Drosophila RpS3a, a novel Minute gene situated between the segment polarity genes cubitus interruptus and dTCF

Abstract

Genetic analysis of the small chromosome 4 of Drosophila has been hampered by the virtual lack of recombination. The segment polarity gene cubitus interruptus (ci) maps to the most intensively studied locus on this chromosome. Up to four complementation groups have been found to be associated with ci. We and others have recently characterized a second segment polarity gene, dTCF or pan, 12 kb upstream of ci, in a head-to-head configuration. During the course of these studies we identified a transcription unit in the intergenic region. We report here the cloning of cDNAs from this transcription unit, which encode the Drosophila homologue of the human ribosomal protein S3a (RpS3a). The RpS3a gene is expressed ubiquitously and throughout development. A Minute allele, M(4)101, linked tightly to ci, was found to harbour an integration of a Doc retroposon in the promotor region of RpS3a. Thus, like other Minute loci, M(4)101 encodes a component of the protein synthesis machinery. These data further unravel the complex genetics surrounding the ci and dTCF loci.

Introduction

Chromosome 4 of Drosophila represents only 3.5% of the total genomic content of the fly. It has an unusual property, in that recombination rarely takes place between homologs. This is most likely the result of the small size of the chromosome, as well as its relatively high content of heterochromatin. The lack of recombination in combination with the presumed paucity of genetic loci has made chromosome 4 difficult to study by genetic means. Nevertheless, a few chromosome 4 loci have been studied in some detail (1,2). One of the regions of chromosome 4 that has received attention contains the cubitus interruptus (ci) locus, encoding a segment polarity gene. The only known Minute locus on chromosome 4 [M(4)101] has been tightly linked to the ci locus (1).

The multiplicity of recessive and dominant phenotypes associated with ci was, until recently, thought to be linked to two different loci (3,4). Locke and Tartoff (5), however, proposed that all these mutants arise from one single complex locus on chromosome 4. These same authors placed all ci mutations in three different complementation groups, based on complementation assays using the prototypic ci^D mutant fly. Recently, we and others have shown that the ci^D fly bears a compound mutation affecting both the ci gene and a novel segment polarity gene termed dTCF or pangolin. One of these proposed ci complementation groups, l(4)13, actually represents dTCF (6,7). The dTCF gene is positioned upstream of ci on the proximal portion of chromosome 4 and is transcribed in the opposite direction (6). A more detailed analysis of this region showed that the transcription start sites of these genes are separated by ∼12 kb (8). Orenic et al. (4) report the presence of an unidentified transcription unit in the intergenic region between ci and dTCF, further complicating the genetics of this area.

In the course of analyzing the dTCF gene, we isolated cDNA clones representing this transcription unit from an embryonic mixed stage cDNA library. Detailed analysis of this gene revealed it to be the Drosophila homolog of the human ribosomal protein S3a. Since Minute phenotypes often result from mutation in structural ribosomal genes, we analysed the status of Drosophila RpS3a in M(4)101^57g flies and found an insertion of a Doc retroposon in the promotor, likely inactivating the gene. Identification of this novel gene may shed light on the complex genetics of proximal chromosome 4.

Materials and Methods

Fly stocks

The fly stocks M(4)101^57g/ci^D, ci^D/Ey^D and df(4)M62f/Ey^D were kindly provided by R. Holmgren and the Bloomington Stock Center (Bloomington, IN).

Cloning of RpS3a

A genomic clone 5′ of dTCF was obtained as described elsewhere (8). A random primed mixed staged embryo cDNA library was kindly provided by B. Hovemann (described in 9). This library was screened at low stringency with a probe derived from the genomic phage clone as indicated in Figure 1 according to standard procedures. Positive clones were subcloned into pBluescript SK and sequenced. Genomic structure was determined by direct sequencing of the original genomic dTCF clones. The full-length sequence is available under GenBank accession no. AF034971.

Northern blot analysis

RNA was isolated according to Chomczynski et al. (10), adjusted for whole organisms. Briefly, adult flies, larvae or timed embryos were homogenized in solution D (6.3 M guanidine thiocyanate, 0.04 M sodium citrate, 0.8% Sarcosyl) using a Dounce homogenizer. After this, RNA extraction was performed according to the standard protocol (10). RNA was subjected to electrophoresis and was blotted onto nitrocellulose and probed with the RpS3a cDNA clone.

In situ hybridization

Whole mount digoxigenin in situ hybridization was performed as described elsewhere (6), using RpS3a cDNA as probe.

Southern blot analysis

Genomic DNA was isolated from adult flies according to protocol 47 of Ashburner (11). An aliquot of 1 µg DNA was digested overnight with either HindIII or EcoRI. The digested DNA was subjected to electrophoresis and blotted onto nitrocellulose and the Southern blot subsequently probed with either the cDNA clone or with the genomic clone.

Genomic size-selected plasmid library of M(4)101^57g/ci^D

Southern blot data revealed a mutant HindIII band on the M(4)101^57g chromosome of 4.2 kb, hybridizing to the 5′-portion of the RpS3a cDNA. Genomic DNA from the M(4)101^57g/ci^D flies was digested with HindIII overnight. After electrophoresis, DNA of ∼4.2 kb was cloned into pBluescript SK. Amongst 2 × 10⁵ clones, ∼40 colonies hybridized to the RpS3a cDNA. Several of these were isolated and sequenced.

Rapid amplification of cDNA ends (RACE)

RNA was isolated from three clones and cDNA was generated using AMV reverse transcriptase and RpS3a-specific primer (VEA, gtctttgcttcgacaatagc; Isogen, The Netherlands). After separation of the reaction products from excess primers by use of microspin columns (Pharmacia Biotech, The Netherlands), cDNAs were extended using dATP and TdT (Pharmacia Biotech). Typically, cDNA samples were incubated for 10 min at 37°C using 1 mM dATP and 1 U TdT in the TdT buffer provided by the manufacturer. Subsequently, 1/50 of the tailed products was subjected to PCR using oligo(dT) and RpS3a-specific primer (MVK, cactttttaaccatcgacc; Isogen). The following PCR program was used: two cycles of 30 s at 94°C, 30 s at 42°C, 30 s at 72°C; 25 cycles of 30 s at 94°C, 30 s at 55°C, 30 s at 72°C; a final extension of 7 min at 72°C. PCR products were cloned in vector pGEM T (Promega, Madison, WI) and sequenced.

Results

During analysis of the 12 kb region between ci and dTCF (6), a mixed stage embryonic cDNA library was screened with a genomic probe located 5′ of dTCF (Fig. 1). Two independent cDNA clones were isolated. Sequencing of these clones revealed that they were derived from the same gene. The largest clone of 802 bp was used for further study. The smaller clone contained 699 bp and was completely internal to the larger clone. Database searches revealed that the encoded protein was the orthologue of a ribosomal protein identified in multiple diverse species (Table 1). Comparison with prototypic KRP-A protein from Aplysia (11) indicated that the reading frame encoded by our largest cDNA clone contained a translation start site, but lacked approximately five codons at the 3′-end. In order to determine the missing sequence, the genomic clone was directly sequenced. The full open reading frame, combining cDNA and genomic sequences, is depicted in Figure 2. Further sequencing revealed the genomic structure of the Drosophila RpS3a gene. It was found that the coding region of the RpS3a gene consists of two exons separated by an intron of 101 bp (Figs 1 and 2). The RpS3a gene was transcribed in the same direction as the ci gene and mapped 2.5 kb upstream of dTCF (Fig. 1).

Expression of RpS3a during development of the fly was examined using northern blot analysis. RNA was isolated at different stages of development: embryos of 0–2, 2–4 and 4–8 h, mixed stage larvae and male and female adult flies. The blotted RNA was probed with the complete cDNA clone. Analysis of the hybridizing bands revealed that at every stage an mRNA species of ∼1.7 kb could be detected (Fig. 3). The intensity of each band, after correction for the amount of RNA loaded, was comparable. The observed expression in the 0–2 h embryos indicated maternal contribution of the RpS3a mRNA. Whole mount in situ hybridization revealed high ubiquitous expression of maternally derived RpS3a mRNA in early cleavage stage embryos (Fig. 4A). Expression also remained ubiquitous during germband extension (Fig. 4B). In order to distinguish between ubiquitous expression and background staining, the offspring of df(4)M62f/Ey^D (see below) were tested. Of this population, 25% should be homozygous for df(4)62f, which deletes a large part of chromosome 4 containing both ci and dTCF. Indeed, 25% of the population showed no staining after germband extension (Fig. 4C), confirming the ubiquitous expression pattern of RpS3a.

We next sought to determine the in vivo function of RpS3a. We performed a Southern blot analysis on total genomic Drosophila DNA using several restriction enzymes. Probing with the cDNA clone revealed simple hybridization patterns with bands of the expected sizes, indicating that the Drosophila genome contained only one RpS3a gene (Fig. 5, lanes 4). In the same experiment df(4)M62f/Ey^D flies were analyzed. The Ey^D locus maps to the distal region of chromosome 4 and is not expected to show abberations in the pertinent region. It was predicted that df(4)M62f deficiency would remove the RpS3a gene. Indeed, df(4)M62f/Ey^D flies showed a 50% reduction in the signal as analyzed by phosphorimaging (not shown), in concordance with the proposed physical map.

Unexpectedly, an ∼4.8 kb insert previously proposed to be present on the ci^D chromosome (5) and located very close to the dTCF promotor turned out to be absent from ci^D/Ey^D flies but was detected in M(4)101^57g/ci^D flies (Fig. 5). This insert, which we tentatively mapped to the M(4)101 Minute chromosome, is indicated in Figure 1. Since mutations in many ribosomal genes cause Minute phenotypes (21,22), we hypothesized that the insertion in M(4)101^57g would affect the RpS3a gene, which likely encodes a component of the Drosophila ribosomal protein complex.

Detailed analysis of the insert by Southern blotting indicated the presence of internal HindIII and EcoRI sites. The insert could be mapped between the KpnI site in exon 2 of the RpS3a gene and a XhoI site ∼1 kb upstream of the RpS3a gene. Upon probing with cDNA, a mutant HindIII fragment of 4.2 kb was observed. This fragment was predicted to cover a region from the HindIII site in exon 2 to a HindIII site located in the insert (Fig. 1). In order to determine the exact localization and nature of the insert, we constructed a size-selected plasmid library from M(4)101^57g/ci^D genomic DNA restricted with HindIII. A primary library of 2 × 10⁵ recombinant clones was obtained and screened with the cDNA. Four independent plasmids were sequenced and found to contain identical inserts. Sequence analysis of these clones and comparison with the wild-type genomic sequences revealed the presence of an insert 139 bp upstream of the translation start site of RpS3a (Fig. 2). Using 5′ RACE, we determined the transcription start site of RpS3a (Fig. 2). A putative TATA box is located ∼30 bp upstream of this start site. The Doc insert was located 43 bp upstream of the TATA box, presumably within the RpS3a promotor. Database searching revealed that the inserted sequence was identical to the 4.8 kb Doc retroposon, at least over the 350 bp that we analyzed. Furthermore, the EcoRI and HindIII sites predicted from Southern blotting coincided with sites in this retroposon (23). The Doc retroposon was originally identified as responsible for a spontaneous mutation in the white-one allele (24).

Discussion

In this report, we describe a molecular and genetic characterization of a Drosophila gene encoding a homolog of the mammalian ribosomal protein S3a. We originally cloned cDNAs encoding a transcription unit located directly between the two segment polarity genes, ci and dTCF. Having identified the nature of the encoded protein, we determined the status of the RpS3a gene in M(4)101^57g, a chromosome harbouring a mutation mapping near ci and causing a Minute phenotype. We could show the insertion of a Doc retroposon in the predicted promotor region of RpS3a, which very likely causes inactivation of the gene. These data add yet another structural ribosomal gene to the list of genes mutated in flies with dominant Minute phenotypes. Furthermore, our report unravels part of the complex and incompletely understood genetics of the pertinent region on proximal chromosome 4.

The Minute phenotype results from a mutation in any of >50 loci scattered throughout the Drosophila genome. The mutations are usually dominant and cause a phenotype consisting of short, thin bristles, slow development, reduced viability, rough eyes, small body size and etched tergites. The mutations causing Minute phenotypes are thought to occur in elements of the ribosomal machinery (21,22,25,26). Most cloned ribosomal proteins have been mapped to chromosomal regions near Minute loci (27,28). For six ribosomal protein genes this correlation has been confirmed: M(3)99D is RP49 (21); M(2)60E is RPL19 (29); M(3)95A is RPS3 (27); M(2)32A is RPS13 (22); M(2)32D is RPL9 (25); M(1)15D is RPS5 (30). Conversely, mutations in ribosomal proteins do not always cause Minute phenotypes, as evidenced by recessive lethal mutations in the RpS14 genes (31) which do not display visible phenotypes in heterozygotes. The current study adds another example of a Minute phenotype caused by a mutation in a gene coding for a ribosomal protein.

Based on the high level of sequence conservation, all genes displayed in Table 1 are likely to code for the ribosomal protein S3a. However, with the exception of human S3a (15), the genes were not cloned in a search for ribosomal components. The Aplysia ortholog was cloned in a search for genes specifically expressed in the large neurons of this organism (12). In the same study, the yeast orthologs were cloned by homology (12). The mouse ortholog, termed TU-11, was originally identified as a TNF-inducible protein (18), whilst the rat ortholog fte-1 was proposed to be an effector of the v-fos oncogene (16). The different settings in which the same gene has apparently been identified may reflect particular limiting requirements for structural ribosomal proteins in individual physiological/developmental processes in the cell.

The starting point of this study was the complex and conflicting genetic and molecular data on the ci locus. The confusion is largely caused by application of the compound ci/dTCF mutant ci^D as the prototypic complementation partner in genetic crosses. Originally, two complementation groups, cell and ciD, were believed to map to this region (3), but were later found to both affect the ci gene (5). The latter authors assign three complementation groups to the ci locus and map the M(4)101 locus directly downstream of the ci gene. We and others have since shown that one of the three proposed ci complementation groups, l(4)13, represents mutations in the independent segment polarity gene dTCF (6,7). Moreover, data from Holmgren and colleagues (32) provide evidence that the l(4)17 complementation group has molecular abberations in upstream regulatory sequences of ci. The third complementation group in ci affects composition of the ci mRNA directly (5,32). Finally, this study now shows the M(4)101 locus to be upstream of ci.

Based on published data from our laboratory and Basler and colleagues (6,7) and on the current study we can now draw a new physical map of this region on proximal chromosome 4, which harbours the three genes ci, RpS3a and dTCF (Fig. 6). Elucidation of the physical map should serve as a framework for the further characterization of other mutations in this region.

Acknowledgements

We thank Drs F. Staal and N. Barker for critically reading the manuscript and the members of the Clevers laboratory for helpful discussions. This work was supported by the Netherlands Organization for Scientific Research-Physics Research Netherlands (NWO-SON).

References