Apoptosis is regulated cell death, whereby cells may be programmed to die at certain times during development and growth. This process is particularly important in development and maintenance of neural cells; a proportion of neural cells are programmed to die during development of the nervous system, and conversely, neural cells require the presence of growth factors, such as NGF (nerve growth factor) throughout their life in order to prevent cell death.

Situations whereby cell death is prematurely or incorrectly induced or accelerated, whether through necrotic or apoptotic mechanisms, occur in many disorders affecting many tissues; for example, in the nervous system cell death occurs in stroke, Alzheimer's and Parkinson's diseases, Huntington's disease and many other neurodegenerative diseases. On the other hand, situations where the normal occurrence of cell death is prevented or reduced can lead to cell proliferation, resulting in diseases such as cancer of various tissues.

Development of mechanisms for modulating cell death, including programmed cell death is therefore of great utility in the investigation and treatment of disorders resulting from abnormal cell death or cell proliferation.

One approach is to look for new genes/proteins which may be involved in modulating cell death.

One method of identifying new gene sequences employs a technique known as subtractive hybridisation.

Subtractive hybridisation provides a method of enriching particular nucleotide sequences prior to cDNA cloning.

The applicants initially used a subtractive hybridisation procedure to isolate cDNAs corresponding to mRNAs expressed at higher levels in newborn rather than adult rat cerebellum. From this information, a rat and corresponding human gene have now been cloned and sequenced.

The rat gene is known to be transcribed in a number of different forms of varying length. Several of these forms are given in the accompanying Figures.

Thus, in a first aspect the present invention provides an isolated polynucleotide fragment comprising the nucleotide sequence depicted in Figure 7 or Figure 15.

More particularly there is provided an isolated polynucleotide fragment comprising a mammalian gene transcript substantially comprising the nucleotide sequence from base 193 to base 2190 in Figure 7 (a) or from base 295 to base 1509 in Figure 7 (b), or from base 120 to base 1403 in Figure 7 (c), or from base 120 to base 455 in Figure 7 (d), or from base 658 to base 990 in Figure 7 (e), or from base 9 to base 2003 in Figure 15 (a), or from base 9 to base 2135 in Figure 15 (b), or sequences complementary thereto.

Hereinafter the mammalian gene will be referred to as nng2.

Figure 7 in fact shows the nucleotide sequences of a number of alternative transcripts of the rat nng2 gene and Figure 15 shows the nucleotide sequence of alternative transcripts of the human nng2 gene. The rat nng2 gene is also known as ruk.

"Polynucleotide fragment"as used herein refers to a polymeric form of nucleotides of any length, both to ribonucleic acid sequences and to deoxyribonucleic acid sequences. In principle, this term refers to the primary structure of the molecule. Thus, this term includes double stranded and single stranded DNA, as well as double and single stranded RNA, and modifications thereof.

In a further aspect the present invention provides a polynucleotide fragment encoding a mammalian NNG2 polypeptide. For example, the polypeptides representing the NNG2 protein variants encoded by the rat nng2 gene variants shown in Figure 7 are shown in Figures 7 and 8.

Also, the polypeptide representing the NNG2 protein encoded by the human nng2 gene is shown in Figure 15.

In general, the term"polypeptide"refers to a molecular chain of amino acids with a biological activity.

It does not refer to a specific length of the product, and if required can be modified in vivo and/or in vitro, for example by glycosylation, myristoylation, amidation, carboxylation or phosphorylation; thus inter alia peptides, oligopeptides and proteins are included. The polypeptides disclosed herein may be obtained by synthetic or recombinant techniques known in the art.

Thus the term extends to cover for example polypeptides obtainable from various transcripts and splice variants of these transcripts from a particular gene. For instance, the polypeptide sequences from three different transcripts of the rat nng2 gene are shown in Figure 8.

Different transcripts are identified according either to their length, or to their tissue-specificity. The rat variants may be designated rukl, rukm and ruk, for the long, medium and short forms of the transcript, with rukh denoting a heart-specific variant. The human transcripts may also be designated nng2l, nng2m, nng2, and nng2h, again denoting the long, medium, short and heart-specific variants. There

also exist further subvariants of the ruk. transcript, denoted as ruk,,, ruks and ruk. Additionally, functional domains may be observed in a particular protein and isolated polypeptides relating to these functional domains may be of particular use. For example three SH3 domains, known to be involved in protein-protein interactions, have been observed in the rat and human NNG2 polypeptides disclosed in Figures 7,8 and 15. The present invention also relates to polynucleotide fragments comprising a nucleotide sequence encoding such functional domain polypeptides.

It will be understood that for the nng2 nucleotide and polypeptide sequences presented herein, natural variations can exist between individuals. These variations may be demonstrated by amino acid differences in the overall sequence or by deletions, substitutions, insertions or inversions of amino acids in said sequence.

As is well known in the art, the degeneracy of the genetic code permits substitution of bases in a codon resulting in another codon for the same amino acid.

Consequently, it is clear that any such derivative nucleotide sequence based on the sequences disclosed herein are also included in the scope of the present invention.

Thus, the present invention further includes nucleotide and polypeptide sequences having at least 80%, particularly at least 90%, and especially at least 95% homology or similarity with the sequences of Figures 7,8 and 15. Similarity a homology may be determined using for

example the GCG Wisconsin sequence analysis set of programs, such as the"gap"program.

Thus, the present invention also includes nucleotide sequences similar to the polynucleotide sequences disclosed herein. It is understood that similar sequences include sequences which remain hybridised to the polynucleotide sequences of the present invention under stringent conditions. Typically a test similar sequence and a polynucleotide sequence of the present invention are allowed to hybridise for a specified period of time generally at a temperature of between 50 and 70°C in double strength SSC (2 x NaCl 17.5g/l and sodium citrate (SC) at 8.8 g/1) buffered saline containing 0.1% sodium dodecyl sulphate (SDS) followed by rinsing of the support at the same temperature but with a buffer having a reduced SSC concentration. Depending upon the degree of similarity of the sequences, such reduced concentration buffers are typically single strength SSC containing 0.1% SDS, half strength SSC containing 0.1% SDS and one tenth strength SSC containing 0.1% SDS. Sequences having the greatest degree of similarity are those the hybridisation of which is least affected by washing in buffers of reduced concentration.

It is most preferred that the similar and inventive sequences are so similar that the hybridisation between them is substantially unaffected by washing or incubation in one tenth strength SSC containing 0.1% SDS.

Furthermore, fragments derived from the NNG2 polypeptides or from the nucleotide sequences depicted in Figures 7,8 and 15 which still display NNG2 specific properties are also included in the present invention.

"NNG2 specific properties"is understood to relate to biological functions which are attributable to the naturally-occurring NNG2 protein.

All such modifications mentioned above resulting in such derivatives of the mammalian NNG2 polypeptide are covered by the present invention so long as the characteristic NNG2 polypeptide properties remain substantially unaffected in essence.

The information presented herein can be used to genetically manipulate the gene or derivatives thereof, for example to clone the gene by recombinant DNA techniques generally known in the art. Cloning of homologous genes from other species of mammal may be performed with this information by widely known techniques; for example, oligonucleotides may be designed to a consensus region and/or functional domains of the sequences shown in Figures 7 and 15 and such oligonucleotides, and/or the polymerase chain reaction products generated using these oligonucleotide primers, can be used as probes for cloning homologous genes from other organisms for example by polymerase chain reaction or by hybridisation.

Moreover mammalian nng2 nucleotide sequences of the present invention are preferably linked to expression control sequences. Such control sequences may comprise

promoters, operators, inducers, ribosome binding sites etc.

Suitable control sequences for a given host may be selected by those of ordinary skill in the art.

A nucleotide sequence according to the present invention can be ligated to various expression-controlling DNA sequences, resulting in a so-called recombinant nucleic acid molecule. Thus the present invention also includes an expression vector comprising an expressible nng2 nucleotide sequence. Said recombinant nucleic acid molecule can then be used for transformation of a suitable host.

Such recombinant nucleic acid molecules are preferably derived from for example, plasmids, or from nucleic acid sequences present in bacteriophages or viruses and are termed vector molecules.

Specific vectors which can be used to clone nucleotide sequences according to the invention are known in the art (eg Rodriguez RL and DT Denhardt, editors, Vectors: A survey of molecular cloning vectors and their uses, Butterworths, 1988).

The methods to be used for the construction of a recombinant nucleic acid molecule according to the invention are known to those of ordinary skill in the art and are inter alia set forth in Sambrook çt « l, Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory, 1989.

The present invention also relates to a transformed cell comprising the nucleic acid molecule in expressible form."Transformation", as used herein, refers to the

introduction of a heterologous nucleic acid sequence into a host cell in vivo, ex vivo or in vitro irrespective of the method used, for example, by calcium phosphate co- precipitation, direct uptake or transduction.

The heterologous nucleic acid sequence may be maintained through autonomous replication or alternatively may be integrated into the host's genome. The recombinant DNA molecules preferably are provided with appropriate control sequences, compatible with the designated host which can regulate the expression of the inserted nucleic acid sequence.

The most widely used hosts for expression of recombinant nucleic acid molecules are bacteria, yeast, insect cells and mammalian cells. Each system has advantages and disadvantages in terms of the vector used, potential ease of production and purification of a recombinant polypeptide and authenticity of product in terms of tertiary structure, glycosylation state, biological activity and stability and will be a matter of choice for the skilled addressee.

In addition to promoting expression of an NNG2 polypeptide in cells, in certain circumstances, it is advantageous to substantially prevent or reduce the expression or activity of the native NNG2 in a host. Thus, according to a further aspect of the invention, there is provided an antisense nucleotide fragment complementary to a nng2 nucleotide sequence of the present invention. Also provided is an antisense nucleotide fragment complementary

to the region of the nng2 nucleotide sequence encoding the SH3 domains. Included in the scope of"antisense nucleotide fragment"is the use of synthetic oligonucleotide sequences, or of equivalent chemical entities known to those skilled in the art, for example, peptide nucleic acids. Also provided is a nucleotide fragment comprising a nucleotide sequence which, when transcribed by the cell, produces such an antisense fragment. Typically antisense RNA fragments will be provided which bind to complementary nng2 mRNA fragments to form RNA double helices, allowing RNAse H to cleave the molecule and rendering it incapable of being translated by the cell into polypeptides.

A further aspect of the present invention provides antibodies specific to the NNG2 polypeptide or epitopes thereof. Production and purification of antibodies specific to an antigen is a matter of ordinary skill, and the methods to be used are clear to those skilled in the art. The term antibodies can include, but is not limited to polyclonal antibodies, monoclonal antibodies (mAbs), humanised or chimeric antibodies, single chain antibodies, <BR> <BR> <BR> <BR> <BR> Fab fragments, F (ab') 2 fragments, fragments produced by a Fab expression library, anti-idiotypic (anti-Id) antibodies, and epitope binding fragments of any of the above. Such antibodies may be used in modulating the expression or activity of the NNG2 polypeptide, or in detecting said polypeptide in vivo or in vitro.

The present invention further provides a recombinant or synthetic NNG2 polypeptide for the manufacture of reagents for use as prophylactic or therapeutic agents in mammals. In particular, the invention provides pharmaceutical compositions comprising the recombinant or synthetic NNG2 polypeptide together with a pharmaceutically acceptable carrier therefor.

It has been shown that the NNG2 polypeptide has an effect on part of the PI3 kinase/Akt pathway. This pathway plays an important role in mediating the survival effects of a number of growth factors in many cell types, including neuronal growth factors. Thus, in another aspect of the present invention there is provided an inhibitor of PI3 kinase.

The polypeptide and nucleotide sequence of the present invention may be used in identification and isolation of further inhibitors of PI3 kinase. The polypeptide and nucleotide sequence may also be used in the identification and isolation of further members of the PI3 kinase/AKT pathway. The polypeptide and nucleotide sequence may also be used in the identification and isolation of further factors which may affect neuronal survival. The procedures and techniques which would be used are a matter of ordinary skill in the art, comprising for example GST fusion, yeast 2-hybrid systems and immunoprecipitation, and are described for example in Current Protocols In Molecular Biology, Ausubel F et al (eds.), Wiley, USA, 1995.

According to a still further aspect of the present invention, there is provided use of a polypeptide or nucleic acid sequence as hereinbefore described for promoting selective cell death. There is also provided use of a polypeptide or nucleic acid sequence as hereinbefore described in preventing, delaying, or inhibiting selective cell death. There is further provided a method of promoting selective cell death comprising the introduction of an NNG2 polypeptide into a cell displaying an abnormal cell cycle or aberrant proliferation. Such a method may find particular application in treatment of cancerous conditions by reducing the number of cancer cells.

Also provided is a method of inhibiting selective cell death comprising the introduction into a cell of a nucleotide sequence or an antibody which substantially prevents or reduces expression or activity of the NNG2 polypeptide.

A yet further aspect of the present invention provides use of a sense polypeptide or nucleic acid sequence as hereinbefore described in the treatment of cell proliferative disorders.

There is also provided the use of a polypeptide or nucleic acid sequence as hereinbefore described in the treatment of degenerative disorders of the nervous system.

The present invention also relates to methods for prognostic and diagnostic evaluation of various cell death- related disorders, and for the identification of subjects who are predisposed to such disorders, for example by

examination of allelic variation by determination of the nng2 DNA sequence in an individual. Furthermore, the invention provides methods for evaluating the efficacy of drugs for such disorders, and monitoring the progress of patients involved in clinical trials for the treatment of such disorders.

The invention further provides methods for the identification of compounds which modulate the expression of nng2 or the activity of NNG2 products involved in processes relevant to the differentiation, maintenance and/or death of cells.

The biological function of the nng2 gene can be more directly assessed by utilizing relevant in vivo and in vitro systems. In vivo systems can include, but are not limited to, animal systems which naturally exhibit the symptoms of cell death disorders, or ones which have been engineered to exhibit such symptoms. Further, such systems can include, but are not limited to transgenic animal systems. In vitro systems can include, but are not limited to, cell-based systems comprising nng2/NNG2 expressing cell types. The cells can be wild type cells, or can be non- wild type cells containing modifications known or suspected of contributing to the disorder of interest.

In further characterising the biological function of the nng2 gene, the expression of the nng2 gene can be modulated within the in vivo and/or in vitro systems, i. e. either overexpressed or underexpressed in, for example, transgenic animals and/or cell lines, and its subsequent

effect on the system can then be assayed. Alternatively, the activity of the product of the identified gene can be modulated by either increasing or decreasing the level of activity in the in vivo and/or in vitro system of interest, and its subsequent effect then assayed.

The information obtained through such characterisations can suggest relevant methods for the treatment or control of cell death disorders. For example, relevant treatment can include a modulation of gene expression and/or gene product activity. Characterisation procedures such as those described herein can indicate whether such modulation should be positive or negative. As used herein,"positive modulation"refers to an increase in gene expression or activity of the gene or gene product of interest."Negative modulation", as used herein, refers to a decrease in gene expression or activity.

In vitro systems can be designed to identify compounds capable of binding the nng2 gene products of the invention.

Compounds identified can for example, be useful for example, in modulating the activity of wild type and/or mutant nng2 gene products, can be useful in elaborating the biological function of the nng2 gene products, or can disrupt normal nng2 gene product interactions.

These and other aspects of the invention shall now be further described, by way of example only, and with reference to the accompanying figures, in which ruk is used as an alternative name for the rat nng2 gene, and which show:

Figure 1. Autoradiograms of a Northern blot of total RNA isolated from newborn rat tissues hybridised with 32p nick-translated rat nng2 (ruk) and L27 probes showing the relative levels of expression of ruk mRNAs in different tissues.

Figure 2. Autoradiogram of a Northern blot of total RNA <BR> <BR> <BR> <BR> isolated from adult rat tissues hybridised with a 32p nick-translated ruk probe showing the relative levels of expression of ruk mRNAs in different tissues.

Figure 3. Autoradiograms of a Northern blot of total RNA isolated from newborn (NB) and adult (AD) rat tissues <BR> <BR> <BR> <BR> hybridised with a P nick-translated ruk probe showing the relative levels of expression of ruk mRNAs in different tissues.

Figure 4. Autoradiogram of a Northern blot of total RNA isolated from rat E7 embryos and E16 embryonic heads and <BR> <BR> <BR> <BR> bodies hybridised with a 32 P nick-translated ruk probe showing the relative levels of expression of ruk mRNAs.

Figure 5. Schematic representation of ruk mRNAs (large boxes) and cDNA clones isolated from skin (prefixed with s), heart (prefixed with h) and cerebellum (prefixed with c) cDNA libraries or as RACE (R) and RT-PCR (P) products (lines below). The regions of differing 5'sequence are shaded by cross-hatch, waves and stipple. Presumptive

translation initiation codons are shown as large flaglike arrowheads above each mRNA. In-frame stop codons are designated by asterisks. The filled boxes inside the longest mRNA show the positions of SH3 domains and the open box the proline-rich region. The small arrows below the mRNAs represent oligonucleotide primers used for RACE (R) or RT-PCR (h, f, and c, see also Figure 6 and text).

Figure 6. Analysis of the products of RT-PCR amplification in a 1.2% agarose gel stained with ethidium bromide.

Primers f, h and c are designated by small arrows on Figure <BR> <BR> <BR> 5. Primers f and c amplify fragments of ru mRNA, primers h and c amplify fragments of rukh mRNA. Poly (A) + RNA isolated from the newborn rat tissues were used as templates for first-strand cDNA synthesis. The amounts of cDNA were normalised by amplification of a fragment of the housekeeping gene encoding ribosomal L27 protein from an aliquot of each sample.

Figure 7. Nucleotide and amino acid sequences of rat ruk cDNA and polypeptides. a) shows the sequences of rukl; b) those of ruk; c) those of ruk; d) those of ruk,; and e) those of rukh. The polypeptide sequences of a), b) and e) are also shown in Figure 8 (a); these of c) and d) are shown in Figures 8 (b) and (c), respectively.

Figure 8. Amino acid sequences of rat NNG2 (Ruk) proteins coded by the different mRNAs shown in Figures 5 and 7. (a) shows the protein coded by rukl mRNA. SH3 domains A, B and C are boxed and the proline-rich domain is underlined. The double-underlined sequence shows the C-terminal region identical in all three proteins shown here. The ATG encoding the methionines marked by asterisks could be used <BR> <BR> <BR> <BR> as translation initiation codons in the 2.5 kb ruk., and<BR> <BR> <BR> <BR> ru mRNAs (Met262) and the 1.8 kb rukh mRNA (Met""),(b)<BR> <BR> <BR> <BR> <BR> shows the protein coded by ruk,,, mRNA. The sequence downstream of amino acid 4 is identical to the sequence in (a) downstream of amino acid 242. (c) shows the protein coded by ruk, mRNA. The first four amino acids are identical to the first four amino acids in (b) and the sequence downstream of amino acid 54 is identical to the sequence in (a) downstream of amino acid 608.

Figure 9. Detection of Ruk proteins in total protein extracted from adult rat tissues (b) and cell lines (c) by Western blotting. (a) shows elimination of specific bands after pre-incubation of the antibody with excess of the peptide used for immunisation. The positions of molecular mass markers are shown.

Figure 10. The survival of trigeminal neurons in culture is affected by overexpression of Rukl protein. P2 mouse trigeminal neurons were plated in medium containing 2 ng/ml of NGF and injected with expression plasmids. The number

of surviving neurons was counted 48 hours later and is expressed as a percentage of the number counted 2 hours after injection. The means and s. e. m. of at least three independent experiments are shown. The Rukl-C protein comprises the polypeptide from amino acid 427 of Rukl to the carboxy terminus. The form of Rukm used in this experiment was Ruk.

Figure 11. The survival of trigeminal neurons in culture is affected by overexpression of antisense ru RNA. P2 mouse trigeminal neurons were injected with expression plasmids and deprived of NGF by extensive washing following injection. The number of surviving neurons was counted 48 hours later and is expressed as a percentage of the number counted 2 hours after injection. The means and s. e. m. of at least three independent experiments are shown.

Figure 12. Reversal of the effect of Rukl overexpression by overexpression of activated forms of Akt and the pllO catalytic subunit of PI3 kinase. P2 mouse trigeminal neurons were plated in medium containing 2 ng/ml of NGF or 5 ng/ml of CNTF and injected with expression plasmids. The number of surviving neurons was counted 48 hours later and is expressed as a percentage of the number counted 2 hours after injection. The means and s. e. m. of at least three independent experiments are shown.

Figure 13. The survival of trigeminal neurons in culture is affected by inhibitors of the PI3 kinase/Akt pathway and Ras/MAP kinase pathways. P2 mouse trigeminal neurons were grown with and without NGF and with NGF in the presence of the PD90859 and LY294002 reagents at two different concentrations alone or in combination. PD90859 is an inhibitor of the MAP kinase pathway, while LY294002 is an inhibitor of PI3 kinase. Some neurons were also injected with the Ruk1 expression plasmid and incubated in the presence or absence of PD90859. The number of surviving neurons was counted 48 hours after addition of reagents.

The means and s. e. m. of at least three independent experiments are shown.

Figure 14. Relative locations of human nng2 clones and rat ruk cDNA.

Figure 15. Nucleotide and predicted amino acid sequences of (a) human nng2 cDNA and (b) the long isoform of human nng2 cDNA.

Figure 16. Functional domains of human NNG2 protein.

Example 1 Cloning of ruk cDNA and analysis of ruk mRNA expression A subtractive cloning procedure [1,2] was used to isolate cDNAs corresponding to mRNAs that are expressed at higher levels in newborn than in adult rat cerebellum. Of

the clones isolated, one which has been named ruk was chosen for further study.

Using the insert of this clone as a probe for Northern hybridisation, a 3.5 kb transcript was detected in the rat cerebellum. The level of this transcript decreased in the cerebellum during the postnatal period, reflecting the way in which the clone was isolated. Interestingly, the level of this transcript did not decrease over the same period of development in other CNS (central nervous system) regions such as the cerebral cortex, hippocampus and brain stem (data not shown). The 3.5 kb transcript was detected in all tissues studied, with the highest level in the nervous system (Figures 1-4). In addition to the 3.5 kb transcript, several other ruk transcripts were detected in some tissues (Figures 1-4). In newborn and adult rat skin in addition to the 3.5 kb transcript two prominent transcripts (2.5 kb and 1.5 kb) could be detected (Figures 1,3). The only ruk transcript in adult testis has the size of 1.3 kb (Figure 2). Very high levels of a specific 1.8 kb ruk transcript are present in newborn and adult rat heart (Figure 3). Whereas the 3.5 kb transcript was the most abundant in RNA samples isolated from E16 embryonic heads, 3.5 kb, 2.5 kb and 1.5 kb ruk transcripts were present at similar levels in E16 embryonic bodies and whole E7.5 embryos (Figure 4).

To isolate full-length clones that represent different ruk transcripts, newborn rat skin, cerebellum and heart cDNA libraries were screened with a ruk probe. Additional

clones were isolated by the RACE technique using newborn rat skin and cerebellum poly (A) + RNAs as templates. The organisation of these clones and the ruk mRNAs is shown in Figure 5. Most of the clones, including the original PCR (polymerase chain reaction) clone, were primed not on the terminal poly (A) sequence of ruk mRNA but on an internal oligo (A) stretch that lies within the 3'UTR (3' untranslated region). Most, if not all ruk mRNAs have the same 3'ends because Northern hybridisation using a 3'-specific probe derived from a 0.4 kb fragment adjacent to the poly (A) sequence revealed the same transcripts as an internal probe in all tissues studied (data not shown).

Thus, all ruk transcripts were generated by alternative splicing and/or different promoter usage in the coding or 5'UTR regions.

The longest mRNA (rukl) corresponds to the 3.5 kb band detected by Northern hybridisation in all tissues (Figures 1-5). This was confirmed by Northern hybridisation with a 5'rukl-specific probe (Figure 3a) and by RT-PCR (reverse- transcriptase-PCR). As shown in Figure 6, amplification using oligonucleotide primer f (located in the 5'region unique to rukl mRNA) and primer c (common for all variants) generated a major fragment of the expected size (1292 bp) in all tissues studied at levels that were consistent with the results of Northern hybridisation (Figures 1 and 6).

It is likely that the minor PCR product (1163 bp) of this reaction was amplified from an mRNA with an internal deletion of 129 nt, similar to the deletion found in cDNA

clone h63 (Figures 5 and 6).

Three mRNAs of intermediate size (ruk<, ruks and ruk-) correspond to the 2.5 kb band detected by Northern hybridisation (Figure 5). Sequence analysis of cDNA clones revealed that these mRNAs differ only in their 5'exons. <BR> <BR> <BR> <P>The shortest mRNA (ruk,) corresponds to the 1.5 kb band detected by Northern hybridisation (Figures 1-5). This is generated by splicing of an internal part of the ruk primary transcript (Figure 5).

All clones isolated from the newborn rat heart cDNA library had a similar organisation. A specific 5'sequence is spliced to the acceptor site downstream to the one used in ruk. mRNAs. In one clone a deletion of 129 bp was found (Figure 5). In agreement with this and the results of Northern hybridisation, RT-PCR using primer h revealed two fragments of the expected sizes in heart, but not in skin or cerebellum. Low levels of these transcripts were detected in muscle and in lung (Figure 6).

Example 2 Proteins coded by ruk mRNA Analysis of ruk mRNA sequences showed that a set of proteins with various features is coded by the ruk gene.

The nucleotide sequences of a number of the clones identified in Figure 5 are given in Figure 7, while the protein sequences are given in Figures 7 and 8. All of these proteins share common C-terminal regions, but their N-termini are very different in length and structure

(Figures 5,7 and 8). In most of the ruk mRNAs analysed, an upstream, in-frame termination codon was found and the first in-frame ATG codon was surrounded by a good Kozak consensus sequence [3]. The common C-terminal region (double underlined in Figure 8a) has no similarity to other proteins in available data banks. Likewise, there are no similarities to other proteins in the upstream region that <BR> <BR> <BR> <BR> <BR> is shared by proteins coded by rukl, ruk. and partly by rukh mRNAs (Figures 5,7 and 8). Interestingly, this region has a substantial number of serine residues (25%). In addition to these novel regions, there were two known motifs in ruk <BR> <BR> <BR> <BR> <BR> and ruk. MRNA. Three typical SH3 domains (A, B, C in Figure 8) and two proline-rich domains (underlined in Figure 8a) are present in the N-terminus of Rukl protein, and one of these SH3 domains (C) and all proline-rich domains are present in the N-terminus of Rukm (Figures 8a, b).

Antibodies were raised by immunising rabbits with a C-terminal peptide that is common to all Ruk isoforms.

Western blot analysis showed various Ruk isoforms in a variety of rat tissues (Figures 9a, b) at levels that correlated with the levels of the corresponding mRNAs shown in Figures 1 and 2. Because the sequence of the immunising Ruk peptide is identical in rat and human, the antibody detected Ruk proteins in both rat and human cell lines (Figure 9c). The mobility of the Rukl protein corresponded to a molecular mass of 85 kDa which is much higher than that predicted from the amino acid sequence, suggesting the existence of post-translational modifications or a

conformation that retards mobility. Pre-incubation of the antibody with the peptide used for immunisation completely eliminated specific Ruk bands on Western blotting.

Staining of cultured fibroblasts and keratinocytes with anti-Ruk antibodies showed that Ruk is a cytoplasmic protein (data not shown).

Example 3 Modulating Ruk expression affects neuronal survival To investigate the potential effect of Ruk isoforms on cell survival, microinjection was used to introduce Ruk sense and antisense expression plasmids into cultured primary sensory neurons from the postnatal day 2 (P2) mouse trigeminal ganglion. At this stage of development, these neurons survive in medium containing either NGF or CNTF and rapidly die in the absence of these factors. Neurons that were incubated in the presence of NGF or CNTF and injected with an plasmid expressing Rukl protein did not survive as well as neurons injected with the empty expression vector plasmid (Figure 10) This survival-inhibiting effect of Ruk is dependent on the N-terminal part of the protein that contains the first two SH3 domains because it was observed with an expression plasmid that has a deletion of the C-terminal part of the Rukl protein, but was not observed with plasmids expressing Ruk, 3 which does not contain the first two SH3 domains (Figure 10). These results raised the possibility that Rukl antagonises the survival machinery of the cell. If so, reducing the expression of endogenous

Ruk would enhance cell survival. Neurons were therefore injected with an expression plasmid containing in the antisense orientation the 5'-UTR and N-terminal part of ru that is not found in other ruk transcripts. Whereas almost all neurons deprived of trophic factors died within 48 hours in vitro, neurons injected with this antisense construct showed significantly enhanced survival similar to <BR> <BR> <BR> <BR> <BR> neurons injected with a plasmid expressing BclXl, a well established anti-apoptotic protein (Figure 11).

It is well established that the PI3 kinase/Akt pathway plays an important role in mediating the survival effects of several growth factors including NGF and CNTF.

Experiments were carried out to ascertain whether Rukl acts in the PI3 kinase/Akt signalling pathway. In cultures containing either NGF or CNTF, the survival-inhibitory effects of Rukl overexpression were prevented by co-injecting plasmids expressing either an activated form of Akt or an activated form of the pllO catalytic subunit of PI3 kinase (Figure 12). However, overexpression of a non-functional form of the PI3 kinase catalytic subunit did not significantly affect the survival-inhibitory action of Rukl overexpression (Figure 12). These results suggest that Ruk1 inhibits the neurotrophic factor survival signal upstream of the catalytic subunit of PI3 kinase.

In addition to the PI3 kinase/Akt pathway, the Ras/MAP kinase pathway plays a role in mediating the survival effects of NGF. Pharmacological inhibitors of either pathway (PD90859 for MAP kinase and LY294002 for the

catalytic subunit of PI3 kinase) reduced the number of neurons surviving with NGF, and have an additive inhibitory effect on the NGF survival response when combined (Figure 12). PD90859 markedly potentiated the survival-inhibitory effects of Rukl overexpression (Figure 13), suggesting that Rukl acts at least in part on a pathway that is distinct from the Ras/MAP kinase pathway.

Example 4 Cloning of human nng2 (ruk) A human substantia nigra XZAPII library was purchased from Stratagene (La Jolla, California, USA). The XZAPII vector allows plasmid rescue of selected clones and is therefore a lot easier to use than a basic lambda system.

The library was manipulated and propagated to manufacturers instructions. For the initial screen 5x104 plaques were plated on each of four plates and plaque lifts were performed in duplicate. The rat Ruk cDNA clone was used to screen the library for the human nng2 cDNA. The clone was digested with the restriction enzymes SstI and XhoI in order to release the clone from its vector and to fragment the ruk sequence itself. XhoI does not cut within the ruk cDNA but two SstI sites exist: one at 1124 nt and the other at 2027 nt. The l. lkb fragment was purified and used to probe the plaque-lift filters (Figure 14). Hybridisations and washes were done under stringent conditions.

Individual positive clones were taken through to a tertiary screen and successfully excised in vivo. Sequence analysis

revealed that one of these matched the rat nng2 (ruk) cDNA sequence. This clone was named SN4.1 and was found to be 1.4kb in length, lacking both 5'and 3'ends of the sequence homologous to the rat ruk cDNA sequence (Figure 14).

Clone SN4.1 was then used to re-screen the library in an attempt to isolate a full-length clone. The library was again plated at 5xl0'plaques per plate but this time only three plates were used. The duplicate filters were hybridised at high stringency and four positives were detected. After a second round of screening the four positive plaques could be individually isolated, and the DNA sequence derived from one clone was found to match the rat nng2 (ruk) cDNA sequence. This clone, SNIII, was 2.5kb in length and, although the 3'nng2 sequences were present, those at the 5'end were missing (Figure 14).

The 5'end of the human nng2 cDNA was finally cloned with the help of a 5'-RACE PCR system purchased from Clontech (Palo Alto, California, USA). Their complete human brain Marathon-ready cDNA was used as a template in the PCR reaction. The 5'RACE primer was supplied by Clontech but the 3'primer was gene-specific and designed from the known human Ruk sequence; GSP1, situated at 1090 nucleotides into the rat Ruk cDNA. The sequence of the primer is anti-sense from the cDNA in order for it to function correctly as a 3'PCR primer: cccaccagcctacgtcgatgcagtcc. The PCR was performed according to the manufacturers recommendations and a smear

of a product was received. Therefore, the product was diluted one in fifty and used in a second round of PCR with nng2 gene-specific primers at both ends. GSP1 was again used as the 3'primer but the 5'primer was one designed from the known rat cDNA sequence. The oligonucleotide is denoted 5ruk2, and encompasses the ATG translational start site of the rat sequence: agctcggcttggcaccaatgg. The PCR reaction conditions were as follows: 94°C, 3m 94°C, 30s then 68°C, 3m for 5 cycles 94°C, 30s then 66°C, 3m for 5 cycles 94°C, 30s then 64°C, 3m for 15 cycles.

A smeared product of lkb was generated and cloned into the pCR2. 1 TA cloning vector (Invitrogen, Leek, The Netherlands). Two populations of clones were obtained, one with a slightly larger size of insert than the other.

Sequencing proved the smaller clone to be the human equivalent of the 5'end of the rat nng2 (ruk) cDNA (Figure 14). Specifically, this is the human equivalent of the longest known transcript from the rat nng2 gene, rat nng21 (rukl). The nucleotide sequence and the predicted amino acid sequence are shown in Figure 15 (a). The larger clone was also human nng2, but a different isoform. This clone had an additional 132 nucleotides beginning at nucleotide 528. The site of insertion of this additional sequence is at an intron/exon splice junction and is therefore likely to be an additional exon. The nucleotide sequence and the predicted amino acid sequence are shown in Figure 15 (b).

The predicted protein sequences were searched for functional sites and domains using uprosite"within Genejockey II (Biosoft, Cambridge, UK). The following sites were detected, although it is unknown whether any of them are functional: N-glycosylation site (x2); cAMP-and cGMP-dependent protein kinase phosphorylation site (xl); protein kinase C phosphorylation site (x15); casein kinase II phosphorylation site (x15); N-myristoylation site (xll); Amidation site (x4); and cell attachment sequence (xl).

The longer isoform included the following additional sites: casein kinase II phosphorylation site (x2); N- myristoylation site (xl); and an ATP/GTP binding site motif A (P-loop).

A second database was used to look for further functional domains. This identified three SH3 domains, thought to be involved in protein-protein interactions, and a RNA hairpin recognition motif (Figure 16). The additional exon of the sequence of Figure 15 (b) is also marked on Figure 16.

A search was done to compare the NNG2 protein sequence with that of known proteins in the internal data base of Glasgow University. Matches were found to other proteins with SH3 domains, but the closest match found was to CD2AP, an adaptor protein linking the CD2 receptor in T-cells to downstream signals. The level of homology between the proteins is 54% nucleotide similarity, 39% amino acid similarity and 66% overall homology.

Experimental Procedures RT-PCR analysis cDNA was synthesised from 1 pg of poly (A) + RNA with 0.5 pg of random hexamer primers using SuperScript reverse transcriptase according to the supplier's recommendations (GIBCO-BRL, Ontario, Canada). The amounts of cDNA generated from different tissues used in PCR reactions were normalised to the RT-PCR amplification product of an mRNA encoding a housekeeping protein (ribosomal L27 protein).

All amplifications were carried out in 50 l using Taq-polymerase and buffer from Promega (Southampton, UK): 35 cycles of 45 sec at 95°C; 30 sec at 58°C; 90 sec at 72°C.

20 Al from each reaction were analysed on an agarose gel.

5'RACE was performed as described [4].

Oligonucleotides used h: 5'-ATATGTACAGAACAGAGGGAC-3' f: 5'-AACTAAGACCAAGGTCGATTG-3' c: 5'-CACCTCCATCTGCAATCGGAG-3' RACE: 5'-GTAATCAAACTCCACTATG-3' Miscellaneous procedures RNA extraction, isolation of poly (A) + RNA, preparation of hybridization probes, Northern hybridization, library screening and plasmid sequencing were performed as described [4,5]. Adult rat skin and heart cDNA libraries in the XZAPII vector were constructed using a kit from Stratagene. Subtractive hybridisation [2] was used to isolate cDNA clones of mRNAs that are enriched in

developing rat cerebellum.

Anti-Ruk antibody Rabbits were immunised with the 15-mer C-terminal peptide of rat Ruk (LQMEVNDIKKALQSK) conjugated with keyhole limpet haemocyanin (Calbiochem, Beeston, Nottingham, UK) activated by MBS (Sigma, Poole, Dorset, UK) [6]. Monospecific antibodies were purified from the antisera by affinity chromatography using the antigen bound to NHS-activated columns (Supelco, Pennsylvania, USA). The rabbit anti-rat monospecific antibody was used at dilutions of 1: 500 for Western blot/ECL detection of NNG2 in total cell protein samples. In some experiments 10 ml of diluted antibody were preincubated with 15 jug of the recombinant rat Ruk protein at room temperature for 2 hours.

Protein extraction and Western blotting/ECL detection Tissues or cultured cells were homogenised directly in SDS-PAGE loading buffer [7] and incubated for 5 min in a boiling water bath. Total protein concentrations were measured by the dotMETRIC assay (Geno Technology, St.

Louis, Missouri, USA). 15 ßg of total protein were used for SDS-PAGE [7]. Electroblotting on Hybond-PVDF membranes was performed in TRIS/glycine/methanol buffer as recommended by the membrane supplier (Amersham International plc, Amersham, UK). Rainbow markers from Amersham were used as protein size standards. After washing with PBS (Phosphate Buffered Saline), the membrane was blocked for 1 hour at room temperature in 4% skimmed milk/0.05% Tween-20TM/PBS. This buffer was also used for

the hybridisation reactions with primary and secondary antibodies and subsequent washes. The final two washes were in 0.1% Tween-20TM/PBS. ECL detection was carried out as recommended by the Amersham protocol.

Immunohistochemistry Cultured cells were fixed on the Petri dish with acetone/methanol mixture (1: 1) at-20°C for 10 minutes, washed with TBT and blocked with 5% goat serum in TBT (20 mM TrisHCl pH7.5; 150 mM NaCl; 0.1% Triton X-100) followed by incubation with the anti-Ruk antibody (1: 50) at 4°C for 16 hours in 1% goat serum/TBT. FITC-conjugated secondary anti-rabbit antibody (Jackson ImmunoResearch Laboratories, Pennsylvania, USA) was used for detection.

Tissue Culture and Microinjection Techniques Purified P2 mouse trigeminal neurons were grown on a polyornithine/laminin substratum, in defined Sato medium [8] either in the presence of 2ng/ml NGF or 5ng/ml CNTF.

After plating, all cultures were maintained in a humidified CO2 incubator at 37°C. Approximately 48 hours after plating, expression plasmids (pRc/CMV vector only, nng2 cDNAs in pRc/CMV, Akt/PKB and PI3 kinase cDNAs in pSG5) were injected into the nucleus of the neurons as described previously [9,10]. The cultures were supplemented with fresh 2ng/ml NGF or fresh 5ng/ml CNTF and the starting number of neurons in each experimental condition was counted two hours after injection. The number of surviving neurons was counted 48 hrs later and expressed as a percentage of the starting number. In some experiments,

cultures were intensively washed with Sato medium without neurotrophic factors two hours after injection and then incubated in the presence of anti-NGF antibody (gift of Y.

Barde.......). The number of surviving neurons was counted 48 hrs later.

REFERENCES 1. Akopyan, A. N., Baka, I. D., Chestkov, A. V., Buchman, V. L., and Georgiev, G. P. (1996). Cloning of genes differentially expressed in the cerebellum during postnatal rat development. Genetika 32,886-895.

2. Baka, I. D., and Buchman, V. L. (1996). Subtractive cDNA cloning from limited amounts of biological material.

Anal. Biochem. 237,155-157.

3. Kozak, M. (1989). The scanning model for translation: an update. J Cell Biol 108,229-241.

4. Buchman, V. L., Ninkina, N. N., Bogdanov, Y. D., Bortvin, A. L., Akopian, H. N., Kiselev, S. L., Krylova, O. Y., Anokhin, K. V., and Georgiev, G. P. (1992). Differential splicing creates a diversity of transcripts from a neurospecific developmentally regulated gene encoding a protein with new zinc-finger motifs. Nucleic Acids Res. 20,5579-5585.

5. Buchman, V. L., Sporn, M., and Davies, A. M. (1994).

Role of transforming growth factor-beta isoforms in regulating the expression of nerve growth factor and neurotrophin-3 mRNA levels in embryonic cutaneous cells at different stages of development Development 120,1621-1629.

6. Harlow, E., and Lane, D. (1988). Antibodies: A Laboratory Manual. Cold Spring Harbor Laboratory Press, New York.

7. Laemmli, E. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4.

Nature 227,680-685.

8. Buchman, V. L., and Davies, A. M. (1993). Different neurotrophins are expressed and act in a developmental sequence to promote the survival of embryonic sensory neurons Development 118,989-1001.

9. Allsopp, T. E., Wyatt, S., Paterson, H. F., and Davies, A. M. (1993). The proto-oncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis Cell 73,295-307.

10. Ninkina, N., Adu, J., Fischer, A., Pinon, L. G., Buchman, V. L., and Davies, A. M. (1996). Expression and function of TrkB variants in developing sensory neurons EMBO J. 15,6385-6393.