Background

Cell death is a fundamental aspect of animal development. Many cells die during the normal develop- ent of both vertebrates (Glucksmann, Biol . Rev.

Cambridge Philos . Soc . 25:59-86 (1951)) and inverte¬ brates (Truman, Ann . Rev . Neurosci . 7:171-188 (1984)). These deaths appear to function in morphogenesis, metamorphosis and tissue homeostasis, as well as in the generation of neuronal specificity and sexual dimorphism (reviewed by Ellis et al . , Ann . Rev. Cell Biol . 7:663- 698 (1991)). An understanding of the mechanisms that cause cells to die and that specify which cells are to live and which cells are to die is essential for an understanding of animal development.

The nematode Caenorhabditiε elegans is an appropriate organism for analyzing naturally-occurring or programmed cell death (Horvitz et al . , Neurosci . Comment . 1 :56-65 (1982 ) ) . The generation of the 959 somatic cells of the adult C. elegans hermaphrodite is accompanied by the generation and subsequent deaths of an additional 131 cells (Sulston and Horvitz, Dev. Biol . 82:110-156 (1977); Sulston et al . , Dev . Biol . 100 : 64-119 (1982)). The morphology of cells undergoing programmed cell death in C. elegans has been described at both the light and electron microscopic levels (Sulston and Horvitz, Dev . Biol . 82:100-156 (1977); Robertson and Thomson, J. Embryol . Exp. Morph . 67:89-100 (1982)).

Many genes that affect C. elegans programmed cell death have been identified (reviewed by Ellis et al . , Ann . Rev . Cell Biol . 7:663-698 (1991)). The activities of two of these genes, ced-3 and ced-4 , are required for the onset of almost all C. elegans programmed cell deaths (Ellis and Horvitz, Cell 44:817-829 (1986)).

When the activity of either ced-3 or ced-4 is eliminated, cells that would normally die instead survive and can differentiate into recognizable cell types and even function (Ellis and Horvitz, Cell 44:817- 829 (1986); Avery and Horvitz, Cell 51:1071-1078 (1987); White et al . , Phil . Trans . R . Soc . Lond . B . 331 : 263-211 (1991)). Genetic mosaic analyses have indicated that the ced-3 and ced-4 genes most likely act in a cell autonomous manner within dying cells, suggesting that the products of these genes are expressed within dying cells and either are cytotoxic molecules or control the activities of cytotoxic molecules (Yuan and Horvitz, Dev . Biol . 138:33-41 (1990)).

Summary of the Invention This invention is based mainly on two experimental findings and their implications: 1) that human interleukin-1/3 convertase (ICE) , which converts pro- interleukin-lβ to the active cytokine and is involved in the inflammatory response in humans, has considerable similarity to the protein encoded by the C . elegans cell death gene, ced-3 ; and 2) that fusion constructs containing amino-terminal portions of the ced-3 gene can prevent cell death in C. elegans . As discovered by Applicant, the human ICE and ne atode Ced-3 proteins have an overall amino acid identity of 28%. A higher . degree of similarity was found in the carboxyl-terminal region, a region shown to be critical for the activities of both proteins. Furthermore, three sequences important for ICE activity, the region surrounding the active cysteine and two autocleavage sites, have been shown to be conserved in the ced-3 gene product.

Thus, significant structural similarity has been shown between two proteins which previously were thought to be unrelated (to have dissimilar physiological

roles) . This finding leads to several implications, some of which are:

1) that the human ICE gene has an activity similar to that of ced-3 in causing cell death; 2) that the Ced-3 protein is also a cysteine protease with a substrate specificity similar to that of ICE;

3) that mutations in the ICE gene corresponding to mutations in the ced-3 gene will produce similar effects, such as inactivation and constitutive activation;

4) that the ced-3 and ICE genes are members of a family of structurally related genes, referred to herein as the ced-3/ICE family, some of which are likely to be cell death genes and some of which may encode substrate- specific proteases;

5) that inhibitors of ICE, such as peptide aldehydes which contain the ICE recognition site or a substituted recognition site and the cowpox virus CrmA protein, may also be useful for inhibiting cell deaths; and

6) that inhibitors of ced-3 , such as inhibitory portions of the gene or encoded product, may also be useful for inhibiting inflammation. This hitherto unknown connection between a cell death protein and a protease involved in the inflammatory response provides a basis for developing novel drugs and methods for the treatment of acute and chronic inflammatory disease, of leukemias in which IL- 1/3 is implicated, and of diseases and conditions characterized by cell deaths (such as myocardial infarction, stroke, traumatic brain injury, viral and other types of pathogenic infection, neural and muscular degenerative diseases, aging, hair loss) . In addition, drugs which increase cell deaths and which are useful

for reducing the size or proliferative capacity of cell populations, such as cancerous cells, infected cells, cells which produce autoreactive antibodies, and hair follicle cells, as well as drugs which incapacitate or kill organisms, such as pests, parasites and recombinant organisms, can be developed using the ced-3 , ICE, and other ced-3/ICE genes and their gene products.

This work also provides probes and methods for identifying additional members of the ced-3/ICE gene family. Genes related to ced-3 and ICE are expected to exist in various organisms. Some of these may be cell death genes and/or proteases. The sequences of these related genes and their encoded products can be compared, for instance, using computer-based analysis, to determine their similarities. Structural comparisons, for example, would indicate those regions or features of the genes or encoded products which are likely to be functionally similar and important. Such information can be used to design drugs which mimic or alter the activity of the ced-3 , ICE, or other ced-3/ICE genes, and which may, thus, be useful in the various medical and agricultural applications mentioned above.

In addition, another mammalian protein, the murine NEDD-2 protein, was also found to be similar to Ced-3. Interestingly, NEDD-2 is not significantly similar to ICE. Thus, another potential mammalian cell death gene was identified.

Also described herein is the discovery that fusion constructs which encode an amino-terminal portion of the Ced—3 protein fused to 3-galactosidase act as inhibitors of cell death in C. elegans . Due to its structural similarity to Ced-3, constructs encoding corresponding portions of the human ICE protein are also expected to inhibit the enzymatic activity of ICE in cleaving interleukin-l . Thus, inhibitors comprising an amino-

terminal portion of the Ced-3 protein, ICE protein or another member of the Ced-3/ICE family and RNAs and DNA constructs which express these protein portions are potentially useful for decreasing cell deaths and/or inflammation involved in various pathologies. Methods for identifying other inhibitory portions of the ced-3 and ICE genes are also described.

Furthermore, deletion of the inhibitory amino- terminal portions of the ced-3 and ICE genes may result in constitutive activation of the genes. Constitutively activated carboxyl-terminal portions of the genes, or their encoded products, may thus be useful in applications where increased cell deaths or an increased inflammatory response are desired.

Brief Description of the Drawings

Figure 1 shows the physical and genetic maps of the ced-3 region on chromosome IV.

Figure 2 summarizes the experiments to localize ced-3 within C48D1. Restriction sites of plasmid C48D1 and subclone plasmids are shown. Ced-3 activity was scored as the number of cell corpses in the head of LI young animals. ++, the number of cell corpses above 10. +, the number of cell corpses below 10 but above 2. - , the number of cell corpses below 2. Figures 3A-H show the nucleotide sequence (SEQ ID N0:1) of ced-3 and deduced amino acid sequence (SEQ ID NO:2). The genomic sequence of the ced-3 region was obtained from plasmid pJ107. The introns and the positions of 12 ced-3 mutations are indicated. The likely translation initiation site is indicated by a solid arrowhead. The SL1 splice acceptor of the RNA is boxed. Repetitive elements are indicated as arrows above the relevant sequences. Numbers on the sides

indicate nucleotide positions. Numbers under the a ino acid sequence indicate codon positions.

Figure 4A shows the genomic structure of the ced-3 gene and the location of the mutations. The sizes of the introns and exons are given in bp. The downward arrows indicate the positions of 12 EMS-induced mutations of cec?-3. The arrow pointing right indicates the direction of transcription. The solid arrowhead indicates the translation initiation site. The open arrowhead indicates the termination codon.

Figure 4B shows the locations of the mutations relative to the exons (numbered 1-7) and the encoded serine-rich region in ced-3 .

Figure 5 is a Kyte-Doolittle hydrophobicity plot of the Ced-3 protein.

Figures 6A-B show the alignment of the amino acid sequences of Ced-3 (SEQ ID NO:2) and human interleukin- Iβ convertase (ICE; SEQ ID NO:4). Vertical bars indicate identical amino acids and single and double dots indicate similar amino acids, where double dots signifies closer similarity than a single dot. The serine-rich region and inactivating mutations of Ced-3 are indicated. The active site and autocleavage sites of ICE are indicated. The portions of the Ced-3 protein encoded by the BGAFQ and PBA constructs are also shown. Figure 7 shows the alignment of the amino acid sequences of Ced-3 (SEQ ID NO:2) and murine NEDD-2 (SEQ ID NO:13). Vertical bars and single and double dots signify degrees of similarity as in Figures 6A-B. Inactivating mutations of Ced-3 are shown.

Figure 8A shows the alignment of the amino-terminal regions of the Ced-3 proteins of three nematode species (C. briσgsae , C . elegans , and C. vulgaris) and mouse (SEQ ID NO:14) and human ICEs. A consensus sequence is also shown. Amino acid positions with the same residue

in more than half of the sequences are shaded. Completely conserved amino acids are also boxed.

Figure 8B shows the alignment of carboxyl-terminal regions of the three nematode Ced-3 proteins, human and mouse ICEs, and the mouse NEDD-2 protein. Except for NEDD-2, these sequences are contiguous with the corresponding sequences shown in Figure 8A. A consensus sequence and amino acid conservation are also shown.

Figure 9 shows a comparison of the Ced-3 proteins of C. elegans (line 1; SEQ ID NO:2) and two related nematode species, C. briggsae (line 2; SEQ ID NO:5) and C. vulgaris (line 3; SEQ ID NO:6). The conserved amino acids are indicated by ".". Gaps inserted in the sequence for the purpose of alignment are indicated by "_••.

Figure 10 is the interleukin-lβ convertase cDNA sequence (SEQ ID NO:3).

Figure 11A is a schematic representation of two fusion constructs that can prevent programmed cell death.

Figure 11B is a schematic representation of the lacZ-containing portion of the fusion constructs.

Detailed Description of the Invention

This invention is based on the discovery that the human enzyme interleukin-1/3 convertase (ICE) has significant structural similarity to the protein product of the C. elegans cell death gene, ced-3 . The activities of ced-3 and another cell death gene, ced-4 , have been shown to be required for almost all the cell deaths which occur during the development of the nematode. ICE is a cysteine protease whose physiological significance has been thought to be related to its role in the maturation of one form of interleukin-1 (IL-1) , a major mediator of the immune and

inflammatory response (Fuhlbrigge et al . , in: The Year in Immunology, Cruse and Lewis (eds.), Karger, Basel, 1989, pp. 21-37) . There are two distantly related forms of IL-1, α and β , of which the β form is the predominant species. ICE selectively converts pro-interleukin-1/3 to the active cytokine, IL-13. The production of active L-lβ has been implicated in acute and chronic inflammatory diseases, septic shock, and other physiological processes, including wound healing and resistance to viral infection (Ray et al . , Cell 69 : 591- 604 (1992)). As a result of this discovery, an enzyme which has been known to be involved in the inflammatory response and inflammatory diseases is implicated as having a role in cell death processes. This discovery is consistent with the notion that cell death genes equivalent to the nematode ced-3 gene function in a variety of organisms. The structural similarity between their gene products suggests that the ICE gene is a human equivalent of the ced-3 cell death gene. As further described below, the conservation of certain features of ICE in the Ced-3 protein further suggests that Ced-3 is a protease with a substrate-specificity similar to that of ICE.

Furthermore, the identification of ced-3 and ICE as structurally related genes (i.e., genes whose encoded products, or which themselves, are structurally similar) presents the possibility that a family of structurally related genes exists and provides probes to identify additional members of this ced-3/ICE gene family. Comparison of the genes within this family could indicate functionally important features of the genes or their gene products, and thus, provide information for designing drugs which are useful for treating conditions characterized by cell deaths and/or inflammatory disease.

This discovery provides novel drugs based on the ced-3 , ICE and other ced-3/ICE genes and encoded products that inhibit the production of IL-lj8 and are useful for treatment (preventive and therapeutic) of acute and chronic inflammatory disease, as well as drugs which reduce cell deaths and are useful for treatment of diseases and conditions involving cell deaths (such as myocardial infarction, stroke, traumatic brain injury, viral and other types of pathogenic infection, degenerative diseases, aging, and hair loss) . These drugs may also be useful for treating leukemias in which IL-l/S has been implicated.

Drugs or agents which increase cell deaths can also be developed based on the ced-3 , ICE, and related genes and gene products; such drugs or agents may be useful for killing or incapacitating undesired cell populations (such as cancerous cells, infected cells, cells which produce autoreactive antibodies and hair follicle cells) or undesired organisms (such as pests, parasites, and genetically engineered organisms) . Drugs are also provided which increase IL-13 production and, therefore, the inflammatory and immune response. These drugs may be helpful for providing increased resistance to viral and other types of infection. Also described herein is the discovery that fusion constructs containing amino-terminal portions of the ced-3 gene can inhibit the activity of the intact gene when expressed in otherwise wild-type worms. Due to the similarity between ICE and Ced-3, it is likely that the corresponding amino-terminal portions of the ICE gene will also inhibit the enzymatic activity of ICE in cleaving interleukin-ljS. Thus, novel inhibitors of the ced-3 and ICE genes are provided which may be useful for decreasing cell deaths and/or inflammation involved in various pathologies.

This work has also shown that Ced-3 and the murine NEDD-2 protein are structurally similar. Thus, drugs for increasing or decreasing cell deaths can be developed based on the NEDD-2 gene and its encoded products.

The above-described discoveries, and their implications, and novel drugs and treatments for diseases related to cell death and/or inflammation arising therefrom are described in further detail below. As used herein, the activity of a gene is intended to include the activity of the gene itself and of the encoded products of the gene. Thus, drugs and mutations which affect the activity of a gene include those which affect the expression as well as the function of the encoded RNA and protein. The drugs may interact with the gene or with the RNA or protein encoded by the gene, or may exert their effect more indirectly.

The ced-3 Gene

The C. elegans ced-3 gene was cloned by mapping DNA restriction fragment length polymorphisms (RFLPs) and chromosome walking (Example 1; Figure 1) . The gene was localized to a 7.5 kb fragment of cloned genomic DNA by complementation of the ced-3 mutant phenotype (Figure 2). A 2.8 kb transcript was further identified. The ced-3 transcript was found to be most abundant in embryos, but was also detected in larvae and young adults, suggesting that ced-3 is expressed not only in cells undergoing programmed cell death.

A 2.5 kb cDNA corresponding to the ced-3 mRNA was sequenced. The genomic sequence cloned in the plasmid pJ107 was also determined (Figure 3; SEQ ID N0:1). A comparison with the cDNA sequence revealed that the ced- 3 gene has 7 introns which range in size from 54 to 1195 bp (Figure 4A) . The four largest introns, as well as

sequences 5' of the start codon, contain repetitive elements (Figure 3) , some of which have been previously characterized in non-coding regions of other C. elegans genes such as fem-1 (Spence et al . , Cell 60:981-990 (1990)), lin-12 (J. Yochem, personal communication), and myoD (Krause et al . , Cell 63 : 901-919 (1990)). The transcriptional start site was also mapped (Figure 3) , and a ced-3 transcript was found to be trans-spliced to a C. elegans splice leader, SL1. Twelve EMS-induced ced-3 alleles were also sequenced. Eight of the mutations are missenεe mutations, three are nonsense mutations, and one is a putative splicing mutation (Table 1) . This identification of ced-3 null alleles, together with results of genetic analysis of nematodes homozygous for these null mutations in ced-3 , indicate that, like ced- 4, ced-3 function is not essential to viability. In addition, 10 out of the 12 mutations are clustered in the carboxyl-terminal region of the gene (exons 6-8, Figure 4B) , suggesting that this portion of the encoded protein may be important for activity.

The ced-3 gene encodes a putative protein of 503 amino acids (Figure 3; SEQ ID NO:2) . The protein is very hydrophilic and no significantly hydrophobic region can be found that might be a transmembrane domain

(Figure 5) . One region of the Ced-3 protein is very rich in serine (Figures 6A-B) . Comparison of the C . elegans protein with the Ced-3 proteins of two related nematodes species, C. briggsae and C. vulgariε , shows conservation of the serine-rich feature without conservation of the amino acid sequence in this region (Figure 9; SEQ ID NO:5-6). This suggests that the exact sequence of this serine-rich region may not be important but that the serine-rich feature is. This hypothesis is supported by analysis of ced-3 mutations: none of 12

EMS-induced ced-3 mutations is in the serine-rich region (Figure 4B) . It is possible that the serine-rich region in Ced-3 is another example of semi-specific protein- protein interaction, similar to acid blobs in transcription factors and basic residues in nuclear localization signals. In all these cases, the exact primary sequence is not important.

The serine-rich region may function as a site for post-translational regulation of Ced-3 activity through protein phosphorylation of the serine residues by a

Ser/Thr kinase. McConkey et al . (J . Immunol . 145 : 1221- 1230 (1990)) have shown that phorbol esters, which stimulate protein kinase C, can block the death of cultured thymocytes induced by exposure to Ca ⁺⁺ ionophores or glucocorticoids (Wyllie, Nature 284:555- 556 (1980); Wyllie et al . , J . Path . 142 : 61-11 (1984)). It is possible that protein kinase C may inactivate certain cell death proteins by phosphorylation and, thus, inhibit cell death and promote cell proliferation. Several agents that can elevate cytosolic cAMP levels have been shown to induce thymocyte death, suggesting that protein kinase A may also play a role in mediating thymocyte death. Further evidence suggests that abnormal phosphorylation may play a role in the pathogenesis of certain cell-degenerative diseases. For example, abnormal phosphorylation of the microtubule- associated protein Tau is found in the brains of Alzheimer's disease and Down's syndrome patients (Grundke-Iqbal et al . , Proc . Natl . Acad . Sci . USA 83:4913-4917 (1986); Flament et al . , Brain Res . 516 : 15- 19 (1990)). Thus, it is possible that phosphorylation may have a role in regulating programmed cell death in C. elegans . This is consistent with the fairly high levels of ced-3 and ced-4 transcripts which suggest that

transcriptional regulation alone may be insufficient to regulate programmed cell death.

Structural Relatedness of the ced-3 and Human Interleukin-lβ Convertase Genes and Functional Implications

A search of GenBank, PIR and SWISS-PROT databases using the Blast program (National Center for Biotechnology Information) revealed that human interleukin-lj8 convertase (ICE) has a 28% amino acid identity with the Ced-3 protein (Figures 6A-B) . A comparable level of overall similarity was found between ICE and the Ced-3 proteins from two other nematode species, C. briggsae and C. vulgariε .

The carboxyl-terminal regions of Ced-3 and ICE (amino acids 250-503 and amino acids 166-404, respectively) were found to be more conserved (33% identity) than the amino-terminal portions of the two proteins (22% identity) . A comparison of human and murine ICEs also indicated a high degree of similarity (80% identity) in the carboxyl-terminal region compared with an overall identity of 62% (Cerretti et al . , Science 256 : 91-100 (1992)). Furthermore, deletion analysis of the ICE cDNA sequence has shown that the amino-terminal 119 amino acids of ICE are not required for enzymatic activity, but that deletions of the carboxyl-terminal region eliminate the enzyme's ability to process pro-IL-1S (Cerretti et al . , 1992 supra) . The observation that most of the inactivating mutations of ced-3 cluster in the carboxyl-terminal region (Figure 4B) suggests that the activity of Ced-3 also resides (at least partially) in this region. Thus, the identification of the carboxyl-terminal regions of the two proteins as functional domains and the marked similarity of these regions suggest that the Ced-3 and

ICE proteins have similar activities, i.e., that ICE has cell death activity similar to Ced-3 and Ced-3 has protease activity similar to ICE.

The possiblity that Ced-3 has protease activity is further supported by the observation that the region surrounding the active cysteine and two autocleavage sites of ICE appear to be conserved in the Ced-3 protein. As shown in Figures 6A-B, the five amino acids (QACRG, amino acids 283 to 287) surrounding the active cysteine of ICE (Thornberry et al . , Nature 356 : 168-114 (1992)) are conserved in amino acids 356 to 360 of Ced- 3; this pentapeptide is the longest conserved sequence between ICE and Ced-3. This peptide is also conserved in the Ced-3 proteins of C. briggsae and C. vulgaris (Figure 9). One inactivating mutation of ced-3 , n2433 , introduces a glycine to serine change near the putative active cysteine (Figures 6A-B) . The human ICE gene encodes a precursor enzyme which is autoproteolytically cleaved at two major sites (amino acids 103 and 297) by the active form of the enzyme (Thornberry et al . , 1992 supra) . The Asp-Ser dipeptides of both autocleavage sites are conserved in Ced-3 (at amino acids 131 and 371) (Figures 6A-B) . The conservation of these functionally important sequences strongly suggests that, like ICE, Ced-3 is a cysteine protease with a similar substrate-specificity. Ced-3 would, therefore, be expected to cleave the Ih-lβ precursor, as well as other substrates of ICE.

The possibility that ICE is a cell death gene is consistent with evidence which suggests that the production of active IL-1/3 is involved with cell death processes. Firstly, a variety of studies has suggested that IL-1/3 can prevent cell death (McConkey et al . , J . Biol . Chem . 265 : 3009-3011 (1990); Mangan et al . , J . Immun . 146:1541-1546 (1991); Sakai et al . , J. Exp . Med .

166:1597-1602 (1987); Cozzolino et al . , Proc . Natl . Acad . Sci . USA 86:2369-2373 (1989)). Secondly, active, mature IL-1/3 appears to be released from cells undergoing cell death. Studies on murine macrophages suggest that release of the active form seems not to be merely due to the lysis of the cells or leaking of cell contents. When murine peritoneal macrophages were stimulated with lipopolysaccharide (LPS) and induced to undergo cell death by exposure to extracellular ATP, mature active IL-1/3 was released into the culture supernatant. In contrast, when the cells were injured by scraping, IL-1/3 was released exclusively as the inactive proform (Hogquist et al . , Proc . Natl . Acad . Sci . USA 88:8485-8489 (1991)). The similarity between ICE and Ced-3 strongly supports the hypothesis that ICE is involved in cell death. Since Ced-3 is necessary for cell death, one suggestion is that ICE is also necessary for cell death. It is possible that IL-13 can cause cell death. Alternatively, ICE could produce products besides IL-13, one or more of which can cause cell death. The observation that the ICE transcript is detected in cells that lack IL-ljS expression (Cerretti et al . , 1992 supra ) supports this idea. The finding of a human gene related to the nematode ced-3 gene is consistent with the idea that cell death genes which are structurally related and/or functionally similar to the nematode ced-3 gene exist in a variety of organisms. This idea is supported by evidence that cell deaths occurring in a variety of organisms, including vertebrates and invertebrates, and possibly microbes and plants, as well as cell deaths observed in various developmental and pathologic situations share a common genetic mechanism. Evidence for this hypothesis is discussed in Example 2. The structural relatedness of

ICE suggests that it is a mammalian equivalent of the nematode cell death gene, ced-3 . The cDNA sequence of ICE is shown in Figure 10 (SEQ ID NO:3).

The ced-3/ICE Gene Family and Uses Thereof The ICE and ced-3 genes can be used to isolate additional structurally related genes, including genes from other organisms. Such genes may be identified using probes derived from both the ced-3 and ICE gene sequences and known techniques, including nucleic acid hybridization, polymerase chain reaction amplification of DNA, screening of cDNA or genomic libraries, and antibody screening of expression libraries. The probes can be all or portions of the genes which are specific to the genes, RNA encoded by the genes, degenerate oligonucleotides derived from the sequences of the encoded proteins, and antibodies directed against the encoded proteins. The sequences of the genes and their protein products can also be used to screen DNA and protein databases for structurally similar genes or proteins.

One strategy for detecting structurally related genes in a number of organisms is to initially probe animals which are taxonomically closely related to the source of the probes, for example, probing other worms with a ced-3-derived probe, or probing other mammals with an ICE-derived probe. Closely related species are more likely to possess related genes or gene products which are detected with the probe than more distantly related organisms. Sequences conserved between ced-3 or ICE and these new genes can then be used to identify similar genes from less closely related species. Furthermore, these new genes provide additional sequences with which to probe the molecules of other animals, some of which may share conserved regions with

the new genes or gene products but not with the original probe. This strategy of using structurally related genes in taxonomically closer organisms as stepping stones to genes in more distantly related organisms can be referred to as walking along the taxonomic tree.

Together, ced-3 , ICE, and related genes obtained as described above would comprise a family of structurally related genes, referred to herein as the ced-3/ICE gene family. It is highly likely that at least some of these additional family members would exhibit cell death and/or protease activity. The new genes can be tested for protease activity using known assay methods. For example, the sequence of the protein encoded by a new gene may indicate an active site and substrate- specificity similar to that of ICE, such as observed in Ced-3. This activity can then be verified using the transient expression assays and purified enzyme assays previously described (Cerretti et al . , Science 256 : 91- 100 (1992); Thornberry et al . , Nature 356 -. 16B-114 (1992)). Cell death activity can be tested in bioassays using transgenic nematodes. A candidate cell death gene, such as the ICE gene, can be injected into Ced-3- deficient mutant animals to determine whether the gene complements the ced-3 mutation. Expression libraries can also be screened for cell death genes by this assay. The ced-3, ICE and other related genes which have cell death activity can be used to develop and identify drugs which reduce or increase cell deaths. Drugs which reduce cell deaths are potentially useful for treating diseases and conditions characterized by cell deaths, such as myocardial infarction, stroke, viral and other pathogenic infections (e.g., human immunodeficiency virus) , traumatic brain injury, neural and muscular degenerative diseases, and aging. Drugs which cause cell deaths can be used to control or reduce undesired

cell populations, such as neoplastic growths and other cancerous cells, infected cells, and cells which produce autoreactive antibodies. Undesired organisms, such as pests, parasites, and recombinant organisms, may also be incapacitated or killed by such drugs.

ICE has been implicated in the growth of certain leukemias (Sakai et al . , J . Exp. Med . 166 : 1591 (1987); Cozzolino et al . , Proc . Natl . Acad . Sci . U.S .A . 86 : 2369 (1989); Estrov et al . , Blood 78:1476 (1991); Bradbury et al . , Leukemia 4:44 (1990); Delwel et al . , Blood 74:586 (1989); Ra baldi et al . , Blood 78:3248 (1991)). The observation that the human ICE gene maps to chromosome location llq23, a site frequently involved in DNA rearrangements seen in human cancers (C. Cerretti et al . , Science 256 : 97-100 (1992)), further suggests that ICE is involved in cancer. The finding that ICE probably functions in cell death implies that ICE and other related genes, like ced-3 , may be used to develop drugs to control cancerous growth. In addition, since cell death plays an important role in mammalian hair growth, it seems likely that by controlling cell death, one could cause or prevent hair loss. It has been found that Jcl-2, a human gene which is structurally related to the gene which prevents cell deaths in nematode development (ced-9 ) , is expressed in the hair follicle in a cell-cycle dependent manner. ced-9 has been shown to act by antagonizing the activities of the cell death genes, ced-3 and ced-4 . Together, these findings suggest that genes equivalent to the ced-3 , ced-4 , and ced-9 genes are involved in the physiology of mammalian hair growth and loss.

Drugs which increase cell deaths may comprise ced-3 , ICE, and other ced-3/ICE family members, their RNA and protein products, constitutively activated mutants of the genes and encoded products, and peptide

and non-peptide mimetics of the proteins. Drugs which decrease cell deaths may comprise antisense RNA complementary to the mRNA of a cell death gene, or mutant cell death genes or encoded products, that no longer cause cell death and interfere with the function of wild-type genes. Furthermore, drugs comprising agonists and antagonists of the cell death genes can be designed or identified using the genes or their gene products as targets in bioassays. The bioassays can be conducted in wild-type, mutant, or transgenic nematodes, in which an alteration in programmed cell deaths is an indicator of an effective agonist or antagonist. Bioassays can also be performed in cultured cells transfected with the target cell death gene, into which the substance being tested is introduced. In bioassays for antagonists of cell death, the cultured cells should be put under conditions which induce the activity of the target cell death gene.

Uses of bioassays utilizing C. elegans are exemplified by the following:

1) use of normal, wild-type nematodes to screen for drugs or genes that inactivate ced-3 and hence, prevent programmed cell deaths;

2) use of normal, wild-type nematodes to screen for drugs or genes that activate ced-3 and hence, cause excess cell deaths;

3) use of mutant nematodes which overexpress ced-3 or which express a constitutively activated ced-3 gene to identify drugs or genes that prevent excess cell deaths caused by the ced-3 mutation;

4) use of mutant nematodes which underexpress σed- 3 or which express an inactivated ced-3 gene to identify drugs or genes that mimic or complement the ced-3 mutation;

5) use of transgenic nematodes (with an inactivated endogenous ced-3 gene) in which either a wild-type or mutant form of ICE or other ced— 3/ICE family member causes excess cell deaths to identify drugs or genes which antagonize the activity of the transgene; and

6) use of transgenic nematodes which carry a transgene that inhibits cell death (e.g., a transgene that expresses an inhibitory fragment of ced-3 , ICE or related gene, as described below) to identify drugs that overcome this inhibition and cause cell death. Drugs can be introduced into nematodes by diffusion, ingestion, microinjection, shooting with a particle gun or other methods. They can be obtained from traditional sources such as extracts (e.g., bacterial, fungal or plant) and compound libraries, or can be provided by newer methods of rationale drug design. Information on functionally important regions of the genes or gene products, gained by sequence comparisons and/or mutational analysis may provide a basis for drug design. Genes can be microinjected into nematodes to produce transgenic nematodes. Individual genes or cDNA and genomic DNA libraries can be screened in this manner. Agonists and antagonists may also be derived from genes which are not cell death genes, but which interact with, regulate or bypass cell death genes. Such interacting genes may be tested by the bioassays mentioned above, as well as by in vivo genetics in nematodes. In this latter method, interacting genes are identified as secondary mutations which suppress or enhance the ced-3 mutation. The sequences of these interacting* genes can then be used to identify structurally related interacting genes in other organisms.

Similarly, anti-inflammatory drugs may be developed or identified using ced-3 , ICE and other family members and their encoded products. Drugs which enhance ICE activity may also be useful for boosting the inflammatory response to viral and other infections. In addition, the availability of a number of structurally related genes makes it possible to carry out structural comparisons. Conserved regions or features of the genes or their encoded products are likely to be functionally significant for cell death and/or protease activity. This information could be helpful in designing or selecting drugs which would mimic or affect the activity of the genes.

Moreover, conservation of functional domains among ced-3/ICE family members or their encoded products suggests not only that these genes have similar activities, but that they and their encoded products function via similar mechanisms. This suggests that mutations in conserved regions, mimetics based on conserved regions, and agonists and antagonists which affect the function of conserved regions of one ced- 3/ICE gene or encoded protein will similarly affect other genes or encoded proteins in the family. This is the rationale behind the use of Ced-3 inhibitors to inhibit ICE and inflammation, and the use of anti- inflammatory drugs which act by inhibiting ICE to inhibit the ced-3 gene and reduce cell deaths (described further below) .

Furthermore, drugs which affect the cell death and/or inflammatory activities of the ced-3 and ICE genes may also affect other as yet undiscovered activities of these genes. The biology of IL-13 and ICE is only incompletely understood at the present time, and it is very likely that other functions of both IL-13 and ICE.may be discovered. These may include new activities

or new physiological processes or diseases in which the respective cytokinetic and proteolytic activities of these molecules are involved. In either case, drugs (such as inhibitory protein portions) which affect ICE activity are likely to affect the new activities and processes, as well.

In addition, mutations and drugs which alter or mimic the activity of one member of the ced-3/ICE family can be engineered based on what is known about mutations and drugs affecting another family member with which it shares a conserved region. Mutations in conserved regions which correspond to those found in another family member could be used to produce similar effects. For example, five out of nine inactivating point mutations analyzed in ced-3 were found to result in alterations of amino acids which are conserved between ICE and Ced-3 (Figures 6A-B) . Amino acid substitutions in ICE corresponding to those in Ced-3 are also expected to result in inactivation. The inhibitory amino- terminal gene portions and constitutively activated carboxyl-terminal gene portions described below are further examples of corresponding mutations which can be made in genes of the ced-3/ICE family.

Comparison of Ced-3, ICE, and related proteins may also provide insights into the substrate-specificity of ICE and related enzymes. Previous studies on ICE have not identified a consistent consensus cleavage site. A comparison of the Ced-3 and ICE autocleavage sites, together with the cleavage site of pro-IL-1/3, reveals that cleavage always occurs after an Asp residue. For this reason, it is likely that Ced-3, ICE, and related proteins are proteases which cleave after some aspartate residues or, perhaps at lower efficiencies, all aspartate residues.

A further use of ced-3/ICE family members is to provide diagnostic probes (DNA, RNA, oligonucleotides and antibodies) for diseases involving cell deaths and inflammation in humans and other organisms. It is likely that such diseases are associated with abnormalities in ced-3/ICE genes and their gene products. The probes can be used to detect abnormalities in the sequence, level and/or activities of the genes and encoded RNA and protein products. The diseases may be genetic, in which case, the probes may be used in patient and pre-natal testing, or non- genetic, in which case, RNAs and proteins may be examined. In particular, the finding that ICE is a putative cell death gene makes this gene and its derivative molecules potentially useful as diagnostic probes for diseases characterized by cell deaths. Similarly, ced-3 and its derivative molecules are potentially useful for detecting abnormalities in pathologies in which inflammation is evident. The usefulness of these probes may be multiplied as more genes with known physiological functions are found to be structurally related to ced-3 and ICE.

Structural Relatedness of ced-3 and the Murine NEDD-2 Gene * Database searches also revealed that another mammalian protein is similar to the Ced-3 protein (Figure 7) . The murine NEDD-2 protein has 27% amino acid identity and 55% similarity to a carboxyl-terminal portion of Ced-3. The NEDD-2 protein is expressed in the brain of mouse embryos and much less in the murine adult brain; the protein is thought to be involved in the development of the murine central nervous system (Kumar et al . , Biochem . Biophys . Res . Comm . 185(3) :1155- 1161 (1992)). The structural similarity between the

NEDD-2 and ced-3 gene products suggests that the NEDD-2 gene is also involved in cell death processes which occur during development, and further supports the hypothesis that genes which are structurally and functionally related to the nematode ced-3 gene function in a variety of organisms. Interestingly, the NEDD-2 amino acid sequence is not significantly similar to that of human ICE.

The similarity of the amino acid sequences of Ced-3 and NEDD-2 further suggests that mutations of the NEDD-2 gene which produce alterations in the protein corresponding to alterations in Ced-3 resulting from the mutations, nll29 , nll64 , n2426 and nll63 (see Figure 1 ) , will inactivate the NEDD-2 gene. This invention includes all and portions of the

NEDD-2 gene, mutated NEDD-2 genes corresponding to known ced-3 mutations, RNAs and proteins encoded by the wild- type and mutated genes, and mimetics and other drugs derived from these genes and gene products, which are useful for controlling cell death.

Figures 8A and 8B show alignments of the amino- terminal and carboxyl-terminal regions, respectively, of the Ced-3 proteins of the three nematode species (C. briggεae , C. elegans , and C. vulgariε) , the human and murine ICEs and the murine NEDD-2 protein (in 8B only) . As shown in these figures (boxed portions) , a number of amino acids are completely conserved among these structurally related proteins, and thus, are likely to be important functionally. Mutations of these sites would be expected to alter the activity of the genes.

Inhibitory Portions of the ced-3 Gene

Fusion constructs containing portions of the ced-3 gene were found to prevent programmed cell death when expressed in wild-type C . elegans . These constructs are

represented schematically in Figure 11A. The BGAFQ construct contains a portion of the ced-3 gene fused 5' of the E. coli lacZ gene and another ced-3 portion fused 3' of lacZ. The 5 ' ced-3 portion is the genomic sequence from a BamHI site located about 300 base pairs upstream of nucleotide 1 of the sequence shown in Figure 3 to a Sail site at nucleotide 5850. This portion spans sequences 5' of the SLl acceptor site (nucleotide 2161) to include the 372 codons of the amino-terminal region. The 3' ced-3 portion of BGAFQ is the genomic sequence from a NotI site at nucleotide 5927 in the ced-3 gene to an Apal site located about 1.5 kb downstream of nucleotide 7653 of the sequence in Figure 3. This portion contains the carboxyl-terminal codons from 398 to the end (codon 503) and 3' untranslated sequences.

The PBA construct has a smaller portion of the ced- 3 gene which is the genomic sequence from the same BatnHI site as in BGAFQ to a Bgrlll site at nucleotide 3020 (Figure 11A) fused 5' of the lacZ gene. This ced-3 portion spans sequences 5' of the SLl acceptor site to include the first 149 codons of the amino-terminal region.

Both constructs were made using the pBluescript vector (Stratagene) and fragments containing the lacZ construct from the pPD vectors of Fire (EMBO J . 5:2673- 2690 (1986) ) . The lacZ-containing portion has the entire lacZ coding sequence except for the first 11 codons. In addition, there is a synthetic intron and a nuclear localization signal upstream of the lacZ gene and a fragment of the 3' end of the unc-54 gene downstream of the lacZ gene (Figure 11B) . Construct PBA was made by inserting a BamHI-Apal fragment containing the lacZ construct shown in Figure 11B from Andy Fire's vector, pPD22.04, into the Bglll-Apal fragment of the ced-3-containing plasmid, pJ40. Construct BGAFQ was

made by inserting a Sall-EagI fragment containing the same lacZ construct from pPD22.04 into the Sall-Notl fragment of pJ40A, which is pJ40 without the NotI site in the vector. Table 2 shows the results of injecting wild-type nematodes with the two constructs. These results indicate that the BGAFQ and PBA fusion constructs pre¬ vent cell deaths which normally occur in the development of the nematodes. These fusion constructs were further observed to prevent cell deaths and the apparently associated inviability caused by a loss-of-function mutation in ced-9 , a gene which functions to keep certain cells from dying during nematode development, and which has been shown to act by antagonizing ced-3 and second cell death gene, ced-4.

Both constructs express /3-galactosidase activity in wild-type nematodes. Since the pBluescript vector does not co _* ntain eukaryotic transcriptional or translational start sites, these signals are probably supplied by the ced-3 gene portions fused 5' of lacZ . Furthermore, since the PBA construct works to prevent cell death, it seems that the ced-3 portion in BGAFQ needed for inhibition is the portion fused upstream of lacZ (as opposed to the portion located downstream of lacZ) . Presumably, only the region from the BamEI site to nucleotide 3020 is needed in BFAGQ, since this is all that is contained in PBA.

A construct that contains the PBA ced-3 portion but not any of the lacZ portion did not prevent cell death, suggesting that fusion to portions of lacZ is needed for expression or action of the inhibitory gene portion.

These observations indicate that the amino-terminal portion of the Ced-3 protein, possibly in conjunction with a portion of E . coli /3-galactosidase, can act to prevent programmed cell deaths in C. elegans . One

plausible mechanism is that this portion of the Ced-3 protein acts in a dominant negative or antimorphic fashion, to prevent the activity of the normal Ced-3 protein. (It is known that inactivation of the Ced-3 protein results in an absence of programmed cell deaths.) Such dominant negative activity could be a result of the partial Ced-3 protein binding to and, thereby, inactivating the normal Ced-3 protein; consistent with this model is the finding that the active form of the structurally similar ICE protein is dimeric. Alternatively, the partial Ced-3 protein may bind to a molecule with which the normal Ced-3 protein must interact to function and by preventing this interaction, inhibits Ced-3 activity. Due to the structural similarity of ICE to the Ced- 3 protein, fusion constructs encoding amino-terminal portions of ICE would also be expected to inhibit the activity of the ced-3 gene. In particular, those portions of the ICE gene corresponding to the ced-3 gene portions in BGAFQ and PBA, i.e., ICE codons 1 to 298 and codons 1 to 111, or active subportions of these, are expected to inhibit ced-3 . A further extension of this reasoning suggests that corresponding gene portions of any structurally related ced-3/ICE family member would also have an inhibitory effect on ced-3 activity.

Furthermore, the structural relatedness of the ced- 3 and ICE genes implies that the ICE enzyme could also be inhibited by fusion constructs containing amino- terminal portions of the ICE gene, as well as corresponding portions of other structurally related genes, such as ced-3 .

Identification of portions of the ced-3 , ICE, and related genes which inhibit the ced-3 gene can be carried out by testing expression constructs containing these gene portions or their encoded products in

bioassays for cell death activity. Identification of gene portions or encoded products which inhibit ICE can be carried out using previously described assays for ICE activity. For example: 1) wild-type worms can be injected with portions of the ced-3 or other structurally related gene, such as ICE, to determine if they prevent programmed cell death; 2) portions of the ICE protein or other structurally similar protein, such as Ced-3, can be co—expressed with ICE and pro-IL-1/3 in nematodes or cultured mammalian cells to see if they inhibit ICE-catalyzed cleavage of the IL-1/3 precursor; and 3) peptides or nucleic acids containing portions of the amino acid or coding sequence of ICE or similar protein, such as Ced-3, can be tested using purified ICE and synthetic substrates.

Inhibitory portions of the ced-3 gene, ICE, and structurally related genes, their encoded RNAs and proteins, and peptide and non-peptide mimetics of the proteins may be used to reduce cell deaths and/or inflammation, and are, thus, useful for the treatment of diseases involving these processes. The encoded proteins and peptide and non-peptide mimetics can be delivered by various known methods and routes of drug delivery. For example, they can be administered orally or by another parenteral route or by a non-parenteral route (e.g., by injection intramuscularly, intraperitoneally or intravenously or by topical administration) . Alternatively, expression constructs containing the gene portions can be made using heterologous transcriptional-and translational signals or signals native to the gene portions. The constructs can be delivered into cells by various methods of gene therapy, such as retroviral infection.

Interestingly, those ICE gene portions corresponding to the ced-3 portions of BGAFQ and PBA

encode approximately the protein fragments which result from cleavage at each of the two autocleavage sites (amino acids 103 and 297) . This observation suggests that autoproteolytic conversion of the proenzyme to active ICE involves cleaving off the inhibitory amino- terminal portions of the protein. Active ICE is a heterodimer composed of subunits of about 20 and 10 kilodaltons (Thornberry et al . , Nature 356: 168-114 (1992)). These subunits have been shown to be derived from the ICE proenzyme and correspond to amino acids 120 to 297 (p20) and 317 to 404 (plO) . Kinetic studies suggest that association of the two subunits is required for activity of the enzyme. It is possible that the amino-terminal region of the protein interferes with this association.

This implies that mutant proteins in which the inhibitory amino-terminal regions are deleted may be constitutively activated. Thus, carboxyl-terminal portions of the Ced-3, ICE, and related proteins, and constructs and RNAs expressing these portions, are potentially useful for increasing cell deaths and/or IL- 1/3 production. Constructs which may be used include those which express the carboxyl region of ICE, which encodes the two subunits of the active enzyme, as well as those which express each of these subunits separately. In addition, it is possible that the amino region of ICE, which is not needed for ICE enzymatic activity in vitro, is important for ICE activity or the regulation of ICE activity in vivo . Consistent with this idea is the finding that two of the ced-3 mutations map in this region. For this reason, a construct which expresses the amino region of Ced-3, ICE or a Ced-3/ICE gene family member may also be used. Furthermore, the NEDD-2 protein, which is similar to a carboxyl-terminal portion of the Ced-3 portion, may also exhibit

constitutive activity in causing cell deaths. Thus, all or active portions of NEDD-2, and DNA and RNA encoding NEDD-2 proteins, would be expected to produce cell death activity when expressed. Drugs comprising activated molecules derived from the carboxyl-terminal regions of Ced-3, ICE and other proteins of the Ced-3/ICE family and from the NEDD-2 protein, DNAs and RNAs encoding these proteins and protein fragments, as well as peptide and non-peptide mimetics, are potentially useful for controlling or reducing the size of undesirable cell populations, such as cancerous cells, infected cells, cells producing autoreactive antibodies and hair follicle cells. Such drugs may also be useful for incapacitating or killing undesired organisms, such as parasites, pests, and genetically engineered organisms. For example, a number of nematodes are human, animal and plant parasites.

ICE Inhibitors As Inhibitors of Cell Death

The conservation of the active site of ICE (active cysteine and surrounding amino acids) in the Ced-3 protein implies that Ced-3 is a cysteine protease which interacts with its substrate by a similar mechanism. Hence, it is likely that inhibitors of ICE which interfere with this mechanism, or chemical analogs of these inhibitors, will also inhibit Ced-3 function.

Peptide aldehydes containing the ICE recognition site:

P4—P3—P2—PI Tyr-Val-Ala-Asp or a substituted site in which P2 is Ala, His, Gin, Lys, Phe, Cha, or Asp, have been shown to be effective, specific, and reversible inhibitors of the protease activity of ICE (Thornberry et al . , Nature 356 : 168-114

(1992)). These molecules are thought to act as transition analogs, which compete for ICE binding to its substrate, pro-IL-13. Three such inhibitors have been described: Inhibitor B (Ac-Tyr-Val-Ala-Asp-CHO) ; Inhibitor C (Ac-Tyr-D-Ala-Ala-Asp-CHO) ; and Inhibitor D (Ac-Tyr-Val-Lys-Asp-CHO) . Of these, Inhibitor B is the most potent, with a K—0.76 nM compared to K*=3 nM for D and K—1.5 μM for C.

In addition, the crmA gene of cowpox virus has been found to encode a serpin which specifically inhibits ICE (Ray et al . , Cell 69 : 591-604 (1992)). The serpin acts by preventing the proteolytic activation of ICE. This inhibitor of ICE is also expected to inhibit structurally similar proteins, such as Ced-3. The crmA gene and methods for obtaining purified CrmA protein have been described (Pickup et al . , Proc . Natl . Acad . Sci . USA 83:7698-7702 (1986); Ray et al . , 1992 supra) . This invention includes the use of inhibitors of ICE, such as peptide aldehydes, and particularly inhibitor B, and the CrmA protein, as drugs for decreasing the activity of cell death genes and, thus, for treatment of diseases characterized by cell deaths.

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims. For example, functional equivalents of DNAs and RNAs may be nucleic acid sequences which, through the degeneracy of the genetic code, encode the same proteins as those specifically claimed. Functional equivalents of proteins may be substituted or modified amino acid sequences, wherein the substitution or modification does not change the activity or function of the protein. A

"silent" amino acid substitution, such that a chemically similar amino acid (e.g., an acidic amino acid with another acidic amino acid) is substituted, is an example of how a functional equivalent of a protein can be produced. Functional equivalents of nucleic acids or proteins may also be produced by deletion of nonessential sequences.

The following examples illustrate the invention and are not intended to be limiting in any way.

EXAMPLE 1

CLONING. SEQUENCING. AND CHARACTERIZATION OF

THE CED-3 GENE

MATERIALS AND METHODS

General Methods and Strains The techniques used for the culturing of C. elegans were as described by Brenner (Genetics 77:71-94 (1974)). All strains were grown at 20°C. The wild- type parent strains were C . elegans variety Bristol strain N2, Bergerac strain EM1002 (Emmons et al . , Cell 32:55-65 (1983)), C. briggsae and C. vulgaris (obtained from V. Ambros) . The genetic markers used are described below. These markers have been described by Brenner (1974 supra) , and Hodgkin et al . (In: The Nematode Caenorhabditis elegans , Wood and the Community of C. elegans Researchers (eds.), Cold Spring Harbor

Laboratory, 1988, pp 491-584). Genetic nomenclature follows the standard system (Horvitz et al . , Mol . Gen . Genet . 175 : 129-133 (1979)): LG I: ced-l (el375) ; unc-54 (r323) LG VI: unc-31 (e928) , unc-30 (e!91) , ced-3 (n717, n718, n!040, n!129, nl!63, n!164, nll65, nl286,

nl949, n2426, n2430, n2433) , unc-26 (e205) , dpy-4 (ell66) LG V: egl-l (n986); unc-76 (e911) LG X: <3py-3 (e27)

Isolation of Additional Alleles of ced-3

A non-complementation screen was designed to isolate new alleles of ced-3. Because animals heterozygous for ced-3 (n717) in trans to a deficiency are viable (Ellis and Horvitz, Cell 44:817-829 (1986)), animals carrying a complete loss-of-function ced-3 allele generated by mutagenesis were expected to be viable in trans to ced-3 (n717) , even if the new allele was inviable in homozygotes. Fourteen EMS mutagenized egl-1 males were mated with ced-3 (n717) unc-26 (e205) ; egl-l (n487) ; dpy-3 (e27) hermaphrodites. egl-1 was used as a marker in this screen. Dominant mutations in egl-1 cause the two hermaphrodite specific neurons, the HSNs, to undergo programmed cell death (Trent et al . , Genetics 104 : 619-641 (1983)). The HSNs are required for normal egg-laying, and egl-1(n986) hermaphrodites, which lack HSNs, are egg-laying defective (Trent et al . , 1983 supra) . The mutant phenotype of egl-1 is suppressed in a ced-3 ; egl-1 strain because mutations in ced-3 block programmed cell deaths, egl-1 males were mutagenized with EMS and crossed with ced-3 (n717) , unc-26 (e205) ; egl-1 (n487) ; dpy-3 (e27) . Most cross progeny were egg- laying defective because they were heterozygous for ced- 3 and homozygous for egl-1 . Rare egg-laying competent animals were picked as candidates for carrying new alleles of ced-3 . Four such animals were isolated from about 10,000 Fl cross progeny of EMS-mutagenized animals. These new mutations were made homozygous to confirm that they carried recessive mutations of ced-3 .

Molecular Biology

Standard techniques of molecular biology were used (Maniatis et al . , Molecular Cloning: A Laboratory Manual , Cold Spring Harbor Laboratory, 1983). Two cosmid libraries were used extensively in this work: a Sau3AI partial digest genomic library of 7000 clones in the vector pHC79 and a Sau3AI partial digest genomic library of 6000 clones in the vector pJB8 (Ish- Horowicz and Burke, Nucleic Acids Res . 9:2989 (1981)). The "right" end of MMM-Cl was cloned by cutting it with Hindlll and self-ligating. The "left" end of MMM- Cl was cloned by cutting it with BglII or Sail and self- ligating.

The "right" end of Jc8 was made by digesting Jc8 with .EcoRI and self-ligating. The "left" end of Jc8 was made by digesting Jc8 by Sa l and self-ligating.

C. elegans RNA was extracted using guanidine isothiocyanate (Kim and Horvitz, Genes & Dev . 4:357-371 (1990)). Poly(A) ⁺ RNA was selected from total RNA by a poly(dT) column (Maniatis et al., 1983 supra) . To prepare stage-synchronized animals, worms were synchronized at different developmental stages (Meyer and Casson, Genetics 106 :29-44 (1986)).

For DNA sequencing, serial deletions were made according to a procedure developed by Henikoff (Gene

28:351-359 (1984)). DNA sequences were determined using Sequenase and protocols obtained from US Biochemicals with minor modifications.

The Tel DNA probe for Southern blots was pCe2001, which contains a Bergerac Tel element (E mons et al . , Cell 32:55-65 (1983)). Enzymes were purchased from New England Biolabs, and radioactive nucleotides were from Amersham.

Primer extension procedures followed the pro- by Robert E. Kingston (In: Current Protocols in Molecular Biology, Ausubel et al . (eds.), Greene Publishing Associates and Wiley-Interscience, Ne p. 4.8.1) with minor modifications.

Polymerase chain reaction (PCR) was carried o : using standard protocols supplied by the GeneAmp Kit (Perkin Elmer) . The primers used for primer extension and PCR are as follows:

Pex2: 5' TCATCGACTTTTAGATGACTAGAGAACATC 3' (SEQ ID NO:7) ; Pexl: 5' GTTGCACTGCTTTCACGATCTCCCGTCTCT 3'

(SEQ ID NO:8) ; SLl: 5' GTTTAATTACCCAAGTTTGAG 3' (SEQ ID NO:9); SL2: 5' GGTTTTAACCAGTTACTCAAG 3' (SEQ ID NO:10);

Log5: 5' CCGGTGACATTGGACACTC 3' (SEQ ID NO:11); and OligolO: 5' ACTATTCAACACTTG 3' (SEQ ID NO:12).

Germline Transformation

The procedure for microinjection basically follows that of A. Fire (EMBO J. 5:2673-2680 (1986)) with modifications: Cosmid DNA was twice purified by CsCl- gradient. Miniprep DNA was used when deleted cosmids were injected. To prepare miniprep DNA, DNA from 1.5 ml overnight bacterial culture in superbroth (12 g Bacto- tryptone, 24 g yeast extract, 8 ml 50% glycerol, 900 ml H ₂ 0, autoclaved; after autoclaving, 100 ml 0.17 M KH ₂ P0 ₄ and 0.72 M KH ₂ P0 ₄ were added) was extracted by alkaline lysis method as described in Maniatis et al. (1983 supra) . DNA was treated with RNaεe A (37°, 30 minutes) and then with protease K (55°, 30 minutes), extracted with phenol and then chloroform, precipitated twice (first in 0.3 M sodium acetate and second in 0.1 M

potassium acetate, pH 7.2), and resuspended in 5 μl injection buffer as described by A. Fire (1986 supra) . The DNA concentration for injection is in the range of .00 ug to 1 mg per ml. All transformation experiments used ced-l (el 735) ; unc-31 (e928) ced-3 (n717) strain, unc-31 was used as a marker for co-transformation (Kim and Horvitz, 1990 supra) . ced-1 was present to facilitate scoring of the Ced-3 phenotype. The mutations in ced-1 block the engulfment process of cell death, which makes the corpses of the dead cells persist much longer than in wild-type animals (Hedgecock et al . , Science 220 : 1211- 1280 (1983)). The Ced-3 phenotype was scored as the number of dead cells present in the head of young LI animals. The cosmid C10D8 or the plasmid subclones of C10D8 were mixed with C14G10 (unc-31(+)-containing) at a ratio of 2:1 or 3:1 to increase the chances that a Unc- 31(+) transformant would contain the cosmid or plasmid being tested as well. Usually, 20-30 animals were injected in one experiment. Non-Unc Fl progeny of the injected animal were isolated three to four days later. About 1/2 to 1/3 of the non-Unc progeny transmitted the non-Unc phenotype to F2 progeny and established a transformant line. The young LI progeny of such non-Unc transformant were checked for the number of dead cells present in the head using Nomarski optics.

RESULTS

Isolation of Additional ced-3 Alleles

All of the ced-3 alleles that existed previously were isolated in screens designed to detect viable mutants displaying the Ced phenotype (Ellis and Horvitz, Cell 44 : 811-829 (1986)). Such screens may have systematically missed. any class of ced-3 mutations that

is inviable as homozygoteε. For this reason, a scheme was designed that could isolate recessive lethal alleles of ced-3 . Four new alleles of ced-3 (nll63 , nll64, nll65, nl286) were isolated in this way. Since new alleles were isolated at a frequency of about 1 in 2500, close to the frequency expected for the generation of null mutations by EMS in an average C. elegans gene (Brenner, Genetics 77:71-94 (1974) ; Greenwald and Horvitz, Genetics 96:147-160 (1980)), and all four alleles are homozygous viable, it was concluded that the null allele of ced-3 is viable.

Mapping RFLPs near ced-3

Tel is a C . elegans transposable element that is thought to be immobile in the common laboratory Bristol strain and in the Bergerac strain (Emmons et al . , Cell 32:55-65 (1983)). In the Bristol strain, there are 30 copies of Tel, while in the Bergerac strain, there are more than 400 copies of Tel (Emmons et al., 1983 supra ; Finney, Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, Massachusetts, 1987) . Because the size of the C. elegans genome is small (haploid genome size 8 x 10 ⁷ bp) (Sulston and Brenner, Genetics 77:95-104 (1976)), a polymorphism due to Tel between the Bristol and Bergerac strains would be expected to occur about once every 200 kb. Restriction fragment length polymorphisms (RFLPs) can be used as genetic markers and mapped in a manner identical to conventional mutant phenotypes. A general scheme has been designed to map Tel elements that are dimorphic between the Bristol and Bergerac strains near any gene of interest (Ruvkun et al . , Genetics 121:501-516 (1989)). Once tight linkage of a particular Tel to a gene of interest has been

established, that Tel can be cloned and used to initiate chromosome walking.

A 5.1 kb Bristol-specific Tel .EcoRI fragment was tentatively identified as containing the Tel closest to ced-3. This Tel fragment was cloned using cosmids from a set of Tcl-containing C. elegans Bristol genomic DNA fragments. DNA was prepared from 46 such Tcl-containing cosmids and screened using Southern blots to identify the cosmids that contain a 5.1 kb .EcoRI Tcl-containing fragment. Two such cosmids were identified: MMM-Cl and MMM-C9. The 5.1 kb EcoRI fragment was subcloned from MMM-Cl into pUC13 (Promega) . Since both ends of Tel contain an .EcoRV site (Rosenzweig et al . , Nucleic Acids Res . 11:4201-4209 (1983)), .EcoRV was used to remove Tel from the 5.1 kb .EcoRI fragment, generating a plasmid that contains only the unique flanking region of this Tcl-containing fragment. This plasmid was then used to map the specific Tel without the interference of other Tel elements. unc-30 (el91) ced-3 (n717) dpy-4 (el!66)/+++ males were crossed with Bergerac (EM1002) hermaphrodites, and Unc non-Dpy or Dpy non-Unc recombinants were picked from among the F2 progeny. The recombinants were allowed to self-fertilize, and strains that were homozygous for either unc-30 (el91) dpy-4 (Bergerac) or unc-30 (Bergerac) dpy-4 (ell66) were isolated. After identifying the ced genotypes of these recombinant strains, DNA was prepared from these strains. A Southern blot of DNA from these recombinants was probed with the flanking sequence of the 5.1 kb .EcoRI Tel fragment. This probe detects a 5.1 kb fragment in Bristol N2 and a 3.4 kb fragment in Bergerac. Five out of five unc-30 ced-3 dpy (+Berg) recombinants, and one of one unc-30 (+Berg) ced-3 dpy-4 recombinants showed the Bristol pattern. Nine of ten unc-30 (+Be£-g) dpy-4 recombinants showed the Bergerac

pattern. Only one recombinant of unc-30 (+Berg) dpy-4 resulted from a cross-over between ced-3 and the 5.1 kb Tel element. The genetic distance between ced-3 and dpy-4 is 2 map units (mu) . Thus, this Tel element is located 0.1 mu on the right side of ced-3.

Cosmids MMM-Cl and MMM-C9 were used to test whether any previously mapped genomic DNA cosmids overlapped with these two cosmids. A contig of overlapping cosmids was identified that extended the cloned region near ced- 3 in one direction.

To orient MMM-Cl with respect to this contig, both ends of MMM-Cl were subcloned and these subclones were used to probe the nearest neighboring cosmid C48D1. The "right" end of MMM-Cl does not hybridize to C48D1, while the "left" end does. Therefore, the "right" end of MMM- Cl extends further away from the contig. To extend this contig, the "right" end of MMM-Cl was used to probe the filters of two cosmid libraries (Coulson et al . , Proc . Natl . Acad . Sci . USA 83:7821-7825 (1986)). One clone, Jc8, was found to extend MMM-Cl in the opposite direction of the contig.

RFLPs between the Bergerac and Bristol strains were used to orient the contig with respect to the genetic map. Bristol (N2) and Bergerac (EM1002) DNA was digested with various restriction enzymes and probed with different cosmids to look for RFLPs. Once such an RFLP was found, DNA from recombinants of the Bristol and Bergerac strains between ced-3 and unc-26 , and between unc-30 and ced-3 was used to determine the position of the RFLP with respect to ced-3 .

The "right" end of Jc8, which represents one end of the contig, detects an RFLP (nP33) when N2 and EM1002 DNA was digested with Hindlll. A Southern blot of DNA from recombinants between three ced-3 (+Berg) unc-26 was probed with the "right" end of Jc8. Three of three

+Berg unc-26 recombinants showed the Bristol pattern, while two of two ced-3 unc-26 (+Berg) recombinants showed the Bergerac pattern. Thus, nP33 mapped very close or to the right side of unc-26. The "left" end of Jc8 also detects a Hindlll RFLP (nP34) . The same Southern blot was reprobed with the Jc8 "left" end. Two of the two ced-3 unc-26 (+Berg) recombinants and two of the three ced-3 (+Berg) unc-26 recombinants showed the Bergerac pattern. One of the three ced-3 (+Berg) unc-26 recombinants showed the

Bristol pattern. The genetic distance between ced-3 and unc-26 is 0.2 mu. Thus, nP34 was mapped between ced-3 and unc-26 , about 0.1 mu on the right side of ced-3 . The flanking sequence of the 5.1 kb EcoRI Tel fragment (named nP35) was used to probe the same set of recombinants. Two of three ced-3 (+Berg) unc-26 recombinants and two of two ced-3 unc-26 (+Berg) recombinants showed the Bristol pattern. Thus, nP35 was also found to be located between ced-3 and unc-26 , about 0.1 mu on the right side of ced-3 .

A similar analysis using cosmid T10H5 which contains the Hindlll RFLP (nP36) , and cosmid B0564, which contains a ffindlll RFLP (nP37) , showed that nP36 and nP37 mapped very close or to the right of unc-30. These experiments localized the ced-3 gene to an interval of three cosmids. The positions of the RFLPs, and of ced-3, unc-30 and unc-26 on chromosome IV, and their relationships to the cosmids are shown in Figure 1. It has been demonstrated by microinjection that cosmids C37G8 and C33F2 carry the unc-30 gene (John Sulston, personal communication) . Thus, the region containing the ced-3 gene was limited to an interval of two cosmids. These results are summarized in Figure l.

Complementation of ced-3 by Ger line Transformation

Cosmids that were candidates for containing the ced-3 gene were microinjected into a ced-3 mutant to see if they rescue the mutant phenotype. The procedure for microinjection was that of A. Fire (EMBO J. 5:2673-2680 (1986)) with modifications, unc-31 , a mutant defective in locomotion, was used as a marker for cotransformation (Kim and Horvitz, Genes & Dev. 4:357-371 (1990)), because the phenotype of ced-3 can be examined only by using Nomarski optics. Cosmid C14G10 (containing unc- 31 (+) ) and a candidate cosmid were coinjected into ced- l (el375) ; unc-31 (e928) ced-3 (n717) hermaphrodites, and Fl non-Unc transformants were isolated to see if the non-Unc phenotype could be transmitted and established as a line of transformants. Young LI progeny of such transformants were examined for the presence of cell deaths using Nomarski optics to see whether the Ced-3 phenotype was suppressed. Cosmid C14G10 containing unc- 31 alone does not rescue ced-3 activity when injected into a ced-3 mutant. Table 4 summarizes the results of these transformation experiments.

As shown in Table 3, of the three cosmids injected (C43C9, W07H6 and C48D1) , only C48D1 rescued the Ced-3 phenotype (2/2 non-Unc transformants rescued the Ced-3 phenotype) . One of the transformants, nEX2 , appears to be rescued by an extra-chromosomal array of injected cosmids (Way and Chalfie, Cell 54:5-16 (1988)), which is maintained as an unstable duplication, since only 50% of the progeny of a non-Unc Ced(+) animal are non-Unc Ced(+) . Since the non-Unc Ced(-t-) phenotype of the other transformant (nISl) is transmitted to all of its progeny, it is presumably an integrated transformant. LI ced-1 animals contain an average of 23 cell corpses in the head. LI ced-1 ; ced-3 animals contain an average of 0.3 cell corpses in the head, ced-1 ; unc-31 ced-3 ; nISl and ced-1 ; unc-31 ced-3 ; nEX2 animals contain an

average of 16.4 and 14.5 cell corpses in the head, respectively. From these results, it was concluded that C48D1 contains the ced-3 gene.

In order to locate ced-3 more precisely within the cosmid C48D1, this cosmid was subcloned and the subclones were tested for the ability to rescue ced-3 mutants. C48D1 DNA was digested with restriction enzymes that cut rarely within the cosmid and the remaining cosmid was self-ligated to generate a subclone. Such subclones were then injected into a ced- 3 mutant to look for completion. When C48D1 was digested with BatnHI and self-ligated, the remaining 14 kb subclone (named C48D1-28) was found to rescue the Ced-3 phenotype when injected into a ced-3 mutant (Figure 2 and Table 4) . C48D1-28 was then partially digested with Bgrlll and self-ligated. Clones of various lengths were isolated and tested for their ability to rescue ced-3 .

One clone, C48D1-43, which did not contain a 1.7 kb Bglll fragment of C48D1-28, was able to rescue ced-3

(Figure 2 and Table 4) . C48D1-43 was further subcloned by digesting with BamHI and ApaJ to isolate a 10 kb BajnHI-Apal fragment. This fragment was subcloned into pBSKII-t- to generate pJ40. pJ40 can restore Ced-3+ phenotype when microinjected into a ced-3 mutant. pJ40 was subcloned by deleting a 2 kb Bglll-Apal fragment to generate pJ107. pJ107 was also able to rescue the Ced-3 phenotype when microinjected into a ced-3 mutant. Deletion of 0.5 kb on the left side of pJ107 could be made by _ExoIII digestion (as in pJ107del28 and pJ107del34) without affecting Ced-3 activity; in fact, one transgenic line, nEX17 , restores full Ced-3 activity. However, the ced-3 rescuing ability was significantly reduced when 1 kb was deleted on the left side of pJ107 (as in pJ107dell2 and pJ107del27) , and the

ability was completely eliminated when a 1.8 kb Sall- Bglll fragment was deleted on the right side of pJ107 (as in pJ55 and pJ56) , suggesting that this Sail site is likely to be in the ced-3 coding region. From these experiments, ced-3 was localized to a DNA fragment of 7.5 kb. These results are summarized in Figure 2 and Table 4.

ced-3 Transcript pJ107 was used to probe a Northern blot of N2 RNA and detected a band of 2.8 kb. Although this transcript is present in 12 ced-3 mutant animals, subsequent analysis showed that all 12 ced-3 mutant alleles contain mutations in the genomic DNA that codes for this mRNA (see below) , thus establishing this RNA as a ced-3 transcript.

The developmental expression pattern of ced-3 was determined by hybridizing a Northern blot of RNA from animals of different stages (eggs, LI through L4 larvae and young adult) with the ced-3 cDNA subclone pJllδ. Such analysis revealed that the ced-3 transcript is most abundant during embryonic development, which is the period when most programmed cell deaths occur, but it was also detected during the LI through L4 larval stages and is present in relatively high levels in young adults. This result suggests that ced-3 is not only expressed in cells undergoing programmed cell death.

Since ced-3 and ced-4 are both required for programmed cell death in C. elegans, one of the genes might act as a regulator of transcription of the other gene. To examine if ced-4 regulates the transcription of ced-3 , RNA was prepared from eggs of ced-4 mutants (nll62, nl416, nl894 , and nl920 ) , and a Northern blot was probed with the ced-3 cDNA subclone pJ118. The presence of RNA in each lane was confirmed with an actin

I probe. Such an experiment showed that the level of ced-3 transcript is normal in ced-4 mutants. This indicates that ced-4 is unlikely to be a transcriptional regulator of ced-3 .

Isolation of a ced-3 cDNA

To isolate cDNA of ced-3 , pJ40 was used as a probe to screen a cDNA library of N2 (Kim and Horvitz, Geneε & Dev. 4:357-371 (1990)). Seven cDNA clones were isolated. These cDNAs can be divided into two groups: one is 3.5 kb and the other 2.5 kb. One cDNA from each group was subcloned and analyzed further. pJ85 contains the 3.5 kb cDNA. Experiments showed that pJ85 contains a ced-3 cDNA fused to an unrelated cDNA; on Northern blots of N2 RNA, the pJ85 insert hybridizes to two RNA transcripts, and on Southern blots of N2 DNA, pJ85 hybridizes to one more band than pJ40 (ced-3 genomic DNA) does. pJ87 contains the 2.5 kb cDNA. On Northern blots, pJ87 hybridizes to a 2.8 kb RNA and on Southern blots, it hybridizes only to bands to which pJ40 hybridizes. Thus, pJ87 contains only ced-3 cDNA.

To show that pJ87 does contain the ced-3 cDNA, a frameshift mutation was made in the Sail site of pJ40 corresponding to the Sa l site in the pJ87 cDNA. Constructs containing the frameshift mutation failed to rescue the Ced-3 phenotype when microinjected into ced-3 mutant animals, suggesting that ced-3 activity has been eliminated.

ced-3 Sequence

The DNA sequence of pJ87 was determined (Figure 3) . pJ87 contains an insert of 2.5 kb which has an open reading frame of 503 amino acids (Figure 3; SEQ ID N0:2). The 5' end of the cDNA contains 25 bp of poly- A/T sequence, which is probably an artifact of cloning

and is not present in the genomic sequence. The cDNA ends with a poly-A sequence, suggesting that it contains the complete 3' end of the transcript. 1 kb of pJ87 insert is untranslated 3' region and not all of it is essential for ced-3 expression, since genomic constructs with deletions of 380 bp of the 3' end can still rescue ced-3 mutants (pJ107 and its derivatives, see Figure 2) . To confirm the DNA sequence obtained from the ced-3 cDNA and to study the structure of the ced-3 gene, the genomic sequence of the ced-3 gene in the plasmid pJ107 was determined (Figure 3; SEQ ID N0:1). Comparison of the ced-3 genomic and cDNA sequences revealed that the ced-3 gene has seven introns that range in size from 54 bp to 1195 bp (Figure 4A) . The four largest introns, as well as sequences 5' of the start codon, were found to contain repetitive elements (Figure 3) . Five types of repetitive elements were found, some of which have been previously characterized in non-coding regions of other C. eleganε genes, such as fem-1 (Spence et al . , Cell 60:981-990 (1990)), lin-12 (J. Yochem, personal communication), and myoD (Krause et al . , Cell 63:907-919 (1990)). Of these, repeat 1 was also found in fem-1 and myoD, repeat 3 in lin-12 and fem-1 , repeat 4 in lin-12 , and repeats 2 and 5 were novel repetitive elements. A combination of primer extension and PCR amplification was used to determine the location and nature of the 5' end of the ced-3 transcript. Two primers (Pexl and Pex2) were used for the primer extension reaction. The Pexl reaction yielded two major bands, whereas the Pex2 reaction gave one band. The Pex2 band corresponded in size to the smaller band from the Pexl reaction, and agreed in length with a possible transcript that is trans-spliced to a C. eleganε splice leader (Bektesh, Geneε & Devel . 2:1277-1283 (1988)) at a consensus splice acceptor at position 2166 of the

genomic sequence (Figure 3) . The nature of the larger Pexl band is unclear.

To confirm the existence of this trans-spliced message in wild-type worms, total C . eleganε RNA was PCR amplified using the SLl-Log5 and SL2-Log5 primer pairs, followed by a reamplification using the SLl-OligolO and SL2-01igolO primer pairs. The SLl reaction yielded a fragment of the predicted length. The identity of this fragment was confirmed by sequencing. Thus, at least some, if not most, of the ced-3 transcript is trans- spliced to SLl. Based on this result, the start codon of the ced-3 message was assigned to the methionine encoded at position 2232 of the genomic sequence (Figure 3). The DNA sequences of 12 EMS-induced ced-3 alleles were also determined (Figure 3 and Table 1) . Nine of the 12 are missense mutations. Two of the 12 are nonsense mutations, which might prematurely terminate the translation of ced-3 . These nonsense ced-3 mutants confirmed that the ced-3 gene is not essential for viability. One of the 12 mutations is an alteration of a conserved splicing acceptor G, and another has a change of a 70% conserved C at the splice site, which could also generate a stop codon even if the splicing is correct. Interestingly, these EMS-induced mutations are in either the N-terminal quarter or C-terminal half of the protein. In fact, 9 of the 12 mutations occur within the region of ced-3 that encodes the last 100 amino acids of the protein. Mutations are notably absent from the middle part of the ced-3 gene (Figure 4A) .

Ced-3 Protein Contains A Region Rich in Serines

The Ced-3 protein is very hydrophilic and no significantly hydrophobic region can be found that might

be a trans-membrane domain (Figure 5) . The Ced-3 protein is rich in serine. From amino acid 78 to amino acid 205 of the Ced-3 protein, 34 out of 127 amino acids are serine. Serine is often the target of serine/threonine protein kinases (Edelman, Ann . Rev. Biochem . 56:567-613 (1987)). For example, protein kinase C can phosphorylate serines when they are flanked on their amino and carboxyl sides by basic residues (Edelman, 1987 supra) . Four of the serines in the Ced-3 protein are flanked by arginines (Figures 6A-B) . The same serine residues might also be the target of related Ser/Thr kinases.

To identify the functionally important regions of the Ced-3 protein, genomic DNAs containing the ced-3 genes from two related nematode species, C. briggsae

(SEQ ID NO:5) and C. vulgar iε (SEQ ID NO:6) were cloned and sequenced. Sequence comparison of the three ced-3 gene products showed that the non-serine-rich region of the proteins is highly conserved (Figure 9) . In C. briggεae and C. vulgariε , many amino acids in the serine-rich region are dissimilar compared to the C. elegans Ced-3 protein. It seems that what is important in the serine-rich region is the overall serine-rich feature rather than the exact amino acid sequence. This hypothesis is also supported by analysis of ced-3 mutations in C. eleganε : none of the 12 EMS- induced mutations is in the serine-rich region, suggesting that mutations in this region might not affect the function of the Ced-3 protein and thus, could not be isolated in the screen for ced-3 mutants.

EXAMPLE 2 A COMMON MECHANISM OF CELL DEATH IN VERTEBRATES AND INVERTEBRATES

Results from previous studies reported in the scientific literature suggest that cell deaths in a variety of organisms, including vertebrates as well as invertebrates, share a common mechanism which involves the activation of genes. These studies are consistent with the hypothesis that genes similar to the C. eleganε ced-3 and ced-4 genes may be involved in the cell deaths that occur in vertebrates, although certain observations have led some to distinguish vertebrate cell deaths from the programmed cell deaths observed in such invertebrates as nematodes and insects. Some vertebrate cell deaths share certain characteristics with the programmed cell deaths in C. eleganε that are controlled by ced-3 and ced-4 . For example, up to 14% of the neurons in the chick dorsal root ganglia die immediately after their births, before any signs of differentiation (Carr and Simpson, Dev. Brain Reε . 2:57-162 (1982)). Genes like ced-3 and ced-4 could well function in this class of vertebrate cell death.

Genetic mosaic analysis has suggested that ced-3 and ced-4 genes are expressed by cells that undergo programmed cell death, so that these genes may not act through cell-cell interactions (Yuan and Horvitz, Dev. Biol . 138 : 33-41 (1990)). Many cell deaths in vertebrates seem different in that they appear to be controlled by interactions with target tissues. For example, it is thought that a deprivation of target- derived growth factors is responsible for vertebrate neuronal cell deaths (Hamburger and Oppenhei , Neuroεci . Comment . 1:39-55 (1982)); Thoenen et al . , in: Selective Neuronal Death , Wiley, New York, 1987, Vol. 126, pp. 82-

85) . However, even this class of cell death could involve genes like ced-3 and ced-4 , since pathways of cell death involving similar genes and mechanisms might be triggered in a variety of ways. Supporting this idea are several in vitro and in vivo studies which show that the deaths of vertebrate as well as invertebrate cells can be prevented by inhibitors of RNA and protein synthesis, suggesting that activation of genes are required for these cell deaths (Martin et al . , J . Cell Biol . 106:829-844 (1988); Cohen and Duke, J. Immunol . 132:38-42 (1984); Oppenheim and Prevette, Neurosci . Abεtr. 14:368 (1988); Stanisic et al . , Invest . Urol . 16 : 19-22 (1978); Oppenheim et al . , Dev. Biol . 138 : 104- 113 (1990) ; Fahrbach and Truman, in: Selective Neuronal Death, Ciba Foundation Symposium , 1987, No. 126, pp. 65- 81) . It is possible that the genes induced in these dying vertebrate and invertebrate cells are cell death genes which are structurally related to the C. elegans ced-3 or ced-4 genes. Also supporting the hypothesis that cell death in C. elegans is mechanistically similar to cell death in vertebrates is the observaiton that the protein product of the C. elegans gene ced-9 is similar in sequence to the human protein Bcl-2. ced-9 has been shown to prevent cells from undergoing programmed cell death during nematode development by antagonizing the activities of ced-3 and ced-4 (Hengartner, et al . , Nature 356:494-499 (1992)). The Jbc -2 gene has also been implicated in protecting cells against cell death. It seems likely that the genes and proteins with which ced-9 and bcl-2 interact are similar as well.

Table 1 Sites of Mutations in the ced-3 Gene

Nucleotide and codon positions correspond to the numbering in Figure 3.

Table 2 ced-3-lacZ Fusions Which

Prevent Programmed Cell Death

Table 3

Summary of Transformation Experiments

Using Cosmids in the ced-3 Region

Animals injected were of genotype: ced-1 (el735) ; unc-31 (e929) ced-3 (n717) .

Table 4

The expression of ced-3 (+) transformants

Average No. No. cell deaths Animals

Genotype DNA injected in LI head scored ced-1 23 20

ced-1 ; ced-3 0.3 10

ced-1; nISl C48D1; 16.4 20 unc-31 ced-3 C14G10

ced-1; unc-31 14.5 20 ced-3 ; nISl/+

ced-1 ; unc-31 C48D1; 13.2 10/14 ced-3 ; nEX2 C14G10 4/14

ced-1 ; unc-31 C48D1-28; 12 9/10 ced-3 ; nEXlO C14G10

1 of 10

ced-1 ; unc-31 C48D1-28; 12 10 ced-3 ; nEX9 C14G10

ced-1; unc-31 C48D1-43 16.7 10/13 ced-3 ; nEXll C14G10

Abnormal cell 3/13 deaths

ced-1; unc-31 pJ40; C14G10 13.75 4/4 ced-3 ; nEX13

Table 4 continued

ced-1; unc-31 pJ107del28, 23 12/14 ced-3 ; nEX17 pJ107del34

C14G10

2/14

ced-1; unc-31 pJ107del28, 12.8 9/10 ced-3 ; nEX18 pJ107dell34 C14G10

1/10

ced-1 ; unc-31 pJ107del28, 10.6 5/6 ced-3; nEX19 pJ107del34

G14G10

1/6

ced-1; unc-31 pJ107dell2, 7.8 12/12 ced-3 ; nEX16 pJ107del27

C14G10

Alleles of the genes used are ced-1 (ell '35) , unc-31 (e928) , and ced 3 (n717) .

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT

(A) NAME: Massachusetts Institute of Technology

(B) STREET: 77 Massachusetts Avenue

(C) CITY: Cambridge

(D) STATE OR PROVINCE: Massachusetts

(E) COUNTRY: U.S.A.

(F) POSTAL CODE: 02139

(ii) TITLE OF INVENTION: Inhibitors of Ced-3 and Related Proteins

(iii) NUMBER OF SEQUENCES: 14

(iv) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: diskette

(2) INFORMATION FOR SEQ ID NO:l:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 7653 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: double

(D) TOPOLOGY: linear

(ii) MOLECULE TYPE: DNA (genomic)

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

AGATCTGAAA TAAGGTGATA AATTAATAAA TTAAGTGTAT TTCTGAGGAA ATTTGACTGT 60

TTTAGCACAA TTAATCTTGT TTCAGAAAAA AAGTCCAGTT TTCTAGATTT TTCCGTCTTA 120

TTGTCGAATT AATATCCCTA TTATCACTTT TTCATGCTCA TCCTCGAGCG GCACGTCCTC 180

AAAGAATTGT GAGAGCAAAC GCGCTCCCAT TGACCTCCAC ACTCAGCCGC CAAAACAAAC 240

GTTCGAACAT TCGTGTGTTG TGCTCCTTTT CCGTTATCTT GCAGTCATCT TTTGTCGTTT 300

TTTTCTTTGT TCTTTTTGTT GAACGTGTTG CTAAGCAATT ATTACATCAA TTGAAGAAAA 360

GGCTCGCCGA TTTATTGTTG CCAGAAAGAT TCTGAGATTC TCGAAGTCGA TTTTATAATA 420

TTTAACCTTG GTTTTTGCAT TGTTTCGTTT AAAAAAACCA CTGTTTATGT GAAAAACGAT 480

TAGTTTACTA ATAAAACTAC TTTTAAACCT TTACCTTTAC CTCACCGCTC CGTGTTCATG 540

GCTCATAGAT TTTCGATACT CAAATCCAAA AATAAATTTA CGAGGGCAAT TAATGTGAAA 600

CAAAAACAAT CCTAAGATTT CCACATGTTT GACCTCTCCG GCACCTTCTT CCTTAGCCCC 660

ACCACTCCAT CACCTCTTTG GCGGTGTTCT TCGAAACCCA CTTAGGAAAG CAGTGTGTAT 720

CTCATTTGGT ATGCTCTTTT CGATTTTATA GCTCTTTGTC GCAATTTCAA TGCTTTAAAC 780

(2) INFORMATION FOR SEQ ID NO:2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 503 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

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

Met Met Arg Gin Asp Arg Arg Ser Leu Leu Glu Arg Asn lie Met Met 1 5 10 15

Phe Ser Ser His Leu Lys Val Asp Glu lie Leu Glu Val Leu He Ala 20 25 30

Lys Gin Val Leu Asn Ser Asp Asn Gly Asp Met He Asn Ser Cys Gly 35 40 45

Thr Val Arg Glu Lys Arg Arg Glu He Val Lys Ala Val Gin Arg Arg 50 55 60

Gly Asp Val Ala Phe Asp Ala Phe Tyr Asp Ala Leu Arg Ser Thr Gly 65 70 75 80

His Glu Gly Leu Ala Glu Val Leu Glu Pro Leu Ala Arg Ser Val Asp 85 90 95

Ser Asn Ala Val Glu Phe Glu Cys Pro Met Ser Pro Ala Ser His Arg 100 105 110

Arg Ser Arg Ala Leu Ser Pro Ala Gly Tyr Thr Ser Pro Thr Arg Val 115 120 125

His Arg Asp Ser Val Ser Ser Val Ser Ser Phe Thr Ser Tyr Gin Asp 130 135 140

He Tyr Ser Arg Ala Arg Ser Arg Ser Arg Ser Arg Ala Leu His Ser 145 150 155 160

Ser Asp Arg His Asn Tyr Ser Ser Pro Pro Val Asn Ala Phe Pro Ser 165 170 175

Gin Pro Ser Ser Ala Asn Ser Ser Phe Thr Gly Cys Ser Ser Leu Gly 180 185 190

Tyr Ser Ser Ser Arg Asn Arg Ser Phe Ser Lys Ala Ser Gly Pro Thr 195 200 205

Gin Tyr He Phe His Glu Glu Asp Met Asn Phe Val Asp Ala Pro Thr 210 215 220

He Ser Arg Val Phe Asp Glu Lys Thr Met Tyr Arg Asn Phe Ser Ser 225 230 235 240

Pro Arg Gly Met Cys Leu He He Asn Asn Glu His Phe Glu Gin Met 245 250 255

Pro Thr Arg Asn Gly Thr Lys Ala Asp Lys Asp Asn Leu Thr Asn Leu 260 265 270

Phe Arg Cys Met Gly Tyr Thr Val He Cys Lys Asp Asn Leu Thr Gly 275 280 285

Arg Gly Met Leu Leu Thr He Arg Asp Phe Ala Lys His Glu Ser His 290 295 300

Gly Asp Ser Ala He Leu Val He Leu Ser His Gly Glu Glu Asn Val 305 310 315 320

He He Gly Val Asp Asp He Pro He Ser Thr His Glu He Tyr Asp 325 330 335

Leu Leu Asn Ala Ala Asn Ala Pro Arg Leu Ala Asn Lys Pro Lys He 340 345 350

Val Phe Val Gin Ala Cys Arg Gly Glu Arg Arg Asp Asn Gly Phe Pro 355 360 365

Val Leu Asp Ser Val Asp Gly Val Pro Ala Phe Leu Arg Arg Gly Trp 370 375 380

Asp Asn Arg Asp Gly Pro Leu Phe Asn Phe Leu Gly Cys Val Arg Pro 385 390 395 400

Gin Val Gin Gin Val Trp Arg Lys Lys Pro Ser Gin Ala Asp He Leu 405 410 415

He Arg Tyr Ala Thr Thr Ala Gin Tyr Val Ser Trp Arg Asn Ser Ala 420 425 430

Arg Gly Ser Trp Phe He Gin Ala Val Cys Glu Val Phe Ser Thr His 435 440 445

Ala Lys Asp Met Asp Val Val Glu Leu Leu Thr Glu Val Asn Lys Lys 450 455 460

Val Ala Cys Gly Phe Gin Thr Ser Gin Gly Ser Asn He Leu Lys Gin 465 470 475 480

Met Pro Glu Met Thr Ser Arg Leu Leu Lys Lys Phe Tyr Phe Trp Pro 485 490 495

Glu Ala Arg Asn Ser Ala Val 500

(2) INFORMATION FOR SEQ ID NO:3:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 1373 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: double

(D) TOPOLOGY: 1inear

(ii) MOLECULE TYPE: cDNA

( ix ) FEATURE :

(A) NAME/KEY: CDS

(B) LOCATION: 18..1232

(D) OTHER INFORMATION: /product= "human interleukin-lβ convertase"

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

AAAAGGAGAG AAAAGCC ATG GCC GAC AAG GTC CTG AAG GAG AAG AGA AAG 50

Met Ala Asp Lys Val Leu Lys Glu Lys Arg Lys 1 5 10

CTG TTT ATC CGT TCC ATG GGT GAA GGT ACA ATA AAT GGC TTA CTG GAT 98 Leu Phe He Arg Ser Met Gly Glu Gly Thr He Asn Gly Leu Leu Asp 15 20 25

GAA TTA TTA CAG ACA AGG GTG CTG AAC AAG GAA GAG ATG GAG AAA GTA 146 Glu Leu Leu Gin Thr Arg Val Leu Asn Lys Glu Glu Met Glu Lys Val 30 35 40

AAA CGT GAA AAT GCT ACA GTT ATG GAT AAG ACC CGA GCT TTG ATT GAC 194 Lys Arg Glu Asn Ala Thr Val Met Asp Lys Thr Arg Ala Leu He Asp 45 50 55

TCC GTT ATT CCG AAA GGG GCA CAG GCA TGC CAA ATT TGC ATC ACA TAC 242 Ser Val He Pro Lys Gly Ala Gin Ala Cys Gin He Cys He Thr Tyr 60 65 70 75

ATT TGT GAA GAA GAC AGT TAC CTG GCA GGG ACG CTG GGA CTC TCA GCA 290 He Cys Glu Glu Asp Ser Tyr Leu Ala Gly Thr Leu Gly Leu Ser Ala 80 85 90

GAT CAA ACA TCT GGA AAT TAC CTT AAT ATG CAA GAC TCT CAA GGA GTA 338 Asp Gin Thr Ser Gly Asn Tyr Leu Asn Met Gin Asp Ser Gin Gly Val 95 100 105

CTT TCT TCC TTT CCA GCT CCT CAG GCA GTG CAG GAC AAC CCA GCT ATG 386 Leu Ser Ser Phe Pro Ala Pro Gin Ala Val Gin Asp Asn Pro Ala Met 110 115 120

CCC ACA TCC TCA GGC TCA GAA GGG AAT GTC AAG CTT TGC TCC CTA GAA 434 Pro Thr Ser Ser Gly Ser Glu Gly Asn Val Lys Leu Cys Ser Leu Glu 125 130 135

GAA GCT CAA AGG ATA TGG AAA CAA AAG TCG GCA GAG ATT TAT CCA ATA 482 Glu Ala Gin Arg He Trp Lys Gin Lys Ser Ala Glu He Tyr Pro He 140 145 150 155

ATG GAC AAG TCA AGC CGC ACA CGT CTT GCT CTC ATT ATC TGC AAT GAA 530 Met Asp Lys Ser Ser Arg Thr Arg Leu Ala Leu He He Cys Asn Glu 160 165 170

GAA TTT GAC AGT ATT CCT AGA AGA ACT GGA GCT GAG GTT GAC ATC ACA 578 Glu Phe Asp Ser He Pro Arg Arg Thr Gly Ala Glu Val Asp He Thr 175 180 185

GGC ATG ACA ATG CTG CTA CAA AAT CTG GGG TAC AGC GTA GAT GTG AAA 626 Gly Met Thr Met Leu Leu Gin Asn Leu Gly Tyr Ser Val Asp Val Lys 190 195 200

AAA AAT CTC ACT GCT TCG GAC ATG ACT ACA GAG CTG GAG GCA TTT GCA 674 Lys Asn Leu Thr Ala Ser Asp Met Thr Thr Glu Leu Glu Ala Phe Ala 205 210 215

CAC CGC CCA GAG CAC AAG ACC TCT GAC AGC ACG TTC CTG GTG TTC ATG 722 His Arg Pro Glu His Lys Thr Ser Asp Ser Thr Phe Leu Val Phe Met 220 225 230 235

TCT CAT GGT ATT CGG GAA GGC ATT TGT GGG AAG AAA CAC TCT GAG CAA 770 Ser His Gly He Arg Glu Gly He Cys Gly Lys Lys His Ser Glu Gin 240 245 250

GTC CCA GAT ATA CTA CAA CTC AAT GCA ATC TTT AAC ATG TTG AAT ACC 818 Val Pro Asp He Leu Gin Leu Asn Ala He Phe Asn Met Leu Asn Thr 255 260 265

AAG AAC TGC CCA AGT TTG AAG GAC AAA CCG AAG GTG ATC ATC ATC CAG 866 Lys Asn Cys Pro Ser Leu Lys Asp Lys Pro Lys Val He He He Gin 270 275 280

GCC TGC CGT GGT GAC AGC CCT GGT GTG GTG TGG TTT AAA GAT TCA GTA 914 Ala Cys Arg Gly Asp Ser Pro Gly Val Val Trp Phe Lys Asp Ser Val 285 290 295

GGA GTT TCT GGA AAC CTA TCT TTA CCA ACT ACA GAA GAG TTT GAG GAT 962 Gly Val Ser Gly Asn Leu Ser Leu Pro Thr Thr Glu Glu Phe Glu Asp 300 305 310 315

GAT GCT ATT AAG AAA GCC CAC ATA GAG AAG GAT TTT ATC GCT TTC TGC 1010 Asp Ala He Lys Lys Ala His He Glu Lys Asp Phe He Ala Phe Cys 320 325 330

TCT TCC ACA CCA GAT AAT GTT TCT TGG AGA CAT CCC ACA ATG GGC TCT 1058 Ser Ser Thr Pro Asp Asn Val Ser Trp Arg His Pro Thr Met Gly Ser 335 340 345

GTT TTT ATT GGA AGA CTC ATT GAA CAT ATG CAA GAA TAT GCC TGT TCC 1106 Val Phe He Gly Arg Leu He Glu His Met Gin Glu Tyr Ala Cys Ser 350 355 360

TGT GAT GTG GAG GAA ATT TTC CGC AAG GTT CGA TTT TCA TTT GAG CAG 1154 Cys Asp Val Glu Glu He Phe Arg Lys Val Arg Phe Ser Phe Glu Gin 365 370 375

CCA GAT GGT AGA GCG CAG ATG CCC ACC ACT GAA AGA GTG ACT TTG ACA 1202 Pro Asp Gly Arg Ala Gin Met Pro Thr Thr Glu Arg Val Thr Leu Thr 380 385 390 395

AGA TGT TTC TAC CTC TTC CCA GGA CAT TAAAATAAGG AAACTGTATG 1249 Arg Cys Phe Tyr Leu Phe Pro Gly His

400 405

AATGTCTGCG GGCAGGAAGT GAAGAGATCG TTCTGTAAAA GGTTTTTGGA ATTATGTCTG 1309

CTGAATAATA AACTTTTTTT GAAATAATAA ATCTGGTAGA AAAATGAAAA AAAAAAAAAA 1369 AAAA 1373

(2) INFORMATION FOR SEQ ID NO:4:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 404 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:4:

Met Ala Asp Lys Val Leu Lys Glu Lys Arg Lys Leu Phe He Arg Ser 1 5 10 15

Met Gly Glu Gly Thr He Asn Gly Leu Leu Asp Glu Leu Leu Gin Thr 20 25 30

Arg Val Leu Asn Lys Glu Glu Met Glu Lys Val Lys Arg Glu Asn Ala 35 40 45

Thr Val Met Asp Lys Thr Arg Ala Leu He Asp Ser Val He Pro Lys 50 55 60

Gly Ala Gin Ala Cys Gin He Cys He Thr Tyr He Cys Glu Glu Asp 65 70 75 80

Ser Tyr Leu Ala Gly Thr Leu Gly Leu Ser Ala Asp Gin Thr Ser Gly 85 90 95

Asn Tyr Leu Asn Met Gin Asp Ser Gin Gly Val Leu Ser Ser Phe Pro 100 105 110

Ala Pro Gin Ala Val Gin Asp Asn Pro Ala Met Pro Thr Ser Ser Gly 115 120 125

Ser Glu Gly Asn Val Lys Leu Cys Ser Leu Glu Glu Ala Gin Arg He 130 135 140

Trp Lys Gin Lys Ser Ala Glu He Tyr Pro He Met Asp Lys Ser Ser 145 150 155 160

Arg Thr Arg Leu Ala Leu He He Cys Asn Glu Glu Phe Asp Ser He 165 170 175

Pro Arg Arg Thr Gly Ala Glu Val Asp He Thr Gly Met Thr Met Leu 180 185 190

Leu Gin Asn Leu Gly Tyr Ser Val Asp Val Lys Lys Asn Leu Thr Ala 195 200 205

Ser Asp Met Thr Thr Glu Leu Glu Ala Phe Ala His Arg Pro Glu His 210 215 220

Lys Thr Ser Asp Ser Thr Phe Leu Val Phe Met Ser His Gly He Arg 225 230 235 240

Glu Gly He Cys Gly Lys Lye His Ser Glu Gin Val Pro Asp He Leu 245 250 255

Gin Leu Asn Ala He Phe Asn Met Leu Asn Thr Lys Asn Cys Pro Ser 260 265 270

Leu Lys Asp Lys Pro Lys Val He He He Gin Ala Cys Arg Gly Asp 275 280 285

Ser Pro Gly Val Val Trp Phe Lys Asp Ser Val Gly Val Ser Gly Asn 290 295 300

Leu Ser Leu Pro Thr Thr Glu Glu Phe Glu Asp Asp Ala He Lys Lys 305 310 315 320

Ala His He Glu Lys Asp Phe He Ala Phe Cys Ser Ser Thr Pro Asp 325 330 335

Asn Val Ser Trp Arg His Pro Thr Met Gly Ser Val Phe He Gly Arg 340 345 350

Leu He Glu His Met Gin Glu Tyr Ala Cys Ser Cys Asp Val Glu Glu 355 360 365

He Phe Arg Lys Val Arg Phe Ser Phe Glu Gin Pro Asp Gly Arg Ala 370 375 380

Gin Met Pro Thr Thr Glu Arg Val Thr Leu Thr Arg Cys Phe Tyr Leu 385 390 395 400

Phe Pro Gly His

(2) INFORMATION FOR SEQ ID NO:5:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 505 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(ix) FEATURE:

(A) NAME/KEY: unsure

(B) LOCATION: at every Xaa

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:5:

Met Met Arg Gin Asp Arg Trp Leu Leu Glu Arg Asn He Leu Glu Phe 1 5 10 15

Ser Ser Lys Leu Gin Ala Asp Leu He Leu Asp Val Leu He Ala Lys 20 25 30

Gln Val Leu Asn Ser Asp Asn Gly Asp Val He Asn Ser Cys Arg Thr 35 40 45

Glu Arg Asp Asn Glu Lys Glu He Val Lys Ala Val Gin Arg Arg Gly 50 55 60

Asp Glu Ala Phe Asp Ala Phe Tyr Asp Ala Leu Arg Asp Thr Gly His 65 70 75 80

Asn Asp Leu Ala Asp Val Leu Met Pro Leu Ser Arg Pro Xaa Xaa Xaa 85 90 95

Asn Pro Val Pro Met Glu Cys Pro Met Ser Pro Ser Ser His Arg Arg 100 105 110

Ser Arg Ala Leu Ser Pro Pro Xaa Tyr Ala Ser Pro Thr Arg Val His 115 120 125

Arg Asp Ser He Ser Ser Val Ser Ser Phe Thr Ser Thr Tyr Gin Asp 130 135 140

Val Tyr Ser Arg Ala Arg Ser Ser Ser Arg Ser Ser Arg Pro Leu Gin 145 150 155 160

Ser Ser Asp Arg His Asn Tyr Met Ser Ala Ala Thr Ser Phe Pro Ser 165 170 175

Gin Pro Xaa Ser Ala Asn Ser Ser Phe Thr Gly Cys Ala Ser Leu Gly 180 185 190

Tyr Ser Ser Ser Arg Asn Arg Ser Phe Ser Lys Thr Ser Ala Gin Ser 195 200 205

Gin Tyr He Phe His Glu Glu Asp Met Asn Tyr Val Asp Ala Pro Thr 210 215 220

He His Arg Val Phe Asp Glu Lys Thr Met Tyr Arg Asn Phe Ser Ser 225 230 235 240

Pro Arg Gly Leu Cys Leu He He Asn Asn Glu His Phe Glu Gin Met 245 250 255

Pro Thr Arg Asn Gly Thr Lys Ala Asp Lys Asp Asn Leu Thr Asn He 260 265 270

Phe Arg Cys Met Gly Tyr Thr Val He Cys Lys Asp Asn Leu Thr Gly 275 280 285

Arg Glu Met Leu Ser Thr He Arg Ser Phe Gly Arg Asn Asp Met His 290 295 300

Gly Asp Ser Ala He Leu Val He Leu Ser His Gly Glu Xaa Asn Val 305 310 315 320

He He Gly Val Asp Asp Val Ser Val Asn Val His Glu He Tyr Asp 325 330 335

Leu Leu Asn Ala Ala Asn Ala Pro Arg Leu Ala Asn Lys Pro Lys Leu

340 345 350

Val Phe Val Gin Ala Cys Arg Gly Glu Arg Arg Asp Asn Gly Phe Pro 355 360 365

Val Leu Asp Ser Val Asp Gly Val Pro Ser Leu He Arg Arg Gly Trp 370 375 380

Asp Asn Arg Asp Gly Pro Leu Phe Asn Phe Leu Gly Cys Val Arg Pro 385 390 395 400

Gin Val Gin Gin Val Trp Arg Lys Lys Pro Ser Gin Ala Asp Met Leu 405 410 415

He Ala Tyr Ala Thr Thr Ala Gin Tyr Val Ser Trp Arg Asn Ser Ala 420 425 430

Arg Gly Ser Trp Phe He Gin Ala Val Cys Glu Val Phe Ser Leu His 435 440 445

Ala Lys Asp Met Asp Val Val Glu Leu Leu Thr Glu Val Asn Lys Lys 450 455 460

Val Ala Cys Gly Phe Gin Thr Ser Gin Gly Ser Asn He Leu Lys Gin 465 470 475 480

Met Pro Glu Leu Thr Ser Arg Leu Leu Lys Lys Phe Tyr Phe Trp Pro 485 490 495

Glu Asp Arg Gly Arg Asn Ser Ala Val 500 505

(2) INFORMATION FOR SEQ ID NO:6:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 480 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(ix) FEATURE:

(A) NAME/KEY: unsure

(B) LOCATION: at every Xaa

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

Thr Val Ser He Ser Leu He He Ala Arg Gin Val Leu Asn Ser Asp 1 5 10 15

Asn Xaa Xaa Met He Asn Ser Cys Gly Xaa Xaa Xaa Xaa Xaa Xaa Xaa 20 25 30

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa

35 40 45

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa

50 55 60

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa

65 70 75 80

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa

85 90 95

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa

100 105 110

Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Xaa Ser Thr Ser

115 120 125

Arg Ser Ser Arg Pro Leu His Thr Ser Asp Arg His Asn Tyr Val Ser 130 135 140

Pro Ser Asn Ser Phe Gin Ser Gin Pro Ala Ser Ala Asn Ser Ser Phe 145 150 155 160

Thr Gly Ser Ser Ser Leu Gly Tyr Ser Ser Ser Arg Thr Arg Ser Tyr 165 170 175

Ser Lys Ala Ser Ala His Ser Gin Tyr He Phe His Glu Glu Asp Met 180 185 190

Asn Tyr Val Asp Ala Pro Thr He His Arg Val Phe Asp Glu Lys Thr 195 200 205

Met Tyr Arg Asn Phe Ser Thr Pro Arg Gly Leu Cys Leu He He Asn 210 215 220

Asn Glu His Phe Glu Gin Met Pro Thr Arg Asn Gly Thr Lys Pro Asp 225 230 235 240

Lys Asp Asn He Ser Asn Leu Phe Arg Cys Met Gly Tyr He Val His 245 250 255

Cys Lys Asp Asn Leu Thr Gly Arg Xaa Met Met Leu Thr He Arg Asp 260 265 270

Phe Ala Lys Asn Glu Thr His Gly Asp Ser Ala He Leu Val He Xaa 275 280 285

Ser His Gly Glu Glu Asn Val He He Gly Val Asp Asp Val Ser Val 290 295 300

Asn Val His Glu He Tyr Xaa Leu Leu Asn Ala Ala Asn Ala Pro Arg 305 310 315 320

Leu Ala Asn Lys Pro Lys Leu Val Phe Val Gin Ala Cys Arg Gly Glu 325 330 335

Arg Arg Asp Val Gly Phe Pro Val Leu Asp Ser Val Asp Gly Val Pro 340 345 350

Ala Leu He Arg Arg Gly Trp Asp Lys Gly Asp Gly Pro Xaa Xaa Asn 355 360 365

Phe Leu Gly Cys Val Arg Pro Gin Ala Gin Gin Val Trp Arg Lys Lys 370 375 380

Pro Ser Gin Ala Asp He Leu He Ala Tyr Ala Thr Thr Ala Gin Tyr 385 390 395 400

Val Ser Trp Arg Asn Ser Ala Arg Gly Ser Trp Phe He Gin Ala Val 405 410 415

Cys Glu Val Phe Ser Leu His Ala Lys Asp Met Asp Val Val Glu Leu 420 425 430

Leu Thr Glu Val Asn Lys Lys Val Ala Cys Gly Phe Gin Thr Ser Gin 435 440 445

Gly Ala Asn He Leu Lys Gin Met Pro Xaa Leu Thr Ser Arg Leu Leu 450 455 460

Lys Lys Phe Tyr Phe Trp Pro Glu Asp Arg Asn Arg Ser Ser Ala Val 465 470 475 480

(2) INFORMATION FOR SEQ ID NO:7:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 30 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:7: TCATCGACTT TTAGATGACT AGAGAACATC 30

(2) INFORMATION FOR SEQ ID NO:8:

(i) SEQUENCE CHARACTERISTICS: (A) ^* LENGTH: 30 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:8: GTTGCACTGC TTTCACGATC TCCCGTCTCT 30

(2) INFORMATION FOR SEQ ID NO:9:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

( i) SEQUENCE DESCRIPTION: SEQ ID NO:9: GTTTAATTAC CCAAGTTTGA G 21

(2) INFORMATION FOR SEQ ID NO:10:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:10: GGTTTTAACC AGTTACTCAA G 21

(2) INFORMATION FOR SEQ ID NO:11:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 19 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:11: CCGGTGACAT TGGACACTC 19

(2) INFORMATION FOR SEQ ID NO:12:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 15 base pairs

(B) TYPE: nucleic acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:12: ACTATTCAAC ACTTG 15

(2) INFORMATION FOR SEQ ID NO:13:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 171 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

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

Met Leu Thr Val Gin Val Tyr Arg Thr Ser Gin Lys Cys Ser Ser Ser 1 5 10 15

Lys His Val Val Glu Val Leu Leu Asp Pro Leu Gly Thr Ser Phe Cys 20 25 30

Ser Leu Leu Pro Pro Pro Leu Leu Leu Tyr Glu Thr Asp Arg Gly Val 35 40 45

Asp Gin Gin Asp Gly Lys Asn His Thr Gin Ser Pro Gly Cys Glu Glu 50 55 60

Ser Asp Ala Gly Lys Glu Glu Leu Met Lys Met Arg Leu Pro Thr Arg 65 70 75 80

Ser Asp Met He Cys Gly Tyr Ala Cys Leu Lys Gly Asn Ala Ala Met 85 90 95

Arg Asn Thr Lys Arg Gly Ser Trp Tyr He Glu Ala Leu Thr Gin Val 100 105 110

Phe Ser Glu Arg Ala Cys Asp Met His Val Ala Asp Met Leu Val Lys 115 120 125

Val Asn Ala Leu He Lys Glu Arg Glu Gly Tyr Ala Pro Gly Thr Glu 130 135 140

Phe His Arg Cys Lys Glu Met Ser Glu Tyr Cys Ser Thr Leu Cys Gin 145 150 155 160

Gin Leu Tyr Leu Phe Pro Gly Tyr Pro Pro Thr 165 170

(2) INFORMATION FOR SEQ ID NO:14:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 402 amino acids

(B) TYPE: amino acid (D) TOPOLOGY: linear

(ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:14:

Met Ala Asp Lys He Leu Arg Ala Lys Arg Lys Gin Phe He Asn Ser 1 5 10 15

Val Ser He Gly Thr He Asn Gly Leu Leu Asp Glu Leu Leu Glu Lys 20 25 30

Arg Val Leu Asn Gin Glu Glu Met Asp.Lys He Lys Leu Ala Asn He

35 40 45

Thr Ala Met Asp Lys Ala Arg Asp Leu Cys Asp His Val Ser Lys Lys 50 55 60

Gly Pro Gin Ala Ser Gin He Phe He Thr Tyr He Cys Asn Glu Asp 65 70 75 80

Cys Tyr Leu Ala Gly He Leu Glu Leu Gin Ser Ala Pro Ser Ala Glu 85 90 95

Thr Phe Val Ala Thr Glu Asp Ser Lys Gly Gly His Pro Ser Ser Ser 100 105 110

Glu Thr Lys Glu Glu Gin Asn Lys Glu Asp Gly Thr Phe Pro Gly Leu 115 120 125

Thr Gly Thr Leu Lys Phe Cys Pro Leu Glu Lys Ala Gin Lys Leu Trp 130 135 140

Lys Glu Asn Pro Ser Glu He Tyr Pro He Met Asn Thr Thr Thr Arg 145 150 155 160

Thr Arg Leu Ala Leu He He Cys Asn Thr Glu Phe Gin His Leu Ser 165 170 175

Pro Arg Val Gly Ala Gin Val Asp Leu Arg Glu Met Lys Leu Leu Leu 180 185 190

Glu Asp Leu Gly Tyr Thr Val Lys Val Lys Glu Asn Leu Thr Ala Leu 195 200 205

Glu Met Val Lys Glu Val Lys Glu Phe Ala Ala Cys Pro Glu His Lys 210 215 220

Thr Ser Asp Ser Thr Phe Leu Val Phe Met Ser His Gly He Gin Glu 225 230 235 240

Gly He Cys Gly Thr Thr Tyr Ser Asn Glu Val Ser Asp He Leu Lys 245 250 255

Val Asp Thr He Phe Gin Met Met Asn Thr Leu Lys Cys Pro Ser Leu 260 265 270

Lys Asp Lys Pro Lys Val He He He Gin Ala Cys Arg Gly Glu Lys 275 280 285

Gin Gly Val Val Leu Leu Lys Asp Ser Val Arg Asp Ser Glu Glu Asp 290 295 300

Phe Leu Thr Asp Ala He Phe Glu Asp Asp Gly He Lys Lys Ala His 305 310 315 320

He Glu Lys Asp Phe He Ala Phe Cys Ser Ser Thr Pro Asp Asn Val 325 330 335

Ser Trp Arg His Pro Val Arg Gly Ser Leu Phe He Glu Ser Leu He 340 345 350

Lys His Met Lys Glu Tyr Ala Trp Ser Cys Asp Leu Glu Asp He Phe 355 360 365

Arg Lys Val Arg Phe Ser Phe Glu Gin Pro Glu Phe Arg Leu Gin Met 370 375 380

Pro Thr Ala Asp Arg Val Thr Leu Thr Lys Arg Phe Tyr Leu Phe Pro 385 390 395 400

Gly His