and protein secretion signal sequences have been elucidated using random peptides (Lemire et al., J. Biol. Chem. 264, 20206 (1989) and Kaiser et al. Science 235, 312 (1987), respectively). Yeast have been engineered to express foreign polypeptide variants to be tested as potential antagonists of mammalian receptors. Libraries encoding mutant glucagon molecules were generated through random misincorporation of nucleotides during synthesis of oligonucleotides containing the coding sequence of mammalian glucagon. These libraries were expressed in yeast and culture broths from transformed cells were used in testing for antagonist activity on glucagon receptors present in rat hepatocyte membranes (Smith et al. 1993).

Summary Of The Invention

The present invention relates to the functional expression of a heterologous adenylyl cyclase in a cell, preferably a mammalian adenylyl cyclase, and to the use of the engineered cells in identifying not only potential inhibitors or activators of the heterologous adenylyl cyclase, but also of other proteins which are naturally or artificially "coupled" to the heterologous adenylyl cyclase in the engineered cell. The term "coupled" here means that inhibition or inactivation of the coupled protein results in inhibition or activation (not necessarily respectively) of the adenylyl cyclase. Functional expression of human adenylyl cyclases is especially desirable. In certain embodiments of the invention, the engineered cells are mammalian cells.

In preferred embodiments, the engineered cells are yeast cells. The adenylyl cyclase of Saccharomyces cerevisiae is a peripheral membrane protein with a structure that is substantially different from that of the cloned mammalian adenylyl cyclases; all six mammalian adenylyl cyclase cDNAs cloned to date encode integral membrane proteins with intricate transmembrane structures forming a significant part of their overall structure. Furthermore, although yeast contain heterotrimeric G proteins, these proteins do not appear to be involved in the regulation of S. cerevisiae adenylyl cyclase, rather, the yeast enzyme is regulated by other members of the family of GTP-binding proteins, Rasl and Ras2. The yeast adenylyl cyclase thus bears scant resemblance to the mammalian enzyme in structure and in regulation and it could not have been assumed that the mammalian enzyme could be made to function in Saccharomyces. Preferably, the yeast cell is a diploid strain, or another strain which does not express yeast Got, Gβ or Gγ.

In one aspect, the present invention provides for a host cell, preferably a yeast or mammalian cell, which expresses a mutant form of a heterologous adenylyl cyclase that is constitutively activated. In one embodiment a host cell further expresses a peptide library. In preferred embodiment, a peptide library of the present invention is expressed intracellularly. In yet another embodiment an engineered cell of the present invention further comprises an intragenic mutation that restores regulation of the mutant form of the heterologous adenylyl cyclase.

In another aspect the present invention provides for a yeast cell expressing a mutant form of a heterologous adenylyl cyclase. In another embodiment the cells further express a peptide library. In a preferred embodiment, the peptide library is expressed intracellularly. In yet another aspect the present invention provides for a yeast cell expressing a heterologous adenylyl cyclase and an intracellularly expressed peptide library.

In still another aspect the present invention provides for a mammalian cell expressing a mutant heterologous adenylyl cyclase and a peptide library. In a preferred embodiment, the peptide library is expressed intracellularly. In particularly preferred embodiments a heterologous adenylyl cyclase of the present invention is of mammalian origin. In particularly preferred embodiments a mammalian adenylyl cyclase is of human origin. ^{:; ■}

In certain embodiments a heterologous adenylyl cyclase of the present invention comprises a mutation in the C 1 a domain of the enzyme to cause constitutive activation of an adenylyl cyclase activity. In other embodiments a heterologous adenylyl cyclase of the invention comprises a mutation in the C2a domain of the enzyme to cause constitutive activation of an adenylyl cyclase activity. In preferred embodiments the heterologous adenylyl cyclase has a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyr265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968 Asp, and Lys998 Asn.

In certain embodiments the endogenous adenylyl cyclase of the host cell is inactivated. For example, in certain embodiments the activity of the endogenous adenylyl cyclase is temperature sensitive. For instance, in certain embodiments the host cell is a yeast cell, which has the mutant allele cdc35- 1. In certain preferred embodiments the heterologous adenylyl cyclase is selected from the group consisting of a type I, a type II, a type III, a type IV, a type V, a type VI, W

- 14 -

a type Vll, a type VIII, a type IX and a type X adenylyl cyclase.

In still other embodiments a host cell of the present invention can be engineered to further expresses one or more heterologous or chimeric G protein subunits. In a preferred embodiment one or more of the heterologous G protein subuints expressed by the host cell is of mammalian origin. In another embodiment the host cell is engineered to express one or more of the chimeric G protein subuints comprises a sequence of mammalian origin.

In certain embodiments of the invention the host cell expresses a reporter gene construct comprising a cyclic AMP responsive promoter operably linked to a reporter gene encoding a selectable or screenable gene product.

In certain preferred embodiments of the invention the peptide library expressed by the host cell is a random or semi-random peptide library. In other embodiments, the peptide library comprises a library of peptides derived from a G protein subunit. In still other embodiments, the library of peptides is derived from Gas. In certain embodiments of the invention, the peptide library is expressed intracellularly. In preferred embodiments the expression of the library is directed by a thioredoxin A expression vector.

In particularly preferred embodiments of the invention the host cell is a yeast cell which is selected from the group consisting of: Kluyveromyces lactis, Saccharomyces kluyveri, Schizosaccharomyces pombe, Ustilago maydis, and Saccharomyces cerevisiae. The cells of the present invention are ideal tools for drug screening assays. In certain embodiments the subject assay provides for contacting the cells of the present invention with test agents and determining the ability of the test agent to alter the activity of adenylyl cyclase compared to its activity in the absence of the test agent. In other embodiments the cyclase may be activated by contacting the reagent cell with an activator of the enzyme, or by stimulation of a receptor, such as a G protein coupled receptor. Accordingly, in one aspect of the invention the subject engineered cells may be used in methods to detect agonists or antagonists of the heterologous adenylyl cyclase. If desired, another exogenous protein may be coupled to the heterologous adenylyl cyclase, so that agonists or antagonists of the other exogenous protein may be detected by virtue of their effect on adenylyl cyclase activity. In a preferred embodiment, the other exogenous protein is a surrogate for a pheromone system protein, as hereafter defined. In an especially preferred embodiment, the other exogenous protein is a mammalian G protein-coupled receptor, which is a surrogate for the yeast pheromone receptor. The ability of particular test agents, be they produced by the reagent cell, produced by a cocultured cell, or exogenously added to the culture, to modulate adenylyl cyclase activity of the enzyme can be measured directly. Alternatively, the use of a reporter gene, coupled to and dependent on the heterologous adenylyl cyclase can provide a convenient readout. In any event, a statistically significant change in the detection signal can be used to facilitate identification of those s test agents which are apparent effectors of the target adenylyl cyclase. By this method, test compounds which are potential therapeutic agents can be identified. For instance, if the test compound does not appear to induce the activity of the enzyme, the assay may be repeated and modified by the introduction of a step in which the recombinant cell is first contacted with a known activator of the enzyme, or engineered to express » — io constitutively activated adenylyl cyclase. By such an embodiment, the test compound can be assayed for its ability to antagonize the adenylyl cyclase activity. Alternatively, the assay can score for compounds which potentiate the induction response generated by the treatment of the cell with a known activator of adenylyl cyclase.

In another embodiment the subject assays provide for the identification of agents

I 5 which indirectly modulate the activity of the target adenylyl cyclase, e.g., by altering signal transduction by a G protein coupled receptor, or otherwise by acting upon the adenylyl cyclase. In certain embodiments the subject assay provides for the use of cells which express, in addition to a heterologous adenylyl cyclase, a heterologous G protein- coupled receptor, and, if necessary, a heterologous G protein or G protein subunit. o Alternatively, an endogenous G protein coupled receptor , such as the yeast STE2 ttr

STE3 could be used.

In another aspect, the invention provides for screening assays which can b$ used to identify a modulator of adenylyl cyclase activity by contacting a cell of the present invention with a test compound and comparing the level of adenylyl cyclase activity in 5 the cell in the presence of the test compound with the level of activity in the absence of the test compound or in a cell lacking the heterologous adenylyl cyclase. Where a change in the level of adenylyl cyclase activity in the presence of the test compound the test compound can be identified as a modulator of adenylyl cyclase activity. -_t

In another aspect of the invention a method of identifying a modulator of o adenylyl cyclase activity is provided in which a cell of the present invention which expresses a peptide library is cultured and the level of adenylyl cyclase activity in the presence of the expressed peptide library is compared with the level of activity in the absence of the expressed peptide library or in the absence of the heterologous adenylyl cyclase activity. Peptides are identified which modulate the activity of adenylyl cyclase 5 by measuring a change in the level of adenylyl cyclase activity in the presence of tiie expressed peptide. In certain embodiments the methods of the present invention are used to identify inhibitors of adenylyl cyclase activity. In other embodiments the methods of the present invention are used to identify activators of adenylyl cyclase activity, ln still another aspect, the present invention provides for a method of identifying an adenylyl cyclase stimulatory domain of a G protein subunit by culturing a cell of the present invention which has been engineered to express a peptide library derived from a G protein subunit. The level of the adenylyl cyclase activity in the presence of the expressed peptide library is compared with the level of activity in the absence of the expressed peptide library, and a change in the level of adenylyl cyclase activity in the presence of the expressed peptide library indicates that the expressed peptide is a cyclase stimulatory domain of a G protein subunit. In a preferred embodiment such a peptide library is derived from the G protein subunit is Gas and the heterologous cyclase is a type IV cyclase.

In still another aspect, the present invention provides for a method of identifying an intragenic mutation that restores regulation of a constitutively active mutant form of an adenylyl cyclase by culturing a yeast cell which expresses a constitutively active mutant form of a heterologous adenylyl cyclase at 23°C and at 30°C. The level of adenylyl cyclase activity is measured at 23°C and at 30°C and an intragenic mutation is identified that restores regulation of the constitutively active mutant form of the heterologous adenylyl cyclase by measuring a difference between the level of adenylyl cyclase activity at 23°C and at 30°C.

In still another aspect, the present invention provides for preparing a pharmaceutical preparation of one or more compounds identified in the subject methods as being a modulator of adenylyl cyclase activity. In yet another aspect the present invention provides for an expression vector encoding a constitutively active mutant form of a mammalian adenylyl cyclase. In a preferred embodiment, the expression vector encodes a mutant form of the mammalian adenylyl cyclase having a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyr265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968 Asp, and Lys998Asn.

In yet another embodiment the invention provides for a nucleic acid encoding a mutant mammalian adenylyl cyclase comprising a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyr265His, Glu313Gly,Val38811e, Gly968Ser, Arg 268 Lys, Gly968Asp, and Lys998Asn. In certain embodiments of the invention the cells can coexpress a heterologous or chimeric G protein subunit. In other embodiments, the heterologous adenylyl cyclase is a mutant form, e.g., a constitutively active mutant.

The practice of the present invention will employ, unless otherwise indicated, conventional techniques of cell biology, cell culture, molecular biology, transgenic biology, microbiology, recombinant DNA, and immunology, which are within the skill of the art. Such techniques are explained fully in the literature. See, for example* Molecular Cloning A Laboratory Manual, 2nd Ed., ed. by Sambrook, Fritsch and Maniatis (Cold Spring Harbor Laboratory Press: 1989); DNA Cloning, Volumes} and II (D. N. Glover ed., 1985); Oligonucleotide Synthesis (M. J. Gait ed., 1984); Mullis et al. U.S. Patent No: 4,683,195; Nucleic Acid Hybridization (B. D. Hames & S. J. Higgins eds. 1984); Transcription And Translation (B. D. Hames & S. J. Higgins eds. 1984); Culture Of Animal Cells (R. I. Freshney, Alan R. Liss, Inc., 1987); Immobilized Cells And Enzymes (IRL Press, 1986); B. Perbal, A Practical Guide To Molecular Cloning (1984); the treatise, Methods In Enzymology (Academic Press, Inc., N.Y.); Gem Transfer Vectors For Mammalian Cells (J. H. Miller and M. P. Calos eds., 1987, Cold Spring Harbor Laboratory); Methods ln Enzymology, Vols. 154 and 155 (Wu et al. eds.), Immunochemical Methods In Cell And Molecular Biology (Mayer and Walker, eds., Academic Press, London, 1987); Handbook Of Experimental Immunology, Volumes I- IV (D. M. Weir and C. C. Blackwell, eds., 1986); Manipulating the Mouse .Embryo, (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y., 1986). The appended claims are to be treated as a non-limiting recitation of preferred embodiments.

Brief Description Of The Drawings

Figure 1. Structural model of mammalian adenylyl cyclase. This figure reproduces the model of mammalian adenylyl cyclase described in Tang et al. (1992) C.S.H. Symposia on Quantitative Biology 57, 135-144. Ml and M2 denote two domains, each containing six putative membrane-spanning sequences. NI andN2 together comprise the short amino terminal tail which is believed to reside intracellularly. N2 sequences are those proximal to the first transmembrane sequence. C 1 a and C 1 b foπn a large cytoplasmic loop that joins the two transmembrane domains. The C2a and C2b sequences form a second large cytoplasmic loop. The C-terminal sequence denoted C2b is present only in type 1 and type 3 adenylyl cyclases.

Figure 2. Outline of successive stages in the development of yeast autocrine systems.

An outline of the normal synthesis and release of mating pheromones is diagrammed in the upper left. Two genes, MFαl and MFα2, encode precursor proteins (MFαlp and MFα2p) containing four and two repeats, respectively, of the tridecapeptide representing mature α-factor. These precursors are processed proteolytically in a series of enzymatic reactions that begin with cleavage of the signal sequence in the endoplasmic reticulum and involve both glycosylation of the leader peptide and cleavage by the proteases KEX2p, STE13p, and KEX1P. The result is the secretion of mature α-factor which, upon binding to STE2p normally expressed on the surface of a cells, elicits a number of changes in the a cells, including growth arrest. The a cells, in turn, express two genes, MFal and MFa2, which encode precursors (MFalp and MFa2p) for a-factor. These precursors undergo farnesylation by RAMl and RAM2, proteolytic trimming of the C- teπninal three amino acids (by a protein tentatively identified as RAM3p), carboxymethylation of the newly exposed C-terminal cysteine by STE14p, and endoproteolytic removal of the N-terminal leader sequence by an activity provisionally identified as STE19p. Upon export of the mature a-factor from the cell via STE6p, it binds to STE3p expressed on the surface of α cells and stops their growth.

Stage 1 (Figure 2a) involves the development of yeast strains in which SST2, FAR1, and HIS3 are inactivated and a suitable reporter construct like fusl ::HIS3 is integrated into the genomes of both α and a cells, α cells are further altered by replacement of the normally expressed STE3p with STE2p, while a cells are further modified by replacement of the normally expressed STE2p with STE3p. The resulting strains should show growth on histidine-deficient media in the absence of exogenous pheromone.

Stage 2 (Figure 2b) involves, first, inactivation of MFαl and MFα2 in cells and inactivation of MFal and MFa2 in a cells developed in Stage 1. These modifications will result in strains which are auxotrophic for histidine. Next, the appropriate expression plasmid will be introduced: the expression plasmid pADC-MF containing an oligonucleotide encoding α-factor should confer upon α cells the ability to grow on histidine-deficient media; the expression plasmid pADC-MFa containing an oligonucleotide encoding a-factor should enable a cells to grow on histidine-deficient media.

Stage 3 (Figure 2b) uses the cells developed in Stage 2 for the insertion of expression plasmids. However, instead of using plasmids which contain oligonucleotides that encode genuine pheromone, the yeast will be transformed with expression plasmids that contain random or semi-random oligonucleotides. Transformants which can grow on histidine-deficient media will be expanded and their plasmids isolated for sequencing of the inserted oligonucleotide.

Detailed Description Of The Invention

5

Definitions

For the purpose of the present invention, an "heterologous" protein is one which sufficiently differs in amino acid sequence from the proteins naturally produced by (he host cell in question so that its closest cognate is a protein produced by a cell other than o the host cell. For example, ifthe host cell is a yeast cell, the cell producing this cognate protein may be a microbial cell (other than a yeast cell), a plant cell, or an animal cell. If an animal cell, it may be of invertebrate (e.g., insect or nematode) or of vertebrate (e.g., avian, piscine or mammalian, especially human) origin. Likewise ifthe host cell is a cell from one mammalian species, e.g., a mouse, a heterologous protein would be a protein s from another species, e.g., a human. A protein is considered to be of, e.g., human origin, regardless of whether it is encoded by the chromosome of a normal human, or by the genome of a virus which infects and replicates in human cells.

A yeast protein which is involved in the post-translational modification, transport, recognition or signal transduction of a yeast pheromone will be referred to as a o "pheromone system protein" (PSP), and cognate non-yeast proteins which are capable of substituting for a PSP sufficiently, to be able, at least under some circumstances, to carry out that role of the yeast protein in the engineered yeast cell as PSP surrogates.

An "activator or agonist" of an adenylyl cyclase is a substance which, in a suitable host cell, causes the adenylyl cyclase to become more active, and thereby 5 elevates the cAMP signal of said cell to a detectable degree. The mode of action of the activator may be direct, e.g., through binding the cyclase, or indirect, e.g., through binding another molecule which otherwise interacts with the cyclase.

Conversely, an "inhibitor or antagonist" of an adenylyl cyclase is a substance which, in a suitable host cell, causes the cyclase to become less active, and thereby o reduces the cAMP signal to a detectable degree. The reduction may be complete or partial, and due to a direct or an indirect effect.

An "activator" of a pheromone system protein surrogate is a substance which, in a suitable host cell, causes the pheromone system protein surrogate to become more active, and thereby elevates the signal transduced by the native or modified pheromone signal pathway of said cell to a detectable degree. The surrogate may be initially nonfunctional, but rendered functional as a result of the action of the activator, or it may be functional, and the effect of the activator is to heighten the activity of the surrogate. The mode of action of the activator may be direct, e.g., through binding the surrogate, or indirect, e.g., through binding another molecule which otherwise interacts with the surrogate. When the PSP surrogate is a substitute for a pheromone receptor, and the 5 activator takes the place of the pheromone, it is customary to refer to the activator as an agonist of the receptor.

Conversely, an "inhibitor" of a pheromone system protein surrogate is a substance which, in a suitable host cell, causes the PSP surrogate to become less active, and thereby reduces the transduced signal to a detectable degree. The reduction may be o complete or partial. When the PSP surrogate is a substitute for a pheromone receptor, and the inhibitor competes with the pheromone for binding to the receptor, it is customary to refer to the inhibitor as an "antagonist".

The term "modulator" includes both "activators" and "inhibitors".

A "mammalian adenylyl cyclase" is a protein which is either identical to an 5 adenylyl cyclase occurring naturally in a mammal, or is a mutant which is substantially homologous with such a mammalian adenylyl cyclase and more similar in sequence to it than to the host adenylyl cyclase. Related terms, such as "primate adenylyl cyclase", or "human adenylyl cyclase", are analogously defined. A heterologous adenylyl cyclase is "functionally homologous" to a host protein if, either alone, or in concert with other o exogenous proteins, or after being modified by a drug, it is able to provide an adenylyl cyclase activity within the engineered host cell. It is not necessary that it be as efficient as the host protein, however, it is desirable that it have at least 10% of the activity of the cognate host protein.

A surrogate PSP protein is "functionally homologous" to a yeast protein if, either 5 alone or after being modified by a drug, it is able to perform the function of the yeast

PSP, or an analogous function, within the engineered yeast cell. It is not necessary that it be as efficient as the yeast protein, however, it is desirable that it have at least 10% of at least one of the pheromone system-related activities of the yeast protein. Nor is it necessary that it have the same spectrum of action as the yeast protein, e.g., if it is a 0 receptor, it may respond to entirely different ligands than does the endogenous receptor, or to some common ligands and some new ones. The receptors of Table 2 are considered functionally homologous with the yeast pheromone receptors, even though they do not respond to yeast pheromones, and may not couple to the unmodified endogenous G proteins, although they are G protein-coupled receptors. This is s considered an "analogous function" .

The PSP surrogate may be a protein which must be modified in some way by a drug to be functional. For example, the drug could cause an allosteric change in the PSP surrogate's conformation, or it could cleave off a portion of the surrogate, the balance of the protein then being a functional molecule.

The PSP surrogate may also be one which is functional only if other modifications are made in the yeast cell, e.g., expression of a chimeric G α subunit to interact with an exogenous G protein-coupled receptor.

As used herein an "intragenic mutation" is a second mutation in a gene that reverts a phenotype caused by a first mutation. For example, in the case of a first mutation that removes regulation of adenylyl cyclase activity a second mutation, an intragenic mutation can restore regulation of adenylyl cyclase activity. For instance, if a mutation in a gene causes a loss of function, and a second mutation restores function, then the second mutation can be referred to as an intragenic suppresser. Intragenic mutations can convey important information about both the structure and mechanism of a protein. A mutation that causes the protein to become activated may do so because it binds a substrate more tightly, and by extension, an intragenic mutation may function by causing the protein to bind the substrate less tightly, often for a distinct reason. Thus, an analysis of both mutations will contribute to a description of how the protein binds its substrate.

As used herein, the term "vector" refers to a nucleic acid molecule capable of transporting another nucleic acid to which it has been linked. One type of vector is a

"plasmid", which refers to a circular double stranded DNA loop into which additional DNA segments may be ligated. Another type of vector is a viral vector, wherein additional DNA segments may be ligated into the viral genome. Certain vectors are capable of autonomous replication in a host cell into which they are introduced (e.g., bacterial vectors having a bacterial origin of replication and episomal mammalian vectors). Other vectors (e.g., non-episomal mammalian vectors) are integrated into the genome of a host cell upon introduction into the host cell, and thereby are replicated along with the host genome. Moreover, certain vectors are capable of directing the expression of genes to which they are operatively linked. Such vectors are referred to herein as "recombinant expression vectors" or simply "expression vectors". In general, expression vectors of utility in recombinant DNA techniques are often in the form of plasmids. In the present specification, "plasmid" and "vector" may be used interchangeably as the plasmid is the most commonly used form of vector. However, the invention is intended to include such other forms of expression vectors, such as viral vectors (e.g. , replication defective retroviruses, adenoviruses and adeno-associated viruses), which serve equivalent functions. As used herein, the term "host cell" is intended to refer to a cell into which a nucleic acid of the invention, such as a recombinant expression vector of the invention, has been introduced. The terms "host cell" and "recombinant host cell" are used interchangeably herein. It should be understood that such terms refer not only to the 5 particular subject cell but to the progeny or potential progeny of such a cell. Because certain modifications may occur in succeeding generations due to either mutation or environmental influences, such progeny may not, in fact, be identical to the parent cell, but are still included within the scope of the term as used herein.

The term "substantially homologous", when used in connection with amino acid o sequences, refers to sequences which are substantially identical to or similar in sequence, giving rise to a homology in conformation and thus to similar biological activity. The term is not intended to imply a common evolution of the sequences.

Typically, "substantially homologous" amino acid sequences are at least 50%, more preferably at least 80%, identical in sequence, at least over any regions known to s be involved in the desired activity, when sequences are aligned by art-accepted methods. For example, a sequence alignment tool that can be used is part of the sequence analysis software package, Version 7.3, from GCG (Genetics Computer Group, 575 Science Drive, Madison, Wisconsin 53711). BESTFIT uses the homology algorithm of Smith and Waterman [Advances in Applied Mathematics 2, 482-489 (1981) ] to identify the o best alignment of two sequences. Parameters can initially be set for the following:

Match=1.0 Gap weight=1.0 Mismatch=-0.9 Length weight=0.0 selecting neither the "LOWroad" nor "HIGHroad" options.

Most preferably, no more than five residues, other than at the termini, are different Preferably, the divergence in sequence, at least in the aforementioned regions, 5 is in the form of "conservative modifications" .

"Conservative modifications" are defined as

(a) conservative substitutions of amino acids as hereafter defined; and

(b) single or multiple insertions or deletions of amino acids at the termini, at interdomain boundaries, in loops or in other segments of relatively high mobility. o Preferably, except at the termini, no more than about five amino acids are inserted or deleted at a particular locus, and the modifications are outside regions known to contain binding sites important to activity.

Conservative substitutions are herein defined as exchanges within one of the following five groups: 5 I. Small aliphatic, nonpolar or slightly polar residues:

Ala, Ser, Thr (Pro, Gly) II. Polar, negatively charged residues, and their amides

Asp, Asn, Glu, Gin ' ι ^■

III. Polar, positively charged residues:

His, Arg, Lys ;" IV. Large, aliphatic, nonpolar residues:

Met, Leu, lie, Val (Cys) V. Large, aromatic residues:

Phe, Tyr, Trp Residues Pro, Gly and Cys are parenthesized because they have special conformational roles. Cys participates in formation of disulfide bonds. Gly imparts flexibility to the chain. Pro imparts rigidity to the chain and disrupts α helices. These residues may be essential in certain regions of the polypeptide, but substitutable elsewhere.

"Semi-conservative substitutions" are defined herein as being substitutions within supergroup I/II/III or within supergroup I V/V, but not within a single one of groups I-V. If a substitution is not conservative, it preferably is semi-conservative.

Two regulatory DNA sequences (e.g., promoters) are "substantially homologous" if they have substantially the same regulatory effect as a result of a substantial identity in nucleotide sequence. Typically, "substantially homologous" sequences are at least 50%, more preferably at least 80%, identical, at least in regions known to be involved in the desired regulation. Most preferably, no more than five bases are different.

"Inactivation" means that production of a functional gene product is prevented or inhibited. Inactivation may be achieved by deletion of the gene, mutation of the promoter so that expression does not occur, mutation of the coding sequence so that tiie gene product is inactive, or failure to provide factors necessary for the biological activity of the gene product Inactivation may be partial or total. "Functional expression" refers to expression of a gene under conditions such that its gene product is not only expressed but is also biologically active within the expressing cell.

The term "autocrine cell", as used herein, refers to a cell which produces a substance which can stimulate a signal transduction pathway of that cell. Wild-type α and a cells are not autocrine with respect to the pheromone pathway. However a yeast cell which produces both α-factor and α-factor receptor, or both a-factor and a-factor receptor, in functional form, is so autocrine. By extension, yeast cells which produce a peptide which is being screened for the ability to activate a G protein-coupled receptor, a surrogate for the yeast pheromone receptor, are called "autocrine cells", though it might be more precise to call them "putative autocrine cells". Of course, in a library of such cells, in which a multitude of different peptides are produced, it is likely that one or more of the cells will be "autocrine" in the stricter sense of the term.

Design of Chimeric Proteins If a mammalian or other exogenous protein, unmodified, cannot be functionally expressed, to the extent desired, in a host cell, e.g., a yeast cell, one may engineer the host cell to express a protein which is a chimera of the exogenous protein of interest and a cognate yeast protein. The term "chimera" implies that one portion of the sequence is more homologous to the host, e.g., yeast protein than to the heterologous, e.g., mammalian protein, and another portion is the reverse. Possible combinations include mammalian/yeast, yeast/mammalian, mammalian/yeast/mammalian, and yeast/mammalian/yeast.

Functional chimeras may be identified by a systematic synthesize-and-test strategy. It is not necessary that all theoretically conceivable chimeras be evaluated directly.

One strategy is described schematically below. The aligned protein sequences can be aligned into two or more testable units. These units may be equal or unequal in length. Preferably, the units correspond to functional domains or are demarcated so as to correspond to special features of the sequence, e.g., regions of unusually high divergence or similarity, conserved or unconserved regions in the relevant protein family or the presence of a sequence motif, or an area of unusual hydrophilicity or hydrophobicity. For example, "Y" can represent a unit of the yeast protein, and "M" a corresponding unit of a mammalian protein. If there are five units (the choice of five instead of two, three, four, six, ten, etc. is arbitrary), any or all of the following chimeras, which will help to rapidly localize the critical regions:

(a) progressive C-terminal substitution of mammalian sequence for yeast sequence, e.g.,

Y Y Y Y Y

Y Y Y Y M

Y Y Y M M

Y Y M M M

Y M M M M MMMMM

(b) progressive N-terminal substitution of mammalian sequence for yeast sequence

YYYYY

MYYYY

MMYYY

MMMYY

MMMMY

MMMMM

(c) dual terminal substitutions, e.g., MMMMM

YMMMY

YYMYY

YYYYY

and

YYYYY

MYYYM

MMYMM

MMMMM, and

(d) single replacement "scans," such as

MY YYY

YMYYY

YYMYY

YYYMY

YYYYM and

YMMMM

MYMMM

MMYMM

MMMYM

MMMMY

Based on the data these tests provide, it may appear that, e.g., the key difference between the yeast and mammalian sequences vis-a-vis, say, display on the yeast membrane, is in the fifth unit. One can then subdivide that unit into subunits and test further, e.g.

MM MM (mm)

MM MM (my)

MMMM (ym) M M M M (yy) where the parenthesis refer to two subunits into which the fifth unit was subdivided.

Design of Functional Mutants GeneraUy A protein is more likely to tolerate a mutation which

(a) is a substitution rather than an insertion or deletion;

(b) an insertion or deletion at the terminus, than internally;

(c) affects a surface residue rather than an interior residue; (d) affects a part of the molecule distal to the binding site;

(e) is a substitution of one amino acid for another of similar and/or hydrophobicity; and

(f) is at a site which is subject to substantial variation among a family of homologous proteins to which the protein of interest belongs. These considerations can be used to design functional mutants.

Surface vs. Interior Residues

Charged residues almost always lie on the surface of the protein. For uncharged residues, there is less certainty, but in general, hydrophilic residues are partitioned to the surface and hydrophobic residues to the interior. Of course, for a membrane protein, the membrane-spanning segments are likely to be rich in hydrophobic residues.

Surface residues may be identified experimentally by various labeling techniques, or by 3-D structure mapping techniques like X-ray diffraction and NMR. A 3-D model of a homologous protein can be helpful.

Binding Site Residues

Residues forming the binding site may be identified by (1 ) comparing the iflects of labeling the surface residues before and after complexing the protein to its target, (2) labeling the binding site directly with affinity ligands, (3) fragmenting the protein and testing the fragments for binding activity, and (4) systematic mutagenesis (e.g., alanine- scanning mutagenesis) to determine which mutants destroy binding. Ifthe binding site of a homologous protein is known, the binding site may be postulated by analogy.

Protein libraries may be constructed and screened that a large family (e.g., 108) of related mutants may be evaluated simultaneously.

Mammalian Adenylyl Cyclases The signals transduced through the heterotrimeric G proteins in mammalian cells influence intracellular events through the action of molecules termed effectors. Among the best characterized of these effector molecules is the hormone-responsive enzyme, adenylyl cyclase. As of early 1994, six full-length and two partial cDNA clones of adenylyl cyclase have been obtained from a variety of mammalian tissues. Sequence analysis of the encoded proteins resulted in the identification of distinct enzyme types, and functional characteristics have determined their grouping into five distinct families (Iyengar (1993) FASEB J. 7, 768-775). The first family is comprised of type 1 adenylyl cyclase; this enzyme is stimulated by hormone receptors through Gas, by forskolin and by Ca2+/calmodulin. A cDNA encoding type 1 was isolated from a bovine brain library [Krupinski et al. (1989) Science 244, 1558-1564]. The activity of the type 1 enzyme is inhibited by the expression of Gβγ subunits. Each of the other cloned mammalian cyclases which have been expressed in cells are also stimulated by Gas and by forskolin. However, it has been shown that responses to Gβγ and to Ca2+/calmodulin vary among those other enzymes.

A second family is comprised of type 2 and type 4 adenylyl cyclases; these enzymes are stimulated by Gβγ, but that stimulation depends on the presence of activated Gas. Enzymes belonging to this second family have been cloned from rat brain [Feinstein et al. (1991) PNAS 88, 10173-77] and testes [Gao and Gilman (1991) PNAS, 88 10178-10182]. This family of adenylyl cyclases is insensitive to

Ca2+/calmodulin. Although related by sequence homology, type 2 and type 4 adenylyl cyclase differ in both distribution and regulation. Type 2 message is found only in brain and lung tissue, while type 4 is more widely expressed, having been detected in brain, kidney, liver, heart, lung and testis. The type 2 enzyme and all other mammalian adenylyl cyclases cloned to date, except type 4, contain potential sites for phosphorylation by protein kinase A. Furthermore, the type 2 enzyme is known to be substantially stimulated by activated protein kinase C, while type 4 adenylyl cyclase is unaffected by that kinase.

Type 3 adenylyl cyclase, cloned from rat olfactory tissue [Bakalyar and Reed (1990) Science 250, 1403-1406] is abundantly expressed in olfactory neuroepithelia, is sensitive to stimulation by Ca2+/calmodulin, but is not affected directly by the presence of Gβγ. The type 3 enzyme may be central to olfactory signal transduction.

A fourth family of adenylyl cyclases has been cloned from a variety of sources including canine heart [Ishikawa et al. (1992) J. Biol. Chem. 267, 13553-13557], rat liver and kidney [Premont et al. (1992) PNAS 89, 9808-9813], mouse lymphoma cells [Premont et al. (1992) Endocrinology 131, 2774-2783] and from a mouse/hamster W

- 29 -

hybrid cell line NCB-20 [Yoshimura and Cooper (1992) PNAS 89, 6716-6720J. These enzymes are termed types 5 and 6 on the basis of sequence, are unaffected by Gβγ in the presence or absence of activated Gas, and are inhibited by low concentrations of Ca2+. Multiple messages for types 5 and 6 have been observed, suggesting that alternatively spliced forms occur. The type 6 enzyme has been shown to exist in long and short forms which differ in the presence or absence of a 14 amino acid stretch at the N-terminus (Iyengar (1993) FASEB J. 7,768-775). Types 5 and 6 show significant overall sequence similarity, including >50% homology in the putative transmembrane regions where the greatest degree of sequence divergence exists among adenylyl cyclase subtypes (Katsushika et al. (1992) Proc. Natl. Acad. Sci. USA 89, 8774-8778). Like type 5, type 6 expression is highest in heart and brain; unlike type 5, type 6 mRNA is also detected in a variety of other tissues.

Type 7 adenylyl cyclase, cloned as a partial cDNA from the S49 mouse lymphoma cell line [Krupinski et al. (1992) J. Biol. Chem. 267, 24858-25862], appears to be related to the type 2 enzyme. A second partial clone, termed type 8 adenylyl cyclase, was obtained from a human brain library, and encodes a protein that is distinct from previously characterized enzymes [Parma et al. (1991) Biochem. Biophys. Res. Comm. 179, 455-462] and may be brain-specific [Krupinski et al. (1992) J. Biol. Chem. 267, 24858-25862]. The majority of mammalian adenylyl cyclases are integral membrane proteins.

Sequence analysis of the cloned enzymes predicts the presence of twelve membrane- spanning regions present in two domains which each contain six hydrophobic sequences. The short N- and the lengthy C-terminal sequences are predicted to reside intracellularly. Overall sequence similarity among the various adenylyl cyclase types is approximately 50%, with greater homologies between some isolates resulting in their categorization in subfamilies. There exist two large cytoplasmic domains, a 350 amino acid loop between the first and second set of six transmembrane sequences (Cl) and an extensive C- terminal tail (250-300 amino acids) following the second set of six transmembrane sequences (C2). Portions of these cytoplasmic domains bear sequence similarity to the catalytic domains that have been identified in cloned guanylyl cyclases [Chinkers and Garbers (1991) Ann. Rev. Biochem. 60, 553-575], are to some degree homologous with one another (60-80% homology between the Cl and C2 domains), and are highly conserved in the mammalian adenylyl cyclases that have been sequenced to date (50- 92%) [Iyengar (1993) FASEB J. 7,768-775; Koesling et al. (1991) FASEB J. 5,2785- 2791; Tang and Gilman (1992) Cell 70, 869-872]. In contrast, the transmembrane regions of the various adenylyl cyclases lack significant sequence homology to one another.

Aside from the sequence homology to the catalytic domains of the guanylyl cyclases, analysis of the Drosophila homolog of the bovine type I adenylyl cyclase yields additional evidence that the C-terminal tail of the mammalian enzyme contributes to catalysis. A point mutation (Gly to Arg) in the C-terminal cytoplasmic domain of the Drosophila enzyme results in the loss of adenylyl cyclase activity. This mutation alters a glycine residue that is absolutely conserved in all mammalian adenylyl cyclase isolates (Levin et al. (1992) Cell 68, 479-489). Point mutations in the conserved domains of Cl and C2 result in decreased catalytic activity of the mammalian enzyme (Tang et al. 1992 Cold Spring Harbor Symp. Quant. Biol. 57, 135-144). In addition, truncated forms of adenylyl cyclase that lack either the central cytoplasmic loop or the C-terminal tail sequence are devoid of enzyme activity. Thus, expression of either half of the molecule yields inactive enzyme while co-expression of both halves partially restores catalytic activity (Tang et al. (1991 ) J. Biol. Chem. 266, 8595-8603). Since co- expression of the two halves of the molecule does not restore activity in full, it has been hypothesized that the transmembrane domains of the enzyme direct optimal interaction between the two large cytoplasmic regions of the protein and, thus, the stable formation of an active site (Iyengar (1993) FASEB J. 7,768-775). Soluble forms of guanylyl cyclase are known to function as heterodimers with each subunit contributing sequence that bears homology to the highly conserved Cl and C2 regions of adenylyl cyclase (Nakane et al. 1990 J. Biol. Chem. 265, 479-489).

Aside from a proposed role of active site stabilization, the contribution of the transmembrane domains to adenylyl cyclase function remains enigmatic. The hydrophobic domains appear to be the site for forskolin action. Forskolin is a lipid- soluble diterpene that specifically binds to and activates adenylyl cyclase in mammalian cells in the absence of G protein-coupled receptor agonists. Forskolin has no effect on the testis-specific cyclase of mammals or on bacterial cyclases; these proteins are cytosolic. However, the ACA cyclase of Dictyostelium, an integral membrane protein, is insensitive to forskolin whereas mammalian type 2 cyclase, when it is expressed in Dictyostelium, is sensitive to stimulation by the diterpene (personal communication from P. Devreotes cited in Iyengar (1993) FASEB J. 7,768-775).

Preferably, a host cell is engineered to express a type 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 adenylyl cyclase. Cyclases of type 2 are especially preferred. Within the aforementioned types, the following cyclases are of particular interest:

Adenylyl Cyclase Type | Characteristics

As many as 10 distinct isoforms of adenylyl cyclase from several species have been cloned. Types I through X have been sequenced and reported in the literature. . Those isoforms whose sequence has been published are: bovine type I (Krupinski, J., et al, Science 244:1558-1564, 1989); human type II (Stengel, D., et al, Hum Genet 90: 126- 130, 1992); rat type II (Feinstein, PG et al, Proc Natl Acad Sci USA 88:10173-10177, 1991); rat type III (Bakalyar, HA, and Reed, RR, Science 250:1403-1406, 1990); rat type IV (Gao, B. and Gilman, A.G., Proc Natl Acad Sci USA 88:10178-10182, 1991); canine type V (Ishikawa, Y. et al, J Biol Chem 267: 13553-13557, 1993); rabbit type V (Wallach, J. et al, FEBS Lett 338:257-263, 1994); canine type VI (Katsushika, S. et al, Proc Natl Acad Sci USA 89:8774-8778, 1992); murine type VI (Yoshimura, M„ and Cooper, D.M., Proc Natl Acad Sci USA 89:6716-6720, 1992); rat type VI (Premont, R.T. et al, Proc Natl Acad Sci USA 89:9808-9813, 1992); murine type VII (Watson* P.A. et al, J Biol Chem 269:28893-28898, 1994); human type VIII (Defer, N., et al

FEBS Lett, 351:109-113, 1994); rat type VIII (Cali, J.J. et al, J Biol Chem 269:12190-5, 1994), type IX (Genbank accession number U30602), and type X (Genbank accession number XZ50190). O 97/40170 P T/US97/0 711

- 32 -

Detailed information is available regarding the regulation of several of the different isoforms. While the isoforms published to date differ in their control by Gαi subunits, Gβγ subunits, calcium, and protein kinase C, all are activated by Gas (Taussig et al, J Biol Chem 269:6093-6100, 1994; Chen, J., and Iyengar, R., J Biol Chem 268:12253-12256, 1993; Cooper et al, Nature 374:421-424. 1995). The present invention can be used to provide these adenylyl cyclases as functional and regulatable activities in host cells.

The present invention is not limited to expression of the presently known adenylyl cyclases, or presently known types of such cyclases. The cyclase is preferably a primate, especially a human cyclase, but may also be a cyclase associated with mammals of the orders Rodenta (mice, rats, rabbits, etc.), Arteriodactyla (goats, pigs, sheep, cows, etc.) or Carnivora (cats, dogs, etc.), or other mammalian orders.

The adenylyl cyclase of the present invention need not be a naturally occurring protein, rather, it may be a mutant, provided that its sequence is more similar to that of a naturally occurring cyclase which is heterologously expressed than to that of the naturally occurring host adenylyl cyclase, for example, in the yeast as encoded by CYR1. Preferably, the mutant is also substantially homologous to a naturally occurring exogenous adenylyl cyclase, or a mutant known to be functional. For instance, as described in greater detail in the appended examples, the reagent cell of the subject assay can be engineered with a mutated adenylyl cyclase which is constitutively active in the absence of factors that otherwise stimulate wild type adenylyl cyclases. For example, mutations may be introduced in the Cla domain or in the C2a domain to produce a constitutively active enzyme. In preferred embodiments such mutations might include Tyr265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968Asp, and

Lys998 Asn in adenylyl cyclase type IV (ACIV), or corresponding mutations in other adenylyl cyclases. The mutation Gly968Ser was originally identified in the priority document of this application as Gly960Ser, based on an error in the wild-type sequence that appeared in Genbank. The S. cerevisiae gene that encodes adenylyl cyclase, C YR1 , was cloned by

Kataoka et al. ((1985) Cell, 43, 493-505). CYR1 encodes a protein consisting of 2026 amino acids; four domains of the protein have been identified and include the N-terminal and C-terminal domains as well as a central, repetitive amphipathic sequence and a catalytic domain. The central repetitive sequence bears homology to a 23 amino acid leucine-rich motif that is found repeated in a family of proteins identified in yeast, mammals and Drosophila melanogaster (Field et al. (1990) Science 247, 464-467). The leucine rich regions and the carboxyl terminus of the enzyme are required for its , interaction with RAS proteins. The catalytic domain of yeast adenylyl cyclase has been localized to the C-terminal 417 amino acids [Kataoka et al. (1985) Cell, 43, 493-505]. Yeast adenylyl cyclase appears to be a peripheral membrane protein; it is found in the insoluble cell fraction after non-detergent extraction of cells. Hydropathic analysis of the sequence does not reveal a hydrophobic, membrane-spanning domain and the coding sequence lacks a signal sequence normally found in secreted or integral membrane proteins (Liao and Thorner (1980) PNAS, 77, 1898-1902; Kataoka et al. (1985) Cell, 43, 493-505; Perlman and Halvorson (1983) J. Mol. Biol. 167, 391-409). However, it is postulated that the repetitive domain which contains amphipathic sequence would permit embedding of the protein into membranes. In support of this hypothesis, truncated yeast adenylyl cyclase molecules that contain the central amphipathic sequence localize to membrane fractions when expressed in E. coli [Kataoka et al. (1985) Cell 43, 493-505]. Two ras genes in S. cerevisiae were originally identified by virtue of sequence homology with probes derived from mammalian ras genes (Broach and Deschennes (1990) Adv. Cancer Res. 54, 79-139). The Ras 1 and Ras 2 proteins of S. cerevisiae are required for vegetative growth of haploid cells through RAS-dependent activation of adenylyl cyclase and the synthesis of cAMP [Toda et al. (1985) Cell 40, 27-36; Brock et al. (1985) Cell 41, 763-769]. It is not known if RAS has any additional essential functions in yeast however, overexpression of adenylyl cyclase in yeast can suppress the lethality that results from the loss of RAS function in ras 1 and ras 2 disruption mutants. Therefore, any additional essential functions of RAS are compensated for by increased production of cAMP (Kataoka et al. (1985) Cell, 43, 493-505).

Mutant proteins which are "substantially homologous" to a naturally occurring adenylyl cyclase may also be of value. In rat type II adenylyl cyclase, possible "neutral" mutations include substitutions in the non-conserved second membrane-spanning sequence in the first transmembrane domain (Leu78Ile;Ile79Leu; Ile93Leu and Leu94Ile). Substitutions that may be made in residues of the non-conserved fifth membrane-spanning sequence in the first transmembrane domain (Uel62Leu; Leul63Ile) could also yield a mutant protein with wild-type activity. It is probable that other conservative amino acid substitutions not specifically cited here may be made in the adenylyl cyclase sequence without any diminishment of wild type protein activity.

Use of the Adenylyl Cyclase in Screening In its natural state, adenylyl cyclase is inactive. However, it can be activated by other molecules, in particular, the free Gα subunit or the Gβγ complex. In addition, the present invention provides for mutated forms of adenylyl cyclase which are constitutively activated or inactivated.

In one embodiment, an engineered cell is used to screen for drugs which, like Gα or in some cases Gβγ, can directly activate the adenylyl cyclase, or increase the activity of a partially activated adenylyl cyclase.

In another embodiment, the engineered cell is used to screen for drugs which inhibit adenylyl cyclase activity. In this situation, the adenylyl cyclase must first be activated. This can be done by engineering the cell to overexpress Gα or Gβγ, or by expressing a mutated adenylyl cyclase. For example, a mutant form of the enzyme which is active in the absence of factors that stimulate wild type adenylyl cyclases. For example, mutations may be introduced in the Cla domain or in the C2a domain. In preferred embodiments such mutations might include Tyr265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968Asp, or Lys998Asn.

Alternatively, the cell may be engineered to co-express both a G protein and a G protein coupled receptor, or any other protein which affects the activity of the adenylyl cyclase, e.g. calmodulin, PKA or PKC or any as yet unknown or uncharacterized proteins which directiy bind/interact with the adenylyl cyclase to affect its function, and the receptor stimulated either by externally added ligand or by a co-expressed ligand. In this case, the receptor could be the yeast pheromone receptor and the ligand the yeast α or a factor. Or the receptor could be a foreign receptor, and the ligand one appropriate to that receptor. In either case, the ligand is a known activator used merely to stimulate activation of the adenylyl cyclase, and the drugs are screened for inhibition of this adenylyl cyclase.

In yet another embodiment, the engineered cell is used to screen for drugs which inhibit or activate adenylyl cyclase indirectly, e.g., by their action upon a G protein- coupled receptor. The receptor activates the G protein subunits act on the adenylyl cyclase. In this case, a compatible G protein-coupled receptor and a compatible G protein would be provided with the heterologous adenylyl cyclase in the same host cell. The subject assays may be used to screen compounds which are added to cultures or, alternatively, the subject cells may be engineered to express a peptide library and the assay used to screen for expressed peptides in the library which have a desired activity, e.g., activating or inhibiting the heterologous adenylyl cyclase. In preferred embodiments, the peptide library is expressed intracellularly. In other embodiments, a peptide library may be secreted. In another embodiment, the engineered cell is used to screen for drugs which inhibit or activate surrogates of pheromone system proteins. In those embodiments in which it is desired that the adenylyl cyclase be activated by Gα or Gβγ, it is necessary that the engineered cell express a form of Gα or Gβγ that can carry out this function. Ifthe yeast Gα or Gβγ will not activate the mammalian adenylyl cyclase, a mammalian or chimeric Gα or Gβγ will be expressed for this 5 purpose. G proteins are discussed in the next section. :

G Proteins

G Protein Alpha Subunits

I o Stimulatory (Gas) Subunits. Through reconstitution analysis of the eye- mutant of S49 murine lymphoma cells, the Gas protein was identified (Ross and Gilman (1977) J. Biol. Chem. 252, 6966-6969) as a stimulatory guanine nucleotide-binding protein that coupled hormone receptors to adenylyl cyclase. Mammalian Gas cDNA clones have been obtained from human brain (Bray et al. 1986, 1987), human liver (Mattera et al. 1986), is bovine brain (Harris et al. 1985), bovine adrenal gland (Robishaw et al. 1986), bovine cerebral cortex (Nukada et al. 1986), hamster lung fibroblasts (Mercken et al. 1990), rat glioma cells (Itoh et al. 1986, 1988), rat olfactory neuroepithelium (Jones and Reed 1987), mouse macrophages (Sullivan et al. 1986), and mouse lymphoma cells (Sullivan et al. 1987; Rail and Harris 1987). Bray et al. (1986) isolated four different Gαs cDNAs

20 from human brain (Gαsl -4); these forms appear to arise from a single Gas gene by alternate splicing. The Gas gene contains 13 exons (Kozasa et al. 1988) which are all present in the long form of Gas. A short form of the molecule lacks the 15 amino acids encoded by exon 3. In addition, two alternate mRNAs arise that differ in the presence or absence of a serine codon at the start of exon 4 when different splice sites are used at the 5 5 ' end of that exon.

Inhibitory (Gai) Subunits. As was the case for receptor-mediated stimulation of adenylyl cyclase, GTP was also found to be required for receptor-dependent inhibition of that enzyme. This pointed to a role for a G protein, distinct from Gas in function, in this o inhibition. The identification of this protein resulted from studies on the mechanism of action of the B. pertussis toxin. This toxin was found to (1) abolish the hormonal inhibition of adenylyl cyclase and (2) to ADP-ribosylate a 41-kd membrane protein. Purification of this toxin substrate permitted its identification as a guanine nucleotide- binding protein related to the mammalian G proteins Gas and transducin. The protein 5 was denoted Gαi (i= inhibitory for adenylyl cyclase).

Three single copy genes encode G protein subunits of the Gαi type and the predicted proteins (Gαi-1 , Gαi-2 and Gαi-3) share 85% sequence identity. In coupling to adenylyl cyclase to signal inhibition of this enzyme, the Gαi proteins function in concert with Gas to control cellular cAMP levels.

Gαi-1 cDNA clones obtained to date are human (Bray et al. 1987), bovine (Nukada et al. 1986) and rat (Jones and Reed 1987). Human (Itoh et al. 1988; Weinstein et al. 1988; Beals et al. 1987; Michel et al. 1986; Didsbury et al. 1987), rat (Jones and Reed 1987; Itoh et al. 1986), mouse (Sullivan et al. 1986) and bovine (Yatomi et al. 1992) Gαi-2 cDNA clones have been isolated. Gαi-3 clones include those from human (Itoh et al 1988; Beals et al. 1088;Suki et al. 1987; Kim et al. 1988) and rat (Itoh et al. 1988; Jones and Reed 1987).

The Gαi clones preferred in this invention are human clones of the subtypes Gαi- 2 and Gαi-3; these subtypes are found to be expressed in inflammatory cells. These clones will be expressed in yeast and will be used as targets for the identification of compounds capable of preventing their inhibition of adenylyl cyclase activity. Inhibitors of Gαi function would be of great utility in the treatment of inflammatory diseases: a large number of cell surface receptors expressed in neutrophils and macrophages mediate signals through Gαi.

It is likely that a subset of the possible amino acid substitutions that could be made in human Gαi could yield fully functional, albeit mutant, protein. It is possible that the following mutations would not alter the wild type activity of the protein: Ala59Asp, Glu64Asp, AsploOGlu, Alal63Ser, Val332Ile. It is probable that other amino acid substitions not specifically cited here could be made without any diminishment of wild type Gαi activity.

Use of Structural Models to Design Chimeric or Other Mutant G Proteins

Models of Gα protein structure may be used to predict amino acid modifications which would not be harmful to activity. Analysis of Gα cDNAs and comparison to conserved sequences present in members of the GTPase superfamily has permitted the identification of five conserved stretches, G1-G5, located throughout a "composite" Gα molecule [Conklin and Bourne (1993); Bourne et al. (1991). In addition, the location of putative α-helices, β strands, loop domains and insertions have been deduced by a comparison of Gα sequences with the known secondary structure of p21 ras. Thus α- helices 1-5, β strands 1-6, loops 1-10 and inserts 1-4 have been assigned position in the primary Gα sequence based on comparisons with Ras proteins. Biochemical and genetic studies as well as sequence analysis have led to the delineation of a conceptual model of the Gα protein (Conklin and Bourne 1993). This conceptual model hypothesizes that W

- 37 -

while the guanine nucleotide binding pocket of Gα is oriented toward the cytoplasm, residues that interact with receptors, effectors and with the Gβγ complex face the plasma membrane. The model also asserts the following:

(1) The N terminus of Gα is a major site for interaction with the Gβγ complex. (2) The α2 helix and insert 1 regions also contribute to the interaction of Gα with Qβγ. (3) At least three regions are hypothesized to interface with receptor: the amino and carboxyl termini and the conserved G5 sequence. In the conceptual model the termini rest on the portion of Gα which faces the plasma membrane while the G5 sequence sits at the "top" of the molecule. „:=,, (4) The sequences purported to be involved in the interaction of Gα with effector molecules are envisioned to reside on the plasma membrane-proximal aspect of Got These sequences include the distal half of the α2 helix, the insert 2-loop 7 sequence and the insert 4-loop 9 sequence.

The orientation of the molecule in this conceptual model is supported in part by experimental evidence that assigns specific amino acids to the GTP binding site based on mutations which have been shown to constitutively activate Gα by inhibiting the GTPase activity of the protein. The mutations in question are homologs of GTPase- inhibiting mutations of p21 ras.

Monoclonal antibodies generated against N-terminal sequence cause the dissociation of the Gαtl heterotrimer; in addition, N-terminally myristilated peptide inhibits the binding of Gαtl to Gβγ in competitive fashion. Chemical cross-linking experiments indicate the close proximity of the α2 helix and Gβγ and a specific Gas mutation (G226A) exhibits two deficiencies: the α2 helical region does not undergo GTP-induced conformational change and GTP does not trigger the dissociation of Gβγ from Gas. The sequence denoted as the α2 helix (analogous to the α2 helix of p21 ras) is more highly conserved than any other sequence in Gα; this strict conservation further supports an involvement of the helix in interaction with Gβγ in that the formation of the heterotrimer underlies signaling in all G protein pathways described to date.

Additional data has contributed to the development of the conceptual model of G α. The amino and carboxyl termini of Gα appear to be in close proximity based on cross-linking studies done using mastoparan and based on the specificity of monoclonal antibodies directed against Gαtl . Experimental evidence also suggests the proximity of the C terminus and the region that is analogous to the α2 helix of p2 Iras. Finally, insert 1 , a large sequence located within loop 2, appears to have GAP function and folds as a domain distinct from the GTPase domain [Markby et al. ( 1993)].

Experimental evidence indicates that three regions of Gα (the N and C termini O 97/40170 PC17US97/06711

- 38 -

and the conserved G5 region) contact the receptor. In addition, Conklin et al. (1993) have obtained data which suggests that amino acid residues at the extreme C-terminus of Gα contribute to the specificity of receptor-G protein interactions. Thus chimeras constructed to replace 4-9 residues at the extreme C-terminus of Gαq with amino acids derived from the same region of Gαi resulted in a Gα protein that can transduce signal from D2 dopamine and Al adenosine receptors to phospholipase C, a Gq-specific effector. These receptors normally couple to Gαi.

A glycine residue at the -3 position relative to the C-terminus is central to the formation of a β-turn in this region of the Gα molecule; the β-turn appears to be the structural signal that specifies interaction between receptors and α subunits of the Gαi, Gαo, Gαt family [Dratz et al. (1993)]. It has been hypothesized that the interaction between receptor and the C-terminus of Gα results in the conformational change that leads to the open conformation of the latter molecule, i.e., the configuration in which nucleotide exchange can occur. Mutagenesis of Gas implicated three regions of the molecule (a portion of the α2 helix, i2-L7 and i4-L9) in the activation of adenylyl cyclase. A second series of experiments utilized peptides derived from Gαtl to deduce the region of that molecule that activates phosphodiesterase; peptides derived from i4-<9 mimicked the ability of G αtl-GTP to stimulate cGMP-phosphodiesterase. The regions identified in effector activation reside on the face of the molecule believed to be oriented toward the plasma membrane; in addition, one of the implicated sequences (the α2 helix) is known to undergo conformational change induced by GTP.

Early crystal structure-based models considered the crystal structure of the GTP- binding domain of E. coli EF-Tu (Jurnak 1985; LaCour et al. 1985), as well as the crystal structure of Ha-ras-p21 (Holbrook and Kim 1989). Recently, a crystal structure of transducin-α (Gtα) complexed with GTPγS has been obtained to a resolution of 2.2 angstrom units (Noel et al. 1993). Analysis of this crystal structure, together with the biochemical and genetic data described above, has been used to derive generalized structure/function relationships applicable to all Gα molecules. In the three-dimensional structure, two domains are most apparent in the Gtα-

GTPγS complex, each flanking a guanine nucleotide binding cleft. These are (1) a highly conserved GTPase domain and (2) a highly helical domain that is unique to heterotrimeric G proteins. The GTPase domain is structurally similar to the GTPase domains of p21 Ras and EF-Tu and consists of five α-helices surrounding a six-stranded β-sheet. The other domain is highly helical, unique to heterotrimeric G proteins, and connected to the GTPase domain by two linker sequences. The helical and GTPase domains appear to enclose the GTPγS molecule and an associated Mg2+ion. This arrangement suggests that a conformational alteration is required of the Gα molecule in order for nucleotide exchange to occur; it is likely that conformational changes in the linker sequences initiate the movement of the helical domain and the opening of the s molecule.

The crystal structure permits delineation of the residues of Gα which interact with the triphosphate portion of the GTP molecule, the essential Mg2ion, and the nucleoside. In Gtα, the residues that contact the nucleoside and the phosphates form part of the helical domain and linker 2. These regions are implicated in receptor- 0 regulated nucleotide exchange. Noel et al (1993) cite extensive interactions between Gα residues and guanosine; a subset of these interactions are unique to G proteins while others are conserved among members of the GTPase superfamily. The linkages between the nucleotide binding sites and the surface of Gα that purportedly interacts with receptor are also described. The authors assert that "a mechanistically important feature s of this system is the elegant manner in which interactions with one portion of the nucleotide support contacts with another. It is likely that these tightly coupled interactions potentiate a highly cooperative receptor-mediated disassembly of the elements that so strongly secure GDP and GTP in the nucleotide-binding cleft" [Noel et al (1993)]. 0 Experimental data exists which implicates specific α-helices and β sheets (α2/β

4, α3/β5, α4/β6) in effector binding and activation. These regions were found to form a series of surface loops in the three-dimensional model derived from analysis of the Gtα- GTPγS crystals. Work done with Gtα suggests interaction of these loops with phosphodiesterase or with the inhibitory γ-subunits of that enzyme. Studies 5 accomplished using Gsα/Giα chimeras suggest that these surface loops play a role in the regulation of adenylyl cyclase. In addition, the crystal structure of Gtα-GTPγS indicates how GTP may effect conformational change in these effector-interactive loops. Glycine residues in the α/γ 2 helix interact with the γ-phosphate of GTP and are believed to be the source of the malleability required for the conformational change which occurs upon o hydrolsis of GTP. The GDP/GTP-induced changes in α2 are hypothesized to transmit to the α3 and α4 loops through a connecting series of interhelical contacts , thus linking changes in the interaction of γ-phosphate with α2 to the effector-binding surface loops. The crystal structure draws attention to two residues that appear to play a role in the hydrolysis of GTP. A conserved arginine residue (Arg 174) contacts the γ-phosphate s directly and may facilitate its release upon hydrolysis. Mutation of the cognate arginine in Gsα and Giα severely compromises GTPase activity and results in a constitutively active Gα. The structure also suggests the glutamine at position 203 as the initiator of the hydrolytic attack on the γ-phosphate. Glu203 appears to be appropriately oriented to activate a water molecule well-positioned for nucleophilic attack on the γ-phosphate. This glutamate resides in the α2 helix and is conserved among the family of Gα subunits.

As indicated by the foregoing models of Gα structure, the function of the molecule is dependent on its interaction with receptor, the βγ complex, GTP or GDP, and effector molecules. Mutation of residues that experimental evidence or crystal structure-derived data have indicated as contributing to these numerous interactions could compromise Gα function. The following sequences, residues and domains have been shown to be particularly important to Gα function: N-terminal residues, residues at the extreme C-terminus (particularly the glycine at position -3), the highly conserved α2 helix, cognates of Arg 174 of Gαt, cognates of glutamine 203 of Gαt the GTPase domain, and the α2/β4, α3/β5, and α4/β6 regions. Other sequences would appear to be important, based on their conservation among members of the GTPase family or in that they are unique to heterotrimeric G protein α-subunits. These include the conserved sequences G1-G5 and the regions identified as inserts through comparisons made between Gα and Ras proteins. It must be stated that a definitive citation of functionally important regions cannot be made as this remains an intense area of research. It is expected that future studies will identify the residues in the larger domains cited above that contribute most to function and identify additional specific residues whose characteristics are central to Gα function. Although conservative mutations in important regions of Gα may leave the function of the molecule intact the more radical the change, the higher the likelihood of interference with protein function. The models as outlined above underscore an important feature of the Gα molecule. Conformational change is inherent to the exchange of nucleotide and that exchange is central to Gα function. The conformational change appears to occur as a wave of signal transmission from one domain of the molecule to another. The models thus emphasize that alterations in any one of the functional domains of the protein can affect the final transduction of signal, i.e., the function of the molecule depends on successful cooperativity of several domains.

It must be stated that mutations can be made, however, which can contribute to the utility of the Gα protein in experimental systems. As an example, mutations which compromise the GTPase function specifically, without affecting Gα interaction with effector proteins, results in a constitutively active protein. In a subset of experimental scenarios, a constitutively active Gα is a desirable molecular reagent. It may be possible to make the following conservative amino acid substitutions in the sequence of human Gas without compromising the wild-type activity of the protein: Ilel83Leu, Aspl84Glu, Leul98Val, Val218Leu, and Ile373Val. It is likely that other conservative amino acid substitutions not specifically cited here may be made in the sequence of Gas without inducing significant change in the activity of the wild type protein.

Gβ and Gγ Subunits

As of early 1994, at least four mammalian Gβ subunits were known and had been cloned. Both human and bovine clones of Gβ 1 (Codina J. et al. ( 1986) FEBS Lett 207,

187-192; Sugimoto K. et al. (1985) FEBS Lett 191, 235-240; Fong H.K.W. et al. (1986) Proc. Natl. Acad. Sci. USA 83, 2162-2166) and Gβ2 (Fong H.K.W. et al. (1987) Proc. Natl. Acad. Sci. USA 84, 3792-3796; Gao B. et al. (1987) Proc. Natl. Acad. Sci, USA 84, 6122-6125) have been isolated. A human Gβ3 (Levine M.A. et al. (1990) Proc, Natl. Acad. Sci. USA 87, 2329-2333) and mouse Gβ4 (Von Weizsacker E. et al. (1992)

Biochem. Biophys. Res. Commun. 183, 350-356) have also been cloned. Five complete mammalian Gγ subunits have been cloned: bovine Gγl (Hurley J.B. (1985) Proc. Natl. Acad. Sci. 81, 6948-6952), bovine Gγ2 (Robishaw J.D. (1989) J. Biol. Chem. 264, 15758-15761), bovine Gγ3 (Cali J.J et al. (1992) J. Biol. Chem. 267, 24023-24027), bovine and rat Gγ5 (Fisher K.J. and Aronson N.N. (1992) Mol. Cell. Biol. 12, 1585- 1591), and bovine Gγ7 (Cali JJ. et al. (1992) J. Biol. Chem. 267, 24023-24027). Part of a sixth Gγ subunit Gγ4, has been isolated (Gautam N. et al. (1990) Proc. Natl. Acad. Sci. USA 87, 7973-7977).

Various combinations of Gβ and Gγ have been observed in vitro and therefore have the potential to be physiologically active. For example, Gβ 1 can dimerize with Gγ

1, Gγ2, Gγ3, and Gγ5 (lώguez-Lluhi J.A. (1992) J. Biol. Chem. 267, 23409-23417), and βlγl, βlγ2, and βlγ5 stimulate phosphoinositide hydrolysis by phospholipase C β2 (Katz A. et al. (1992) Nature 360, 686-688). Other combinations are not observed. For example, Gβ2 does not dimerize with Gγl (Schmidt CJ. et al. (1992) J. Biol. Chem. 267, 13807-13810; Pronin A.N. and Gautam N. (1992) Proc. Natl. Acad. Sci. USA 89, 6220-6224). The effects of βγ dimers on adenylyl cyclases depends on both the isoform of adenylyl cyclase and the particular βγ dimer in question. While type 1 adenylyl cyclase is inhibited to varying degrees by different βγ dimers, the same dimers will potentiate the stimulatory effect of Gas on type 2 adenylyl cyclase (Ifliguez-Lluhi J.A. (1992) J. Biol. Chem. 267, 23409-23417). In both cases the potencies of βlγ2, βlγ3, β2 γ2, and β2γ3 are reported to be equivalent and greater than that of βlγl . The following additional references may be of value:

Gbeta5: A fifth member of the mammalian G-protein beta-subunit family. Expression in brain and activation of the beta 2 isotype of phospholipase C. Watson AJ, Katz A Simon MI, 19994, J Biol Chem, 269:22150-6.

Ggammaβ: A novel GTP-binding protein gamma-subunit, Ggammaβ, is expressed during neurogenesis in the olfactory and vomeronasal neruoepithelia. Ryba NJP, Tirindelli R, J Biol Chem, 270:6757-6767, 1995.

Ggamma7: Selective Tissue Distribution of G protein gamma subunits, Including a new form of the gamma subunits identified by cDNA Cloning. Cali JJ, Balcueva EA, Rybalkin I, and Robishaw JD, J Biol Chem, 267:24023-24027, 1992.

Ggammaό: Existence of two gamma subunits of the G proteins in brain. J.D. Robishaw,

V.K. Kalman, CR. Moomaw, CA. Slaughter, J Biol Chem, 264:15758-15761, 1989. Mutant Gβ and Gγ subunits may be designed in a manner analogous to that set forth with respect to Gα subunits.

Drug Screening

The identification of biological activity in new molecules has historically been accomplished through the use of in vitro assays or whole animals. Intact biological entities, either cells or whole organisms, have been used to screen for anti-bacterial, anti¬ fungal, anti-parasitic and anti-viral agents in vitro. Cultured mammalian cells have also been used in screens designed to detect potential therapeutic compounds. A variety of bioassay endpoints are exploited in mammalian cell screens including the stimulation of growth or differentiation of cells, changes in cell motility, the production of particular metabolites, the expression of specific proteins within ceils, altered protein function, and altered conductance properties. Cytotoxic compounds used in cancer chemotherapy have been identified through their ability to inhibit the growth of tumor cells in vitro and in vivo. In addition to cultures of dispersed cells, whole tissues have served in bioassays, as in those based on the contractility of muscle. In vitro testing is a preferred methodology in that it permits the design of high- throughput screens: small quantities of large numbers of compounds can be tested in a short period of time and at low expense. Optimally, animals are reserved for the latter stages of compound evaluation and are not used in the discovery phase; the use of whole animals is labor-intensive and extremely expensive. Microorganisms can also be easily exploited for use in rapid drug screens. Yeast provide a particularly attractive test system; extensive analysis of this organism has revealed the conservation of structure and function of a variety of proteins active in basic cellular processes in both yeast and higher eukaryotes. ■

The functional expression of a heterologous adenylyl cyclase in a host cell provides for the design of inexpensive screens useful in the identification of modulators of tiiis enzyme. Any chemical entity, or combination of chemical entities, whether natural or synthetic, may be screened for the ability to modulate the heterologous adenylyl cyclase. These modulators may act directly on the cyclase to alter the activity of the enzyme or may affect the ability of certain molecules to alter adenylyl cyclase activity, including, for example, Gas, Gai, or Gβγ.

Overview of the assay

The present invention makes available a rapid, effective assay for screening and identifying pharmaceutically effective compounds that specifically interact with and modulate the activity of recombinant adenylyl cyclase. In one embodiment the subject assay enables rapid screening of large numbers of compounds to identifying those

'Wn compounds of the library which induce or antagonize adenylyl cyclase bioactivity.

In certain embodiments, the assay is characterized by the use of recombinant cells to sample compounds for adenylyl cyclase agonists or antagonists. The reagent cells express a heterologous adenylyl cyclase capable of producing a detectable signal in the reagent cell.

In certain embodiments, the host (reagent) cell also produces the test compound which is being screened. For instance, the reagent cell can produce a test polypeptide, a test nucleic acid and/or a test carbohydrate which is screened for its ability to modulate the heterologous adenylyl cyclase activity. In such embodiments, a culture of such reagent cells will collectively provide a library of potential adenylyl cyclase effectors and those members of the library which either agonize or antagonize the adenylyl cyclase function can be selected and identified. Moreover, it will be apparent that the reagent cell can be used to detect agents which directly alter the activity of the heterologous adenylyl cyclase, or which act on some target upstream or downstream of the heterologous adenylyl cyclase.

In other embodiments, the test compound is exogenously added. In such embodiments the test compound is contacted with the reagent cell. Exemplary compounds which can be screened for activity include peptides, nucleic acids, carbohydrates, small organic molecules, and natural product extract libraries. In such embodiments, both compounds which agonize or antagonize the adenylyl cyclase function can be selected and identified. Moreover, it will be apparent that the reagent cell can be used to detect agents which directly alter the activity of the heterologous adenylyl cyclase, or which act on some target upstream or downstream of the heterologous adenylyl cyclase. In still other embodiments, the test compound is produced by cells which are cocultured with the reagent cells expressing a heterologous adenylyl cyclase.

In another preferred embodiment, the reagent cell is engineered to express an adenylyl cyclase which is has some degree of constitutive activity above the basal level of the wild-type enzyme. As described above, mutation to the adenylyl cyclase sequence can provide consitutively active adenylyl cyclase mutants. Alternatively, the cell can express an autocrine factor, such as a peptide, which causes activation of the adenylyl cyclase activity, ln still another embodiment an endogenous agent, such as forskolin, which causes activation of the enzyme can be applied to the reagent cell of the assay, preferably before or contemporaneously with the test compound. A salient feature of such embodiments of the subject assay is the enhanced sensitivity resulting from the higher basal level of adenylyl cyclase activity, and the concomitant improved ability to detect either agonist or antagonists of that activity. In particular, such ceils will generally have a greater dynamic range for detecting inhibitors of the adenylyl cyclase activity. Thus, inhibitors which produce weak signals can be detected, as well as discerned from those inhibitors which are more potent. Moreover, while not wishing to bound by any particular theory, it is possible that adenylyl cyclase activity requires a threshold activation signal. The constitutively active system, by overcoming any such threshold requirement, would more sensitive to, and hence more readily permit the detection of modulators of the adenylyl cyclase activity which, under the assay - conditions, seem only weakly active. It will be understood that any molecule that causes a statistically significant change in the adenylyl cyclase activity is of potential value. Hits which are apparently weak inhibitors or potentiators may nevertheless provide the pharmacophoric core to engineer more potent molecules.

Accordingly, the present invention provides a convenient format for discovering drugs which can be useful to modulate adenylyl cyclase activity, as well as to understand the pharmacology of compounds that specifically interact with the enzyme.

In particularly preferred embodiments, the ability of particular test compounds to modulate the enzymatic activity of the adenylyl cyclase can be scored for by detecting inhibition or activation of the adenylyl cyclase activity directly. Alternatively, the use of a reporter gene can provide a convenient readout. In any event a change, e.g., a statistically significant change, in the detection signal can be used to facilitate identification of those test compounds which are effectors of the target adenylyl cyclase. In certain alternative embodiments, it may be desirable to exploit the differences between yeast and mammalian adenylyl cyclase in order to screen for drugs which inhibit a yeast enzyme and not a mammalian enzyme. Such a differential screen, e.g., side-by-side comparison of otherwise identical cells expressing either a mammalian or yeast adenylyl cyclase, can lead to the identification of compounds which may be useful as antifungal agents.

Drugs, GeneraUy

Suitable chemical entities, from among which modulators of adenylyl cyclase may be identified, include nucleotide analogs (in particular, analogs of ATP, the natural substrate of adenylyl cyclase, and analogs of GTP, an activator of Gas). Forskolin, a diterpene, binds directiy to adenylyl cyclase and is a potent stimulator of that molecule. Therefore, forskolin-like structures, forskolin derivatives, and the diterpene class of compounds as a whole would be suitable chemical entities to test for effect on adenylyl cyclase activity. Synthetic peptides are also of interest. By way of example, peptides based on the calmodulin-binding domain of calmodulin-dependent adenylyl cyclases could serve as modulators of cyclase activity. In addition, peptides or molecules of any structure which inhibit the interaction between the cyclase and known endogenous modulators of adenylyl cyclase activity are of interest. Known endogenous adenylyl cyclase modulators include Ca2+, Ca2+/calmodulin, protein kinase C, protein kinase A, Gas, Gai, Gβγ, and adenosine.

Activation of protein kinase C can stimulate adenylyl cyclase activity and adenylyl cyclase has been shown to be a direct target for phosphorylation by protein kinase C (Yoshimasa et al. (1991) Nature 327,67-70). P2 purinergic and M5 muscarinic receptors, stimulators of the protein kinase C pathway, activate adenylyl cyclase (Johnson et al. (1991) J. Pharmacol. Exp. Ther. 39, 539-546). The adenylyl cyclases that have been cloned to date have been tested for susceptibility to regulation by protein kinase C activation: the basal activity of the type 2 enzyme is greatly increased by activation of protein kinase C whereas the activities of types 1 and 3 are affected to a lesser degree. In contrast, adenylyl cyclase types 4, 5, and 6 are not stimulated by protein kinase C activation (Jacobowitz et al. (1993) J. Biol. Chem. 268, 3829-3832; Yoshimura and Cooper (1993) J. Biol. Chem. 268, 4604-4607).

Studies have indicated a decrease in forskolin-stimulated adenylyl cyclase activity following exposure of mammalian cell membranes to protein kinase A (Premont et al. )1992) Endocrinology 131, 2774-2783; Yoshimura and Cooper (1992) PNAS 89, 6716-6720). A subset of the identified mammalian adenylyl cyclases appear to be susceptible to negative regulation by protein kinase A-dependent phosphorylation. Putative protein kinase A phosphorylation sites have been identified in the sequence of each of the enzymes, with the exception of type 4 adenylyl cyclase. The locations of the putative sites of phosphorylation are conserved in types 5 and 6 but vary among the other cyclases (Iyengar (1993) FASEB J. 7,768-775.

Peptide Drugs One class of potential modulators of particular interest is the peptide class. The term "peptide" is used herein to refer to a chain of two or more amino acids, with adjacent amino acids joined by peptide (-NHCO-) bonds. Thus, the peptides of the present invention include oligopeptides, polypeptides, and proteins. Preferably, the peptides of the present invention are 2 to 200, more preferably 5 to 50, amino acids in length. The minimum peptide length is chiefly dictated by the need to obtain sufficient potency as an activator or inhibitor. The maximum peptide length is only a function of synthetic convenience once an active peptide is identified. When peptide drugs are being assayed, the host cells may be engineered to express the peptides, rather than being exposed to the peptides simply by adding the peptides to the culture medium.

5 Peptide Libraries

In one embodiment, the host cells are engineered to express a peptide library. A "peptide library" is a collection of peptides of many different sequences (typically more than 1000 different sequences), which are prepared essentially simultaneously, in such a way that, if tested simultaneously for some activity, it is possible to characterize the io "positive" peptides. The peptide library of the present invention takes the form of a cell culture, in which essentially each cell expresses one, and usually only one, peptide of the library. While the diversity of the library is maximized if each cell produces a peptide of a different sequence, it is usually prudent to construct the library so there is some v. redundancy. Moreover, each sequence should be produced at assayable levels.

I 5 The peptides of the library are encoded by a mixture of DNA molecules of different sequence. Each peptide-encoding DNA molecule is ligated with a vector DNA molecule and the resulting recombinant DNA molecule is introduced into a host cell. Since it is a matter of chance which peptide-encoding DNA molecule is introduced into a particular cell, it is not predictable which peptide that cell will produce. However,

2 o based on a knowledge of the manner in which the mixture was prepared, one may make certain statistical predictions about the mixture of peptides in the peptide library.

It is convenient to speak of the peptides of the library as being composed of constant and variable residues. Ifthe nth residue is the same for all peptides of the library, it is said to be constant. Ifthe nth residues varies, depending on the peptide in 25 question, the residue is a variable one. The peptides of the library will have at least one, and usually more than one, variable residue. A variable residue may vary among any of two to any of all twenty of the genetically encoded amino acids; the range of possibilities may be different, if desired, for each of the variable residues of the peptide. Moreover, the frequency of occurrence of the allowed amino acids at particular residue

3 o positions may be the same or different. The peptide may also have one or more constant residues.

There are several ways in which to prepare the required DNA mixture that are well known in the art. For example, the DNAs are synthesized a base at a time. When variation is desired, at a base position dictated by the Genetic Code, a suitable mixture 35 of nucleotides is reacted with the nascent DNA, rather than the pure nucleotide reagent of conventional polynucleotide synthesis. In another method more control over the amino acid variation is provided. First, trinucleotide reagents are prepared, each trinucleotide being a codon of one (and only one) of the amino acids to be featured in the peptide library. When a particular variable residue is to be synthesized, a mixture is made of the appropriate trinucleotides and reacted with the nascent DNA.

Once the necessary "degenerate" DNA is complete, it can be joined with the DNA sequences necessary to assure the expression of the peptide, as discussed in more detail elsewhere, and the complete DNA construct must be introduced into the host cell.

Cytoplasmic Expression

Cells that express heterologous adenylyl cyclases can be used to discover novel activators or inhibitors of the various isoforms of adenylyl cyclase. Rich sources of potentially novel modulators of cyclase activity are semi-random or completely random peptide libraries. Indeed, suitably engineered, adenylyl cyclase-expressing cells may be transformed with plasmids encoding structurally diverse peptides, and the resulting transformants may be subjected to conditions that select for the growth of those particular transformants that encode a peptidic activator or inhibitor of the adenylyl cyclase. The plasmid encoding the active peptide may be isolated from any colony surviving the selection, the phenotype conferred by the plasmid may be confirmed with additional transformations, and the sequence predicted for the plasmid-encoded peptide may be determined by DNA sequencing. In this way, peptides may be identified that function as lead compounds for the design of small molecules that activate or inhibit the adenylyl cyclase.

The catalytic and regulatory regions of all known isoforms of adenylyl cyclase map to regions of the enzyme that are known to be on the cytoplasmic side of the plasma membrane. Accordingly, in one embodiment, peptides that target these regions are produced in the cytoplasm. Furthermore, given the proclivity of small peptides to be degraded when produced intracellularly, a strategy to ensure their stability is necessary. Finally, such strategies to protect the peptides from degradation should not interfere with their presentation to the adenylyl cyclase. To achieve intracellular production of stable peptides that can be efficiently presented to cytoplasmic targets like the adenylyl cyclases, an expression vector, such as for example, using the bacterial protein, thioredoxin A (trxA) can be used.

The tertiary structure of trxA reveals that its active site sequence, -CGPC-, forms a tight, disulfide-constrained loop on the surface of the protein (Katti, SK, DM LeMaster, H Eklund. 1990. Crystal structure of theioredoxin from E. coli at \jξβ angstroms resolution. J. Mol. Biol. 212:167-184.). This loop permits insertion of a wide diversity of peptide sequences without untoward effects on peptide folding (LaVaiiie, 5 ER, EA Diblasio, S Kovacic, KL Grant, PF Schednel, JM McCoy. 1993. A thioredoxin gene fusion expression system that circumvents inclusion body formation in the £ coli cytoplasm. Bio/Technology 1 1 : 1187-1 193). The peptides that are inserted into the active site loop should therefore be stable and available to interact with other proteins. As described in more detail in the appended examples, these features of trxA were o successfully exploited in a two-hybrid system, in which a fusion protein containing trxA, a nuclear localization sequence, a hemagglutinin epitope tag, and a transcriptional activation domain was used as for insertion of peptides of variable sequence (Colas, P, B Cohen, T Jessen, I Grishina, J McCoy, R Brent. 1996. Genetic selection of peptide aptamers that recognize and inhibit cyclin-dependent kinase 2. Nature 380:548-556). 5

Autocrine Cells

In other embodiments host cells, preferably yeast cells, are engineered to produce both peptide drugs and the receptors. Assays using cells engineered to produce the receptor, but that do not produce the drugs themselves, are inefficient To utilize them o one must bring a sufficient concentration of each drug into contact with a number of cells in order to detect whether or not the drug has an action. Therefore, a microtiter plate well or test tube must be used for each drug. The drug must be synthesized in advance and be sufficiently pure to judge its action in the assay. When the cell produces the drug, the effective concentration is higher. 5

Periplasmic Secretion

The cytoplasm of the yeast cell is bounded by a lipid bilayer called the plasma membrane. Between this plasma membrane and the cell wall is the periplasmic space. o Peptides secreted by yeast cells cross the plasma membrane through a variety of mechanisms and thereby enter the periplasmic space. The secreted peptides are then free to interact with other molecules that are present in the periplasm or displayed on the outer surface of the plasma membrane. The peptides then either undergo re-uptake into the cell, diffuse through the cell wall into the medium, or become degraded within tiie periplasmic space.

The peptide library may be secreted into the periplasm by one of two distinct mechanisms, depending on the nature of the expression system to which they are linked. In one system, the peptide may be structurally linked to a yeast signal sequence, such as that present in the α-factor precursor, which directs secretion through the endoplasmic reticulum and Golgi apparatus. Since this is the same route that the receptor protein s follows in its journey to the plasma membrane, opportunity exists in cells expressing both the receptor and the peptide library for a specific peptide to interact with the receptor during transit through the secretory pathway. This has been postulated to occur in mammalian cells exhibiting autocrine activation. Such interaction would likely yield activation of the linked pheromone response pathway during transit, which would still o allow identification of those cells expressing a peptide agonist. For situations in which peptide antagonists to externally applied receptor agonist are sought, this system would still be effective, since both the peptide antagonist and receptor would be delivered to the outside of the cell in concert. Thus, those cells producing an antagonist would be selectable, since the peptide antagonist would be properly and timely situated to prevent s the receptor from being stimulated by the externally applied agonist.

An alternative mechanism for delivering peptides to the periplasmic space is to use the ATP-dependent transporters of the STE6/MDR1 class. This transport pathway and the signals that direct a protein or peptide to this pathway are not as well characterized as is the endoplasmic reticulum-based secretory pathway. Nonetheless, o these transporters apparently can efficiently export certain peptides directly across the plasma membrane, without the peptides having to transit the ER/Golgi pathway. At least a subset of peptides can be secreted through this pathway by expressing the library in context of the a-factor prosequence and terminal tetrapeptide. The possible advantage of this system is that the receptor and peptide do not come into contact until both are 5 delivered to the external surface of the cell. Thus, this system strictly mimics the situation of an agonist or antagonist that is normally delivered from outside the cell. Use of either of the described pathways is within the scope of the invention.

The present invention does not require periplasmic secretion, or, if such secretion is provided, any particular secretion signal or transport pathway. 0

Detection of Inhibition or Activation of Adenylyl Cyclase Activity

For the engineered cells to be useful in screening drugs for the ability to inhibit or activate a adenylyl cyclase, there must be a detectable change in adenylyl cyclase activity. This change (the signal) must be detectable against the background (the basal 5 level of adenylyl cyclase activity in the absence of the drug). The signal may be change in the growth rate of the cells, or other phenotypic changes, such as a color change or W

- 51 -

luminescence. -ΛCΪ*

The endogenous adenylyl cyclase contributes to background. This contribution may be reduced by using host cells in which the endogenous gene has been mutated and the endogenous protein is not functional or the yeast cell is a conditional mutant For eample, in the case of yeast, cells which bear a mutant allele of yeast adenylyl cyclase, cdc 35-1, as the endogenous adenylyl cyclase.

This allele encodes a temperature-sensitive enzyme that is active when the Cells are grown at room temperature; at 30°C or higher, the cyclase is inactive and the yeast cells are incapable of growth. This strain also preferably displays the Cam phenotype, which reflects an ability for growth rescue at the higher temperatures through the addition of exogenous cAMP (since yeast displaying the Cam phenotype are capable of taking up and utilizing cAMP)(Matsumoto et al. (1982) J. Bacteriol. 150, 277-28%

Thus, in a preferred embodiment, the background of the test strain provides for a simple metric of the function of the mammalian cyclase when that protein is introduced into these cells via an expression plasmid. If the heterologous adenylyl cyclase is active, the yeast grow at temperatures greater than 30°C, in a range where the yeast cyclase is non-functional. Growth of the test yeast cells is a simple and elegant indicator of the activity of the mammalian enzyme.

Alternatively, one may use a host strain in which the gene encoding the endogenous adenylyl cyclase is completely, unconditionally inactivated, e.g., by deletion, e.g., in the case of yeast the CYR1 mutant. Such yeast could grow in the presence of glucose provided that they also exhibited the Cam phenotype and were provided with exogenous cAMP. Alternatively, these cells would also be capable of growth if they expressed rat adenylyl cyclase and Gas. Thus, hosts other than the cdc 35-1 mutant strain that was used in the invention reported here could be used and may have certain advantages. For example, spontaneous reversions in the cdc35-l allele could give rise to adenylyl cyclase with wild type activity. Such reversions can be virtually eliminated using deletion mutants of the endogenous adenylyl cyclase. Rather than relying on the effect of the adenylyl cyclase on endogenous components of the cell for signal generation, it is also possible to select or screen for adenylyl cyclase activity by means of a marker gene engineered into the host cell. A marker gene is a gene whose expression causes a phenotypic change which is screenable or selectable. Ifthe change is selectable, the phenotypic change creates a difference in the growth or survival rate between cells which express the marker gene and those which do not. If the change is screenable, the phenotype change creates a difference in some detectable characteristic of the cells, by which the cells which express the marker may be distinguished from those which do not. Selection is preferable to screening.

Situations may occur wherein it would be advantageous to use a readout other than growth as an assay of cyclase activity. While growth rates require at least a day to measure, direct readouts of transcriptional activity offer the possibility of more rapid assays. For example, in yeast, by placing the bacterial gene encoding lacZ under the control of the FUS1 promoter, activation of the yeast pheromone response pathway can be detected in less than an hour by monitoring the ability of permeabilized yeast to produce color from a chromogenic substrate. The rapidity of such a readout would, in itself, be advantageous. And such a readout would be necessary to monitor cyclase activity under conditions where the yeast do not grow.

In mammalian cells cAMP influences transcription from a panel of genes by activating protein kinase A (PKA), which phosphorylates and thereby activates transcription factor CREB (reviewed in Brindle, PK and Montminy, MR (1992) Curr. Opinion Gen. Dev., 2:199-204). In yeast, however, only a few genes are known to be affected by PKA: ADH2 (Denis, CL et al (1992) Mol Cell Biol 12:1507-1514), UBI4 (Tanaka, K et al (1988) EMBO J. 7:495-502), CTT1 (Marchler, G (1993) EMBO J 12:1997-2003), and various ribosomal protein genes like RPS13 (Sussel, L and Shore, D (1991) Proc Natl Acad Sci USA 88:7749-7753). The cAMP-dependent activation of the ribosomal genes is mediated by the yeast RAP1 transcription factor, which binds to the DNA sequence RMACCCANNCAYY in a wide variety of yeast promoters (Klein, C and Struhl, K (1994) Mol Cell Biol 14:1920-1928). Indeed, the RAP1 binding site can increase transcription from a heterologous HIS3 promoter (Klein, C and Struhl, K (1994) Mol Cell Biol 14:1920-1928), suggesting that an upstream regulatory sequence containing RAP1 binding sites could be constructed and linked to a reporter gene such as lacZ. Such a construct could provide a rapid, colorimentric readout of the activity of mammalian adenylyl cyclase in yeast. Other useful reporters include such genes as alkaline phosphatase, chloramphenicol acetyl transferase, luciferase and fluorescent green protein (FGP), which can be used to generate colorimetric, luminescent, fluorescent or radio-isotopic readouts. (The latter requires a radioisotopic substrate.) Thus, a marker gene may be coupled to the heterologous adenylyl cyclase so that expression of the marker gene is dependent on activity of the adenylyl cyclase. This coupling may be achieved by operably linking the marker gene to a cyclic AMP- responsive promoter. The term "cyclic AMP-responsive promoter" indicates a promoter which is regulated by either cyclic AMP or a metabolic product produced as a consequence of cyclic AMP production. For example, the cauliflower mosaic virus 35S RNA promoter appears to be regulated by cAMP in S. cerevisiae (Ruth et al. (1992) Mol. Gen. Genet. 235, 365-372). The promoter could be one which is natively ,, responsive to cyclic AMP, or one engineered to be so responsive by incoφoration of a suitable operator. , *

In one embodiment, the promoter is activated upon activation of the cyclase, in which case, for selection, the expression of the marker gene should result in a benefit to the cell. A preferred marker gene is the imidazoleglycerol phosphate dehydratase gene (HIS3). If a cyclic AMP responsive promoter is operably linked to a beneficial gene, the cells will be useful in screening or selecting for adenylyl cyclase activators. If ifjJH linked to a deleterious gene, the cells will be useful in screening or selecting for inhibitors.

Alternatively, the promoter may be one which is repressed by cyclic AMP, thereby preventing expression of a product that is deleterious to the cell. With a cyclic AMP-repressed promoter, one screens for agonists by linking the promoter to a deleterious gene, and for antagonists, by linking it to a beneficial gene. Repression may be achieved by operably linking a cyclic AMP-induced promoter to a gene encoding mRNA that is antisense to at least a portion of the mRNA encoded by the marker gene (whether in the coding or flanking regions), so as to inhibit translation of that mRNA. Repression may also be obtained by linking a cyclic AMP- induced promoter to a gene encoding a DN A-binding repressor protein, and incorporating a suitable operator site into the promoter or other suitable region of the marker gene.

Suitable positively selectable (beneficial) genes that can be used in yeast include the following: URA3, LYS2, HIS3, LEU2, TRP1; ADE1,2,3,4,5,7,8; ARGI,3,4A6\8; HISl.4,5; ILV1,2,5; THR1,4; TRP2,3,4,5; LEU1,4; MET2,3,4,8,9,14,16,19; URAl.2,4,5,10; HOM3.6; ASP3; CHOI; ARO 2,7; CYS3; OLEl; INOl.2,4; PR01,3 Countless other genes are potential selective markers. The above are involved in well- characterized biosynthetic pathways.

The imidazoleglycerol phosphate dehydratase (IGP dehydratase) gene (HΪS3) is preferred because it is both quite sensitive and can be selected over a broad range of expression levels. In the simplest case, the cell is auxotrophic for histidine (requires histidine for growth) in the absence of activation. Activation leads to synthesis of the enzyme and the cell becomes prototrophic for histidine (does not require histidine). Thus the selection is for growth in the absence of histidine. Since only a few molecules per cell of IGP dehydratase are required for histidine prototrophy, the assay is very sensitive.

In a more complex version of the assay, yeast cells can be selected for resistance to aminotriazole (AT), a drug that inhibits the activity of IGP dehydratase. Cells with low, fixed level of expression of HIS3 are sensitive to the drug, while cells with higher levels are resistant. The amount of AT can be selected to inhibit cells with a basal level of HIS3 expression (whatever that level is) but allow growth of cells with an induced level of expression. In this case selection is for growth in the absence of histidine and in the presence of a suitable level of AT.

In appropriate assays, so-called counterselectable or negatively selectable genes may be used. For example, when yeast cells are used, suitable genes include: URA3 (orotidine-5'-phosphate decarboxylase; inhibits growth on 5-fluoroorotic acid), LYS2 (2- aminoadipate reductase; inhibits growth on α-aminoadipate as sole nitrogen source), GAL1 (encodes galactokinase; expression of GAL 1 is toxic in the presence of galactose in strains that contain mutations in either GAL7 (encodes galactotransferase) or GAL 10 (encodes epimerase) genes); CYH2 (encodes ribosomal protein L29; cycloheximide- sensitive allele is dominant to resistant allele), CAN1 (encodes arginine permease; null allele confers resistance to the arginine analog canavanine), and other recessive drug- resistant markers.

The marker gene may also be a screenable gene. The screened characteristic may be a change in cell morphology, metabolism or other screenable features. Suitable markers include beta-galactosidase (Xgal, C12FDG, Salmon-gal, Magenta-Gal (latter two from Biosynth Ag)), acid or alkaline phosphatase, horseradish peroxidase, exo¬ glucanase (product of yeast exbl gene; nonessential, secreted); luciferase; and chloramphenicol transferase. Some of the above can be engineered so that they are secreted (although not β-galactosidase). The preferred screenable marker gene is beta- galactosidase; yeast cells expressing the enzyme convert the colorless substrate Xgal into a blue pigment. Again, the promoter may be cyclic AMP-induced or cyclic AMP- inhibitεd.

To identify inhibitors of adenylyl cyclase, in one embodiment the gene encoding MATα2p is placed under the control of a promoter containing RAP1 binding site, so that transcription of MATα2 is cAMP dependent. A lacZ gene construct is then placed under the control of any haploid-specific promoter. In a yeast cell transformed with these two constructs, the expression of lacZ — and therefore the development of blue color -- will then be under the control of MATα2p, which in turn will be controlled by cAMP levels. Specifically, in the presence of active adenylyl cyclase, the yeast will be white due to expression of MATα2p and consequent inhibition of lacZ expression. If cyclase is inhibited, MATα2p will decline, resulting in increased lacZ expression from the haploid-specific promoter. Examples of haploid-specific promoters include but are not limited to promoters for the GPA 1 , STE4, STE 18 genes.

Another strategy to discover inhibitors of adenylyl cyclase is to use promoters that are repressed by cAMP. cAMP-repressible elements are found in the promoters of SSA3 (Boorstein, WR and Craig, EA, 1990, EMBO J, 9:2543-2553) and CTTI ! 5 (Marchler, GC et al 1993, EMBO J, 12: 1997-2003), and may be present in the UB14 promoter (Tanaka, KK et al, 1989, EMBO J, 7:495-502). These promoters, if engineered to direct the transcription of a screenable marker like lacZ, can provide a readout of inhibition of adenylyl cyclase activity. Specifically, inhibition of adenylyl cyclase will result in reduced cAMP levels that will relieve repression of lacZ o expression, resulting in the development of blue color within several hours in a culture of host cells exposed to an inhibitor. Such readouts can be very sensitive to small changes in levels of cAMP caused by agonists or antagonists and do not have some of the potential drawbacks of growth readouts discussed above.

5 Detection of Inhibition or Activation of Proteins Other Than Adenylyl Cyclase

It is possible to detect inhibitors or activators of proteins other than adenylyl cyclase provided that the host cell expresses or is engineered to express the protein of interest in such a manner that it is functionally "coupled", directly or indirectly, to the adenylyl cyclase. o For example, the cells could be used to screen for inhibitors or activators of a cyclase associated protein (CAP). CAP has been identified in yeast which also Interacts with the yeast cytoskeleton (Kawmukai et al, 1992, Mol. Biol. Cell, 3:167-180) and may be a regulator of the yeast adenylyl cyclase activity.

Similarly, a human homolog of the yeast CAP gene has been identified and this 5 may function as a regulator of the human adenylyl cyclase. Such regulators may potentially feed in signals from other signal transduction pathways. These and other as yet unidentified regulators which interact with adenylyl cyclase(s) and modify their activity may be used in further embodiments of the present invention for the discovery of agonist or antagonist activities affecting any step of the signal transduction o pathway(s) which are coupled to the cloned mammalian adenylyl cyclase.

The best characterized regulator of adenylyl cyclase is the G protein, or, more precisely, the Gα subunit and/or the Gβγ complex resulting from the dissociation Of the G protein. Consequently, the host cells of the present invention may be used to test drugs for the ability to interact with an exogenous (or chimeric) G protein-coupled s receptor or, in the case of yeast cells, another PSP surrogate. The yeast cells must express both the exogenous G protein-coupled receptor (or other PSP surrogate), and a complementary (usually exogenous or chimeric) G protein (or other PSPs necessary for the PSP surrogate to function in the pheromone system, if need be after activation by a drug), and these molecules must be presented in such a manner that adenylyl cyclase activity is affected. For example, if a yeast cell expresses a G protein-coupled receptor, and the stimulation of this receptor results in the dissociation of a G protein also expressed by the yeast cell, and either the Gα subunit or the Gβγ complex then interacts with the mammalian adenylyl cyclase, to increase or decrease its level of activity, the G protein- coupled receptor (and the G protein) may be said to be coupled to the adenylyl cyclase. That is, inhibitors or activators of the receptor will affect adenylyl cyclase activity.

Thus, a host cell may be engineered so it can be used to detect inhibitors or activators of an exogenous (usually mammalian) G protein coupled receptor by virtue of their effect on the activity of a coupled adenylyl cyclase. This coupling may be facilitated by the use of corresponding exogenous (or chimeric) Gα, Gβ and/or Gγ subunits, and the signal-to-noise ratio may be improved by partial or total inactivation of the endogenous genes (or their products).

An exogenous G protein coupled receptor is one example of a PSP surrogate, the corresponding yeast protein being the α- or a-factor receptor. However, it is possible to screen for inhibitors or activators of surrogates of other PSPs, provided that they directly or indirectly affect the stimulation of an endogenous or exogenous G protein-coupled receptor, and thereby of the G protein-coupled adenylyl cyclase.

Examples of these "upstream" PSPs include: Farnesyltransferases and carboxymethyltransferases. After expression, a-factor is famesylated by RAMlp and RAM2p and carboxymethylated by Stel4p. These modifications are required for activity.

RAMlp and RAM2p are homologous to the subunits of the heterodimeric mammalian famesyltransferase, which itself is necessary for membrane association of mammalian Ras proteins. If a yeast cell is engineered to express the mammalian famesyltransferase, it may be used to identify drugs which interact with that enzyme by determining whether a functional a-factor is produced. Similarly, Stel4p is homologous to mammalian carboxymethyltransferases, which play regulatory roles in controlling the function of low molecular weight G proteins (Ras, Rho, Rab).

Proteases. The PSP may be a yeast protease, such as KEX2, STE13 or KEX 1. Yeast α- factor pheromone is generated through the controlled and limited proteolysis of precursor proteins by these proteases. A yeast cell may be engineered to express an inactive precursor of yeast α-factor in which a peptide linker, corresponding to the cleavage site of a surrogate non-yeast protease, is incorporated so that cleavage will liberate mature α-factor (or its functional homologue). For example, the PSP surrogate may be HIV protease, with the cleavage site of HIV protease being substituted for tiie yeast protease cleavage sites in the α-factor precursor. The precursor and the HIV protease are co-expressed in the yeast cell. Proteolysis by HIV protease will give rise to production of mature α-factor and initiation of signal transduction. This system may be used to identify inhibitors of HIV protease.

Preferably, unlike yeast cells occurring in nature, the yeast cell is engineered not only to express the α-factor precursor, but also the α-factor receptor, so that a sittgb haploid type of yeast is all that is required to conduct the assay.

ABC Transporters. Ste6 is the yeast ABC transporter necessary for the export of a- factor. The yeast cell is engineered to express both a-factor and a foreign ABC transporter. This transporter may be one which is not by itself, able to transport a- factor, but which in the presence of a drug of interest, is capable of doing so, or it may be one which is already functional. Preferably, the yeast cell is engineered to express not only a-factor, but also the a-factor receptor.

If an endogenous pheromone receptor (or other cognate PSP) is produced by the yeast cell, the assay may not be able to readily distinguish between peptides which interact with the pheromone receptor (or other cognate PSP) and those which interact with the exogenous receptor (or other PSP surrogate). It is therefore desirable tiiat the endogenous gene be deleted or otherwise rendered nonfunctional.

The present invention may be used to identify inhibitors or activators of many mammalian receptors, including but not limited to, receptor tyrosine kinases and cytokine receptors (such as those for IL-3, IL-5, GM-CSF, M-CSF and EPO etc.), G protein-coupled chemokine receptors (such as RANTES, MCP-3, MCP-1, MlP-lα and IL-8 receptor), growth factor receptors (such as EGFR and PDGFR etc.), and multi- subunit immune recognition receptors also known as MIRRs (such as FcεRI, and FcγRII etc.). Further receptors useful in the current invention include the G protein-coupled C5a peptide receptor, the thrombin peptide receptor (PAR1 ) and its homolog PAR2, formyl peptide and bradykinin receptors, muscarinic receptors, adrenergic receptors, melatonin, galanin, glucagon and orphan receptors and transporter proteins such as the multidrug resistance protein (MDR).

G Protein-Coupled Receptors The host cells of the present invention may be used to identify drugs which modulate the activity of a heterologous, e.g., mammalian G protein-coupled receptor. In this embodiment, the yeast cell is engineered to express a mammalian G protein-coupled receptor, Most G protein-coupled receptors are comprised of a single protein chain that is threaded through the plasma membrane seven times. Such receptors are often referred to as seven-transmembrane receptors (STRs). More than a hundred different STRs have been found, including many distinct receptors that bind the same ligand, and there are likely many more STRs awaiting discovery.

In addition, STRs have been identified for which the natural ligands are unknown; these receptors are termed "orphan" G protein-coupled receptors. Examples include receptors cloned by Neote et al. Cell 72, 415 (1993); Kouba et al. FEBS Lett. 321, 173 (1993); Birkenbach et al. J. Virol. 67, 2209 (1993).

The "exogenous G protein-coupled receptors" of the present invention may be any G protein-coupled receptor which is exogenous to the wild-type yeast cell which is to be genetically engineered for the purpose of the present invention. This receptor may be a plant or animal cell receptor. Screening for binding to plant cell receptors may be useful in the development of, e.g., herbicides. In the case of an animal receptor, it may be of invertebrate or vertebrate origin. If an invertebrate receptor, an insect receptor is preferred, and would facilitate development of insecticides. The receptor may also be a vertebrate, more preferably a mammalian, still more preferably a human receptor. The exogenous receptor is also preferably a seven transmembrane segment receptor.

Suitable receptors include, but are not limited to, dopaminergic, muscarinic cholinergic, α-adrenergic, β-adrenergic, opioid (including delta and mu), cannabinoid, serotoninergic, and GABAergic receptors. Other suitable receptors are listed in Table 2 5 of WO94/23025. The term "receptor," as used herein, encompasses both naturally occurring and mutant receptors.

Many of these G protein-coupled receptors, like the yeast a- and α-factor receptors, contain seven hydrophobic amino acid-rich regions which are assumed to lie within the plasma membrane. Specific human G protein-coupled STRs for which genes o have been isolated and for which expression vectors could be constructed include those listed in Table 2 of WO94/23025. Thus, the gene would be operably linked to a promoter functional in yeast and to a signal sequence functional in yeast. Suitable promoters include Ste2, Ste3 and gal 10. Suitable signal sequences include those of Ste2, Ste3 and of other genes which encode proteins secreted by yeast cells. Preferably, the 5 codons of the gene would be optimized for expression in yeast. See Hoekema et al., Mol. Cell. Biol., 7:2914-24 (1987); Sharp, et al., 14:5125-43 (1986). The homology of STRs is discussed in Dohlman et al., Ann. Rev. Biochem., 60:653-88 (1991). When STRs are compared, a distinct spatial pattern of homology is discemable. The transmembrane domains are often the most similar, whereas the N- and C-terminal regions, and the cytoplasmic loop connecting transmembrane segments V 5 and VI are more divergent.

The functional significance of different STR regions has been studied by introducing point mutations (both substitutions and deletions) and by constructing; chimeras of different but related STRs. Synthetic peptides corresponding to individual segments have also been tested for activity. Affinity labeling has been used to identify I o ligand binding sites.

It is conceivable that a foreign receptor which is expressed in yeast will ■ functionally integrate into the yeast membrane, and there interact with the endogenous yeast G protein. More likely, either the receptor will need to be modified (e.g., by replacing its V-VI loop with that of the yeast STE2 or STE3 receptor), or a compatible G 15 protein should be provided.

If tiie wild-type exogenous G protein-coupled receptor cannot be made functional in yeast it may be mutated for this purpose. A comparison would be made of tiie amino acid sequences of the exogenous receptor and of the yeast receptors, and regions of high and low homology identified. Trial mutations would then be made to distinguish 20 regions involved in ligand or G protein binding, from those necessary for functional integration in tiie membrane. The exogenous receptor would then be mutated in the latter region to more closely resemble the yeast receptor, until functional integration was achieved. If this were insufficient to achieve functionality, mutations would next be made in the regions involved in G protein binding. Mutations would be made in regions

25 involved in ligand binding only as a last resort, and then an effort would be made to preserve ligand binding by making conservative substitutions whenever possible.

Preferably, the yeast genome is modified so that it is unable to produce the „" endogenous a- and α-factor receptors in functional form. Otherwise, a positive assay score might reflect the ability of a peptide to activate the endogenous G protein-coupled o receptor, and not the receptor of interest

When the PSP surrogate is an exogenous G protein-coupled receptor, the yeast cell must be able to produce a G protein which is activated by the exogenous receptor, and which can in turn activate the mammalian adenylyl cyclase. It is possible that die endogenous yeast Gα subunit (e.g., GPA) will be sufficiently homologous to the 5 "cognate" Gα subunit which is natively associated with the exogenous receptor for coupling to occur. More likely, it will be necessary to genetically engineer the yeast cell to produce a foreign Gα subunit which can properly interact with the exogenous receptor. For example, the Gα subunit of the yeast G protein may be replaced by the G α subunit natively associated with the exogenous receptor.

Dietzel and Kurjan, Cell, 50:1001 (1987) demonstrated that rat Gas functionally coupled to the yeast Gβγ complex. However, rat Gαi2 complemented only when substantially overexpressed, while GαO did not complement at all. Kang, et al., Mol.

Cell. Biol., 10:2582 (1990). Consequently, with some foreign Gα subunits, it is not feasible to simply replace the yeast Gα.

Therefore, alternatively, the yeast Gα subunit is replaced by a chimeric Gα subunit in which a portion, e.g., at least about 20, more preferably at least about 40, amino acids, which is substantially homologous with the corresponding residues of the amino terminus of the yeast Gα, is fused to a sequence substantially homologous with the main body of a mammalian (or other exogenous) Gα. While 40 amino acids is the suggested starting point, shorter or longer portions may be tested to determine the minimum length required for coupling to yeast Gβγ and the maximum length compatible with retention of coupling to the exogenous receptor. It is presently believed that only the final 10 or 20 amino acids at the carboxy terminus of the Gα subunit are required for interaction with the receptor.

Ifthe yeast cell is engineered to express a mammalian or chimeric Gαi, one may screen for specific modulators of the inhibitory activity of that Gαi. The following receptors have been shown to activate adenylyl cyclase via signalling to Gas: βl -adrenergic histamine H2 β2-adrenergic glucagon

Dl dopamine calcitonin D5 dopamine glucagon-like peptide 1 vasoactive intestinal peptide adrenocorticotropic follicle stimulating hormone melanocyte stimulating hormone luteinizing hormone/choriogonadotropin secretin vasopressin V2 adenosine A2 thyroid stimulating hormone

The following receptors have been shown to be involved in the inhibition of adenylyl cyclase; this inhibition is believed to occur via signalling to a G protein member of the Gi/Go family: adenosine A 1 angiotensin II type 1 adenosine A3 cannabinoid α-adrenergic bradykinin muscarinic acetylcholine receptor type 2 GAB A muscarinic acetylcholine receptor type 4 galanin -*-^■

D2 dopamine metabotropic glutamate type 2 D4 dopamine metabotropic glutamate type 3

5HT1 a metabotropic glutamate type 4

5HTlb melatonin

5HTld NPY1

5HTle NPY2 5HTlf somatostatin 2 formylMet-Leu-Phe somatostatin 3 delta opioid somatostatin 4

The yeast cell may also engineered to express mammalian or chimeric Gβ and/or Gγ subunits as well as mammalian or chimeric Gas. The Gas and Gβγ subunits will associate to form a heterotrimeric G protein to which the receptor is coupled. Stimulation of the receptor will greatly increase the rate of activation of Gas causing dissociation of the Gas and Gβγ; the Gas will subsequently activate the mammalian adenylyl cyclase.

Farnesyltransferases

The activity of yeast a-factor requires its farnesylation (mediated by protein famesyltransferase, comprised of Ramlp and Ram2p), proteolytic cleavage of the C* terminal 3 amino acids of the primary translation product (mediated by an as yet unidentified enzyme), and carboxymethylation of the C-terminal cysteine (mediated by

Stel4p). The yeast and mammalian farnesyltransferases are structurally and functionally similar (Gomez R et al., Biochem. J. 289:25-31, 1993; Kohl NE et al., J. Biol. Chem. 266:18884-8, 1991). Sequence homologies exist between the genes encoding the α and β subunits of the yeast famesyltransferase (RAM2 and RAMl, respectively) and the genes encoding the α and β subunits of the mammalian famesytransferase (Kohl NE et al., J. Biol. Chem.266:18884-8, 1991; Chen WJ et al., Cell 66:327-34, 1991). It has been observed that the β subunit of mammalian famesytransferase and Ramlp are 37% identical in amino acid sequence (Chen WJ et al., Cell 66:327-34, 1991).

The importance of a screen for inhibitors of famesyl-transferase is suggested by the facts that mammalian p21 ras, a preeminent regulator of the growth and differentiation of mammalian cells that is involved in a variety of cancers, is a substrate for the famesyltransferase and that farnesylation of p2 Iras is required for its activity. In fact, a synthetic organic inhibitor of farnesyl protein transferase has been shown to selectively inhibit ras-dependent cell transformation (Kohl et al., Science 260, 1934 (1993). Of the two subunits of famesyltransferase, the β subunit is a more attractive 5 target for inhibitors, since it is apparently dedicated to farnesyla-tion. The α subunit in contrast, is shared by geranyl-geranyltransferase I, an enzyme involved in the modification of the Gγ subunits of heterotrimeric G proteins and small molecular weight G proteins of the Rho/Rac family. While the β subunit is dedicated to farnesylation, the mammalian famesyltransferase has a variety of substrates in addition to p21 ras. The o effect of inhibitors of the β subunit on the farnesylation of these other substrates, e.g., lamin proteins, transducin-γ and rhodopsin kinase, will be considered in the design and use of potential famesyltransferase inhibitors.

It has not yet been demonstrated that the homologous mammalian gene will functionally substitute for yeast Ramlp, however, this can be formally tested using raml s mutants and a vector expressing the mammalian gene encoding the β subunit of the famesyltransferase. Ifthe mammalian β subunit can function in place of Ramlp, test cells will be both viable (as a result of farnesylation of Ras) and competent for mating (as a result of farnesylation of a-factor).

Ifthe mammalian gene encoding the β subunit of farnesy 1-transferase o complements raml , yeast would provide a test system for the discovery of potential inhibitors of mammalian farnesyl-transferase. Specifically, MATa yeast tester cells could be exploited that: 1. carry the gene for the β subunit of mammalian famesyltransferase in lieu of RAMl ; 2. carry the cam mutation that renders the strains resistant to loss of Ras function in the presence of c AMP; 3. respond to a-factor which s they export by virtue of heterologous expression of Ste3p; 4. respond to autocrine a- factor such that they cannot grow on media containing galactose. The latter characteristic will require expression of GAL 1 under the control of a pheromone- responsive promoter and cells engineered to contain mutated GAL7 or GAL 10 genes. Expression of GAL 1 is toxic in the presence of galactose in strains which contain ₀ mutations in either the GAL7 or GAL 10 genes. Signaling through the pheromone response pathway would render cells so engineered galactose-sensitive. Exposure of such strains to compounds which inhibit the β subunit of farnesyl-transferase will confer upon these cells the ability to grow on media containing galactose and cAMP.

Ifthe mammalian gene encoding the β subunit of famesyltransferase (and all 5 modified versions of the gene) fails to complement ram 1 , the wild-type Raml p may be used as a surrogate target for potential effectors of mammalian famesyltransferase. Specifically, tester cells MATa yeast strains that: 1. carry the cam mutation that renders the strains resistent to loss of RAS function in the presence of c AMP; 2. respond to a- factor which they export by virtue of heterologous expression of Ste3p; 3. respond to autocrine a-factor such that they cannot grow on media containing galactose may be s used. Exposure of such strains to compounds which inhibit the β subunit of famesy_¬ transferase will confer upon these cells the ability to grow on media containing galactose and cAMP.

In the strategies outlined above, it is desirable to dis-criminate inhibitors of famesytransferase from compounds that either directly block the negative response to a- o factor, e.g. by interfering with the interaction of the Ste4-Ste 18 complex with its effector, or by blocking the production of a-factor by a mechanism that does not involve famesyltransferase. Controls would identify such false positives. Candidate agents will be tested on a MATa strain that is engineered to secrete α-factor and to respond to me secreted a-factor by failing to grow on galactose-containing media, as in the negative s selection scheme outlined above. The strain will express wild type Ramlp. Any agent that enables these cells to grow on media containing galactose and cAMP will not be acting as an inhibitor of famesyltransferase.

Candidate compounds which pass the foregoing test may act by targeting SteI4p, Ste6p, or otiier proteins involved in the maturation and export of a-factor, rather titan ⁰ farnesyl-transferase. (Note, however, that compounds that inhibit processes critical to cell survival will not give rise to false positives. For example, since the protease -«^" responsible for the endoproteolytic removal of the C-terminal tripeptide of the a-factor precursor likely participates in the processing of Gg and members of the Rho/Rac iamily of proteins, inhibitors of this enzyme may not permit growth of the tester cells). Of the 5 proteins involved in the production of a-factor, only the famesyltransferase is also a major determinant of RAS function. Due to this effect, raml mutants are defective for growth at 30°C and completely unable to grow at 37 (He B et al., Proc Natl Acad Sci 88:11373-7, 1991). Tester cells (described above) can be grown in the presence of a candidate inhibitor on galactose-containing media ± cAMP. Ifthe test compound o inhibits fariiesyltransferase, cells will be capable of growth on galactose + cAMP but not on galactose in the absence of cAMP. This difference may be most obvious at 37°. If, on the other hand, the test compound inhibits other proteins involved in a-factor production, cells will grow on galactose-containing media regardless of the presence or absence of cAMP. 5 Compounds which pass the above tests are likely inhibitors of famesyltransferase. This can be confirmed and their potencies determined with direct in O 97/40170 PC1 US / 1

- 64 -

vitro enzyme assays. Note that the strategies outlined will identify famesyltransferase inhibitors which affect Ramlp. Agents which block Ram2p would likely fail to grow under all conditions. Indeed, ram2 null mutations are lethal (He B et al., Proc Natl Acad Sci 88:11373-7, 1991), perhaps due to the fact that Ram2p also functions as a component of geranylgeranyltransferase 1.

Carboxymethyltransferases

In yeast, methylation of the C-terminal amino acid of a-factor, Ras proteins, and presumably Rho/Rac proteins is catalyzed by Stel4p. Although MATa stel4 mutants are unable to mate, reflecting the requirement of carboxymethylation for the activity of a-factor, stel4 disruptions are not lethal and do not affect the rate of cell proliferation. Carboxymethylation appears to be dispensible for the function of Ras proteins and Stel 8p (the yeast homologue of the Gγ subunit). Although Ras function in yeast can apparently tolerate the absence of carboxymethyl modification, it is nonetheless possible that inhibitors of mammalian methyltransferases could alter the activity of mammalian p21ras.

It could be determined if yeast stel 4 mutations can be complemented by the homologous mammalian gene, or a modified version of it. One would use an episomal vector to express the mammalian gene encoding the methyltransferase in yeast that are genotypically stel 4. The strain would be a modified MATa strain that expresses the a- factor receptor in lieu of the normal a-factor receptor and that contains an integrated fusl-HIS3 construct so that the a-factor secreted by the cell confers autocrine growth on histidine-deficient media. Ifthe mammalian methyltransferase can function in place of Stel4p, the tester cells will be capable of mating. That is, the mammalian methyltransferase will permit synthesis of active a-factor in ste 14 mutants.

Ifthe mammalian gene encoding the methyltransferase will complement stel 4, tester strains can be constructed to test for potential inhibitors of mammalian methyltransferase. In one embodiment, tester MATa yeast strains will: 1. carry a mammalian carboxymethyltransferase gene in lieu of STE14; 2. respond to a-factor which they export by virtue of heterologous expression of Ste3p; 3. respond to autocrine a-factor such that they cannot grow on media containing galactose as in the negative GAL1 selection scheme outlined above. Exposure of such strains to compounds which inhibit the methyltransferase will confer upon these cells the ability to grow on media containing galactose. It is desirable to discriminate inhibitors of carboxy-methy ltransferase activity from compounds that either directly block the negative response to a-factor, e.g. by interfering with the interaction of the Ste4-Stel 8 complex with its effector, or block the production of a-factor by a mechanism that does not involve methyltransferase. The following control experiments will identify such false positives. Candidate inhibitors will be tested on a MATa strain that is engineered to secrete a-factor and to respond to s the secreted a-factor by failing to grow on galactose-containing media. Any agent that enables these cells to grow on media containing galactose will be not be acting as an inhibitor of carboxymethyltransferase. Candidate compounds which pass the foregoing test may be targeting the carboxy-methyltransferase, famesyltransferase, Ste6p, or otiier proteins involved in the maturation and export of a-factor. In order to discriminate the o target of the compounds, a combination of in vitro biochemical and in vivo genetic assays can be applied: both the carboxymethyltransferase and the famesyltransferase can be assayed in vitro to test the effect of the candidate agent. Furthermore, ifthe target is Stel4p its overexpression on high-copy plasmids should confer resistance to the effect of the compound in vivo. 5

Proteases

Mature yeast α-factor is a thirteen amino acid peptide that is derived from a polyprotein precursor in much the same manner as mature mammalian melanocyte_* stimulating hormone (MSH) or calcitonin are derived from precursor polyproteins. Two 0 genes in the yeast genome encode prepro-α-factor, MFαl and MFα2. MFαl encodes a precursor polypeptide containing four copies of mature α-factor embedded in a polypeptide of the following structure: hydrophobic pre-sequence / hydrophilic pro- sequence / α-factor / α-factor / α-factor / α-factor. MFα2 encodes a polyprotein precursor of a similar structure containing only two copies of mature α-factor. 5 Pre-pro-α-factor is synthesized in the cytoplasm and is then transported from the cytoplasm to the endoplasmic reticulum and then to the Golgi along the classical secretory pathway of S. cerevisiae. The signal sequence of prepro-α-factor is cleaved during transit into the ER by signal peptidase and asparagine-linked oligosaccharides are added (in the ER) and modified (in the Golgi) on the pro-segment of the precursor as it o transits the secretory pathway. Once in the Golgi, three distinct proteolytic processing events occur. First the Kex2 protease cleaves at dibasic residues (-KR-) near the amino terminus of each α-factor repeat Kex2 is homologous to the subtilisin-like endoproteases PC2 and PC1/PC3 involved in prohormone processing in mammalian cells (Smeekens and Steiner 1990; Nakayama et al. 1991). Additional mammalian s Kex2-like processing endoproteases include PACE, isolated from a human hepatoma, PC4, expressed in testicular germ cells and PC6, a candidate protease for the processing

niEln of gastrointestinal peptides (Barr et al. 1991 ; Nakayama et al. 1992; Nakagawa et al. 1993). It appears that Kex2-like proteins comprise a large family of tissue-specific endoproteases in mammalian cells.

Once Kex2 has released the immature α-factor peptides, two additional proteases act to complete processing. Kexl is a specific carboxypeptidase that removes the carboxy-terminal-KR remaining after cleavage by Kex2. Like its mammalian counterparts carboxypeptidases B and E, Kexl is highly specific for peptide substrates with carboxy-terminal basic residues. The final proteolytic processing event that occurs is the removal of the spacer dipeptides at the amino terminus of each pro-α-factor peptide. This is accomplished by the product of the STE13 gene, dipeptidyl aminopeptidase A. This enzyme is a type IV dipeptidyl aminopeptidase: it is capable of cleaving on the carboxyl side of either -x-A- or -x-P- sites in vitro.

Other type IV dipeptidyl aminopeptidases are believed to be active in the processing of a variety of pre-peptides in animal cells (Kreil 1990). In addition, functional similarity has been proved between yeast Kex 1 and Kex2 and their mammalian counter-parts in that both yeast enzymes will proteolytically cleave endogenous precursors when expressed in mammalian cells deficient in the native enzyme (Thomas et al. 1988, 1990). It appears likely, then, that mammalian homologs of the yeast proteases Kexl , Kex2 and Stel 3 p, when expressed in yeast, will function to process a synthetic α-factor pheromone precursor bearing appropriate cleavage sites.

Human proteases that may so function in yeast include PC2 and PC1/PC3 (or other Kex2 homologs), carboxypeptidases B and E (Kexl homologs) and type IV dipeptidyl aminopeptidases (Stel3p homologs).

Yeast would provide a facile assay system for the discovery of inhibitors of proteases able to process synthetic α-factor. The yeast could be engineered to express both the potential inhibitor and the exogenous protease, and, preferably, not the latter's yeast cognate.

Furthermore, this means of exploiting yeast pheromone processing to identify protease inhibitors can be expanded to encompass any protease that can be expressed to function in yeast provided an appropriate cleavage recognition site is included in a synthetic α-factor precursor. In the latter approach, novel proteolytic activities will be added to yeast; these enzymes will substitute for proteases in the α-factor maturation pathway but will not be "catalytic homologues" of Kexl, Kex2 or Stel3p. Production of mature α-factor will become dependent on the activity of the novel protease through removal of the recognition site(s) for a selected yeast enzyme from a synthetic MFα gene and insertion of the recognition sequence for the novel protease(s). ExogenousABC Transporters

The majority of proteins destined for transport to the extracellular environment proceed through a secretory pathway that includes translation initiation in the cytoplasm, 5 transport to the lumen of the endoplasmic reticulum, passage through the Golgi to secretory vesicles and subsequent exit from cells. Other proteins leave the cell by an alternative mechanism, which involves mediation by an "ABC transporter". The ABC transporters form a family of evolutionarily conserved proteins, share a similar overall structure, and function in the transport of large and small molecules across cellular io membranes (Higgins 1992). The characteristic component of members of this protein family is a highly conserved sequence that binds ATP (Higgins et al., 1986; Hyde et al. 1990); these intrinsic membrane proteins are ATPases, deriving energy from the hydrolysis of that nucleotide to effect the transport of molecules. This family includes over 50 prokaryotic and eukaryotic proteins: transporters of amino acids, sugars,

I 5 oligosaccharides, ions, heavy metals, peptides, or other proteins belong to this superfamily. Representative transmembrane transporters are included in Table I of WO94/23025. Typically, ABC transporters use the energy of ATP hydrolysis to pump substrate across a cell membrane against a concentration gradient. Some import substrate, others export it. See Higgins, Ann. Rev. Cell, Biol., 8:67-113 (1992).

2 o The prototypical structure of an ABC transporter includes four membrane- associated domains: two hydrophobic, putative membrane-spanning sequences, each predicted to traverse the membrane six times, and two nucleotide binding domains that couple ATP hydrolysis to transport In prokaryotes, the domains of an ABC transporter are often present on separate polypeptides. Various permutations of domain fusions

25 have been described: the E. coli iron hydroxamate transporter contains the two membrane-spanning domains in a single polypeptide and the ribose trans-porter of the same organism bears two nucleotide-binding domains on one molecule. The major histocompatibility complex (MHC) peptide transporter is composed of two polypeptides, Tapl and Tap2. The N-teπninus of each protein contains a hydrophobic membrane- 30 spanning domain while the C-terminus contains an ATP-binding sequence. Together Tapl and Tap2 form a functional complex. The heavy metal tolerance protein, HMT1 , expressed in the fission yeast Schizosaccharomyces pombe, consists of a polypeptide containing a single hydrophobic domain and a C-terminal ATP-binding sequence (Ortiz et al. 1992). It may be that the HMT1 transporter functions as a homodimer. The

35 Saccharomyces cerevisiae Ste6 a-factor transporter is expressed as a single polypeptide containing two membrane-spanning domains and two nucleotide-binding domains. When Ste6 is expressed as two half-molecules, the protein complex which apparently forms retains function at a level greater than 50% that of the wild type, single polypeptide (Berkower and Michaels 1991). In other eukaryotic ABC transporters, including Mdrl, CFTR and MRP, the four domains are also contained within a single 5 polypeptide. Thus, the ABC transporter may be a single multidomain polypeptide, or it may comprise two or more polypeptides, each providing one or more domains.

In general, transporters contain six transmembrane segments per each hydrophobic domain, for a total of twelve segments. The minimum number of transmembrane segments required for formation of a translocation complex appears to o be 10. Thus the histidine transporter of S. typhimurium lacks an N-terminal transmembrane segment from each of its hydrophophic domains and therefore contains five transmembrane segments per domain (Higgins et al., Nature 298, 723-727 (1982). The MalF protein of the E. coli maltose transporter contains an N-terminal extension of hydro-phobic sequence which bears two additional transmembrane segments, bringing s the total for this hydrophobic domain to 8 (Overduin et al. 1988). The N-terminal extension can be deleted, however, without loss of function of this transporter (Ehrmann et al. 1990). Although the number of segments required for formation of a functional translocator is suggested by these studies, there exists no data on the precise structure of the transmembrane segments themselves. These sequences are assumed to have an α- o helical form, but this has not been proven and the structure of the entire translocation complex within the plasma membrane remains to be elucidated.

In order to span the lipid bilayer, a minimum of 20 amino acids is required and sequences believed to form transmembrane segments have been identified using hydrophobicity scales. Hydrophobicity scales assign values to individual amino acid 5 residues indicating the degree of hydrophobicity of each molecule (Kyte and Doolittie

1982; Engleman et al. 1986). These values are based on experimental data (solubility measurements of amino acids in various solvents, analysis of side chains within soluble proteins) and theoretical considerations, and allow prediction of secondary structure in novel sequence with reasonable accuracy. Analysis using hydrophobicity measurements o indicates those stretches of a protein sequence which have the hydrophobic properties consistent with a transmembrane helix.

With a few exceptions, there is little or no significant amino acid sequence similarity between the transmembrane domains of two different transporters. This lack of sequence similarity is not inconsistent with the apparent function of these 5 hydrophobic domains. While these residues must be capable of forming the hydrophobic α-helical structures believed to transverse the plasma membrane, many amino acid residues are hydrophobic and can contribute to the formation of an α-helix.

Considerable, if as yet inexplicable, sequence similarity has been detected ill comparisons of the transmembrane domains of the yeast STE6, human MDR and E. coli HlyB hemolysin transporters [Gros et al., Cell 47, 371 (1986); McGrath and 5 Varchavsky, Nature 340, 400 ( 1989); Kuchler et al., EMBO J. 8, 3973 ( 1989)]. Otiier sequence similarities can be explained by gene duplication, as in the case of the transmembrane domains of rodent P-glyco-proteins (Endicott et al. 1991 ). The transmembrane domain of the histidine transporter of S. typhimurium bears homology to that of the octopine uptake system of Agrobacterium tumefaciens, the latter two_^ o transporters translocate chemically similar substrates 21 (Valdiva et al. 1991 ). , »,

Study of mutant transport proteins has pointed to a role for the transmembrane sequences in the recognition of substrate. Thus maltose transporters in E. coli which gain the ability to translocate p-nitrophenyl-α-maltoside bear mutations in the transmembrane domain (Reyes et al. 1986). A mutation in transmembrane segment 11 s of MDR has been shown to change the substrate specificity of that transporter (Gros et al. 1991) and mutation of charged residues in the transmembrane domain of CFTR changes its ion selectivity (Anderson et al. 1991 ). _■,..

Some aspects of the involvement of extramembrane loop sequences in transport function are being elucidated. In a number of bacterial transporters a short conserved o motif is present on the cytoplasmic loop which connects transmembrane segments 4 and

5 [Dassa and Hofπung (1985)]. It has been hypothesized that this sequence interacts with the ATP-binding domains of these transport proteins; mutation of this conserved sequence will abolish transport function (Dassa 1990). Cytoplasmic loops may also be involved in substrate recognition. Thus the sequences following transmembrane 5 segments 7 and 12 of the yeast a-factor transporter resemble sequences in the a-factor receptor, Ste3p, and may interact with the pheromone substrate (Kuchler et al. 1989). In fact, mutations in the cytoplasmic loops are known to alter the substrate specificity of a given transporter. The GI 85 V mutation of human MDR, located in the loop between transmembrane segments 2 and 3, alters the interaction of that transporter with o vinblastine and colchicine (Choi et al. 1988).

The ATP-binding domains are about 200 amino acids long, and domains Horn different transporters typically have a sequence identity of 30-50%. The conserved sequences include the "Walker motifs" which are associated with many nucleotide binding proteins. Walker, et al., EMBO J. 1:945-951 (1982). s Sequence conservation extends over the length of the ATP-binding domain, not bemg limited to the Walker motifs. Furthermore, the ATP-binding domains of a single transporter exhibit greater sequence identity to one another than to the domains from two different transporters. Not all proteins containing a conserved ATP-binding domain are involved in transport, however. The cytoplasmic enzyme UvrA functions in DNA repair and the EF-3 protein of yeast is an elongation factor. Yet both proteins contain ATP- binding cassettes identifiable by sequence comparison.

ATP-binding domains are highly hydrophilic and, in the case of transporters, appear to reside at the cytoplasmic face of the membrane, anchored there via an association with the membrane-spanning domain of these proteins. The points of interaction between the transmembrane and ATP-binding domains have not been experimentally determined. Models of the structure of the nucleotide binding domain indicate that loop sequences may extend from the core of the structure to interface with the hydrophilic sequences which transverse the membrane (Hyde et al. 1990; Mimura et al. 1991). The two structural models, one based on adenylate cyclase and the other on ras p21 structure, predict a core nucleotide binding fold composed of five β-sheets with the Walker A motif (a glycine-rich loop) positioned to interact with ATP during hydrolysis. In addition, loop structures (two loops in one model, one large loop in the other) are predicted to extend from the core to couple the ATP-binding domain to other domains of the transporter. The coupling sequences transmit most likely through conformational change, the energy of ATP hydrolysis to those portions of the molecule which are involved in transport. Ste6 function is required for mating but the protein is not necessary for yeast survival (Wilson and Herskowiz 1984; Kuchler et al. 1989; McGrath and Varshavsky 1989). Ste6 is structurally homologous to the mammalian MDRs. Furthermore, it has been demonstrated that two mammalian MDR proteins, murine Mdr3 and human Mdrl, will substitute functionally for the yeast transporter in cells deleted for STE6 (Raymond et al. 1992; Kuchler and Thomer 1992). Yeast strains deleted for STE6 serve as a starting point for the design of screens to discover compounds that modulate the function of exogenous ABC transporters.

Two different yeast screens can be used to identify modulators of ABC transporter function. In the first instance, a mammalian protein that transports a-factor will serve as a target for potential inhibitors of transporter function. Thus, a yeast strain will be engineered to express a functional transporter, e.g. mammalian MDR1 , which substitutes for the yeast Ste6 protein in the transport of a-factor. Furthermore, this strain will be engineered to respond in autocrine fashion to a-factor: e.g., so that the cells will be unable to grow on media containing galactose. This negative selection will depend on the expression of the GAL 1 gene under the control of a pheromone-responsive promoter in a strain background which includes mutated versions of the GAL7 or W

- 71 -

GAL 10 genes. Expression of GAL 1 in the presence of galactose in such a strain background is toxic to cells. In the absence of a-factor transport, signaling down the pheromone response pathway would cease as would the consequent expression of the toxic gene. Cell growth in the presence of a test compound, or upon expression oft s specific random peptide, would signal inhibition of transport function and the identification of a potential therapeutic.

In addition to inhibitors of MDR, compounds may be identified which interfere with the interaction of a-factor with the a-factor receptor. Such compounds can be discriminated by their inhibition of a-factor-induced growth arrest in a wild type Matα o strain. Compounds may also impact at other points along the pheromone response pathway to inhibit signaling and these compounds will prevent signal transduction in a wild type Matα strain.

In a second screen, a mutant heterologous transporter (e.g., mutant CFTR) that is initially incapable of transporting a-factor or an a-factor-like peptide can be expressed in s autocrine yeast deleted for endogenous Ste6. The cells will be capable of an autocrine response to the a-factor which those cells produce. Thus a pheromone-responsive promoter will control expression of a gene that confers an ability to grow in selective media. Such cells will permit identification of compounds which correct defects in the transporter and permit it to function in the export of pheromone analogues to the 0 extracellular space. In this way, therapeutic peptides or other classes of chemical - compounds could be identified which stabilize a mutant protein and allow noπnai processing, transport, localization to the plasma membrane and function. This strategy, if successful, may eliminate the need to "replace" some mutant genes with normal sequence, as envisioned in gene therapies, by recovering the function of mutant proteins 5 through the correction of processing and/or localization defects.

In addition to "activators" of the mutant transporter, compounds may also be identified which are capable of initiating signalling from the a-factor receptor in the absence of transport by the endogenously expressed pheromone. These compounds will be distinguished by their ability to cause growth arrest in a wild type Matα strain. <L o Compounds may also impact at other points along the pheromone pathway and can be discerned via an ability to initiate signalling in a wild type Matα strain in the absence of a-factor.

In a preferred embodiment the exogenous protein produced by the yeast cells is one of the exogenous ABC transporters listed in Table 1 of WO94/23025. 5

Gene Expression O 97/40170 PC17US97/06711

- 72 -

Vectors

Another aspect of the invention pertains to vectors, preferably recombinant expression vectors, containing a nucleic acid encoding adenylyl cyclase. The expression vectors of the invention comprise a nucleic acid of the invention in a form suitable for expression of the nucleic acid in a host cell, which means that the recombinant expression vectors include one or more regulatory sequences, selected on the basis of the host cells to be used for expression, which is operatively linked to the nucleic acid sequence to be expressed. Within a recombinant expression vector, "operably linked" is intended to mean that the nucleotide sequence of interest is linked to the regulatory sequence(s) in a manner which allows for expression of the nucleotide sequence (e.g., in an in vitro transcription/translation system or in a host cell when the vector is introduced into the host cell). The term "regulatory sequence" is intended to includes promoters, enhancers and other expression control elements (e.g., polyadenylation signals). Such regulatory sequences are described, for example, in Goeddel; Gene Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, CA (1990).

Regulatory sequences include those which direct constitutive expression of a nucleotide sequence in many types of host cell and those which direct expression of the nucleotide sequence only in certain host cells (e.g., tissue-specific regulatory sequences). It will be appreciated by those skilled in the art that the design of the expression vector may depend on such factors as the choice of the host cell to be transformed, the level of expression of protein desired, etc. The expression vectors of the invention can be introduced into host cells to thereby produce adenylyl cyclase as proteins or peptides, including fusion proteins or peptides.

The recombinant expression vectors of the invention can be designed for expression of adenylyl cyclase protein in prokaryotic or eukaryotic cells. For example, adenylyl cyclase can be expressed in bacterial cells such as E. coli, insect cells (using baculovirus expression vectors) yeast cells or mammalian cells. Suitable host cells are discussed further in Goeddel, Gene Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, CA (1990). Alternatively, the recombinant expression vector may be transcribed and translated in vitro, for example using T7 promoter regulatory sequences and T7 polymerase.

Expression of proteins in prokaryotes is most often carried out in E. coli with vectors containing constitutive or inducible promotors directing the expression of either fusion or non-fusion proteins. Fusion vectors add a number of amino acids to a protein encoded therein, usually to the amino terminus of the recombinant protein. Such fusion vectors can serve one or more purposes: 1 ) to increase expression of recombinant protein; 2) to increase the solubility of the recombinant protein; 3) to aid in the _ ; purification of the recombinant protein by acting as a ligand in affinity purification; 4) to provide an epitope tag to aid in detection and/or purification of the protein; and/or 5) to provide a marker to aid in detection of the protein (e.g., a color marker using b- 5 galactosidase fusions). Often, in fusion expression vectors, a proteolytic cleavage site is introduced at the junction of the fusion moiety and the recombinant protein to enable separation of the recombinant protein from the fusion moiety subsequent to purification of the fusion protein. Such enzymes, and their cognate recognition sequences, include Factor Xa, thrombin and enterokinase. Typical fusion expression vectors include pGEX o (Pharmacia Biotech Inc.; Smith, D.B. and Johnson, K.S. (1988) Gene 67:31-40), pMAL (New England Biolabs, Beverly, MA) and pRIT5 (Pharmacia, Piscataway, NJ) which fuse glutathione S-transferase (GST), maltose E binding protein, or protein A, respectively, to the target recombinant protein. Recombinant proteins also can be expressed in eukaryotic cells as fusion proteins for the same purposes discussed above. s Examples of suitable inducible non-fusion E. coli expression vectors include pTrc (Amann et al.. (1988) Gene £9:301-315) and pET 1 Id (Studier etal.. Gem Expression Technology: Methods in Enzymology 185. Academic Press, San Diego, California (1990) 60-89). Target gene expression from the pTrc vector relies on host RNA polymerase transcription from a hybrid trp-lac fusion promoter. Target gene o expression from the pET 11 d vector relies on transcription from a T7 gnl 0-lac fusion promoter mediated by a coexpressed viral RNA polymerase (T7 gnl). This viral polymerase is supplied by host strains BL21(DE3) or HMS 174(DE3) from a resident λ prophage harboring a T7 gnl gene under the transcriptional control of the lacUV 5 promoter. 5 One strategy to maximize recombinant protein expression in E. coli is to express the protein in a host bacteria with an impaired capacity to proteolytically cleave the recombinant protein (Gottesman, S., Gene Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, California (1990) 119-128). Another strategy is to alter the nucleic acid sequence of the nucleic acid to be inserted into an o expression vector so that the individual codons for each amino acid are those preferentially utilized in E. coli (Wada et al., (1992) Nuc. Acids Res. 20:2111-2118)- Such alteration of nucleic acid sequences of the invention can be carried out by standard DNA synthesis techniques.

In another embodiment the adenylyl cyclase expression vector is a yeast s expression vector. Examples of vectors for expression in yeast S. cerivisae include pYepSecl (Baldari. et al., (1987) EMBOJ. 6:229-234), pMFa (Kurjan and Herskowitz, (1982) Cell 30:933-943), _PJRY88 (Schultz et al., (1987) Gene 54:1 13-123), and pYES2 (Invitrogen Coφoration, San Diego, CA). The vector may be a DNA which is integrated into the host genome, and thereafter is replicated as a part of the chromosomal DNA, or it may be DNA which replicates autonomously, as in the case of a plasmid. In the latter case, the vector must include an origin of autononomous replication that is functional in the host. There are two types of yeast origins of replication in common use: those derived from the yeast 2 micron circle which permit replication of plasmids to 40-50 copies per yeast cell; and those derived from genomic CEN ARS sequences, which are maintained at lower copy number, typically only one or two plasmids per yeast cell. In the case of an integrating vector, the vector may include sequences that facilitate integration, e.g., sequences homologous to host sequences, or encoding integrases.

Besides being capable of replication in yeast cells, it is convenient ifthe vector can also be replicated in bacterial cells, as many genetic manipulations are more conveniently carried out therein. Shuttle vectors capable of replication in both yeast and bacterial cells include YEps, Yips, and the pRS series.

The expression of a peptide-encoding gene in a yeast cell requires a promoter which is functional in yeast. Suitable promoters include the promoters for metallothionein, 3-phosphoglycerate kinase (Hitzeman et al., J. Biol. Chem. 255, 2073 (1980) or other glycolytic enzymes (Hess et al., J. Adv. Enzyme Reg. 7, 149 (1968); and Holland et al. Biochemistry 17, 4900 (1978)), such as enolase, glyceraldehyde-3- phosphate dehydrogenase, hexokinase, pyruvate decarboxylase, phosphofructokinase, gIucose-6-phosphate isomerase, 3-phosphoglycerate mutase, pyruvate kinase, triosephosphate isomerase, phosphoglucose isomerase, and glucokinase. Suitable vectors and promoters for use in yeast expression are further described in R. Hitzeman et al., EPO Publn. No. 73,657. Other promoters, which have the additional advantage of transcription controlled by growth conditions, are the promoter regions for alcohol dehydrogenase 2, isocytochrome C, CUP1 (inducible by copper), acid phosphatase, degradative enzymes associated with nitrogen metabolism, and the aforementioned metallothionein and glyceraldehyde-3-phosphate dehydrogenase, as well as enzymes responsible for maltose and galactose utilization. Finally, promoters that are active in only one of the two haploid mating types may be appropriate in certain circumstances. Among these haploid-specific promoters, the pheromone promoters MFal and MFαl and the GPA1 promoter are of particular interest. In constructing suitable expression plasmids, the termination sequences associated with these genes, or with other genes that are efficiently expressed in yeast, may also be ligated into the expression vector 3 ' of the heterologous coding sequences to provide polyadenylation and termination of the mRNA.

Alternatively, adenylyl cyclase can be expressed in insect cells using baculovirus expression vectors. Baculovirus vectors available for expression of proteins in cultured 5 insect cells (e.g., Sf 9 cells) include the pAc series (Smith et al, (1983) Mol Cell Biol. 3:2156-2165) and the pVL series (Lucklow, V.A., and Summers, M.D., (1989) Vlmlogy 170:31-39).

In yet another embodiment, a nucleic acid of the invention is expressed jq,-. mammalian cells using a mammalian expression vector. Examples of mammalian l o expression vectors include pCDM8 (Seed, B., ( 1987) Nature 329:840) and pMT2PC (Kaufman et al (1987), EMBOJ. 6:187-195). When used in mammalian cells, the expression vector's control functions are often provided by viral regulatory elements. For example, commonly used promoters are derived from polyoma, Adenovirus 2, cytomegalovirus and Simian Virus 40. is In another embodiment the recombinant mammalian expression vector is " capable of directing expression of the nucleic acid preferentially in a particular cell type (e.g., tissue-specific regulatory elements are used to express the nucleic acid). Tissue- specific regulatory elements are known in the art. Non-limiting examples of suitable tissue-specific promoters include lymphoid-specific promoters (Calame and Eaton 0 (1988) Adv. Immunol. 42:235-275), in particular promoters of T cell receptors (Winoto and Baltimore (1989) EMBOJ. 5:729-733) and immunoglobulins (Banerji et al. (1983) Cell 22:729-740; Queen and Baltimore (1983) Cell 23:741-748), the albumin promoter (liver-specific; Pinkert etal. (1987) Genes Dev. 1:268-277), neuron-specific promoters (e.g., the neurolϊlament promoter, Byrne and Ruddle (1989) Proc. Natl Acad Sci VSA 5 86:5473-5477), pancreas-specific promoters (Edlund et al. ( 1985) Science 222:9! 2_*916), and mammary gland-specific promoters (e.g., milk whey promoter, U.S. Patent No. 4,873^16 and European Application Publication No. 264,166). Developmentally- regulated promoters are also encompassed, for example the murine hox promoters (Kessel and Gruss (1990) Science 242:374-379) and the a-fetoprotein promoter (Campes 0 and Tilghman ( 1989) Gems Dev. 2:537-546).

Moreover, inducible regulatory systems for use in mammalian cells are known in the art, for example systems in which gene expression is regulated by heavy metal ions (see e.g.. Mayo et al. (1982) Cell 29:99-108; Brinster et al. (1982) Nature 296:39_*42; Searle e/fl/. (1985) Mol Cell. Biol. 5:1480-1489), heat shock (see e.&, Nouer ettd_, 5 (1991) in Heat Shock Response, e.d. Nouer, L. , CRC, Boca Raton , FL, ppl67-22§)_» hormones (see e.g., Lee et al. (1981) Nature 294:228-232; Hynes β/α/. (198I) Λw. Natl. Acad. Sci. USA 78:2038-2042; Klock et al. (1987) Nature 329:734-736; Israel & Kaufman (1989) Nucl. Acids Res. 17:2589-2604; and PCT Publication No. WO 93/23431), FK506-related molecules (see e.g., PCT Publication No. WO 94/18317) or tetracyclines (Gossen, M. and Bujard, H. (1992) Proc. Natl. Acad. Sci. USA 89:5547- 5551 ; Gossen, M. et al. (1995) Science 261:1766-1769; PCT Publication No. WO 94/29442; and PCT Publication No. WO 96/01313). Accordingly, in another embodiment, the invention provides a recombinant expression vector in which DNA encoding adenylyl cyclase is operatively linked to an inducible eukaryotic promoter, thereby allowing for inducible expression of cyclase in eukaryotic cells.

Host Cells

Another aspect of the invention pertains to recombinant host cells into which a vector, preferably a recombinant expression vector, of the invention has been introduced. A host cell may be any prokaryotic or eukaryotic cell. For example, adenylyl cyclase protein may be expressed in bacterial cells such as E. coli, insect cells, yeast or mammalian cells (such as Jurkat T cells, Chinese hamster ovary cells (CHO) or COS cells). Other suitable host cells are known to those skilled in the art. Vector DNA can be introduced into prokaryotic or eukaryotic cells via conventional transformation or transfection techniques. As used herein, the terms "transformation" and "transfection" are intended to refer to a variety of art-recognized techniques for introducing foreign nucleic acid (e.g., DNA) into a host cell, including calcium phosphate or calcium chloride co-precipitation, DEAE-dextran-mediated transfection, lipofection, or electroporation. Suitable methods for transforming or transfecting host cells can be found in Sambrook et al. (Molecular Cloning: A Laboratory Manual, 2nd Edition, Cold Spring Harbor Laboratory press ( 1989)), and other laboratory manuals.

In a preferred embodiment the host cell of the present invention is a mammalian cell transfected with a vector appropriate for expression of heterologous proteins in mammalian cells. For stable transfection of mammalian cells, it is known that, depending upon the expression vector and transfection technique used, only a small fraction of cells may integrate the foreign DNA into their genome. In order to identify and select these integrants, a gene that encodes a selectable marker (e.g. , resistance to antibiotics) is generally introduced into the host cells along with the gene of interest. Preferred selectable markers include those which confer resistance to drugs, such as G418, hygromycin and methotrexate. Nucleic acid encoding a selectable marker may be introduced into a host cell on the same vector as that encoding adenylyl cyclase or may be introduced on a separate vector. Cells stably transfected with the introduced nucleic acid can be identified by drug selection (e.g., cells that have incorporated the selectable marker gene will survive, while the other ceils die).

A host cell of the invention, such as a prokaryotic or eukaryotic host cell in culture, can be used to produce (i.e., express) adenylyl cyclase protein. In particularly preferred embodiments the host cell is a yeast cell. For example, the yeast may be of any species that require cyclic AMP for growth and which are cultivatable. Suitable species include Kluyveromyces lactis, Saccharomyces kluyveri, Schizosaccharomyces pombe, and Ustilago maydis; Saccharomyces cerevisiae is preferred. The term "yeast", as used herein, includes not only yeast in a strictly -.? taxonomic sense (i.e., unicellular organisms), but also yeast-like multicellular fungi.

The host cell may be a diploid a/α or a haploid cell. Preferably, to eliminate, in the first instance, any possible effects of the yeast G proteins on the mammalian cyclase, a diploid strain is used. Diploid yeast cells, in contrast to haploid cells, do not express GPA1, which encodes the yeast homolog of Gα, nor do they express STE4 or STE18, which encode yeast Gβ and Gγ, respectively.

In addition, crosses are preferably made to derive a diploid strain that bears a mutant allele of yeast adenylyl cyclase, cdc35- 1 , as the endogenous cyclase. _« .. _t

The yeast cell is preferably of a cAMP dependent strain such as the caml, cam2, cam3 strain. -

Structural Sequences

The structural gene encoding the adenylyl cyclase may be the wild-type mammalian gene, or a modified gene. "Silent" modifications may be made to improve expression, by, e.g., (1) eliminating secondary structures in the corresponding mRNA, or (2) substituting codons preferred by the host cell for codons that are not so preferred, or to facilitate cloning, e.g., by introducing, deleting or modifying restriction sites. The gene may also be modified so that a mutant adenylyl cyclase is encoded.

The sequences of nucleic acids which encode the isotypes of adenylyl cyclase are known in the art, and exemplary citations are provided supra. A nucleic acid molecule encoding adenylyl cyclase can be isolated using standard molecular biology techniques and the sequence information known in the art. For example, aadenylyl cyclase cDNA can be isolated from a cDNA library (e.g., a human cDNA library, prepared from human cells (commercially available from Stratagene) using all or portion of a known sequence as a hybridization probe and standard hybridization techniques (e.g., as described in Sambrook, J., et al. Molecular Cloning: A Laboratory Manual 2nd, ed, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1989). Moreover, a nucleic acid molecule encompassing all or a portion of a known adenylyl cyclase sequence can be isolated by the polymerase chain reaction using oligonucleotide primers designed based upon the sequence of a known sequence. For example, mRNA can be isolated from cells (e.g., human cells, by the guanidinium-thiocyanate extraction procedure of Chirgwin et al. (1979) Biochemistry 18: 5294-5299) and cDNA can be prepared using reverse transcriptase (e.g., Moloney MLV reverse transcriptase, available from Gibco/BRL, Bethesda, MD; or AMV reverse transcriptase, available from Seikagaku America, Inc., St. Petersburg, FL). Synthetic oligonucleotide primers for PCR amplification can be designed based upon a nucleotide sequence known in the art. A nucleic acid of the invention can be amplified using cDNA or, alternatively, genomic DNA, as a template and appropriate oligonucleotide primers according to standard PCR amplification techniques. The nucleic acid so amplified can be cloned into an appropriate vector and characterized by DNA sequence analysis. Furthermore, oligonucleotides corresponding to an adenylyl cyclase nucleotide sequence can be prepared by standard synthetic techniques, e.g., using an automated DNA synthesizer.

In addition to known adenylyl cyclase seqences, it will be appreciated by those skilled in the art that DNA sequence polymorphisms that lead to changes in the amino acid sequences of adenylyl cyclase may exist within a population (e.g., the human population). Such allelic varients can also be used in the practice of the instant invention. Nucleic acid molecules corresponding to natural allelic variants and homologues of adenylyl cyclase cDNAs can be isolated based on their homology to known cyclase nucleic acid molecule disclosed herein using a known sequences, or a portion thereof, as a hybridization probe according to standard hybridization techniques under stringent hybridization conditions. The skilled artisan will further appreciate that changes may be introduced by mutation into the nucleotide sequence of known adenylyl cyclase genes, thereby leading to changes in cyclase amino acid sequence. In certain embodiments, this may not alter the functional activity of the cyclase protein. For example, nucleotide substitutions leading to amino acid substitutions at "non-essential" amino acid residues, e.g., not in the cytoplasmic domains known to be important in protein function. Such changes may not alter the functional activity of cyclase, such as its ability to be regulated by a G protein subunit(s), whereas an "essential" amino acid residue is required for functional activity. Accordingly, another aspect of the invention pertains to nucleic acid molecules encoding adenylyl proteins that contain changes in amino acid residues that alter adenylyl cyclase activity. Exemplary mutants are described in more detail in the appended Examples. Given the teachings of the instant specification, one of ordinary skill in the art could select other amino acid residues which are amenable to mutation. For example, sequences of different cyclases, or sequences from the same type of cyclase from _k different sources could be aligned for optimal comparison purposes (e.g., gaps may be 5 introduced in the sequence of one protein for optimal alignment with the other protein). The amino acid residues at corresponding amino acid positions are then compared. When a position in one sequence is occupied by the same or a similar amino acid residue as the corresponding position in the other sequence, then the molecules are homologous at that position (i.e., as used herein amino acid "homology" is equivalent to amino acid o identity or similarity). As used herein, an amino acid residue is "similar" to another amino acid residue ifthe two amino acid residues are members of the same family of residues having similar side chains. Families of amino acid residues having similar side chains have been defined in the art, including basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side s chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine) and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine). The percent homology between two sequences, therefore, is a function of the number of o identical or similar positions shared by two sequences (/^'. e. , % homology = # of identical or similar positions/total # of positions x 100).

An isolated nucleic acid molecule encoding a mutant adenylyl cyclase can be created by introducing one or more nucleotide substitutions, additions or deletions into the nucleotide sequence of a known sequence such that one or more amino acid 5 substitutions, additions or deletions are introduced into the encoded protein. Mutations can be introduced by standard techniques, such as site-directed mutagenesis and PCR- mediated mutagenesis. Conservative amino acid substitutions or non-conservative substitutions can be made at one or more amino acid residues. A "conservative amino acid substitution" is one in which the amino acid residue is replaced with an amino acid o residue having a similar side chain. Families of amino acid residues having similar side chains are defined above. Alternatively, in another embodiment, mutations can be introduced randomly along all or part of an adenyl cyclase gene, such as by saturation mutagenesis, or by the expression in bacterial cells lacking in DNA repair enzymes, as described in the examples. The resultant mutants can be screened, for example, using 5 selective media, or in a complementation assay as described herein.

In addition, codon usage may also be modifed to reflect that used in the host cell chosen for expression. For example, analysis of yeast codon usage indicates that there exists a preferred codon set consisting of the most abundant isoaccepting tRNAs present in yeast and that this preferred set (25 out of the 61 possible coding triplets) is the same for all yeast proteins (Bennetzen and Hall ( 1981 ) J. Biol. Chem. 257, 3026-3031 ). The 5 rapid translation rate required for abundant proteins is believed to provide the selective pressure for the existence of the preferred set of codons. As the extent of biased codon usage in specific genes correlates directly with the level of gene expression (Hoekma et al. (1987) Mol. Cell. Biol. 7, 2914-2924), experimental strategies aimed at the expression of heterologous genes in yeast exploit the codon bias that has been described l o for that organism (Sharp et al. ( 1986) Nuc. Acids Res. 14, 5125-5143).

In engineering the sequences with which to express mammalian adenylyl cyclase in Saccharomyces cerevisiae, a chimeric coding sequence was constructed. The initial 27 codons of the rat type 2 adenylyl cyclase are contributed by an oligonucleotide that was inserted into the expression vector, while the remainder of the coding sequence, is beginning with codon 28, was derived from the cDNA clone obtained from rat brain. By means of the oligonucleotide, codon usage at the N-terminus of the enzyme was altered to optimize translation of the sequence in yeast.

Pharmaceutical Preparations of Identified Agents

20 After identifying certain test compounds as potential modulators of adenylyl cyclase activity, the practitioner of the subject assay will continue to test the efficacy and specificity of the selected compounds both in vitro and in vivo. Whether for subsequent in vivo testing, or for administration to an animal as an approved drug, agents identified in the subject assay can be formulated in pharmaceutical preparations for in vivo

25 administration to an animal, preferably a human.

The subject compounds selected in the subject or a pharmaceutically acceptable salt thereof, may accordingly be formulated for administration with a biologically acceptable medium, such as water, buffered saline, polyol (for example, glycerol, propylene glycol, liquid polyethylene glycol and the like) or suitable mixtures thereof.

3 o The optimum concentration of the active ingredient(s) in the chosen medium can be determined empirically, according to procedures well known to medicinal chemists. As used herein, "biologically acceptable medium" includes any and all solvents, dispersion media, and the like which may be appropriate for the desired route of administration of the pharmaceutical preparation. The use of such media for pharmaceutically active ₃₅ substances is known in the art. Except insofar as any conventional media or agent is incompatible with the activity of the compound, its use in the pharmaceutical preparation of the invention is contemplated. Suitable vehicles and their formulation inclusive of other proteins are described, for example, in the book Remington's Pharmaceutical Sciences (Remington's Pharmaceutical Sciences. Mack Publishing 5 Company, Easton, Pa., USA 1985). These vehicles include injectable "deposit formulations". Based on the above, such pharmaceutical formulations include, although not exclusively, solutions or freeze-dried powders of the compound in association with one or more pharmaceutically acceptable vehicles or diluents, and contained in buffered media at a suitable pH and isosmotic with physiological fluids. In preferred l o embodiment the compound can be disposed in a sterile preparation for topical and/or systemic administration. In the case of freeze-dried preparations, supporting excipients such as, but not exclusively, mannitol or glycine may be used and appropriate buffered solutions of the desired volume will be provided so as to obtain adequate isotonic buffered solutions of the desired pH. Similar solutions may also be used for the is pharmaceutical compositions of compounds in isotonic solutions of the desired volume and include, but not exclusively, the use of buffered saline solutions with phosphate or citrate at suitable concentrations so as to obtain at all times isotonic pharmaceutical preparations of the desired pH, (for example, neutral pH).

20

Exemplification Example 1.

Yeast require the catalytic activity of the enzyme adenylyl cyclase to grow.

25 There exists a mutant form of the Saccharomyces cerevisiae adenylyl cyclase encoded by the mutant allele cdc35-l. The cdc35-l mutation describes a temperature sensitive cell division cycle mutant (Casperson et al. (1985) Proc. Natl. Acad. Sci. USA 82, 5060- 5063). This mutation results in a phenotype resembling that of temperature-sensitive cyrl-1 mutants which bear an alteration in the yeast adenylate cyclase. Haploids bearing

30 cdc35-l growth arrest at GI, as unbudded cells, at temperatures above 30°C. Diploids homozygous for cdc35-l sporulate in rich medium, in nutrient conditions that prevent sporulation of wild-type yeast cells (Shilo et al. (1978) Exp. Cell Res. 112, 241-248). The cdc35-l mutation can be complemented by sequence encoding yeast CYR1 or tiie cells can be rescued from growth arrest by the addition of exogenous cAMP, provided 5 the cells also exhibit the cam phenotype. cdc35- 1 maps to the same locus as does cyrl-1 (Boutelet and Hilger 1980). Extracts of cdc35-l cells that have been growth arrested at the restrictive temperature exhibit in vitro adenylyl cyclase activity approximately equal to that of wild type cells incubated at room temperature. These data suggest that the cdc35-l mutation affects a portion of the adenylyl cyclase molecule outside of the catalytic site, perhaps in sequence involved in the interaction between the cyclase and co-factors or regulatory molecules. A diploid strain (CY1106) bearing this mutant allele (genotype: MATa/MATα cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 Ieu2-3,112/leu2 trpl/trpl his7/+) was transformed with plasmids containing sequences encoding rat type 2 adenylyl cyclase and rat Gas. The designation cam/(cam?) indicates that this strain was derived from a cross in which the cam phenotype of one parent was uncharacterized. Expression of the rat adenylyl cyclase was constitutive and that of Gas was inducible by copper. This strain grew at 34°C in the presence, but not in the absence, of copper. Similar strains, in which only the rat adenylyl cyclase-encoding plasmid or only the rat Gαs-encoding plasmid were present and expressed, failed to grow at 34°C. This experiment demonstrated that, in the presence of mammalian Gas, type 2 rat adenylyl cyclase enables the growth at 34°C of yeast harboring the mutant adenylyl cyclase encoded by cdc35-l.

This experiment is described in more detail below. 1. Construction of adenylyl cyclase expression plasmid ln order to obtain a plasmid expressing rat type II adenylyl cyclase, a yeast expression plasmid that is based on YEp51 (Broach et al. (1983) in "Experimental Manipulation of Gene Expression" (M. Inouye, ed.) pp. 83-1 17, Academic Press, NY, 1983) can be used. Salient features of this vector (Cadus 1284) are the following: first, it contains the replication determinant of the yeast 2μ circle plasmid; this permits the plasmid to replicate to high copy number in yeast (typically 10 to 40 copies per cell). It also contains a yeast gene which permits selection for the presence of the plasmid in yeast that lack a functional genomic copy of the same gene; specifically, in the absence of the amino acid leucine, leu2 yeast which carry the vector will grow while those cells lacking the vector will not. Finally, in lieu of the GAL 10 promoter sequences present in the parental Yep51 , Cadus 1284 contains the promoter sequences of the yeast phosphoglycerol kinase (PGK) gene. Ncol and BamHI restriction enzyme sites that permit the insertion of genes to be expressed from the plasmid are present downstream of this constitutively active promoter.

An oligonucleotide of approximately 100 base pairs was inserted into the Ncol- and BamHI -restricted vector, Cadus 1284. This oligonucleotide encodes the first 27 amino acids of rat cyclase with codon usage which favors translation in yeast. The oligonucleotide was constructed using the following single stranded oligonucleotides:

oligo 066:

5'GCCGTCTCACATGAGAAGAAGAAGATACTTGAGAGATAGAGCTGAAGCTG 5 CTGCA 3^* (SEQ ID NO: 1 ) oligo 069:

5^,GCAGCTTCAGCTCTATCTCTCAAGTATCTTCTTCTTCTCATGTGAGACGGC 3^, (SEQ ID NO: 2) oligo 070: 5OCTGCTGCTGCTGGTCK^}TGGTGAAGGTTTGCAAAGATCCCGβ 3^* o (SEQ ID NO: 3) oligo 071:

5'GATCCCGGGATCTTTGCAAACCTTCACCACCACCAGCAGCAGCAGCTC3€A 3' (SEQ ID NO: 4)

5 Oligo 066 was annealed to phosphorylated oligo 069; oligo 071 was annealed to phosphorylated oligo 070. The two double-stranded oligonucleotides were mixed, ligated, digested with Espl, and the resulting approximately 92 base pair oligonucleotide was gel purified and ligated to Ncol- and BamHI-digested Cadus 1284. The resulting modified vector, Cadus 1464, contains a unique Xmal site that overlaps the BamHI site o and is contributed by the oligonucleotide.

The gene encoding type 2 adenylyl cyclase from rat brain was obtained from Randall R. Reed (Johns Hopkins School of Medicine) as a 6.4 kilobase plasmid clone that contains Xmal sites 80 bases downstream of the start codon and approximately 220 base pairs downstream of the cyclase stop codon (Feinstein et al. (1991) Proc. Natl. 5 Acad. Sci. USA 88, 10173-10177). These restriction enzyme sites were used to isolate a 3.4 kilobase fragment that contains nearly the entire coding region of the cyclase, beginning at codon 28 and continuing through the stop codon. When this fragmem is ligated into the Xmal site of the modified vector described above, a chimeric cyclase gene is created, wherein the initial 27 codons are contributed by the oligonucleotide o present in tiie vector and the remaining codons are contributed by the authentic rat adenylyl cyclase type 2 gene sequence. The consequences of this strategy for tiie construction of the cyclase expression plasmid include the following: 1. The N-terminal 27 amino acids encoded by the chimeric cyclase gene are identical to those encoded by the native rat gene, but the triplets encoding these amino acids are those that are 5 efficiently translated in yeast; 2. The chimeric cyclase gene is expressed under the control of the PGK promoter which has high, constitutive activity; 3. The cyclase gene will be in high copy number in yeast cells.

2. Construction of Gas expression plasmid

The plasmid used to express Gas contains a full-length rat Gas cDNA under the control of the copper-inducible yeast promoter, CUP1 (plasmid described in Kang et al. Mol. Cell. Biol. 10:2582-2590, 1990). Expression of Gas was induced by plating yeast bearing this plasmid on solid media containing lOOμM copper sulfate.

3. Derivation of yeast strain

The strain used to test the activity of mammalian adenylyl cyclase in yeast was a diploid strain generated from a haploid strain of genotype MATa cdc35-l cam leu2 trpl ura3 his7 (Y1777). Y1777, bearing the mutant cdc35-l allele, was obtained from the laboratory of J.R. Broach at Princeton University. Y1777 was crossed with CY5 (genotype MATα ura3 lys2 ade2 his3 leu2), the resulting diploids were sporulated, and haploid progeny of genotype MATα cdc35-l (cam?) ura3 trpl leu2 were selected for mating with Y1777 to generate the diploid strain CY1 106 (genotype cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 leu2-3, 1 12/leu/2 trpl/trpl his7/+). CY1106, like the haploids from which it was generated, is capable of growth at room temperature but does not grow at 30°C or at higher temperatures due to the temperature sensitivity of the endogenous yeast adenylyl cyclase encoded by the cdc35-l mutant allele.

The diploid strain CY1106 was transformed with the following pairs of plasmids: (1) PGK promoter-driven rat adenylyl cyclase expression plasmid (Cadus 1470) and CUP1 promoter-driven rat Gas expression plasmid (Cadus 1284) and CUP1 promoter- driven rat Gas expression plasmid (Cadus 1046) to yield strain CY1251; (2) PGK promoter-driven expression plasmid lacking adenylyl cyclase sequence (Cadus 1046) to yield strain CY 1248; (3) PGK promoter-driven adenylyl cyclase expression plasmid (Cadus 1470) and CUP 1 -driven expression plasmid lacking Gas (Cadus 1136) yielding CY 1249; (4) PGK promoter-driven expression plasmid lacking adenylyl cyclase sequence (Cadus 1284) and CUP 1 -driven expression plasmid lacking Gas sequence (Cadus 1136) yielding CY 1246. The four types of double transformants, CY 1251, CY 1248, CY 1249 and CY 1246, (genotype MATa/MATα cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 Ieu2-3,1 12/leu2 trpl/tφl his7/+), each carrying a high-copy LEU2- marked plasmid containing a PGK promoter and a high-copy TRP 1 -marked plasmid containing a CUP1 promoter, were plated onto synthetic solid media which lacked leucine and tryptophan ± lOOμM CuSO4. Plates were incubated for three days at RT, 30 °C, 34°C, and at 37°C. At RT and at 30°C, the four strains grew at approximately the same robust rate both in the presence and in the absence of copper, reflecting the activity of the endogenous yeast adenylyl cyclase encoded by the cdc35-l allele.

At 34°C in the absence of copper, all four strains failed to grow. But at 34°C, in the presence of copper sulfate, CY1251 (containing the constitutively expressed mammalian adenylyl cyclase and the copper-inducible mammalian Gas) grew rapidly 5 while other strains failed to grow. Thus, at 34°C the mammalian adenylyl cyclast could compensate for the inactive, mutant yeast adenylyl cyclase, provided that mammalian G as was also expressed.

At 37°C all four strains showed little growth in the presence of copper, likely due to the poor viability of yeast at this temperature. o 4. Expression of mammalian adenylyl cyclase in haploid yeast cells

Our initial result of functional expression of a mammalian adenylyl cyclase in yeast was obtained using diploid cells (genotype MATa/MATα cdc35-l/cdc35-I cam/(cam?) ura3-52/ura3-52 Ieu2-3,112/leu2 txpl/trpl his7/+). To determine if similar results would be obtained in haploid yeast, haploid cells of genotype MATa s cdc35-l cam ura3-52 leu2-3,112 trpl his7 and MATα cdc35-l cam uιa3-52 leu2- 3,112 trpl his7 were transformed with the identical plasmids used to transform diploid cells: one containing a copper-inducible rat Gas gene and a second containing a constitutively expressed rat type 2 adenylyl cyclase. Provided that Gas was expressed, the rat adenylyl cyclase was able to rescue growth of each haploid strain at 34^βC, Minor o differences in results obtained with diploid cells and haploid cells were observed, however. Gαs-stimulated growth at 34°C was slightly greater and growth in the absence of Gas at 34°C was slightly lower (i.e. there was lower background growth of cells) in diploids expressing the rat adenylyl cyclase.

5 5. Expression of mammalian adenylyl cyclase from low copy plasmids

Initial results were obtained with the expression of rat type 2 adenylyl cyclase from a high copy plasmid that is typically present at 10 to 40 copies per cell. In older to determine ifthe same phenotype would be observed ifthe gene encoding rat adenylyl o cyclase was present at only one or two copies per yeast cell, a low copy plasmid was utilized to express the rat cyclase. Yeast harboring the cdc35-l allele were transfoimed with a low copy plasmid encoding the type 2 adenylyl cyclase. Specifically, sequence encoding rat type 2 adenylyl cyclase under the control of the constitutive PGK promoter was moved from a LEU2-containing high copy-plasmid to an equivalent low-copy 5 plasmid. This low-copy plasmid and appropriate negative controls were transformed into both haploid and homozygous diploid cdc35-l yeast containing the rat Gas gene under the control of the copper-inducible CUP1 promoter. When both Gas and rat adenylyl cyclase are expressed in these cells, the cells acquire an ability to grow at 34°C. These results demonstrate that cyclase activity sufficient for the growth of cells is obtained when the adenylyl cyclase gene is present on plasmids that replicate at one to two copies per cell. This result was obtained in both haploid and diploid yeast.

6. Expression of rat adenylyl cyclase from an integrated sequence

The results obtained using low copy plasmids suggest that rat adenylyl cyclase will likely rescue growth at 34°C in cdc35-l yeast in which the cyclase gene is integrated into the yeast genome. It is preferable to have the cyclase expressed from an integrated copy of the gene rather than from extrachromosomal plasmids: genes are more stable when integrated and integration frees the selectable LEU2 marker for use on another plasmid in subsequent experiments. The gene encoding the rat type 2 cyclase can be inserted into the lys2 locus of strain CY732 (genotype MATa cdc35-l cam lys2 leu2 trpl ura3) using an integrating plasmid marked with URA3 and containing the rat adenylyl cyclase gene inserted into the LYS2 gene. CY732 will be transformed with this plasmid, and URA+ transformants will be selected and grown in the presence of 5- fluoroorotic acid (FOA) to select for the loss of URA3. Colonies that grow on FOA will be picked, transformed with a plasmid encoding Gas, and tested for an ability to grow at 34°C. Those yeast exhibiting Gαs-dependent growth at 34°C should bear the rat adenylyl cyclase gene integrated at the LYS2 locus. Their genotype will be designated MATa cdc35-l cam lys2::AC2 leu2 trpl ura3.

7. Screen for Activators and Inhibitors of type 2 adenylyl cyclase

Yeast bearing cdc35-l and expressing rat type 2 adenylyl cyclase can be used to screen for agents that stimulate the mammalian cyclase since activators of the latter protein will promote the growth of yeast at 34°C. (Accordingly, rat Gas is an example of an activator of adenylyl cyclase.) Specifically, haploid yeast that carry cdc35-1 and an integrated copy of rat type 2 adenylyl cyclase (genotype MATa cdc35-l cam lys2::ACII leu2 trpl ura3) will be used to screen libraries of natural or synthetic compounds to identify candidate activators of the mammalian adenylyl cyclase. Candidates will be capable of stimulating growth of the test strain at 34°C, but incapable of stimulating growth of the parental strain lacking the rat adenylyl cyclase gene (genotype MATa cdc35-l cam lys2 leu2 tipl ura3). As an extension of this approach, the test strain will be transformed with a library of URA3-containing ptasmids that encode peptides of random sequence. The transformants will be plated on uncD* deficient media and incubated at 34°C. Cells that express peptides that activate the mammalian adenylyl cyclase will form colonies due to "autocrine" stimulation of that ₅ enzyme. These peptides can be identified by isolating the peptide-encoding plasmids and sequencing the region that encodes the random peptide. Candidate activators of the mammalian adenylyl cyclase will be further tested in an in vitro biochemical screen with purified enzyme in order to confirm direct stimulation of the cyclase.

Haploid cdc35-l yeast bearing an integrated copy of rat type 2 adenylyl cyclase, io and a plasmid encoding Gas can be used in primary screens for inhibitors of the mammalian cyclase. Agents that reduce Gαs-dependent growth at 34°C will be considered candidate inhibitors of the cyclase and will be tested in a secondary biochemical screen using purified enzyme. This secondary screen will discriminate between agents that directly inhibit type 2 adenylyl cyclase and those that act indirectly,

I ₅ for example by interfering with the ability of Gas to stimulate the mammalian cyclase. Note that compounds that act by blocking the interaction of Gas with adenylyl cyclase are, in their own right of interest and will be characterized independently.

8. Expression of mammalian Gαi in yeast

20

While Gas can stimulate all known forms of adenylyl cyclase, type 1, type S and type 6 adenylyl cyclase have been shown to be inhibited directly by Gαi-1 [Taussig et al (1994)]. Haploid yeast tiiat carry cdc35-l, an integrated copy of type 5 adenylyl Cyclase, and a plasmid encoding Gas will be transformed with a high-copy plasmid tiiat encodes,

2₅ for example, Gαi-1. Yeast expressing the Gαi gene would be expected to grow more slowly at 34°C than the parental strain which lacks Gαi-1, due to the inhibitory effect of the Gαi-1 subunit. If this proves to be the case, this strain, expressing nrømmaiian adenylyl cyclase, Gas, and Gαi-1, can be used as a test strain to screen for compounds that interfere with the inhibition of cyclase by Gαi-1.

3 o Compounds that stimulate growth of the test strain at 34°C may be exerting this effect by blocking the interaction of Gαi-1 and adenylyl cyclase. However, the same growth-stimulatory effect would also be exhibited by compounds that directiy activate adenylyl cyclase, enhance the stimulatory effect of Gas on adenylyl cyclase, activate a downstream component whose activity was dependent on c AMP production, etc. To

35 distinguish the growth-stimulating compounds that directly affect the inhibitory influence of Gαi-1 on adenylyl cyclase from those that act elsewhere, all candidate compounds will be tested on a battery of isogenic control strains. One control strain will lack Gαi-1 (yet contain type 5 adenylyl cyclase and Gas); compounds that accelerate the growth of this strain or enable its growth over a wider range of temperatures will be considered to affect targets other than Gαi-1. Other control strains will be those without adenylyl cyclase type 5 or without Gas or without both adenylyl cyclase and Gas. Compounds that stimulate the growth of any of these control strains will be excluded as inhibitors of the interaction between Gαi-1 and adenylyl cyclase. Note that these control tests can lead to the identification of compounds that directly stimulate adenylyl cyclase or promote the stimulatory effect of Gas on adenylyl cyclase.

9. Expression of activated mammalian Gas in yeast

The Gas subunit exists in either of two forms, designated Gαs-GTP and Gαs- GDP. Under the conditions of our experiments, the predominant form of mammalian G as in yeast is expected to be Gαs-GDP. The ability of Gas to stimulate rat type 2 adenylyl cyclase in yeast to the presence of a relatively small pool of the GTP-bound form can be tested. As there may be experimental situations in which it is desirable that a larger pool of the activating species is available to stimulate adenylyl cyclase activity, will be exploited a constitutively active, mutant form of Gas. Haploid and diploid yeast of genotypes MATa cdc35-l cam lys2 leu2 tφl ura3 and MATa/MATα cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 Ieu2-3,112/leu2 tφl/tepl his7/+, respectively, will be transformed with the following two plasmids: a low-copy, LEU2-containing plasmid that encodes rat type 2 adenylyl cyclase driven by the PGK promoter and a high-copy, TRP I -containing plasmid that uses the CUP1 promoter to direct expression of a mutant Gas that is predominantly, if not exclusively, in the Gαs-GTP form. This activated form of Gas (GasQ227L), generated by conventional oligonucleotide-directed mutagenesis of the wild-type Gas allele, was identified as a GTPase-inhibiting mutation that results in constitutive activation of Gas and persistent stimulation of adenylyl cyclase in pituitary adenomas (Landis et al. (1989) Nature 340, 692-696). This mutation causes a 95% decrease in GTPase activity and the mutant Gas therefore exists predominantly in the GTP-bound fomi. In a strain co- expressing the rat adenylyl cyclase and the mutant Gas, a greater adenylyl cyclase activity exhibited by more rapid growth at 34°C, greater temperature range of growth, or greater sensitivity to the inductive effects of copper is expected.

10. Expression of mammalian Gβγ subunits in yeast In addition to expressing mammalian Gα subunits, yeast will be engineered to express specific mammalian Gβγ combinations. Coincident expression of mammalian G as or Gai, Gβ, and Gγ subunits in yeast will result in the reconstitution of mammalian heterotrimeric G proteins in yeast. In mammalian cells, these heterotrimeric G proteins couple a subset of seven-transmembrane receptors to adenylyl cyclase, causing the stimulation or inhibition of that enzyme. Therefore, expression in yeast of an appropriate G protein-coupled receptor and adenylyl cyclase, together with the components of a heterotrimeric G protein will duplicate a complete mammalian signal transduction pathway in that organism. Upon reconstitution of this pathway in appropriate strains, agents that activate or inhibit the seven-transmembrane receptor will influence cAMP-dependent growth of the yeast at 34°C. In addition, through appropriate choice of the particular isoforms of adenylyl cyclase, Gβ, and Gγ that are expressed in yeast, agents that affect the function of various Gβγ dimers will affect c AMP-dependent growth at 34°C.

Judicious choice of the Gβ and Gγ subtypes that are expressed in yeast will influence the utility of the strains that express mammalian adenylyl cyclase. While Gβl and Gγl can form a functional complex that binds Gas, the βlγl dimer exhibits little ability to activate type 2 adenylyl cyclase in the presence of Gas (J Biol Chem 267:23407, 1992). It is expected that expression of β 1 and γl in yeast that simultaneously express both mammalian Gas and type 2 adenylyl cyclase would, by forming a complex with Gas, lower the production of cAMP by the cyclase by preventing the stimulation of the cyclase by Gas. This, in turn, should be reflected in slower growth of the yeast when compared with the growth that occurs upon expression of Gas and the cyclase alone. Thus, slower growth of yeast simultaneously expressing the three mammalian G protein subunits and type 2 adenylyl cyclase will indicate that the mammalian G proteins form a heterotrimer in yeast. This result would provide ft yeast strain in which functional, mammalian heterotrimeric G proteins can be studied. Gβl and Gγl will be amplified by the polymerase chain reaction using as templates plasmids obtained from Dr. Mel Simon at Cal Tech. These plasmids contain the bovine genes encoding Gβl (Fong et al. (1986) Proc. Natl. Acad. Sci. USA 83, 2162- 2166) and Gγl (Hurley et al. (1985) Proc. Natl. Acad. Sci. USA 81, 6948-6952). Each of the two amplified products will be cloned into high-copy number plasmids marked with URA3 and LEU2. In these plasmids expression will be controlled by the copper- inducible CUP1 promoter. Haploid cdc35-l yeast bearing the rat type 2 adenylyl cyclase gene integrated at the LYS2 locus and expressing mammalian Gas under the control of the GPA1 promoter (which is constitutively active in haploid cells) from a TRP1 -marked high-copy plasmid will be transformed with the high-copy plasmids containing Gβ and Gγ marked with URA3 and LEU2, respectively. Thus, the yeast will carry three plasmids, each expressing a unique selectable marker and a mammalian G 5 protein subunit. These yeast should grow at 34°C in the absence of copper due to stimulation of rat type 2 adenylyl cyclase by Gas. Addition of copper to the growth medium will boost expression of Gβ and Gγ and should thereby inhibit growth as the Gβ γ dimer complexes Gas and prevent stimulation of the cyclase by the alpha subunit.

Upon demonstration that mammalian Gα, Gβ, and Gγ can form heterotrimers in l o yeast, yeast strains will be constructed whose growth is extremely sensitive to the degree of association of the βγ dimer with Gas. It has been documented that, in vitro, Gβl and Gγ2 form functional complexes that bind to Gas. But, in contrast to the βlγl complex, βlγ2 acts synergisticaily with Gas to activate type 2 adenylyl cyclase (Iniguez-Lluhi et al. (1992) J. Biol. Chem. 267, 23407-23417). That is, the ability of free Gas to stimulate is type 2 cyclase is enhanced by free βlγ2 dimer. Note that expression of equal levels of mammalian Gas, Gβl, and Gγ2 in yeast that also express type 2 adenylyl cyclase should result in little cyclase activity due to the formation of the heterotrimeric G protein. But these yeast are poised to exhibit greatly enhanced adenylyl cyclase activity, and therefore growth at 34°C, in response to interventions that dissociate Gas and Gβγ. For

20 example, these yeast will be extremely sensitive to compounds that interfere with the association of Gas and βγ.

Yeast of genotype MATa cdc35-l cam lys2::ACII leu2 tφl ura3 will be transformed with the following three high-copy number plasmids:

1. a TRP 1 -marked plasmid with the mammalian Gas under the control of the CUP 1 25 promoter;

2. a LEU2-marked plasmid with the mammalian Gβl under the control of the GPA1 promoter;

3. a URA3 -marked plasmid with the mammalian Gγ2 under the control of the GPA1 promoter.

30 Induction of Gas expression with CuSO4 should not enable cells to grow at 34°C due to the binding of Gas by Gβγ. If, however, growth at 34°C is stimulated by copper, it will indicate a probable excess of Gas subunits relative to Gβγ. Accordingly, the expression of Gas can be reduced by lowering the concentration of copper in the medium until no growth is observed. (Parallel titrations of a control strain lacking Gβl and Gγ2 with

₃ s CuSO4 will be conducted to confirm that this concentration of CuSO4 can induce expression of sufficient Gas to induce growth at 34°C in the absence of Gβl and Gγ2.) At this copper concentration there should be approximately equimolar intracellular concentrations of the three mammalian G protein subunits. While this strain will be unable to grow at 34°C in the presence of this critical concentration of copper, growth should be extremely sensitive to agents that dissociate α from βγ. Accordingly, this s strain will be further engineered to express mammalian seven-transmembrane receptors, such that activation of these receptors at the critical copper concentration results in the growth of the strain at 34°C.

Low copy and integrated ACII can rescue growth of cdc35-l strains, o We have, for the reasons specified above, integrated rat AC2 (adenylyl cyclase 2) into the yeast genome. The plasmid directing integration of AC2 was constructed by cloning the PGK promoter-driven AC2 into a vector (Cadus 1294; genotype CmR Tys2 lys2') that directed integration of the PGKp-AC2 into the LYS2 locus of strain CY1789 (genotype MATa tbtl-1 cdc35-l ura3 his3 trpl leu2 (cam?). Specifically, PGKp-AC2 s was excised from Cadus plasmid 1512 as a 4.4 kb EcoRI-to-Bglll fragment the Bgl II site was blunt-ended, and the fragment was cloned into the EcoRI and Sma I sites in the polylinker of Cadus plasmid 1294. The resulting construct (Cadus plasmid 1633) was linearized at the unique Bgl II site, and strain CY1789 was transformed with the linearized DNA. Since this targeted integration of AC2 disrupts wild-type LYS2, o integrants were selected on plates containing α-aminoadipate (2 g/1), a compound that confers a growth advantage to yeast lacking fully functional LYS2 (Chatoo et al. Genetics 93:51, 1979). One such integrant CY1936 (genotype MATa lys2::PGKp-ACII tbtl-1 cdc35-l ura3 his3 trpl leu2 (cam?), was transformed with a plasmid encoding rat Gas whose expression is under the control of the copper-inducible CUP1 promoter. The 5 resulting strain exhibited growth at 34°C in the presence of 1 OOμM CuSO4, reflecting the ability of the integrated AC2 to complement the temperature-sensitive cdc35-l allele, provided that Gas is also expressed. An isogenic diploid strain was made from CY1936, and this strain was also shown to express functional rat AC2, as demonstrated by its ability to grow at the restrictive temperature as long as Gas was coexpressed. 0

AC type IV confers temperature resistance to cdc35-l strains in a Gαs-dependent manner, both episomally and integrated.

A cDNA encoding rat type 4 adenylyl cyclase was provided by Dr. Al Oilman of the University of Texas Southwestern Medical Center. A plasmid was constructed for expressing AC4 in yeast by subcloning the 3.2 kb Spel to Bglll fragment, which contained all but the 14 N-terminal amino acids of AC4 open reading frame, into a LEU- marked 2μ vector (Cadus plasmid 1849) that contained the PGK promoter followed by a synthetic oligonucleotide encoding the N-terminal 14 amino acids of AC4. Insertion of the Spel to Bglll fragment resulted in Cadus plasmid 1856, in which the PGK promoter directs transcription of the entire open reading frame of a version of AC4 in which the 14 N-terminal codons have been optimized for expression in yeast. Diploid yeast that are homozygous for the cdc35-l allele, which encodes a temperature-sensitive yeast adenylyl cyclase, and that contain a plasmid encoding a CUP1 promoter-driven wild- type rat Gas were transformed with Cadus plasmid 1856. Transformants were tested for their ability to grow at the restrictive temperature (34°C) in the presence of copper. Strain C Y2128 and C Y2129, representatives of these transformants, exhibited copper- dependent temperature-resistant growth, indicating that AC4 can function in these yeast in the presence of mammalian Gas. As was observed with AC2, integration of PGK promoter-driven AC4 at the LYS2 locus yielded strains that retained the ability to grow at 34°C when Gas is coexpressed.

GasQ227L shows greater growth than wild-typet Gas, reflected in greater background growth, more rapid growth, and greater temp resistance. Experiments with GαsG226A confirm that Gαs.GTP is the activating species.

We have constructed both high copy and low copy plasmids that express wild- type rat Gas from the haploid-specific GPA1 promoter, the copper-inducible CUP1 promoter, and the constitutive PGK promoter. In each case, expression of Gas together with either AC2 or AC4 resulted in temperature-resistant growth of cdc35-l yeast. As described above, Gas can exist in two forms, distinguished by the guanine nucleotide bound to it: Gαs»GTP and Gαs^GDP. Which of the two species responsible for stimulating adenylyl cyclase was determined by comparing the temperature-resistant growth of strains expressing the wild-type Gas with isogenic strains expressing either of two mutant forms of Gas, GasQ227L and GαsG226A. The GαsQ227L mutant has only 5% of wild-type GTPase activity, so it exists predominantly in the Gαs»GTP form. In contrast, GαsG226A shows severely compromised exchange of GTP for GDP, so it exists in predominantly in the Gαs»GDP form. Ifthe Gαs«GTP form is the active species, it might indicate that strains expressing mammalian adenylyl cyclases would show the greatest temperature resistant growth with coexpressed GαsQ227L and the least temperature-resistant growth with coexpressed GαsG226A.

Cadus plasmid 1536 is analogous to Cadus plasmid 1046 (see above), except that it contains a constitutively active Gas (GasQ227L) in lieu of the wild type Gas. Cadus plasmid 1843 is also analogous to Cadus plasmid 1046, except that it contains Gα sG226A in place of the wild type Gas. Diploid temperature-sensitive yeast strains carrying PGK promoter-driven AC2 on a LEU2-marked high copy plasmid (Cadus plasmid 1512) were transformed with these plasmids to yield strains CY1429 and-. CY1430 (genotype cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 leu2-3,l 12/tøiiS! tφl/Upl his7/+ [TRP1 CUPlP-ratGαs REP3 2mu-ori AmpR / LEU2 2mu-ori R$ϊ»3 AmpR PGKp-ratACII]); CY1773 and CY1774 (genotype cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 1eu2-3,112Λeu2 tτpl/tφl his7/+ [TRP1 CUPlP-ratGαsQ227L BBP3 2mu-ori AmpR / LEU2 2mu-ori REP3 AmpR PGKp-ratACII]); and CY 2052 and CY2053 (genotype cdc35-l/cdc35-l cam/(cam?) ura3-52/ura3-52 leu2-3,l 12/leu2 trpl/tφl his7/+ [TRP1 CUPlP-ratGαsG226A REP3 2mu-ori AmpR / LEU22rmiiθri REP3 AmpR PGKp-ratACII]). Each of these strains was plated on synthetic media with and without lOOμM CuSO4 and incubated at 34°C. While the strain expressing wild type Gas grew in the presence but not the absence of copper, the strain expressing GαsG226A showed no growth at 34°C in the presence or absence of copper. The strain expressing GαsQ227L grew in the presence and absence of copper, indicating that the specific activity of GαsQ227L is sufficiently high to enable growth at 34°C even at levels resulting from the basal activity of the CUP1 promoter. The greater activity of G αsQ227L compared to wild type Gas is also reflected by a greater range of temperature resistance of strains CY1773 and CY1774 compared to CY1429 and CY1430: AU2- expressing strains that contain GαsQ227L will grow at temperatures as high as 37*C if expression is induced with added copper, while strains containing wild type Gas fail to show such temperature resistance under any conditions.

These experiments indicate that Gαs»GTP is the species that activate AC2 expressed in yeast. Similar results are obtained with AC4 in place of AC2. The results are consistent with in vitro studies that suggest that Gαs*GTP is the stimulatory form of Gas. The results therefore indicate that mammalian adenylyl cyclases and mammalian G proteins expressed in yeast exhibit normal physiological behavior.

Human and rat Gas proteins are equally effective in activating rat ACII. The genes encoding these proteins are expressed at different levels, however, due to a lower translatability of the 3' 1/3 of the coding sequence of the human Gas mRNA.

Human and rat Gas subunit protein differ at a single amino acid: the rat protein contains an asparagine at position 6 while this residue is a threonine in the human protein. To determine ifthe human and rat Gas have different activities on adenylyl cyclase type 2 (AC2), the rat and human coding sequences were expressed from high copy plasmids containing the CUP1 promoter in haploid cdc35-l yeast expressing AC2 from a promoter on a LEU2-marked, high-copy plasmid (Cadus plasmid 1512). The resulting strains showed different growth rates at the restrictive temperature: the strains expressing rat Gas (CY1635 and CY1636) grew more rapidly than those expressing the 5 human protein (CY1703 and C YI 704). Similar results were observed in homozygous dipoid cdc35-l yeast, with different promoters driving the expression of the two Gas's (viz., PGK and GPA promoters), and with the Gas's expressed from both high and low- copy plasmids. While these results suggested that the single amino acid difference between the rat and human proteins was responsible for the difference, additional o experiments indicated that in fact this was not the case: yeast expressing a human Gas from a chimeric gene, wherein the 5' approximately 30% (from the start codon through the Eco RI site) of the human Gas coding sequence was replaced with the analogous region encoding the rat Gas, grew at rates equal to that of the yeast expressing the rat gene. The lower activity of the human Gas was mapped to the 3' one-third of the s human Gas coding region, from the Bglll site to the stop codon, which encodes amino acids identical to those encoded by the rat gene in this region. The 5' end of the human coding sequence seems to be less efficiently expressed than the equivalent region of the rat gene, since there is no reason to doubt that the activities of the rat and human Gas's that are expressed (i.e., their specific activities) are indistinguishable with respect to o stimulation of adenylyl cyclases.

Construction of yeast expression vector for mouse adenylyl cyclase type 6 and rat adenylyl cyclase type 3.

s We obtained mouse adenylyl cyclase type 6 from Gary Johnson as a 5 kb cDN A in a plasmid designated Cadus plasmid 1757. This was subcloned in three steps into a LEU2-marked high-copy yeast expression vector containing the PGK promoter (Cadus plasmid 1284). The first step consisted of : 1 ) amplifying from Cadus 1757 by the polymerase chain reaction (PCR) a fragment containing the N-terminal 1.2 kb of the o AC6 open reading using oligo 1 10 (5' CAGACATGTCTTGGTTTCGTGGCCTCCTG 3') (SEQ ID NO:5) and oligo 111 (5'

GCGGATCCAAGGTCATGACCAGTTCCTGTGCAGTGC 3') (SEQ ID NO:6), 2) cutting the PCR product with Afl III and BamH I, and 3) cloning the amplified product into Ncol- and BamH I-digested Cadus plasmid 1284 (LEU2 PGKp 2mu-ori REP3 ₅ AmpR). This yielded Cadus plasmid 1918. Note that the 1.2 kb PCR-amplified product contains an Sph I site near its 3'end that derives from AC6. This endogenous Sph 1 site was used in the second step, which consisted of: 1 ) excising the 3' 574 nucleotides of the AC6 open reading frame from Cadus 1757 as part of a 2.1 kb Sph I-to-BamH I fragment and 2) cloning the fragment into Sph I- and BamH I-digested Cadus 1918. The result was Cadus plasmid 1919. The final step involved: 1 ) excising the middle 2 5 kb of AC6 open reading frame as an Sph I fragment, 2) cloning it into Sph I-digested Cadus 1919, and 3) screening recombinants for clones that contained a plasmid containing a reconstructed AC6 open reading frame. The resulting plasmid was designated Cadus plasmid 1950.

We obtained rat adenylyl cyclase type 3 (rat AC3) from Gary Johnson as a 4.5 kb o cDNA in a plasmid designated Cadus plasmid 1756. The expression plasmid for expression of AC3 in yeast was constructed as follows. Oligos 1 12 (5' CATGACTGAAGATCAAGGTTTCTCG 3') (SEQ ID NO:7)and 113 (5' GATCCGAGAAACCTTGATCTTCAGT 3') (SEQ ID NO:8)were annealed, and the double-stranded oligonucleotide, which encodes the N-terminal 9 amino acids of AC3, s was cloned into Neo I- and BamH I-digested Cadus plasmid 1284 (LEU2 PGKp 2mu- ori REP3 AmpR) to yield Cadus plasmid 1894. The rest of the AC3 open reading frame was then inserted by cloning the 3.5 kb BamH I-to-HindlH fragment from Cadus 1756 into BamH I- and Hindlll-digested Cadus 1894. The resulting plasmid, Cadus 1916, contains an AC3-encoding gene whose N-terminal 8 amino acids are optimized for 0 expression in yeast and whose transcription is directed by the PGK promoter.

In cam 1,2,3 strains expressing Gas and various cyclase isoforms: ACI shows a temperature optimum at Room Temρerature(RT), with some growth at 30° C; AC IV and ACVI show growth over a broad temp range; the optimum for AC4 may 5 be RT, for ACVI may be 30°C.

ACIII shows no growth at any temp.

Yeast strains that contain the cdc35-l temperature-sensitive allele can grow at 30 °C and lower temperatures, reflecting the activity of the mutant yeast adenylyl cyclase at o these temperatures. When these strains are used to assess the capacity for mammalian adenylyl cyclases to complement the cdc35-l mutation, complementation tests can only be done at temperatures above 30°C. In practice, the temperature range available for such tests is about 33^βC to 37°C. If for some reason the mammalian enzyme is not functional in this relatively narrow temperature range, then successful complementation s will not be observed in this yeast background. It is therefore advantageous to have a strain background that permits determination of functional cyclase activity at temperatures below 30°C.

We have obtained a yeast strain from Dr. Al Gilman at the University of Texas Southwestern Medical Center that is derived from a strain (TC41 ) constructed by Warren Hiedeman at the University of Wisconsin. Strain TC41 does not encode a yeast cyclase as a result of deletion of CYR 1 and carries three uncharacterized mutations (caml, cam2, and cam3) that enable growth of the strain on media containing cAMP. Dr. Gilman' s group has modified this strain in part by integrating rat GαsQ227L under the control of the CUP1 promoter at the TRP1 locus. The resulting strain (CY2828; genotype MATa cyrl ::ura3 t^χpl-l ::CUPlp-GαsQ227L caml cam2 cam3 leu2-3 leu2- 112 his3-532 his4) requires added cAMP to grow.

To examine the ability of various mammalian adenylyl cyclases to enable growth of CY2828 in the absence of added cAMP, CY2828 were transformed with: 1) Cadus plasmid 1856 (described above), which encodes AC4, to yield sibling strains CY2906 and CY2907 (genotype MATa cyrl ::ura3 tφl-1 ::CUPlp-G@sQ227L caml cam2 cam3 leu2-3 leu2-l 12 his3-532 his4 [LEU2 2mu-ori REP3 AmpR PGKp-AC4]); 2) Cadus plasmid 1916 (described above), which encodes AC3, to yield sibling strains CY2908 and CY2909 (genotype MATa cyrl::ura3 tφl-1 ::CUPlp-G@sQ227L caml cam2 cam3 leu2-3 leu2-l 12 his3-532 his4 [LEU2 2mu-ori REP3 AmpR PGKp-ACIII]); 3) Cadus plasmid 1950 (described above), which encodes AC6, to yield sibling strains CY2910 and CY2911 (genotype MATa cyrl ::ura3 trpl-1 ::CUPlp-G@sQ227L caml cam2 cam3 leu2-3 leu2-l 12 his3-532 his4 [LEU2 2mu-ori REP3 AmpR PGKp-ACVI]); and 4) Cadus plasmid 2129 (constructed in Dr. Giman's laboratory), which encodes AC1, to yield sibling strains CY2906 and CY2907 (genotype MATa cyrl ::ura3 tφl-1 ::CUPlp- G@sQ227L caml cam2 cam3 leu2-3 leu2-l 12 his3-532 his4 [AmpR LEU2 CEN ARS CUP 1 p-ste6-AC 1 ]). Growth of each strain was determined in the absence of exogenous cAMP at room temperature, 30°C, 34°C, and 37^βC. Both AC4 and AC6 showed effective complementation of the cyrl deletion at 30°C and 34°C; AC1 complemented the deletion best at room temperature; and AC3 was unable to complement the deletion at any temperature tested.

Evidence that βγ can form a functional heterotrimer reduction of the stimulation of cyclase by Gα.

We constructed URA3-marked CEN ARS plasmids that contained various pairs of Gβ and Gγ on the same plasmid. For example, a plasmid that expresses both Gβ 1 and Gγ2 was constructed as follows. The open reading frame of bovine Gγ2 was PCR- ampiified from Cadus plasmid 1319 (provided by Dr. Melvin Simon and containing the Gγ2 cDNA) using primers A 14652 (5'

GGGCGTCTCCCATGGCCAGCAACAACACCGC 3') (SEQ ID NO:9and AI4653 (5^* GGGGTCGACCGAGGCTCCTCAGGTTCCTC 3 ) (SEQ ID NO: 10 The amplified s product was digested with Esp3I and Sal I and cloned into the Neo I and Sal I sites of Cadus plasmid 1449. The resulting plasmid, Cadus 1705, uses the PGK promoter to direct expression of Gγ2. The PGKpromoter-Gγ2 unit was then excised from Cadus 1705 as a 1 kb Not I-to-Xho I fragment and cloned into Not I- and Xho I-digested Cadus plasmid 1460. The resulting plasmid, Cadus 1781, then received a 422 bp fragment o containing the ADHl promoter. Specifically, the ADHl promoter was removed from Cadus 1625 by cutting with Nhel, filling in the overhang, cutting with Spe I, and isolating the 422 bp fragment. This fragment was ligated to Cadus 1781 that had been cut with Xbal, the 5' overhang filled in, and digested with Spel. A recombinant (Cadus plasmid 2209) that contained the ADHl promoter together with PGK promoter-driven G s γ2 was selected and used as the recipient for the Gβl open reading frame. In particular, the bovine Gβl open reading frame was PCR-amplified from Cadus 1315 (provided by Dr. Mel Simon and containing the Gβl cDNA) using primer 123(5' CGGCTAGCATCTATATACAATGAGTGAACTTGACCAGTTACGGC 3^*) (SEQ ID NO:l 1) (5' CGAGCGGCCGCTCAGTTCCAGATTTTGAGGAAGCTGTCC 3*) (SEQ 0 ID NO: 12). The amplified product was digested with Not I and Nhe I and cloned into Notl- and Nhe I-digested Cadus 2209, yielding Cadus plasmid 2254. Thus, Cadus 2254 is a URA3-marked low-copy plasmid that directs the expression of Gγ2 from the PGK promoter and Gβl from the ADHl promoter. A similar construction strategy yielded analogous plasmids encoding Gβl and Gγl (Cadus plasmid 2255), Gβ2 and Gγl (Cadus 5 plasmid 2257), Gβ2 and Gγ2 (Cadus plasmid 2256), Gβ3 and Gγl (Cadus plasmid

2259), Gβ3 and Gγ2 (Cadus plasmid 2258), Gβ4 and Gγl (Cadus plasmid 2363), and Gβ 4 and Gγ2 (Cadus plasmid 2361).

We have tested the ability of Gβl and Gγ2 (encoded by Cadus 2254) and Gβl and Gγl (encoded by Cadus 2255) to form a complex with Gas by assessing the effect 0 of coexpression of Gβ and Gγ on the stimulation of AC2 by Gas. If a Gαβγ complex can be formed upon coincident expression of the three G protein subunits, it could indicate that the level of free Gα would be lower in yeast expressing all three subunits compared to yeast expressing only the Gα subunit. This would be reflected as slower growth of a cdc35-l strain expressing a mammalian adenylyl cyclase together with the s three G protein subunits compared to an equivalent strain expressing the mammalian cyclase with Gα alone. Yeast strain CY2065 (genotype MATa/α lys2::PGKp-ACII /lys2::PGKp-ACII tbtl-l/tbtl-1 cdc35-l/cdc35-l cam/(cam?) ura3/ura3 Ieu2/leu2 tφl/tφl) was transformed with Cadus plasmid 2081 (TRP1 CEN6 ARSH4 AmpR CUPlp-ratGαs) and Cadus plasmid 2254 (URA3 PGKp-Gγ2 CEN6 ARSH4 AmpR ADHlp-Gβl) to get sibling strains CY3845 and CY3846. These strains will express mammalian proteins Gβ 1 , Gγ2, and AC2 constitutively and Gas in the presence of added copper. Equivalent strains were constructed that express mammalian proteins Gas, Gβl, Gγl, and AC2 (CY3847 and CY3848) or Gas and AC2 alone (CY3861 and CY3862). These 6 strains were tested for temperature-resistant growth at 34°C by spotting one thousand cells on solid media that selects for maintenance of the plasmids. The strains containing Cadus plasmid 2254 (encoding Gβl and Gγ2) or Cadus plasmid 2255 (encoding Gβl and Gγl) grew more slowly than those not expressing a Gβγ. The three G protein subunits results in the formation of a G protein heterotrimer.

Example 2.

This example describes the isolation of constitutively active, mutant forms of Type IV adenylyl cyςlase (herein "ACIV" or "AC4").

Although the cAMP signaling pathway is of fundamental importance to cellular physiology, few chemicals have been identified that influence the activity of the enzymes that produce cAMP, the adenylyl cyclases (ACs). The only pharmacological agents that directly activate the ACs are forskolin and its analogs (DeSouza, N.J. 1993. Industrial development of traditional drugs: the forskolin example. A mini-review. J. Ethnopharmacol. 38:177-180), while the known inhibitors consist of derivatives of adenosine that bind to the P-site on the enzyme (Londos, C, and J. Wolff. 1977. Two distinct adenosine-sensitive sites on adenylate cyclase. Proc. Natl. Acad. Sci. USA 74:5482-5486; Desaubry, L., I. Shoshani, and R.A. Johnson. 1996.2',5'- dideoxyadenosine 3 'polyphosphates are potent inhibitors of adenylyl cyclases. J. Biol. Chem. 271.2380-2383). Neither of these classes of modulators is specific for any of the nine different isoforms of membrane-bound mammalian ACs. Nor have the regions of the cyclases that interact with these modulators been identified. Indeed, the catalytic site of the ACs has yet to be mapped. Thus, an impasse exists for developing drugs that target ACs, created by two reinforcing deficiencies: the lack of mechanistic knowledge of cyclase enzymology has slowed development of pharmacological agents, and the lack of multiple pharmacological tools has hindered structure-function analysis of ACs. Speciβc Background - Expression of Mammalian ACs in Yeast

The present invention enables rapid screening of libraries of compounds for activators and inhibitors of human adenylyl cyclases. The technology involves the expression of mammalian isoforms of adenylyl cyclases and Gα subunits in yeast with either of two genetic backgrounds. One of these is strain CY2827 (genotype MATΛ cyrl::ura3 caml cam2 cam3 leu2-3 leu2-112 his3-532 his4 ura3) which cannot grow on media without supplemental cAMP due to a deletion of CYR1, the gene encoding the yeast adenylyl cyclase. However, due to the presence of the triple caml cam2 cam3 mutations the strain can grow on media that has been supplemented with cAMP. Ihe second approach uses strain CY1789 (genotype MAT*/a cdc35-l/cdc35-l ura3-$2fura3- 52 leu2-3, 112/leu2 trpl/trpl his7/+) which harbors the mutant cdc35-l allele at the CYR1 locus, which encodes a temperature-sensitive yeast adenylyl cyclase. The cyclase encoded by cdc35-l is active at room temperature, permitting cell growth, but inactive at 34°C, resulting in growth arrest. Expression of Type I, II, IV, V, or VI marnmalian adenylyl cyclase together with mammalian Gas in both of these genetic backgrounds enables their growth under the appropriate restrictive conditions. That is, co-expression of mammalian adenylyl cyclase and Gas in the CY2728 background enables grøjih on media lacking supplemental cAMP, while co-expression in the CY1789 background enables growth at 34°C. Complementation of mutant cyrl by all mammalian adenylyl cyclases ths far examined requires co-expression of mammalian Gas, as is true in mammalian cells (Taussig, R. and A.G. Gilman. 1995. Minireview: Mammalian membrane-bound adenylyl cyclases. J. Biol. Chem. 270:1-4). Indeed, GTP-bound Gas is likely the form of Gas that stimulates mammalian AC expressed in yeast Co- expression of the GTPase deficient mutant GαsQ227L instead of wild-type Gas results in more robust growth under the restrictive conditions, in contrast, growth is not observed when adenylyl cyclases are coexpressed with the GαsG226A mutant, which exists predominantly in the GDP-bound form.

The requirement for expression of Gas along with AC for growth under restrictive conditions is not optimal for two reasons. First robust growth, reflecting acceptable activity of AC, requires expression of Gas from high-copy plasmids. Expression from low-copy centromeric episomes or from genes integrated at either the STE2 or the LYS2 locus yields modest growth. The apparently low AC activity obtained when Gas is integrated necessitates expression of Gas from a high copy episome, which requires maintenance and use of the yeast on synthetic media. This is less desirable than strains that can be used on rich media, where their growth is more rapid and growth- dependent phenotypes are generally more robust. The second drawback to requisite co- expression of Gas and ACs is that a compound identified by the yeast screen as an inhibitor may target not the AC directly, but rather Gas or the interaction between Gas and the AC. Determining the true target of the compound would require re-testing in separate secondary assays.

Thus, yeast-based screens that use a constitutively active (i.e., Gαs-independent) adenylyl cyclase enable rapid identification of compounds that target AC directly. Also, when used in conjunction with yeast that coexpress Gas and AC, screens using a Gαs- independent AC can identify compounds that affect Gas or the interaction between AC and Gas.

Structural Studies of ACs.

The nine cloned isoforms of mammalian AC are integral membrane proteins with the same predicted topology. The enzymes contain in succession: a cytosolic N- terminus (designated N) that varies from about 30 to 180 amino acids in length among the isoforms; an intensely hydrophobic domain (Ml) predicted to contain 6 transmembrane helices; a large cytosolic domain of about 40 kD (Ci); a second hydrophobic region (M2) that, like the first, is predicted to include 6 transmembrane helices; and a cytosolic C-terminus (C2) that varies from about 30 to 50 kD. ACs consist of two topologically similar halves, where neither half alone exhibits enzymatic activity. Thus, no cyclase activity is observed in Sf9 cells infected with recombinant baculovirus encoding NM1C1 or M2C2. However, concurrent expression of the two halves in Sf9 cells results in substantial activity (Tang, W.-J., J. Krupinski, and A.G. Gilman. 1991. Expression and characterization of calmodulin-activated (type I) adenylyl cyclase. J. Biol. Chem. 266:8595-8603; Tang, W.-J., M. Stanzel, and A.G. Gilman. 1995. Truncation and alanine-scanning mutants of type I adenylyl cyclase. Biochemistry 34:14563-14572).

Domain Cl is conserved among isoforms, as is C2, and some homology is even observed between the two cytosolic regions of the same AC. The homology observed among isoforms in their C 1 and C2 regions is attributable to a subdomain within each. Thus, the 250 amino acids that immediately follow M] exhibit about 90% similarity among isoforms. Likewise, the 250 amino acids after M2 of different isoforms are similar. These subdomains of high homology are designated Cla and C2a. respectively. Each is bordered on its C-terminal side by a more divergent region, termed C 1 b and

C2b. so that Cib separates Cι_a from M2, and C2b constitutes the genuine C-terminal o 97/

- 101 -

domain. The importance of Ci_a and C2a for catalytic and regulatory properties of ACs is suggested by the observation that co-expression of NM i C i _a and M2C2a •" Sf9 cells results in substantial AC activity that was Gas- and forskolin-responsive (Kawabe, J., T. Ebina, S. Ismail, D.B. Kitchen, CJ. Homey, Y. Ishikawa. A novel peptide inhibitor of adenylyl cyclase. 1994. J. Biol. Chem 269:24906-2491 1 ). These regions are important to enzyme activity since: a synthetic 20 amino acid peptide corresponding to a region within Cia of ACV inhibits forskolin- and Gαs-stimulated ACV activity (Kawabe, j., T. Ebina, S. Ismail, D.B. Kitchen, CJ. Homey, Y. Ishikawa. A novel peptide inhibitor of adenylyl cyclase. 1994. J. Biol. Chem 269:24906-2491 1 ); C la and C2a are homologous to a region in the adenylyl cyclase of Saccharomyces cerevisiae that contains the catalytic domain (Krupinski, J., F. Coussen, H.A. Bakalyar, W.-J.Tang, P.G. Feinstein, K. Orth, C. Slaughter, R.R. Reed, and A.G. Gilman. 1989. Adenylyl cyclase amino acid sequence: possible channel-or transporter-like structure. Science 244:1558-1564); Cia and C2a show homology with a number of diverse nucleotide-binding proteins (Krupinski, J., F. Coussen, H.A. Bakalyar, W.-J.Tang, P.G. Feinstein, K. Orth, C. - < Slaughter, R.R. Reed, and A.G. Gilman. 1989. Adenylyl cyclase amino acid sequence: possible channel-or transporter-like structure. Science 244: 1558-1564); and a synthetic peptide corresponding to a C2a region inhibits effects of Gβγ on ACII (Chen, J., M. DeVivo, J. Dingus, A. Harry, J. Li, J. Sui, D.J. Carty, J.L. Blank, J.H. Exton, R.H.. Stoffel, J. Inglese, R.J. Lefkowitz, D.E. Logotiietis, J.D. Hildebrandt R. Iyengar. l£95. A region of adenylyl cyclase 2 critical for regulation by G protein βγ subunits. Science 268: 1166-1169). Cib also appears to have some regulatory function, at least in ACI: the calmodulin binding domain, which mediates calcium activation of the enzyme, maps to Cib (Vorherr, T., L. Knopfel, F. Hofmann, S. Mollner, T. Pfeuffer, E. Carafoli. I9f 3- The calmodulin binding domain of nitric oxide synthase and adenylyl cyclase.

Biochemistry 32:6081-6088; Wu, Z., S.T. Wong, D.R. Storm. 1993. Modificatioftof the calcium and calmodulin sensitivity of the Type I adenylyl cyclase by mutagenesis of its calmodulin binding domain. J. Biol. Chem. 268:23766-23768).

Thus, Cib and C2b seem to be expendable for the catalytic and many of regulatory properties of ACs. Additional studies indicate that N, M i , and M2 are also dispensable for catalysis and regulation by Gas and forskolin. These additional studies exploit an extensively deleted, chimeric adenylyl cyclases, consisting of Cia of Type I AC linked via short peptides to C2a of Type II AC, expressed in E. coli (Tang, W.J. and A.G. Gilman. 1995. Construction of a soluble adenylyl cyclase activted by Gsα and forskolin. Science 268:1769-1772). Such soluble adenylyl cyclases are active and stimulated by both forskolin and Gas. Continuing studies will define the minimal catalytic and regulatory regions in such soluble cyclases.

While the foregoing studies have delimited the regions of AC involved in catalysis and Gas regulation, additional structural information may be gained from ACs with phenotypically interesting point mutations. In fact, alanine scanning mutagenesis 5 of particular, conserved residues suggests that both Cia and C2a are required for catalytic activity and that these two domains are not functionally equivalent (Tang, W.J. and A.G. Gilman. 1995. Construction of a soluble adenylyl cyclase activted by Gsα and forskolin. Science 268:1769-1772). Yet, despite the nonequivalence of Cι_a and C2a_> it has not been possible to ascribe unique functions to each. It is has been suggested that o one of these highly conserved domains is predominantly regulatory, while the other is predominantly catalytic (Tang, W.J. and A.G. Gilman. 1995. Construction of a soluble adenylyl cyclase activted by Gsα and forskolin. Science 268:1769-1772). Any model of AC activity, however, remains speculative due to the lack of informative AC mutants. Constitutively active ACs are particularly informative mutants. They enable the 5 identification of specific residues that participate in enzyme catalysis and regulation. They also make significant contributions to a mechanistic understanding of the human adenylyl cyclases. In addition to their utility for constructing yeast strains for screening compound libraries for modulators of human adenylyl cyclases, they are useful tools for deciphering the enzymology of ACs. In turn, better structure-function understanding of 0 ACs may enable additional, unanticipated improvements in yeast- and bacteria-based assays of AC activity.

Generation and Selection of Constitutively Active Mutants of Type IV Adenylyl Cyclase

s Cadus plasmid 1856(described herein) contains the coding region of rat Type IV adenylyl cyclase (ACIV) under the control of the PGK promoter, along with the yeast LEU2 gene, which enables genetic selection of yeast harboring the plasmid, and the 2 micron origin of replication, which allows replication of the plasmid to about 40 copies per cell. In addition, Cadus 1856 carries bacterial sequences that allow replication of the 0 plasmid in E. coli and selection of bacteria that harbor the plasmid. Cadus 1856 was subjected to random mutagenesis by propagating the plasmid in bacterial strain XL1- Red (Stratagene Cloning Systems, La Jolla, California), which is deficient in three of the primary DNA repair pathways due to mutations mutS, mutD, and mutT. Specifically, 100 ml of manufacturer-supplied, electrocompetent, Epicurian Coli XL 1 -Red TM were 5 transformed with 100 ng Cadus 1856, and transformants were plated on 5 ampicillin- containing plates. After growth at 37°C for about 30 hours, ampicillin-resistant colonies were scraped into two 500 ml cultures of ampicillin-containing 2xYT. The cultures were incubated at 37°C overnight with shaking, and fresh 500 ml cultures were started by inoculating 500ml 2xYT + ampicillin with 2.5 ml of each of the overnight cultures. Plasmid DNA was prepared from the remaining overnight cultures (total volume_! nearly 5 1 liter) and then combined. This procedure was repeated on each of the following two days, yielding a total of 3 preparations of plasmid DNA that had been propagated m the XL 1 -Red strain for 2, 3, or 4 days. These preparations are designated library 241 , library 242, and library 243, respectively. ,,-g.

Yeast strain CY2827 (cyrl::ura3 caml cam! cam3 leu2-3 Ieu2-U2 his3-532 o his4 ura3), obtained from Warren Heideman at the University of Wisconsin, Madison, was transformed by electroporation (Becker, D.M., and L. Guarente. 1991. High-efficiency transformation of yeast by electroporation. Methods Enzymol. 194:182-187) with the each of the three libraries, and transformants were subjected to a variety of selection protocols. According to Protocols 1 and 2, transformants obtained with each library were plated on 5 medium containing 100 mM cAMP but deficient in leucine and incubated at 30°C. After 4 days leucine prototrophs were collected from the plates by washing with water and aliquots of yeast harboring each of the three libraries were subjected to two different protocols. With Protocol 1, 6x10 cells were re-plated onto solid media that lacked both cAMP and leucine for incubation at 30°C for 7 days, after which growing colonies were o collected and their DNA extracted for amplification in bacterial strain DH 10B. Yeast strain CY2827 was transformed with this bacterial plasmid DNA, and transformants were selected on media lacking both cAMP and leucine. Colonies of the resulting prototrophs for both cAMP and leucine were picked, expanded, and their plasmid DNA

7 extracted for DNA sequencing and functional analysis. With Protocol 2, 6x10 ceils 5 were re-plated onto solid media that lacked both cAMP and leucine for incubation at 30°

C for 3 days, and plasmid DNA from the c AMP and leucine prototrophs was extracted and amplified in DHl 0B. Yeast strain CY2827 was transformed with the resulting preparation of bacterial plasmid DNA, transformants were selected on media without cAMP or leucine, and plasmid DNA was recovered for another round of amplification in o bacteria, introduction into C Y2827, and recovery of clonal DNA from particular cAMP and leucine prototrophs for sequencing and functional analysis.

According to Protocol 3, yeast CY2827 was transformed with libraries 241, 242, and 243 and transformants were plated directiy onto media without leucine or cAMP. DNA was extracted from growing colonies, amplified in bacteria, introduced into 5 CY2827, and recovered from cAMP and leucine prototrophs. The resulting DNA pools, enriched for the ability to confer cAMP-independent growth on yeast strain CY2827, were again introduced into CY2827 for selection of cAMP-independent clones whose plasmid DNA was sequenced and functionally analyzed. Finally, according to Protocol 4, CY2827 was transformed with libraries 241, 242, and 243, and the transformants selected on leucine-deficient medium containing cAMP. After 3 days at 30°C, the plates 5 were replica-printed to medium lacking both leucine and cAMP, followed by incubation for 4 days at 30°C Plasmid DNA was recovered from the cAMP prototrophs, amplified in bacteria, introduced into CY2827, again recovered from cAMP and leucine prototrophs, amplified in bacteria, and transformed into CY2827 for clonal selection and analysis. ι o Plasmids that confer cAMP-independent growth on CY2827 and that therefore encode constitutively active mutants of Type IV adenylyl cyclase are analyzed as follows:

1. CY2827 is transformed with a plasmid encoding an apparently constitutively active 15 ACIV, wild-type ACIV, or with empty vector, transformants are plated directly on media without leucine or cAMP, and the number of c AMP prototrophs are compared. The occurrence of cAMP prototrophs arising from transformation of CY2827 must be at least 1, 000-fold greater with candidate mutants than with wild-type ACIV.

20 2. CY2827 is transformed with a plasmid encoding an apparently constitutively active adenylyl cyclase or wild-type ACIV, and transformants are selected on media containing cAMP. Transformants are picked, expanded on cAMP -containing media, and tested by streaking on media without cAMP. Transformants carrying candidate mutants must grow on cAMP-deficient media while those harboring wild-type ACIV do not.

25

3. To ensure that the phenotype conferred by candidate mutants is not haploid-specific, the mutants must also confer cAMP-independent growth on CY7785 (genotype MATa/MATa cyrl ::ura3/cyrl ::ura3 caml/caml cam2/cam2 cam3/cam3 leu2-3/leu2-3 Ieu2-U2/Ieu2-U2 his3-532/his3-532 his4/his4 ura3/ura3 Iys2/lys2

3 o trpl ::L YS2/trpl::LYS2), a diploid derivative of C Y2827.

4. To ensure that the phenotype conferred is due to mutations within the coding region of the cyclase, plasmids constructed by subcloning the coding region of candidate mutants into a vector that has not been exposed to mutagenesis must be able to confer

35 the constitutively activated phenotype. 5. Discreet coding regions of the candidate mutants are cloned into wild-type backgrounds, and the resulting recombinants are tested for ability to confer the mutant phenotype. Specifically, Bgl II-to-Sfi I, Sfi 1-to-Stu I, and Stu I-to-Spe I fragments, covering codons 17 to 414, 414 to 924, and 924 through the stop codon, respectively, are subcloned into a wild-type background. In the case of all candidate mutants of ACIV studied, the mutant phenotype is observed to segregate with one of these fragments.

6. Regions of mutant cyclases that confer the constitutively active phenotype are sequenced. 0

Using this protocol a number of mutants have been isolated. For example* Tyr265His mutant from library 241, a Glu313Gly mutant from library 242, and Gly968Ser and Val388Ile mutants from library 243 have been isolated. All mutants lie in regions of cyclase that are highly conserved among different isoforms of the enzyme 5 and among different species. Tyr265His, Glu313Gly, and Val388Ile lie in the conserved region Cu, while Gly968Ser is in the conserved C2a domain. When the 9 cloned isoforms of mammalian adenylyl cyclase are aligned, Tyr265 of ACIV is conserved across 8 of the 9 isoforms. It is noteworthy that Tyr265 lies within a region corresponding to a synthetic 20 amino acid peptide that inhibits forskolin- and Gαs- 0 stimulated ACV activity (Kawabe, J., T. Ebina, S. Ismail, D.B. Kitchen, CJ. Homey, Y. Ishikawa. A novel peptide inhibitor of adenylyl cyclase. 1994. J. Biol. Chem 269:24906-24911). Glu313 is conserved across three of the nine isoforms, while His is present at the equivalent position among 5 isoforms and Lys is present in one isoform. Thus, this position is charged in all wild-type ACs. Val388 is invariant among all ACs. 5 Gly968, in contrast is not conserved among the isoforms: 3 have Asp, 2 have Lys, and Gin, Thr, and Asn are each represented once among the isoforms. Each of these mutant cyclases were expressed at levels equivalent to wild-type ACIV. Three other mutants Arg268Lys, GIy968Asp, Lys998Asn, showed elevated expressen when compared to wild-type cyclase. o When transiently expressed in HEK293 cells, the Tyr265His, Glu313Gly, Val

388Ile, and Gly 968Ser mutants result in levels of c AMP that are higher than wild-type ACIV, as measured both by SPA and a CRE-SEAP reporter construct.

These mutants are the first reported constitutively active mammalian adenylyl cyclases. These data indicate that Yeast that express mammalian adenylyl cyclases can 5 be used to identify mutant cyclases that exhibit Gαs-independent activity.

To confirm the Gαs-independent activity of mutants, membranes of yeast expressing the mutants are prepared for measurement of AC activity. The AC mutants are tested in three independent systems:

1. The mutant ACs are expressed in baculovirus- infected Sf9 cells for measuring membrane-associated cyclase activity. cDNA encoding each of the mutant genes are

5 transferred to pVL1393, and the resulting plasmids cotransfected into Sf9 cells with baculoviral DNA for isolation of recombinant viruses by plaque purification (Taussig, R., L.M. Qarmby, and A.G. Gilman. 1993. Regulation of purified type 1 and type II adenylylcyclasses by G protein βγ subunits. J. Biol. Chem. 268:9-12). Sf9 cells are infected with the recombinant baculovirus at 1 pfu/cell, and cells harvested for lysis by nitrogen o cavitation approximately 50 hours post-infection. Membranes are prepared and AC activity measured as described (Tang, W.-J., M. Stanzel, and A.G. Gilman. 1995. Truncation and alanine-scanning mutants of type I adenylyl cyclase. Biochemistry 34:14563- 14572). In addition to measuring basal AC activity, the effects of Gas, Gai and Gβγ on the activities of the mutants are examined. 5

2. Mutants are expressed in mammalian cells for testing cellular cAMP levels and membrane-associated cyclase activity. Genes encoding wild-type and mutant ACIVs are cloned into the eucaryotic expression vector pCEP4 (Invitrogen, San Diego, CA) for transient expression in both COS7 and HEK293 cells. Cellular cAMP levels are o measured by scintillation proximity assay, and adenylyl cyclase activity of membrane preparations measured directly. Those mutations that result in constitutive ACIV activity in mammalian cells are introduced at the equivalent position in other cyclase isoforms, which are then be expressed in mammalian cells to test the generality of the results obtained with ACIV. 5

3. The constitutively active mutations are produced in soluble adenylyl cyclases for expression in E. coli. Gilman and coworkers have developed a soluble, Gαs-responsive AC construct in which Cia of Type 1 AC is attached via a short peptide linker to C2a of Type II AC (Tang, W.J. and A.G. Gilman. 1995. Construction of a soluble adenylyl o cyclase activated by Gsα and forskolin. Science 268: 1769- 1772). The resulting construct can be expressed in E. coli for production of sufficient quantities of the enzyme for biochemical and crystallographic studies.

Constitutively active mutants that exhibit the greatest activity in the Sf9 and mammalian cell assays are subjected to another round of mutagenesis and selection for 5 more robust Gαs-independent growth. These more active mutants are further mutagenized and selected, leading to a lineage of related mutants which are both useful

predicted to be synthesized by the yeast expression vector is identical to trxA sequences present in Genbank with the following exceptions:

The second amino acid is predicted to be valine in the protein . This amino acid is actually encoded by one of the synthetic oligonucleotides designed for PCR amplification of the gene and its subsequent cloning into the expression plasmid. Accordingly, the second residue differs from that present in the authentic protein expressed in E coli, viz. leucine. The predicted C-terminal amino acid is valine, rather than the alanine reported in the Genbank sequences. This latter difference was observed in several of our PCR products, and presumably represents a polymorphism present among various E. coli strains.

The resulting trxA expression vectors (Cp3594, carrying HIS3; Cp3595, carrying

TRPI ; and Cp3960, carrying URA3) were then modified to permit insertion of peptide- encoding oligonucleotides and DNA fragments into the active site loop of trxA. Specifically the vectors were digested with Rsr II, which cuts in the middle of the active site loop, and the double-stranded oligonucleotide formed by annealing single-stranded oligonucleotides o284 (5'

GTCCGAGATCTCTTAAGGACGCGTTGCTAGCAGCATGCG) (SEQ ID NO: 15) and o285 (5' GACCGCATGCTGCTAGCAACGCGTCCTTAAGAGATCTCG) (SEQ ID NO: 16) were ligated into the Rsr II site of each vector. The plasmid products of these ligations (Cp3768 [carrying HIS3], Cp3770 [carrying TRPI], and Cp3962 [carrying URA3]) contain a multiple cloning site that permits the efficient insertion of DNA fragments into the active site loop of trxA.

Mapping the region of Gas that activates adenylyl cyclase.

A previous study attempted to define the region of Gas that activates mammalian adenylyl cyclase (Berlot, CH, HR Bourne. 1992. Identification of effector-activating residues of Gas. Cell 68:91 1-922). Both these approaches exploit the homologies between Gαi, the G protein that inhibits certain adenylyl cyclases, and Gas, by making various Gai-Gαs hybrid proteins and testing the chimeras for cyclase-stimulatory activity. The yeast trxA expression vector offers an alternative approach to mapping the region of Gas that activates adenylyl cyclase.

A DNA fragment containing the complete coding region of rat Gas is isolated by agarose gel electrophoresis following digestion of preparative quantities of a Gαs- containing plasmid with the appropriate restriction enzymes. Replicate samples of gel- purified, Gαs-containing fragment are subjected to DNase I digestion in the presence of 10 mM MnC-2 for various times as described by Maniatis (Sambrook J, EF Frisch, T Maniatis. 1989. Molecular Cloning, A Laboratory Manual. Cold Spring Harbor Laboratory Press, pl 3.28). DNase digestion is terminated by adding EDTA, and samples whose DNA has been degraded to an average size of about 100 nucleotides will be subjected to preparative gel electrophoresis to isolate all digestion products from approximately 50 to 150 nucleotides. The fragments are blunted-ended and then ligated to adapters for subsequent cloning into the multiple cloning site of a trxA expression vector (e.g.Cp3770). Electro-competent bacteria DHl 0B (Electromax DHl 0B, Gibco- BRL) are transformed by electroporation with the ligation products, and the plasmids are grown in the bacterial host overnight at 37C, followed by isolation of the plasmid DNA. This DNA are then used to transform yeast whose growth is dependent on an active rat type IV adenylyl cyclase (genotype a/a cdc35-l ura3 his 3 trpl leu2 CANs lys2::PGKp- ratACIV), an activity that is normally dependent on coexpressed Gas. Those - transformants that express Gαs-derived peptides in the context of trxA that can stimulate adenylyl cyclase will form colonies. Isolation of the trxA-encoding plasmids from tiiese colonies, followed by sequencing of the peptide inserts in trxA, enable mapping of tiie cyclase-stimulatory domain of Gas.

Identifying peptides in random libraries that stimulate adenylyl cyclase.

Completely or partially random oligonucleotides are cloned into the multiple cloning site of a trxA expression vector (Cp3768, Cp3770, or Cp3962), the resulting plasmid library is transformed into yeast whose growth is dependent on an active rat type IV adenylyl cyclase (genotype a/a cdc35-l ura3 his 3 trpl leu2 CANs lys2:;PGKp- ratACIV), and the trxA expression plasmid is isolated from growing colonies for determination of the peptide responsible for cyclase activity. Each peptide-encoding insert in trxA is subjected to partial randomization and these derivitive libraries will be subjected to selection for those peptides that confer more robust growth. Such peptide stimulators of adenylyl cyclase can then serve as models for medicinal chemical efforts to design and synthesize small organics that can activate adenylyl cyclase.

Example 4 Temperature Dependence of Growth Phenotypes of Yeast Expressing Mammalian Adenylyl Cyclases and the Use of Intragenic Suppressors in Genetic Analysis.

This example describes the isolation of mutant forms of Type IV adenylyl cyclase that contain both a constitutively active mutation (isolated and described previously in Example 2 of this application), and a second mutation that restores proper regulation by Gas to the mutant cyclase, ie. Gas stimulation of ACIV activity.

All drugs that modify the activity of a protein do so by changing a physical property of the protein. For example, competitive inhibitors prevent an enzyme from functioning by blocking access of the substrate to the enzyme. In the case of adenylyl cyclase, information is emerging that its activity is regulated by formation of a dimer that is composed of the two cytoplasmic domains. The two cytoplasmic domains have been shown to interact in vitro, and the extent of activation is correlated with the fraction of dimers formed in solution. One critical aspect of this model that relates to the regulation of adenylyl cyclase concerns the role of Gas. Addition of Gas to a solution of the free cytoplasmic domains of adenylyl cyclase increases the affinity of the domains for each other, and increases the activity of the solution at low concentrations of the cytoplasmic domains. This data indicates that Gas activates adenylyl cyclases by promoting dimer formation. Additional information concerning the role of dimer formation as the mechanism of cyclase activity comes from the recent structural analysis of adenylyl cyclase type II complexed with the small molecule activator forskolin. The catalytic site is formed by a pocket in the interface between the two cytoplasmic domains, and forskolin is tightly coordinated between the monomers. Thus, forskolin contributes to stability of the dimer. If regulation of cyclase activity is a direct consequence of regulation of dimer formation, then (a) understanding the regulation of dimer formation becomes the means by which an understanding of enzyme regulation may be based, and (b) molecules that regulate dimer formation will regulate cyclase activity.

The observations described in Example2 concerning the requirement of Gas for sufficient activity of ACIV to allow growth of a yeast strain that carries a deletion of the cyrl gene in the absence of exogenous cAMP, were made at 30C, the optimal temperature for yeast growth. However, it was observed that there is no requirement for Gas at 23C (ACIV itself is still required). As described below, this observation is critical for the effective design of a screen for the isolation of intragenic suppressors of the constitutive mutations in ACIV described in Example 2.

An intragenic suppressor is a second mutation in a gene that reverts the phenotype caused by the first mutation. If a mutation in a gene causes a loss of function, and a second mutation restores function, then the second mutation can be referred to as an intragenic suppressor. Intragenic suppressors can convey important information about both the structure and mechanism of a protein. A mutation that causes the protein to become activated may do so because it binds a substrate more tightly, and by extension, an intragenic suppressor may function by causing the protein to bind the 5 substrate less tightly, often for a distinct reason. Thus, an analysis of both mutations will contribute to a description of how the protein binds its substrate.

Regarding functional analysis of ACIV, intragenic suppressors can provide critical information about the regulation of cyclase activity. Ifthe mutations isolated in Example 2 are constitutive because they stabilize the dimer, then the isolation of o intragenic revertants can identify additional residues that are important for dimer formation. An intragenic suppressor for the constitutive mutations isolated in Example 2 would be a mutation that restores regulation of cyclase activity by Gas.

The recent results that suggest the regulation of adenylyl cyclase occurs through the regulation of dimer formation strongly suggests that modulation of cyclase activity s may be achieved explicitly by agents that directly stimulate or inhibit dimer formation. Analysis of both constitutive mutations and their intragenic suppressors can provide fundamental information about regions of ACIV that play a critical role in this These regions may then be developed into binding assays or other technologies that can be used to identify small molecule inhibitors or activators. The identification of a large o number of critical residues greatly increases the likelihood that a unique region of tiie target isoform of adenylyl cyclase may be identified, and therefore an drug may be developed that is isoform specific.

5 Intragenic suppressors are isolated as follows:

1. A plasmid that contains an activating mutation in the ACIV gene (described in Example2), is randomly mutagenized with hydroxylamine and transformed into Cadus yeast strain CY 2827. 0

2. 10,000 to 20,000 transformants of the strain carrying the mutated plasmid are plated at a density of 300 transformants per plate and allowed to grow at room temperature for 10 days on plates that lacked leucine and are not supplemented with exogenous cAMP. This allows for selection of plasmids that have not incorporated a mutation in the ACIV s gene that results in a complete loss of function. - 112 -

3. After 10 days, the transformants are replica printed onto new plates and the new plates are incubated at 30C. This allows for the identification of colonies that contain plasmids which encode mutant adenylyl cyclases that can not function constitutively to be identified as patches of non-growing cells when compared to the corresponding s colony that had grown at 23C.

4. Colonies identified as growing at 23°C but not at 30°C were recovered and the plasmids contained in them were considered as candidates for intragenic suppressors.

o 5. Plasmids are recovered and retransformed to check that the phenotypes were plasmid dependent and therefore likely to result from mutations in the ACIV gene.

6. Complete intragenic suppression is determined by cotransforming the mutated ACIV containing plasmid with either Cadus plasmid 1015 (a plasmid that contains the URA3 s gene), or Cadus plasmid 1069 (which contains the URA3 gene and the mammalian Gas gene, transcribed from a yeast promoter), and measuring growth at 23°C and 30°C on plates that lack leucine, uracil or exogenous cAMP. Complete intragenic suppression is scored as an ACIV containing plasmid that can confer growth at 23°C regardless of the presence of Gas, but can only grow at 30°C when Gas is coexpressed. 0

7 The ACIV gene from each candidate plasmid is isolated, recloned to determine the restriction fragment that confers intragenic suppression (as described in Example 2), and sequenced.

s Using this protocol, Cadus plasmid 3190 (LEU22mu-ori REP3 AmpR PGKp-

ACIV-31C G968S) has been mutagenized to identify fifteen plasmids that met the requirements of Steps 1 through 5. Seven of these fifteen fulfulled Step 6. The remaining plasmids didnot grow at 30°C, even when mammalian Gas was coexpressed. These plasmids are candidates for mutations that affect the catalytic activity of ACIV, 0 but not its regulation. Of the plasmids that are candidates for intragenic suppressors, four have been restriction mapped and sequenced to the point where it can be determined that the mutation that confers intragenic suppression lies in the Cia region. Mutations in Cadus plasmid 3118 (LEU2 2mu-ori REP3 AmpR PGKp-ACIV-13N Y265H) has also been screened in this protocol and 24 plasmids recovered through Step 5. Six of these 5 have fulfilled Step 6. The mechanism by which these intragenic suppressors and catalytic mutants affect adenylyl cyclase activity is addressed in several experimental protocols including those described in Example 2 and methods for quantitating the extent of dimer formation including in vitro binding assays (such as described by Gorman et al. 1996. J. Biol. Chem. 27:6713) and yeast dihybrid analysis (such as described in U.S. Patents 5283,173 and 5,468,614 by Fields et al., and U.S. Patent 5,580,736 by Brent et al.).

Other References cited herein

Aiba H., Mori K., Tanaka M., Ooi T., Roy A., Danchin A (1984) The complete nucleotide sequence of the adenylyl cyclase gene of Escherichia coli. Nucl. acids Res. 12, 9427-9440.

Alexandre S., Paindavoine P., Tebabi P., Pays A., Halleux S., Steinert M., Pays E. (1990) Differential expression of a family of putative adenylate/guanylate cyclase genes in Trypanosoma brucei. Mol. Biochem. Parasitol. 43, 279-288. Arkinstall, S., Payton, M. and Mondrell, K., (1995) Activation of Phospholipase

C gamma in Schizosaccharomyces pombe. Mol. Cell. Biol. 15, 3, 1451 :1458.

Bakalyar H.A. and Reed R.R. (1990) Identification of a specialized adenylyl cyclase that may mediate odorant detection. Science 250, 1403-1406.

Beals, C. R., Wilson, C. B. and Perlmutter, R. M. (1987) A small multigene family encodes Gi signal-transduction proteins. Proc. Natl. Acad. Sci. U.S. A 84, 7886- 7890.

Bennetzen J.L., and Hall B.D. (1982) Codon selection in yeast. J. Biol. Chem. 257, 3026-3031.

Boutelet F. and Hilger F. Tenth International Conference on Yeast Genetics and Molecular Biology, Sept. 8-12, 1980, Louvain-la-Neuve, Belgium, p. 177 (abstract).

Bourne H.R., Sanders D.A., McCormick F. (1991) THe GTPase superfamily conserved structure and molecular mechanism. Nature 349, 117-127.

Bray, P., Carter, A., Guo, V., Puckett, C, Kamholz, J., Spiegel, A., and Nirenberg, M. (1987) Human cDNA clones for an α subunit of Gi signal-transduction protein. Proc. Natl. Acad. Sci. USA 84, 5115-5119.

Bray, P., Carter, A., Simons, C, Guo, V., Puckett, C, Kamholz, J., Spiegel, A., Nirenberg, M., (1986) Human cDNA clones for four species of Gas signal transduction protein. Proc. Natl. Acad. Sci. U.S.A. 83, 8893-8897.

Broach J.R., Li Y.-Y., Wu L.-C.C, Jayarum M. (1983) in "Experimental Manipulation of Gene Expression" (M. Inouye, ed.) pp. 83-1 17. Academic Press, New York.

Broek D., Samiy N., Fasano O., Fujiyama A., Tamanoi F., Northup J., Wigler M. (1985) Differential activation of yeast adenylyl cyclase by wild-type and mutant RAS proteins. Cell 41, 763-769. Cali J.J., Balcueva E. A., Rybalkin I., Robishaw J.D. (1992) Selective tissue distribution of G protein γ subunits, including a new form of the γ subunits identified by cDNA cloning. J. Biol. Chem. 267, 24023-24027.

Casperson G.F., Walker N., Bourne H.R. (1985) Isolation of the gene encoding adenylate cyclase in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 82, 5060- 5063. Chinkers M. and Garbers D.L. ( 1991 ) Signal transduction by guanyl cyclases.

Ann. Rev. Biochem. 60, 553-575.

Codina, J., Stengel, D., Woo, S. L. C, and Birnbaumer, L. (1986) Beta-subunits of the human liver Gs/Gi signal-transducing proteins and those of bovine retinal rod cell transducin are identical. FEBS Lett 207, 187-192. Colicelli J., Field J., BaJlester R., Chester N., Young D., Wigler M. (1990).

Mutational mapping of RAS-responsive domains of the S. cerevisiae adenylyl cyclase. Mol. Cell. Biol. 10, 2539-2543. _m \

Conklin B.R., Farfel Z., Lustig K.D., Julius D. and Bourne H.R. (1993) Substitution of three amino acids switches receptor specificity of Gαq to that of Gαi. Nature 363, 274-276.

Cull MG; Miller JF; Schatz PJ (1992) Screening for receptor ligands using large libraries of peptides linked to the C terminus of the lac repressor. Proc Natl Acad Sci U S A 89, 1865-1869.

Cwirla S. E., Peters E. A., Barrett R. W., Dower W. J. (1990) Peptides on phage: A vast library of peptides for identifying ligands. Proc. Natl. Acad. Sci. U.S.A. 87, 6378- 6382.

Devlin J. J., Panganiban L. C, Devlin P. E. (1990) Random peptide libraries: A source of specific binding molecules. Science 249, 404-406.

Devreotes P. (1989) Dictyostelium discoideum: A model system for cell-cell interactions in development. Science 245, 1054- 1058.

Didsbury, J. R. and Snyderman, R. (1987) Molecular cloning of a new human G protein: evidence for two Giα-like protein families. FEBS Lett. 219, 259-263.

Didsbury, J. R., Ho, Y.-S., and Snyderman, R. (1987) Human Gi protein α subunit: deduction of amino acid structure from a cloned cDNA. FEBS Lett. 211, 160- 164.

Dietzel C, Kurjan J. (1987) The Yeast SCG1 Gene: A Gα-like protein implicated in the a-and α-factor response pathway. Cell 50, 1001-1010.

Dratz E. A., Furstenau J.E., Lambert C.G., Thireault D.L., Rarick H., Scheners T., Pakhlevaniants S., Hamm H.E. (1993) NMR structure of a receptor-bound G-protein peptide. Nature 363, 276-281.

Feinstein P.G., Schrader K.A., Bakalyar H.A., Tang W.J., Koprinski J., Gilman A.G., Reed R.R. (1991) Molecular cloning and characterization of a Ca2+/calmodulin- insensitive adenylyl cyclase from rat brain. PNAS 88, 10173-10177.

Felici F., Castagnoli L., Musacchio A., Jappelli R., Cesareni G. (1991) Selection of antibody ligands from a large library of oligopeptides expressed on a multivalent exposition vector. J. Mol. Biol. 222, 301 -310.

Field J., Xu H.-P., Michaeli T., Ballester R., Sass P., Wigler M., Colicelli J. (1990) Mutations of the adenylyl cyclase gene that blocks RAS function in Saccharomyces cerevisiae. Science 247, 464-467.

Fisher, K. J. and Aronson, N. N. Jr. (1992) Characterization of the cDNA and genomic sequence of a G protein γ subunit (γ5). Mol. Cell. Biol. 12, 1585-1591.

Florent I., Raibaud A., Eisen H. (1991) A family of genes related to a new expression site-associated gene in Trypanosoma equiperdum. Mol. Cell. Biol. 11, 2180- 2188.

Fodor S. P .A., Read J. L., Pirrung M. C, Stayer L., Lu A. T., Solas D. (1991) Light-directed, spatially addressable parallel chemical synthesis. Science 249, 386-390.

Fong, H. K. W., Amatruda, T. T., Ill, Birren, B. W., and Simon, M. I. (1987). Distinct forms of the β subunit of GTP-binding regulatory proteins identified by molecular cloning. Proc. Natl. Acad. Sci. U.S.A. 84, 3792-3796.

Fong, H. K., Hurley, J. B., Hopkins, R. S., Miake-Lye, R., Johnson, M. S., Doolittie, R. F., and Simon, M. I. (1986) Repetitive segmental structure of the transducin beta subunit: Homology with the CDC4 gene and identification of related mRNAs. Proc. Natl. Acad. Sci. U.S.A. 83, 2162-2166.

Gao B. and Gilman A.G. (1991) Cloning and expression of a widely distributed (type IV) adenylyl cyclase. PNAS 88, 10178-10182 Gao, B., Gilman, A. G., and Robishaw, J. D. (1987). A second form of the β subunit of signal-transducing G proteins. Proc. Natl. Acad. Sci. U.S.A. 84, 6122-6125.

Gautam, N., Northup, J., Simon, M. I. (1990) G protein diversity is increased by associations with a variety of γ subunits. Proc. Natl. Acad. Sci. U.S.A. 87, 7973-7977.

Gilman A.G. (1984) G proteins and dual control of adenylate cyclase. Cell 36, 577-579.

Gilman A. (1984) G proteins and dual control of adenylyl cyclase. Cell 36, 577- 579.

Hadwiger J.A., Wilkie T.M., Stratmann M., Firtel R.A. (1991) Identification of Dictyostelium Gα genes expressed during multicellular development. Proc. Natl. Acad. Sci. USA 88, 8213-8217.

Harris, B. A., Robishaw, J. D., Mumby, S.M., and Gilman, A. G. (1985) Molecular cloning of complementary DNA for the alpha subunit of the G protein that stimulates adenylate cyclase. Science 229, 1274-1277.

Hashimoto C, Hudson K.L., Anderson K.V. (1988) The Toll gene of Drosophila, required for dorsal-ventral embryonic polarity, appears to encode a transmembrane protein. Cell 52, 269-279.

Hoekma A., Kastelein R.A., Vasser M., de Boer H.A. (1987) Codon replacement in the PGKl gene of Saccharomyces cerevisiae: Experimental approach to study the role of biased codon usage in gene expression. Mol. cell. Biol. 7, 2914-2924.

Holbrook S. and Kim S.-H. (1989) Molecular model of the G protein α subunit based on the crystal structure of the HRAS protein. Proc. Natl. Acad. Sci. USA 86, 1751-1755.

Houghten R.A., Pinilla C, Blondelle S.E., Appel J.R., Dooley C.T., Cuervo J.H.

(1991) Generation and use of synthetic peptide combinatorial libraries for basic research and drug discovery. Nature 354, 84-86. Hurley, J. B., Fong, H. K, Teplow, D. B., Dreyer, W. J., Simon, M. I. (1984)

Isolation and characterization of a cDNA clone for the γ subunit of bovine retinal " transducin. Proc. Natl. Acad. Sci. U.S.A. 81, 6948-6952.

Iftiguez-Lluhi J.A., Simon, M.I., Robishaw J.D., Gilman A.G. (1992) G protein β γ subunits synthesized in Sf9 cells. J. Biol. Chem. 267, 23409-23417. Ishikawa Y., Katsushika S., Chen L., Halnoon N.J., Kawabe J., Homey C J.

(1992) Isolation and characterization of a novel cardiac adenylyl cyclase cDNA. J. Biol. Chem.267,13553-13557.

Itoh, H., Kozasa, T., Nagata, S., Nakamura, S. I., Katada, T., Ui, M., Iwai, S., Ohtsuka, E., Kawasaki, H., Suzuki, K., and Kaziro, Y. ( 1986) Molecular cloning and sequence determination of cDN As for CE ' subunits of the guanine nucleotide-binding proteins Gs, Gi, and Go from brain. Proc. Nati. Acad. Sci. U.S.A. 83, 3776-3780.

Itoh, H., Toyama, R., Kozasa, T., Tsukamoto, T., Matsuoka, M., Kaziro, Y. (1988) Presence of three distinct molecular species of Gi protein α subunit. Structure of rat cDNAs and human genomic DNAs. J. Biol. Chem. 263, 6656-6664.

Iyengar R. (1993) Molecular and functional diversity of mammalian Gs- stimulated adenylyl cyclases. FASEB J. 7, 768-775.

Jacobowitz O., Chen J., Premont R.T., Iyengar R. (1993) Stimulation of specific types of adenylyl cyclases by phorbol ester treatment. J. Biol. Chem. 268, 3829-3832. Johnson J.A., Friedman J., Halligan R.D., Bimbaumer M, Clark R.D. (199))

Sensitization of adenylyl cyclase by P2 purinergic and M5 muscarinic receptors in L

ip, /

- 1 18 -

cells. J. Pharmacol. Exp. Ther. 39, 539-546.

Jones, D.T. and Reed, R.R. (1987) Molecular cloning of five GTP binding protein cDNA species from rat olfactory neuroepithelium. J. Biol. Chem. 262, 14241- 14249. Jurnak F. ( 1985) Structure of the GDP domain of EF-Tu and location of the amino acids homologous to ras oncogene proteins. Science 230, 32-36.

Kaiser C.A., Preuss D., Grisafi P. and Botstein D. (1987) Many random sequences functionally replace the secretion signal sequence of yeast invertase. Science 235, 312-317. Kang Y.-S., Kane J., Kurjan J., Stadel J.M., Tipper D.J. (1990) Effects of expression of mammalian Gα and hybrid mammalian-yeast Gα proteins on the yeast pheromone response signal transduction pathway. Mol. Cell. Biol. 10, 2582-2590.

Katsushika S., Chen L., Kawabe J.-I., Nilakantan R., Halnon N.J., Homey C.J., Ishikawa Y. (1992) Cloning and characterization of a sixth adenylyl cyclase isoform: Types V and VI constitute a subgroup within the mammalian adenylyl cyclase family. Proc. Natl. acad. sci. USA 89, 8774-8778.

Katz, A., Wu, D., and Simon, M. I. (1992) Subunits βγ of heterotrimeric G protein activate β2 isoform of phospholipase C. Nature 360, 686-689. Kim, S., Ang, S.-L., Bloch, D. B., Bloch, K. D., Kawahara, Y., Tolman, C, Lee,

R., Seidman, J. G., and Neer, E. J. (1988) Identification of cDNA encoding an additional α subunit of a human GTP-binding protein: Expression of three αi subtypes in human tissues and cell lines. Proc. Natl. Acad. Sci. U. S. A. 85, 4153-4157.

Kozasa, T., Itoh, H., Tsukamoto, T., Kaziro, Y., (1988). Isolation and characterization of human Gsα gene. Proc. Natl. Acad. Sci. U.S.A. 85: 2081-2085.

Kesbeke F., Snaar-Jagalska E., Van Haastert P.J.M., (1988) Signal transduction in Dictyostelium Frd A mutants with a defective interaction between surface cAMP receptor and a GTP-binding regulatory protein. J.Cell. Biol. 107, 521-528.

Koesling D., Boheme E., Schultz G. (1991) Guanylyl cyclases, a growing family of signal-transducing enzymes. FASEB J. 5, 2785-2791.

Krupinski J., Coussen F., Bakalyar H.A., Tang W.-J., Feinstein P.G., Orth K., Slaughter C, Reed R.R., Gilman A.G. (1989) Adenylyl cyclase amino acid sequence: possible channel- or transporter-like structure. Science 244, 1558-1564.

Krupinski J., Lehman P.C., Frankenfield CD., Zwaagstra J.C, Watson P.A. (1992) Molecular diversity in adenylyl cyclase family evidence for eight forms of the enzyme and cloning of type 6. J. Biol. Chem. 267, 24858-25862. LaCour T.F.M., Nyborg J., Thirup S., Clark B.F.C (1985) Structural details of the binding of guanosine diphosphate to elongation factor Tu from E. coli as studied by X-ray crystallography. EMBO J. 4, 2385-2388.

Lam K.S., Salmon S.E., Hersh E.M., Hruby V.J., Kazmierski W.M., Knapp R.J. s (1991 ) A new type of synthetic peptide library for identifying ligand-binding activity. Nature 354, 82-84.

Lemire, B. D., Fankhauser, C, Baker, A., Schatz, G. (1989) The mitochondria! targeting function of randomly generated peptide sequences correlates with predicted I o helical amphiphilicity . J. Biol. Chem. 264, 20206-20215.

Levine, M. A., Smallwood, P. M., Moen, P. T. Jr., Helman, L. J., Ann, T. G. (1990) Molecular cloning of the β3 subunit, a third form of the G protein β-subunit polypeptide. Proc. Natl. Acad. Sci. U.S.A. 87, 2329-2333.

Levin L.R., Han P.-Y., Hwang P.M., Feinstein P.G., Davis R.L., Reed R.R. is (1992) The Drosophila learning and memory gene rutabaga encodes a Ca2+/calmodulin responsive adenylyl cyclase. Cell 68, 479-489.

Lopez J.A., Chung D.W., Fujikawa K, Hagen F.S., Davie E.W., Roth GJ. (1988) The α and β chains of human glycoprotein lb are both transmembrane proteins containing a leucine-rich amino acid sequence. Proc. Natl. Acad. Sci. USA 85, 2135- 20 2139.

Markby D. W., Onrust R., Bourne H.R. (1993) Separate GTP binding and GTPase activating domains of a Gα subunit. Science 262, 1895-1901.

Matsumoto K., Uno I., Toh-E. A., Ishikawa T., Osbima Y. (1982) Cyclic AMP may not be involved in catabolite repression in Saccharomyces cerevisiae: Evidence 25 from mutants capable of utilizing it as an adenine source. J. Bacteriol. 150, 277-285.

Mattera, R., Codina, J., Crozat, A., Kidd, V., Woo, S.L.C, and Bimbaumer L. (1986) Identification by molecular cloning of two forms of the α-subunit of tiie human liver stimulatory (Gs) component of adenylate cyclase. FEBS Lett. 206, 36-41.

Mercken L., Moras V., Tocque B., Mayaux J.F. (1990) The cDNA sequence of 30 the alpha subunit of the Chinese hamster adenylate cyclase-stimulatory G-protein. Nuc. Acids Res. 18, 662.

Michel T., Winslow J.W., Smith J.A., Seidman J.G., Neer EJ. (1986) Molecular cloning and characterization of cDNA encoding the GTP-binding protein alpha i and identification of a related protein alpha h. Proc. Natl. Acad. Sci. U. S. A. 83, 7663-7667.

35

Mitts M.R., Grant D.B., Heideman W. (1990) Adenylate cyclase in O 97/40170 PC17US97/06711

- 120 -

saccharomyces cerevisiae is a peripheral membrane protein. Mol. Cell. Biol. 10, 3873- 3883.

Nakane M., Arai K., Saheki S., Kuno T., Buechler W., Murad F. (1990) Molecular cloning and expression of cDNAs coding for soluble guanylate cyclase from rat lung. J. Biol. Chem. 265, 479-489.

Noel J.P., Hamm H.E., Sigler, P.B. (1993) The 2.2 A crystal structure of transducin-alpha complexed with GTP-gamma-S. Nature 366, 654-663.

Northup J.K., Sternweis P.C, Smigel M.D., Schleifer L.S., Ross E.M., Gilman A.G. (1980) Purification of the regulatory component of adenylate cyclase. Proc. Natl. Acad. Sci. USA 77, 6516-6520.

Nukada T., Tanabe T., Takahashi H., Noda M. et al. (1986) Primary structure of the alpha-subunit of bovine adenylate cyclase-stimulatory G-protein deduced from the cDNA sequence. FEBS Lett. 195, 220-224.

Paindavoine P., Rolin S., Van AsselS., Geuskens M et al. (1992) A gene from the variant surface glycoprotein expression site encodes one of several transmembrane adenylyl cyclases located on the flagellum of Trypanosoma brucei. Mol. Cell. Biol. 12, 1218-1225.

Parma J., Stengel D., Gannage M.-H., Poyard M., Barouki R., Hanoune J. (1991) Sequence of a human brain adenylyl cyclase partial cDNA. Evidence for a consensus cyclase specific domain. Biochem. Biophys. Res. Comm. 179, 455-462.

Pitt G.S., Milona N., Borleis J., Lin K.C, Reed R.R., Devreotes P.N. (1992) Structurally distinct and stage-specific adenylyl cyclase genes play different roles in Dictyostelium development. Cell 69, 305-315.

Premont R.T., Chen J., Ma H.-W., Ponnapalli M., Iyengar R. (1992) Two members of a widely expressed subfamily of hormone stimulated adenylyl cyclases. PNAS 89, 9808-9813.

Premont R.T., Jacobowitz O., Iyengar R. (1992) Lowered responsiveness of the catalyst of adenylyl cyclase to stimulation by Gs in heterologous desensitization: a role for cAMP-dependent phosphorylation. Endocrinology 131, 2774-2783. Pronin, A. N. and Gautam, N. (1992) Interaction between G-protein β and γ subunit types is selective. Proc. Natl. Acad. Sci. U. S. A. 89, 6220-6224.

Pupillo M., Insall R., Pitt G.S., Devreotes P.N. (1992) Multiple cyclic AMP receptors are linked to adenylyl cyclase in Dictyostelium. Molec. Biol. of the Cell 3, 1229-1234. Rail T. and Harris B.A. (1987) Identification of the lesion in the stimulatory

GTP-binding protein of the uncoupled S49 lymphoma. FEBS Lett. 224, 365-371. Robishaw, J. D., Kalman, V. K., Moomaw, C R., Slaughter, C A. (1989) Existence of two γ subunits of the G proteins in brain. J. Biol. Chem. 264, 15758rI5761.

Robishaw, J.D., Russell, D. W., Harris, B. A. Smigel, M. D. and Gilman, A. G- (1986a) Deduced primary structure of the α subunit of the GTP-binding stimulatory protein of adenylate cyclase. Proc. Natl. Acad. Sci. U.S.A. 83, 1251-1255.

Ross, E.M., and Gilman, A.G. (1977). Resolution of some components of adenylate cyclase necessary for catalytic activity. J. Biol. Chem. 252, 6966-6969.

Ross D.T., Raibaud A., Florent LC, Sather S., Gross M.K., Storm D.R., Eisen H. (1991 ) The trypanosome VSG expression site encodes adenylyl cyclase and a leucine- rich putative regulatory gene. EMBO J. 10, 2047-2053.

Ruth J., Hirt H., Schweyen RJ. (1992) The cauliflower mosaic virus 35S promoter is regulated by cAMP in Saccharomyces cerevisiae. Mol. Gen. Genet. 235, 365-372.

Schmidt C. J., Thomas, T. C, Levine, M. A., Neer, E. J. (1992) Specificity of G protein β subunit and γ subunit interactions. J. Biol. Chem. 267, 13807-13810.

Scott J. K. and Smith G. P. (1990) Searching for peptide ligands with an epitope library. Science 249, 386-390.

Sharp P.M., Tuohy T.M.F., Mosurski K.R. (1986) Codon usage in yeast: cluster analysis clearly differentiates highly and lowly expressed genes. Nuc. Acids Res. 14, 5125-5143.

Shilo V., Simchen G., Shilo B. (1978) Initiation of meiosis in cell cycle initiation mutants of Saccharomyces cerevisiae. Exp. Cell Res. 112, 241-248.

Smith R.A., Sisk R., Lockhart P., Mathewes S., Gilbert T., Walker K., Piggot J. (1993) Isolation of glucagon antagonists by random molecular mutagenesis and screening. Mol. Pharmacol. 43, 741-748.

Sugimoto K., Nukada, T., Tanabe, T., Takahashi, H., Noda, M., Minamino, N., Kangawa, K., Matsuo, H., Hirose, T., Inayama, S., Numa, S. (1985) Primary structure of the β-subunit of bovine transducin deduced from the cDNA sequence. FEBS Lett 191, 235-240.

Suki, W. N., Abramowitz, J., Mattera, R., Codina, J. and Bimbaumer, L. (1988) The human genome encodes at least three non-allelic G proteins with αi-type subunits. FEBS Lett. 220, 187-192.

Sullivan K.A., Liao Y.C, Alborzi A., Beiderman B. et al. (1986) Inhibitory and stimulatory G proteins of adenylate cyclase: cDN A and amino acid sequences of tite alpha chains. Proc. Nati. Acad. Sci. U. S. A. 83, 6687-6691. Sullivan, K. A., Miller, R. T., Masters, S. B., Biederman, B., Heideman, W., and Bourne, H. R. (1987). Identification of receptor contact site involved in receptor-G protein coupling. Nature 330, 758-760.

Takahashi N., Takahashi Y., Putnam F.W. (1985) Periodicity of leucine and 5 tandem repetition of a 24-amino acid segment in the primary structure of leucine-rich α 2-glycoprotein of human serum. Proc. Natl. Acad. Sci. USA 82, 1906-1910.

Tang W.-J. and Gilman A.G. (1992) Adenylyl cyclases. Cell 70, 869-872. Tang W.-J., Iftiguez-Lluhi J.A., Mumby S.M., Gilman A.G. (1992) Regulation of o mammalian adenylyl cyclases by G protein α and βγ subunits. Cold Spring Harbor Symp. Quant. Biol. 57, 135-144.

Tang W.-J., Krupinski J., Gilman A.G. (1991) Expression and characterization of calmodulin activated adenylyl cyclase. J. Biol. Chem. 266, 8595-8603.

Taussig R., Iftiguez-Lluhi J.A., Gilman A.G. (1993) Inhibition of adenylyl s cyclase by Giα. Science 261 , 218-221.

Taussig R., Iftiguez-Lluhi J.A., Gilman A.G. (1993) Inhibition of adenylyl cyclase by Gαi. Science 261, 218-221.

Taussig R., Quarmby L.M., Gilman A.G. (1993) Regulation of purified type 1 and 2 adenylyl cyclases by G protein βγ subunits. J. Biol. Chem. 268, 9-12. o Taussig R., Tang W.J., Helper J.R., Gilman A.G. ( 1994) Distinct patterns of bidirectional regulation of mammalian adenylyl cyclases. J. Biol. Chem. 269, 6093- 6100.

Toda T., Uno I., Ishikawa T., Powers S., Kataoka T., Broek D., Cameron S., Broach J., Matsumoto K., and Wigler M. (1985) In yeast, RAS proteins are controlling s elements of adenylyl cyclase. Cell 40, 27-36. von Weizsacker, E., Strathmann, M. P., and Simon, M. I. (1992) Diversity among the beta subunits of heterotrimeric GTP-binding proteins: characterization of a novel beta-subunit cDNA. Biochem. Biophys. Res. Commun. 183, 350-356. Weinstein, L.S., Spiegel, A. M., Carter, A. D. (1988) Cloning and 0 characterization of the human gene for the α subunit of Gi2, a GTP-binding signal transduction protein. FEBS Lett. 232, 333-340.

Whiteway M., Hougan L., Dignard D., Thomas D.Y., Bell L., Saari G.C, Grant F.J., O'Hara P., MacKay V.L. (1989) The STE4 and STE18 genes of yeast encode potential β and γ subunits of the mating factor receptor-coupled G protein. Cell 56, 467- 5 477.

Wu L. and Devreotes P.N. (1991) Dictyostelium transiently expresses eight distinct G W

- 123 -

protein α-subunits during its developmental program. Biochem. Biophys. Res. Commun. 179, 1 141-1147.

Yatomi Y., Arata Y., Tada S., Kume S., Ui M. (1992) Phosphorylation of tiie inhibitory guanine-nucleotide-binding protein as a possible mechanism of inhibition by protein kinase C of agonist-induced Ca2+ mobilization in human platelet. Eur. J. Biochem. 205, 1003-1009.

Yoshimasa T., Sibley D.R., Bouvier M., Lefkowitz R.J., Caron M.G. (1991) Cross-talk between cellular signalling pathways suggested by phorbolester induced adenylyl cyclase phosphorylation. Nature 327, 67-70. i Yoshimura M. and Cooper D.M.F. ( 1992) Cloning and expression of a Ca2+ inhibitable adenylyl cyclase from NCB-20 cells. PNAS 89, 6716-6720

Yoshimura M. and Cooper D.M.F. (1993) Type specific stimulation of adenylyl cyclase by protein kinase C J. Biol. Chem. 268, 4604-4607.

Additional References

Anderson M.P., Gregory R.J., Thompson S., Souza D.W., Paul.S. et al. (1991) Demonstration that CFTR is a chloride channel by alteration of its anion selectivity. Science 253, 202-205.

Artemeyev N.O., Rarick H.M., Mills J.S., Skiba N.P., and Hamm H.E. (1992) Sites of interaction between rod G-protein α-subunit and cGMP-phosphodiesterase γ- subunit J. Biol. Chem. 267, 25067-25072.

Banerjee S., Anderson G.D., Luthra H.S., David CS. (1989) Influence of complement C5 and V beta T cell receptor mutations on susceptibility to collagen- induced arthritis in mice. J. Immunol. 142: 2237-2243. Barr P.J., Mason O.B., Landsberg K.E., Wong P.A. et al. (1991) cDNA and gene structure for a human subtilisin-like protease with cleavage specificity for paired basic amino acid residues. DNA Cell Biol. 10, 319-328.

Benjannet S., Rondeau N., Day R., Chretien M., Seidah N.G. (1991) PCI and PC2 are proprotein convertases capable of cleaving proopiomelanacortin at distinct pairs of basic residues. Proc. Nati. Acad. Sci. USA 88:3564-3568.

Bianchi A.B., Fischer S.M., Robles A.L, Rinchik E.M., Conti CJ. (1993) Overexpression of cy clin Dl in mouse skin carcinogenesis. Oncogene 8, in press. Birkenbach M., Josefsen K., Yalamanchili R., Lenoir G., Kieff E. (1993) Epstein-Barr Virus-induced genes: First lymphocyte-specific G protein-coupled peptide receptors J. Virol. 67, 2209-2220.

Buratowski S., Hahn S., Sharp P. A., Guarente L. (1988) Function of a yeast TATA element-binding protein in a mammalian transcription system. Nature 334, 37.

Burkholder A.C. and Hartwell L.H. (1985) The yeast α-factor receptor: Structural properties deduced from the sequence of the STE2 gene. Nuc. Acids Res. 13, 8463. Cavallini B., Huet J., Plassat, J.-L., Sentenac A., Egly J.-M., Chambon P. (1988)

A yeast activity can substitute for the HeLa cell TATA box factor. Nature 334, 77.

Chen WJ, Andres D.A., Goldstein J.L., Brown M.S. (1991) Cloning and expression of a cDNA encoding the subunit of rat p21ras protein famesyltransferase. Cell 66:327-334. Choi K, Chen C.-J., Kriegler M., Roninson LB. (1988) An altered pattern of cross-resistance in multidmg resistant human cells results from spontaneous mutation in the mdrl (P-glycoprotein) gene. Cell 53, 519-529.

Clark KL, Dignard D., Thomas DY, Whiteway M. (1993) Interactions among the subunits of the G proteins involved in Saccharomyces cerevisiae mating. Mol. Cell. Biol. 13:1-8.

Coleman D.E., Berghuis A.M., Lee E., Linder M.E., Gilman A.G., Sprang S.R. (1994) Structures of Active Conformations of Giαl and the Mechanism of GTP Hydrolysis. Science 265:1405-1412. Conklin, B.R., Farfel, Z., Lustig, K.D., Julius, D., and H.R. Blurne (1993)

Substitution of three amino acids switches receptor specificity of Gqα to that of Giα. Nature 363,274-276.

Crawford M.H., Graver F.L., Kolb W.P., McMahan CA. et al. (1988) Complement and neutrophil activation in the pathogenesis of ischemic myocardial injury. Circulation 78: 1449-1458.

Crews CM., Allessandrini A., Erikson R.L. (1992) The primary structure of MEK, a protein kinase that phophorylates the ERK gene product. Science 258, 478-480.

Cross F. (1988) DAF1, a mutant gene affecting size control, pheromone arrest, and cell cycle kinetics of Saccharomyces cerevisiae. Mol. Cell. Biol. 8, 4675-4684. Dassa E. (1990) Cellular localization of the MALg protein from the maltose transport system in Escherichia coli K12. Mol. Gen. Genet. 222:33-36.

Dassa E. and Hofhung M. (1985) Sequence of gene malG in E. coli K12: homologies between integral membrane components from binding protein-dependent transport systems. EMBO J. 4:2287-2293. Dmochowska A., Dignard D., Henning D., Thomas D. Y., Bussey H. ( 1987)

Yeast KEX1 gene encodes a putative protease with a carboxypeptidase B-like function invoived in killer toxin and α-factor precursor processing. Cell 50, 573.

Duchateau J., Haas M., Schreyen H., Radoux L. et al. (1984) Complement activation in patients at risk of developing the adult respiratory distress syndrome. Am. Rev. Respir. Dis. 130: 1058-1064. s Elirmann M., Boyd D., Beckwith J. ( 1990) Genetic analysis of membrane protein topology by a sandwich gene fusion approach. Proc. Natl. Acad. Sci. 87, 7574-7578.

Elledge SJ. and Spottswood M.R. (1991) A new human p34 protein kinase, CDK2, identified by complementation of a cdc28 mutation in Saccharomyces cerevisiae, is a homolog of Xenopus Egl. EMBO J. 10, 2653-2659. o Emter O., Mechler B., Achstetter T., Muller H., Wolf D.H. (1983) Yeast pheromone α-factor is synthesized as a high molecular weight precursor. Biochem. Biophys. Res. Commun. 116, 822-829.

Endicott J.A., Sarangi F., Ling V. (1991) Complete cDNA sequences encoding the Chinese hamster P-glycoprotein gene family. DNA Sequence 2, 89-101. 5 Engleman D. A., Steitz T. A., Goldman A. ( 1986) Identifying nonpolar transbilayer helices in amino acid sequences of membrane proteins. Ann. Rev. Biophys. Chem. 15, 321-353.

Etienne G., Armau E., Tiraby G. (1990) A screening method for antifungal substances using Saccharomyces cerevisiae strains resistant to polyene macrolides. J. 0 Antibiot.43, 199-206.

Fikes J.D., Becker D.M., Winston F„ Guarente L. (1990) Striking conservation of TFIID in Schizosaccaharomyces pombe and Saccharomyces cerevisiae. Nature 346, 291.

Franke A.E., Andrews G.C, Stimler-Gerard N., Gerard C J. and Showell RJ. (1988) Human C5a anaphylatoxin: Gene synthesis, expression, and recovery of biologically active material from Escherichia coli. Methods in Enzymology 162: 653- 668.

Gallego C, Gupta S.K., Winitz S., Eisfelder B J., Johnson G.L. (1992) Myristoylation of the Gαi2 polypeptide, a G protein α subunit, is required for its signaling and transformation functions. Proc. Natl. Acad. Sci. USA89, 9695-9699.

Garritsen, A., van Galen, P.J.M., and W.F. Simonds (1993) The N-terminal coilded-coil domain of β is essential for γ association: A model for G-protein βγ subunit interaction. Proc. Nati. Acad. Sci. USA 90, 7706-7710.

Gasch A.A., Hoffman M., Horikoshi M., Roeder R., Chua N. (1990) Arabodopsis thaliana contains two genes for TFIID. Nature 346, 390.

Gelfand J.A., Donelan M., Hawiger A., Burke J.F. (1982) Alternative complement pathway activation increases mortality in a model of bum injury in mice. J. Clin. Invest 70: 1170-1176.

Gerard N.P. and Gerard C. (1990) Construction and expression of a novel recombinant anaphylatoxin, C5a-N19, a probe for the human C5a receptor. Biochemistry 29, 9274-9281.

Glotzer M., Murray A.W., Kirschner M.W. (1991 ) Cyclin is degraded by the ubiquitin pathway. Nature 349, 132-138.

Gomez R., Goodman LE; Tripathy SK; O'Rourke E; et al. Purified yeast protein famesyltransferase is structurally and functionally similar to its mammalian counterpart, (1993) Biochem. J. 289, 25-31.

Gotoh Y., Nishida E.,Shimanuki M., Toda T., Imai Y., Yamamoto M. (1993) Schizosaccharomyces pombe SPK1 is a tyrosine-phosphorylated protein functionally related to Xenopus mitogen-activated protein kinase. Mol. Cell. Biol. 13, 6427-6434.

Graf R., Mattera R., Codina J., Estes M., Birnbaumer L. (1992) A truncated recombinant α subunit of Gi3 with a reduced affinity for βγ dimers and altered guanosine 5'-3-0- (Thio) triphosphate binding. J. Biol. Chem. 267, 24307-24314.

Gros P., Dhir R., Croop J., Talbot F. (1991) A single amino acid substitution strongly modulates the activity and substrate specificity of the mouse mdrl and mdr3 drug efflux pumps. Proc. Natl. Acad. Sci. 88, 7289-7293. Guarente L. (1983) Yeast promoters and lacZ fusions designed to study expression of cloned genes in yeast. Methods Enzymol. 101, 181-191.

Guarente L. (1988) UASs and enhancers: Common mechanism of transcriptional activation in yeast and mammals. Cell 52, 303.

Guarente L. in The molecular and cellular biology of the yeast Saccharomyces: Gene Expression, Jones E.W., Pringle J.R., Broach J.R., eds., Cold Spring Harbor Laboratory Press, New York, 1992, p49-98.

Hadwiger J.A., Wittenberg C, Richardson H.E., de Barros Lopes M., Reed S.I. (1989) A family of cyclin homologs that control the GI phase in yeast. Proc. Natl. Acad. Sci. 86, 6255-6259. Hagen D.C, McCaffrey G., Sprague G.F. (1986) Evidence the yeast STE3 gene encodes a receptor for the peptide pheromone a-factor: gene sequence and implications for the structure of the presumed receptor. Proc. Natl. Acad. Sci. 83, 1418.

Hammerschmidt D.E., Weaver L.J., Hudson L.D., Craddock P.R., Jacob H.S. (1980) Association of complement activation and elevated plasma-C5a with adult respiratory distress syndrome. Pathophysiological relevance and possible prognostic value. Lancet 1:947-949. Hara M., Akasaka K., Akinaga S., Okabe M., Nakano H., Gomez R., WoβAD., Uh M., Tamanoi F. (1993) Identification of Ras famesyltransferase inhibitors by microbial screening. Proc. Natl. Acad. Sci. 90, 2281-2285.

Harbury P.B., Zhang T., Kim P.S. Alber T. (1993) Aswitch between two-, three-, 5 and four-stranded coiled coils in GNC4 leucine zipper mutants. Science 262, 1401- 1407.

Harshman K.D., Moye-Rowley W.S., Parker CS. (1988) Transcriptional activation by the SV40 AP-1 recognition element in yeast is mediated by a factor similar to AP- 1 that is distinct from GCN4. Cell 53, 321. o He B., Chen P., Chen S.-Y., Vancura K.L., Michaelis S., Higgins CF., Haag

P.D., Nikaido K., Aedeshir F., Garcia G. et al. (1982) Complete nucleotide sequence and identification of membrane components of the histidine transport operon of S. typhimurium. Nature 298, 723-727. »

Hoey T., Dynlacht B.D., Peterson M.G., Pugh B.F., Tjian R. (1990) Isolation and 5 characterization of the Drosophila gene encoding the TATA box binding protein, TFIID. Cell 61, 1179.

Hoffman A., Sinn E., Yamamoto T., Wang J., Roy A., Horikoshi M., Roeder R.G. (1990) Highly conserved core domain and unique N terminus with presumptive regulatory motifs in a human TATA factor (TFIID). Nature 346, 387. 0 Howe OP.H., Draetta G., Leof E.B. (1991) Transforming growth factor 1 inhibition of p34cdc2 phosphorylation and histone HI kinase activity is associated with Gl/S-phase growth arrest. Mol. Cell. Biol. 11, 1185-1 194.

Hrycyna CA., Sapperstein S.K., Clarke S., Michaelis S. (1991) The Saccharomyces cerevisiae STE14 gene encodes a methyltransferase that mediates C_* 5 terminal methylation of a-factor and RAS proteins. EMBO J. 10, 1699.

Hughes D.A., Ashworth A., Marshall CJ. (1993) Complementation of byrl in fission yeast by mammalian MAP kinase kinase requires coexpression of Raf kinase. Nature 364, 349-352.

Hyde S.C, Emsley P., Hartshorn M., Mimmack M.M., Gileadi U. et al. (1990) o Structural model of ATP-binding proteins associated with cystic fibrosis, multidmg resistance and bacterial transport. Nature 346, 362-365.

Jabbar, M. A., Sivasubramanian, N., Nayak, D. P. (1985) Influenza viral (A/WSN/33) hemagglutinin is expressed and glycosylated in the yeast Saccharomyces cerevisiae. Proc. Natl. Acad. Med. U.SΛ. 82, 2019-2023. s Journot L., Pantaloni C, Poul M.-A., Mazarguil H., Bockaert J., Audigier Y.

(1990) Amino acids 367-376 of the Gsα subunit induce membrane association when fused to soluble amino-terminal delted Gilα subunit. J. Biol. Chem. 265, 9009-9015.

Julius D., Brake A., Blair L., Kunisawa R., Thomer J. (1984) Isolation of the putative structural gene for the lysine-arginine-cleaving endopeptidase required for processing of yeast prepro-α-factor. Cell 37, 1075-1089. Julius D., Schekman, Thomer J . ( 1984) Glycosylation and processing of prepro- α-factor through the yeast secretory pathway. Cell 36, 309-318.

Julius D., Blair L., Brake A., Sprague G., Thomer J. (1983. Yeast α-factor is processed from a larger precursor polypeptide: the essential role of a membrane-bound dipeptidyl aminopeptidase. Cell 32, 839. Kakidani H, and Ptashne M. (1988) GAL4 activates gene expression in mammalian cells. Cell 52, 161.

Kao CC, Lieberman P.M., Schmidt M.C, Zhou Q., Pei R., Berk AJ. (1990) Cloning of a transcriptionally active human TATA binding factor. Science 248, 1646.

Kay B.K., Adey N.B., He Y.-S., Manfredi J.P., Mataragnon A.H., Fowlkes D.F. (1993) An M13 phage library displaying random 38-amino-acid peptides as a source of novel sequences with affinity to selected targets. Gene 128, 59-65.

Keyomarsi K. and Pardee A.B. (1993) Redundant cyclin overexpression and gene amplification in breast cancer cells. Proc. Natl. Acad. Sci. 90, 1112-1 116.

Khavari P. A., Peterson CL., Tamkun J.W., Mendel D.B., Crabtree G.R. (1993) BRGl contains a conserved domain of the SWI2/SNF2 family necessary for normal mitotic growth and transcription. Nature 366, 170-174.

King K., Dohlman H.G., Thomer J., Caron M.G., Lefkowitz RJ. (1990) Control of yeast mating signal transduction by a mammalian β2-adrenergic receptor and Gsα subunit. Science 250, 121-123. Kingsman S.M., Kingsman A.J., Mellor J. (1987) The production of mammalian proteins in Saccharomyces cerevisiae. TIBTECH 5, 53-57.

Koff A., Cross F., Fisher A., Schumacher J., Leguellec K., Phillipe M., Roberts J.M. (1991) Human cyclin E, a new cyclin that interacts with two members of the CDC2 gene family. Cell 66, 1217-1228. Koff A., Ohtsuli M., Polyak K., Roberts J.M. Massague J. ( 1993) Negative regulation of GI in mammalian cells: inhibition of cyclin E-dependent kinase by TGF-β. Science 260, 536-539.

Kohl N.E., Mosser S., deSolms S.J., Giuliani E.A., Pompliano D.L., Graham S.L., Smith R.L., Scolnick E.M., Oliff A., Gibbs J.B. (1993) Selective inhibition of ras- dependent transformation by a famesyltransferase inhibitor. Science 260, 1934-1937.

Kohl NE, Diehl R.E., Schaber M.D., Rands E., Soderman D.D., He B., Moores S.L., Pompliano D.L., Ferro-Novick S., Powers S. et al. (1991) Structural homology among mammalian and Saccharomyces cerevisiae isoprenyl-protein transferases. J. Biol. Chem. 266, 18884-18888.

Komer, J., Chun J., Harter D., Axel R. (1991) Isolation and functional expression 5 of a mammalian prohormone processing enzyme, murine prohormone convertase 1. Proc. Natl. Acad. Sci USA 88:6834-6838.

Kouba M; Vanetti M; Wang X; Schafer M; Hollt V (1993) Cloning of a novel putative G-protein-coupled receptor (NLR) which is expressed in neuronal and lymphatic tissue. FEBS Lett 321, 173-178. o Kramer R.A., Schaber M.D., Skalka A.M., Ganguly K., Wong-Staal F.„ Reddy

E.P. (1086) HTLV-III gag protein is processed in yeast cells by the virus pol-protease.

Kreil G. (1990) Processing of precursors by dipeptidylaminopeptidases: a case of molecular ticketing. Trends Biochem. Sci. 15, 23.

Kuchler K., Sterne R.E., Thorner J. (1989) Saccharomyces cerevisiae STE6 gene s product: a novel pathway for protein export in eukaryotic cells. EMBO J. 8, 3973.

Kurjan, J., Herskowitz, I. (1982) Structure of a yeast pheromone gene (MF): a putative α-factor precursor contains four tandem copies of mature α-factor. Cell 30_# 933- 943.

Kurjan J. (1985) α-factor structural gene mutations in yeast: effects on α-factor 0 production and mating. Mol. Cell. Biol. 5, 787-796.

Kyte and Doolittie (1982) A simple method for displaying the hydropathic character of a protein. J. Molec. Biol. 157, 105-132.

Lambright DG, Noel JP, Hamm, HE, Sigler, PB (1994) Structural determinants for activation of the α-subunit of a heterotrimeric G protein. Nature 369:621-628. 5 Lammie G.A., Fanti V., Smith R., Schuuring E., Brookes S., Michalides R.,

Dickson C, Arnold A. Peters G. (1991) Dl 1S287, a putative oncogene on chromosome 1 lql3, is amplified and expressed in squamous cell and mammary carcinomas and linked to BCL-1. Oncogene 6, 439-444.

Leberer E., Dignard D., Hougan L., Thomas DY, Whiteway M. (1992) o Dominant-negative mutants of a yeast G-pβotein b subunit identify two functional regions involved in pheromone signalling. EMBO J. 1 1 :4805-4813.

Lee E., Taussig R., Gilman A.G. (1992) The G226A Mutant of Gsα Highlights the Requirement for Dissociation of G Protein Subunits. J. Biol. Chem. 267:1212-1218.

Lee K.S., Irie K., Gotoh Y., Watanabe Y., Arakai H., Nishida E., Matsumoto K., 5 Levin D.E. (1993) A yeast mitogen-activated protein kinase homolog (Mpkl p) mediates signalling by protein kinase C Mol. Cell. Biol. 13, 3067-3075. Lee M.G. and Nurse P. (1987) Complementation used to clone a human homologue of the fission yeast cell cycle control gene cdc2. Nature 327, 31-35.

Lew DJ., Dulic V., Reed S.I. (1991) Isolation of three novel human cyclins by rescue of GI cyclin (Cln) function in yeast. Cell 66, 1 197-1206. Linder M.E., Pang I.-H., Duronio R.J., Gordon J.L, Stemweis P.C, Gilman A.G.

(1991) J. Biol. Chem. 266, 4654-4659.

Lupas, A.N., Lupas J.M., Stock J.B. (1992) Do G protein subunits associate via a three-stranded coiled coil? FEBS Lett. 314, 105-108.

Ma J., Przibilla J., Bogorad L., Ptashne M. (1988) Yeast activators stimulate plant gene expression. Nature 334, 631.

Manney T.R., Duntze W., Betz R. (1981) The isolation, characterization, and physiological effects of the S. cerevisiae sex pheromones. In Sexual interactions in eukaryotic microbes (ed. D.H. O'Day et al.), p 21. Academic Press, New York.

Masters, Stroud, and Bourne (1986) Protein Engineering 1:47-54. Matsushime H., Roussel M.F., Ashmun R.A., Scherr CJ. (1991) Colony- stimulating factor 1 regulates novel cyclins during the GI phase of the cell cycle. Cell 65, 701-713.

McDonnell D.P., Nawaz Z., Densmore C, Weigel N.L. et al. (1991) High level expression of biologically active estrogen receptor in Saccharomyces cerevisiae. J. Steroid Biochem. Mol. Biol. 39, 291-297.

Metzger, D., Losson R., Bornert J.M., Lemoine Y., Chambon P. (1992) Promoter specificity of the two transcriptional activation functions of the human oestrogen receptor in yeast. Nuc. Acids Res. 20, 2813-2817.

Metzger D., White J.H., Chambon P. (1988) The human oestrogen receptor functions in yeast. Nature 334, 31.

Milano, C.A., Allen, L.F., Rockman, H.A., Dolber, P.C, McMinn, T.R., Chien, K.R., Johnson, T.D., Bond R.A., Lefkowitz RJ. (1994) Enhanced myocardial function in transgenic mice overexpressing the β2-adrenergic receptor. Science 264, 582-586.

Mimura CS., Holbrook S.R., Ames G.F.-L. (1991) Structural model of the nucleotide binding conserved component of periplasmic permeases. Proc. Natl. Acad. Sci. 88, 84-88.

Moir DT; Davidow LS. (1991) Production of proteins by secretion from yeast. Methods Enzymol 194, 491-507.

Motokura T., Bloom T., Kim H.G., Juppner H., Ruderman J.V., Kronenberg H.M., Arnold A. (1991) A novel cyclin encoded by a bcll-linked candidate oncogene. Nature 350, 512-515. Moye-Rowley W.S., Harshman K.D., Parker CS. (1989) Yeast YAP1 encodes a novel form of the jun family of transcriptional activator proteins. Genes Dev. 3,283.

Mumby, S.M., Heukeroth R.O., Gordon J.L, Gilman A.G. (1990) G protein Gα- subunit expression, myristoylation, and membrane association in COS cells. Proc, Natl. 5 Acad. Sci. USA 87, 728-732.

Nakafuku M., Itoh H. Nakamura S., Kazioro Y. (1987) Occurrence in Saccharomyces cerevisiae of a gnee homologous to the cDNA coding for the α subunit of mammalian G proteins. Proc. Natl. Acad. Sci. 84, 2140-2144.

Nakagawa T., Hosaka M., Torii S., Watanabe T. et al. (1993) Identification and o functional expression of a new member of the mammalian Kex-2-like processing endoprotease family: its striking structural similarity to PACE4. J. Biochem. (Tokyo) 1 13, 132-135.

Nakayama K., Hosaka M., Hatsuzawa K., Murakami K. (1991) Cloning and functional expression of a novel endoprotease involved in prohormone processing at s dibasic sites. J. Biochem. 109, 803-806.

Nakayama K, Kim W.S., Torii S., Hosaka M. et al. (1992) Identification of a fourth member of the mammalian endoprotease family homologous to the yeast Kex2 protease. Its testis specific expression. J. Biol. Chem. 267, 5897-5900.

Nakayama N., Miyajima A., Arai K. (1985) Nucleotide sequences of STE2 and o STE3, cell type-specific sterile genes from Saccharomyces cerevisiae. EMBO 3. 4, 2643.

Nash R., Tokiwa G, Awand S., Erickson K., Futcher A.B. (1988) WHI1+ gene of Saccharomyces cerevisiae tethers division to cell size and is a cyclin homolog. EMBO J. 7, 4335-4346.

5 Neer E J., Pulsifer L., Wolf L.G. ( 1988) The amino terminus of G protein a subunits is required for interaction with βγ. J. Biol. Chem. 263, 8996-9000.

Neiman A.M., Stevenson BJ., Xu H.P., Sprague G.F. et al. (1993) Functional homology of protein kinases required for sexual differentiation in Schizosaccharomyces pombe and Saccharomyces cerevisiae suggests a conserved signal transduction module 0 in eukaryotic organisms. Mol. Biol. Cell. 4, 107- 120.

Neote, K. DiGregorio, D., Mak, J. Y., Horuk, R., and Schall, T. J. (1993) Molecular cloning, functional expression, and signaling characteristics of a C-C chemokine receptor. Cell 72, 415-425.

Noel J.P., Hamm H.E., Sigler P.B. (1993) The 2.2 A crystal structure of transducin-α complexed with GTPγS. Nature 366, 654-663.

Norman C, Runswick M., Pollock R., Treisman R. (1988) Isolation and - 132 -

properties of cDNA clones encoding SRF, a transcription factor that binds to the c-fos serum response element. Cell 55, 989.

Oeda, K., Sakaki, T., Ohkawa, H. (1985) Expression of rat liver cytochrome P- 450MC cDNA in Saccharomyces cerevisiae. DNA 3. 203-210. s Ogden, J.E., Stanway, C,Kuim, S.,Mellor, J., Kingsman, A.J., and Kingsman,

S.M. (1986) Efficient expression of the Saccharomyces cerevisiae PGK gene depends on an upstream activation sequence but does not require TATA sequences. Mol. Cell. Biol. 6,4335.

Olson L.M., Moss G.S., Baukus O., Das Gupta T.K. (1985) The role of C5 in o septic lung injury. Ann. Surg. 202: 771-776.

Overduin P., Boos W., Tomassen J. (1988) Nucleotide sequence of the ugp genes of Escherichia coli K-12: homology to the maltose system. Mol. Microbiol. 2, 767-775.

Peterson M.G., Tanese N., Pugh B.F., Tjian R. (1990) Functional domains and upstream activation properties of cloned human TATA binding protein. Science 248, s 1625.

Pines J. and Hunter T. (1990) Human cyclin A is adenovirus El A-associated protein p60 and behaves differently from cyclin B. Nature 346, 760-763.

Powers S. (1991) RAM2, an essential gene of yeast, and RAMl encode the two polypeptide components of the famesyltransferase that prenylates a-factor and Ras 0 proteins. Proc Natl Acad Sci 88:11373-11377.

Pronin, A.N., and N. Gautam (1992) Interaction between G-protein β and γ subunit types is selective. Proc. Natl. Acad. Sci. USA 89: 6220-6224.

Rarick H. M., Artemyev, N. O., and Hamm, H.E. A site on rod G protein α subunit that mediates effector activation. (1992) Science 256, 1031-1033. s Reyes M., Treptow M.A., Schuman H. A. (1986) Transport of p-nitrophenyl- maltoside by the maltose transport system of Escherichia coli and its subsequent hydrolysis by a cytoplasmicα -maltosidase. J. Bacteriol. 165, 918-922.

Rogers S., Wells R., Rechsteiner M. (1986) Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science 234, 364-368. 0 Russell M. and Johnson G. L. (1993) G protein N-terminal αi2/αs chimeras reveal amino acids important in regulating as activity. Mol. Pharmacol. 44:255-263

Scharer, E. and R. Iggo (1992). "Mammalian p53 can function as a transcription factor in yeast." Nuc. Acids Res. 20 (7): 1539-1545.

Schafer W.R., Kim R., Steme R., Thomer J., Kim S.-H., Rine J. (1989) Genetic 5 and pharmacological suppression of oncogenic mutations in RAS genes of yeast and humans. Science 245, 379. Schafer W.R., Trueblood C.E., Yang C-C, Maayer M.P., Rosenberg S._» Coulter CD., Kim S.-H., Rine J. (1990) Enzymatic coupling of cholesterol intermediates to a mating pheromone precursor and to the Ras protein. Science 249, 1133.

Schena M. and Yamamoto K.R. (1988) Mammalian glucocorticoid receptor s derivatives enhance transcription in yeast. Science 241 , 965.

Scherr CJ. (1993) Mammalian GI cyclins. Cell 73, 1059-1065.

Schultz R.M., Silberman S., Persky B. et al. (1988) Inhibition by human recombinant tissue inhibitor of metalloproteinases of human amnion invasion and lung colonization by murine B16-F10 melanoma cells. Cancer Res. 48, 5539. o Seideh N.G., Fournier H., Boileau G., Benjannet S. et al. ( 1992) The cDtøjjk structure of the procine pro-hormone convertase PC2 and the comparative processing by PCI and PC2 of the N-terminal glycopeptide segment of porcine POMC FEBS Lett,, 310, 235-239.

Singh A., Chen E.Y., Lugovoy J.M., Chang C.N., Hitzman R.A., SeeburgJPJrϊ. s (1983) Saccharomyces cerevisiae contains two discrete genes coding for the α-factor pheromone. Nuc. Acids Res. 11, 4049-4063.

Slepak V. Z., Wilkie T. M., Simon, M. 1.(1993) Mutational analysis of G protein α subunit Goα expressed in Escherichia coli. J. Biol. Chem. 268, 1414-1423.

Smeekens S.P. and Steiner D.F. (1990) Identification of a human insulinoina o cDN A encoding a novel mammalian protein structurally related to the yeast dibasic processing protease Kex2. J. Biol. Chem. 265, 2997-3000.

Speigel A. M., Backlund P. S., Jr., Butrynski J. E., Zones T. L. J., Simonds W. F.(1991) The G protein connection:molecular basis of membrane association, TIBS 16, 338-3441. s Steube K; Chaudhuri B; Marki W; Merryweather JP; Heim J. α-factor-leader- directed secretion of recombinant humaninsulin-like growth factor I from Saccharomyces cerevisiae. Precursor formation and processing in the yeast secretory pathway. (1991) Eur J Biochem 198, 651-657.

Strubin M. and Struhl K. (1992) Yeast and human TFIID with altered Dϊf A- o binding specificity for TATA elements. Cell 68, 721 -730.

Struhl, K. (1986) Constitutive and inducible Saccharomyces cerevisiae promoters: Evidence for two distinctive molecular mechanisms. Mol. Cell. BioL 6, 3847.

Struh, K. (1989) Molecular mechanisms of transcriptional regulation in yeast. Annu. Rev. Biochem. 58, 1051. 5 Sullivan, K. A., et al(1987) Nature 330, 758-760

Takahashi, H., Hakamata Y., Watanabe Y., Kikuno R. et al. (1983) Complete nucleotide sequence of the human corticotropin-beta-lipotropin precursor gene. Nucleic Acids Research 11 :6847-6858.

Thomas G., Thome B.A., Thomas L., Allen R.G., Hruby D.E., Fuller R., Thomer J. (1988) Yeast KEX2 endopeptidase correctly cleaves a neuroendocrine prohormone in 5 mammalian cells. Science 241 , 226.

Thomas L., Cooper A., Bussey H., Thomas G. (1990) Yeast KEX1 protease cleaves a prohormone processing intermediate in mammalian cells. J. Biol. Chem. 265, 10821.

Valdiva R.H., Wang L., Winans S.C. (1991) Characterization of a putative o periplasmic transport system for octopine accumulation encoded by Agrobacterium tumefaciens T: plasmid pTi46. J. Bacteriol. 173, 6398-6405.

Vogt P.K., Bos TJ., Doolittie R.F. (1987) Homology between the DNA-binding domain of the GCN4 regulatory protein of yeast and the carboxy-terminal region of a protein coded for by the oncogene jun. Proc. Natl. Acad. Sci. 84, 3316. 5 Waters M.G., Evans E. A., Blobel G. ( 1988) Prepro-α-factor has a cleavable signal sequence. J. Biol. Chem. 263, 6209.

Webster N., Jin J.R., Green S., Hollis M., Chambon P. (1988) The yeast UASG is a transcriptional enhancer in human HeLa cells in the presence of the GAL4 transactivator. Cell 52, 169. o Weisman H.F., Bartow T., Leppo M.K., Marsh H.C Jr. et al. ( 1990) Soluble human complement receptor type 1 : in vivo inhibitor of complement suppressing post¬ ischemic myocardial inflammation and necrosis. Science 249: 146-151. West, J. P. et al (1985) J. Biol. Chem. 260, 14428-14430. Whiteway M., Clark K.L, Leberer E., Degnard D., and Thomas D.Y. (1994) 5 Genetic Identification of Residues Involved in Association of α and β G-Protein Subunits. Mol. Cell. Biol. 14:3233-3239.

Wood C.R., Boss M.A., Kenten J.H., Calvert J.E., Roberts N.A., Emtage J.S. (1985) The synthesis and in vivo asembly of functional antibodies in yeast Nature 314, 446-449. o Xiong Y., Connolly T., Futcher B., Beach D. ( 1991 ) Human D-type cyclin. Cell

65, 691-699.

Equivalents

5 Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, numerous equivalents to the specific polypeptides, nucleic acids, methods, assays and reagents described herein. Such equivalents are considered to be within the scope of this invention and are covered by the following claims.

All of the above-cited references, issued patents and patent publications are hereby incorporated by reference.

SEQUENCE LISTING (1) GENERAL INFORMATION:

(l) APPLICANT:

(A) NAME: Cadus Pharmaceutical Corporation

(B) STREET: 777 Old Sawmill Road (C) CITY: Tarrytown

(D) STATE: New York

(E) COUNTRY: USA

(F) POSTAL CODE (ZIP) : 10591

(G) TELEPHONE: (914) 345-3344 (H) TELEFAX: (914) 345-3565

(ii) TITLE OF INVENTION: Functional Expression of Heterologous Adenylyl Cyclase in Cells (iii) NUMBER OF SEQUENCES: 16

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: LAHIVE & COCKFIELD

(B) STREET: 60 State Street, Suite 510 (C) CITY: Boston

(D) STATE: Massachusetts

(E) COUNTRY: USA

(F) ZIP: 02109 (V) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: Patentln Release #1.0, Version #1.25

(vi) CURRENT APPLICATION DATA:

(A) APPLICATION NUMBER:

(B) FILING DATE:

(C) CLASSIFICATION:

(vii) PRIOR APPLICATION DATA:

(A) APPLICATION NUMBER: 08/636,596

(B) FILING DATE: 23 April 1996 (viii) ATTORNEY/AGENT INFORMATION:

(A) NAME: Kara, Catherine J.

(B) REGISTRATION NUMBER: P41.106

(C) REFERENCE/DOCKET NUMBER: CPI-014CPPC (ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: (617) 227-7400

(B) TELEFAX: (617) 227-5941 (2) INFORMATION FOR SEQ ID NO:l:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 55 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: CDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:l:

GCCGTCTCAC ATGAGAAGAA GAAGATACTT GAGAGATAQA GCTGAAGCTG CTGCA 55

(2) INFORMATION FOR SEQ ID NO:2: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 51 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:

GCAGCTTCAG CTCTATCTCT CAAGTATCTT CTTCTTCTCA TGTGAGACGG C 51

(2) INFORMATION FOR SEQ ID NO:3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 42 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: CDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:3:

GCTGCTGCTG CTGGTGGTGG TGAAGGTTTG CAAAGATCCC GG 42 (2) INFORMATION FOR SEQ ID NO:4:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 50 base pairs

(B) TYPE: nucleic acid (c) STRANDEDNESS: single O 97/40170 PC17US97/06711

- 138 -

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: GATCCCGGGA TCTTTGCAAA CCTTCACCAC CACCAGCAGC AGCAGCTGCA 50

(2) INFORMATION FOR SEQ ID NO:5:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 29 base pairs (B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5: CAGACATGTC TTGGTTTCGT GGCCTCCTG 29

(2) INFORMATION FOR SEQ ID NO:6:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 36 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:6:

GCGGATCCAA GGTCATGACC AGTTCCTGTG CAGTGC 36

(2) INFORMATION FOR SEQ ID NO:7: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 25 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: CATGACTGAA GATCAAGGTT TCTCG 25

(2) INFORMATION FOR SEQ ID NO:8:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 25 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8:

GATCCGAGAA ACCTTGATCT TCAGT 25 (2) INFORMATION FOR SEQ ID NO:9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 31 base pairs

(B) TYPE: nucleic acid (C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: GGGCGTCTCC CATGGCCAGC AACAACACCG C 31 (2) INFORMATION FOR SEQ ID NO:10:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 29 base pairs (B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10; GGGGTCGACC GAGGCTCCTC AGGTTCCTC 29

(2) INFORMATION FOR SEQ ID NO:11:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 44 base pairs O 97/40170 PC17US97/06711

- 140 -

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: CGGCTAGCAT CTATATACAA TGAGTGAACT TGACCAGTTA CGGC 44

(2) INFORMATION FOR SEQ ID NO:12: (i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 39 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA

(Xi) SEQUENCE DESCRIPTION: SEQ ID NO:12:

CGAGCGGCCG CTCAGTTCCA GATTTTGAGG AAGCTGTCC 39

(2) INFORMATION FOR SEQ ID NO:13:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 28 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: CDNA

(Xi) SEQUENCE DESCRIPTION: SEQ ID NO:13:

CCCCCCATGG TTCACCAACA ACGAAACC 28 (2) INFORMATION FOR SEQ ID NO:14:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 27 base pairs

(B) TYPE: nucleic acid (C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: cDNA (xi) SEQUENCE DESCRIPTION: SEQ ID NO:14:

CCGGATCCTT ACGCCAGGTT AGCGTCG 27

(2) INFORMATION FOR SEQ ID NO:15:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 39 base pairs (B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear ^"

(ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:15: GTCCGAGATC TCTTAAGGAC GCGTTGCTAG CAGCATGCG 39

(2) INFORMATION FOR SEQ ID NO:16:

(i) SEQUENCE CHARACTERISTICS: (A) LENGTH: 39 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: cDNA

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:16: GACCGCATGC TGCTAGCAAC GCGTCCTTAA GAGATCTCG 39

Claims

We Claim:

1. The yeast or mammalian cell which expresses a mutant form of a heterologous adenylyl cyclase that is constitutively activated.

2. The cell of claim 1 , which further expresses a peptide library.

3. The cell of claim 2, wherein the peptide library is expressed intracellularly.

4. The cell of claim 1 , which further comprises an intragenic mutation that restores regulation of the mutant form of the heterologous adenylyl cyclase.

5. The cell of claim 4, which further expresses a peptide library.

6. The cell of claim 5, wherein the peptide library is expressed intracellularly.

7. The cell of any of claims 1 -6, which is a yeast cell.

8. The cell of any of claims 1-6, which is a mammalian cell.

9. A yeast cell expressing a mutant form of a heterologous adenylyl cyclase.

10. The yeast cell of claim 9, which further expresses a peptide library.

1 1. The yeast cell of claim 10, wherein the peptide library is expressed intracellularly.

12. A yeast cell expressing a heterologous adenylyl cyclase and an intracellularly expressed peptide library.

13. A mammalian cell expressing a mutant heterologous adenylyl cyclase and a peptide library.

14. The mammalian cell of claim 13, wherein the peptide library is expressed intracellularly.

15. The cell of any one of claims 1 -6 or 9-12, wherein the heterologous adenylyl cyclase is of mammalian origin.

5 16. The cell of any one of claims 15, wherein the mammalian adenylyl cyclase is of human origin.

17. The cell of any of claims 1 -6, wherein the heterologous adenylyl cyclase comprises a mutation in the Cia domain of the enzyme to cause constitutive activation o of an adenylyl cyclase activity.

18. The cell of any of claims 1-6, wherein the heterologous adenylyl cyclase comprises a mutation in the C2a domain of the enzyme to cause constitutive activation of an adenylyl cyclase activity. 5

19. The cell of claim 17, wherein the heterologous adenylyl cyclase comprises a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyr265His, Glu313Gly,Val388Ile, GIy968Ser, Arg 268 Lys, Gly968Asp, and Lys998Asn. 0

20. The cell of any one of claims 1 -6 or 9- 14, wherein an endogenous adenylyl cyclase of the cell is inactivated.

21. The cell of claim 20, wherein activity of the endogenous adenylyl cyclase is 5 temperature sensitive.

22. The cell of claim 20, which comprises the mutant allele cdc35-l .

23. The cell of any one of claims 1-6 or 9-14, wherein the heterologous adenylyl o cyclase is selected from the group consisting of a type I, a type II, a type III, a type IV, a type V, a type VI, a type VII, a type VIII, a type IX and a type X adenylyl cyclase.

24. The cell of any one of claims 1-6 or 9-14, wherein the cell further expresses one or more heterologous or chimeric G protein subunits. 5

25. The cell of claim 24, wherein one or more of the heterologous G protein subuints O 97/40170 PC17US97/06711

- 144 -

is of mammalian origin.

26. The cell of claim 24, wherein one or more of the chimeric G protein subuints comprises a sequence of mammalian origin.

27. The cell of any one of claims 1-6 or 9-14, wherein the cell further comprises a reporter gene construct comprising a cyclic AMP responsive promoter operably linked to a reporter gene encoding a selectable or screenable gene product.

28. The cell of any one of claims 2, 3, 5, 6 or 10-14 wherein the peptide library comprises a random or semi-random peptide library.

29. The cell of any of claims 2, 3, 5, 6 or 10-14, wherein the peptide library comprises a library of peptides derived from a G protein subunit.

30. The cell of claim 29, wherein the G protein subunit is Gas.

31. The cell of claim 28, wherein the expression of the peptide library is directed by a thioredoxin A expression vector.

32. The cell of any one of claims 1-6 or 9-12, wherein the cell is a yeast cell selected from the group consisting of: Kiuyveromyces lactis, Saccharomyces kluyveri, Schizosaccharomyces pombe, Ustilago maydis, and Saccharomyces cerevisiae.

33. A method of identifying a modulator of adenylyl cyclase activity which comprises:

(i) contacting the cell of any of claims 1 -6 or 9-14 with a test compound, (ii) comparing the level of adenylyl cyclase activity in the cell in the presence of the test compound with the level of activity in the absence of the test compound or in a cell lacking the heterologous adenylyl cyclase, and

(iii) identifying the test compound as a modulator of adenylyl cyclase activity, wherein a change in the level of adenylyl cyclase activity in the presence of the test compound indicates that the test compound is a modulator of adenylyl cyclase activity.

34. A method of identifying a modulator of adenylyl cyclase activity which comprises: (i) culturing the cell of any one of claims 2, 3, 5, 6 or 10-14, (ii) comparing the level of adenylyl cyclase activity in the presence of the expressed peptide library with the level of activity in the absence of the expressed peptide library or in the absence of the heterologous adenylyl cyclase activity, and 5 (iii) identifying peptides which modulate the activity of adenylyl cyclase, wherein a change in the level of adenylyl cyclase activity in the presence of tiie expressed peptide indicates that the expressed peptide is a modulator of adenylyl cyclase activity.

o 35. The method of claim 33 or 34, which is used to identify inhibitors of adenylyl cyclase activity. j_r_μ

36. The method of claim 33 or 34, which is used to identify activators of adenylyl cyclase activity. 5

37. A method of identifying an adenylyl cyclase stimulatory domain of a G protein subunit, comprising:

(i) culturing the cell of claim 29,

(ii) comparing the level of the adenylyl cyclase activity in the presence of the o expressed peptide library with the level of activity in the absence of the expressed peptide library, wherein a change in the level of adenylyl cyclase activity in the presence of the expressed peptide library indicates that the expressed peptide is a cyclase stimulatory domain of a G protein subunit. 5

38. The method of claim 37, wherein the G protein subunit is Gas and the cyclase is a type IV cyclase.

39. A method of identifying an intragenic mutation that restores regulation of a constitutively active mutant form of an adenylyl cyclase, comprising

(i) culturing a yeast cell which expresses a constitutively active mutant form of a heterologous adenylyl cyclase at 23°C and at 30°C; (ii) measuring the level of adenylyl cyclase activity at 23°C and at 30°C; and (iii) identifying an intragenic mutation that restores regulation of the constitutively active mutant form of the heterologous adenylyl cyclase, wherein a difference between the level of adenylyl cyclase activity at 23°C and at 30°C is indicative of the presence of an intragenic mutation.

40. The method of claim 33 or 34, which further comprises of preparing a pharmaceutical preparation of one or more compounds identified as being a modulator of adenylyl cyclase activity.

41. An expression vector encoding a constitutively active mutant form of a mammalian adenylyl cyclase.

42. The expression vector of claim 41 , wherein the mutant form of the mammalian adenylyl cyclase comprises a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyr265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968Asp, and Lys998Asn.

43. A nucleic acid encoding a mutant mammalian adenylyl cyclase comprising a mutation corresponding to a mutation in adenylyl cyclase type IV selected from the group consisting of Tyτ265His, Glu313Gly,Val388Ile, Gly968Ser, Arg 268 Lys, Gly968Asp, and Lys998Asn.