AU2249400A - Multispecific chimeric receptors - Google Patents

Description

Our Ref:7465662 P/00/011 Regulation 3:2

AUSTRALIA

Patents Act 1990

ORIGINAL

COMPLETE SPECIFICATION STANDARD PATENT Applicant(s): Address for Service: Invention Title: Cell Genesys, Inc 342 Lakeside Drive Foster City California 94404 United States of America DAVIES COLLISON CAVE Patent Trade Mark Attorneys Level 10, 10 Barrack Street SYDNEY NSW 2000 Multispecific chimeric receptors The following statement is a full description of this invention, including the best method of performing it known to me:- 5020 MULTISPECIFIC CHIMERIC RECEPTORS

INTRODUCTION

15 Technical Field The field of the invention relates to the construction and use of novel multispecific chimeric receptors to overcome the obstacles presented by drug-resistance and genetic variation in 20 infectious disease, cancer and autoimmune disease.

Backaround Agents designed to selectively inhibit the replication of a rapidly growing pathogen .or cancer inevitably face a challenge from the development of drug-resistance. This problem is viewed by many clinicians as one of the major impediments to the effective management of malignant disease tor the control of infectious agents which undergo -major genetic variation. -A related dilemma is presented by "i.autoi intie 4ease; where a disease-causing cell population may be heterogeneous with respect to marker -antigens which could be targeted as part of a therapeutic strategy.

The challenge of genetic variation to disease therapy is well illustrated by the problem. of antiviral drug-resistance, though a similar situation holds true for resistance to anti-microbials and chemotherapeutics. The clinical emergence of drug-resistant virus has been documented in most instances in which antiviral therapy has been applied, including HIV, herpes simplex virus, varicella zoster virus, cytomegalovirus, influenza A virus, and rhinovirus (Richman, Curr. Opin. Infect. Dis. 3:819-823 (1990)). HIV infection provides a clear example of this problem, given its chronic, persistent nature, the high rate of viral replication, and the error-prone character of reverse transcriptase (RT).

Resistance has been observed for every HIV antiviral tested, including nucleoside analogs (AZT, ddl, ddC, d4T and TSAO), nonnucleoside RT inhibitors (Merck's L-697,639 and Boehringer 5 Ingelheim's nevirapine), and a protease inhibitor (Richman, Annu.

Rev. Pharmacol. Toxicol. 32:149-164 (1993)).

e The clinical significance of drug-resistance in HIV infection is best documented for AZT. While AZT reduced the rate of mortality 20 in AIDS patients by over half during the first 12 months of treatment, ongoing therapy out to 24 months did not provide any additional advantage to the treated group. The apparent loss of S. AZT clinical efficacy correlates with the finding that by 12 months of .therapy, approximately. 90% of individuals with late-stage disease have developed AZT resistant virus (Richman et al., J. AIDS 3:743-746 (1990)). Similarly, the loss of -antiviral activity observed with the non-nucleoside RT inhibitors within two months of their administration is associated with the.-rapid appearance of Sdrug-resistant virus (Nunberg et al., J. Virol., 65:4887-4892 .(1991); _Richman et al., Proc. -Natl. Acad.- Sci 88:11241-11245 S(1991)." Thus, while antiviral therapy with single agents can be quite effective for short periods, extended treatment of chronic or latent infections like HIV and the herpes viruses may require the application of combination therapies. Such a strategy is, however, often impractical .due to the additional .problems. of crossresistance as well as the cumulative side-effects of multiple agents. It is therefore desirable to design an therapeutic agent which can attack the pathogen at multiple points, in a fashion that minimizes cross-resistance, and has the safety profile of a single active agent. The present invention achieves this goal by providing multispecific chimeric receptors.

SUMMARY OF THE INVENTION The present invention provides novel multispecific chimeric receptors and their applications to human disease therapy. The multispecific chimeric receptors of the present invention are single proteins possessing more than one antigen-binding and/or ligand-binding domain linked to an effector signaling domain and/or a proliferation signaling domain. A principal application of the novel multispecific chimeric receptors of the present invention is to combine the therapeutic benefits of two or more monospecific chimeric effector function receptors in a single protein for the treatment of a disease. In this manner, a multispecific protein product? is provided which has the pharmacological profile of a single agent.

DES.--CRIPTION OF THE DRAWINGS Figure 1 illustrates the structures of the multispe'cific chimeric receptors discussed in the detailed description.

Figure 2 is a listing of oligonucleotides (SEQ ID NOS: 1-22) as described in the Examples, infra.

.jj Figure 3 is an autoradiogram of -immunoprec ipi tat ions of .culture supernatants from 293 cells transfected with the following monospecific and multispecific antibody constructs as described in Examples 1 and 2: lane a, CD4-Fc;. lane b, SAb(cxgp4l) -Fc; lane c, SAb (agpl20) -Fc; lane d, SAb(cxgp4l)-SAb(agpl2O)-Fc; lane e, SAb(cagp4l) -Ll-SAb(cagpl2O) -Fc; lane f, SAb(agp4l) -L2-SAb(cxgpl2O) -Fc; lane g, SAb(agp4l)-L3-SAb(agpl2O)-Fc; lane h, SAb(cgp4l)-L4- SAb(agpl2O)-Fc; lane i, SAb(agpl2O)-CD4-Fc; lane j, SAb(agpl2O)-L1- CD4-Fc; lane k, SAb(cagp12)-L2-CD4-Fc; lane 1, SAb(agpl2O)-L3-CD4- Fc; lane m, Sab(agpl2O)-L4-CD4-Fc.

Figure 4 illustrates the cytolytic activity of CD8* T lymphocytes expressing chimeric. receptors on 5 1 Cr-labeled 293 target cells as ::..described in Example 10F. The cytolytic effector cells were either untransduced (triangles), or transduced with retroviral. constructs encoding either SAb(agp12O)-L2-CD4-. (squares),, SAb(agp41)- (circles) or CD4- (diamonds). The target cells were -29-3. cells transfected with either the neo gene (293neo), the wild-type .HIV-l env gene (293 env), or mutant env genes (293/CD4- and 293/447D-).

The Y axis is specific lysis and the X axis if the effector to target ratio.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As noted above, the present invention generally relates to novel multispecific chimeric receptors, namely, chimeric proliferation receptors, chimeric effector function receptors, and hybrids thereof, each comprising at least two ligand and/or antigen binding domains of different specifities, and DNA sequences encoding these novel chimeric receptors. Further aspects of the present invention %J will be discussed in detail below following a definition of terms employed herein.

Definitions: The term "extracellular inducer-responsive clustering domain" or "ECD" refers to the portion of a protein of the present invention which is outside of the plasma membrane of a cell and binds to at least one extracellular inducer molecule as defined below. The ECD may include the entire extracytoplasmic portion of a-transmembrane protein, a cell surface or membrane associated protein, a secreted 20 protein, a cell surface targeting protein, a cell adhesion molecule, or a normally intracytoplasmic inducer-binding domain, and truncated or modified portions thereof. In addition, after :*..binding one or more extracellular inducer molecule(s)defined below, the ECDs will become associated with each other by dimerization or oligomerization, "cluster".

The term "multispecific e xtracellular. inducer-responsive clustering domain" or "MSECD" refers to more, -than one ECD, as defined above, linked in tandem, either covalently or noncovalently, on the same protein. The term "intracellular inducer-responsive clustering domain" or "ICD" refers to the portion of a .protein which is inside of the plasma membrane of a cell, that binds to at least one intracellular inducer molecule as defined below. After binding one or- more intracellular -inducer nolecule(s), -the ICDs will become associated with each other by dimerization or oligomerization, i.e., "cluster".

The term "transmembrane domain" or "TM" refers to the domain of the protein which crosses the plasma membrane and is derived from the inducer-binding ECD domain, the effector function signaling domain, the proliferation signaling domain or a domain associated with a totally different protein.'Alternatively, the transmembrane domain may be an artificial hydrophobic amino acid sequence which spans the plasma cell membrane.

The term "proliferation signaling domain" or "PSD" refers to a protein domain which signals the cell to enter mitosis and begin cell growth. Examples include the human or mouse Janus kinases, including but not limited_ to, JAK1,- JAK2, JAK3, Tyk2, Ptk-2, homologous members of the Janus kinase family from other mammalian Sor eukaryotic specien, the IL-2 Receptor B and/or y chains and other subunits from the cytokine receptor superfamily of proteins ~that may interact -with the Janus kinase family of proteins to transduce a signal, and the cytoplasmic domains from the members of the superfamily of tyrosine kinase -growth factor receptors, or portion's, modif ications or conbinations thereof.

The term "effector function" refers to the specialized function of a differentiated cell. Effector function of a T cell, for example, iiay be cytolytc Activity or eiper' activity including the secretion of cytokines.- The term "effector function signaling domain or "EFSD" refers to the portion of a protein which transduces the effector function signal and directs the cell 6 perform 'its specialized fu nction.

While usually the entire EFSD will b emplyred, in many cases it .T1 .4 -~will not be necessaryc-,totisethe entire chain.- To -the extemitthat a truncated portion of the EFSD may find use, such truncated portion may be used in place of the intact chain as long as it .still transduces the:.-effector. .function-.signial. Examples are the* chain -of the T. cell,!.receptor or .any .of its .homologs n chain, FceRl -y and-f chains MBl .7(Igc) chain;11 B29 -(Igo).,chain, etc.) CD3 polypeptides 5 'and -syk..f amily -tyrosine kinases (Syk, ZAP etc.), -src. family. tyrosine kinases (Lck,; Fyn, Lyn, etc.) and other molecules involved in T.'cell'transduction,* such as CD2, and CD28.

The term "linker" or' "linker 'region" refers- to an oligo- or polypeptide region.of -from about 1 ,to 100 amino acids that -links together -any of the above described domains of the MSCRs defined above. The amino acid seqence is typically derived from a protein other than the ICDs, ECDs, EFSDs, PSDs or TM domains. Examples of linker regions are* linktr--'212- and -linker 205 as referenced in Bedzyk et Biol. 'Chem.', 265:18615-18620 (1990) and Gruber et al. J. Immxunol., 152:5368-5'374 (1994), respectively.

The term "memnbrane hinge" refers to a hydrophilic polypeptide *sequence found -in m~embran'e-b6"und'immunoglobulin heavy chains, where it is attached 'to the' extrace lular side of the TM domain (Bensmana &Lefranc, Imu~ !ej 32:321-'330' As used in the 25 present invention, th'nrbaehinge may be-considered A' subset of linke rs.

i*dcr -ep6~ -pr 1 if e The term OchimericIide-rposv pofeaion receptor" or 2 CPR" refers to a chimieric-receptor that comprises'~ extracellular inducer responsi~ 'cfitediii d~~niE ;atasmembrane domain TM adatleatt on*Okl signaling' dom'in (PSD) The ECD and PSD-i n:J6Una4 0-1' ~id~together on a 'single" protein -receptor. Otionilly t r' receptor may aldcnanan effector> -function signaling .domain to -f orm "hybrid MSCR as defined below. The term "chimeric effector.- function receptor" or- "CEFR" .refers to a chimerici receptor. that *comprises an extracellular domain, transmembrane domain, and at least -one cytoplasmic domain as.described in U.S. -Patent #5,359,046 or at least one-EFSD-domain as described above. The extracellular domain serves to bind to an extracellular inducer and transmit a signal to the cytoplasmic domain which transduces an effector function signal to the cell.

The term "multispecific chimeric effector function receptor" or "MSCEFR" refers to a chimeric effector function receptor which contains a multispecific extracellular inducer-responsive clustering domain (MSECD).

The term "multispecific chimeric. proliferation receptor" or "MSCPR" refers to a chimeric inducer-responsive proliferation 2* receptor which contains ,a multispecific extracellular inducerresponsive clustering domain (MSECD).

The term "hybrid MSCR" refers to a chimeric_ receptor that comprises a MSECD, a TM and at least one. EFSD and at least one PSD linked together, in either order, directly or via a linker region to the transmembrane domain-in either order.

The term "multispecific chimeric receptor" or "MSCR" refers to a chimeric receptor that comprises a multispecific ECD. (MSECD), a transmembrane domain (TM) and at least -one effector function signaling domain (EFSD) ,and/or at least one proliferation signaling domain,(PSD (i e. MSCEFRs, MSCPRs and hybrid MSCRs)). In addition, the MSCEFRs, MSCPRs and hybrid MSCRs of the.present invention may also have one or .more .ICDs attached. to, -one or -more of their cytoplasmic domains.

The.term "extracellular inducer:molecule" refers to a ligand or antigen.which binds to an ECD.and induces-the clustering of-the ECD or MSECD as described above, or portions or modifications of the extracellular inducer molecule that-are:still capable of binding to the ECD and inducing the clustering of an MSECD. To facilitate clustering,. the extracellular inducer molecule may be intrinsically bivalent or multivalent; or it may be :presented to the ECD in a bivalent or multivalent form, eg., on the surface of a cell or a virus.

The term "intracellular inducer molecule" refers to a natural or synthetic ligand that can be delivered to the cytoplasm of a cell, and binds to and induces the clustering of an intracellular-inducer responsive domain (ICD). To facilitate clustering, the *16 intracellular inducer molecule may be intrinsically bivalent or e multivalent.

The term "multispecific antibody" refers to an antibody molecule, or truncations or modifications thereof, that comprises two or more ECDs of different specificities.

0 The term "modifications" refers to an addition of one or more amino acids to either or both of the C- and N-terminal ends of the extracellular or intracellular inducer molecules (in the case where these are proteins) or, the ECDs, PSDs, EFSDs, ICDs or TMs, a substitution -of one or more amino acids at one or more sites -throughout these proteins, a deletion of one or more amino acids within or at either or both ends 'of these proteins, or an insertion of- one or more amino acids at one or more sites in these proteins such that the extracellular or' iritracellular inducer molecule binding to the ECD or ICD is retained or improved as measured by binding'assays known in theart, for example, Scatchard plots, or such that 'the PSD,- EFSD, ICD or TM domain activities are retained or improved as measured by one or more of the proliferation assays .or effector assays described below. .In addition, modifications can Sbe .made to the extracellular or- intracellular inducer molecules (where they are proteins) .and to the corresponding ECDs zor ICDs to create an improved receptor-ligand binding pair.

The term "variant" refers to a DNA fragment encoding an.

extracellular or intracellular inducer molecule, or an ECD, PSD, EFSD, ICD or TM domain that may further contain an addition of one or more nucleotides internally or at the 5' or 3' end of the DNA "0 fragment, a deletion of one or more nucleotides internally or at the 5' or 3' end of the DNA fragment or a substitution of one or more nucleotides internallly or at the 5' or 3' end of the DNA fragment such that the extracellular or intracellular inducer molecule binding to the ECD or ICD is retained or improved as measured by binding assays known in the art, for example, Scatchard plots, or such that the PSD, EFSD, ICD or TM domain activities are retained or improved as measured by one or more of the proliferation assays or effector assays described below. In addition, variants of the DNA sequences encoding the extracellular and intracellular inducer molecules (where they are proteins) and the corresponding ECDs and ICDs can be made so as to create an improved receptor-ligand binding pair.

In a general embodiment, the present invention relates to novel :25 multispecific chimeric receptors, nucleic acid sequences encoding the receptors, the vectors containing the nucleic acids encoding the receptors, the host cells expressing these novel multispecific chimeric receptors, and methods of using these novel multispecific chimeric receptors as therapeutics. Three .types of multispecific chimeric receptors (MSCR) are provided herein, .namely, multispecific chimeric proliferation receptors .(MSCPR) (Figure multispecific chimeric effector function receptors (MSCEFR) (Figure and hybrid .MSCRs. comprising both _an effector signaling domain and a proliferation signaling domain (Figure l(c) and In each.category of receptors, the multispecific binding domain, the effector function signaling domain, and the proliferation signaling domain do not-naturally exist together on a single protein.

In one particular. embodiment, the present invention relates to a multispecific chimeric proliferation receptor (MSCPR) designed to -be expressed in cells, -which in turn proliferate :in response to at least one specific extracellular inducer molecule. The three domains which comprise a MSCPR are: a multispecific binding domain comprising at least two extracellular inducer-responsive clustering domains (ECDs) which serves to bind to -at least one specific extracellular molecule, a transmembrane domain, which crosses the plasma membrane and, at least one proliferation 15 signaling domain that -signals the cell to divide (Figure The cells bearing the MSCPRs of the present invention will expand in number in response to the binding of one or more different specific extracellular molecules to an extracellular inducerresponsive clustering domain of the MSCPR. In each instance, the S extracellular inducer molecule binds to at least one ECD, which results in the dimerization or oligomerization of the MSECDs to each other. The clustering of these MSECDs signals activation of the proliferation domain(s).' *5 In another embodiment, the present invention relates to a novel multispecific.chimeric .effector function receptor (MSCEFR) designed to be expressed .in cells,- which when :activated the binding of at least one specific.exjctracellular inducer molecule', will induce a specialized effector function ;of a differentiated cell. The three domains:that comprisel a -MSCEFR- are: a 'multispecific binding Sdomain :comprisig at :least-two rextracellular 'inducer-responsive clustering idomains -(ECDs) which serves tb -bihd 'to -at least one specific extracellular -inducer": molecule, a transmembrane domain, .which. crosses -the plasma membrane and, at leat one effector function signaling domain which tranduces the effector function signal. and .directs the cell to perform its specialized function (Figure The cells bearing the MSCEFRs of the present invention will express effector function in response to the binding of one or more different specific extracellular inducer molecules to an extracellular inducer-responsive clustering domain of the MSCEFR.

l In each instance, the extracellular inducer molecule binds to at least one ECD, which results in the dimerization or oligomerization of the MSECDs to each other. The clustering of these MSECDs signals activation of the effector function signaling domain(s).

oe In yet another .embo4i ie4 the present invention relates to a novel hybrid multispecific chimeric receptor .(hybrid MSCR) containing a proliferation signaling domain and an effector function signaling domain together on the same multispecific receptor. In this particular embodiment, the hybrid receptor comprises a MSECD and TM described above, and additionally comprises at least one effector function signaling domain and at least one proliferation:siii~aling domain joined together on the same protein (Figure:- Ai:{ ad-d Thus, the multispecific extracellular inducer responsive clustering domains (MSECDs) of the 25 hybrid MSCR are linked via a transtmembrane domain to two different types of signal trahnsduing dbmains. Either the proliferation signaling domain or the effector function signaling-domain may be linked to" the transmemi~a ed o;mjain -which :is further linked on its 3' end to the second signalIng ,domain -either directly or through a linker region. It-is contemplated that the preparation :dof this novel hybrid MSCR:will avS.te, proliferation- and effector. function simultaneously inb~a! t ll .upon >the -binding of at least one extracellular i-nd', r -:iBe :.to one -or -more the ECDs -of the hybrid MSCR of the present. invention.

In yet another; aspect of the present invention, a novel multispecific chimeric .proliferation receptor containing a multispecific extracellular -inducer-responsive clustering domain (MSECD), -and a proliferation signaling domain (PSD) is provided together. in the same receptor protein with an intracellular inducer-responsive clustering domain (ICD). In this embodiment, a receptor is constructed as one protein comprising in the N-terminal to C-terminal direction a multispecific ECD, TM domain, an ICD and a PSD. Alternatively, a receptor may be constructed as one protein comprising in the N-terminal to C-terminal direction an MSECD, TM domain, PSD and an ICD. In preparing the multispecific chimeric inducer binding receptors of the present embodiment, one may separate one or more domains of each receptor with a linker or 1 5 membrane hinge region. Additionally, more than one ICD and PSD may be attached directly or via a linker or membrane hinge region to each other to form multiple ICDs and PSDs. Upon introduction of these novel inducer-binding chimeric proliferation receptors into a host cell, one may modulate proliferation of the cell by either an extracellular inducer, an intracellular inducer or a combination of these two different inducer molecules. The embodiment of this aspect of the invention may be modified even further by attaching an effector function signaling domain (EFSD) at the N-terminal or C-terminal end of the PSD or the "ICD. A MSECD and multiple ICDs and/or PSDs may be employed in the construction of these receptors.

Upon expression of these- novel hybrid multispecific receptors containing both an MSECD and an ICD in a host cell, one may modulate proliferation-and effector signaling by adding either an extracellular inducer, an intracellular inducer or a combination of these two different inducer molecules.

The extracellular inducer-responsive clustering domain (ECD) may be obtained from any of the wide variety of extracellular domains of eukaryotic transmembrane proteins, secreted proteins or-other proteins associated with ligand binding and/or signal transduction.

The ECD may be part of a protein which is monomeric, homodimeric, heterodimeric, or associated with a larger number of proteins in a non-covalent or disulfide-bonded complex.. -To create the multispecific binding.domains of the present invention, two or more individual binding domains are connected to each other on a-single protein, either covalently or noncovalently. An -oligo- or polypeptide linker region may be used to connect these domains to each other.

In particular, the ECDs may consist of monomeric, dimeric or tetrameric immunoglobulin molecules, or portions or modifications thereof, which are prepared in the following manner.

The full-length IgG heavy chain comprising the VH, CH1, hinge, and the CH2 and CH3 (Fc) Ig domains may be fused to a EFSD, PSD or ICD of an MSCR via the appropriate transmembrane domain. If the VH domain alone is sufficient to confer antigen-specificity (so-called "single-domain antibodies"), homodimer formation of the MSCR is expected to be functionally bivalent with regard to antigen binding sites. If both the VH domain and the VL domain are necessary to .generate a fully. active antigen-binding site, both the VHcontaining MSCR and the full-length IgL chain are introduced into :25 cells to generate an active antigen-binding site. Dimer formation resulting from the intermolecular Fc/hinge disulfide bonds results in the assembly of MSCRs with extracellular domains resembling those of IgG antibodies. Derivatives of these MSCRs include those in which only non-Fc regions of the heavy chain are employed in the fusion. For example, the VH domain (and the CH1 domain) of the heavy'chain can be retaind in the extracellular domain of the MSCR, but MSCR.dimers are not formed. As above, -:the full-length IgL Schain can be.introduced into cells to generate an active antigenbinding site.

As indicated, the ECD may consist of an Ig heavy chain which may in turn be covalently associated with Ig light chain by virtue of the presence of the CH1 region, or may become covalently associated with other Ig heavy/light chain complexes by virtue of the presence of hinge, CH2 and CH3 domains. The two heavy/light chain complexes '0 may have different specificities, thus creating a MSCR which binds two distinct antigens. Depending on the function of the antibody, the desired structure and the signal transduction, the entire chain may be used or a truncated chain may be used, where all or a part of the CH1, CH2, or CH3 domains may be removed or all or part of the hinge region may be removed.

Because association of both the heavy and light V domains are required to generate a functional antigen binding site of high affinity, in order to generate an antibody-containing MSCR with the 20 potential to bind antigen, a total of two molecules will typically need to be introduced into the host cell. Therefore, an alternative and preferred strategy is to introduce a single molecule bearing a functional antigen binding site. This avoids the technical difficulties that may attend the introduction and :25 coordinated expression of more than one gene construct into host cells. This "single-chain antibody" (SAb) is created by fusing together the variable domains of the heavy and light chains using an oligo- or polypeptide linker, thereby reconstituting anh antigen binding site on a single molecule..

Single-chain antibody variable fragments (SAbFv) in which the Cterminus of one variable domain (VH or VL) is tethered to the Nterminus of the other (VL or VH, respectively), via a oligo- or polypeptide linker, have been developed without significantly disrupting antigen binding or specificity of the binding (Bedzyk et Al. (1990) J. Biol. Chem.-, 265:18615; Chaudhary et al. (1990) Proc.

Natl. Acad. Sci., 82:9491). The SAbFvs used in the present invention may be of two types depending on the relative order of the VH and VL:domains: VH-l-VL or VL-l-VH (where represents the linker). These SAbFvs :lack the constant regions (Fc) -present in the -heavy and .light chains of the native antibody. In another aspect of the present invention, the SAbFv fragment may be fused to all or a portion of the constant domains of the heavy-chain, and 0 the resulting ECD is joined to the EFSD, PSD or ICD via an appropriate transmembrane domain that will permit expression in the host cell: The resulting MSCRs differ from the SAbFvs, described above, in that upon binding of antigen they initiate signal transduction via their cytoplasmic domain.

Single-chain derivatives of T cell receptors (SCTCRs) in which the variable regions of the T cell receptor a and 0 chains are joined together by an appropriate oligo- or polypeptide linker may also be employed as one or more of the ECDs contained within an 20 MSCR.

To aid in the proper folding and efficient expression of the MSCRs, including MSCEFRs, MSCPRs and hybrid MSCRs, the antibodyderived ECDs may be connected at their C-terminal end to one of a number of membrane hinge regions which are a normal part of membrane-bound immunoglobulin molecules. For example, the eighteen amino acids of the IGHG3 M1 exon may be used (Bensmana and Lefranc, Immunogenet., 32:321-330 (1990)) The TM domain is attached to the C-terminal end of the membrane'hinge. It.is also contemplated that membrane hinge sequences may be used to connect non-antibody derived ECDs -to the transmembrane domains to increage CPR expression.- Diabodies -may also be used as ECDs ini the present invention.

Diabodies contain -two chimeric immunoglobulin chains, one of which comprises a VH domain connected to a VL domain on the same polypeptide- chain (VH-VL) A linker-that is too short to allow pairing of the VH and VL domains on this chain with each other is used so that the domaiis will pair with the complementary VH and VL domains on the other chimeric immunoglobulin chain to create two antigen-binding sites olli er et al., Proc. Natl. Acad Sci.

90:6444-6448 (1993)). As described above, one of these chains is linked to the membrane hinge and/or the TM domain, which in turn is linked to the EFSD, PSD and/or ICD. The other chain (not connected covalently to a EOD', PSD tor iD) will be co-expressed in the same cell to create a MSCR with a diabody ECD which will respond to two different extracellular inducer molecules.

In the present invention, the SCFv fragment or the single domain 2O antibody may be fused to all or a portion of the constant domains of- the heavy chain or lA4iqit chain before being linked to each other to- form the -m4t i fic BCD. The MSECDs Used in -the present invention may comiprise' to C-terminal) a SCFv fragment linked to another SCFv doiaiin that in.turn is linked to all or part of the constant domains (CHil, hinge, CH2 and CH3).

-To aid in the proper iding and efficient expression of the MSCRs of the_ present invention,.: the antibody-derived ECDs may be connected at their -C-terminal end to one of a number of-"mimbrane hinge regions-1 ywhich4 a; normal part of minembrane-bound immunoglobulin *ot oe TMh-;Idomain is attached -to the Cterminalend For example, eighteen amino acids -from the N-teeid -of 'the IGHG3 M1- exon may -be used (Bensmana and Lefranc, Tmihnunoaenet., 32:321-330 (1990)). It is contemplated that membrane.-hinge sequences may be used to connect non-antibody derived ECDs to the transmembrane domains to increase CPR and CEFR expression. In the present invention, the membrane hinge region may also be employed like other linker regions, eg., be.:attached on either side; of a TM,-PSD, or ECD.

Ligand-binding domains from naturally occurring receptors may also be used as ECDs to prepare -the MSECDs of the present invention, where, the receptors are surface membrane proteins, S including cell differentiation antigens such as CD4 and CD8, cytokine or hormone receptors or cell adhesion molecules such as ICAM or LFA-1. The receptor may be responsive to a natural ligand, an antibody or fragment thereof, a synthetic molecule, drug, or any other agent which is capable of inducing a signal. In addition, either member of a ligand/receptor pair, where one is expressed on a target cell such as a cancer cell, a virally infected cell or an autoimmune disease causing cell, may also be used as an ECD in the present invention. In addition, the receptor-binding domains of soluble ligands or portions thereof could be employed as ECDs in the MSECDs of the present invention.

In addition, binding portions of antibodies, cytokines, horm6nes, or serum proteins can be used. Furthermore, secreted targeting molecules such as IL-14 or the soluble components of the cytokine receptors such as..IL-6R, IL-4R, and IL-7R can be used as ECDs (Boulay and Paul Current Biology 3: 573-581, 1993).

Where a receptor is a molecular complex of proteins, where only one chain has the major -role obinding to :the ligand, it will usually be desirable to use solely the "extracellular portion of the ligand binding protein. Where the textracellular portion may complex with other- extracellular-portions of other proteins or form covalent bonding through disulfide linkages, one -may also provide S for the .formation of- such dineric 'r multimeric- extracellular regions. Also, where the entire extracellular region is not required, -truncated portions thereof may be employed, where-such truncated portion is functional. In particular, when the extracellular region of CD4 is employed, one may use only those sequences required for binding of gpl20, .the HIV envelope glycoprotein. in the case in which immunoglobulin-derived sequences are used to create a muitispecific ECD, one may simply use the antigen binding regions of the antibody molecule and dispense with the constant regions of the molecule (for example, the Fc region consisting of the CH2 and CH3 domains).

To create the multispecific ECDs of the present invention, two or more individual ECDs are connected to each other, either *C covalently or noncovalently, on a single protein molecule. An oligo- or polypeptide linker, an Fc hinge or membrane hinge region may be used to connect these domains to each other. The MSECDs of the present invention may-comprise two or more of the different ECDs described above connected together in different combinations.

For example, two or more ECDs'containing immunoglobulin sequences, SCFvs and/or single-domain antibodies) may be linked to each JO other. In another example, two or more ECDs from membrane proteins cytokine receptors and/or CD antigens) may be linked to each other. In yet another example,. a- MSECD may consist of a mixture of ligand-binding domains and immunoglobulin-derived domains eg., an ECD from CD4 may be. joined to a SCFv.

-The proliferation signaling domains (PSDs) that comprise the SMSCPRs and hybrid MSCRs of -the present invention ,may be obtained from the- cytoplasmic a-signal-transducing -domains of: the cytokine/hematopoietin receptor superfamily. The -members this mammalian receptor superfamily, can transduce proliferative .signals in a wide variety of cell types. The cytoplasmic domains of the signal-transducing subunits may .contain conserved motifs that are critical for the transduction of proliferative .signals (Bazan, Current Bioloav, 3:603-606 (1993); Boulay and.Paul, Current :573-58.,(1993); -Wells,: Current nCl-Bog 6:163-173--.(1994); -Sato andA- iyajima,- cuiint O'jpiiofn Cell 'Bi1Q 6:174-179 (1994); -Stahl- and'Yahcopbul0s2IriU,-174:587-59O (19 93); Minami et' a 'Ann". ReV iiun6!jf;11ii:245-267 *(1993); _Kihiot et'a.,Qe,' (199 The s-ignal-transducing components of these _cytokine..receptors to ,be used -as.PSDs in the present invention include,.but are-,not limited to, Interleukin-2 receptor 0 (IL-2RO3), IL-2Ry, IL-3R3,.IL-4R', IL-5Ra, IL- 5RJ3, IL- 6R,.

IL-6R gpl'36,' IL-7R, IL-9R, IL-12R, IL-13R, IL-15R, EPO-R J (erythropoietin receptor), G-CSFR (granulocyte -colony stimulating factor receptor), GM-CSFRa (granulocyte macrophage colony stimulating factor receptor ax), GM-CSFR3, LIFRa .(leukemia inhibitory factor receptor GHR (growth hormone receptor), PRLR **(prolactin receptor) CNTFR (ciliary neurotrophic f actor receptor) OSMR (oncostatin M receptor) IFNRa/3 (interferon a/3 receptor) IFNRy, TFR (tissue factor. receptor), and TPOR (thrombopoietin or mM.l-ligand receptor) (Minami et al., J. Immunol., 152:5680-5690 (1994); Boulay and Paul, Current-Biology 3:573-581 (1993); Wells, Current Opinion in Cell Biolo, 6:163-173:(1994)).

SThe proliferation signaling-. domains (PSDs) that* comprise the MSCPRs and hybrid.MSCRs.of the present invention .may'be- -obta ined from the signal-transducing. domains; of the -tyrosin6 kiniase 'growth factor receptor superfamily. or from oncogenes or proto-oncogenes which- are related_.to this-growth-. factor family, (schlbsginger and Ullrich, Cgfl,:61:203-212 -(1990) J-Ullrich'aiid Schiless inger,: eur 9:383-391h(992) .Th-"eimbeiz.s' of rniam-mallil receptor superfamily cii -transdu-ie' pro ieraitiVe 'si.h&ls in Th~wie var iety -of b -ell types The cytoasmic- doiairis 'of -thd' sigial-transducing -subunits'66ntainiyrosine kinas ordainis thit are cri c se crtia for the ofb -iL1 h 7- transducti~h-o rlf~tv-ias Ti growth factor ep or,-poo ncge-s recetor ~and oncogen~es tobe used as PSDs in -the peetienininclude, but-are nbt limited t -o pi d e--rmiz g i .o wth factor receptor.(EGF-R), HER2/neu, HER3/c-erbB-3, Xmrk, InstHin-R, IGF-1-R (insulin-like growth factor-.1 receptor) IRR, PDGF-R-A (platelet-derived growth factor receptor-A), PDGF-R-B (plateletderived growth *factor receptor-B) CSF-1-R -(colony-stimulating factor-i receptor), c-kit, FGF-R -_(fibroblast growth factor receptor), acidic FGF-R, and basic FGF-R -(Ullrich and Schlessinger, CIll, 61:203-212 (1990)).- The proliferation signaling domains employed in constructing the MSCPRs and hybrid MSCRs of the present invention may also be obtained from any member of the Janus or JAK:eukaryotic family of tyrosine kinases, including Tyk2, JAK1, JAK2, JAK3 and Ptk-2.

Members of the Janus kinase family are found in all cell types.

They associate with various signal transducing components of the cytokine receptor superfamily discussed above and respond to the binding of extracellular inducer by the phosphorylation of Styrosines on cytoplasmic substrates (Stahl and Yancopoulos, Cell, 74:587-590 (1993)). They are thus an integral part of the control of cell proliferation in many different kinds of cells. The members of this family are marked by similar multidomain structures and a high degree of sequence conservation. Unique among tyrosine kinases, the Janus kinase family may have two non-identical tandem kinase-like domains, only one of which may have catalytic activity (Firmbach-Kraft et al., Oncodene, 5:1329-1336 (1990);-Wilks et al., Mol.- Cell. Biol., 11:2057-2065 (1991); Harpur et al., Oncooene, 7:1347-1353 (1992)) e kinase activity of the Janus kinases is usually activated after the binding of ligands to their associated cytokine family receptors and the oligomerization of the receptors (Stahl and Yadricopoulos, C11, 74:587-590 (1993)). This activation, in turn, triggers the initiation of intracellular signaling casdcdes (Witthiihn et al., Nature, 370:153-157 (1994); Russell et al., Science, 366:1042-1044 (1994); Kawamura. et al., Proc. Natl.

Acad. Sci., 91:6374-6378 (1994); Miyazaki et Science, .2W6:1045-.1047: (1994) Johnston :et. al, Nature, 370:151-153 -1994); Asao et-al., FESq Lept tern, -351:201-206 -The effector function signaling -domains (EFSDs), employed in the MSCEFRs and hybrid MSC2Rs oe teprsn ivetion may be derived from a proteinw-.hic6h ii- ii .rt&ac6tivate- various second messenger pathways. One pathway of interest is -that involving phosphatidylinositol-specific phospholipase hydrolysis of phosphat idyl inds itol 5 -biphosphate, and production of, inositol- 3 1,4,5-trisphosphate and diacylglycerol. The calcium mediated patway te troie an&serine/threonine kinase and-phosphts -pathway, the adey t, cyclase, and the guanylate cyclase pathways may also be sec~ii:or meso,,rsenger ,.'pathwa ys. EFSDs of interest include proteins with 41A M iotf (eh NiZ, 338:383-384 (1989); WeskU 32~2'C~,freample, the chain of the T-cell receptor, the il -chain, which -di ff ers from the chain only in its most C-terminal exon as, a result' of alternative splicing of the nRNA, the y and 13subunit of the FceRl- receptor, ,the MB1 (Iga) and B29 (Igo3) chains of the B -cell. receptor, and the 6, y, :0 and e chains of the T-cell receptor (CD3 chains), other proteins homologous to the above. -protein. subunits including synthetic Amt. MO A.? polypeptides with ARq4tus ndsc other cytoplasmic -regions which are capable f aittng a signal as.. result of interacting with other. p-rote.nsh cg p,!Ole of binding. to.-.an -inducer (Romeo et al., C-P110 6t;O84-4 7 (92 Weiss, -l 73:209-212 (1993)). The syk 'fam "ily ef 'tyros'ine kinases may also-.be -used as effector function signt ig, 4 L in the sn ivnin. h clui~teinig-do 0i e4 fro6m Syk and ZAP-70 leads, to. the -activation of T cel y~~i activity (Kolanus et. Crf 74:'171-183 na 1i te arc family of-tyros-ine kinases (Lck,' pFn, Lyn -j Today 15:225-234 (1994) and jfD5 and CD28, which -are involved -in T:cl irni eue as EFSDs"'ih the present inetion. A-ub'-ofES~o functina rgmn&rmunt thereof may be employed, generally ranging from -about 50 t- 1500 amino acids each, where-the entire naturally occurring cytoplasmic region may be-employed or only an active portion thereof.

The intracellular clustering domain (ICD)-.can be-obtained from the inducer binding domains of:a variety of -intracellular proteins.

For example, -eukaryotic steroid receptor molecules can be used as ICDs the receptors for estrogen, progesterone, androgens, glucocorticoids, thyroid hormone, vitamin D, retinoic acid, 9-cis retinoic acid and ecdysone). In addition, variants of steroid and other receptors which fail to bind their native inducer, but still bind to an antagonist, can be prepared by one skilled in the art and used to make the CPRs of this invention. For example, a Cterminal deletion mutant of the human progesterone receptor, which fails to bind progesterone, can be clustered by the addition of progesterone antagonists, including RU 486 (Wang et al., Proc Natl S Acad Sci 91: 8180-8184, 1994). Binding domains from the eukaryotic immunophilin family of molecules may also be used as ICDs.

Examples include but are not limited to members of the cyclophilin family: mammalian cyclophilin A, B and C, yeast cyclophilins 1 and 2, Drosophila cyclophilin analogs such as ninaA; and members of the FKPB family: the various mammalian isoforms of FKBP and the FKBP analog from Neurospora (Schreiber, Science, 251:283-287 (1991), McKeon, ell., 66:823-826, -(1991), Friedman and Weissman, Cell, 66:799-806,- (1991), Liu et al., C~eU, 66:807-815 (1991)). For example, the inducer binding portion of the immunophilin, FKBP12, which can be clustered in the cytoplasm by the addition of FK1012, a synthetic dimeric form of the immunosuppressant FK506 (Spencer et al. Sciece 262: 1019-1024 (1993) can be used as an ICD.

The transmembrane domain may be contributed by the protein Scontributing the minultispecific extracellular inducer clustering domain, the protein contributing the effector function signaling domain;.. the r protein contributing the proliferation signaling portion, or by a totally different protein. For the most part it will be convenient to have: the -transmembrane domain naturally associated with one of the domains. In some cases it will be desirable to employ .the transmembrane domain of the n or FceRly chains which contain a cysteine residue capable of disulfide bonding, so that the resulting chimeric protein will be able to form disulfide linked dimers with itself, or with unmodified versions of the 5, n or FceRly chains or related -proteins. In some instances, the transmembrane domain will be selected or modified by amino acid substitution to avoid binding of such domains to the transmembrane domains of the same or different surface membrane proteins to minimize interactions with other members of the S receptor complex. In other cases it will be desirable to employ the transmembrane domain of n, FceRl-y and MB1 (Iga), B29 or CD3-y, 56,or e, in order to retain physical association with other members of the receptor complex.

Extracellular inducer molecules employed in the present invention 20 can be antigens which bind the immunoglobulin-derived ECDs, described above. These may include viral proteins, gpl20" and gp41 envelope proteins of HIV, envelope proteins from the Hepatitis B and C viruses, the gB and other envelope glycoproteins of human cytomegalovirus, the envelope proteins from the Kaposi's sarcomaassociated herpesvirus), and surface proteins found on cancer cells in a specific or amplified fashion, the IL-14 receptor, CD 19 and CD 20 for B cell lymphoma, the Lewis Y and CEA antigens for a variety of carcinomas, the Tag72 antigen for breast and colorectal cancer, EGF-R for lung cancer, and the HER-2 protein which is often amplified in human breast and ovarian carcinomas). For other receptors/ the receptors and ligands of particular interest are CD4, where the ligand is the HIV gpl20 envelope glycoprotein, and otherviral receptors,; for example CAM, which is the receptor for the human rhihovirus, and the related receptor molecule for poliovirus..

The intracellular Ainducer *molecules -employed in the present invention must be molecules which can be delivered to the cytobplasm. For example,- the inducer may be lipophilic, nor be transported-into the cell by active transport or pinocytosis, by fusion with a liposome carrying the- inducer, or by semipermeabilization of the cell membrane. The intracellular inducers cluster the ICDs which make up the CIPRs of the present invention.

~Examples of inducers include, but are not limited to synthetic dimeric molecules such as FK1012 (Spencer et al., Science, 262:1019-1024 (1993)) or dimeric derivatives of the binding domains .:of other immunophilin binding molecules such as cyclosporin, rapamycin and. 506BD (Schreiber, Science, 251:283-287 (1991), McKeon, Ce.1, 66:823-826, (1991)) Steroids, such as estrogen, progesterone, the androgens, glucocorticoids, thyroid hormone, vitamin D, retinoic acid, 9-cis retinoic acid or ecdysone, or antagonists or derivatives of these molecules may also be used as intracellular inducer molecules. In particular, the steroid *fee antagonist RU 486 may be used (Wang et al., Proc. Natl. Acad. Sci., 91:8180-8184 (1994)).

9* In -'some instances, a few amino acids at the joining region of the natural protein domain may be deleted truncated), usually not emoe thtn 30, more usually not more than 20. Also, one may wish to introduce a small numbr of amineacids At the borders, usually not more than -30, more usually not more thah' 20 (linkers or the membrane hinge region). The'deletion or insertion famiino acids 11will uiualy be as a result of the heeds of the coisitruction, p1:ovidigfor convenient rest'riction sites, 'ease of manipulation, provemtinlevels of expression, proper' fojldingof -'the molecule or te ike. In addition, one m.y.ish to subs'titute one or more amino acids with a different amino acid a modification) for similar reasons, usually not substituting more than about--five amino acids in any one domain. The PSDs and EFSDs will generally be from about 50 to -1500 amino acids depending ,upon the particular domain employed, while the transmembrane domain will generally have from about 20 to 35 amino acids. The individual ECDs will-generally be .from about 50 to 1500 amino acids, depending .on the particular domain employed. The MSECDs will usually contain between two and twenty ECDs, more preferably between two and ten ECDs, and most preferably between two and five ECDs.

Normally, the signal sequence at the 5' terminus of the open reading frame (ORF) which directs the chimeric protein to the surface membrane will be the signal sequence of the ECD. However, in some instances, one may wish to exchange this sequence for a different signal sequence. However, since the signal sequence will be removed from the protein during processing, the particular signal sequence will normally not be critical to the subject invention..- 0 The chimeric construct, which encodes the chimeric protein according to this invention will be prepared in conventional ways.

Since, for the most part, natural sequences may be employed, the natural genes may be isolated and manipulated, as appropriate, so as to allow for the proper joining of the various domains.- Thus, one may prepare the truncated portion,of the sequence by employing the polymerase chain reaction (PCR), using appropriate .primers which result in deletion of the undesired portions of the gene.

Alternatively, one may use primer repair, where the sequence of interest.may be cloned in an appropriate .host. In either case, primers may be .employed which .result in .termini, .which allow for annealing of the sequences. to result in the desired -open reading frame encoding the chimeric protein. Thus,. the sequences may be selected to provide for restriction sites.which are blunt-ended, or have complementary overlaps. i If desired, the multispecific -extracellular:'diomaiii may also include the transcriptional initiation region, which will allow for .expression in the. Xa6 get'. host.- Alternat ively,:> one may wihto.

-for. a.-dif feremt-trascriptional initiation: -region,: wich -may: al low-f or-constiti z-r'inducible expression, depending upon the .target host,-c .the Purpose for the .intro'dufction: of :the'subj ect .chimeric protein-.into such host, -the :levelgl-of -expression 'desired, the nature of the target -host, and the like. Thus, one may provide 7' for expression upon differentiation or maturation' of the -target host, activati on-.of..th4etarg.##Ot #oat,:.-or the like.' A .wide variet 'y f -pirbio -Oip have 'been described in. the literature,, which"476 c*are '.otitu,tiLve. or inducible, where induction may be. assoqiated,L -cell 'type or a spedcif ic level of expression. Altentrrl, i.uer of viral promoters are known which may also find use.-Promoters of interest include the 13-actin promoter,: SV40 early-OW.d late promoters, immunoglobulin promoter, human cytomegalovirus promoter, and the Friend spleen 'focus-f orming 420 virus promoter. The promoters may or may not be associated with enhancers, where the enh~ner may be naturally associated -with' the particular promotert.-~'' Pr,, wihadfeetp oter.' radh frm -my- ba .The sequence, of; the on redn rm a e obaned -f rom genornic DNA,. cDNA, rbesnesized, or combinat ions "thereof.

Depending upon: the, qte enomic DNA and )tenumber of introns,,one may wis -Q Aoracmiaion thereof; In many instaIhces;*-it. i. iironis stabilize -the mRNk.-." Also, one may provide for nan- COPMi regioqns which stabilize the mRNUA.

A~ terminatipn'. trgQZ, I e~rvdd. othei cytoplasmic doan weel ion may-be natu-ra1lly", aifociated with he.Cyp 4~ e-drved, f kom "a-4 dif ferent -source. For- tC ~st et r-the .termination -regions -&re not 271 critical and a wide variety of termination regions may be employed without adversely affecting expression. The, various manipulations may be carried out in Yitro or may be introduced into: vectors .for cloning in an appropriate -host, e.g., coli. Thus, :after each manipulation; the resulting construct from joining -of -the. DNA sequences may be cloned into an -expression vector. The sequence may. be screened by restriction analysis, sequencing, or the like to insure that it encodes the desired chimeric protein.

The chimeric construct may be introduced into the target cell in any convenient manner. Techniques include calcium phosphate or DEAE-dextran mediated DNA transfection, electroporation, protoplast fusion, liposome fusion, biolistics •B using DNA-coated particles, and infection, where the chimeric construct is introduced into an appropriate virus retrovirus, S adenovirus, adeno-associated virus, Herpes virus, Sindbis virus, papilloma virus), particularly a non-replicative form of the virus, or the like. In addition, direct injection of naked DNA or proteinor lipid-complexed DNA may also be used to introduce DNA into cells.

*c *q Once the target host has been transformed, integration will usually result. However, by appropriate choice of vectors, one may provide for.:episomal-maintenance. A large number of vectors are known which :are based on viruses, where the copy number of the virus -maintained .in the cell is low enough -to maintain the viability of the cell.- Illtstrative vectors include SV40, EBV'and BPV It is also contemplated that the introduction :of the chimeric constructs. of:-the present. invention into: cells may result in the ,transient -expression of -the MSCRs. Such transient .expression may be preferable :if a: short-term therapeutic effect is 'desired.

Unstable replication or the absence -of DNA replication may result, for example, from adenovirus infection or transformatiobn with naked DNA. The MSCRs of the present invention are employed in a wide variety of target host cells, normally cells from vertebrates, more particularly, mammals, desirably domestic animals 'or primates, particularly humans. In particular, the subject invention may also find application in regulating the expansion and effector activity '7 of lymphoid cells, T lymphocytes (including CD8+ T cells, CD4+ T cells, cytotoxic T cells and helper T cells), B lymphocytes, cytotoxic lymphocytes (CTL), natural killer cells tumorinfiltrating-lymphocytes (TIL) or other cells which are capable of killing target cells when activated. In addition, suitable host cells in which to introduce MSCRs of the present invention include hematopoietic stem cells, which develop into cytotoxic effector cells with both myeloid and lymphoid phenotype including granulocytes, mast cells, basophils, macrophages, natural killer (NK) cells and T and B lymphocytes.

Once one has established that the transformed host cell- expresses the MSCRs of the present invention in accordance with the desired regulation and at a desired level, one may then determine whether the MSCPRs or hybrid MSCRs are functional in the host cell in -providing for the desired proliferation signal. One may use established methodology for measuring proliferation to verify the functional capability of the .above -MSCRs. .The 'prol iferative response of cells can be measured by a variety -f techniques known -to those skilled in the art. For example, DNA synthesis' can be measured by the -incorporation of 'either -tritiatd: thymidine or _orotic acid. :The incorporation of bromodeoxyuridine into newly synthesized DNA can -be measured -by immunological -staining and the detection- of: dyes, -or::by ELISA (Enzyme-linked -immunosorbent assay) (Doyle_ et al., Cell and Tissue -Culture: Laboratory Procedures, Wiley, Chichester, England, (1994)). The mitotic-kndex of -cells can be. determined by -staining and microscopy, by the fraction labeled mitoses method or by FACS analysis (Doyle et al., isura, (1994); Dean, Cell Tissue Kinet. 13:299-308 (1980); Dean, Cell Tissue Kinet. 13:672-681::(1980)).- The increase in cell size which accompanies progress.through the cell cycle can be measure by centrifugal elutriation (Faha .et. al., J Virol.- 67:2456-2465 (1993)). Increases in the number of cells may also be measured by counting the cells, with or without the addition of vital dyes. In addition, signal transduction can also be measured by the detection of phosphotyrosine, the in vitro activity of tyrosine kinases from activated cells, c-myc induction, and calcium mobilization as S described in the Examples below. In the case of MSCRs containing EFSDs, one may determine whether the host cell has been provided with an effector signal in a variety of ways well known to those skilled in the art, depending on the EFSD and the cell type. For example, the activity of MSCEFRs and hybrid MSCRs in signaling cytotoxic effector function in engineered cytotoxic T cells can be measured by the release of sCr from labeled cells displaying 2D extracellular inducer molecules, while the activity of MSCEFRs and hybrid MSCRs in signaling helper effector function in engineered j* helper T cells can be measured by the release of cytokines in the presence of inducer.

In the present invention, a host cell may express two different MSCRs, one containing an:effector function signaling domain and the other containing a proliferation signaling domain MSCEFR and MSCPR). Both MSCRs maycontain the: same MSECDs. Alternatively, a host :cell may express hybrid MSCR combining -both signaling domains (EFSD and.:PSD) _on the :same chain. In both situations, the binding of-:an extracellular inducer molecule to the MSECD will Sstimulate host cells to:act. as therapeutic -agents at 'the "same time they are expanding inr--response' to- binding to an extracellular inducer molecule, _gpl20 for HIV or cancer-specific antigens.

The specific .targets of cells 7expressing the multispecific chimeric. receptors (MSCRs) of the 'present invention- include diseased cells, such as cells infected with HIV, HTLV-I or II, cytomegalovirus, hepatitis A, B or C viruses, ijcobacterium avium, Mycobacterium tuberculosis, Mvcobacterium leprae etc., neoplastic cells, or autoimmune disease-causing cells where the diseased cells have a surface marker associated with-the diseased state. Since MSCRs of the present invention have more than one antigen-binding and/or ligand-binding domain, the present invention may have advantages- over monospecific chimeric receptors in treating human diseases such as infectious disease, cancer and autoimmune disease.

S" The application of the present invention for targeting more than one epitope of a given pathogen, more than one pathogen, or a 15 heterogeneous population of disease-causing cells that may be found in malignant or autoimmune disease,: is designed to have the therapeutic benefit of a combination. therapy, while having the safety profile and pharmacological properties of a single therapeutic agent.

The MSCRs of the present invention are envisioned as a strategy for overcoming the ability of viruses such as HIV .to become drugresistant, and may be applicable in treating, other diseases in which drug-resistance or antigenic variation is a significant problem (eg. cancer, bacterial and parasitic.infections). Through the application of the-present invention, the :probability that a given viral variant will eventually arise which is:not targeted by a chimeric receptor expressing cell would, decrease .in -direct relation to .the number of antigenic specificities .recognized by the multispecific chimeric receptor. _In particular, -single-chain antibodies which recognize many. different. HIV :antigens can be employed to create.MSCRs to use as anti-AIDS therapeutics. Similar strategies can be applied for other types of pathogens, such as chronic or recurrent bacterial infections, for which tht-the problem of drug .:resistance-in the-face -of. ongoing antibiotic therapy: is particularly The-HSCRs of the t;ivni~ myalso -be: useful in the development of-chimeric receptor-expressing cells for the treatment of multiple .infections,. -'For example-, 'A:an MSCR-containing T cell which,.recognizes HIV as well as another opportunistic pathogen such as. CMV, herpesviruses, etc. can -also' be used -as- an anti-AIDS therapeutic.. since s~opd.--mse individuals are also susceptible to vg-o' c$e4 "nfcios neutrophils mayas muW peiticcr64 be armed with ofl i'i reeptors which recognize the mjrclasses of e~atp~t~o~s The.MSCRs of the pr~ert ixntinmy~s be- usful in treating *viral and bacterial., inifecdtions where aniftigen'ic variation and the existence of multiple ftiains -has limited traditional,- single agent therapies. For example, MSCRs containing kSECDs which recognize multiple antigens or epitopes. from a single pathogen can be used to treat HIV, hepatitip. d n C viruses, Kaposils sarcomaassociated herpes. vir, I0 S 6kle viruses, Herpes Zoster virus,. CMV, 'papiJlloma viAi* respirat o~kr s4yn cyt i a 1virus, and inf luenza viruses..' The MSRcnann cisof "the present invention may alsoft hdappliation rO' thrap, Since resistance to chemotherapeutic agen6 h4'b an impbrtindtzobstacle' in the successful :ueo ~imens andcminto ch~othrapyhas:b~~ igifficai_'-side-efet of tbhe ag en ts U11 Oi res is tance., In *-addition,' -the -HAL6'll increases the potentialf or:;th genic:-vats-or for resistance due to antigen or epitope loss. As an example, B cell lymphoma can be treated with MSCRs which recognize -more than one pan-B cell surface marker such as CD19, CD20 or CD22, and/or-markers specific for malignant but not resting B cells, such as the interleukin-14 receptor.

The present invention may also be useful as a cancer therapy in the MSCR's ability to target cytotoxic cells to cancers such as melanoma in.which the host can mount a tumor antigen-specific T cell response. The availability of CTL clones from patients to MHC-restricted epitopes on melanoma has permitted the molecular analysis of the T cell receptors responsible for tumor killing (Mandelboim et al., Nature, 369:67-71 (1994); Cox et al., Science, 264:716-719 (1994)). However, in employing a single-chain T cell 15 receptor (scTCR) of sufficient affinity to redirect cytolysis, a given chimeric receptor would still be MHC-restricted and active in only a small fraction of the patient population. Multispecific scTCR's could be developed which would recognize the relevant tumor-specific peptide antigen in the context of many HLA haplotypes. Such 'semi-universal' receptors may be capable of Sdealing with a particular disease target in most affected individuals.

The MSCR technology may also be used to treat autoimmune disorders such as multiple sclerosis, rheumatoid arthritis, type 1 diabetes mellitus, myasthenia gravis and Graves' disease, where the autoreactive'T 'cells and/or B cells re oligo- or polyclonal.

-immune cells bearing' MSCRs which recogniize the members of the restricted T cell receptor (TCR) repertoire expressed by the disease-causing -autoreactive T cells can be. used to treat autoimmune disease. In particular, restricted TCR expression has been observed for the T -:cells .found at -the sites of disease- in multiple.sclerosis (brain) (Oksenberg et al., Proc. -Natl. Acad.

Sci.,-86:988-992 (1989)) and rheumatoid arthritis (joint synovium) (Stamenkovic-et al., Proc. Natl. Acad. Sci., 85:1179-1183 (1988)).

High-titer retroviral producer lines are used to transduce the chimeric proliferation receptor constructs into autologous or allogeneic human T-cells, hematopoietic stem cells or other cells, described above through the process of retroviral mediated gene transfer as described by Lusky t al. in (1992) Blood 80:396. In addition to.the gene encoding the chimeric proliferation receptor, additional genes may be included in the retroviral construct.

These include genes such as the thymidine kinase or cytosine deaminase genes (Borrelli et al. (1988) Proc. Natl. Acad. Sci. USA 85:7572) which acts as a suicide gene for the marked cells if the patient is exposed to gancyclovir or 5'-fluorouracil respectively. Thus, if the percentage of marked cells is too high, gancyclovir or 5FU may be administered to reduce the percentage of cells expressing the chimeric receptors. In addition, if the "2D percentage of marked cells needs to be increased, the multi-drug resistance gene can be included (Sorrentino et al. (1992) Science 25.7:99) which functions as a preferential survival gene for the S marked cells in the patients if the patient is administered a dose of a chemotherapeutic agent such as taxol. Therefore, the percentage of marked cells in the patients can be titrated to obtain the maximum therapeutic benefit. In addition, high-titer retroviral, adenoviral or other viral or non-viral producer lines may be. used -to _transduce -the .chimeric proliferation receptor constructs into.autologous -or allogeneic nerve cells, hematopoietic cells including stem cells, islets of Langerhans, keratinocytes, muscle cells or other .cells following the methods of retroviral, adenoviral or other viral or non-viral meiediad t gene transfer as described in Finer et al., Blood 83:43-48 (1994) and U.S. Patent Application No. 08/333,680. Similar-tM the procedure described: above, other :.genes may be included in the retroviral, adenoviral or other viral or non-viral constructs in addition to the chimeric proliferation receptor in the recipient cell. After introduction of the construct into the cell type of interest, the cells may be expanded in an appropriate medium well known in the -art and used in a variety of ways previously described.

oo The following examples are:by way of illustration and not by way of .limitation.- Example 1: Multispecific antibodies comprising multiple antibody extracellular clustering domains and an Ig-Fc effector function domain.

S Multispecific antibodies were created which contain two or more *extracellular clustering domains (ECDs) which are derived from antibodies and/or single-chain antibodies, or modifications thereof. The constructs described in this example contain two SCFv domains which were derived from the 98.6 human monoclonal antibody (Mab), which is specific for the HIV-1 gp41 envelope glycoprotein, and the 447D human MAb, which is specific for the HIV-1 envelope glycoprotein. The 98.6 light-linker-heavy (LLH) SCFv consists of (from N- to C-terminus) 1) the VK signal sequence and VK variable domain (residues 1-107 of the mature protein) of the 2 0 98.6

M

Ab, 2) the 14 amino acid L212 peptide linker (Gly-Ser-Thr- Ser-Gly-Ser-Gly-Lys-Ser-Ser-Glu-Gly-Lys-Gly) (SEQ ID NO:23) (Bedzyk et al J.Biol Chem. (1990) 265: 18615-18620), and 3) the VH variable domain (residues 1-113 of the mature protein) of the 98.6 MAb. The 447D light-linker-heavy (LLH) SCFv consists of (from N- to Cterminus): 1) the VA signal sequence and VX variable domain (residues 1-107 of the. mature protein) of the 447D MAb, 2) the 14 amino acid L212 peptide linker, and 3) the VH variable (residues 1- 113 of the mature protein) of the 447D MAb. Alternatively, the 98.6 and 447D SCFv's are created as heavy-linker-light (HLL) constructs in which the heavy chain variable domain precedes the light chain variable domain, connected by a suitable oligo- or polypeptide linker. Both LLH or HLL SCFv derivatives of the 98.6 and 447D MAbs may be constructed using a variety of oligo- and polypeptide linkers. In this example, the 98.6 LLH SCFv was joined at -its. C- terminus ;(residu~e.:113, of the VH variable domain) to -th'e Nterminus of. -the 44DLLH SCFV (res idue -1 _of the. VX -variable domain). Alternatie y, the %447D:,LLH SCFv isLjoihed at its Cterminus. (res idue -113:- of beVHivariable domain) to. :the-.N-termi 'nu. s of the 98.6 LLH SCTFv M' r ae 1 -of -the, Vx -variable domain):. Either LLH SCFv may be suaii~ or by the corresponding HLL SCFv, or modifications thereof. -SCFv'Is.--were. joined either directly, or via an oligo- or: poly~peptide linker. The C-terminus of the .second SCFv was-fused in turn .'to the hinge and Fc region (residues *3Jn 226-478) of the. human IgqG2. heavy -chain. mammalian expression vectors for the 0_*6 Vt/467 SCFv .multispeci-fic. antibodies described in this'." E4I]e 6 cfitrted using plasmid pMSAblint.

This intermediate-- cted fromh thrde DNA fragments: 1) a .3.9 kb vecto:k'.&i~~ $t ed- by didgestion of pIK1.1 with EcoRI 'and. Sf M ro~i h 86LHS~ domain, obtained yo I9 lCM2wt mI modification of the oe.i' v4p ,end with T4 DNA--polymierase and dNTPs to create a blunt entd-, 46oll0w4d by. digestion with EcoRI, and 3) a 1.9 kb fragment encoding. the. 447D' LLH SCFv domain and the human O IgG2 hinge and Fc domains, obtained by'digestion of pIK447DLLHy2 with EcoRI, modif ication of ,the6 c911esive end with T4 DNA polymerase and dN~ to crefo1I;.W0 by digestiz vith Sf ii.

pMSAblint was idenii Or" t. r Ctibn ezZVeanhla Wend used to prepare single-sr44 :~la for 6-ligonucleotide-directed mutagenes is. -In. ed& t orrect expression plasmid was identified by rest4TX4 ndistucre was confre by-DNA sequencing. .c Example IA: BAb(jO*4)~LO-O:....

pIK- SAb (cxgp.4I) -SAb~ 9)y-x ~i ts the expression o f ;a'hybrid protein.- consistin'''-' :Oequence- arid .SCFv -domain of 9_8.,,'6-LL86 joi reioUeA1),.to the Nterminus-of Ptheeide followed by th huma Ig~h ead. Fc domains (residues 226-478).

This plasmid is- -:constructed by oligonucleotide-directed mutageniesis using singlestaddpSbit-N as the- template with .oligonucleotide. 1 -(SEQ 'ID NO: 1) the*: -primer. The* -correct expression-- plasmid -was -:identif ied 'by -colony. hybridization -using L.-oligonucleotide-2-.(SEQ ID.NO:2)as a probe'.-- Example 2.3: SAb(agp41) -L1-SAb(cxgp12O) -Fc pIK-S b(ag4l)-L1-SAb(agpl2O)-Fc directs the expression of a hybrid protein consisting of the Vx signal sequence and SCFv domain of 98.6-LLH Joined at its C-terminus (98.6-VH residue 113) by a 14 ::.:amino acid linker. (Gly-Ser-Thr-Ser-Gly- Ser-Gly-Lys-Ser- Setr-Glu-Gly- Lys-Gly) (SEQ ID NO:23) to the N-terminus of the SCFv-domain of 447D-LLH (447D-V\ -residue followed by the human IgG2 hinge and *Fc domains (residues 226-478). This plasmid is constructed by oligonucleotide-directed mutagenesis using single-stranded pMSAblint DNA as the template with oligonucleotide 3 (SEQ ID NQ:.3)as the primer. The correct expression plasmid was identified by colony hybridization using oligonucle'otide 4 (SEQ ID NO:4)as a probe.

Example 1C: SAb(cgp4l)-L2-SAb(cxgpl2O)-Fc,.

pIK-SAb(cxgp4l)-L2-SAb(agpl2O)-Fc directs the expression of a hybrid protein consisting of the VK signal'sequence and SCFv' -domain of 98.6-LLH-joined-at its C-terminus (98.6-VH residue 113) by a 25 amino acid linker (Ser-Ser-Ala-Asp-Asp-Ala-Lys- -Lys -Asp-Ala-Ala -Lys -Lys -Asp-Asp-Ala-Lys -Lys -Asp-Asp-Ala-Lys Lys-Asp-Gly) (SEQ ID NO:_2 4) to. the N-terminus of -the'*SCFv' domain of 447D-LLH (447D-Vk residue 1) followed by the human IgG2 hinge and Fc domains -(residues 226 r478).: -,This plasmrid is constructed -by :oligoinucleotide-direcdted- :-mutageriesis Ziising single-stranded -,pMSAblift DNA the ->:-template -with" (SEQ ID NO_:5)as -th~e primer.". :The correctexpression plasmid was: identified-by colony -hybridiz-ation-u-sing oligonucleotide 6- (SEQ. IDNO: 6) as *a probe..; Example.ID: SAb(agp4l)-L3-SAb(agp2)-Fc., pI.K-SAb (agp4 1) -L3 7SA~b (agpl 20): -Fc directs the- expresion of a hybrid,: protein "consisting of -the VYK s ignal -sequence and SCFV domain of 98.6-LLH joined at its C-terminus (98.6-VH -re'sidue 113) by a 10 amino acid linker (Asp-Lys-Thr-His-Thr-Ser-Pro- Pro-Ser-Pro) (SEQ ID NO: 25) -to the N-terminus of -the -SCFv domain -of 447D-LLH (447D-VX residue followed *bythe human IgG2 hinge and Fc domains (residues 226-478).'- This plasmid is constructed -by oligonucleotide-directed mutagenesis using single-stranded plSAblint DNA *as .the :template with oligonucleotide 7 (SEQ ID NO:7)as the primer. The correct expression plasmid was identified by 'colony hybridization using oligonucleotide 8 (SEQ ID NO:8) as a probe.

Example lE: SAb (agp4 1) -L4-SAb (agpI20) -Fc a. pIK-SAb(agp4l)-L4-SAb(ctgpl2O)-Fc directs the expression of a hybrid protein consisting of theVK, signal sequence and SCFv domain of 9.8.6-LLH joined at its C-terminus-(98.6-VH residue 113) by an 18 amino acid linker (Glu-Leu-Gln-Leu-Glu-Glu-Ser- Ser-Ala-Glu-Ala-Gln-Asp-Gly-Glu-Leu -Asp) (SEQ ID NQ:26) to the N-terminus of the SCFv domain of 447D-tLH (447D-VX\ residue 1), .:followed-by the human IgG2 hinge and Fc domains (residues 226- .This plasmid is'~constructed by oligonucleotide-directed mutagenesis using single-stranded .,pMSAb1int'_ DNA _as the template with :oligonucleotide 9 :(SEQ -ID NO:9) 'as: the primer.

The correct-'.-expression- _plasidi,-was:- identified by colony hybridization -using-'- oligonucleotide 10 (SEQ ID NO:10).as a probe. 7

I--W.

Example 2: Multispecific antibodies comprising an antibody extracellular clustering domain, a ligand-receptor (CD4) extracellular _clustering domain, and an Ig-Fc -effector function .domain. Multispecific antibodies were created which contain two or more extracellular clustering domains. (ECDs), at least one of which is derived from an antibody and/or single-chain antibody, or modifications thereof, and at least one of which is. derived from a ligand-receptor binding domain, or modifications thereof. The constructs described in this example contain the 447D LLH SCFv domain (described in example and the human CD4 Vl V2 domains, which bind with high affinity to the HIV-1 gp120 envelope glycoprotein.

Alternatively, the 447D SCFv's are created as HLL constructs 'I in which the heavy chain variable domain precedes the light chain variable domain, connected by a suitable oligo- or polypeptide linker. Both LLH or HLL SCFv derivatives of the 447D MAb may be constructed using a variety of oligo- and polypeptide linkers. Other portions of CD4 may be employed in such constructs including the entire CD4 EXT domain (residues 1-371 of the mature polypeptide) as well /as various truncations and/or modifications thereof. In this example, the 447D LLH SCFv was joined at its C-terminus -(residue 113 of the VH variable domain) -to the N-terminus of the CD4. protein (residue 1 of the mature polypeptide). Alternatively,; the CD4 protein is joined at the C-terminus of its.EXTidomain (residue: 371 of the mature polypeptide, or truncations thereof, e.g: residue 180 which resides at the C-terminus of the CD4 VI V2 domains) to the N-terminus of the 447 LLH SCFv (residue 1 of the VA variable domain). The 447D LLH SCFv may be substituted for by the corresponding HLL SCFv, or modifications thereof.

The 447D LLH SCFv was joined to the CD4 protein either directly, or-: via an_ oligo- or polypeptide -linker. -The Cterminus of the CD4 V1 .V2 domains, was fused in turn to the hinge and Fc region (residues 226-478) of the human IgG2 heavy chain. Mammalian expression vectors for the 447D SCFv/CD4 VlV2 multispecific antibodies described. in'this example were constructed using plasmid -pMSAb2int. This intermediate plasmid was constructed from three DNA fragments: 1) a 3.9 kb vector fragment obtained by digestion of pIK1.1 with EcoRI and SfiI, 2) a 1.0kb fragment encoding the 447D LLH SCFv domain, obtained by digestion of pIK447DLLHy2 with PmlI, modification 006@ of the cohesive end with T4 DNA polymerase and dNTPs to create.

a blunt end, followed by digestion with EcoRI, and 3) a 1.9 kb fragment encoding the CD4 V1 V2 domains linked at their Cterminus to the .hinge and Fc regions of the human IgG2 heavy chain, obtained by digestion of pKCD4y2 with EcoRI, modification of the cohesive end with T4 polymerase and dNTPs S to create a blunt end, followed by digestion with SfiI.

pMSAb2int was identified by restriction analysis, and used to 0% prepare single-stranded DNA. template for oligonucleotide- 2O directed mutagenesis. In each example, the correct expression plasmid was identified by restriction mapping anrid its structure was confirmed by DNA sequencing.

Example 2A:_ SAb(agpl20)-CD4-Fc pIK-SAb(agpl20) -CD4-Fc directs. the expression of a hybrid protein consisting of the VX signal sequence and SCFv domain of 447D-LLH joined at its C-terminus (447D-VH residue -113)_ to the N-terminus of human CD4 (residues 1-180 of the 'mature polypeptide) followed by the.human IgG2 hinge -and -Fc domains (residues -226-478) This plasmid >.is constructed :by oligonucleotide-directed :mutagenesis busing -Single .stranded' pMSAb2int DNA as the .template with olioiiicleotide -11 (SEQ ID NO: 11) as the. primer._ The correct .expression -plasmid was identified by colony hybridization usnbligonucleotidde 12 (SEQ. ID NO:12)as-a 00~&p Example 21i: pIK-SAb(agp12O) -L1- 4'i ts 'the expression hybrid protein consisting oft~V inlsequence and SCFv domain of 447D-LLH joined at its td-td tM nus.(447D-VH residue 113) by a14 amino acid linker. (Gy4r-TrSrGySrGyL sr nf Ser-Glu-Gly-Lys-Gly) NO:23)- to the N-terminus of human CD4. (residues 1-8016 poletd) -f ollowed by the human 1gG2 -4106-; (esdus226-479). This plasmid is constu bmuagnei uigsingle SN As the template with oligonucleot~d )4 h iriir. The correct expression plasmid, %44' b colony hybridization using oligonucleotide, 4-N:as* a probe.

Example 2C: Shl (agpl2O1)-.L2-CDpIK-SAb(agpl2O) -L2-CD4-Fc d'iiects the expression- of a hybrid protein consisting of th -1.,-lsequence and SCFv domain of 447D-LLH joined;.q± ,47D-VH,'o$ redu 1 3 by a 25 amino acid lii&~ Ap Ala-Ala-Lys-Lys-Aspo4pL rA-Ps-Aa-Lys-Lys-Asp-* Gly) (SEQ ID ofhianCD4 *residue's 1-180 of -the -matg-Urett, -'..1owed by .the' humadi IgG2 hine ad .Pc4omi Thi plaismid is constructed by WON ~4iet mtgnss uig single. stranded the epae wi th oligonucleotid 14'S_ st epie.-The. correct expression plid P, )f colon'y -hybridization uising oligonucleotie6 Example'..-: SAI,(dgp2) -L3-CD4'-Fc-.

pIK-SAb(cagp120) -L3-CD4-Fc- directs' the' .expression of -a hybrid protein consisting :of -the VX _-signal -'sequence and -SCFv -domain of -447D-LLH -joined -at. its .C-terminus '-(447D VH residue :113)-by a 10 amino acid linker (Asp-Lys-Thr-His-Thr-Ser-Pro-Pro-Ser- Pro) (SEQ ID NO:25) to the N-terminus, of-human CD4 (residues 1-180 of the mature polypeptide), followed by the human IgG2 hinge and Fc domains (residues 226-478). This plasmid is constructed by oligonucleotide-directed mutagenesis using single stranded *pMSAb2int DNA as the template -with oligonucleotide 15 (SEQ ID NO:15) as the primer. The correct expression plasmid was identified by colony hybridization using oligonucleotide 8 (SEQ ID NO:8) as a probe.

Example 23: SAb(agp2)-L4-CD4-Fc pIK-SAb (cgpl2o) -L4-CD4-Fc directs the expression of -a hybrid protein consisting of the'VX signal sequence and SCFv domain of 447D-LLH joined at its C-terminus (447D-VH residue 113) by an 18 amino acid linker (Glu-Leu-Gln-Leu-Glu-Glu-Ser-Ser-Ala- Glu-Ala-Gln-Asp-Gly-Glu-Leu-Asp) (SEQ ID NO:26) to theNterminus of human- CD4 (residues .1-180 of the mature polypeptide), followed by the human IgG2 hinge and Fc domains (residues 226-478) This plasmid is constructed by oligonucleotide-directed .mutagenesis using single stranded pMSAb2int DNA- as:'the template _with oligonuicl'eotide 16 -(SEQ ID NO: 16) as the. primer. Thie lcor. ect expression -plasmid was identified -by-'-colony hybridization -using -oligonucleotide (SEQ ID NO:1O) as.-a probe.- Example'3:. Expression .basxicterization of nmltispecif ic antibodies To -determine .,whether -7 each Jultispecbific. ant'ibddy- can b efficiently expressed and secreted, -ard thus properly folded, each corresponding mammalian expression vector was transfected~ into a model mammalian cell, .using the :human 293 embryonic kidney cell line (ATCC _CRL 1573). -Following transfection, the expression and corresponding apparent molecular mass of each polypeptide was evaluated by radioimmunoprecipitation (RIP), and: the level of secretion .was quantitated using an enzymelinked immunosorbent assay (ELISA).

Example 3A: Transfection of human 293 cells with multispecific antibody expression vectors.

For..transfection, 293 cells were grown in DMEM:F12 media (JRH Biosciences) containing 10% fetal calf serum, and passaged at a 1:8 to 1:12 split ratio every 3 to 4 days. Forty-eight hours prior to transfection, cells were plated by passaging 1 the contents of one subconfluent 10 cm tissue culture dish onto twenty 6 cm tissue culture dishes. Five wg of each expression plasmid DNA was transfected onto a 6 cm dish by the •gee calcium phosphate coprecipitation method (Wigler et al. (1979) Cell 16:777). Sixteen hours after transfection, the 20 transfected cells were fed with fresh complete medium. The expression of multispecific antibody polypeptides was evaluated by RIP analysis and ELISA at 48 hours posttransfection.

Example 3B: Radioimmunoprecipitation analysis of multispecific antibodies expressed in transfected 293 cells Forty hours after transfection, -293 .cells were fed with 2 ml of methionine- and cysteine-deficient RPMI media containing dialysed fetal calf serum, supplemented with methionine and [35S]-cysteine (Tran35Slabel, 1160 Ci/mMol, ICN Biomedicals, -Inc., Irvine,.CA) .Cells -were. cultured for, an-.

additional 8 hours, the conditioned medium harvested, .and the labelled cells.-lysed in RIPA buffer .(50.mM Tris pH 7.5, 150 mM NaCI; 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate For '.radio-inmunoprecipitatio6n labelled conditioned medium was mixed .with:an :equal :volume- of 2X RIPA buffer. .35S-labelled multispecific .antibodies, -which bind to the Staphylococcus aureus Protein. A-via their -IgG2 Fc domain, .:were then precipitated -with:, 10 l Pansorbin (Calbiochem, -La LJolla,- CA) Immunoprecipitates 'were washed three times in RIPA buffer and ronce in .distilled .water, and boiled for several minutes in SDS sample buffer (50 mM Tris- HC1, pH.6.8, 150 mM 3-mercaptoethanol, 2% SDS, 10% glycerol).

Immunoprecipitates were also analysis :-following boiling in non-reducing SDS. sample buffer. (lacking 0-mercaptoethanol).

S Samples were analysed by 10% SDS polyacrylamide gel S electrophoresis (SDS-PAGE).. Gels were fixed in 20% methanol and 10% acetic acid, soaked in Enlightening solution (NEN Research Products, Boston, MA) for .15 minutes, dried and Ssubjected to autoradiography. This analysis revealed that multispecific antibodies of the predicted molecular mass are expressed and secreted into the culture medium (Figure 3).

Samples analysed by SDS-PAGE under non-denaturing conditions revealed multispecific antibodies with the molecular mass expected for the corresponding homodimeric proteins.

Example 3C: ELISA analysis of multispecific antibodies secreted by transfected 293 cells Human 293 cells were transfected as described above and were fed fresh medium 16 hours later. Forty-eight hours after transfection the conditioned medium from the transfected cells was harvested and analysed by ELISA. Nunc Maxisorp microtiter plates (Nunc Inc., Naperville, IL) were coated for 16 hours at o 4 C with an (Fab')2 fragment derived from a rabbit anti-human IgG polyclonal antisera (Accurate Chemical and Scientific Corporation Westbury, NY), that was diluted 1:1000 in 0.05 M sodium carbonate buffer pH 9.6. Plates were then washed three times in PBS containing 0.05% Tween-20 (PBS-Tween), followed by.- :blocking with .PBS containing 1% bovine serum alburmi (PBSA) ,at7 room -temperature -for .1 hour. Human IgG2 (Calbiochem, La Jolla, 1 CA) .was empldyed -as :a positivecontrol standard in the. assays. Samples and standards were diluted-in PBSA, -added to the antibody-coated -plates, and the plates-then 0 incubatedfor 16 hours at 4 C. :Plates were washed three times with PBS-Tween, and an enzyme-linked detection antibody was added to .the plates -and incubated for 1 hour at. room temperature. The detection antibody used was horseradish peroxidase-conjugated goat -anti-human IgG polyclonal antisera (Boehringer .Mannheim Biochemicals, -Indianapolis, IN) that-was diluted 1:50,000 in PBSA. Plates were:-then washed 3 times with PBS-Tween and an indicator solution consisting of 1.7.

mg/ml o-phenylenediamine hydrochloride in 16 mM citrate buffer S pH. 4.5 was added. Color-dvel6pment was stopped by the addition of 1 M sulfuric- acid, and the results read on a UVmax S microplate reader (Molecular Devices Corporation, Menlo Park, S CA). As shown.in Table 1, significant levels of each of the multispecific antibodies described in Examples 1A-E and 2A-E were detected in the culture supernatants of transfected 293 cells.

*i

S

rJ r i:

I

t

I:

-Table.: ~Dnat-1 -j CD4 -Fc SAb(cxgp4l).-Fc: SAb(agpl2O-Fc one t~rAtion Iucr ml)I -1.6 9 1.2 1.3 SAl, (cxgp4l1) -SAb (cxgp120-) SAb (agpI) SAb(axgpl)-C;.-A)( SAb (cgpl2O)-C4c SAb (Cgpl2') -LI I 4 -L3-CD 4" Fc SAb (agpl20) -L4-CD4.-F~c- 0.-23 0.65 0.75 0.40 0.48 0.05 .0.17 0.19 0.10 o0.12 Em~~le 4: i pr ris of Muti~pecif ic antibodies Multispecif ic antib. To4*o the HIV-1 binding domains of CD4. and. the 447Dv~ 00id atJ~ ne f urther 'characterized f or their ability to -bJA~vt0o ,;n1 Cg120 and gp4l proteins which are -the proteolYdtC.>',. 4 4 t s of"-the protein,: using-ceall *4 exress .:wild-type: mutant forms of--gp16O v1isel f oki or' another. ,binding -epitope 1is impaired. 7 7t*- 2 4 ~-lL 0-- To facilitate this analysis, cell lines are generated whith efficiently express surface gpl20 and gp41 polypeptides. Mutants of gpl60env are generated which have significantly diminished binding to at -least one of the ECDs present 'in a given multispecific antibody to permit detection of binding mediated by the other ECD(s) present. Levels of binding are determined by FACS analysis using the multispecific antibodies to stain the cells.

Example 4A: Vectors for efficient -expression of HNV-1 pIKenv+/rev+/tat- is a vector designed to allow the efficient expression of. gpl60env in mammalian cells, based on the observation that while expression of the rev gene product is S. essential for efficient expression of gpl60env, expression of the 15 tat gene product may be inhibitory (Bird et al. J. Biol. Chem.

(1990) 265: 19151-19157). This plasmid was made in two steps.

The first step was the construction of pIKenv, which directs the expression of three HIV gene products: env, rev and tat. This plasmid was constructed from two DNA fragments: 1) a 4.3 kb vector fragment obtained by digestion of pIK1.1 with BglII, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by digestion with EcoRI, and 2) a 3.1 kb fragment encoding the env, rev and tat gene products of the HXB2 isolate of HIV-1 obtained by digestion of pCMVenv (U.S.

Patent #5,359,046) with Xhol, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by:. digestion with EcoRI.. ,pIKen.was identified by restriction analysis, and: use'd in the :-second step .to6prepare a single- -stranded DNA template for oligornucleotide directed mutagenesis using oligonucleotide 17 (SEQ ID NO:17)as a primer to remove the initiator methionine codon (ATG) and the adjacent arginine codon (GAG) of the tat gene, and to replace these sequences with a novel PstI site (CTGCAG). The correct plasmid, pIKenv+/rev+/tatwas identified by colony hybridization using oligonucleotide 18_ (SEQ _ID NO: 18) as a.-probe and its.structure-was-confirmed by DNA: sequencing.*< Example 4B: Expression -of an Ely- I gpi6Oenv gene with a mutation in the CD4 binding-site pIKenvG37OR/rev+ /tat- directs- the -expression of a mutant- polypeptide in which Glu-370 is -replaced with Arg, :-which results in a loss of -CD4 binding (Olshevsky et al., J. viral. 64: 5701- 5707). This plasmid is constructed by oligonucleotide-directed mutagenesis using single-stranded plKenv+/rev+/tat- -DNA'with oligonucleptide 19 '(SEQ ID NO:19) as the primer. -The" correct expression plasmid is identified by colony -hybridi zat ion using oligonucleotide 20 (SEQ ID NO:20)as a probe, and its structure is confirmed by DNA sequencing.

Example 4C: Expression of an NWV-i gpi6Oenv gene with a mutation in the 447D Mhb binding site pIKenvR3l5Q/rev+/tat- directs the expression of a mutant polypeptide in which Arg-315 is replaced with Gln, which results 020 in a loss of 447D MAb binding (Gamny et al, J. Viral. 66: -7538- 7542). This plasmid is constructed by oligonucleotide-directed mutagenesis using single stranded pIKenv+/rev+/tat- DNA with *oligonucleotide 21 (SEQ ID NO:21)*-as the primer. The correct expression plasmid is -identified by colony hybridization using oligonucleotide 22 (SEQ.ID NO:22) as a probe, and its* structure is confirmed by DNA sequencing.: Example 4D: Binding of 'multispecif ic antibodies -to muitant and wild-type Ely-i gpi6oen* polypeaptides Human 293 cells '.are- tranfsf ected -with. -ve'tors d ir ecdtin-g the expression of the -multispecific antibodies describ6d. in. ExAmples 1 a-s- well: as.--vetr- directing the" expressionj of the' parent monospeif ic antibodies -(pI1KCD4y2' for- CD4-F pIK447D-LLHy2 for 447D-LLH-Fc, and pIK98.6KLHy2 for 98.6 LLH-iFc)>Supe~iataht~s 'from transfected z293 cells are analysed by ELISA to determine -tne concentration of the recombinant proteins secreted into the culture medium. Known amounts of each recombinant protein are -then used to determine their ability .to -bind to cell surfaceexpressed wild-type and mutant gpl60env polypeptides, by FACS analysis of human 293 cells which have been transfected with ,pIKenv+/rev+/tat-, pIKenvG370R/rev+/tat- and pIKenvR315Q/rev+/tat-. Following the binding of each multispecific or monospecific antibody, the cells are stained with FITC-conjugated mouse anti-human IgG MAb (Becton Dickinson Immunocytometry Systems, San Jose, CA) The difference in the S degree of fluorescence intensity obtained. with multispecific C. S antibodies versus each of the parent monospecific antibddies on transfected cells expressing either mutant or wild-type 15 polypeptides reveals the ability of each ECD contained within a given multispecific antibody to bind to 9 Example 5: MSCRs comprising multiple antibody extracellular clustering domains and a r family signaling domain.

Multispecific chimeric receptors (MSCRs) were created from multispecific antibodies which contain two or. more extracellular clustering domains (ECDs) derived from antibodies and/or single-chain antibodies, or modifications thereof. The constructs described in this example contain two

S.

LLH .SCFv domains which were derived from the 98.6 and. 447D human MAbs, respectively, as described. in Example 1.

Alternatively, the 98.6 and..447D SCFv's are created as heavylinker-light (HLL) -constructs r:in which ,the heavy chain variable domain precedes the -light: chain -yariable domain, connected-by a suitable.oligo- or -polypeptide -linker. Both LLH .or HLL ;SCFv derivatives of the 98.6 :and :447D MAbs may be constructed- using a variety of oligp- and polypeptide linkers.

.In this .example, the 98 .6 LLH SCFv was. ;joined at its: Cterminus- (residue 113 of the VH variable domain) to the Nterminus of the 447D LLH SCFv -(residue 1 of the V -variable domain). Alternatively, the 447D LLH SCFv is joined at its Cterminus (residue 113 :of the VH variable domain) to the Nterminus of the 98.6 LLH SCFv (residue 1 :of the VK-variable domain). Either LLH SCFv may be substituted for by the corresponding HLL SCFv, or modifications thereof. The two SCFv's. were joined either directly, or via an. oligo- or polypeptide linker. The C-terminus of the second SCFv was fused in turn to the hinge and Fc region (residues 226-477) of the human IgG2 heavy chain, followed by (from N- to Cterminus) the 18 residue human IgG3 M1 membrane hinge, the CD4 TM domain (residues 372-395 of the mature polypeptide), and the CYT domain (residues 31-142 of the mature polypeptide). Mammalian transduction-expression vectors for the 98.6 SCFv/447D SCFv MCRsi described in this example were constructed using plasmid pRT43.2F16, a retroviral vector S which directs the expression of the SAb-( monospecific chimeric receptor F16 Patent #5,359,046) which contains the 98.6 LLH SCFv domain fused to (from N- to C-terminus): the Fc domain of human IgG1 heavy chain, the IgG3 M1 membrane.

hinge domain, the CD4 TM domain, and the CYT domain.

Plasmid pRT43.2F16 was constructed from three fragments: 1) a 6.7 kb vector fragment obtained by digestion of pRT43.2F3 (as .i described in U.S. Patent Application 08/258,152) with EcoRI and Apa I, 2) a 1.4 kb fragment encoding the 98.6 LLH SCFv domain and the N-terminal portion of the IgGl Fc domain, obtained by digestion of.pIKF16 Patent #5.359,046) with EcoRI and NsiI, and 3) a 0.7 kb fragment encoding -the remainder of the IgG1 Fc domain, the -IgG3 -inembrane hinge domain, the CD4 TM. domain and the CYT domain, obtained by digestion of pIKF16 with NsiI and Apal;- The construction of MSCRs with an IgG2 -Fc domain using a monosspecific chimeric receptor with .an IgG1 Fc domain was facilitated by- the fact that. the CH3 regions of the human IgG1 and IgG2 Fc domains share a- uni.que-,NSiI site,-,and the amino :acid-seuerices of ea~r which follow this:. restrict ion ,site are' -identical. In :each example, -:the. _correct-. .:epression:plasmid was :idehtif ied -by restriction mappijig. Example 5A:- T7 toW kkt 2)F_ pRT-SAb.(agp4l) -SAb(. directs the expression of -a hybrid protein consisting of the -Vrsignal sequence and SCFv domain of 98.6 -LLH.jointed. at Its C-terminus (98.6-VH residue '0 113) to the N-ter~i1o th SF dain of47DLH(7-v residue ter nus) thehuman IgG2 -hknge. ahd" d* 226-48.):, the gG3 1 membran inge tedxa.n(~~de 372-395 of the mature pol C CYT, dqmai (residues 31-142 of the mi tuae- i 148s is: corlsk!tr*ed from two fragments: k ~trfamn encoigteC terminal portion'. of;- gl c domain (identical to the corresponding IgG2-r~o) the,:IgG3 'MI domain, the :CD4 TM domain and the t CT' domvain,: -obtained by-.digestion* of* pIK43.2F16 with EcoRI anid Ms iI, and a 2. 0 kb f ragmept encoding the 98.6-L14:Wx S~i oi ,te '447D-LLH SC~v,. domain IA A and the N-termi.hin*.p. Ec don,,ain, by digestion. of Kk 42~-cwtand NsiI. Exmle 5B: a "At pRT- SAb (gp41) -Ll-A V 4Aecsthe epVre ssion of a hybrid pr otein 6r4i~i i~l sequenceand SC~v domain of 986LHjj~ rinuis (98. 6--VH -re sidue 113) b y.:a..14 am~ino, it,4411 ].rkr. ySeir-Thr-'Ser-Gly-; S -GyLseSerer y t (EQID 23) to--nthe terminus :of-~e followed by* kd h hmn~g2 ighn Fc domains: (resldie 2S. 48 the- IgG3 -membrane hinge domain, the CD4 TM domain (residues 372-395 of the maturi polypeptide) and the .domain-_ (residues- 31-142 of:-.themature. polypept'ide),.. -This -plasmid 'Was constructed. if rom -two f ragments: a .7.5 kcb vector.fragmelt encoding-the C-teriminal pqrtion--of the IgGl:Fc- domain .(identical.-tb the corresponding IgG2 -region) -the IgG3 -Ml domain,. the CD4 TM-.domain and -the l CYT domain,-obtained by digestion of pIK43.2F16.with EcoRI and Ns iI, and 2) -a 2.-0 -kb f ragment -encoding the -9 8. 6 -LLH -SCFv domain, the. 14 amino acid peptide linker,, the 447D-LLH SCFv it domain and.-the N- terminal -portion- of the IgG2 Fc domain, obtained by_ digestion of pIK-.SAb(agp4l)-L1-SAb(cvgpl2O)-Fc with EcoRI and NsiI.

Exaz~l 5C: BAb(cxgp41)-L2-SAb(agP12O)-Fc-_ .*015 pRT-SAb(agp41)-L2-SAb(cxgp12O)-Fc- directs the expression of a hybrid protein consisting' of 'the Vv. signal -sequence- and SCFv domain of 98.6-LLH joined at its C-terminus (98.6-VH residue 0 113) by a 25 amino acid linker -(Ser-Ser-Ala-Asp'-Asp-Ala-Lys- Lys-Asp- Gly) (SEQ ID NQ:24) to the N-terminus of the SCFv domain of 447D-LLH (447D-VX residue 1),-followed by (from Nto C-terminus): the human IgG2 hinge and Fc domains (residues 22 6-478), the IgG3 Ml membrane hinge domain, the CD4 TM domain 0 6 0 (residues 372-395 of the.-mature,, polypeptide). and the CYT domain (--residues 31-l42--of the-mature. polypeptide) This plasmid was- constructed. from two-_fragmentsi: -a .7.5 -kb vector f ragment- encoding the C-teprminal, portion of '-the IgGi Fc domain (identical -to -the corresponding IgG2 -region) ;the -IgG3 Ml' domain,- the. CD4 JMdomain,- and.-the CYIT domain, "obtained by digestion.' of pIK43.2F16 .withEcR an NsI% an )a2k fragment encoding ;the:_.98. 6-LLH! .SCFv domain, *h-2 amioad peptide linker, -the* 447D7LLH:SCFv domaiin:-and' the;N-terminal portion.::of the. IgG2 FcC. domain;-:= obtained -by -digestion o0f p'IK-- SAb(cgp4l) L2-SAb (cgpl20) -Fc -with EcoRI'I and NsiI.- Example SD:' SAb (xgV4 1)-L3 'S~b (agp12)-F- 2: pRT-Sb(agp4) -L3 -SAb(gp12 0) -Fc- :directs -the expression of a -hybrid protein .consisting ,of -the Wk signal -sequence: and SCFVdomain- of. 98.6-LLH joi ned- at-i1ts C-terminius -_(98.6-VH-'residue 113) 'by-a- 10 amino acid linker (Asp -Lys--Thr-His-Thr-Ser-Pr o- Pro-Ser-Pro) -(SEQ 'ID -NO: 25) to- -the 'N-terminus of the:-SCFv domain,-of 447D-LLH (447D-VX residue followed by *-(from' Nto-C-terminus): the human IgG2 hinge and Fc domains. (residues 226-478), the IgG3 M1 membrane hinge-domain," the .CD4 TM domain (residues 372-395' of the mature polypeptide) and the-- CYT domain (residues 31-142 of the mature polypeptide). This plasmid is constructed from two fragments: 1) a 7.5 kb vector fragment encoding the C-.!terminal. portion of the IgGi Fc domain 3:5 (identical to the corresponding IgG2 region), the IgG3 -M1 domain, the CD4 TM domain sand .the. -CYT domain,-.obtained by *digestion of pIK43.2F16 with EcoRI and NsiI, and 2) a 2.0 kb ~*fragment encoding -the- 98.6-LLH SCFV domain, the 10 amino 'acid peptide linker, the 447D-LLH SCFv domain and the N-terminal portion of the IgG2 Fc domain, obtained bydgsinof pK SAb(agp41)-L3-SAb(agpl2O)-Fc with EcoRI and NsiI.

Examle SE: S~b(cxgp4l)-.L4;-BAI(agpl2O)-Fc-( .7 pRT-SAb(ap4l)-L4-SAb(gpl2O)-Fc-a 'directs the expression- of: a hybrid protein consisting of'.the'Vx~signal seguence-and -SCFv domain :of-98.6-LLH joine51 at its4 terminus' (98 .6-VH residue-.

113) by an 18 'amino acid linkerAGlu-Leu-Gln-Leu-Glu -Glu- Ser- Ser-Ala-Glu-Ala-Gln-Asp-Gly Glu- Leu- Asp) (SEQ *ID NO: 26) -to- the N-terminus of the SCFv'ddmain f -447D LLH :(447D-VX- r7esidue'-71), followed :by (frfom N- to' C-terminius) Cthe' human -i~g~2 :hin§6 Fcddmains .,(residues.-':226-"478)-., the- -IgG3 MI membrane hin'ge domain,':the.CD4: TM -domain I-residiesj'372-395--o f the- -maturepolypeptide) and 'the ~~Y.domain sidues 31-142 .of 'the mature polypeptide) This. plasmid is constructed-: from two fragments: 1) a 7.5 kb vector fragment encoding the C- terminal portion of the IgG1 .Fc domain (identical to the corresponding IgG2. region), the IgG3- M1 domain,- the CD4 -TM domain and -the CYT domain, obtained by digestiori of .pIK43.2F16 with EcoRI and NsiI, and 2) a 2.0 kb fragment encoding the 98.6-LLH SCFv domain, .the 18 amino acid peptide linker, the 447D-LLH SCFv domain and the N-terminal -portion of the '.IgG2 Fc domain, obtained by digestion of pIK-SAb(agp41)-L4-SAb(agpl20)-Fc with EcoRI and NsiI.

Example 6: MSCRs comprising an antibody extracellular clustering domain, a ligand-receptor (CD4) extracellular.

clustering domain, and a C family signaling domain.

.5 Multispecific chimeric receptors (MSCRs) were created from multispecific antibodies which contain two or more extracellular clustering domains (ECDs), at least one of which is derived from an antibody and/or single-chain antibody, or modifications thereof, and at least one of which is derived from a ligand-receptor binding domain, or modifications thereof. The constructs described in this example contain the 447D LLH SCFv domain, and the human CD4 V1-V4 domains, which i bind with high affinity to the HIV-1 gpl20 envelope glycoprotein. Alternatively, the 447D SCFv's are created as HLL constructs in which the heavy :chain variable domain.

precedes the light chain variable domain, connected by a suitable.oligo- or polypeptide linker. Both LLH or HLL SCFv derivatives of the 447D MAb may be constructed using a variety of -oligo- and polypeptide linkers. .Portions of .CD4 other- than the, entire -CD4 EXT "domain (residries 1-371 of the inmature polypeptide) :made be employed, including various -truncations and/or -modifications thereof. this -exanple; the: 447D-~LLH SCFv:, was -joined :at its :C-terminus (residue: 113 -'of the VH variable domain) to the N-terminus of the CD4 protein residue 1 of -the mature polypeptide).. Alternatively, the -CD4 .protein is joined at the C-terminus of "its'.entire 'EXT domain (residue 371: of the: mature polypeptide;, or -truncations thereof, e.g.

residue.180 which resides at the C-terminus.of the CD4 V1 V2 domains) ,to- the N-terminus of::the 447 LLH SCFv (residue 1' of the' VX variable domain). -The 447D LLH SCFv may be substituted for by the corresponding' HLL SCFv, or-modifications thereof.

The 447D. 'LLH SCFv was joined 'to the CD4 protein either directly,.or. viaan oligo- or polypeptide linker. The -Cterminus of the CD4 EXT domain, was fused in turn to the CD4 TM*'domain, (residues 372-395 of the mature polypeptide), and S the. CYT domain (residues 31-142 of the mature polypeptide).

S Mammalian transduction-expression vectors for the 447D.

SCFv/CD4 MSCRs described in this example were constructed using -pRT43.2F3, a retroviral vector which directs 'the expression of the CD4-( monospecific chimeric receptor F3 Patent #5,359,046) comprised of the CD4 EXT and TM domains fused to the. CYT domain. The construction of MSCRs with a C-terminal CD4 EXT domain was facilitated by using a unique NheI restriction site in CD4 which is present in both the .447D SCFv/CD4 multispecific antibody and the CD4monospecific 'chimeric receptor. In each example, the correct expression plasmid was identified by restriction mapping.

Example 6A: SAb(agp120)-CD4pRT-SAb(agp120)-CD4-( directs .the expression' of a hybrid protein consisting .of the..VA.secretion leader and SCFv :domain of_447D-LLH:joined at its C-terminus (447D-VH residue .113) 'to the. CD4 .EXT.vand TM domains (residues .L1-395 of i the mature polypeptide) and the (:CYT:domain.(residues 31-142 of the mature,-polypeptide). This plasmid is constructed from three fragments:.:1) a..6,.7kb vector fragment :obtained by digesting pRT43.:-2F3 -with .EcoRI and Apal, kb.fragment encoding the -447D-LLH SCFv domain and the.N-terminal portion of the' CD4 EXT -domain,-- obtained-,b:y rdigesition :of -pIKSAb (cgp120) -CD4 FC with EcoRI and-NheI,', and 3) -a -1'.2kcb -fragment :ecdn the Cterminal portion- of .thei CD4. mo domain,'- -the -CD4 TM 'domain -and the CYT domain; bbtii 4 dgetion -of _pI43. 2F3 -with 1NheI and AkpaI.pRT-SAb(axgpl2O)-Ll-C4. .ct the expression of a hybrid protein consisting of- the. VX s~ tionlae n Cvdmi of 447D-LLH joineda t it~ ~~(4DV eiu 113) by a 1.4 aminot acid.- 1:Vr' y-SrGyLsSr Se*GuGlLy to theCD4 EXT and TM domains (rsiduesk.KA,"3 ~4 y tide) And the doai oy tide). This .1i plasmid was. a ~g ns.. 1) a. 6.7kb vector fragment oti' j '3.23 wihEcoRI and ApaI, 2) a 1.5 kb frai e ng he4D-LLH SCFv domain, the 14 amino acid p"p'eli r: and the N-terkminal portion of the CD4 EXT domain, obtainfed -by- digiestion, of pIKSAb(cxgpl2O)- L1-CD4-Fc with EcoRTL A* NiI, aid a 1..2kb f ragment encoding the C-terd n CfbETiman h D TM domain and. th bW by13i no pIK43.2F3 wt h B"amle GC: .11 37 axPres of a 'hybrid protein consistii n l~4' qn CFv domain of 447D-LLH joi1nd (47-vf si 113!)* bk a 25 amino acid 1inAkt:"-J *9 13 PAu -Aa- ys-Ly~p Ala-Ala-Lys-Lys A~~p~LaLsLsAp Gly) (SEQ:ID NOi*-.2 41,' 1- MLIs qsde -395-of .'the mAtuk". Y diin-(sidiieLs 1-14 2 o f h~.11A'1- Lasinid was constructed .from- -tte )a :6 .7kb vector -fragment obtained by.digesting pRT43.2F3 with EcoRI-and Apal, 2) a kb fragment encoding the 447D-LLH SCFv domain the 25 amino acid peptide linker and the N-terminal portion of the CD4 EXT domain, obtained by digestion of pIKSAb(agpl2O)-L2-CD4-Fc with EcoRI and NheI, and 3) a 1.2kb fragment encoding the C-terminal portion of the CD4 EXT domain, the CD4 TM domain and the CYT domain, obtained by digestion of pIK43.2F3 with NheI and Apal.

Example 6D: SAb(cagpl2O)-L3-CD4-l pRT-SAb(agpl20)-L3-CD4- directs the expression of a hybrid protein consisting of the VX secretion leader and SCFv domain of 447D-LLH joined at its C-terminus (447D-VH residue 113) by a 10 amino acid linker (Asp-Lys-Thr-His-Thr-Ser-Pro-Pro-Ser- Pro) (SEQ ID NO:25) to the CD4 EXT and TM domains (residues I- 395 of the mature polypeptide) and the CYT domain (residues 31-142 of the mature polypeptide). This plasmid is constructed from three fragments: 1) a 6.7kb vector fragment obtained by digesting pRT43.2F3 with EcoRI and Apal, 2) a 1.4 kb fragment encoding the 447D-LLH SCFv domain, the 10 amino S acid peptide linker and the N-terminal portion of the CD4. EXT domain, obtained by digestion of pIKSAb(agpl20)-L3-CD4-Fc with EcoRI and NheI, and 3) a 1.2kb fragment encoding the Cterminal portion of the CD4 EXT domain, the CD4 TM domain and the CYT domain, obtained by digestion of pIK43.2F3 with NheI and Apal.

Example 6E: SAb(agpl20)-L4-CD4- pRT-SAb(agpl20)-L4-CD4-. -directs the expression of a hybrid protein consisting ,of the VX secretion leader and SCFv. domain of 447D-LLH joined at..its-C-terminus (447D-VH residue 113) by an 18 amino acid linker (Glu-Leu-Gln-Leu-Glu-Glu-Ser-Ser-Ala-Glu-Ala- Gln-Asp-Gly-Glu-Leu-Asp) -(SEQ IDNO:26) -to the CD4 EXT and TM domains (residues.1-395 of the mature polypeptide) and the CYT domain .(residues 31-142 of the mature polypeptide) This plasmid is constructed: from,;three. fragments-:.1). -a 6.7kb .vector fragment obtained by digesting pRT43 .2F3 with EcoRI .andL Apal, a;l 5 kb fragment -encoding:the :447D-LLH :SCFv.idomain, -the 18: amino acid peptide-linker and -the N-terminal portion :of=the CD4 EXT domain, obtained by digestion of pIKSAb(agpl20)-L4-CD4-Fc. with EcoRI and NheI, and 3) a 1.2kb fragment encoding the-C-terminal portion of the CD4 EXT domain, the CD4 TM domain and the CYT domain, obtained by digestion of pIK43.2F3 with NheI and Apal.

Example 7: Multispecific antibodies 1KSCRs comprising- two antibody extracellular clustering domains, a ligand-receptor (CD4) extracellular clustering domain, and a family signaling domain.

Multispecific antibodies and MSCRs are created which contain three or more extracellular clustering domains (ECDs), at least one of which is derived from an "antibody and/or singlechain antibody, or modifications thereof, and at least one of which is derived from a ligand-receptor binding domain, or i modifications thereof. -The constructs described in this example contain the 98.6 LLH SCFv domain, the 447D LLH SCFv domain, and the human CD4 V1-V2 domains, in the case of the multispecific antibodies, and the entire CD4 EXT domains',-' in the case of the MSCRs. Alternatively,. the 98.6 and 447D SCFv's are created:as.HLL constructs- in which the heavy chain variable domain .precedes the light chain variable domain, connected by.a suitable..oligo- -oripolypeptide linker. :-Both LLH or HLL SCFv derivatives of the 98.6 and.447D MAb may- be constructed using a variety..of oligo- and polypeptide linkers and substituted accordingly. in- the multispecific -antibodies and MSCRs herein described.. Portions -of. CD4 other. tihan the entire CD4 :.EXT .or .V1-V2- domains :.may.:similarly be biplyed, including various truncations .and/or :modificationsi thereof.

In this.example, the order of ECDs is' (from N-to C-terminal): the..98.'6 ILLH SCFv..domain, the -447D LLH.-SCFv domain n-h:.

CD4- -EXT domain.- :Alternatively,~: i1t- i.s' Possible to create variants of -all of :the possible 7permnutations -of teorder.: of _these tbree.as well Asother-d aiisj'-he9 LHSCv 14D LLH -SCFv,. CD4 and other :-ECDs many- be- -liziked either -directly, :orT via yvarious. oligo- -or pjolypeptide linkers.. In ithe event that a.SCFv is the most C-terminal -ECD in an, MSCR,' -it -may bef-. fused in-turn -to the-.hinge and Fc region (residu es-226-477) of the human IgG2 heavy chain, followed--by- (from-N- to C-terminus):' the 18 residue human IgG3 Ml membrane hinge, the CD4 TM domain (residuesy'372-395--of the mature polypep:tide), and the CYT domain (regidues 31-142 of the mature polypeptide). In theevent that CD4 is the most C-terminus ECD of the MSCR, *it may be fused in turn to the CD4 TM domain (residues -372-395 of the mature polypeptide), and the CYT domain'-(residues 31-142 of the mature. polypeptide)*. In each example, -the correct expression -plasmid was identif ied by restriction mapping.

Example 7A: Shb (agp4l) -SAb(a~gP120) -CD4 -Fc ro~o A series of plasmids of the general structure pIKSAb(cxgp4l)- (Lx)-SAb(agp12O)-(Ly)-CD4-Fc, where Lx and Ly are any one of a number of various oligo- and,-polype4ptide linkers including Li,, L2, L3, L4 or no -linker, direct -the expression of a series hybrid. proteins' consisting -(from to terminus)' of 1)theVK signal sequence.and 98 -6-LLH-tSCFvdomnain,,- 2) linker. Lx, 3) the. -447D-LLH -SCFv;'. 4) linker Ly, 5) -at--portion:-of the' CD4 EXT. domain.,)(residues: 1-180 -of -the -mature pblypeptide)7 and: the-human -IgG2 -hinge- -and. Fa -domains l(residues 226Z:478) -The-se. plasmids: are.- constructed from'-:three::fragments: :1)z a' -4*.3kbvector. fragment obtainediby diqestion-_ofi'pIK1:1l with EcoRI- and BglII, a-1.7 kb--ftagment.:encodirig-the CterminusbI the 447D-LLH- scFv .domain,- linker.s-Ly, the..CD4.71.-& M2diafs andthe, :IgG2 Fc domain,.: obtaind- by.-fdigestion of one -of the pIKSAb (agpl2 0)-Ly-CD4 -Fc -series o:plasmids _-with- SpeI 'and- BglII, and 3) a 1.2 kb fragment encoding the entire 98.6-LL..

SCFv domain,. linker LX, and, -the N-termtinus of -the 447D-LLH SCFv domain, obtained by digestion of :one-_-ofth pIKSAb(cgp4l) -Lx-SAb(ctgpl2O) -Fc -series -of ,plasmids -'with-EcoRI and.SpeI.

A -series of plasmids of the general structure pIKSAb (cgp4l) (Lx) SAb(agp120)-(Ly)-CD4- where Lx and Ly are any-one of a number of various oligo- and polypeptide linkers including L1, L2., L3, L4 or no linker, direct the expression of a series of hybrid proteins consisting (from N- to C-terminus) of the VK-signal sequence anid 98.6-LLH SCFv domain, 2) linker -Lx, thi -447D-LLH SCFv, 4) linker Ly, 5) the CD4 EXT and TM domains -(residues 1-395 of the mature polypeptide), and 5) the t CYT domain (residues 31- 015 142 of the mature polypeptide). These plasmids are constructed from three fragments: 1) a 6.7kb vector fragment obtained by digesting pRT43.2F3 with EcoRI and ApaI, a 2.2..kb fragment encoding the* the entire 98.6-LLH SCFv domain, linker Lx, -the entire 447D-LLH SCFv -domain, linker Ly, and the N-terminal portion of the CD4 EXT domain, obtained by digestion of pIK- SAb(cgp4l)-Lx-SAb(cxgpl2O)-Ly-CD4-Fc with EcoRI and -NheI, and 3) .a 1.2kb fragment encoding the C-terminal portion of the CD4 EXT domain, the CD4 TN 'domain and -the CYT domain, obtained by a :digestion of pIK43.-2F3 with NheI and ApaI.* Example 'Expression &-charatritation of 118CR6 To determine whether each _MSCR polypeptide can-be efficiently expressed and transported to, the cell surface, and:, ths prcperly folded, a corresponding mammalian transduction-expression vector Is used. to -transf ect human 293 ;embryonic.kidneyei:.ie Following transf ection, the expression of each constrfit -is evaluated .by -rddidimminoprecipitation#- and :Its -tianispIcrt td -the cell surface is evaluated -by f luoresc6'nt-activatea cell soiting IFACS) analysis.-..- *4

C

C

C.

a C.

C

C.

C

a Examiple. SA: TrazmfeatLon -of: human 293:- cells with MSCR Fo ransfcin 293 L1 er grown -in: DMMI: F12* media (JRH Biosciences) conta, O~~al calf serum, ndasgeda a110 split ratio 8. .Twenty-four. Ihurs-*prior to transf ectiont- -293 it plated at:5x1 0. cells per 10 cm culture dish. Ter M~o ach: expression plasmid DNA is transfected onto-+4 9 1O x ih by the calcium phosphate coprecipitation metbo4 (O0 t al. (1979) Cell' 16:717).

Tet-or pm~ q~ p the -transfeacted cells were fed g bp, APiediLa.- The'expression of MSCR polyet IyFACS analysis and adomuprect at 48' hours pot Exapl 8: ~c M8~Rexpreson'on 293 cells- Transfected 293 ce1- v Zinsed once with PBS and indiubated in PBS containing 10 mMt -VT- or 5..minutes at room" 'temperature.

20 Cells are collected frft',tes. centr ifuged 'and rspeded in 6 PBS containing f serum. Aroximatoly ixib cells/sample tuat~n g~i tratlon, of -FITC-cOnj3~ ellf'CD M (Becton Dikis W a Jopse CV). -Mouse FIC-gA ndP.II'control l(Abs. All FACS. anlssB~tt~n(ecton Dickinson)

P~

-as~ prevoul tb; Exp Wed., *-160:1284-1299. 4 P"ml.C Xq Ltow. of IKCRs-expressedi 9 cells:,.z> Trans M vt PI .medii=m lacking methionin. nal -8 hours in;2 ml- of methionin-efc tf peete ih200.ui 62 methionine -,(1160 C/mmol, ICN Biomedicals, Inc., Irvine, CA). -_Tf-fe labelled cells are lysed in RIPA buffer, and-the cell lysates -are incubated at 4 0 C f or I..hour .with either'-no antibody :(Class 1 MSCRs -contain the IgG2 Fc domain and bind Protein A -directly) or mouse-anti-CD4 OKT4A.MAb (Ortho Diagnostic Systems.; -Raritan, NJ).

Ten microliters of Pansorbin is added to .the .lysates to precipitate the -MSCR. lImmunoprecipitates -are awashed three 'times in RIPA buffer, boiled in SDS sample buffer and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Gels are fixed -in 20% methanol/ 10% acetic acid, then soaked in Enlightning solution for 15 min, dried and subjected to autoradio~raphy. SDS-PAGE-analysis reveals the molecular mass of MSCRs expressed in 293 cells.

Example 9: Biochemical and biological properties of human CD8+ T cells expressing MSCRB Example 9A: Infection of 1humnan CD8 +T cells with MSCRexpressing retroviral vectors :9 0 Human CD8 'T lymphocytes are isolated from peripheral. blood lymphocytes (PBL) obtained from healthy donors by purificatio~n with the CEPRATE LC system (CellPro, Inc., Bothell, WA), followed negative selection against CD4 cells using a MicroCELLector (AIS, Inc., Santa.Clara, CA). Immediately after purification, cells are stimulated for 24 hours with an- equal number of y-irradiated autologous PBMCs in AIM-V media (GibcoBRL, _iGrand Island, NY) containing 10 ng/ml of OKT3.MAb and 100 units of hmanIL- Chirn Crp. Emryville, CA). Cells' are* then washed f ree :o f .QOKT3 and cul tured in AR media 0 %.AIM-V I 50 RPMI,-.--4 4-mM> Glutamine, 20 mM -Hepes, -1 3mM Na-Pyruvate, :nonessential :-amino .acids, and -1.00. units human IL-2)..-supplemented -with 45%. ,heat inactivated human -AB plasma (Sigma,~ St.: Louis, MO) A.Retrovirus--.is prepared. in _the -TIN-4. cell line-'derived '.from thymidinew- kinase-expressing- -human _293 -cells.- Fdr -the transduction- of human CD8..cells,-'TIN-4 cells are seeded at -5X10 celplate7_=in:_6-well plates Corning- Glass,- Corning, NY) in complete -DMEM- medium 48 hours- prior to transfection. Ten micrograms of' MSCEFR'_construct-.in :the 'retroviral vector pRT43.2 (as:-_.:described -in Patent- Application. 08/258,152) -are transf ected per plate -in the absence or presence of -packaging plasmids by the calcium phosphate coprecipitation method.

Following transfection, 1.5 ml-of fresh AR medium containing 100 units/ml ofhuman IL-2 is added to each well of the plate. Three hours later, 5X10 5 of CDE8-- T cells in AR media containing 100 units/ml of human IL-2 and 2 ug/ml of polybrene are added to each well of the plate. CD8 T cells are removed from the 6-well plates 24 hours later and then-transduced a second time by the same procedure. Newly transduced CD8 T cells are maintained in AR media.

Example 9B: FACS analysis of 118CR expression in human CD8 T 'Cells- At various times f ollowing. transduction,.+ CD8+ T cells are harvested and washed with PBS containing 1% FCS. Approximately *6 1x10 CD8 T cells are stained with specific antibodies for FACS analysis as described.intExample 8B.

_*ENAmple.9C: 2 -Immunprecipitation-asalysis of 118CR expression 'in huian.CD8. T Vells- various. [times f ol lowing* transduction, human CD8 T cells -are -harvested-and plated in methionine-depleted AR media -supplemented with -200_4iCi [35S]-methionine -(11610 .Ci/immol, 'ICN Biomedicals, Inc. 7.'-Cells are lysed in RIPA buff er'--and:.te~nuae t 0 -for 1'.hour.Ewith either. with.-nol antibody jFc-containing MSCRs) -or mouse Lanti-CD4 OKT4A "-MAb :-CD4 -containing MSCRs). Ten microliters of Pansorbin. are then wadded. to the;:ysates.-to -precipitate the MSCR polypeptide. -The ixmunoprecipitates .,are washed thkee-times -in RIPA buffer, boiled in.SDS sample buff er and' Analyzed by 8% -SDS-polyacrylamide gel electrophoresis. Gels'.are f ixed ::methanol/. _10% acetic- acid,- -then soaked -in. LEnlightning solution minutes,-- dried and'subjected to autoradiography.. SDS-PAGE analysis reveals- the molecular mass of MSCRs -expressed -in human CD8 T cells.

Example 9D: Cytolytic activity of 318CR-expressing .human CD8+ T. cells.- To determine-the cytolytic activity of MSCR-expressing human CD8+ 310 T cells, in vitro cytolytic assays are carried out with target cells expressing wild-type and mutant HIV-l antigens. HIV-l infected hiiman T cells or gpl6 0-expressing *293 cells are compared with uninfected human T- cells or untransfected 293 cells for their ability to be cytolytic targets for MSCR-expressing.CD8+ T cells. Plasmids plKenv+/rev+/tat-, pIKenvG37OR/rev+ /tat- and are used to generate stably transfected 293 cells as described in Patent #5,359,046,.which express wildtype or mutant env proteins. These gpl6O -expressing 293 cells or HIV-1 infected human T cells -are labeled at 37C for 18 hours 3 with C H]TdR (Roberts et al, Blood 84:2878-2889 washed *and aliquoted to 96-well V-bottom plates at 1x1 4 /well. -Seriall dilutions of MSCR-expressing human CD8+ T cells 'are made up to achieve an effector to target ratio ranging from 100:1 to 0.1:1. Samples -are set, up in triplicate and incubations are carried out- for 6 hours at 37 0 C. Following incubation,..aliquots of the culture :supernatant are removed and counted ina liquid scintillation-counter.. Spontaneous, release is obtained -in a negative control -sampie---acking MSCR-expressing human _CD8.- T cells; maximum release (MR) is obtai'ned .froma -positive c-ontrol sample by lysing target cells with IN HRd. The percent specific iyss.iscalculated from thejfol'1owirg qtion: specific lysis=(SR~ Sample )/Sample,..:M) lO: The pytolyticS;activity of CD8?. T -cells xpsin vaios.MSCRs .and monospecif ic chimeric.-receptors -as -ef fector cells on target .cells..-.inf ected. with various :.HIV-1. isolates target -:b~ls .transfected-with wild-type ,or mutant'-efv ge .nes are -compared.- -In _particular,..MSCRs which: direct the ef ficient cytolysis' of a range -of .primary HIV-l isolates-are. considered good -candidates, for therapeutic application.

Aftdmvle Mvolvtic activity of human CDB T cells ex ressincr EIV-Specif ic Mltifflecif ic ahimeric regqM-taor. -ab(&r10-2C4 This example illustrates that both of the extracellular inducer- *responsive cluttering 'domains on a multispec ific chimeric receptor (MSCR) are able to' -independently recognize their res pective target ligands -and. -signal a cytolytic res ponse in human T -cells. The "chimer ic". construct SAb'(agpl20) -L2-CD4- ,described above in Eale6C, 'contains two ECDs, a single chain antibody domain prepared'--from monoclonal -antibody 447D which recognizes a -specific e'pitope on -the HIV-1 'env glycoprotein and a CD4 extracellular -dofidiin which recogni -z e s a different 20 -region of the env glycoprotein.- To analyze the activity of-'each of these ECDs' separately, target cell lines were'prepar .ed which express-mutant env proteins that 'can be recognized by only one n -ua-C of 'the ECDS. _-SAb'(dgpl2) L2-CDb4-(- was transduced in'tohua D Tlymphocytes' and. the* 2 cytolyti ac.t iv ity 'o 'tee tasduced cells was" tested 'on targ6e cell ine-s 'expriBs si =4e ither w- i type -'env protein or-+the nitAnt' env iroteins.'T&reut of "cytolytic assays-1showed that -ind~pbfidenit ekngag~ment' b-f etr-o'the ECbs ori 'a .MSCR 'was 'sffcen irge vyo~i tiiyi -transduced hum'an T-'l 66h6't'6s Example 10A: vrprain f t~isafbi effiAicn&e jr a ion ~'V~'ctors je e rpared Vh -hallow the 4jffi~i'4Rc 76636io th HIV-1 gp16Oenv Prf6l it IMl~ Ilev/e4tt s :plasmid- which expresses :the- HIV!-1 gpl60env.and rev.-Proteins, bu~t -is missing thef:-HIV-4-l tat: protein:-.- This vector...was! 4es§igned to allow the -ef-ficient bw~reSAOn~d o- env-iinamalian cells, based on the.9Pbservation h 4etteiOf ;the rev gene- produc t isesetalfr pion -of gpl6oehv, expl~ession of the -tat_,.gene prdc ~i~~btisepeso Bird-:et al.

-(1990). LT.- Biol. Chemn.? .1-19157). This plasidaSme -in two steps. The -first 'stop -was the construction of plKenv, which directs the expressioawof three HIV-1 gene products: env, rev- and tat.- Thi's -plakmid -was constructed from' two DNA fragments:. 1),a 4. m eat.obtained by digestion of pIKi. 1 with Eglil "1od the.--cohesive end with T4 DNA polymerase anid.dW m t# en, :followed by digestion wit EoR, ad 3. rt' ecoding -'the env, rev and tat gene producit~ <~eQ 1, -obtained by dgesionofpCMVenv P4s r' &p net al. In USPatent No.

5,359,046) with XhoI, 14.6 aUf o6f the cohesive end with T4 DNA polymerase. and dZP _tc "cr ate a -blunt end,- followed by digestion with EcoRI. T he 'structurei§ of .pIKenv was confirmed by resritin aalsi. I t~esecond step, plKenv+/rev+/tat- was constructed from two 6.8kb vector. fragment obtained by diiidfiaio of the.

cohesive end with'~ ~IP to- create"':'a blunt end, followed by 71 Si 2) a 0.4 kb fragment obtained by digsri )U 04ok et al. (1986) Natur-e 322: 470-474) with, ",0.of the cohesive end with T4 DNA polymeras a blunt end, followed by digestion with Asi8.ki ~r of plKenv+/rev+ /tat- was confirmed by restriction '4 q- Example 10B: 4.4 0-M IZ gpl6Oenv gene with a Mtation in the.

An expression HV-1 gpl.60env protein containing a muain 4bning site was constructed.

67 This :mutant protein can be recognized by -the single chain antibody domain prepared from monoclonal antibody 447D -but not by-the CD4 extracellular domain. pIKenvG370R/rev+/tat- directs the expression of a mutant gpl60env polypeptide in which Glu-370 is -replaced with Arg, -which results :in a -loss of CD4 binding (Olshevsky et al.,-(1990) J. Virol. 64: 5701-5707). :This plasmid was constructed in two steps. In the first step, pIKenvG370R -was made by oligonucleotide-directed mutagenesis using singlestranded pIKenv DNA with oligonucleotide 19 (SEQ ID NO: 19) as the primer. The correct expression plasmid was identified by colony hybridization using oligonucleotide 20 (SEQ ID NO: 20) as a probe, and its structure was confirmed by DNA sequencing. In the second step, pIKenvG370R/rev+/tat- was constructed from two S: DNA fragments: 1) a 6.8kb vector fragment obtained by digesting 15 pIKenvG370R with EcoRI, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by digestion with Asp718, and 2) a 0.4 kb fragment obtained by Sdigesting pIII env 3.1 (Sodroski et al. (1986) Nature 322: 470- 474) with Sal I, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by digestion with Asp718. The structure of pIKenvG370R/rev+/tat- was confirmed by restriction analysis.

6 68 Example 10C: Expression of an HIV-1 .gpl60env gene with a mutation in the 447D binding site An expression .vector :encoding- an HIV-1 gpl60env protein containing a mutation in the binding site for monoclonal antibody 447D was constructed. This mutant protein can be recognized by the .CD4 extracellular domain of the MSCR; but not by. the single chain antibody domain of the MSCR prepared from monoclonal antibody 447D. pIKenvR315Q/rev+/tat- -directs the expression of a mutant gpl60env polypeptide in which Arg-315 is replaced with Gln, which results in a loss of 447D binding (Gorny et al (1992), J. Virol. 66: 7538-7542). This plasmid was constructed in two S steps. In a first step, pIKenvR315Q was made by oligonucleotidedirected mutagenesis using single stranded pIKenv with 15 oligonucleotide 21 (SEQ ID NO: 21) as.the primer. The correct expression plasmid was identified by colony hybridization using oligonucleotide 22 (SEQ ID NO: 22) as a probe, and its structure was confirmed by DNA sequencing. In the second step, pIKenvR315Q/rev+/tat- was constructed from two DNA fragments: 1) 0 a 6.8kb vector fragment obtained by digesting pIKenvR315Q with EcoRI, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by digestion with Asp718, and 2) a 0.4 kb fragment obtained by digesting pIII env 3.1 (Sodroski et al. (1986) Nature 322: 470-474) with Sal I, modification of the cohesive end with T4 DNA polymerase and dNTPs to create a blunt end, followed by digestion with Asp718. The .:structure of pIKenvR315Q/rev+/tat- was confirmed by restriction analysis Example 10D: Infection of :human CD8 T cells with MSCRexpressing retroviral vectors Recombinant -retroviruses -'dncoding -SAb(agpl20)-L2-CD4- were prepared and used to infect purified human CD8* T lymphocytes.

Several populations ofT lymphocytes were also infected with recombinant retroviruses encoding either Sab(agp41) -(pRT43.2E16) or.CD4-< (pRT43.2F3) as control samples.

-A retroviral vector encoding SAb(agpl20)-L2-CD4- was prepared -as -described above in Example 6C :and used to prepare recombinant retrovirus according to the -methods disclosed by Finer et al. in WO-94/29438. Briefly, retroviral stocks were prepared as follows. The packaging vector pIK6 .MCVampac. UT and the retroviral vectors expressing SAb(agpl20) -L2-CD4-;, Sab((agp41) and CD4-( were transiently co-transfected by the calcium phosphate method coprecipitation method into the human tsa54 cell line. tsa54 cells, derived from 293 cells by the transfection of .Large SV40 T antigen (Heinzel et al. (1988) J. of Virol. 62 3738-3746), were grown in DMEM (JRH Biosciences, Lenexa, .5 Kansas), 1 g/l glucose, 10% Donor calf serum (Tissue Culture Biologicals) and split 1:10 every three days. Twenty-four hours following transfection the medium was changed. After an addition twenty-four hours, the viral supernatants were harvested and filtered through a .45 Mm filter and frozen on dry ice.

Human CD8* T lymphocytes were isolated from peripheral blood lymphocytes (PBL) obtained from healthy donors by purification with the CEPRETE LC system (CellPro, Inc.. Bothell, WA), followed by negative selection against CD4 cells using a MicroCELLector (AIS, Inc., Santa Clara,. CA). Immediately after purification, cells were stimulated for twenty-four hours with an equal number of y-irradiated autologous PBMCs in AIM-V media (GibcoBRL, Grand Island, NY) containing 10 ng/ml. of OKT3 monoclonal antibody and 100 units of human IL-2 (Chiron Corp., Emeryville, CA). .Cells are then washed -free of OKT3 and cultured in AR media (50% AIM-V, 50% RPMI,:4 .mM Glutamine,- 20 -mMHepes, -1 mM Na-Pyruvate, non-essential amino acids, and 100: units human IL-2) supplemented -with 5% heat :inactivated -human :AB plasma (Sigma,-St. Louis, MO) The purified CD8* *T lymphocytes, were infected with--the recombinant retroviruses as follows. The T,'-cells were plated at 6 1 x 10 -cells per.well in 24-well tissue culture-dishes in 1 ml AR medium for twenty-four hours.. 0.5 ml -medium .was then removed from each well and replaced with 0.75 ml retroviral .supernatant and 0.75 ml AR medium containing Wg/ml polybrene: (Sigma Chemical Co., Saint Louis, MO) and- 1000 units/ml IL-2 and incubated for an additional twenty-four hours.. 1.5 ml medium was then removed and replaced with 0.75 ml retroviral supernatant and 0.75 ml AR medium containing 2 ug/ml polybrene and 1000 units/mi IL-2 and reincubated for twenty-four. hours.. .This last step was then repeated. After incubation with the viral supernatants, the transduced cells .were harvested by 6 centrifugation and replated in AR medium at 1 x 10 cells/ml, 2 5 ml/well in 24-well dishes. -After five days, the cells were analyzed by FACS as described in Example 8B.

SExample 10E: Purification of CD8+ T lymphocytes expressing chimeric receptors Transduced CD8* T lymphocytes were partially-purified by the use of magnetic beads or by panning on antibody-coated plates. T cells expressing CD4-( or SAb(agpl20)-L2-CD4-. were purified using Dynabeads M450 CD4 magnetic beads and DETACHaBEAD (Dynal Inc., Lake Success, NY) .following the manufacturer.' s protocol.

Cells expressing SAb(agp4l)-tWere -purified by panning on coated plates as follows. Tissue culture flasks were coated with 100 Mg/ml Goat-anti-human IgG (Fc specific) (Caltag Inc., South San

O

Franciscco, -CA) for -2 hours at 37 and then.washed three times 6 with PBS. -Cells were -resuspended in PBS: FCS -at-3.75 X cells/ml, .added to the .washed- flasks, and allowed to. adhere for 1 hour -at room temperature. Flasks were -,then washed. gently four times with.PBS -to remove inon-adherent- cells Adherent cells were released by incubating ,the flasks in AR. medium for two days. _---Example loF: Cytob.ytIa Aati*Lty of SIb(axgp12)-L2-Cb4expressing humian -C68S ate To -determine'; that otV fthe extracellulAr domains of SAb(agpl20)--C4c#i ietyrcgnz theirresptive ligands -on:*the*.*sur et elsadtrigger a -speic -cytolytic rsoa i- ressingT y cytolytic 'assays W6,, iQt' With target cells expressing wild-type and mutant anigens.

93celsstbl e4n~fg(-,w~jL4type and mutant HIV-i gpl6Oenv *were prepared 1o6 ~according 'to -the methods described by C)dZ -P~tefit No. 5,359,046. Plasmids p I Ke nv +r 6%~V3T0R/rev+/tat'-, and pI~enV315Q/.~e'vt/ generate stably transfected 293~ cell 2 9~n 3 env"), enV with a muato in tt' "93C4- and ed with mutation in the 4~~~binding site ("293/447D-"), respectively. The ~~iepesng -mutan orwl-type env were prepared as cytolYltict, ,,'argets to be used 'in a chromium release assay to m ea slr uk"4.,-c'y"0 l activity (Current Protocols In Immunolog,p ,S Johh Wiley So"~,In.

1992). 293. tran"_ ge0 alond C(m2' 3n were also prip( a~T~a237 C for 1-2 hours'it L.6~(%S/L~'cii 500. mCi/mg, AmgiW', #06 Ci2X 10 -cells), washed and aliqUoted to -P at -1l 0 /well.

To perform theat 9--1 ~il :diiutions-of L2 -CD4- -expressin~ lymphocyte& (h efectorcells) were add t d-;.so cieeailffcort targetamj IET le s-were set inh tripli& 0r11 6.r~ie otor hours -at 37*C. .7 1~n 1h lultur supernatant were removed iq i liquid scintillation coutr -Spontaneous :release is obtained in.a negative control. sample lacking-effector .cells; maximum:release (MR) is obtained from a .positive .control sample by -lysing: target. cells with -lNHC1. The percent specific lysis.:is calculated from the .following -equation: specific lysis=(SR Sample-)/(Sample^ MR )xl00O%..

The results of .the cytolytic assays involving -the CD8 effector cells and 293 target cells bearing the wild-type or mutant env proteins are shown in Figure CD8 T lymphocytes expressing .the multispecific- chimeric_ receptor Sab(agpl20)-L2-CD4-( specifically lysed 293 cells expressing wild-type..env (293 env) as well as CD8 T lymphocytes expressing the monospecific chimeric receptors Sab(agp41)-1 or CD4-(. Untransduced. T lymphocytes were incapable of lysing any of the target cells.

The ligand which was recognized by these cytoloytic cells was on the HIV-1 env protein, since target cells lacking that molecule (293 neo) were not lysed by any of the cytolytic cells.

Cytotoxic T lymphocytes expressing the multispecific chimeric receptor SAb(agpl20)-L2-CD4- .were also able to specifically lyse both of the 293 targets which expressed the mutant proteins on their cell surfaces. These MSCR-expressing cells were able to kill target cells expressing mutant gpl60 molecules lacking the binding site for CD4. In contrast, 293/CD4- cells were resistant to killing by CD4- -expressing cells. The SAb(agpl20)-L2-CD4- -expressing cells were therefore able to lyse the 293/CD4- target cells by virtue of the binding of their 447D single chain antibody extracellular domain to the CD4- mutant env proteins. CD8* T-cells expressing SAb(agpl20)-L2-CD4-( were also able to lyse 293/447D- target cells expressing mutant molecules lacking the epitope for monoclonal antibody 447D (Gorny et al, (1992)J. Virol. 66: 7538-7542). The SAb(agpl20)-L2-CD4- ?-expressing cells were therefore able to lyse the 293/447Dtarget cells by virtue of the binding of their extracellular CD4 domain to the 447D- mutant envelope proteins. Accordingly, this example demonstrates that the two extracellular binding domains on :the multispecific chimeric: receptor, SAb(agp20) -L2-CD4-.

were each able to independently bind to the -target envrmolecules and .signal the cytoplasmic domain to initiate the cytolytic activity of the CD8 human T lymphocytes,- All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent- as if each individual publication or patent .0 application was specifically and individually indicated to be incorporated by reference.

The invention now being fully described, it will be apparent to one of ordinary skill in the art that many changes and **35 modifications can be made thereto without departing from the spirit or scope of the appended claims.

*7 o* a.

S. 74 SEQUENCE LISTING GENERAL INFORMATION: APPLICANT: CAPON, DANIEL J SMITH, DOUGLAS H TIAN, HUAN WINSLOW; GENINE A SIEKEVITZ, MIRIAM *0 (ii) TITLE OF INVENTION: MULTISPECIFIC CHIMERIC RECEPTORS (iii) NUMBER OF SEQUENCES: 26 (iv) CORRESPONDENCE

ADDRESS:

ADDRESSEE: CELL GENESYS, INC.

STREET: 322 LAKESIDE DRIVE CITY: FOSTER CITY STATE: CALIFORNIA COUNTRY: US ZIP: 94404 COMPUTER READABLE FORM: MEDIUM TYPE: Floppy disk COMPUTER:-IBM PC dompatible- OPERATING SYSTEM: PC-DOS/MS-DOS SOFTWARE: PatentIn Release Version #1.25 (vi) CURRENT APPLICATION DATA: APPLICATION

NUMBER:

FILING DATE:-

CLASSIFICATION:

(Viii) ATTORNEY /AGENT INFORMATION: NAME: KRUPEN, KAREN I REGISTRATION NUMBER: 34, 647 REFERENCE/DOCKET NUMBER: CELL 18 (ix) TELECOMMUNICATION INFORMATION: TELEPHONE: (415) 358-9600 x131 TELEFAX: (415) 349-7392- 0*00 *to be&* (2 NOMTO FRSQI 01 :*of.

:0(i EUNC HRCTRSIS 215 (i)MLEUETYE-. DA .gnmc (iFORMQTINC FORSEQIPINO:,: EI.NO CTCTCA 0.GCGGAGAAGTGCA .36 INFORMATION FOR, S.Q ID 2: Wi SEQUENCEt C~ W cs: LEIITE: TYPE: nu'; q Ac aid STRAWOft S3 TOPmOOG:1iea (ii JqEUE (xi) SEQUIENCZ b~ 0 i EQI O2 ACAGACTGTG AGGAGA 16 20 INFORMATION. FORS ()SEQUENf TY.t 4 1 w.

(ii) MOLECULE TYPE (etoa (xi) SEQUENCE-V -hSR1~~~ ID NO -3: CTGCGTCAAC ACAGACTGAC CCTTACCCTC AGAAGATTA CCCGACCCCG AGGTCGAOCC TGAGGAGACG GTGACCAG 78 INFORMATION FOR SEQ ID NO:4: SEQUENCE CHARACTERISTICS: LENGTH: 16 base pairs vIB) TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear .00o 15(ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:4: AGAAGATTTA CCCGAC 16 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 111 base pairs TYPE: nucleici-:acid) STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION:.SEQ ID CTGCGTCAAC ACAGACTGAC CGTCCTTCTT AGCGTCGTCC TTCTTAGCGT CGTCCTTCTT AGCAGCGTCC TTCTTAGCGT CGTCAGCGGA AGATGAGGAG, ACGGTGACCA G INFORMATION FOR SEQ I D NO:6: Wi SEQUENCE CHARACTERISTICS: LENGTH: 16 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (i)MOLECULE TYPE: DNA (genomic) o (xi) SEQUENCE DESCRIPTION: SEQ ID NO:6: GCGTCGTCCT TCTTAG 16 INFORMATION FOR SEQ ID.NOi7::- T t SEQUENCE CHARACTERISTICS: LENGTH: 66 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) SEQUENCE DESCRIPTION: SEQ ID.NO:.7*- In CTGCGTCAAC ACAGACTGTG GGGACGGT4GG GGATGTGTGA GTTTTGTCTG AGGAGACGGT

GACCAG

:66 INFORMATION FOR SEQ ID NO-:8: (i)SEQUENCE CHARACTERISTICS: LENGTH: 16 base pairs, TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: :SEQ D CGGTGGGGAT GTGTGA 16 INFORMATION FOR SEQ :ID NO: 9:- SEQUENCE CHARACTERISTICS: LENGTH: 87 base pairs TYPE: nucleic. acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) 20 a. (Xi) SEQUENCE DESCRIPTION: SEQ ID NO:9: CTGCGTCAAC ACAGACGGT CCAGCTCCCC GTC!CTGCGCT TCGGCqCTCG ATTCTTCCAG TTGCAGCTCT GAGGAGACGG

TGACCAG

87 INFORMATION FOR SEQ-IID"NO:1O-e SEQUENCE CHARACTERISTICS: LENGTH: 16 base iairs' TYPE: nucleic acid STRANDEDNESS: single' TOPOLOGY: linear- (ii) MOLECULE TYPE: DNA. (geriomi-c) (xi) SEQUENCE. D .i EQ -ID TCGGCGCTCG ATTCTT 16 1 INFORMATION FAD0., SEQ O~1 TOPOLG:lna (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCtE bigo'e NO jib.

*GCCCAGCACC ACTTT ~4 ~CGT 36 N INFORMATIONW~R~ I Q:2~

SEQUENCE,!C

:82 TOPOLOGY: linear (ii) MOLECULE TYPE: DNA -(genomic)-:- (xi) SEQUENCE DESCRIPTION:. SEQ ID NO:12: ACTTTCTTTG, AGCTCA 0 16 INFORMATION FOR SEQ ID NO:13: SEQUENCE CHARACTERISTICS: ~L5(A) LENGTH: 78 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear o 20 (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:13: GCCCAGCACC ACTTTCTTAC CCTTIACCCTC -AGAAGATTTA CCCGACCCCG AGGTCGACCC TGAGCTCACG GTGACCGT 78 INFORMATION FOR SEQ ID NO:14: SEQUENCE CHARACTERISTICS: LENGTH: 111 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) .1 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:14: GCCCAGCACC ACTTTCTTAC CGTCCTTCTT AGCGTCGTCC TTCTTAGCGT CGTCCTTCTT AGCAGCGTCC TTCTTAGCGT CGTCAGCGGA AGATGAGCTC ACGGTGACCG T 111 a. INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 66 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID GCCCAGCACC AC'1?rCT'TG, GGGACGGTGG GGATGTGTGA GT'ITTGTCTG AGCTCAE43GT

GACCGT

97. *60020 INFORMATION FOR SEQ ID NO:16: SEQUENCE CHARACTERISTICS: LENGTH: 87 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:16: GCCCAGCACC ACTTTCTTGT CCAGCTCCCC GTCCTGCGCT TCG GCGCTCG ATTCTTCCAG TTGCAGCTCT GAGCTCACGGTGACCGT 87 INFORMATION FOR SEQ ID N0:17: SEQUENCE CHARACTERISTICS:.

LENGTH: 42 base pairs TYPE: nucleic acid STRAkNDEDNESS:, single (D)LTPOLinear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:17: TAGTCTAGGA TCTACTGGCT GCAGTTCTTG*CTCTCCTCTG TC 42 INFORMATION FOR SEQ ID NO:18: SEQUENCE CHARACTERISTICS: LENGTH: 16 bAse pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear 0 (ii) MOLECULE TYPE: DNA (genomic) SEQUENCE DECRIPTRTI:SQ-: N~ 8 LENGTH: 3 9 base pairs TYPE: pucleic (C)STMDDES:snl (ii) moLjECULE. Azo c) (xi) SEQUEN~CE DESCRIPfTION: SEQ ID NO: 19: LCTGTGCE GTTQCMW7 ZCc.. .TC CTGAGGA 39 INFORMAWI9T LE 1'1par tE ntcle c. acid- STRANWOE,,"gl (ii) MOLECbtLk," (xi) SEQUW ID NO:t2 0: o o SEQUENCE CHARACTERISTICS: li.. LENGTH: 39 base.pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:21: TCCTATTGTA ACAAATGCTT GCCCTGGTCC TCTCTGGAT 39 INFORMATION FOR SEQ ID NO:22: SEQUENCE CHARACTERISTICS: LENGTH: 16 base pairs TYPE: nucleic acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: DNA (genomic) (xi) SEQUENCE DESCRIPTION: SEQ ID NO:22: 20 AAATGCTTGC CCTGGT 16 -I -n, c' u INFORMATION FOR SEQI-ID.NO:23:'---* i)SEQUENCE CHARACTERISTICS:...

LENGTH: 14 amino acids TYPE:' amino acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: peptide *j5

S

S

20 (xi) SEQUENCE DESCRIPTION: SEQ ID NO:23: Gly Ser Thr Ser Gly Ser Gly Lys Ser Ser Glu Gly Lys Gly INFORMATION FOR SEQ ID NO:24: SEQUENCE CHARACTERISTICS: LENGTH: 25 amino acids TYPE: amino acid STR.ANDEDNESSt single TOPOLOGY:' linear S. S.

S

S (ii) MOLECULE TYPE;- peptide-.

(xi) SEQUENCE DESCRIPTION: SEQ-IDNO:24: Ser Ser Ala Asp Asp Ala Lys yLs Asp Ala Ala Lys Lys Asp Asp Ala 1 5 10 Lys Lys Asp Asp Ala Lys Lys Asp Gly '0 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 10 amino acids TYPE: amino acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: peptide •0 (xi) SEQUENCE DESCRIPTION: SEQ ID. Asp Lys Thr His Thr Ser Pro Pro Ser Pro 1 5 INFORMATION FOR SEQ ID NO:26: SEQUENCE CHARACTERISTICS: LENGTH: 17 amino acids TYPE: amino acid STRANDEDNESS: single TOPOLOGY: linear (ii) MOLECULE TYPE: peptide (xi) SEQUENCE DESCRIPTION: SEQ ID NO:26: Glu Leu Gin Leu Glu Glu Ser Ser Ala Glu Ala Gin Asp Gly Glu Leu S. *5

S

S

S

Asp Throughout this specification and the claims which follow, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

1.1 1 u.

r.

i i 1.

c

Claims (46)

1. A chimerj ,Is" e.;e encoding a membrane bound protein, said DNA seqelK' fram. .Nie: a DNA sequence &N sigal se6quenkce;' a DNA sequence" eW\ .t least two extracellular inducer- responsive cludto~t41 no to form a multispecific extracellular iri4,,q J*ive clustering domain that binds spciic:Ly' to c $nUer mlcle which results in the- 4imeiz J 'M 4~4n of said m~ultispecific extracellular 4 :.~t~~nembrae domuain; and uction signalitig domain which encodes' a" gt~idces an ef fect or function signal in a hobt I-.-O whereinc~i sad"e'xtracellular domain and effector funcionsignaling. d~ml, V I: e riot. naturally joined together, and when said chimeric DNA, qI#We -is expressed as a membrane bound protein in a sner odi i. su*~a~ o expression, Ca4~ qaaitatel for uci'~~Y~ ~1J'effector fpo uncding. to eas t one inducer molecule

2. The 1* rgt, claim 1 wereinsi ef fector -fud~i is eleted from the group consisting of thnef 03191 of. the. T-cell receptor, the bet and, 10mu rCetr, the MB1(Ig alpha) and B2 9 (Ig beta), th.p th CD3 the T-cell: receptor, and th y bieIinases 92 A chimeric DNA sequence :encoding a membrane bound protein, said DNA sequence -comprising iin reading -frame: a DNA sequence :encoding a signal sequence; a DNA sequence encoding at least two extracellular inducer- responsive clustering domains- to form a multispecific extracellular inducer-responsive -clustering -domain that binds specifically to at least one inducer molecule which results in the dimerization or oligomerization of said multispecific extracellular domain; a DNA sequence encoding a transmembrane domain; and a DNA sequence encoding a proliferation signaling domain which encodes a polypeptide that signals the cells to proliferate, wherein said multispecific extracellular domain and proliferation signaling domain are not naturally joined together, and when said chimeric :DNA sequence is.expressed as a membrane bound protein in a selected host cell under conditions suitable for expression, said membrane bound protein initiates a signal for proliferation in said host cell upon binding to at least one inducer molecule.

4. The DNA sequence according to -claim 3, wherein said S proliferation signaling domain is the cytoplasmic portion of a member-of the cytokine "receptor protein superfamily that does not contain a kinase domain.- The DNA -sequence according: to -claim 4, wherein said cytokine receptor protein superfamily is selected from the group consisting :-6f :the- interleukin-2 receptor subfamily,- the interleukin-3 subfamily, the .interleukin- 6 receptor: subfamily and combinations thereof r i. T.

6. The DNA sequence according to claim 3, wherein -said proliferation signaling domain is. selected from the group consisting of interleukin-2 receptor beta, protein interleukin-2 receptor gamma protein, and combinations thereof. The DNA sequence :according to 'claim 3 wherein said proliferation signaling domain is .selected from the eukaryotic family of Janus tyrosine kinases. 0 8. The DNA sequence according to claim 3, wherein said prliferation signaling domain is the cytoplasmic portion of a member of the eukaryotic tyrosine kinase growth factor receptor S: superfamily. S

9. The DNA sequence according to claim 3 wherein at least one said extracellular inducer responsive clustering domain is an antibody or single-chain antibody or portions or modifications thereof containing inducer binding and clustering activity.

10. The DNA sequence according to claim 3 wherein at least one said extracellular inducer responsive clustering domain is a cell differentiation antigen.

11. The DNA sequence according to claim 1 or 3 wherein at least one said extracellular inducer-responsive clustering.domain is CD4 and at least another said-extracellular inducer-responsive clustering domain is an antibodycor single chain.antibody? S12. The ,DNA sequence according to- claim 9 wherein at-:least one said antibody or single chain .antibody recognizes .an antigen from a virus selected from the group consisting of HIV, hepatitis A, B and C viruses, Kaposi's sarcoma associated Herpes virus, Herpes Simplex -viruses, Herpes Zoster-virus, cytomegalovirus, papilloma virus, respiratory .syncytial virus and influenza virus.

13. The DNA sequence according to claim 11 wherein at least said antibody or said single chain antibody recognizes an antigen on a cancer .cell selected from the group -consisting of interleukin 14 receptor, CD19, CD20, Lewis Y-antigen, CEA, Tag72 antigen, EGF-R and HER-2.

14. The DNA sequence according to claim 1 or 3 wherein said transmembrane domain is naturally associated with said extracellular inducer responsive clustering domain.

15. The DNA sequence according to claim 1 or 3 wherein said transmembrane domain is naturally associated with said proliferation signaling domain. a>

16. A chimeric DNA sequence encoding a membrane bound protein, said DNA sequence comprising in reading frame: a DNA sequence encoding a signal sequence; a DNA sequence encoding at least two extracellular inducer- S: responsive clustering domains to form a multispecific extracellular inducer-responsive binding clustering domain, that binds specifically to at least one inducer molecule which results in the dimerization or oligomerization of said multispecific extracellular domain; a DNA sequence encoding:a transmembrane.-domain; a DNA sequence encoding a proliferation signaling domain which encodes a polypeptide that signals -the cells to proliferate; and a DNA sequence encoding a- cytoplasmic :effector .function signaling domain which encodes a polypeptide that transduces an effector function signal in a host cell; wherein said multispecific extracellular domain and .proliferation signaling domain are not naturally joined together, and when said chimeric DNA sequence is expressed as a membrane bound protein in a selected host cell under conditions suitable for expression; said membrane bound protein initiates .a signal for proliferation and-effector function in said host cell upon binding to at least one-inducer molecule.._

17. A chimeric DNA sequence encoding a membrane bound protein, 3 said DNA sequence comprising in reading frame: a DNA sequence encoding a signal sequence; S.a DNA sequence encoding at least two extracellular inducer- responsive clustering domains to form a multispecific extracellular inducer-responsive clustering domain that binds specifically to at least one inducer molecule which results in the dimerization or oligomerization. of said multispecific extracellular domain; a DNA sequence encoding a transmembrane domain; a DNA sequence encoding a cytoplasmic effector function signaling domain which encodes a polypeptide that transduces an effector function signal in a-host cell; and a DNA .sequence encoding a proliferation signaling domain which encodes a polypeptide that signals.the cells to proliferate; Swherein said multispecific .extracellular domain and proliferation domain are not naturally joined together, and when said chimeric DNA sequeinie is expressed as a membrane bound protein in .a.-selected- host cell under conditions suitable for Sexpression,- said membrane bound protein initiates a signal for .proliferation and effector- function .in said host: cell. upon binding to _at ?least one inducer molecule. The DNA sequen0ce accordintg to -claim .16 or '17, wherein said proliferation8 a-0 toPni~ ~esberof the-..ytoplAsmic protiofao memo" 'o _4 ,n reetrpoeiuefml cytokne rcepto proL~4 s~4amly i slce from teroupi consting of -ambet-o" iniJin-2receptor sufmit th ha dobpos nt co 'Ai.OM

20-h 17, wereai sai i s-slcted from the group 1*okn recepto proqloe.n, 4nterleukin- consisting of ith nelui-*P k-pjd- receptor ubfamil and~I&onbiain teef Te NAseuece&cori tWcai m 16 or 17, wherein i 2 The 71~ 1 1 sai OX is46 th"A"p"mcpoto 1 cofnaimeine of pr o+n knse groth factor receptor supe. Pr q 44in h I f

23. The DNA seqneI, kcria to claim 16 or 17, wherein satd iss. ane aztdy-t-orliM 'portion o i modifiif erqcrbndn idclxtrn activity.d gro -fato 97 :24. The DNA sequence according to claim 16 or 17, wherein at least one said extracellular inducer responsive clustering domain is-a cell differentiation -antigen. The DNA sequence according to claim 16 or 17, wherein at least one said extracellular inducer-responsive clustering domain is CD4 and at least one other said extracellular inducer- responsive clustering domain is a single chain antibody.

26. The DNA sequence according to claim 16 or 17, wherein said single chain antibody recognizes an antigen from a virus selected from the group consisting of HIV, hepatitis A, B and C viruses, Kaposi's sarcoma associated Herpes virus, Herpes Simplex viruses, Herpes Zoster virus, cytomegalovirus, respiratory syncytial virus, influenza virus and papilloma virus.

27. The DNA sequence according to claim 16 or 17, wherein at V least one said single chain antibody recognizes an antigen on a cancer cell selected from the group consisting of interleukin 14 receptor, CD19, CD20, Lewis Y antigen, CEA, Tag72 antigen, EGF-R S. and HER-2.

28. The DNA sequence according to claim 17 wherein said transmembrane domain is naturally associated with said multispecific extracellular inducer-responsive clustering domain or said cytoplasmic effector function signaling domain. The DNA sequence according.to claim 16 wherein said transmembrane domain -is:;naturally joined to. said multispecific extracellular inducer-responsive clustering domain or: said proliferation signaling domain. 98 The DNA sequence according: to claim 16 or- 17, .wherein said effector function signaling domain is selected from the .group -consisting of:- the .:zeta and -eta -chains- of the T-cell receptor, the beta and gamma:chains -of :FceR. receptor, the MB1(Ig alpha) and B29 (Ig beta) chain of the B cell receptor, the protein, the CD3 delta, gamma and epsilon chains of the T-cell :receptor, and the syk and src families-of -tyrosine kinases.

31. A chimeric DNA sequence encoding an extracellular and intracellular multispecific chimeric receptor protein, said DNA sequence comprising in reading frame: a DNA sequence encoding a signal sequence; a DNA sequence encoding at least two extracellular inducer- responsive .clustering domain to form a -multispecific extracellular inducer-responsive clustering domain that binds specifically to at least one inducer molecule which results in the dimerization or oligomerization of said multispecific extracellular domain; *E a DNA sequence encoding a transmembrane domain; a DNA sequence encoding a proliferation signaling domain which encodes a polypeptide that signals a host cell to.proliferate; *o and Sa DNA sequence encoding a intracellular inducer-responsive clustering domain that binds specifically to at least one inducer molecule.which results in .the dimerization or oligomerization of said intracellular domain; I wherein said multispecific extracellular domain and proliferation domain are.not naturally joined together; and when said chimeric DNA sequence is expressed as.a'hybrid :extracellular .and intracellular receptor protein in a'.selected host cell under conditions suitable for expression, said .receptor:-protein initiates a signal for proliferation in said host cell upon -99 Sbinding to at least one inducer-molecule.- -32. A chimeric DNA sequence encoding. a hybrid multispecific -chimeric receptor protein, said .DNA isequence;-comprising in reading frame: z' sa a DNA sequence encoding a signal sequence; a DNA sequence encoding at least two extracellular inducer- responsive clustering domains to form a multispecific extracellular inducer-responsive clustering .domain that binds specifically to at least one inducer molecule which results in the dimerization or oligomerization of said multispecific extracellular domain; a DNA sequence encoding a transmembrane domain; a DNA sequence encoding a proliferation signaling domain which encodes a polypeptide that signals a host cell to proliferate; .a DNA sequence encoding a effector function signaling domain which encodes a polypeptide that transduces an effector function signal in a host cell 0. a DNA sequence encoding a intracellular .inducer-responsive clustering domain that binds specifically to at least one inducer molecule. which results in the dimerization or oligomerization of said intracellular domain; Swherein: said multispecific extracellular .domain .said proliferate.domain said effector function signaling-domain and intracellular induced-responsive clustering domain -are -not naturally joined together; and-when.said chimeric.DNA sequence is expressed as a hybrid multispecific :chimeric receptor protein in a selected-host -cell under:conditions suitable .for expression, said .receptor protein. initiates a.isignalfor:proliferation and effector function -in said .host cell- upon -binding .tosat least one Sinducer molecule. r 1 "1 DI_ o 7.: 100

33. The DNA sequence according to-claim 31 or 32 wherein said intracellular inducer -responsive clustering domain binds to a -natural or synthetic .inducer that is cell membrane permeable and induces the clustering of said intracellular inducer responsive domain.

34. The DNA sequence according to claim 31 or 32 wherein said intracellular inducer responsive clustering domain is selected from the group of immunophilins, cyclophilins and steroid 3 receptors.

35. The DNA sequence according to claim 32 wherein said effector function signaling domain is selected from the group consisting of the zeta and eta chains of the T-cell receptor, the 15 beta and gamma chains of FceR1 receptor, the.MBl(Ig alpha) and o B29 (Ig beta) chain of the B cell receptor,the BLVgp30 protein, the CD3 delta, gamma and epsilon chains of the T-cell receptor, and the syk and src families of tyrosine kinases.

36. The DNA sequence according to claim 3. or 32, wherein said proliferation signaling domain is the cytoplasmic portion of a member of the cytokine receptor protein superfamily that does not contain a kinase domain.

37. The DNA sequence according to claim 31 or 32 wherein said proliferation signaling -domain is selected from the eukaryotic family of Janus tyrosine kinases.

38. An expression cassette comprising a transcriptional initiation region,- a DMA sequence according to claim 1 under the transcriptional control of said transcriptional initiation region and a transcriptional termination region. '"101 .0 1 0 0 0 3. n.expression cassette comprising;* a trans cript iroea 1 -initiation regionF--:a, ,,,,quece -according :t6 claim:3..under the -transcriptional obtI70L- p~-ran criptional initiation' region and a transcriptionail- t. M.11Mation1 -region. An expresqi9~ et comprising a transcriptional initiation region, A. 4W e;accordinmg 'to claim* 16 under the transcriptional oi.r2?Q 4tncriptional initiation region and. a transcriptiqnal. trn4kjli# region. l. An e'c.0. om: i sing a transcriptional initiation, req.~ )J cording to claim 17 under the transcriptiona, .1,0 itiit~*ariptional initiation region and a tra%$e..4 e* A42. An expreuip cmpisn a transcriptional initatin rgio, a~ quxi6 accordngt claim 31 under the transcriptional cotrol 6fi,- $id'transcriptional initiation region and a transcriptional tlr~&tonrgin

43. An expt Mi ig .Otoa initiation regioni,'&' 4e od t li~~n~.the transcriptional d641 to-5t 1~ 1' initiat ion -region alternative:~e. functional in a WMwu ~x&~gto claims 38"A43 in the Otioital. .initiat ion region is ,.4.nce ;:accord ngrtb .claim 1. Oquen ce according to claim 3.

45.- A -celL,:h.4

46. A cell c, 4> 0 2

47. A cell comprising a DNA sequence according to claim-46.

48. A cell comprising a DNA sequence according to claim 17.

49. A cell comprising a DNA sequence according to claim 31. A cell comprising a DNA sequence according to claim 32.

51. A cell comprising a DNA sequence that encodes a chimeric effector function receptor comprising an extracellular inducer- responsive clustering domain, a transmembrane domain, and a effector function signaling domain and a second DNA sequence according to claim 1. 15 52. A cell comprising a DNA sequence that encodes a chimeric effector function receptor comprising an extracellular inducer- responsive clustering domain, a transmembrane domain, and a effector function signaling domain and a second DNA sequence according to claim 3.

53. A cell comprising a DNA sequence that encodes a chimeric effector function receptor comprising an extracellular inducer- responsive clustering domain, a transmembrane domain, and a effector function signaling domain, and a second DNA sequence according to claim 16. A cell comprising a DNA sequence that encodes -a chimeric effector: function receptor comprising an extracellular inducer- responsive clustering domain, a transmembrane domain, and a effector function signaling domain, and-a-second DNA sequence according to claim 31. 103

55. Thecell, according :to claims 45-54 in the alternative, wherein said cell is a mammalian cell.

56. The cell according to claims 45-54 in the alternative, wherein said cell is a .human cell.

57. A chimeric protein comprising in the N-terminal to C- terminal direction: a multispecific extracellular inducer-responsive clustering 9 domain consisting of a portion of a surface membrane protein or secreted protein that binds specifically to at least one inducer- molecule which results in the dimerization or oligomerization of said multispecific extracellular domain; a transmembrane domain; and *15 a cytoplasmic effector function domain polypeptide which transduces an effector signal -in a host cell; wherein said extracellular domain and cytoplasmic effector function domain are not naturally joined together, and when said chimeric protein is expressed as a membrane bound protein in a selected host cell under conditions suitable for expression, said membrane bound protein initiates a signal for effector function in said host cell upon binding to at least one inducermolecule.

58. A chimeric protein comprising in the N-terminal to C- terminal direction: a multispecific extracellular inducer-responsive clustering domain.consisting ,of a portion of a -surface membrane protein or secreted protein that binds, specifically to :at .least one -inducer- molecule which results in the.dimerization or: oligomerization of said multispecific -extracellular domain; a transmembrane domain; and a proliferation signaling domain of a polypeptide that signals 104 .the.:cells to.proliferate, wherein said.: multispecific extracellular domain and proliferation domain are not naturally joined .together, .and when said chimeric protein is expressed as a-.membrane-bound protein in a-selected host cell under :conditions suitable for .expression, said membrane bound protein-initiates a signal for proliferation in said host cell upon binding to at least one -inducer molecule. A chimeric protein comprising in the N-terminal to C- terminal direction: a multispecific extracellular inducer-responsive clustering domain consisting of a portion of a surface membrane protein or secreted protein that binds specifically to at least one inducer- molecule which results in the dimerization or oligomerization of 15 said multispecific extracellular domain; a transmembrane domain; a proliferation signaling domain of a polypeptide that signals the cells to proliferate; and a cytoplasmic effector function domain polypeptide which transduces an effector signal in a host cell; wherein said extracellular domain and proliferation domain are not naturally joined together, and when said chimeric protein is expressed as a membrane bound protein in a selected host cell under conditions suitable for expression, said membrane bound protein initiates a signal-for proliferation: and effector function in said host cell upon-binding to at least one inducer molecule. A .chimeric protein: comprising -in -the N-terminal to C- terminal direction: a ::multispecific .extracellular -inducer-responsive clustering :.domain consisting of a portion-of a:surface membrane protein or '105 secreted protein that binds specifically -to .at least one inducer- -molecule which .'results in the dimerization or oligomerization of said--multispecific extracellular domain; irb -a transmemzbrane domain; cytoplasmic effectdr .functi6n domain 'polypeptide which transduces an effector :function signal.. in a host cell; and azproliferation signaling domain of a polypeptide -that -signals the cells to proliferate; .wherein said multispecific extracellular domain and proliferation domain -are not naturally joined together,-. and when Ssaid..chimeric protein is expressed -as a membrane bound protein in a selected host cell under conditions suitable for expression, said membrane bound protein initiates a signal for proliferation and effector function in said host cell upon binding at least one 15 inducer molecule.

61. A chimeric hybrid binding .:proliferation protein comprising in the N-terminal to C-terminal direction: a multispecific extracellular inducer-responsive clustering domain consisting of a portion of a surface membrane protein or secreted protein that binds specifically, to. at least one inducer molecule which results in-:the :dimerization or oligomerization of said multispecific extracellular domain;...... a transmembrane domain; a. aproliferation signaling doinari of a polypeptide that signals the cells to proliferate;-::i§ c. r i: a an intracellular inducer-responsive clustering doniain-ithat binds specifically to at least one inducer molecule which results in the -dimerization i or::bligomerization of- said :iritracellular domain protein; and L 0 o :wherein: i said multi specific d extracellular adoimain and proliferation :domain are not naturally j oiedtogether;- and when 106 said-,,chimeric hybrid-. biridig p rolifeiation protein is expres-sed as-apoi -receptor. in a eelected -host cell-under conditions -suitable:. -for :e t oj-n~receptor: initiates a" signal for proliferation iki 3t ell: upon binding to either-said inducer molecu'le.W rof.

62. A chmrc b~s idin poliferatio protein comprising in the. Io C-tierminal direction: a multispecific indae-responsive clustering *domain: ot1; msulrface membrane protein or Ssicp te t,1eo' ait least .One -inducer molecule wik ~zibn or oliomerization of said mult isp eil proliferat'ii, "0 5 Y of aplptiethat signals the cells to p*4' acytoplas-mi i- domain polypeptide which transduces an eff ectA ti si.Lgnal in a host cell; an intracellular i11 Monsive clustering domain that binds specif icakl' ":ce moeule which results in. the. d"O4Pro sadie~ua domain protein* wherein sal54 -xdiuar domain and proliferation.''~yj~edtgehr and when said- chimeric i, 1 r66se a a aerti receptor qr.Lo bnitins. suitable" for expression, 84"1 *o-ihidtiates a signal for 3nducer- mlecuJ~ a~ef 3 0 .63.:_"__'A4me4t ~4~r ctidn-in -a mammalian host comprisin a. introducing a multispecific 'chimeric receptor -constract into autologous CD8* cytotoxic T cells under conditions suitable for expression to produce receptor expressing cytotoxic T cells; and b. introducing said receptor expressing cytotoxic T cells into a mammal such that said receptor expressing cytotoxic T cells proliferate and kill cells infected with at least one virus.

64. The method of claim 63 wherein said virus is selected from the group consisting of HIV, hepatitis A, B and C virus, hepatitis B virus, Herpes Simplex viruses, Herpes Zoster virus, Kaposi's sarcoma associated Herpes virus, cytomegalovirus, respiratory syncytial virus, influenza virus and papilloma virus.

65. The method of claim 63 wherein said receptor is the receptor of claims 1, 3, 16, 17, 31 or 32.

66. A method of inducing a cell to proliferate comprising introducing a multispecific chimeric receptor construct into a cell under conditions suitable for expression, to produce a receptor expressing cell and contacting said receptor expressing cell with a target inducer.

67. -The-method of claim 66 wherein said cell is selected from the group consisting of a nerve cell, a keratinocyte cell, islet of Langerhans cell, a muscle cell, or a.hematopoietic cell.

68. The DNA sequence according to claim 1 wherein at least one said extracellular inducer responsive clustering domain is an antibody or single-chain antibody or portions or modifications thereof containing inducer binding and clustering activity. 108

69. The DNA sequence according to claim 1 wherein at least-one said extracellular inducer responsive clustering domain is a cell differentiation antigen.

70. The DNA sequence according to claim 68 wherein at least one said antibody or single chain antibody recognizes an antigen from a virus selected from the group consisting of HIV, hepatitis A, B and C viruses, Kaposi's sarcoma associated Herpes virus, Herpes Simplex viruses, Herpes Zoster virus, cytomegalovirus, "0 papilloma virus, respiratory syncytial virus and influenza virus. *f

71. The DNA sequence according to claim 69 wherein at least said antibody or said single chain antibody recognizes an antigen on a cancer cell selected from the group consisting of interleukin 14 receptor, CD19, CD20, Lewis-Y antigen, CEA, Tag72 antigen, EGF-R and HER-2. f DATED this 23rd day of March, 2000. Cell Genesys, Inc By its Patent Attorneys DAVIES COLLISON CAVE 109