WO1999036102A9 - GENETIC MANIPULATION BY 3' UNTRANSLATED SEQUENCES OF mRNA - Google Patents

Abstract

The invention provides methods for the alteration of mRNA function and the regulation of gene expression by 3' untranslated nucleotide sequences of mRNA. The invention further provides competitive binding assays for 3'UTR binding proteins employing 3'UTR sense nucleotide sequences.

Description

GENETIC MANIPULATION BY 3 ' UNTRANSLATED SEQUENCES OF mRNA

BACKGROUND OF THE INVENTION

The 3 ' untranslated regions (UTRs) of messenger RNAs have been demonstrated to have biological functions, including regulation of mRNA stability, mRNA translational efficiency, pattern formation in Drosophila , and mRNA localization. Most recently, the 3 'UTRs of specific muscle structural genes have been shown to act in trans as regulators of growth and differentiation. In addition, it has been demonstrated that exogenous expression of the 3'UTR of the skeletal muscle tropomyosin mRNA in chicken embryonic fibroblasts induce the cells to become spindle shaped, fuse, and express markers of muscle differentiation. The mechanism by which certain 3 'UTRs act to complement differentiation-defective mutants or to mediate differentiation is unknown.

Expression of the gene for the neuronal growth associated protein GAP-43 (B-50, Fl, pp46, neuromodulin) constitutes an important step in the differentiation of PC12 cells in response to nerve growth factor (NGF) . A major mechanism for the control of GAP-43 expression in PC12 cells involves the modulation of GAP-43 mRNA stability by selective changes, this process is mediated by the interaction of highly- conserved sequences in the GAP-43 3'UTR with neuronal-specific RNA-binding proteins. These observations suggested to the inventors that overexpression of an RNA comprising the GAP-43 3'UTR in PC12 cells could compete for the binding of GAP-43 mRNA- binding proteins causing the destabilization of the endogenous GAP-43 mRNA and leaving it unprotected against ribonuclease attack, thereby depleting GAP-43 protein in cells. Several independent PC12 cell transfectants expressing the GAP-43 3'UTR were generated; levels of GAP-43 mRNA and protein were significantly diminished in these lines. These lines show that transient expression of the GAP-43 3'UTR in NGF-treated PC12 cells and neurons reduces GAP-43 protein and neuronal differentiation, evidence the specificity of effect of the 3'UTR on GAP-43 gene expression, and demonstrate that a 225-base sequence at the 3' end of the 3'UTR is responsible for these effects .

An increasing number of developmentally regulated mRNAs are found to be controlled by post-transcriptional mechanisms. These processes control RNA processing, nuclear-cytoplasmic transport, mRNA stability and translation efficiency and are mediated by specific sequences localized to the 3'UTR. Since 3'UTR sequences are often recognized by specific RNA-binding proteins, it is likely that RNA-protein interactions play an important role in the control of developmentally regulated genes. Herein, different experimental procedures aimed at the specific displacement of GAP-43 mRNA-binding proteins by competition with an excess of exogenous GAP-43 3'UTR are described. The results indicate that this approach is effective in blocking GAP-43 gene expression and the NGF-induced differentation of PC12 cells.

1- Field of Art

Messenger RNAs (mRNAs) are intermediary molecules active in the cell processes that convert genetic information encoded by DNA to proteins. mRNAs have three well-defined domains: the 5' untranslated region (5 'UTR), the coding region, and the 3' untranslated region (3'UTR). Nucleotide sequences in the coding region direct the amino acid sequence of the protein, while sequences in the 5' and 3 'UTRs variously mediate the ability of mRNAs to effect protein synthesis in accordance with the supplied genetic information. The untranslated regions of mRNA are thus a significant factor in the expression of eukaryotic genes .

Recently, it has been recognized that the 3 'UTRs of mRNAs contain regulatory nucleotide (nt) sequences which mediate temporal or spatial regulation of defined structures, properties and/or functions of mRNA including mRNA translation efficiency, mRNA localization in the cytoplasm, sorting and intracellular transport of mRNA, and mRNA stability against degradation (see, e.g., Decker et al. , Current Opinions in Cell Biol . 7:386-392, 1995, incorporated herein by reference). Alterations in these and other mRNA properties or functions are effected by binding interactions between these mRNA regulatory ("tag") sequences and members of a class of endogenous eukaryotic cell proteins having at least one binding domain which recognizes these sequences; this class includes a superfamily of RNA binding proteins containing RNA recognition motifs (RRMs) . This class of proteins is also referred to herein as "3'UTR mRNA binding proteins" (3 'UTRBP) . Binding complexes formed between mRNA 3' UTRs and

3'UTRBPs mediate both up-regulation and down-regulation of mRNA functions. While the complex regulatory mechanisms involved are only beginning to be understood, they tend to currently be defined by the art in terms of the 3'UTR regions affected by the binding events. That is, 3'UTR regulatory sequences have been conveniently characterized as "negative" or "positive" regulatory sequences which in their unbound (active) state inhibit (repress) or stimulate a given mRNA function, respectively, and which are neutralized or deactivated when protein-bound, thereby derepressing or destimulating their normal function (see, e.g., Current Opinion in Cell Biol., op. cit.) .

In attempts to more fully understand 3'UTR and 3 'UTRBP interactions, many binding studies directed to the identification and characterization of specific 3'UTR recognition sites and their BP ligands have been conducted. While a large number of 3'UTRBPs have been reported to date, identification of their complementary 3'UTR binding sites has generally been elusive. Exemplary reported 3'UTRBPs having 3'UTR recognition sites characterized by specific nt sequences and/or secondary structures include the iron-responsive element (IRE) in transferrin receptor mRNA, the stem loop structure in histone mRNAs, the AU-rich elements (AREs) in the mRNAs for several proto-oncogenes and cytokines and the U-rich regions (URRs) of fibroblast c-fos mRNAs (Mol . Brain Res. 3_6:240-250, 1996) . Additional 3'UTRBPs are described in Current Opinions in Cell Biol , op. cit . , and throughout the patent art. Neural elav- like proteins are of particular note. This family of RNA building proteins includes the neural-specific proteins HuC, HuD and Hel-Nl, and a ubiquitous protein, HuR, which specifically bind to U-rich elements in the 3'UTR of many mRNAs. The neural- specific elav-like proteins are necessary for neuronal differentiation and maintenance. As a specific example, HuD has been shown to be a significant regulator of the neuronal growth associated protein GAP-43 (B-50, 71, neuromedulin) , a neuron- specific phosphoprotein required for the regeneration and remodeling of neuronal connections. HuD selectively binds to GAP-43 mRNA 3'UTR, stabilizing the molecule and increasing message volume for GAP-43 gene expression (Chung et al, J. Biol. Chem. 272:6593-6598, 1997; Kohn et al, Mol . Brain Res. 36:240- 250, 1996; Neve et al, "cis-Acting regulatory elements in the GAP-43 mRNA 3 ' -untranslated region can function in trans to suppress endogenous GAP-43 gene expression", Mol . Brain Res . , 1998, in press; each of these publications is incorporated herein by reference) . It is reported that a majority of 3'UTRBPs are ubiquitous in (eukaryotic) tissues and are highly conserved in phylogeny, although tissue-specific BPs are also known. (Kennan et al, Trends Biochem. Sci . 16:214-220, 1991, cited in U.S. Patent 5,444,149 to Keene et al) .

2. Discussion of Related Art

The accumulated body of knowledge represented by the above brief discussion has formed the basis for various strategies to control gene expression by enhancing or inhibiting mRNA transcriptional and/or post-transcriptional activity.

Foremost among these strategies has been the application of antisense techniques employing oligonucleotides complementary to mRNA target region nt sequences to hybridize with and deactivate the targeted regions. Other gene control strategies available include the use of hybrid mRNAs comprising heterologous UTR nt sequences (U.S. Patent 5,807,707; European Patent 726,319; WO 96 11,265); the use of purified mRNA binding proteins to deactivate specific mRNA binding sites (U.S. Patent 5,773,246); transfection of cells to overexpress selected mRNAs (WO 98 27,220); and the use of mutated native UTR sequences (WO 95 29,244; WO 88 02,029) .

BRIEF DESCRIPTION OF THE DRAWING

Figure 1. Immunoblot analysis of control and GAP-43 3 ' UTR- expressing PC12 cell lines, using an antibody to GAP-43. Note that the parental tTA-transfeeted lines (tTA13 and tTAl8) and UHD vector-transfected subclones of these lines (13V3 and 18V7) express GAP-43 protein, whereas transfectants expressing the GAP- 43 3'UTR (13UTR22, 18UTR20, 18UTR4, and 18UTR8) do not display detectable levels of GAP-43. All lanes had equal amounts of total protein loaded as verified by Coomasie Blue staining (not shown) .

Figure 2. RNA blot analysis of control and GAP-43 3'UTR- expressing PC12 cells lines. Figure 2A. GAP-43 mRNA in a parental tTA-transfected line (tTA18) and a UHD vector-transfected subclone of this line relative to transfectants expressing the GAP-43 3'UTR (18UTR4, 18UTR20, 13UTR22). Ethidium bromide staining of the blot verifies that the samples comprised comparable amounts of RNA, as shown by the intensity of fluorescence of the 28S rRNA in each sample. The blot was exposed to film for 13.5 hours.

Figure 2B. Expression of GAP-43 and other mRNAs in a parental tTA-transfected line (tTAlδ), a UHD vector-transfected subclone of this line (18V4), and a transfectant expressing the GAP-43 3'UTR (18UTR4) . The 3 ' UTR-transfected cells display significantly lower levels of the endogenous GAP-43 mRNA than do vector- transfected controls. In contrast, levels of APP, fos, and tau mRNAs are not altered in the transfectants expressing the GAP-43 3'UTR. Exposure times: GAP-43, 2 days; APP, 5 days; fos, 13 days; tau, 3 days. Ethidium bromide staining of ribosomal RNA (not shown) confirmed the presence of equivalent amounts of RNA in the lanes . Figure 3. Effect of GAP-43 3'UTR expression via HSV vectors on GAP-43 protein and mRNA levels in primary neuronal control cultures and NGF-treated PC12 cells.

Figure 3A. PC12 cells were infected (m.o.i.=5) with the empty HSV-1 vector HSVPrpUC or with recombinant vector containing GAP- 43 3'UTR four hours prior to treating the cultures with NGF (100 ng/ml) . Cells were incubated for an additional 20 hours and the proteins were isolated and aliquots containing 20 μg of total protein analyzed by immunoblots . Control lanes show GAP-43 levels in untreated cultures.

Figure 3B. Embryonic day (E)21 rat cortical neurons were plated on poly-L-lysine-coated dishes as described in the Methods. Five days post-plating each culture was infected (m.o.i.=l) either with HSV-3'UTR, with the negative control HSVlac or mock- infected. Two days after infection, the cells were harvested and proteins processed for Western blot analysis using an antibody to GAP-43.

Figure 4. GAP-43 mRNA levels in PC12 cells infected with HSV-1 vectors containing the GAP-43 3'UTR. PC12 cells were infected with HSV-3'UTR virus as described in Figure 3A. Four hours post- infection, the media was changed and NGF was added to some of the cultures. After an additional incubation for 20 hours, cells from control, NGF-treated and HSV-3'UTR infected NGF-treated cultures were harvested and RNA was isolated and processed for Northern blot analysis. The identity of the bands was verified by separate hybridizations with specific probes for the GAP-43 5 'UTR and coding region (Panel A) or the GAP-43 3'UTR (Panel B) . Panel C shows the bands corresponding to the 28S and 18S ribosomal RNA stained with ethidium bromide. Blots hybridized with the 5 'UTR probes were exposed for 2 days while those hybridized with the 3 -UTR probes were exposed for lh. Arrows indicate the migration of the endogenous GAP-43 mRNA (endo) and the exogenous GAP-43 3'UTR (exo) . Figure 5. Effects of overexpression of the GAP-43 mRNA 3'UTR via HSV-3'UTR infection on GAP-43 expression and neurite outgrowth in NGF-treated PC12 cells. PC12 cells were infected (m.o.i.=5) with the empty HSV-1 vector (non-recombinant HSV-PrpUC) (negative control) or with HSV/GAP-43 3'UTR. NGF was added to the cells four hours after infection. Following 20 hours of exposure to NGF, cells were fixed and immunostained with a specific antibody to GAP-43. Micrographs show that overexpression of GAP-43 3'UTR sequences specifically suppresses GAP-43 immunoreactivity and process outgrowth (phase contrasts) in NGF-treated PC12 cells. Control panels show GAP-43 IF and cell morphology in untreated PC12 cells.

Figure 6. Quantification of GAP-43 mRNA and protein and extent of differentiation and neurite outgrowth in PC12 cells stably transfected with pMEP4/3'UTR, pMEP4/B, and pMEP4/C, expressing the GAP-43 3'UTR or the B or C fragment, respectively. Cells were treated with NGF in the presence or absence of cadmium (which induces expression of the transgene) for 48 hours before they were used for RNA blots, immunofluorescence, and morphometric analyses . Determination of percentage of inhibition of these four parameters by transgene expression is described in the experimental procedures section. Values are the mean + s.d.

Figure 7. GAP-43 immunoreactivity and neurite outgrowth in PC12 cells transfected with the pMEP4 vector containing the GAP-43 3'UTR. PC12 cells were transfected with pMEP4 constructs containing the entire GAP-43 3'UTR (A) or with the empty pMEP4 vector (B) as described in the Methods. Some cultures were treated with NGF in the presence of cadmium to induce the transient expression of the transgene (+NGF+Cd) , for 48 hours before cultures were fixed and cells analyzed by CAP-43 immunostaining. Control cultures were only treated with NGF

(+NGF) . Top panels show GAP-43 immunofluorescence and bottom panels the morphology of the cells as visualized by phase contrast .

Figure 8. The structure of GAP-43 mRNA, indicating the 3'UTR and poly (A) tail and displaying the sequence of GAP-43B with its 26 nt HuD binding site bolded (J. Biol. Chem. 272 , o . cit . ) .

SUMMARY OF THE DISCLOSURE

The invention provides, inter alia, a method for the manipulation of gene expression in eukaryotic cells or tissues comprising contacting cells or tissues containing a target gene with exogenous 3'UTR oligonucleotide sequences corresponding to 3'UTR nucleotide sequences of endogenous mRNA for this gene as described infra, and sequestering binding proteins in the cell or tissue environment that would otherwise be available to interact with the endogenous mRNA present. Gene manipulation both in vivo and in vitro is contemplated. The method applies competitive binding techniques otherwise known in the biotechnology art

(competitive immunoassays are exemplary) to sequester endogenous binding proteins away from their recognition sites on the endogenous 3'UTR mRNA and/or to preferentially bind with such proteins so that normal 3'UTRBP/mRNA binding interactions are inhibited. According to the inventions, the exogenous 3'UTR nucleotides thus serve as molecular decoys for cell proteins essential for the established flow of genetic information in the cell for protein synthesis, disrupting this flow in a selected manner to control expression of the corresponding genes both in vivo and in vitro.

Applications of the inventions include gene up- regulation and down-regulation, gene knockout and gene therapy, and provide gene constructs for overexpression of 3'UTR sequences in cells . Other applications include competitive binding assays for the identification and/or characterization of specific mRNAs and their BP ligands, useful in diagnostics and as research tools . The inventions are particularly useful for manipulating tissue type-specific or cell type-specific genes or gene subsets whose expression is directly or indirectly controlled by specific interactions between 3'UTRBPs and gene 3'UTR mRNAs. Mechanisms include direct gene control by 3'UTR feedback interaction with gene products, and indirect control by interaction with remote cell proteins. DETAILED DESCRIPTION OF THE INVENTION

The inventions herein are predicated on the concept of an approach to the modulation of gene expression which permits the selection of genetic regulators based on function. By augmenting 3'UTR function of selected mRNAs according to the inventions, followed by assaying for altered target gene product, painstaking binding studies directed to characterizing and identifying mechanisms by which genes are translated into proteins are obviated. Other functional approaches to gene expression have been used in the art (see, e.g., Rastinejad et al, Cell 72:903-917, 1993, which describes the evaluation of gene expression by DNA complementation) but not extensively, for reasons set forth in this paper. According to one aspect of the invention, exogenous sense poly- or oligonucleotides corresponding to all or a portion of the nucleotide sequence of the 3 ' untranslated region of an endogenous mRNA are introduced in excess into a cell or tissue composition comprising the endogenous mRNA and one or more binding proteins for its 3'UTR, to bind the proteins with the exogenous nucleotides in competition with the 3'UTR sequence of the endogenous mRNA. By depletion of free binding proteins in competition with the mRNA molecule, function and/or properties of endogenous mRNA dependent upon the interaction of the mRNA molecule with proteins now bound to the sense nucleotides are altered.

Sense nucleotide sequences useful in the practice of the invention comprise sequences corresponding to the complete 3'UTR of the selected mRNAs, or a portion thereof. To be useful, the sense sequences must contain at least one effective recognition sequence for at least one binding protein; as discussed supra , the recognition sequence may comprise secondary structural elements, multiple binding sites, or other factors necessary for a protein/3 'UTR binding event to occur. The sense nucleotide sequences may be selected to provide binding sites for specific classes of binding proteins such as tissue-specific or cell-specific proteins (e.g., brain-specific or neuronal-specific BPs) , to provide binding sites containing only negative regulatory sequences or only positive regulatory sequences, to provide binding sites for specific protein species, or as otherwise selected to provide alteration of desired mRNA function (s). Particularly useful nucleotide sequences are those which contain recognition sequences for binding proteins which, when bound, regulate the stability of the mRNA molecule. First, this is a function/property common to all (or perhaps nearly all) mRNAs, and mRNA down regulation or up regulation can be broadly achieved by decreasing or increasing the stability of the molecule against normal enzymatic degradation such as by alteration of mRNA to increase or decrease its decay rate or turnover. Second, the mRNA 3'UTR dominates the regulation of mRNA stability, and destabilization can typically be affected by 3'UTR sense polynucleotides alone, without benefit of regulatory regions in the remainder of the mRNA molecule..

The sense nucleotide sequences may be selected to contain recognition sequences including motifs known to recognize one or more binding protein species, such as the ARE, IRE, and RRM binding regions described supra . Of particular interest are sequences rich in a single RNA nucleotide such as the U-rich sequences described supra , especially sequences containing at least about 50% by number of the selected nucleotide; or those rich in two nucleotide repeating units, such as the AU-rich ARE sequences .

For regulation of selected gene expression, the sense nucleotides are selected to correspond to the 3 'UTRs of mRNAs for selected gene, or polynucleotide portions thereof, which essentially contain recognition sites such as those described above which alter normal mRNA functions necessary for protein translation and thus effect up-regulation or down-regulation of the gene in the context of the inventions. For clinical applications it will be generally appropriate to select 3'UTR sequences having good-to-high specificity for the gene of interest to avoid undesirable side-reactions; in other applications, 3'UTR sequences having less specific or nonspecific properties, or 3'UTR sequences of ubiquitous mRNAs, may be more useful.

The principles of competitive binding are well-known in the art, for example, for antigen/antibody immunoassays , and these principles are broadly applicable to the present inventions . By supplying an excess of polynucleotide competitor having binding properties comparable to those of the target mRNA molecule to be competed with, potential ligands (binding proteins) are competed out of the system, inhibiting formation of complexes between ligand and target. As known, the competitor may be selected for particular binding properties such as high binding affinity specificity, or avidity to obtain e.g., preferential binding or displacement of bound ligand from the target molecule.

In the present invention, the sense nucleotides are introduced into compositions containing intact mRNA, in an amount sufficient to significantly increase the 3'UTR nt concentration in the composition over that of the corresponding 3'UTR present in the mRNA molecules. Typically, increases in 3'UTR concentration over 3'UTR mRNA concentration of at least about 2 fold, usually from about 10 fold to 50 fold are used. The compositions may be preliminarily assayed for mRNA content by methods known in the art, using detectable anti-mRNA ligands such as anti-sense oligonucleotides, or RNA-binding assays such as described by Kohn et al. , Mol . Brain Res . , op. cit.

Compositions containing mRNA for use in the inventions include in vitro compositions such as cell and tissue cultures and extracts. Other mRNA compositions according to the invention comprise in vivo tissues and cells, particularly for clinical applications. As research tools, for example, the methods of the invention may employ mRNA manufactured compositions consisting essentially of one or more isolated mRNAs. For introduction of the 3'UTR sense polynucleotides into compositions of cells or tissues in vivo , the present invention exploits techniques well-known and used in the art of antisense hybridization. Briefly, for in vitro applications, the mRNA compositions of the inventions may be incubated with the sense nucleotides under conditions which permit contact and binding reactions between intracellular binding proteins and the competitor nucleotides; for in vivo applications such as gene therapy or experimental use in organisms, the sense nucleotides are conveniently administered parenterally, especially intravenously, in amounts at least about equal to or somewhat in excess of those typically used for antisense gene therapy or experimentation, as well-described in the literature and patent art. For in vivo use, the nucleotides may be administered in compositions also comprising conventional carriers or other active ingredients. Also, for jm vitro applications, synthetic RNAs can be generated by in vitro transcription assays as described in Kohn, et al . , 1996, supra . Mixtures of different nucleotides may be added to the compositions, if desired, as may be mixtures of binding proteins.

Nucleotides for use in the inventions comprise 3'UTR poly- or oligonucleotides from donor cell mRNA, synthetic nucleotides processed from nucleic acid synthesizers commonly used in the art, and nucleotides genetically engineered according to known techniques. "Complete" or "entire" naturally-occurring 3'UTR sequences will usually start after the coding sequence 3'UTR stop codon (i.e., just after the open reading frame as exemplified for GAP-43 3'UTR in Fig. 8) and end with the 3'UTR terminal nucleotide unit. The poly (A) tail may be superfluous in some applications. Naturally-occurring 3'UTR sequences should generally have an interspecies homology of at least 70%, more preferably, at least about 90% for interspecies use according to the invention; since the mRNA 3'UTR sequence is highly conserved as mentioned above, heterologous 3'UTR RNA sequences are broadly useful. Polynucleotide sequences within the scope of the invention comprise nucleotide sequences containing one or more binding protein recognition sites as described above subtended by the starting and ending 3'UTR bases. The sequences will generally contain at least about 7 nt units. The sequences may contain non-naturally occurring nucleotides or modified natural nucleotides, instead of, or in addition to, nucleotides naturally present in the mRNA 3'UTR, so long as function necessary for the purposes of the invention, especially binding properties, are not substantially adversely affected. See, e.g., U.S. Patent 5,703,054 for exemplary non-naturally occurring nucleotide bases potentially useful in the nucleotide sequences of the present invention. Non-naturally occurring bases or sequences thereof which confer stability on the RNA may be used to advantage; stabilizing thio derivatives are known and exemplary. "Gapped" nucleotide sequences of the type used in antisense techniques can also be used; such sequences comprise two or more isolated 3'UTR recognition sequences linked by benign nucleotide sequences.

Additionally, cells may be engineered as exemplified in the Examples infra, to overexpress 3'UTR or portions thereof in host cells containing cDNA expression constructs including DNA copies of the desired 3'UTR nucleotide sequences for use in the present inventions. Viral expression vectors were found to be particularly useful herein; in addition to HSV, adenoviruses , adeno-associated viruses, and herpes virus are exemplary. 3 'UTRs useful in the invention further comprise complete 3'UTR sequences or portions thereof derived or copied from any mRNA presently known or discovered in future, such as the GAP-43 B segment used in the Examples and described in Kohn, et al., 1996, and Chung, et al . , 1997 (ops.cit. ) . Isolated recognition sequences, such as the HuD recognition sequences shown in FIG. 8 may also be derived from mRNA, or synthesized.

The inventions are illustrated by the Examples which are directed to the regulation of GAP-43 gene expression. The procedures of the Examples are, however, broadly applicable to the regulation of gene expression according to the inventions. EXAMPLES

1. Materials and Methods

Construction of Recombinant Vectors and Generation of Transfected PC 12 Cells

A tetracycline-responsive expression system (Proc . Natl. Acad. Sci . USA 8£: 5547-5541, 1992) was used to generate PC12 cell lines expressing the GAP-43 3'UTR. PC12 cells were co- transfected with pUHDl5-l, encoding the tetracyline-controlled transactivator (tTA) , and pJ7hygro (a variant of pJ7Ω (Nucl . Acids Res . 1068, 1990) into which we inserted the gene encoding hygromycin resistance) . Transfectants were selected in the presence of 150 μg/ml hygromycin, and those transfectants expressing robust levels of the tTA RNA were identified using the reverse transcription-polymerase chain reaction (RT-PCR) (Receptor Molecule Biology. Methods in Neuroscience. 25 : 163-174, 1995, Sealfon, (Ed), Academic Press, San Diego, CA) . Two of these transfectants, tTA13 and tTAlδ, were subcloned by limiting dilution and expanded. Meanwhile, the rat GAP-43 3'UTR, was inserted into pUHDlO-3, containing a multiple cloning site and SV40 polyadenylation signal downstream of the hCMV minimal promoter with heptamerized tet-operators in two different orientations bp 871-1406, (EMBO J. 6:3641-3646, 1987) PNAS USA 89:5547-5541, 1992). This promoter is activated in the absence of tetracycline (Tc) , and inhibited in the presence of relatively high (1 μg/ml) levels of tetracycline. The pUHD/GAP3 'UTR plasmid was transfected (together with RSVneo (J. Neuro Sci. 14 : 1943- 1952, 1994) into both the tTAl3 and tTA18 PC12 cell lines; G418 colonies were selected in the presence of 1 μg/ml Tc; and those expressing detectable levels of the GAP3'UTR RNA in the absence of Tc were identified by RT-PCR (Receptor Molecule Biolocrv, op. cit. ) . Three of these subclones, one in tTA#13 (13UTR22) and three in tTA#18 (18UTR4, 18UTR8 and 18UTR20) were chosen for further analysis. In parallel transfections, tTA#13 and tTA#18 were transfected with the UHD10-3 vector alone to generate the vector-transfected control lines 13V3 and 18V7.

To make the herpes simplex virus (HSV) vector, the GAP- 43 3'UTR was subcloned into the amplicon vector pHSVPrpUC and termed HSV/3'UTR. This recombinant vector, together with the control vectors HSVPrpUC and HSVlac, were packaged into virus as described (Bio Techniques 20:460-469, 1996; Neurosci . 79:435-447, 1997) . The titer of the recombinant virus HSVlac, as measured by expression of the transgene in infected PC12 cells, was 3-4 x 10 infectious units (iu)/ml for all of the three preparations used in this study. HSV/3'UTR and HSVPrpUC virus could not be titered on the basis of expression; therefore, viral DNA was isolated from these vectors and HSVlac, and amplicon virus DNA copy number as evaluated by Southern blot analysis was found to be similar for all of the isolates (data not shown) . The titer of the defective virus component of each viral preparation as measured on 2-2 cells was consistently between 1 and 1.2 x 10 virus particles/ml.

Subclones of domains of the GAP-43 3'UTR in the vector pMEP4 were used to define the sequences in the 3'UTR that mediate the suppression of endogenous GAP-43 synthesis. Briefly, the entire GAP-43 3'UTR and the B and C fragments thereof were inserted into pMEP4 that had been cleaved with Xhol + Xbal . The resulting constructs were called pMEP4/3'UTR, pMEP4/B, and pMEP4/C. To prepare constructs in the pMEP vector (Invitrogen) the entire GAP-43 3'UTR (bp 871-1406) supra was amplified by PCR and cloned into the Nhel/Xhol sites of pMEP4. The resulting construct was called pMEP4-3'UTR.

Cell Cul ture

For the initial studies, PC12 cells were stably transfected with the GAP 3'UTR recombinants in the tetracycline- responsive vectors using lipofectamine (Life Technologies) . These transfectants were maintained as described in J. Neurochem. 6_0: 626-633, 1993). For subsequent experiments involving infection of PC12 cells with HSV recombinants and analysis of stable PC12 transfectants of the pMEP4 recombinants, PC12 cells were cultured in RPMI medium (ICN) supplemented with 7.5% heat- inactivated horse serum, 2.5% fetal bovine serum, 2 mM L- glutamine, and 30 μg/ml gentamycin. Cells were plated on poly-L- lysine (100 μg/ml) pre-coated dishes and maintained at 37° C in 5% C0₂. Cells were transfected with pMEP4-3'UTR or vector alone using electroporation, and stable transfectants were selected with hygromycin B (150 μg/ml) (J. Neurosci. 17:1950-1958, 1997).

Primary neuronal cultures were established from the cortices of embryonic day 21 (E21) rat embryos. Cortices were triturated briefly in MEM and were treated with trypsin and DNAse for 1 hr at 37° C, after which the tissue was dissociated mechanically in Neurobasal medium containing B27 supplements

(Life Technologies) , 10% fetal bovine serum, and 5% horse serum, using a fire-polished glass pipet. Cells were plated in 100-mm poly-L-lysine-coated dishes (1 x 10⁷ cells/dish) or on poly-L- lysine-coated coverslips punched out from ACLAR plastic film (Ted Pella, Inc.) in 24-well dishes (4.0 x 10⁴ cells/well). Three days post-plating, half of the medium in the 100-mm dishes was exchanged for serum-free Neurobasal medium containing B27 supplements. Five days post-plating, the medium was exchanged completely for the serum-free medium and infected with virus. Cells were harvested for immunoblot analysis 48 hours later. The cells in the 24-well dishes were infected either 4 hours or 16 hours post-plating and fixed for analysis 24 or 48 hours after infection.

Blot Analyses and Immuno cytochemistry

Transfected PC12 cells or neurons were lysed in 1% sodium dodecyl sulfate (SDS) . Protein determinations and immunoblot analyses were conventional (J. Neurochem 60 : 626-633 , 1993; PNAS USA 93:14182-14187, 1996; J. Neurosci. 17:1950-1958, 1997) . GAP-43 immunoreactivity was detected with monoclonal antibody m9lEl2 (Boehringer-Mannheim) or with a sheep polyclonal antibody to GAP-43 (J. Neurosci. 8:339-352, 1988) . Blotts were stained with Coomasie Brilliant Blue to control for the loading of protein in each lane.

RNA was isolated from cells by the guanidinium thiocyanate procedure (Mol. Brain Res. 1:271-280, 1986) . 20 μg of total RNA from each sample was subjected to electrophoresis on agarose/ formaldehyde gels, transferred to Biotrans membrane (ICN) , and hybridized with radiolabeled probe (Mol. Brain Res. 1, op. cit.) . The blots were exposed to Kodak X-Omat AR film for

1 to 13 days. The APP, GAP-43, and tau cDNA probes used are described in Neuron 1:669-677, 1988a; PNAS USA 85:3638-3642, 1988b; Mol. Brain Res. 1, op. cit.). For the immunocytochemical studies, cells were fixed in 4% paraformaldehyde (PFA) at room temperature for 15 minutes, pre-incubated for 30 minutes in blocking buffer (10% normal rabbit serum in phosphate buffered saline [PBS] ) , incubated for

24 hours at 4°C with a sheep anti-GAP-43 antibody (supra) , and then incubated with FITC-conjugated donkey anti-sheep IgG diluted

1:50. Photomicrographs were taken on a Zeiss Axiovert 35 microscope. Quantification of the immunofluorescence was performed using a BioQuant image analysis program (BioQuant/TRW, Nashville, TN) .

Morphological Analyses

Stable transfectants of pMEP4/3'UTR, pMEP4/B, pMEP4/C, and pMEP4 vector were plated in 4 well chamber Permanox slides precoated with poly-L-lysine . Cells were treated with cadmium (1.5 μM, to induce transgene expression) in the presence of NGF (100 ng/ml) , or with NGF alone, for 48 hours. The cells were then fixed in 4% PFA for 15 minutes at room temperature and observed under a light microscope. Cells were defined as differentiated if they possessed polygonal or bipolar morphology and had neurites, or as nondifferentiated if they were round and devoid of neurites. Neurites were defined as being long if they were longer than the diameter of the cell body. All of the cells in at least three randomly-chosen high-power fields were evaluated. A minimum of 100 cells was analyzed for each experiment, and at least two separate experiments were performed.

The analyses were performed blind by two different individuals.

The percentage of inhibition for each parameter in the transfectants was expressed relative to the levels observed in cells transfected with pMEP4 vector alone using the following equation: level transfectant (NGF+Cd) / level transfectant (NGF) level pMEP4 transfectant (NGF+Cd) /level pMEP4 transfectant (NGF)

The value obtained was subtracted from 100 to obtain percentage of inhibition .

2 . Results

Stable PC12 cell transfectants expressing the GAP-43 3 ' UTR show decreased levels of GAP-43 protein and mRNA

A tetracycline-responsive expression system (supra) was used to generate PC12 cell lines expressing the GAP-43 3'UTR. Immunoblot (Western blot) analysis (Fig. 1) of the positively- expressing transfectants and of tTA and vector-transfected tTA controls (all grown in the absence of Tc) were carried out using an antibody to GAP-43. GAP-43 protein was immunodetected in lines tTA13 and tTAlδ and vector-transfected controls (13V3, 18V7), but not in lines expressing the GAP-43 3'UTR (UTR22, UTR20, and UTR4) .

To determine whether levels of endogenous GAP-43 mRNA were altered in these lines, as would be predicted if the exogenous GAP-43 3'UTR were competing with the endogenous GAP-43 mRNA for the binding of RNA-stabilizing proteins, RNA Northern blot analysis was carried out on selected cell lines. As shown in Fig. 2, the endogenous 1.6-kb GAP-43 mRNA was far less prominent in the UTR-transfected lines than in the tTA and vector-transfected tTA lines. This demonstrates that observed decreases in GAP-43 expression in 3 'UTR-transfected cells in mediated by a reduction in the steady-state levels of its mRNA.

The levels of other mRNAs that share 3 ' UTR sequences and binding proteins wi th the GAP-43 mRNA are not al tered in GAP-43 3 ' UTR- transfected PC12 cells

mRNAs in addition to the GAP-43 mRNA share similar 3'UTR sequences. For example, the amyloid precursor protein (APP) mRNA possesses U-rich elements resembling those in the GAP- 43 3'UTR, as do the mRNAs for the proto-oncogene c-fos and microtubule protein tau. Multiple proteins interact with each of these 3'UTR sequences, and at least one of these proteins also binds to the GAP-43 3'UTR in vi tro . Examination of the levels of APP mRNA in the GAP-43 3'UTR transfectants showed no alteration of APP, c-fos, or tau mRNA abundance in these lines relative to controls in contrast to the observed down-regulation of GAP-43 mRNA (Fig. 2B) . One of the proteins that binds to the GAP-43 3'UTR is

HuD, a 40 kDa protein that belongs to the elav family of neuronal-specific RNA binding proteins. This protein also binds to the 3'UTR of the mRNAs encoding fos and the neuronal microtubule associated protein tau. The levels of fos and tau mRNAs in the transfected lines were examined, and it was found that expression of these mRNAs was not altered in the lines expressing the GAP-43 3'UTR (Fig. 2). Thus, the presumptive destabilization of endogenous GAP-43 mRNA by the exogenously- expressed GAP-43 3'UTR does not affect basal levels of APP, fos, and tau mRNAs. The effect of exogenously-expressed GAP-43 3'UTR thus appears to be selective for its corresponding mRNA and does not affect the levels of other mRNAs that contain similar structural motifs in their 3 'UTRs.

Transient expression of the GAP-43 3 ' UTR in NGF- treated PC12 cells and primary cortical neurons in vi tro causes decreased levels of GAP-43 protein and inhibi ts process outgrowth The inhibition of GAP-43 synthesis in the stably- transfected PC12 lines expressing the GAP-43 3'UTR is an indirect effect of expressing the endogenous 3'UTR chronically in these cells. Expression of the tetracycline-responsive system was not tetracycline-regulatable in our PC12 cells, a result that has been reported for other cell lines as well. An alternative acute intervention was therefore tested, in which PC12 cells and cortical neurons were infected with a recombinant HSV vector expressing the GAP-43 3'UTR or with a nonrecombinant HSV vector. PC12 cells were infected, at a multiplicity of infection (m.o.i.) of five, with HSV-1 vectors containing pUC vector sequences (pHSVPrpUC; vector control) or expressing the GAP-43 3'UTR. Four hours later, NGF was added to some of the cultures. Cells were then incubated for an additional 20 hours, following which they were processed for protein or RNA by Western blot analysis. As shown in Fig. 3, the HSV recombinant expressing GAP-43 3'UTR was capable not only of down-regulating GAP-43 protein and mRNA levels in nondifferentiated PC12 cells, but also of preventing their induction upon treatment with NGF (Fig. 3A) . These effects were not due to viral infection, as HSV alone (PrpUC) did not affect GAP-43 levels in either control or NGF-treated cells.

The effects of the GAP-43 mRNA 3'UTR on neurite outgrowth in NGF-treated PC12 cells (Fig. 4) was examined, PC12 cells were infected (m.o.i. =5) with control and recombinant HSV-1 vectors, and NGF was added four hours later. Following 20 hours of exposure to NGF, cells were fixed and analyzed for GAP-43 immunoreactivity. The results shown in Fig. 4 reveal that overexpression of GAP-43 3'UTR mRNA in NGF-treated PC12 cells specifically blocked both GAP-43 immunoreactivity and neurite outgrowth.

Comparable results were obtained upon infection of primary rat cortical neuron cultures with HSV/GAP3'UTR (Fig. 3B) . Embryonic day (E)21 rat cortical neurons were plated on poly-L- lysine-coated 10-cm dishes at 10 x 10⁶ viable cells/plate, and five days post-plating each plate was infected (m.o.i.=l) with the appropriate virus. 48 hours after infection, the cells were harvested and immunoblotted (Western blot) with an antibody to GAP-43, GAP-43 protein was greatly reduced in neurons infected with HSV/GAP-43 3'UTR, compared to mock-infected neurons or neurons infected with the negative control HSVlac. Although it is known that the virion host shutoff ( vhs) component of the HSV virion, encoded by UL41, facilitates the nonspecific degradation of host mRNAs, the inclusion of HSVPrpUC and HSVlac control and mock infections in studies indicate that GAP-43 mRNA and protein levels do not appear to be altered by infection with HSV recombinant viruses alone.

As an additional control, HSV vectors were used to overexpress the APP 3'UTR in PC12 cells. No effects on neurite outgrowth or on GAP-43 expression were detected as a result of this intervention.

To further investigate the mechanism by which exogenous GAP-43 3'UTR sequences interfere with GAP-43 gene expression, we measured the levels of endogenous and exogenous GAP-43 mRNA sequences in HSV-3'UTR infected PC12 cells by Northern blots

(Fig. 4) . The identify of the bands was verified by separate hybridizations of the same blots with specific probes for either the GAP-43 5 'UTR and coding region, which detects only the endogenous mRNA (Fig. 4A) , or the GAP-43 3'UTR, which detects both exogenous and endogenous 3'UTR sequences (Fig. 4B) . As shown in Fig. 4A, NGF-treated PC12 cells infected with HSV-3'UTR (NGF+3'UTR) contained lower levels of the endogenous GAP-43 mRNA than NGF-treated uninfected cells. Ethidium ro ide staining of the blot (Fig. 4C) indicated that these changes were not due to a reduction in total RNA levels or to non-specific RNA degradation. Analysis of the same blot with the 3'UTR probe demonstrated the presence of a band corresponding to the size of the exogenous GAP-43 3'UTR in HSV-3'UTR infected cells (Fig. 4B) . Because this band was detected at much shorter exposure times than those with the 5 'UTR probe (Fig. 4A) , the band corresponding to the endogenous GAP-43 mRNA could not visualize at this short exposure. Comparison of exposure times and specific activities of the probes for the GAP-43 5' and 3 'UTRs revealed that the viral infection resulted in a 20-50 fold molar excess of exogenous 3'UTR sequences.

Overexpression of GAP-43 3 ' UTR reduces NGF-induced neuri te outgrowth in PC12 cells

The effects of the exogenous GAP-43 3'UTR or GAP-43 expression and neurite outgrowth in NGF-treated PC12 cells were also examined (Fig. 5) . Four hours prior to the addition of NGF, PC12 cells were infected (m.o.i. =5) with control and recombinant HSV-1 vectors. Following 20 hours of exposure to NGF, cells were fixed and analyzed for GAP-43 immunoreactivity. As shown in Fig. 5, HSV-3'UTR infection specifically suppressed GAP-43 levels and neurite outgrowth in NGF-stimulated PC12 cells. In agreement with the results of the Western blot analysis (Fig. 3), we found that nonrecombinant HSV (HSV-PrpUC) had no effect on either GAP-43 immunoreactivity or the NGF-induced neuronal differentiation. Similar results were obtained using the mammalian expression system pMEP4 to express the GAP-43 3'UTR in PC12 cells (Fig. 7) . Treatment of pMEP4 transfectants with cadmium activates the metallothionein promoter in the vector leading to a 5-10 fold induction of transfected sequences (J. Neurosci. 17:1950-1958, 1997). In the absence of cadmium, pMEP4- 3 'UTR transfectants treated with NGF exhibited normal process outgrowth and GAP=43 expression (+NGF panels in Fig. 7A) . Yet, transient expression of the GAP-43 3'UTR by treatment with cadmium completely blocked the outgrowth of neurites in response to NGF (+NGF+Cd panels) . These results are not due to cadmium exposure, as this treatment had no effect in cells transfected with empty pMEP4 vector (Fig. 7B) . Measurement of GAP-43 immunofluorescence (IF) *Fig. 5) with the BioQuant image analysis program indicated that transient expression of the GAP-43 3'UTR in pMEP4-3'UTR transfected PC12 cells decreased GAP-43 IF to 40% of the levels found in cells transfected with the vector alone.

A 225-base sequence at the 3 ' end of the GAP-43 3 ' UTR is primarily responsible for the reduction in GAP-43 levels and inhibi tion of process outgrowth in the presence of NGF

To map the sequences in the GAP-43 3'UTR that act in trans to suppress endogenous GAP-43 gene expression, specific subdomains of the GAP-43 3'UTR in PC12 cells were expressed. A 225-base sequence at the 3' end of the GAP-43 3'UTR (the "B" fragment) mediates the increased stability of the GAP-43 mRNA in PC12 cells treated with the phorbol ester TPA. The ability of this fragment, and of the adjacent upstream 114-base "C" fragment to suppress endogenous GAP-43 synthesis when expressed in PC12 cells was tested. The B and C fragments were cloned into the pMEP4 vector under control of the cadmium-inducible metallothionein promoter. Cells stably transfected with the B or C fragments of the GAP-43 3'UTR in pMEP4 , or with the pMEP4 vector alone, were treated for 48 hours with NGF alone or in the presence of 1.5 uM cadmium. The levels of GAP-43 mRNA and protein, and the extent of differentiation and neurite outgrowth of the cells, were then quantified (Fig. 6). Cadmium-treated cells transfected with vectors expressing the entire GAP-43 3'UTR or the B fragment demonstrated levels of NGF-induced GAP-43 mRNA and protein that were about 35% less than levels in vector- transfected cells. In contrast, C fragment transfectants showed only about 20% inhibition of NGF-induced GAP-43 mRNA and protein. Extent of inhibition of differentiation and neurite outgrowth was correspondingly greater in the cadmium-treated 3'UTR and B fragment transfectants than in the cadmium-treated C fragment transfectants. Thus, inhibition of GAP-43 gene expression correlated with inhibition of NGF-induced differentiation and neurite outgrowth in PC12 cells, and was predominantly mediated by the B fragment. 3. Discussion

The 3 'UTRs of mRNAs have been demonstrated to regulate multiple cellular processes, including RNA processing and export from the nucleus, mRNA translation efficiency, and intracellular localization and stability. That expression of the GAP-43 gene is controlled partly by selective changes in the stability of its mRNA, and that this process is mediated by the interaction of specific sequences in the GAP-43 mRNA 3'UTR with neuronal RNA- binding proteins has been previously shown. Given the recent discovery that the 3 'UTRs of specific muscle structural proteins can suppress cell growth and overcome deficiencies in differentiation of myoblast cells, whether expression of the GAP- 43 mRNA 3'UTR would have an effect on differentiation of PC12 cells was tested. An excess of exogenous GAP-43 3'UTR did indeed interfere with the expression of the endogenous mRNA and thus modulate the state of differentiation of PC12 cells treated with NGF. Furthermore, analysis of subdomains of the GAP-43 mRNA 3'UTR revealed that the B-element, a 225-base sequence at the 3' end of the 3'UTR, can mimic the effects of the entire 3'UTR on GAP-43 expression and PC12 differentiation.

In NGF-treated PC12 cells and in developing neurons, the B-element mediates the stabilization of GAP-43 mRNA by association with HuD, a neuronal-specific RNA-binding protein from the elav family which binds to a 26-nucleotide U-rich motif within a regulatory element in the GAP-43 3'UTR (J. Biol. Chem. , 272 , op. cit). It is likely that overexpression of the entire 3'UTR or the B-element titrates this protein, leaving the endogenous GAP-43 mRNA unprotected against the action of ribonucleases . Supporting this idea, it has been found that downregulation of HuD expression in PC12 cells by antisense techniques (J. Cell Biol. 121:417-429, 1993; J. Cell. Biol. 128:647-660, 1995) causes a dramatic decrease in GAP-43 mRNA and protein levels and blocks the NGF-induced neuronal differentiation of these cells. This has suggested to the inventors herein that overexpression of 3'UTR sequences are an alternative tool to antisense applications for controlling gene expression in neuronal and other cells.

To determine the specificity of the reduction of GAP-43 mRNA and protein by exogenous expression of the GAP-43 3'UTR, the expression of other mRNAs that share similar 3'UTR sequences and/or bind to similar RNA-binding proteins was examined as described above. Besides binding to the GAP-43 3'UTR, the neuronal-specific RNA binding protein HuD is known to interact with U-rich elements in the 3 'UTRs of the mRNAs for APP, c-fos and the microtubule associated protein tau. Examination of the levels of APP, fos, and tau mRNAs in the GAP-43 3 'UTR-transfected lines, however, revealed that the abundance of these mRNAs was not altered in the lines expressing the GAP-43 3'UTR. Thus, the effect of excess GAP-43 3'UTR sequences appears to be selective for this mRNA and does not extend to at least three other mRNAs that share 3'UTR sequences and binding proteins such as HuD with the GAP-43 mRNA 3'UTR. Besides HuD, other neuronal-specific RNA- binding proteins interact with sequences in the Gap-43 mRNA. It is therefore possible that, in addition to HuD, other transacting factors mediate the effect of exogenous GAP-43 3'UTR sequences on GAP-43 gene expression and cell differentiation. Of these, the 95 kDa GAP-43 mRNA-binding protein described by Kohn et al (Mol. Brain Res. 36, op. cit.) is a very likely candidate for it also interacts with U-rich sequences within the stability-conferring element in the GAP-43 3'UTR. Furthermore, we have shown that this RNA-binding protein binds to the GAP-43 3'UTR in its phosphorylated state and that this interaction is enhanced by phorbol ester treatment. Since the binding properties of this protein correlate well with the stabilization of the GAP=43 mRNA in the presence of phorbol esters or NGF, this factor is probably also involved in the control of GAP-43 mRNA stability.

The function of the neuronal growth associated protein GAP-43, a phosphoprotein of the presynaptic membrane that has been implicated in developmental and synaptic plasticity, has been assessed both in null-mutant mice and also in neuronal cultures with antisense techniques. In vivo , GAP-43 appears to mediate decision-making aspects of neuronal pathfinding. In "knock-out" mice lacking GAP-43, retinal axons remain trapped in the chasm, .evidently unable to find their way past this midline decision point. It has been suggested that GAP-43 acts to amplify pathfinding signals from the growth cone.

In complementary work using antisense oligonucleotides, primary sensory neurons of GAP-43 were incapable of extending neurites. It is shown herein that PC12 cells acutely depleted of GAP-43 protein by exogenous expression of the GAP-43 mRNA 3'UTR are deficient in neurite outgrowth in response to NGF, although increased sensitivity to NGF and accelerated neurite outgrowth. These findings provide further evidence that sequestration of appropriate binding proteins by sense 3'UTR sequences according to the invention have effects on gene function similar to those obtained by antisense technologies.

Post-transcriptional mechanisms are found to control an increasing number of developmentally-regulated mRNAs; many of which are localized to specific subcellular compartments. Thus, it is not unreasonable to suggest that 3'UTR sequences are involved in local regulation of mRNA stability and translation efficiency. Furthermore, given that these sequences are often recognized by specific RNA-binding proteins, it is likely that RNA-protein interactions play an important role in the control of their developmental expression. An additional implication of the results described here is that a new tool for probing the function of a protein has been discovered. Inhibition of expression of specific neuronal mRNAs by overexpression of their

3'UTR sequences may be a useful genetic intervention for deciphering the roles of specific molecules in neural development .

As previously noted, expression of the GAP-43 gene is controlled partly by selective changes in the stability of its mRNA, a process mediated by the interaction of specific sequences in the 3' untranslated region (3'UTR) with neuronal- specific RNA-binding proteins. Limiting amounts of these transacting factors for GAP-43 mRNA stability are available in the cell, and the expression of an excess of exogenous 3'UTR sequences affects the levels of the endogenous mRNA via competitive binding to specific RNA-binding proteins. Chronic expression of exogenous GAP-43 3'UTR sequences in PC12 cells and in primary neurons in culture causes depletion of the endogenous mRNA and consequent reduction of GAP-43 protein levels. The levels of other neuronal mRNAs that exhibit similar 3'UTR sequences and share the same RNA-binding proteins with the GAP-43 mRNA (c-fos. amyloid precursor protein, and the microtubule associated protein tau) were unchanged; thus the effect of excess GAP=43 3'UTR appears to be specific for its corresponding mRNA.

Transient expression of the GAP-43 3'UTR in nerve growth factor

(NGF) -treated PC12 cells and in primary neurons using herpes simplex virus (HSV) vectors and mammalian expression vectors with inducible promoter to acutely express a 10 to 50 fold excess of 3'UTR sequences, caused not only a decrease in GAP-43 levels but also an inhibition of neurite outgrowth. Stable expression of subdomains of the GAP-43 3'UTR in PC12 cells under control of an inducible promoter revealed that a 225-base sequence at the 3 ' end of the UTR was responsible for the reduction in GAP-43 levels and inhibition of neuritogenesis in the presence of NGF. Suppression of specific neuronal mRNAs by overexpression of 3'UTR sequences is a potential tool for probing the function of a protein by inhibition of its synthesis.

The term "polynucleotide" as used in the following claims and otherwise herein includes oligonucleotides unless otherwise specified.

Claims

WHAT IS CLAIMED IS;

1. A method for altering the function or properties of messenger RNA (mRNA) in a composition also containing regulatory binding proteins for the 3' untranslated region (3'UTR) of the mRNA comprising contacting the composition with one or more sense polynucleotide species corresponding to the mRNA 3'UTR or a portion thereof containing at least one recognition site for one or more 3'UTR binding proteins, to competitively bind one or more binding protein species and inhibit the formation of mRNA-regulatory mRNA/binding protein complexes.

2. The method of Claim 1, wherein the composition comprises a cell or tissue culture.

3. The method of Claim 1, wherein the composition comprises a cell or tissue extract.

4. The method of Claim 1, wherein the composition comprises cells in vivo .

5. The method of Claim 1, wherein the composition comprises tissue in vivo .

6. A method for altering gene expression in cells or tissues comprising contacting the cells or tissues with one or more exogenous sense polyonucleotide species corresponding to the 3'UTR of mRNA for a target gene, or a portion thereof, and containing at least one recognition site for one or more 3'UTR binding protein, to competitively bind one or more binding protein species and inhibit the formation of gene-regulatory mRNA/binding protein complexes .

7. The method of Claim 6, wherein the composition comprises a cell or tissue culture.

8. The method of Claim 6, wherein the composition comprises a cell or tissue extract.

9. The method of Claim 6, wherein the composition comprises cells in vivo .

10. The method of Claim 6, wherein the composition comprises tissue in vivo .

11. A gene therapy method for regulating gene expression in a mammal comprising administering to the mammal one or more exogenous sense polynucleotide species corresponding to the 3'UTR of mRNA for at least one target gene, or a portion thereof, and containing at least one recognition site for one or more 3'UTR gene-regulatory binding proteins, to form binding complexes between the polynucleotides and at least one binding protein species which alter function or properties of mRNA required for gene expression.

12. A method for altering gene expression in intact cells comprising engineering the cells to over-express the 3'UTR of one or more selected mRNAs or a portion thereof containing at least one recognition site for one or more 3'UTR gene-regulatory mRNA binding protein, to competitively bind one or more gene- regulatory binding protein species and inhibit the formation of gene-regulatory mRNA/binding protein complexes .

13. The method of Claim 12, wherein the engineered cells are in vivo cells.

14. The method of Claim 12, wherein the engineered cells are in vitro cells.

15. The method of Claim 12, wherein at least one selected mRNA is endogenous .

16. The method of Claim 12, wherein at least one selected mRNA is exogenous .

17. A competitive binding assay for 3'UTR mRNA/binding protein interactions, comprising contacting a composition containing mRNA and one or more putative 3'UTR mRNA binding proteins with one or more sense polyoligonucleotide species corresponding to a 3'UTR of an mRNA or a portion thereof and containing at least one recognition site for one or more 3'UTR mRNA binding protein species, and detecting mRNA/binding protein complexes .

18. The method of Claim 17, wherein the composition comprises a cell or tissue culture.

19. The method of Claim 17, wherein the composition comprises a cell or tissue extract.

20. The method of Claim 17, wherein the oligonucleotide carries a detectable label.

21. A competitive binding assay for 3'UTR mRNA/binding protein complexes comprising contacting a composition containing one or more polynucleotide species putatively containing at least one recognition site for one or more mRNA 3'UTR binding protein species with one or more binding protein species, and detecting any mRNA/binding protein complexes.

22. The method of Claim 21, wherein the composition comprises a cell or tissue culture.

23. The method of Claim 21, wherein the composition comprises a cell or tissue extract.

24. The method of Claim 21, wherein the protein species carries a detectable label.

25. The method of Claim 21, wherein the polynucleotide species corresponds to the 3'UTR of one or more selected mRNAs or a portion thereof.

26. The method of Claim 1, wherein the mRNA-regulatory complexes regulate mRNA stability.

27. The method of Claim 1, wherein the 3'UTR recognition site comprises an ARE, IRE, or URR, or a combination thereof .

28. The method of Claim 1, wherein at least one recognition site comprises a nucleotide sequence containing at least 50% by number of one of A, T, C or U.

29. The method of Claim 1, wherein the mRNA 3'UTR or portion thereof is cell or tissue specific.

30. The method of Claim 1, wherein the mRNA 3'UTR or portion thereof is specific to neural cells or brain tissues.

31. The method of Claim 1, wherein the mRNA 3'UTR or portion thereof is specific to GAP-43.