We have identified recessive overproliferation mutations by screening and examining clones of mutant cells in genetic mosaics of the fruitfly Drosophila melanogaster. This type of screen provides a powerful approach for identifying and studying potential tumor suppressors. One of the identified genes, lats, has been cloned and encodes a putative protein kinase that shares high levels of sequence similarity with three proteins in budding yeast and Neu-rospora that are involved in regulation of the cell cycle and growth. Mutations in lats cause dramatic overproliferation phenotypes and various developmental defects in both mosaic animals and homozygous mutants.

Tumorigenesis in humans is a complex process involving activation of oncogenes and inactivation of tumor suppressor genes (Bishop, 1991). Tumor suppressor genes in humans have been identified through studies of genetic changes occurring in cancer cells (Ponder, 1990; Weinberg, 1991). In Drosophila, tumor suppressor genes have been previously identified by recessive overproliferation mutations that cause late larval and pupal lethality (Gateff, 1978; Gateff and Mechler, 1989; Bryant, 1993; Török et al., 1993). Mutations of interest were identified when dissection of dead larvae and pupae revealed certain overproliferated tissues. Several genes identified in homozygous mutants have been cloned including l(1)discs large-1(dlg; Woods and Bryant, 1991, 1993), fat (Mahoney et al., 1991), l(2)giant larvae (lgl. Lützelschwab et al., 1987; Jacob et al., 1987), expanded (ex; Boedigheimer and Laughon, 1993; Boedigheimer et al., 1993), hyperplastic discs (hyd; Mansfield et al., 1994) and the gene encoding the S6 ribosomal protein (Watson et al., 1992; Stewart and Denell, 1993).

Although examining homozygous mutant animals has allowed the successful identification of overproliferation mutations that cause late larval and pupal lethality, mutations that cause lethality at early developmental stages are unlikely to be recovered by this approach. To circumvent this problem, we have searched for recessive overproliferation mutations in mosaic animals (Fig. 1A). Flies that carry small groups of somatic cells mutated for negative regulators of cell proliferation or tumor suppressors are viable, yet the overproliferated mutant tissues can be readily identifiable.

Fig. 1.

Identifying overproliferation mutations in mosaic flies. (A) Although animals that are homozygous for a lethal mutation could die at an early developmental stage, mosaic flies carrying clones of cells that are homozygous for the same mutation could live. One could identify potential tumor suppressors by generating and examining clones of overproliferated mutant cells in mosaic animals. The genetic constitution of these mosaic flies is similar to the mosaicism of the tumor patients. (B) Genetic scheme. The P-element insertions carrying the FLP recombinase (hsFLP; Golic and Lindquist, 1989), its target site, FRT (solid arrows, Xu and Rubin, 1993), the yellow+and miniwhite+ marker genes (y+ and mini-w+, open arrows) were indicated. Mutagenized males were crossed to females to produce heterozygous embryos. Clones of cells homozygous for the induced mutations were generated in developing first-instar larvae by mitotic recombination at the FRT sites induced with the FLP recombinase. Mosaic adults were examined for overproliferated mutant patches (w−, y−). Individuals carrying clones of interest were then mated to recover the mutations of interest in the next generation (Xu and Rubin, 1993; Xu and Harrison, 1994). Clones of ommatidia derived from fast proliferating mutant cells were identified since they were larger than their darkly pigmented wt twin-spot clones (mini-w+/mini-w+).

Fig. 1.

Identifying overproliferation mutations in mosaic flies. (A) Although animals that are homozygous for a lethal mutation could die at an early developmental stage, mosaic flies carrying clones of cells that are homozygous for the same mutation could live. One could identify potential tumor suppressors by generating and examining clones of overproliferated mutant cells in mosaic animals. The genetic constitution of these mosaic flies is similar to the mosaicism of the tumor patients. (B) Genetic scheme. The P-element insertions carrying the FLP recombinase (hsFLP; Golic and Lindquist, 1989), its target site, FRT (solid arrows, Xu and Rubin, 1993), the yellow+and miniwhite+ marker genes (y+ and mini-w+, open arrows) were indicated. Mutagenized males were crossed to females to produce heterozygous embryos. Clones of cells homozygous for the induced mutations were generated in developing first-instar larvae by mitotic recombination at the FRT sites induced with the FLP recombinase. Mosaic adults were examined for overproliferated mutant patches (w−, y−). Individuals carrying clones of interest were then mated to recover the mutations of interest in the next generation (Xu and Rubin, 1993; Xu and Harrison, 1994). Clones of ommatidia derived from fast proliferating mutant cells were identified since they were larger than their darkly pigmented wt twin-spot clones (mini-w+/mini-w+).

One way to generate mosaic animals is to induce mitotic recombination in developing heterozygous individuals (Fig.1B). Recently, it was found that the site-specific recombination system from yeast, the FLP recombinase and its target site FRT, can be used to induce high frequency of mitotic recombination in Drosophila (Golic and Lindquist, 1989; Golic, 1991). To produce and analyze genetic mosaics, a series of special Drosophila strains was constructed, containing the FLP/FRT recombination system on genetically marked chromosomes (Xu and Rubin, 1993). Using these strains, high frequencies of mosaicism can be produced for more than 95% of the Drosophila genes. We have used these strains to identify overproliferation mutations in mosaic animals. Our results show that screening for overproliferation mutations in mosaic animals is a powerful way to identify negative regulators of cell proliferation and potential tumor suppressor genes. Molecular and genetic characterization of large tumor suppressor (lats), one of the identified genes, revealed that it encodes a predicted novel protein kinase.

Genetics

Fly stocks and crosses were grown on standard medium at 25°C unless otherwise indicated. The F1 mosaic screens were modified from the one described in Xu and Rubin (1993) and in Xu and Harrison (1994). Briefly, the F1 mosaic individuals were produced from three crosses. Mutagenized y w hsFLP1; P[ry+; hs-neo; FRT]40A males were mated to the y w hsFLP1; P[ry+; y+]25F, P[mini-w+; hs-NM]31E, P[ry+; hs-neo; FRT]40A females. Mutagenized y w hsFLP1; P[ry+; hs-neo; FRT]42D males were mated to the y w hsFLP1; P[ry+; hs-neo; FRT]42D, P[ry+; y+]44B, P[mini-w+; hs-NM]46F/CyO females. Finally, mutagenized y w hsFLP1; P[ry+; hs-neo; FRT]82B males were mated to the y w hsFLP1; P[ry+; hs-neo; FRT]82B, P[mini-w+; hs-πM]87E, Sb63b, P[ry+; y+]96E females. The male parents were irradiated with X-rays (4000 r) and were removed from the crosses after 4 days of mating. The eggs from the crosses were collected for every 12 hours and aged for another 30 hours before being incubated in a 38°C water bath for 60 minutes. The F1 animals were then returned to normal culture conditions until eclosion. About 25,000 F1 adults from these crosses were examined. Each P-induced lethal mutation was recombined onto one of the FRT-carrying arms using the neoR and w− double selection as described in Xu and Harrison (1994) before examining its clonal phenotype.

The latsx1 mutation was meiotically mapped to the right of claret. It was further localized to the 100A1-5 region since it complemented Df(3R)tlle(100A2-5; 100C2-3) and failed to complement Df(3R)tllpgx(100A1-2; 100B4-5) and Df(3R)tll20(100A1-3; 100B1-2). A saturation genetic screen had previously been performed for this interval, and three lethal complementation groups, l(3)100Aa, l(3)100Ab and the zfh-1, were isolated (Lai et al., 1993). The latsx1 mutation failed to complement the EMS-induced mutations in l(3)100Aa (latsa1-a15), but complement mutations in l(3)100Ab and zfh-1. The clonal phenotypes were examined for latsx1,P1,a1,a2,a6 and a10 induced either with the FLP/FRT-marker system or X-ray irradiation.

The latsP1 allele was recovered from a mosaic male produced from the cross of y w hsFLP1; P[ry+; hs-neo; FRT]82B x y w P[lacZ; w+]5; P[ry+; hs-neo; FRT]82B/delta2-3, Sb. The mutant chromosome was cleaned up before performing complementation tests and an excision screen (Robertson et al., 1988). Two hundred and fifteen excision lines were established that had lost the w+ gene in the P[lacZ; w+] element (Bier et al., 1989). In about 50% of these lines, the pupal lethality had been reverted completely to wild type, indicating the mutant phenotype is caused by the P-element insertion. Five lines were found to cause lethality at late embryonic and/or early first instar larval stages. The remaining lines were found to cause lethality at larval and pupal stages or to produce viable mutant animals. All of these mutant excision lines (except one which is located outside the 100A1-5 region) failed to complement latsx1and latsP1, but do complement mutations in the zfh-1 and l(3)100Ab loci.

The insert in lats cDNA A2 was cloned into the pCaSpeR-hs vector (Thummel and Pirrotta, 1992) for germ-line transformation. Three of the transform lines were tested and were able to rescue the lethality of the latsa1/latsx1, latsP1and latse26-1 animals after one hour heat shock for every 24 hours during larval and pupal development.

Histology

Fixation and sectioning (2 mm) of adult Drosophila tissues was performed as described (Tomlinson and Ready, 1987). Scanning electron microscopy was performed as described (Xu and Artavanis-Tsakonas, 1990).

Nucleic acid manipulation

A P1 genomic clone (DS02640) mapped in the 100A1-7 region was obtained from the Berkeley Drosophila Genome Center (personal communication; Hartl et al., 1994). DNA fragments from this P1 clone and genomic DNA obtained by plasmid rescue from the latsP1 mutant (Bier et al., 1989) were used to isolate several overlapping cosmids including CLT-52 from the genomic library prepared by J. Tamkun. Genomic DNA from +7.5 (BglII) to −4.2 (EcoRI; Fig. 3) was used to screen a total imaginal disc cDNA library prepared by A. Cowman. Screening approximately 2 million phage yielded three groups of cDNAs (five lats cDNAs; fifteen T1 cDNAs; fourteen T2 cDNAs). The sizes of the inserts in the lats cDNAs are as follows: 5.6 kb in A2; 5.1 kb in B1; 1.1 kb in 9 and 4; and 0.9 kb in B3.

Genomic DNA from latsx1/TM6B, latsa1-15/TM6B, latsP1/TM6B, latse7-2/TM6B, latse78/TM6B, latse100/TM6B, latse119/TM6B and latse148/TM6B flies was digested with a combination of the EcoRI, BamHI, BglII and XhoI restriction enzymes for Southern analysis.

DNA sequencing

DNA sequence was determined by the dideoxy chain termination method (Sanger et al., 1977) using Tag polymerase (Perkin Elmer) and Sequenase (US Biochemical). The sequences of lats cDNAs were determined from both strands using templates generated from plasmids containing EcoRI fragments inserted into the pBlueScriptII vector. Templates generated from DNase 1 deletion subclones were also used. The complete sequences of cDNAs A2 and 9 were determined; partial sequences were determined for cDNAs B1 and 4. Templates of genomic DNA were generated from plasmids containing EcoRI fragments and were sequenced on one strand using synthetic oligonucleotide primers. Mutant DNA from the latsa1 allele was amplified with PCR reactions using synthetic oligonucleotide primers and cloned in the pBlueScript II vector for sequencing.

Screening for overproliferation mutations in mosaic animals

We have screened individuals carrying clones of cells that were homozygous for either X-ray or P-element-induced mutations for overproliferation phenotypes. (Fig. 1B; Materials and Methods). Two types of overproliferation phenotypes were sought: (a) Clones of mutant cells formed overproliferated, outgrowth tissues in a non-position-dependent fashion; (b) Clones of mutant cells formed normal structures, but proliferated faster than wild-type cells such that the sizes of the mutant clones were larger than their wt twin-spot clones. Three independent mutations were identified that caused the first type of phenotype (Fig. 2A-E). A mutation which was allelic to one of the original mutations was later found to cause the second type of phenotype (see below). All three mutations in the first class caused embryonic and/or early larval lethality and they represented single alleles of different loci since they had different chromosome locations. One of them was identified among 215 randomly chosen lethal mutations in which each were caused by a P-element insertion in a different essential gene (Karpen and Spradling, 1992; Berkeley Drosophila Genome Center, personal communication). In addition to these overproliferation mutations, one P-induced mutation was found to cause both unpatterned outgrowth and duplications of patterned structures in mosaic animals, suggesting that this mutation may not directly affect cell proliferation.

Fig. 2.

Mutant phenotypes. (A) A clone of unpatterned, overproliferated lats mutant cells in the eye. (B) Induced at the same stage, the 93B mutant cells formed a less overproliferated clone. (C) A third instar latse26-1 larva (right) was much larger than a wt sibling (left; at 18°C). (D) Wing discs from the larva in (C) (wt, top; latse26-1, bottom). (E) Dissected central nervous systems (wt, top; latse26-1, bottom). (F) A SEM view of a lats clone near the eye. (G) A closer view of a region in F showing the irregularity of the sizes and shapes of the mutant cells. (H) A plastic section of a mutant clone similar to the one in (F). Cells seem to be ‘budding’ out of the surface to form new proliferating lobes (arrows). (I) A lats clone on the back. The boxed area is shown in (J). The bristles in the mutant clone are short, bent and often split (arrows). (K) A closer view of the hairs in a lats clone on the body showing enlarged bases and bent tips. (L) A section of a lats clone on the back showing extra cuticle deposits (arrows). All the mutant clones were induced with latsx1 unless stated differently.

Fig. 2.

Mutant phenotypes. (A) A clone of unpatterned, overproliferated lats mutant cells in the eye. (B) Induced at the same stage, the 93B mutant cells formed a less overproliferated clone. (C) A third instar latse26-1 larva (right) was much larger than a wt sibling (left; at 18°C). (D) Wing discs from the larva in (C) (wt, top; latse26-1, bottom). (E) Dissected central nervous systems (wt, top; latse26-1, bottom). (F) A SEM view of a lats clone near the eye. (G) A closer view of a region in F showing the irregularity of the sizes and shapes of the mutant cells. (H) A plastic section of a mutant clone similar to the one in (F). Cells seem to be ‘budding’ out of the surface to form new proliferating lobes (arrows). (I) A lats clone on the back. The boxed area is shown in (J). The bristles in the mutant clone are short, bent and often split (arrows). (K) A closer view of the hairs in a lats clone on the body showing enlarged bases and bent tips. (L) A section of a lats clone on the back showing extra cuticle deposits (arrows). All the mutant clones were induced with latsx1 unless stated differently.

Fig. 3.

Organization of the lats gene. The genomic restriction map of the lats region is aligned with the lats 5.7 kb transcript unit. The direction of transcription is indicated with large arrows. The sizes of the lats introns are as follows: intron 1 (5.0 kb), intron 2 (5.8 kb), intron 3 (68 bp), intron 4 (63 bp), intron 5 (64 bp), intron 6 (61 bp), intron 7 (62 bp). The genomic DNA from +7.5 (BglII) to −4.2 (EcoRI) was used to screen a total imaginal disc cDNA library, which isolated three groups of cDNAs: lats, T1, T2. The introns in the T2 transcript are not labeled. Only parts of the zfh-1 (18) and T1 transcripts are indicated. The locations of the P-element insertion (latsP1), the deletions in the five excision alleles (latse7-2, e78, e100, e119, e148) and in latsa1, latsa4 are indicated at the bottom. The slash indicates a gap in the genomic map. Restriction sites: EcoRI (small open arrow), BglII(open box) and BamHI (open circle). The BglII site at the −0.5 position of the CLT-52 clone is not present in other genomic DNA. A scale is labeled under the restriction map.

Fig. 3.

Organization of the lats gene. The genomic restriction map of the lats region is aligned with the lats 5.7 kb transcript unit. The direction of transcription is indicated with large arrows. The sizes of the lats introns are as follows: intron 1 (5.0 kb), intron 2 (5.8 kb), intron 3 (68 bp), intron 4 (63 bp), intron 5 (64 bp), intron 6 (61 bp), intron 7 (62 bp). The genomic DNA from +7.5 (BglII) to −4.2 (EcoRI) was used to screen a total imaginal disc cDNA library, which isolated three groups of cDNAs: lats, T1, T2. The introns in the T2 transcript are not labeled. Only parts of the zfh-1 (18) and T1 transcripts are indicated. The locations of the P-element insertion (latsP1), the deletions in the five excision alleles (latse7-2, e78, e100, e119, e148) and in latsa1, latsa4 are indicated at the bottom. The slash indicates a gap in the genomic map. Restriction sites: EcoRI (small open arrow), BglII(open box) and BamHI (open circle). The BglII site at the −0.5 position of the CLT-52 clone is not present in other genomic DNA. A scale is labeled under the restriction map.

The lats locus is defined by a single complementation group of mutations that cause defects throughout development

The mutations caused different levels of overproliferation. One mutation (latsx1) produced much more dramatic overproliferated clones than the ones produced by the other mutations (Fig. 2A,B). The lats mutant clones induced in first instar larvae can be as large as 1/5 of the body size. The latsx1 mutation was genetically mapped in the 100A1-5 region and the locus was further defined by a single complementation group of over fifty alleles including mutations induced by X-ray, EMS, P-element insertion and imprecise excision of the P-element (Table 1; Materials and Methods). Removing the P-element insertion reverted the lethal chromosome into wild type, indicating the P-element insertion is responsible for the mutant phenotype. Furthermore, five of the imprecise excision lines caused late embryonic and early larval lethality which were stronger than the pupal lethality phenotype caused by the latsP1 mutation. These five excision lines failed to complement latsx1, but complemented the mutations in two other complementation groups (l(3)100Ab and zfh-1) in the 100A1-5 region, indicating that these two genes were not affected by the excision alleles.

Table 1.

The alleles of the lats locus

The alleles of the lats locus
The alleles of the lats locus

The lats alleles can be classified into three main groups (Table 1). Strong alleles caused homozygous animals to die at a late embryonic stage or shortly after hatching with no obvious cuticular defect. Mutations in the group of medium alleles cause lethality at different times in larval and pupal development. This group was further divided into two subgroups because three of the excision alleles not only caused pupal lethality, but the sizes of the homozygous mutant animals were also significantly larger than wt animals (Fig. 2C). The weak mutations caused either one or a combination of the following phenotypes: held out wings with broadened blades, rough eye with ventral outgrowth, outgrowth on the dorsal-anterior region of the head and partial to complete sterility (Table 1).

Proliferation defects were observed in both mutant clones in mosaic animals and homozygous mutants. Clones of cells on the head that were homozygous for strong or medium alleles formed unpatterned, overproliferated tissues with many lobes or folds. The mutant cells seemed to be ‘budding out’ of the surface to form new proliferation centers or lobes (Fig. 2A,F,H). The sizes and the shapes of these mutant cells were very irregular. Cells several times larger than their neighbors were often seen in mutant clones, indicating problematic cell division (Fig. 2F,G). Furthermore, lats mutant clones behaved differently from clones mutant for the previously identified Drosophila tumor suppressor genes such as dlg, lgl and hyd. The dlg, lgl or hyd mutant cells proliferated slower than wt cells and, thus, the mutant clones induced in first instar larvae were competed away during growth and did not form detectable clones in the adults (Bryant, 1987; Woods and Bryant, 1989; Mansfield et al., 1994; Allen Shearn, personal communication). In contrast, the lats mutant clones induced at similar developmental stages formed dramatic overproliferated tissues, suggesting the mutant cells proliferated faster than wt cells. Consistent with this notion, clones of cells mutant for a weak lats allele (latsa10) produced normal looking tissues, but the mutant clones were significantly larger than their wt twinspot clones. In homozygous animals, the imaginal discs and the central nervous system in many of the pupal lethal mutants were dramatically overproliferated (Fig. 2D,E). The discs lost the single layer of epithelial structure and formed multi-layer, deformed tissues. The lats overproliferation phenotype was not caused by prevention of differentiation. Cells in the overproliferated mutant clones on the body differentiated and produced bristles and hairs, although the morphologies of these structures were not wild type (Fig. 2I-L). Careful examination of multiple mutant clones confirmed that lats caused mutant cells (w cells in the eye, y bristles and enlarged-base hairs on the body) to overproliferate and did not affect the surrounding wt tissues. Finally, the frequency of overproliferated clones was similar to wt clonal frequency induced with the same FRT element, indicating that loss of the lats function alone is sufficient to initiate the overproliferation process.

Cloning of the lats gene

Genomic DNA from the 100A1-5 region was isolated using probes mapped to this region (Materials and Methods). A restriction map of the relevant genomic region is illustrated in Fig. 3. Genomic DNA flanking the P-insertion site (+7.5 to −4.2) was used to screen a total imaginal disc cDNA library. A group of cDNAs corresponding to a 5.7 kb transcript (lats) was found to contain sequence from the region where the Pelement was inserted (Fig. 3). Two other groups of cDNAs were also isolated (T1 and T2). The 5.7 kb transcript was located in an intron of the T1 gene (Fig. 3). The intron-exon structure of the 5.7 transcription unit was determined by Southern and sequence analysis of the cDNA clones and the corresponding genomic DNA (Materials and Methods). The zfh-1 gene was found to be located at the left side of the 5.7 kb transcription unit (Fig. 3; Fortini et al., 1991).

In addition to latsP1, genomic DNA from the five strong excision alleles was analyzed. All of them deleted exon sequences from the 5.7 kb transcript and, in addition, three of them also deleted sequences in the next transcript (T2; Fig. 3). Furthermore, DNA from the X-ray and EMS-induced mutants was analyzed with cDNA probes made from the 5.7 kb, T2 and T1 transcripts. In two cases alterations were detected in the 5.7 kb transcription unit: a 0.4 kb and a 0.3 kb deletions associated with latsa1 and latsa4, respectively (Fig. 3). The 446 bp deletion in latsa1 was revealed by sequencing. It removed codons 92 to 238 of the open reading frame and caused a frame shift from codon 239 (Fig. 5). Finally, transformants containing a cDNA corresponding to the 5.7 transcript driving by the hsp70 promoter rescued the lethality of both strong and medium lats alleles. These findings indicate that the 5.7 kb transcription units which correspond to the lats gene and strong lats alleles including latsa1 were either amorphic or nearly amorphic alleles.

The lats gene encodes a putative proteinserine/threonine kinase

The 5.7 kb lats transcript was detected throughout development (Fig. 4) and in both adult males and females (data not shown). In addition, probes from the 5.7 kb transcript also detected a second transcript, which is about 1 kb shorter (4.7 kb), in young embryos (0–4 hours; Fig. 4) and in adult males and females. Northern analysis showed there was more maternally deposited 4.7 kb transcripts than 5.7 kb transcripts in young embryos (02 hours; Fig. 4). The 5.7 kb transcript became the dominant message at the embryonic stage (4–6 hours), known to have zygotic gene expression (Fig. 4). No effort was made to isolate cDNA clones corresponding to the 4.7 kb transcript; thus the exact sequence of this short transcript is not known. However, a polyadenylation signal consensus sequence was found at nucleotide position 4655–4660 in the 5.7 kb transcript and in the corresponding genomic DNA (Fig. 5) and a 0.51 kb probe from the 3′ end of the 5.7 kb transcript did not hybridize to the 4.7 kb transcript while a 1 kb probe from the 5′ untranslated region of the 5.7 kb transcript hybridized to both the 5.7 kb and 4.7 kb transcripts. This suggests that the 4.7 kb transcript may be a truncated version of the 5.7 kb transcript. The genomic and cDNA sequence corresponding to the 5.7 kb transcript was determined (Materials and Methods). The entire 5720 bp cDNA sequence, which is interrupted by seven introns, and the putative lats product (LATS), deduced from the long open reading frame, are illustrated in Fig. 5. An interesting feature of the 5.7 kb transcript is the existence of a 141 bp segment located in the 3′ untranslated region (Fig. 5), which is identical to the first 141 bp of the 5′ untranslated region of the class I transcript from the Drosophila phospholipase C gene, plc-21 (Shortridge et al., 1991). The functional significance of this sequence motif is unknown. It could be a regulatory target sequence that is shared by both genes.

Fig. 4.

RNA blot analysis of the lats mRNA. 5 μg of poly(A)+ RNA isolated from various developmental stages was separated on a 1% agarose gel and hybridized with 32P-labeled 5′ end 1 kb probe from the lats cDNA. E0-2 hours, E2-4 hours, E4-6 hours, E6-8 hours, E816 hours and E16-24 hours indicate the age of the embryos in hours. RNA from first, second and third instar larvae is denoted by L1, L2 and L3, respectively. The numbers and arrows on the right correspond to the size and location of the RNA standards. A 5.7 kb RNA is found in all the developmental stages, whereas a 4.7 kb RNA is predominantly presented in 0to 4-hour-old embryos. The blot was also hybridized with DNA from the ribosomal protein gene, RPA1.

Fig. 4.

RNA blot analysis of the lats mRNA. 5 μg of poly(A)+ RNA isolated from various developmental stages was separated on a 1% agarose gel and hybridized with 32P-labeled 5′ end 1 kb probe from the lats cDNA. E0-2 hours, E2-4 hours, E4-6 hours, E6-8 hours, E816 hours and E16-24 hours indicate the age of the embryos in hours. RNA from first, second and third instar larvae is denoted by L1, L2 and L3, respectively. The numbers and arrows on the right correspond to the size and location of the RNA standards. A 5.7 kb RNA is found in all the developmental stages, whereas a 4.7 kb RNA is predominantly presented in 0to 4-hour-old embryos. The blot was also hybridized with DNA from the ribosomal protein gene, RPA1.

Fig. 5.

Composite cDNA sequence of the lats gene. The entire cDNA sequence corresponding to the 5.7 kb lats RNA is shown. This nucleotide sequence is a composite of two cDNA clones (nucleotide 1–191 from cDNA 9 and the rest from cDNA A2). The sequence of the corresponding genomic DNA has been determined and is identical to the cDNA sequence except where indicated (above the cDNA sequence). The predicted amino acid sequence is shown below the cDNA sequence. The opa repeat is indicated by the heavy bar. The location of the putative SH3binding site and the RERDQ peptide are designated by dashed lines. The two sites that match the polyadenylation signal consensus sequence are underlined. The second site is located at 12 bp away from the 3′ end of the cDNA. The locations of the introns are indicated by vertical arrows. The underlined 141 bp sequence at the 3′ end of the lats transcript is identical to the 5′ end untranslated sequence of the class I transcript of the Drosophila phospholipase C gene, plc-21. The location of the 446 bp deletion in the latsa1 allele is also indicated. There are 34 differences between the cDNA and genomic sequences and 31 of them do not affect the deduced amino acid sequence. In the remaining three differences, one changes the serine 206 in cDNA into a cysteine. The second change in the genomic sequence adds an additional glutamine in the polyglutamine opa repeat (Fig. 6; Wharton et al., 1985). The third is the addition of a 15 bp sequence in the genomic DNA after the nucleotide 2644 of the cDNA. This sequence could be translated into another copy of the Arg-Glu-Arg-Asp-Gln peptide. However, this sequence is not present in the two independent cDNA clones that were sequenced (Materials and Methods).

Fig. 5.

Composite cDNA sequence of the lats gene. The entire cDNA sequence corresponding to the 5.7 kb lats RNA is shown. This nucleotide sequence is a composite of two cDNA clones (nucleotide 1–191 from cDNA 9 and the rest from cDNA A2). The sequence of the corresponding genomic DNA has been determined and is identical to the cDNA sequence except where indicated (above the cDNA sequence). The predicted amino acid sequence is shown below the cDNA sequence. The opa repeat is indicated by the heavy bar. The location of the putative SH3binding site and the RERDQ peptide are designated by dashed lines. The two sites that match the polyadenylation signal consensus sequence are underlined. The second site is located at 12 bp away from the 3′ end of the cDNA. The locations of the introns are indicated by vertical arrows. The underlined 141 bp sequence at the 3′ end of the lats transcript is identical to the 5′ end untranslated sequence of the class I transcript of the Drosophila phospholipase C gene, plc-21. The location of the 446 bp deletion in the latsa1 allele is also indicated. There are 34 differences between the cDNA and genomic sequences and 31 of them do not affect the deduced amino acid sequence. In the remaining three differences, one changes the serine 206 in cDNA into a cysteine. The second change in the genomic sequence adds an additional glutamine in the polyglutamine opa repeat (Fig. 6; Wharton et al., 1985). The third is the addition of a 15 bp sequence in the genomic DNA after the nucleotide 2644 of the cDNA. This sequence could be translated into another copy of the Arg-Glu-Arg-Asp-Gln peptide. However, this sequence is not present in the two independent cDNA clones that were sequenced (Materials and Methods).

The predicted LATS product contains 1099 amino acid residues. The C-terminal half of LATS shares extensive sequence similarity with a group of six proteins including the Dbf20 and Dbf2 cell cycle protein-ser/thr kinases from Saccharomyces cerevisiae (Johnston et al., 1990; Toyn et al., 1991; Toyn and Johnston, 1994), the COT-1 putative protein kinase from Neurospora crassa (Yarden et al., 1992; Fig. 6A, 6B). The sequence similarity between the kinase domains of LATS and these proteins (39–49% identity) is much higher than the sequence similarity observed between the different subgroups of protein-ser/thr kinases (20-25% identity; Hanks et al., 1988). However, there is an insertion of about 40 amino acid residues within the kinase domains of these proteins, sharing little sequence similarity (denoted by a black bar in Fig. 6B). The human myotonic dystrophy protein kinases (MDPK) also have significant similarity with the C-terminal region of LATS (Brook et al., 1992; Fu et al., 1993; Mahadevan et al., 1993), but their kinase domains do not contain this ∼40 amino acid insertion. In addition, LATS and these proteins also share significant levels of sequence similarity in the two regions (each contains ∼100–150 a.a.) flanking the kinase domain (2028% identity; Fig. 6A, 6B). In the case of Dbf20, its entire sequence except for the 20 C-terminal most residues can be aligned with LATS, indicating LATS is a close relative of Dbf20. A polyglutamine opa repeat is located near the middle of the protein (Fig. 5; Wharton et al., 1985). The N-terminal half of LATS contains many short homopolymeric runs including polyproline which makes up about 15% of the residues. At least one of the proline-rich stretches closely matches the consensus of SH3-binding sites (Fig. 3B; Ren et al., 1993), raising the possibility that it may interact with SH3containing proteins.

Fig. 6.

Schematic of the lats predicted gene product and the related proteins (A) and sequence comparison of the proteins homology to LATS (B). In panel A, solid, hatched, open and shaded boxes denote putative SH3-binding site, opa repeat, RERDQ peptide and kinase domain in the lats product, respectively. The Dbf20, Dbf2 and COT-1 proteins are illustrated at the bottom. The regions that are homologous to LATS are indicated by shaded boxes. The degrees of sequence similarity (identical sequences) between LATS and the three related proteins are indicated above the corresponding regions of these proteins. In panel B, the C-terminal half of LATS is compared to the six most related proteins that are revealed by blastp search as of Sept. 1, 1994. Amino acid residues identical to LATS are highlighted. Numbers at the beginning of every sequence refer to the position of that amino acid within the total protein sequence. The boundary of the kinase domain is defined according to Hanks et al., (1988). The kinase domain of LATS is more similar to protein-serine/threonine kinases than to protein-tyrosine kinases, especially in the sequences of the domains VI and VIII defined by Hanks et al (1988; protein-serine/threonine kinase consensus in domain VI: Asp-LeuLys-Pro-Glu-Asn. LATS sequence in domain VI: Arg-Asp-Ile-Lys-Pro-Asp-Asn (836-842); protein-serine/threonine kinase consensus in domain VIII: Gly-Thr/Ser-X-X-Tyr/Phe-X-Ala-Pro-Glu. LATS sequence in domain VIII: Gly-Thr-Pro-Asn-Tyr-Ile-Ala-Pro-Glu (917-925)). The location of a region of ∼40 amino acid residues that is not conserved among the proteins is indicated by the heavy bar above the sequence. The sequence of PKTL7 from tobacco, Nicotiana tabacum, was submitted to Genbank by Huang,Y. (X71057). Both the sequence of the protein kinase from spinach, Spinacia oleracea, and the sequence of the protein kinase from common ice plant, Mesembryanthemum crystallinum, were submitted to Genbank by Baur, B., Winter, K., Fischer, K. and Dietz, K. (Z30329 and Z30330).

Fig. 6.

Schematic of the lats predicted gene product and the related proteins (A) and sequence comparison of the proteins homology to LATS (B). In panel A, solid, hatched, open and shaded boxes denote putative SH3-binding site, opa repeat, RERDQ peptide and kinase domain in the lats product, respectively. The Dbf20, Dbf2 and COT-1 proteins are illustrated at the bottom. The regions that are homologous to LATS are indicated by shaded boxes. The degrees of sequence similarity (identical sequences) between LATS and the three related proteins are indicated above the corresponding regions of these proteins. In panel B, the C-terminal half of LATS is compared to the six most related proteins that are revealed by blastp search as of Sept. 1, 1994. Amino acid residues identical to LATS are highlighted. Numbers at the beginning of every sequence refer to the position of that amino acid within the total protein sequence. The boundary of the kinase domain is defined according to Hanks et al., (1988). The kinase domain of LATS is more similar to protein-serine/threonine kinases than to protein-tyrosine kinases, especially in the sequences of the domains VI and VIII defined by Hanks et al (1988; protein-serine/threonine kinase consensus in domain VI: Asp-LeuLys-Pro-Glu-Asn. LATS sequence in domain VI: Arg-Asp-Ile-Lys-Pro-Asp-Asn (836-842); protein-serine/threonine kinase consensus in domain VIII: Gly-Thr/Ser-X-X-Tyr/Phe-X-Ala-Pro-Glu. LATS sequence in domain VIII: Gly-Thr-Pro-Asn-Tyr-Ile-Ala-Pro-Glu (917-925)). The location of a region of ∼40 amino acid residues that is not conserved among the proteins is indicated by the heavy bar above the sequence. The sequence of PKTL7 from tobacco, Nicotiana tabacum, was submitted to Genbank by Huang,Y. (X71057). Both the sequence of the protein kinase from spinach, Spinacia oleracea, and the sequence of the protein kinase from common ice plant, Mesembryanthemum crystallinum, were submitted to Genbank by Baur, B., Winter, K., Fischer, K. and Dietz, K. (Z30329 and Z30330).

Screening for mutations in mosaic animals to identify and study potential tumor suppressors

The comparison between mosaic flies and tumor patients is simplistic yet useful. Tumor patients contain wt tumor suppressor genes in most of their cells and only small groups of cells sustain mutations in tumor suppressors. We have searched for recessive overproliferation mutations in mosaic animals. Flies that carry somatic cells mutated for tumor suppressors or negative regulators of cell proliferation are viable, yet the overproliferation mutant phenotype is readily identifiable. Therefore, mosaic flies, which are in a fashion analogous to tumor patients, provide a mean to screen for potential tumor suppressors. Three overproliferation mutations were identified in our screen. They were not identified as ‘interesting’ mutations in screens for embryonic lethal mutations. Identifying overproliferation mutations in homozygous mutant larvae and pupae is not only biased against embryonic lethals, but also laborious, since it requires establishment of individual lines before examining the potential phenotypes. Further screens for overproliferation mutations in mosaic animals will allow us to identify other important players in pathways that negatively regulate cell proliferation.

The overproliferation phenotypes that we observed were caused by loss of function in a single gene. In humans, it was suggested that most retinoblastomas are caused by defects in a single tumor suppressor (Knudson, 1971). In contrast, evidence indicates that tumorigenesis in other human tissues (e.g. colon cancer) is a multistep process which involves inactivation of more than one gene (Fearon and Vogelstein, 1990; Vogelstein and Kinzler, 1993). Overproliferation caused by defects in multiple genes is unlikely to be detected in our screens unless these genes are located on the same chromosome arm. To identify this type of gene, one could perform a modified mosaic screen that induces clones of cells to become homozygous for more than one mutagenized chromosome arm.

lats affects many tissues throughout development

The lats gene is genetically defined by a single complementation group that consists of various alleles causing a wide range of defects. Different alleles were found to cause lethality at almost every stage during development: embryo, early larvae, late larvae, early pupae, late pupae and pharate-adult. The embryonic lethality occurs in the pharate first instar stage. The early embryonic requirements for lats could well be masked by the wt products that are maternally deposited in the egg. Weak lats alleles produce viable animals with phenotypes ranging from rough eye to sterility. The lats transcripts were detected throughout development up to adult stage, consistent with the observation that lats mutants affect all these stages. Although mutations at lats cause many defects, affecting cell proliferation could cause most of the phenotypes including overproliferation in mutant clones, lethality at the various stages, tissue overproliferation on the head, broadened wing blade and sterility in homozygous mutants. However, phenotypes such as extra cuticle deposits and malformed bristles and hairs are evidence of defects in differentiation.

The different behavior of the lats mutant clones and clones mutant for other previously identified Drosophila tumor suppressors is interesting. Cells mutant for dlg, lgl or hyd seem to fail to receive growth regulation signals. They proliferated slower than wt cells during larval stages when the cells were instructed to proliferate, and they failed to terminate proliferation in late larval and pupal stages when the wt cells have ceased proliferation. In contrast, the lats mutant clones induced during the larval stages were overproliferated and, later, the mutant cells on the body were differentiated to form adult cuticular structures. Thus, lats could be a negative regulator that monitors the rate of proliferation.

The lats gene is located in a complex region. The 5′ end of the lats 5.7 kb transcript (cDNA) is only about 550 bp away from the T2 transcript and its 3′ end is about 1.5 kb away from the zfh-1 transcript. Furthermore, all three of these closely located transcripts are located in an intron of the T1 transcription unit. Thus, a sizable deletion in the 5.7 kb transcription unit could affect the function of any of the genes in the region, which makes it difficult to determine which transcript is responsible for the lats phenotype. The fact that P-element transform lines carrying a cDNA from the 5.7 kb transcript under the hsp70 promoter rescued all types of lats alleles demonstrated that the 5.7 kb transcription unit is the lats gene. The determination of the molecular lesions associated with the various lats alleles will be informative for understanding the structure and function of LATS, and for answering questions such as why only some of the pupal lethal mutants and not others produce giant animals.

The LATS putative protein-ser/thr kinase shares homology with proteins that are involved in regulation of cell cycle and growth in budding yeast and Neurospora

All 11 subdomains of the kinase domain that are found in previously identified protein kinases (Hanks et al., 1988) are conserved in LATS. This predicts that LATS is a protein kinase. Furthermore, the sequence comparisons suggest LATS to be a ser/thr kinase as the LATS kinase domain is more similar to protein-ser/thr kinases than to protein-tyr kinases. The C-terminal half of LATS shares extensive sequence similarity with a group of six proteins. Mutations are known for three of these genes and in each case they affect either cell cycle or growth. The cot-1 (colonial temperature sensitive-1) gene of Neurospora was identified by a temperature-sensitive mutant that causes compact colony growth (Mitchell and Mitchell, 1954; Galsworthy, 1966). Wild-type filamentous ascomycete Neurospora grows on solid media by continuous hyphal elongation and branching to form spreading colonies. Strains lacking functional cot-1 gene are viable, but their hyphae branch extensively, resulting in compact colonial growth (Yarden et al., 1992). This extensive branching phenotype is somewhat similar to the growth property of the lats mutant clones: the lats mutant cells continue to ‘bud’ out of the surface to form new proliferation lobes. Another homologous gene, the DBF2 gene of the budding yeast, was identified in a genetic screen for mutations causing defects in DNA synthesis (Johnston and Thomas, 1982). The temperaturesensitive alleles of DBF2 were found to both delay the initiation of S phase and also to arrest the cell cycle during nuclear division (Johnston et al., 1990). The DBF20 gene was identified through cross hybridization with DBF2 DNA (Toyn et al., 1991). Strains carrying deletions for either DBF2 or DBF20 are viable; however, deleting both genes in the same strain causes lethality. The kinase activities of both proteins have been shown to be specific for serine/threonine residues and are regulated during the cell cycle (Toyn and Johnston, 1994). In the case of Dbf20, its entire sequence except the 20 most C-terminal residues can be aligned with LATS. The mutant phenotype of lats and its sequence homology with the cell cycle protein kinases is consistent with the notion that lats might be directly involved in regulation of the cell cycle. The N-terminal half of LATS contains many proline-rich stretches and at least one of them closely matches the consensus sequence of SH3-binding sites (Ren et al., 1993), raising the possibility that this region could be a regulatory domain for the LATS kinase, which binds to SH3 domain-containing proteins.

In recent years, many protein kinases have been identified to be involved in regulation of the cell cycle and cell proliferation. While Wee1 is an inhibitor of the Cdc2 kinase (Russell and Nurse, 1987; Featherstone and Russell, 1991), all other previously identified protein kinases are positive regulators of cell proliferation. They are either required for completion of the cell cycle or for signalling cells to proliferate. LATS is the first predicted protein-ser/thr kinase that has been shown to cause overproliferation when its function is removed. Studies of lats and other overproliferation mutations in Drosophila will provide a better understanding of how cell proliferation is regulated during development and how mutations could lead to abnormal growth.

We wish to thank G. Rubin for his encouragement of this project. We are grateful to S. Artavanis-Tsakonas, L. Cooley, I. Dawson, E. Fearon, S. Weissman and members of our laboratory for their helpful suggestions regarding the manuscript, Z.C. Lai for information about the 100A1-5 region and the mutant strains, R. Diederich and S. Mahajan for RNA blots, T. Li for DNA sequencing, K. Smith for proof reading, J. Tamkun and A. Cowman for libraries, E. Grell and Y. N. Jan for the y w P[lacZ; w+]5 strain, and the Berkeley Drosophila Genome Center for strains. S. Z. and R. A. S. were supported by a Yale University predoctoral fellowship. This work was supported in part by a grant from the Lucille P. Markey Charitable Trust and a grant from the American Cancer Society (IN-31-35).

Bier
,
E.
,
Vaessin
,
H.
,
Shepherd
,
S.
,
Lee
,
K.
,
McCall
,
K.
,
Barbel
,
S.
,
Ackerman
,
L.
,
Carretto
,
R.
,
Uemura
,
T.
,
Grell
,
E. H.
,
Jan
,
L. Y.
and
Jan
,
Y. N.
(
1989
).
Searching for pattern and mutations in the Drosophila genome with a P-lacZ vector
.
Genes Dev
.
3
,
1273
1287
.
Bishop
,
J. M.
(
1991
).
Molecular themes in oncogenesis
.
Cell
64
,
235
248
.
Boedigheimer
,
M.
,
Bryant
,
P.
and
Laughon
,
A.
(
1993
).
Expanded, a negative regulator of cell proliferation in Drosophila, shows homology to the NF2 tumor suppressor
.
Mechan. Dev
.
44
,
83
84
.
Boedigheimer
,
M. J.
and
Laughon
,
A. S.
(
1993
).
Expanded: a gene involved in the control of cell proliferation in imaginal discs
.
Development
118
,
1291
1301
.
Brook
,
J. D.
,
McCurrach
,
M. E.
,
Harley
,
H. G.
,
Buckler
,
A. J.
,
Church
,
D.
,
Aburatani
,
H.
,
Hunter
,
K.
,
Stanton
,
V. P.
,
Thirion
,
J. P.
,
Hudson
,
T.
,
Sohn
,
R.
,
Zemelman
,
B.
,
Snell
,
R. G.
,
Rundle
,
S. A.
,
Crow
,
S.
,
Davies
,
J.
,
Shelbourne
,
P.
,
Buxton
,
J.
,
Jones
,
C.
,
Juvonen
,
V.
,
Johnson
,
K.
,
Harper
,
P. S.
,
Shaw
,
D. J.
,
Housman
,
D. E.
(
1992
).
Molecular basis of myotonic dystrophy: Expansion of a trinucleotide (CTG) repeat at the 3′ end of a transcript encoding a protein kinase family member
.
Cell
68
,
799
808
.
Bryant
,
P. J.
(
1987
).
Experimental and genetic anaysis of growth and cell proliferation in Drosophila imaginal discs
.
In Genetic Regulation of Development pp. 339-372
.
New York
:
A. R. Liss
.
Bryant
,
P. J.
(
1993
).
Towards the cellular functions of tumor suppressors
.
Trends Cell Biol
.
3
,
31
35
.
Featherstone
,
C.
and
Russell
,
P.
(
1991
).
Fission yeast p107wee1 mitotic inhibitor is a tyrosine/serine kinase
.
Nature
349
,
808
811
.
Fearon
,
E. R.
and
Vogelstein
,
B.
(
1990
).
A genetic model for colorectal tumorigenesis
.
Cell
61
,
759
767
.
Fortini
,
M. E.
,
Lai
,
Z.
and
Rubin
,
G. M.
(
1991
).
The Drosophila zfh-1 and zfh-2 genes encode novel proteins containing both zinc-finger and homeodomain motifs
.
Mechan. Dev
.
34
,
113
122
.
Fu
,
Y. H.
,
Friedman
,
D. L.
,
Richards
,
S.
,
Pearlman
,
M. B.
,
Scarlato
,
G.
,
Fenwick
,
R. G.
and
Caskey
,
C. T.
(
1993
).
Decreased expression of mytonin-protein kinase messenger RNA and protein in adult form of myotonic dystrophy
.
Science
260
,
235
238
.
Galsworthy
,
S. B.
(
1966
).
Ph. D. thesis, University of Wisconsin, Madison
.
Diss. Abstr
.
26
,
6348
.
Gateff
,
E.
(
1978
).
Malignant neoplasms of genetic origin in Drosophila melanogaster
.
Science
200
,
1448
1459
.
Gateff
,
E. A.
and
Mechler
,
B. M.
(
1989
).
Tumor-suppressor genes of Drosophila melanogaster
.
CRC Crit. Rev. Oncogen
.
1
,
221
245
.
Golic
,
K. G.
(
1991
).
Site-specific recombination between homologous chromosomes in Drosophila
.
Science
252
,
958
961
.
Golic
,
K. G.
and
Lindquist
,
S.
(
1989
).
The FLP recombinase of yeast catalyzes site-specific recombination in the Drosophila genome
.
Cell
59
,
499
509
.
Hanks
,
S. K.
,
Quinn
,
A. M.
and
Hunter
,
T.
(
1988
).
The protein kinase family: conserved features and deduced phylogeny of the catalytic domains
.
Science
241
,
42
52
.
Hartl
,
D. L.
,
Nurminsky
,
D. I.
,
Jones
,
R. W.
and
Lozovskaya
,
E. R.
(
1994
).
Genome structure and evolution in Drosophila: Applications of the framework P1 map
.
Proc. Natl. Acad. Sci. USA
91
,
6824
6829
.
Jacob
,
L.
,
Opper
,
M.
,
Metzroth
,
B.
,
Phannavong
,
B.
and
Mechler
,
B. M.
(
1987
).
Structure of the I(2)gl gene of Drosophila and delimitation of its tumor suppressor domain
.
Cell
50
,
215
225
.
Johnston
,
L. H.
,
Eberly
,
S. L.
,
Chapman
,
J. W.
,
Araki
,
H.
and
Sugino
,
A.
(
1990
).
The product of the Saccharomyces cerevisiae cell cycle gene DBF2 has homology with protein kinases and is periodically expressed in the cell cycle
.
Mol. Cell. Biol
.
10
,
1358
1366
.
Johnston
,
L. H.
and
Thomas
,
A. P.
(
1982
).
The isolation of new DNA synthesis mutants in the yeast Saccharomyces cerevisiae
.
Mol. Gen. Genet
.
186
,
439
444
.
Karpen
G. H.
and
Spradling
A. C.
(
1992
).
Analysis of subtelomeric heterochromatin in the Drosophila minichromosome Dp1187 by single P element insertional mutagenesis
.
Genetics
132
,
737
53
.
Knudson
,
A. G.
, Jr
. (
1971
).
Mutation and cancer: statistical study of retinoblastoma
.
Proc. Natl. Acad. Sci. USA
68
,
820
823
.
Lai
,
Z.-C.
,
Rushton
,
E.
,
Bate
,
M.
and
Rubin
,
G. M.
(
1993
).
Loss of function of the Drosophila zfh-1 gene results in abnormal development of mesodermally derived tissues
.
Proc. Natl. Acad. Sci. USA
90
,
4122
4126
.
Lützelschwab
,
R.
,
Klämbt
,
C.
,
Rossa
,
R.
and
Schmidt
,
O.
(
1987
).
A protein product of the Drosophila recessive tumor gene, l (2) giant gl, potentially has cell adhesion properties
.
EMBO J
.
6
,
1791
1797
.
Mahadevan
,
M. S.
,
Amemiya
,
C.
,
Jansen
,
G.
,
Sabourin
,
L.
,
Baird
,
S.
,
Neville
,
C. E.
,
Wormskamp
,
N.
,
Segers
,
B.
,
Batzer
,
M.
,
Lamerdin
,
J.
,
de Jong
,
P.
,
Wieringa
,
B.
and
Korneluk
,
R. G.
(
1993
).
Structure and genomic sequence of the myotonic dystrophy (DM kinase) gene
.
Hum. Mol. Genet
.
2
,
299
304
.
Mahoney
,
P. A.
,
Weber
,
U.
,
Onofrechuk
,
P.
,
Biessmann
,
H.
,
Bryant
,
P. J.
and
Goodman
,
C. S.
(
1991
).
The fat tumor suppressor gene in Drosophila encodes a novel member of the Cadherin gene superfamily
.
Cell
67
,
853
868
.
Mansfield
,
E.
,
Hersperger
,
E.
Biggs
,
J.
and
Shearn
,
A.
(
1994
).
Genetic and molecular analysis of hyperplastic discs, a gene whose product is required for regulation of cell proliferation in Drosophila melanogaster imaginal discs and germ cells
.
Dev. Biol
.
165
,
507
526
.
Mitchell
,
M. B.
and
Mitchell
,
H. K.
(
1954
).
A partial map of linkage group D in Neurospora crassa
.
Proc. Natl. Acad. Sci. USA
40
,
436
440
.
Ponder
,
B. A. J.
(
1990
).
Inherited predisposition to cancer
.
Trends Genet
.
6
,
213
218
.
Ren
,
R.
,
Mayer
,
B. J.
,
Cicchetti
,
P.
and
Baltimore
,
D.
(
1993
).
Identification of a ten-amino acid Proline-rich SH3 binding site
.
Science
259
,
1157
1161
.
Robertson
,
H. M.
,
Preston
,
C. R.
,
Phillis
,
R. W.
,
Johnson-Schlitz
,
D. M.
,
Denz
,
W. K.
and
Engels
,
W. R.
(
1988
).
A stable genomic source of P element transposase in Drosophila melanogaster
.
Genetics
118
,
461
470
.
Russell
,
P.
and
Nurse
,
P.
(
1987
).
Negative regulation of mitosis by wee1+, a gene encoding a protein kinase homolog
.
Cell
49
,
559
567
.
Sanger
,
F.
,
Nicklen
,
S.
and
Coulson
,
A.
(
1977
).
DNA sequencing with chain terminating inhibitors
.
Proc. Natl. Acad. Sci. USA
74
,
5463
5467
.
Shortridge
,
R. D.
,
Yoon
,
J.
,
Lending
,
C. R.
,
Bloomquist
,
B. T.
,
Perdew
,
M. H.
and
Pak
,
W. L.
(
1991
).
A Drosophila phospholipase C gene that is expressed in the central nervous system
.
J. Biol Chem
.
266
,
12474
12480
.
Stewart
,
M. J.
and
Denell
,
R.
(
1993
).
Mutations in the Drosophila gene encoding ribosomal protein S6 cause tissue overgrowth
.
Mol. Cell. Biol
.
13
,
2524
2535
.
Thummel
,
C. S.
and
Pirrotta
,
V.
(
1992
).
New pCaSpeR P element Vectors
.
Drosophila Inform. Service
71
,
150
.
Tomlinson
,
A.
and
Ready
,
D. F.
(
1987
).
Cell fate in Drosophila ommatidium
.
Dev. Biol
.
123
,
264
275
.
Török
,
T.
,
Tick
,
G.
,
Alvarado
,
M.
and
Kiss
,
I.
(
1993
).
P-lacW insertional mutagenesis on the second chromosome of Drosophila melanogaster: Isolation of lethals with different overgrowth phenotypes
.
Genetics
135
,
71
80
.
Toyn
,
J. H.
,
Araki
,
H.
,
Sugino
,
A.
and
Johnston
,
L. H.
(
1991
).
The cell-cycleregulated budding yeast gene DBF2, encoding a putative protein kinase, has a homologue that is not under cell-cycle control
.
Gene
104
,
63
70
.
Toyn
,
J. H.
and
Johnston
,
L. H.
(
1994
).
The Dbf2 and Dbf20 protein kinases of budding yeast are activated after the metaphase to anaphase cell cycle transition
.
EMBO J
.
13
,
1103
1113
.
Vogelstein
,
B.
and
Kinzler
,
K. W.
(
1993
).
The multistep nature of cancer
.
Trends Genet
.
9
,
138
141
.
Watson
,
K. L.
,
Konrad
,
K. D.
,
Woods
,
D. F.
and
Bryant
,
P. J.
(
1992
).
Drosophila homolog of the human S6 ribosomal protein is required for tumor suppression in the hematopoietic system
.
Proc. Natl. Acad. Sci. USA
89
,
11302
11306
.
Weinberg
,
R. A.
(
1991
).
Tumor suppressor genes
.
Science
254
,
1138
1146
.
Wharton
,
K. A.
,
Yedvobnick
,
B.
,
Finnerty
,
V. G.
and
Artavanis-Tsakonas
,
S.
(
1985
).
Opa: a novel family of transcribed repeats shared by the Notch locus and other developmentally regulated loci in Drosophila melanogaster
.
Cell
40
,
55
62
.
Woods
,
D. F.
and
Bryant
,
P. J.
(
1989
).
Molecular cloning of the lethal(1)Discs large-1 oncogene of Drosophila
.
Dev. Biol
.
134
,
222
235
.
Woods
,
D. F.
and
Bryant
,
P. J.
(
1991
).
The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions
.
Cell
66
,
451
464
.
Woods
,
D. F.
and
Bryant
,
P. J.
(
1993
).
ZO-1, DlgA and PSD-95/SAP90: homologous proteins in tight, septate and synaptic cell juctions
.
Mechan. Dev
.
44
,
85
89
.
Xu
,
T.
and
Artavanis-Tsakonis
,
S.
(
1990
).
deltex, a locus interacting with the neurogenic genes, notch, delta and mastermind in Drosophila melanogaster
.
Genetics
126
,
665
677
.
Xu
,
T.
and
Harrison
,
S.
(
1994
).
Mosaic analysis using FLP recombinase
.
Methods in Cell Biology
44
,
655
682
.
Xu
,
T.
and
Rubin
,
G. M.
(
1993
).
Analysis of genetic mosaics in developing and adult Drosophila tissues
.
Development
117
,
1223
1237
.
Yarden
,
O.
,
Plamann
,
M.
,
Ebbole
,
D. J.
and
Yanofsky
,
C.
(
1992
).
cot-1, a gene required for hyphal elongation in Neurospora crassa, encodes a protein kinase
.
EMBO J
.
11
,
2159
2166
.