CA2210251A1

CA2210251A1 - The c. elegans clock gene gro-1

Info

Publication number: CA2210251A1
Application number: CA002210251A
Authority: CA
Inventors: Thomas Barnes; Siegfried Hekimi; Jason Lemieux; Bernard Lakowski
Original assignee: McGill University
Current assignee: Individual
Priority date: 1997-08-25
Filing date: 1997-08-25
Publication date: 1999-02-25
Also published as: NZ503078A; AU8849598A; ATE365207T1; DE69837967D1; EP1009813A1; AU735436B2; EP1009813B1; WO1999010482A1

Abstract

The invention relates to the identification of gro-1 and to show that the gro-1 gene is involved in the control of a central physiological clock.

Description

CA 022102~1 1997-08-2~

THE C. ~T~r.ANS gro-l GENE

RACR~OUND OF THE INVBNTION
(a) Field of the Invention The invention relates to the identification of gro-1 and to show that the gro-1 gene is involved in the control of a central physiological clock.
(b) DescriPtion of Prior Art The gro-1 gene was originally defined by a spontaneous mutation isolated from of a Caenorhabditis elegans strain that had recently been established from a wild isolate (J. Hodgkin and T. Doniach, Genetics 146: 149-164 (1997)). We have shown that the activity of the gro-1 gene controls how fast the worms live and how soon they die. The time taken to progress through embryonic and post-embryonic development, as well as the life span of gro-1 mutants is increased (Lakowski and Hekimi, Science 272:1010-1013, (1996)). Further-more, these defects are maternally rescuable: when homozygous mutants ( gro-1/gro-1) derive from a heterozygous mother ( gro-1/+), these animals appear to be phenotypically wild-type. The defects are seen only when homozygous mutants derive from a homozygous mother (Lakowski and Hekimi, Science 272:1010-1013, (1996)).
In general, the properties of the gro-1 gene are simi-lar to those of three other genes, clk-1, clk-2 and clk-3 (Wong et al., Genetics 13g: 1247-1259 (1995);
Hekimi et al., Genetics, 141: 1351-1367 (1995);
Lakowski and Hekimi, Science 272:1010-1013, (1996)), and this combination of phenotypes has been called the Clk ("clock") phenotype. All four of these genes interact to determine developmental rate and longevity in the nematode. Detailed examination of the clk-1 mutant phenotype has led to the suggestion that there exists a central physiological clock which coordinates CA 022102~1 1997-08-2~

all or many aspects of cellular physiology, from cell division and growth to aging. All four genes have a similar phenotype and thus appear to impinge on this physiological clock.
It would be highly desirable to be provided with the molecular identity of the gro-l gene.

SUMMARY OF THE INVENTION
One aim of the present invention is to provide the molecular identity of the gro-l gene.
In accordance with the present invention there is provided a gro-l gene which has a function at the level of cellular physiology involved in developmental rate and longevity, wherein gro-l is located within an operon and gro-l mutants have a longer life and a altered cellular metabolism relative to the wild-type.
In accordance with the present invention there is also provided a GRO-l protein which has a function at the level of cellular physiology involved in devel-opmental rate and longevity, wherein said GRO-l protein is encoded by the gro-l gene identified above.
In accordance with the present invention there is also provided a method for the diagnosis and/or prognosis of cancer in a patient, which comprises the steps of:
a) obtaining a tissue sample from said patient;
b~ analyzing DNA of the obtained tissue sample of step a) to determine if the human gro-l gene is altered, wherein alteration of the human gro-l gene is indicative of cancer.
In accordance with the present invention there is also provided a mouse model of aging and cancer, which comprises a gene knock-out of murine gene homolo-gous to gro-l.
In accordance with the present invention there is provided the use of compounds interfering with enzy-CA 022102~1 1997-08-2~

matic activity of GRO-l for enhancing longevity of a host.
In accordance with the present invention there is provided the use of compounds interfering with enzy-matic activity of GRO-l for inhibiting of tumorous growth.
In accordance with the present invention there is provided the use of gro-l to identify at least one other gene within or in the proximity of the gro-l operon.
In accordance with the present invention there is provided a hap-l gene which may have a function at the level of cellular physiology involved in develop-mental rate and longevity.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. lA illustrates the genetic mapping of gro-l ;
Fig. lB illustrates the physical map of the gro-1 region;
Fig. 2A illustrates cosmid clones able to res-cue the gro-l fe2400) mutant phenotype;
Fig. 2B illustrates the genes predicted by Genefinder, the relevant restriction sites and the fragments used to subclone the region;
Fig. 3A illustrates the genomic sequence and translation of the C. elegans gro-l gene;
Fig. 3B illustrates the predicted mutant pro-tein;
Fig. 4 illustrates the five genes of the gro-l operon;
Fig. 5 illustrates the alignment of the pre-dicted GRO-l amino acid sequence with homologues from other species (see text for the origin of these sequences);

CA 022102~1 1997-08-2~

Fig. 6 illustrates the biosynthetic step cata-lyzed by DMAPP transferase (MiaAp in E. coli, Mod5p in S. cerevisiae, and GRO-l in C. elegans); and Fig. 7 illustrates the alignment of the pre-dicted HAP-l amino acid sequence with homologues from other species.

DETATr~n DESCRIPTION OF THE INVENTION

The gro-1 phenotYPe In addition to the previously documented pheno-types, we recently found that gro-1 mutants were tem-perature-sensitive for fertility. At 25~C the progeny of these mutants is reduced so much that a viable strain cannot be propagated. In contrast, gro-1 strains can easily be propagated at 15 and 20~C.
We also discovered that the gro-l(e2400J muta-tion increases the incidence of spontaneous mutations.
As gro-l(e2400) was originally identified in a non-standard background (Hodgkin and Doniach, Genetics 146:149-164 (1997)), we first backcrossed the mutations 8 times against N2, the standard wild type strain. We then undertook to examine the gro-1 strain and N2 for the occurrence of spontaneous mutants which could be identified visually. We focused on the two class of mutants which are detected the most easily by simple visual inspection, uncoordinated mutants (Unc) and dumpy mutants (Dpy). We examined 8200 wild type worms and found no spontaneous visible mutant. By contrast, we found 6 spontaneous mutants among 12500 gro-1 mutants examined. All mutants produced entirely mutant progeny indicating that they were homozygous.
Positional cloninq of gro-1 gro-1 lies on linkage group III, very close to the gene clk-1. To genetically order gro-1 with respect to clk-1 on the genetic map, 54 recombinants in CA 022102~1 1997-08-2~

the dpy-l 7 to lon-l interval were selected from among the self progeny of a strain which was unc-79 (elO30) +
+ clk-l (e2519) lon-l (e678) +/+ dpy-17(el64) gro-1 (e2400) + sma-4 (e729). Three of these showed neither the Gro-l nor the Clk-l phenotypes, but carried unc-79 and sma-4, indicating that these recombination events had occurred between gro-l and clk-l. From the dispo-sition of the markers, this showed that the gene order was dpy-l 7 gro-l clk-l lon-l, and the frequency of events indicated that the gro-l to clk-l distance was 0.03 map units. In this region of the genome, this corresponds to a physical map distance of ~20 kb.
Several cosmids containing wild-type DNA span-ning this region of the genome were tested by microin-jection into gro-l mutants for their ability to comple-ment the gro-l (e2400) mutation (Fig. 1). gro-l was mapped between dpy-l 7 and lon-l on the third chromo-some, 0.03 m.u. to the left of clk-l (Fig. lA) .
Based on the above genetic mapping, gro-l was estimated to be approximately 20 kb to the left of clk-1. Eight cosmids (represented by medium bold lines) were selected as candidates for transformation rescue (Fig. lB). Those which were capable of rescuing the gro-l (e2400) mutant phenotype are represented as heavy bold lines (Fig. lB).
Of these, only B0498, C34E10 and ZC395 were able to rescue the mutant phenotype. Transgenic ani-mals were fully rescued for developmental speed. In addition, the transgenic DNA was able to recapitulate the maternal rescue seen with the wild-type gene, that is, mutants not carrying the transgenic DNA but derived from transgenic mothers display a wild type phenotype.
The 7 kb region common to the three rescuing cosmids had been completely sequenced, and this sequence was publicly available.

CA 022102~1 1997-08-2~

We generated subclones of ZC395 and assayed them for rescue (Fig. 2A). The common 6. 5 kb region is blown up in part B. B0498 has not been sequenced and therefore its ends can not be positioned and are there-fore represented by arrows.
One subclone pMQ2, spanned 3.9 kb and was also able to completely rescue the growth rate defect and recapitulate the maternal effect. The sequences in pMQ2 potentially encodes two genes. However, a second subclone, pMQ3, which contained only the first of the potential genes (named ZC395.7 in Fig. 2A), was unable to rescue.
Furthermore, frameshifts which would disrupt each of the two genes' coding sequences were con-structed in pMQ2 and tested for rescue. Disruption of the first gene (in pMQ4) did not eliminate rescuing ability, but disruption of the second gene (in pMQ5) did. This indicates that the gro-l rescuing activity is provided by the second predicted gene.
pMQ2 was generated by deleting a 29.9 kb SpeI
fragment from ZC395, leaving the left-most 3.9 kb region containing the predicted genes ZC395.7 and ZC395. 6 (Fig. 2B). pMQ3 was created in the same fash-ion, by deleting a 31.4 kb NdeI fragment from ZC395, leaving only ZC395.7 intact. In pMQ4, a frameshift was induced in ZC395.7 by degrading the 4 bp overhang of the ApaI site. A frameshift was also induced in pMQ5 by filling in the 2 bp overhang of the NdeI site found in the second exon of ZC395. 6. These frameshifts pre-sumably abolish any function of ZC395.7 and ZC395. 6 respectively. The dotted lines represent the extent of frameshift that resulted from these alterations.
To establish the splicing pattern of this gene, cDNAs encompassing the 5' and 3' halves of the gene CA 022102~1 1997-08-2~

were produced by reverse transcription-PCR and sequenced (Fig. 3).
This revealed that the gene is composed of 9 exons, spans ~2 kb, and produces an mRNA of 1 3 kb. To confirm that this is indeed the gro-l gene, genomic DNA
was amplified by PCR from a strain containing the gro-1 (e2400) mutation and the amplified product was sequenced. A lesion was found in the 5th exon, where a 9 base-pair sequence has been replaced by a 2 base-pair insertion, leading to a frameshift (Fig. 3B). Fig. 3B
illustrates those residues which differ from wild type are in bold.
The reading frame continues out-of-frame for another 33 residues before terminating.
Fig. 3A illustrates the coding sequence in capital letters, while the introns, and the untrans-lated and intergenic sequence are in lower case let-ters. The protein sequence is shown underneath the coding sequence. Position 1 of the nucleotide sequence is the first base after the SL2 trans-splice acceptor sequence. Position 1 of the protein sequence is the initiator methionine. All PCR primers used for genomic and cDNA amplification are represented by arrows. For primers extending downstream (arrows pointing right) the primer sequence corresponds exactly to the nucleo-tides over which the arrow extends. But for primers extending upstream (arrows pointing left) the primer sequence is actually the complement of the sequence under the arrow. In both cases the arrow head is at the 3' end of the primer. The sequence of the two primers which flank gro-l (SHP93 and SHP92) are not represented in this figure. Their sequences are: SHP93 TTTCTGGATTTTAACCTTCC and SHP92 GATAGTTCCCTTCGTTCGGG.
The wild type splicing pattern was determined by sequencing of the cDNA. Identification of the e2400 CA 022102~1 1997-08-2~

lesion was accomplished by sequencing the e2400 allele.
The e2400 lesion consists of a 9 bp deletion and a 2 bp insertion at position 1196, resulting in a frameshift.
gro-l is part of a comPlex oPeron Amplification of the 5' end of gro-l from cDNA
occurred only when the trans-spliced leader SL2 was used as the 5' primer, and not when SLl was used. SL2 is used for trans-splicing to the downstream gene when two genes are organized into an operon (Spieth et al., 10 Cell 73: 521-532 (1993); Zorio et al., Nature 372: 270-272 (1994)). This indicates that at least one gene upstream of gro-l is co-transcribed with gro-l from a common promoter. We found that sequences from the 5' end of the three next predicted genes upstream of gro-l 15 (ZC395.7, C34E10.1, and C34E10.2) all could only be amplified with SL2. Sequences from the fourth predicted upstream gene (C34E10.3), however, could be amplified with neither spliced leader, suggesting that it is not trans-spliced. The distance between genes in operons appear to have an upper limit (Spieth et al., Cell 73: 521-532 (1993); Zorio et al., Nature 372: 270-272 (1994)), and no gene is predicted to be close enough upstream of C34E10.3 or downstream of gro-1 to be co-transcribed with these genes. Our findings sug-gest therefore that gro-1 is the last gene in an operon of five co-transcribed genes (Fig. 4).
Nested PCR was used to amplify the 5' end of each gene. SLl or SL2 specific primers were used in conjunction with a pair of gene-specific primers. cDNA
generated by RT-PCR using mixed stage N2 RNA was used as template in the nested PCR. Fig. 4A illustrates a schematic of the gro-1 operon showing the coding sequences of each gene and the primers (represented by flags) used to establish the trans-splicing patterns.

CA 022102~1 1997-08-2~

Fig. 4B illustrates the products of the PCR
with SLl and SL2 specific primers for each of the five genes. The sequences of the primers used are as fol-lows: SLl: TTTAATTACCCAAGTTTGAG, SL2:
TTTTAACCCAGTTACTCAAG, SHP141: AAAACTTCTACCAACAATGG, SHP142: CGTAATCTCTCTCGATTAGC, SHP143:
CCGTGGGATGGCTACTTGCC, SHP144: TGGATTTGTGGCACGAGCGG, SHP145: TTGATTGCCTCTCCTCGTCC, SHP146:
ATCAACATCTGATTGATTCC, SHP130: CATCCAAAAGCAGTATCACC, SHPll9: ACATCTTTATCCATTTCTCC, SHP95:
TACAGGAATTTTTGAACGGG, SHP99: ATCGATACCACCGTCTCTGG.
The gene immediately upstream of gro-1, has homology to the yeast gene HAM1, and we have renamed the gene hap-1. We have established its splicing pat-tern by reverse transcription PCR and sequencing. Thisrevealed that hap-1 is composed of 5 exons and produces an mRNA of 0.9 kb. We also found that sequences which were predicted to belong to ZC395.7 (now hap-1) are in fact spliced to the exons of C34E10.1. This is consis-tent with our finding that hap-1 is SL2 spliced as it puts the end of the C34E10.1 very close to the start of hap-1 ( Fig. 4).
The gro-1 qene product Conceptual translation of the gro-1 transcript indicated that it encodes a protein of 430 amino acids highly similar to strongly conserved cellular enzyme:
dimethylallyldiphosphate:tRNA dimethylallyltransferase (DMAPP transferase). Fig. 5 shows an alignment of gro-1 with the published sequences of the E. coli ( Caillet and Droogmans, J. Bactriol. 70: 4147-52 (1988)) and yeast (Najarian et al., Molecular & Cellular Biology 7:185-91 (1987)) enzymes, as well as with a sequence inferred from a human expressed sequence tag (Genbank ID: Z40724). The human clone has been used to derive a sequence tagged site (STS). This means that the CA 022102~1 1997-08-2~

genetic and physical position of the human gro-1 homo-logue is known. It maps to chromosome 1, 122.8 cR from the top of Chr 1 linkage group and between the markers DlS255 and DlS2861. This information was found in the UniGene database or the National Center for Biotechnol-ogy Information (NCBI). Fig. 5 illustrates residues where the biochemical character of the amino acid is conserved are shown in bold. Identical amino acids are indicated further with a dot. The ATP/GTP binding site and the C2H2 zinc finger site are predicted and not experimental. The zinc finger site occurs only in the worm sequence. The point at which the gro-l(e2400) mutation alters the reading frame of the sequence is shown. The two alternative initiator methionines in the yeast sequence, and the putative corresponding methionines in the worm sequence, are underlined.
Additional homologue can be found for Leishmania major (Genbank accession T93363) and Arabidopsis thaliana (Genbank accession B09117). In E.
coli and other bacteria, the gene encoding DMAPP trans-ferase is called miaA ( a.k.a trpX) and is called mod5 in yeast. DMAPP transferase catalyzes the modification of adenosine 37 of tRNAs whose anticodon begins with U
(Fig. 6).
In these organisms the enzyme has been shown to use dimethylallyldiphosphate as a donor to generate dimethylallyl-adenosine (dma6A37), one base 3' to the anticodon (for review and biochemical characterization of the bacterial enzyme see Persson et al., Biochimie 76: 1152-1160 (1994)i Leung et al., J Biol Chem 272:
13073-13083 (1997); Moore and Poulter, Biochemistry 36:604-614 (1997)). In earlier literature this modifi-cation is often referred to as isopentenyl adenosine (i6A37).

CA 022102~1 1997-08-2~

The high degree of conservation of the protein sequence between GRO-l and DMAPP in S. cerevisiae and E. coli suggest that GRO-l possesses the same enzymatic activity as the previously characterized genes. The sequence contains a number of conserved structural motifs (Fig. 5), including a region with an ATP/GTP
binding motif which is generally referred to as the 'A' consensus sequence (Walker et al., EMBO J 1: 945-951 (1982)) or the 'P-loop' (Saraste et al., Trends Biochem Sci 15: 430-434 (1990)).
In addition, at the C-terminal end of the GRO-l sequence, there is a C2H2 zinc finger motif as defined by the PROSITE database. This type of DNA-binding motif is believed to bind nucleic acids (Klug and Rhodes, Trends Biochem Sci 12: 464-469 (1987)).
Although there appears to be some conservation between the worm and yeast sequences in the C-terminus end of the protein (Fig. 5), including in the region encom-passing the zinc finger in GRO-l, the zinc finger motif per se is not conserved in yeast.
In yeast DMAPP transferase is the product of the MOD5 gene, and exists in two forms: one form which is targeted principally to the mitochondria, and one form which is found in the cytoplasm and nucleus.
These two forms differ only by a short N-terminal sequence whose presence or absence is determined by differential translation initiation at two "in frame"
ATG codons. (Gillman et al., Mol ~ Cell Biol 11: 2382-90 (1991)). The gro-1 open reading frame also contains two ATG codons at comparable positions, with the coding sequence between the two codons constituting a plausi-ble mitochondrial sorting signal (Figs. 3 and 5). It is likely therefore that DMAPP transferase in worms also exists in two forms, mitochondrial and cytoplasmic.

CA 022102~1 1997-08-2 The hap-1 qene Product hap-1 is homologous to the yeast gene HAM1 as well as to sequences in many organisms including bacte-ria and mammals (Fig. 7).
The origin of the worm and yeast sequence is as described above and below. The human sequence was inferred from a cDNA sequence assembled from expressed sequence tags (ESTs); the accession numbers of the sequences used were: AA024489, AA024794, AA025334, 10 AA026396, AA026452, AA026502, AA026503, AA026611, AA026723, AA035035, AA035523, AA047591, AA047599, AA056452, AA115232, AA115352, AA129022, AA129023, AA159841, AA160353, AA204926, AA226949, AA227197 and D20115. The E. coli sequence is a predicted gene 15 (accession 1723866).
Mutations in HAM1 increase the sensitivity of yeast to the mutagenic compound 6-N-hydroxylaminopurine (HAP), but do not increase spontaneous mutation fre-quency (Nostov et al., Yeast 12:17-29 (1996)). HAP is an analog of adenine and in vitro experiments suggest that the mechanism of HAP mutagenesis is its conversion to a deoxynucleoside triphosphate which is incorporated ambiguously for dATP and dGTP during DNA replication (Abdul-Masih and Bessman, J Biol Chem 261 (5): 2020-2026 (1986)). The role of the Hamlp gene product in increasing sensitivity to HAP remains unclear.

CA 022102~1 1997-08-2~

Explaininq the Pleiotropy of miaA and gro-1 Mutations in miaA, the bacterial homologue of gro-1, show multiple phenotypes and affect cellular growth in complex ways. For example, in Salmonella typhimurium, such mutations result in 1) a decreased efficacy of suppression by some suppressor tRNA, 2) a slowing of ribosomal translation, 3) slow growth under various nutritional conditions, 4) altered regulation of several amino acid biosynthetic operons, 5) sensi-tivity to chemical oxidants and 6) temperature sensi-tivity for aerobic growth (Ericson and Bjork, J. Bacte-riol. 166: 1013-1021 (1986); Blum, ~. Bacteriol. 170:
5125-5133 (1988)). Thus, MiaAp appears to be important in the regulation of multiple parallel processes of cellular physiology. Although we have not yet explored the cellular physiology of gro-1 mutants along the lines which have been pursued in bacteria, the appar-ently central role of miaA is consistent with our find-ings that gro-1, and the other genes with a Clk pheno-type, regulate many disparate physiological and meta-bolic processes in C. elegans (Wong et al., Genetics 139: 1247-1259 (1995) ; Lakowski and Hekimi, Science 272: 1010-1013 (1996); Ewbank et al., Science 275: 980-983 (1997)).
In addition to the various phenotypes discussed above, miaA mutations increase the frequency of sponta-neous mutations (Connolly and Winkler, J Bacteriol 173(5):1711-21 (1991); Connolly and Winkler, J Bacte-riol 171: 3233-46 (1989)). As described in the previ-ous section we have preliminary evidence that gro-l(e2400J also increases the frequency of spontaneous mutations in worms.
How can the alteration in the function of MDAPP
transferase result in so many distinct phenotypes?
Bacterial geneticists working with miaA have generally CA 022102~1 1997-08-2~

suggested that this enzyme and the tRNA modification it catalyzes have a regulatory function which is mediated through attenuation (e.g. Ericson and Bjork, J. Bacte-riol. 166: 1013-1021 (1986)). Attenuation is a phe-nomenon by which the transcription of a gene is inter-rupted depending on the rate at which ribosomes can translate the nascent transcript. Ribosomal transla-tion is slowed in miaA mutants, and thus, through an effect on attenuation, could affect the expression of many genes whose expression is regulated by attenu-ation.
gro-l(e2400) also produces pleiotropic effects and, in addition, displays a maternal-effect, suggest-ing that it is involved in a regulatory process (Wong et al., Genetics 139: 1247-1259 (1995). However, attenuation involves the co-transcriptional translation of nascent transcripts, which is not possible in eukaryotic cells were transcription and translation are spatially separated by the nuclear membrane. If the basis of the pleiotropy in miaA and gro-1 is the same, then a mechanism distinct from attenuation has to be involved. Below we argue that this mechanism could be the modification by DMAPP transferase of adenine resi-dues in DNA in addition to modification of tRNAs.
A role for gro-1 in DNA modification?
We observed that gro-1 can be rescued by a maternal effect, so that adult worms homozygous for the mutation, but issued from mother carrying one wild type copy of the gene display a wild type phenotype, in spite of the fact that such adults are up to 1000 fold larger than the egg produced by their mother. It is unlikely that enough wild type product can be deposited by the mother in the egg to rescue a adult which is 1000 times larger. This observation suggests therefore that gro-1 can induce an epigenetic state which is not CA 022102~1 1997-08-2~

altered by subsequent somatic growth. One of the best documented epigenetic mechanisms is imprinting in mam-mals (Lalande, Annu Rev Genet 30: 173-196 (1996)) which is believed to rely on the differential methylation of genes (Laird and Jaenisch, Annu Rev Genet 30: 441-464;
Klein and Costa, Mutat Res 386: 103-105 (1997)). Modi-fication of bases in DNA have also been linked to regu-lation of gene expression in the protozoan Trypanosoma brucei. The presence of beta-D-glucosyl-hydroxy-methyluracil in the long telomeric repeats of T. bruceicorrelates with the repression of surface antigen gene expression (Gommers-Ampt et al., Cell 75: 112-1136 (1993); van Leeuwen et al., Nucleic Acids Res 24:
2476-2482 (1996)).
gro-1 and miaA increase the rate of spontaneous mutations, which is generally suggestive of a role in DNA metabolism, and can be related to the observation that methylation is linked to spontaneous mutagenesis, genome instability, and cancer (Jones and Gonzalgo, Proc. Natl. Acad. Sci. USA, 94: 2103-2105 (1997)).
Does gro-1 have access to DNA? Studies with mod5, the yeast homologue of gro-1, have shown that one form of Mod5p is localized to the nucleus as well as to the cytoplasm (Boguta et al., Mol. Cell. Biol. 14:
2298-2306 (1994)), and this in spite of the fact that the tRNA modification is believed to occur exclusively in the cytoplasm (reviewed in Boguta et al., Mol. Cell.
Biol. 14: 2298-2306 (1994)). Furthermore, studies of a gene mafl have shown that when mod5 is mislocalized to the nucleus, the efficiency of certain suppressor tRNA
is decreased, an effect known to be linked to the absence of the tRNA modification (Murawski et al., Acta Biochim. Pol. 41: 441-448 (1994)). Finally, as described in the previous section, gro-1 contains a zinc finger, a nuclei acid binding motif. The zinc CA 022102~1 1997-08-2 finger could bind tRNAs, but as it is in the C-terminal domain of gro-l which has no equivalent in miaA, it is clearly not necessary for the basic enzymatic function.
We speculate that it might be necessary to increase the specificity of DNA binding in the large metazoan genome.
m A and gro-1 are found in comPlex operons We have found that gro-l is part of a complex operon of five genes (Fig. 4). It is believed that genes are regulated coordinately by single promoters when they participate in a common function (Spieth et al., Cell 73: 521-532 (1993)). In some cases, this is well documented. For example, the proteins LIN-15A and LIN-15B which are both required for vulva formation in C. elegans, are unrelated products from two genes tran-scribed in a common operon (Huang et al., Mol Biol Cell 5(4): 395-411 (1994)). One of the genes in the gro-l promoter is hap-l, whose yeast homologue has been shown to be involved in the control of mutagenesis (Nostov et al., Yeast 12: 17-29 (1996)). Under the hypothesis that gro-l modifies DNA, it suggest an involvement of hap-1 in this or similar processes. The presence in the same operon also suggest that all five genes might collaborate in a common function. The phenotype of gro-l suggests that this function is regulatory. In this context, it should be noted that miaA also is part of a particularly complex operon (Tsui and Winkler, Biochimie 76: 1168-1177 (1994)), although, except for miaA/gro-l, there are no other homologous genes in the two operons.
A role for gro-l in a central mechanism of physioloqi-cal coordination We have speculated that the genes with a Clk phenotype might participate in a central mechanism of physiological coordination, probably including the CA 022102~1 1997-08-2~

regulation of energy metabolism. clk-1 encodes a pro-tein potentially involved in a diversity of processes, one of which being the regulation of the biosynthesis of ubiquinone (Ewbank et al., Science 275: 980-983 (1997)). Ubiquinone, also called coenzyme Q, is cen-tral to the production of ATP in mitochondria. How might gro-1 fit into this picture?
One link is that dimethylallyldiphosphate is known to be the precursor of the lipid side-chain of ubiquinone. In bacteria, ubiquinone is the major lipid made from DMAPP. In eukaryotes cholesterol and its derivatives are also made from DMAPP. Interestingly, C. elegans requires cholesterol in the growth medium for optimal growth. This link, however, remains tenu-ous, in particular in the absence of an understandingof the biochemical function of CLK-l.
In several bacteria, the adenosine modification carried out by DMAPP transferase is only the first step in a series of further modification of this base (Persson et al., Biochimie 76: 1152-1160 (1994)).
These additional modifications have been proposed to play the role of a sensor for the metabolic state of the cell (Buck and Ames, Cell 36: 523-531 (1984);
Persson and Bjork, J. Bacteriol. 175: 7776-7785 (1993)). For example, one of the subsequent steps, the synthesis of 2-methylthio-cis-ribozeatin is carried out by a hydroxylase encoded by the gene miaE. When the cells lack miaE they become incapable of using intermediates of the citric acid cycle such as fumarate and malate as the sole carbon source.
Another link to energy metabolism springs from the recent biochemical observations of Winkler and co-workers using purified DMAPP transferase ( E. coli MiaAp) (Leung et al., J Biol Chem 272: 13073-13083 (1997)). These investigators observed that the enzyme CA 022102~1 1997-08-2 in competitively inhibited by phosphate nucleotides such as ATP or GTP. Furthermore, using their estimation of Km of the enzyme and its concentration in the cell, they calculate that the level of inhibition of the enzyme in vivo, would exactly allow the enzyme to mod-ify all tRNAs but any further inhibition would leave unmodified tRNAs. This suggests that the exact level of modification of tRNA (or of DNA) could be exqui-sitely sensitive to the level of phosphate nucleotides.
Superficially, this is consistent with the phenotypic observations. The state of mutant cells which lack DMAPP transferase entirely would be equivalent of cells where very high levels of ATP would completely inhibit the enzyme. Such cells might therefore turn down the ATP generating processes in response to the signal pro-vided by undermodified tRNAs (or DNA~.
While the invention has been described in con-nection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any varia-tions, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth, and as follows in the scope of the appended claims.

Claims

1. A gro-1 gene which has a function at the level of cellular physiology involved in developmental rate and longevity, wherein gro-1 is located within an operon and gro-1 mutants have a longer life and a altered cellular metabolism relative to the wild-type.

2. A GRO-1 protein which has a function at the level of cellular physiology involved in developmental rate and longevity, wherein said GRO-1 protein is encoded by the gene of claim 1.

3. A method for the diagnosis and/or prognosis of cancer in a patient, which comprises the steps of:
a) obtaining a tissue sample from said patient;
b) analyzing DNA of the obtained tissue sample of step a) to determine if the human gro-1 gene is altered, wherein alteration of the human gro-1 gene is indicative of cancer.

4. A mouse model of aging and cancer, which comprises a gene knock-out of murine gene homologous to gro-1 according to claim 1.

5. The use of compounds interfering with enzymatic activity of GRO-1 of claim 2 for enhancing longevity of a host.

6. The use of compounds interfering with enzymatic activity of GRO-1 of claim 2 for inhibiting of tumorous growth.

7. The use of gro-1 of claim 1 to identify at least one other gene within or in the proximity of the gro-1 operon, wherein said at least one other gene has a related function.

8. A hap-1 gene which may have a function at the level of cellular physiology involved in developmental rate and longevity.