The bas-1 gene maps to chromosome III, between dpy-17 and unc-32. When this region was sequenced by the C. elegans Genome Sequencing Consortium, two AADC-homologous predicted genes, designated C05D2.4 and C05D2.3, were found to be located close together on a single cosmid, C05D2 (Fig. 2A). This suggested that one (or both) of these sequences comprised the gene mutated in bas-1 mutants. To test this hypothesis, we injected bas-1 mutants with the cosmid C05D2 plus rol-6 (dom) plasmid DNA as a co-injection marker. We isolated transgenic Roller progeny (expressing the rol-6 (dom) marker phenotype) of the injected worm and propagated strains that transmitted the marker, then tested these worm strains using serotonin antibody staining. We found that 3 of 3 independent Roller transgenic lines were rescued for serotonin immunoreactivity, confirming that the bas-1 gene was located within this 46 kb of genomic DNA (Fig. 2B). We then injected plasmid subclones of C05D2, each of which still contained both the predicted C05D2.4 and C05D2.3 genes. A 15.1 kb plasmid subclone (C05D2XN) also rescued bas-1 mutants (n = 4/4), as did smaller subclones of C05D2XN, including an 11.3 kb subclone (pCL3001, n = 11/11) and an 8.8 kb subclone (pCL7001, n = 1/1). These results confirm that at least one of AADC-homologous genes likely corresponds to the bas-1 gene.

To determine which of the two predicted AADC sequences was needed to rescue bas-1 mutants, we prepared two constructs from C05D2XN, one mutated in C05D2.4, the other in C05D2.3 (Fig. 2B). In each case, a mutation was created by eliminating a unique restriction site early in the predicted coding region, creating a frameshift resulting in premature stop codons. We found that constructs mutated in C05D2.3 when injected rescued serotonin immunoreactivity in bas-1 mutants (n = 3/3), whereas the construct mutated in C05D2.4 failed to rescue (n = 0/5 rescued). A construct containing a GFP gene inserted into the C05D2.3 coding sequence (and disrupting the gene) also rescued bas-1 mutants (n = 2/2). In Roller transgenic lines lacking rescue, we confirmed the presence of the injected construct by PCR. Therefore, an intact C05D2.4 gene is necessary to rescue bas-1 mutants, whereas the C05D2.3 gene is not. In all rescued transgenic lines, we saw the complete set of known serotonergic neurons, although not necessarily all cells in every animal – mosaicism from somatic loss of extrachromosomal DNA is expected in these transgenics. This result suggests that no critical cell-specific regulatory sequences were missing from even the smallest construct we injected.

To confirm further that C05D2.4 is the bas-1 gene, we identified the mutations in four bas-1 mutant alleles; we also examined the phenotypes of deletion mutants in C05D2.4 and C05D2.3 generated by the C. elegans Gene Knockout Consortium (GKC). We found that the bas-1 alleles pa4, n2948, and n3008 contained point mutations in C05D2.4 coding sequence resulting in premature stop codons (Fig. 2A). We found that the original bas-1 allele (ad446) had a 4268 bp deletion from the second exon of C05D2.4 to the final intron of C05D2.3; therefore, ad446 is a knockout of both predicted genes. We examined the phenotypes of GKC-generated deletion mutants in each predicted gene. The C05D2.4 knockout (tm351) removes the entire predicted second exon. We found that both tm351 homozygotes and tm351/ad446 worms were deficient in serotonin immunoreactivity. On the other hand, a knockout of C05D2.3 (ok703) is wildtype for serotonin staining. Therefore, tm351 is a fifth mutant allele of the bas-1 gene, and C05D2.4 corresponds to the gene bas-1.

To continue our characterization of the bas-1 gene, we isolated cDNAs using RT-PCR; we also obtained cDNA clones from the C. elegans EST/Transcriptome project (courtesy of Yuki Kohara) and the ORFeome project 40. We found that C05D2.4/bas-1 cDNAs are trans-spliced to SL1 just 15 nucleotides upstream of the predicted translation start site. The consensus sequence from our clones and others we examined predicts a 514 amino acid, 58 kDa protein product (Fig. 3A). This is similar in size to other known AADC/dopa decarboxylase proteins such as those of Drosophila (510 aa) and human (480 aa). The predicted protein possesses a conserved lysine PLP binding site at residue 343, and has other amino acids identical to those shown to be essential for rat AADC function 54142. A number of possible phosphorylation sites can be predicted, including three serines and one tyrosine that are conserved in all known AADCs and HisDCs (Fig. 3A).

We found two splice variants different from the Genefinder-predicted cDNA described above, which was the predominant form. About 20% of clones we sequenced had a 27 bp microexon inserted between the predicted exons 2 and 3 (Fig. 3B). The 27 bp microexon is found within what is the second intron in the more commmon splice form. This intron is not conserved among other AADCs, and is inserted within a region of the BAS-1 protein that is not conserved among AADC proteins. Modeling of BAS-1 protein structure, based on a recent crystal structure of porcine DDC 43, indicates that this region is located at the surface of the protein where it would not interfere with the conserved enzymatic function of the protein (data not shown). We observed that this additional exon is conserved in the C. briggsae ortholog of bas-1 in genomic sequence (Fig. 3B), although we did not isolate any splice variants with this exon among C. briggsae bas-1 cDNAs we sequenced (see also below). We found a single clone that used an alternative splice acceptor 60 bp upstream of the usual splice site for exon 3; this alternate splice introduces a premature stop codon in the coding sequence. This transcript may be a rare, aberrant splice form without functional significance.

We examined the pattern of expression of a GFP reporter construct with ~4500 bp upstream of the predicted bas-1 translation start site and an in-frame fusion with the 2nd exon, injected with rol-6(dom) plasmid into wild-type worms (kindly provided by Ian Hope). Two independent transgenic Roller lines with extrachromosomal arrays had the same pattern of expression. The reporter was reliably expressed in several easily identified cells including the paired serotonergic neurons NSM and HSN and the dopaminergic PDE postdeirid sensory neurons (Fig. 4). NSM processes studded with varicosities were apparent in the isthmus of the pharynx labeled with GFP (Fig. 4A,4D). The egg-laying neuron HSN normally expresses serotonin only in adulthood, and we found the reporter to be expressed in adult hermaphrodites and sometimes late L4 larvae. Often the HSN processes were apparent extending to vulval muscles and anteriorly within the ventral nerve cord (Fig. 4C,4F). We saw a cell we identified as PDE, which is born during L2, only after this stage. In some worms, we saw a PDE process and dendrite, confirming our identification (Fig. 4F).

The bas-1::GFP reporter was also expressed in other neurons in the head, around the nerve ring. We believe that all of these cells are known serotonergic and dopaminergic neurons. It was somewhat more difficult, however, to be certain about these identifications since we saw few processes, and even when present we could not always unambiguously associate a process with a particular neuronal soma. Nevertheless, the reporter was expressed in probable dorsal and ventral cephalic sensilla neurons CEPD and CEPV; we sometimes observed as many as four processes extending to the tip of the nose (Fig 4A,4B,4E). We also saw expression in the anterior deirid sensory neurons ADE (Fig. 4A,4B,4E). Less frequently we saw expression in probable ADF and AIM neurons (Fig 4A,4E). We saw as many as 12 neurons (6 bilateral pairs) expressing the reporter in the head of young larvae. This includes all the identified serotonergic (NSM, ADF, AIM) and dopaminergic (CEPD, CEPV, ADE) head neurons excepting the unpaired RIH neuron 10. In a small number of males examined, we saw expression in male-specific serotonergic and dopaminergic neurons, including up to 6 pairs of ray sensory neurons (RNs) in both adults and late L4 larvae (Fig. 4G,4H). (There are three pairs of serotonergic, and three pairs of dopaminergic RNs among the 18 RNs.) Expression in CP neurons, male-specific ventral cord motoneurons controlling tail curling during mating, was limited and usually weak in the male worms we examined. Six CP neurons are strongly serotonin-immunoreactive in males 16. At most we saw three posterior cells staining, and usually only one or two posterior cells (CP5, CP6) weakly stained, when expression was present at all (Fig. 4I). We never saw CP staining in L4 animals, and often none even in male worms expressing GFP strongly in the RNs.

Just downstream of the bas-1/C05D2.4 gene is C05D2.3, the product of an ancient tandem duplication event. The two genes have diverged considerably – being only 59% identical at the amino acid level (Table 1). The genomic structures of the two genes have also diverged. The two genes share four introns, but C05D2.4 has one and C05D2.3 has three introns not found in the other (Fig. 7). Nevertheless, comparisons with other AADC proteins showed that bas-1/C05D2.4 is most similar to C05D2.3 and the predicted gene F12A10.3 (Fig. 5, Table 1). The predicted amino acid sequence of C05D2.3 contains one noteworthy gap: it is missing six amino acids from a highly conserved region found in all other PLP-dependent decarboxylases. This sequence, the consensus of which is VDAAYA, contains an aspartate (D) residue that is absolutely essential for function of Rat DDC. Substitution of an alanine or asparagine completely abolishes enzymatic activity, and even the conservative substitution of a glutamate at this site reduces activity to 2% of wildtype 5. It is therefore unlikely that a C05D2.3 protein could function enzymatically as a typical AADC.

Because the bas-1 and C05D2.3 genes are so close together – only 369 bp from predicted translation stop to predicted translation start – we considered whether they might be expressed as an operon. In C. elegans and other nematodes, genes that are very close together (and often functionally related) may be expressed from a single promoter initially as a single primary transcript 44. Operon transcripts are subsequently processed to yield separate mRNAs. The first gene in an operon is trans-spliced to the leader sequence SL1; downstream genes are typically spliced to a slightly different leader sequence termed SL2. We would expect to find C05D2.3 transcripts trans-spliced to SL2 if it is a downstream gene in an operon with bas-1. We were unable to isolate either SL1 or SL2-spliced transcripts from C05D2.3 by RT-PCR, although we did isolate a partial cDNA using internal primers. DNA microarray experiments suggest the gene is not expressed above background levels, unlike C05D2.4/bas-1 (Table 2). Furthermore, a global analysis of expression specifically designed to identify operons did not select C05D2.4 and C05D2.3 as likely members of an operon 45. Since genes comprising an operon should be expressed at similar levels, these data provide no support for the idea that bas-1 and C05D2.3 constitute an operon.

We compared the predicted amino acid sequences of five other C. elegans AADC-like genes revealed by deletion mapping 46 and by whole genomic sequencing 24, along with a previously identified C. elegans glutamate decarboxylase (GAD) gene, unc-25 47 to related PLP-dependent decarboxylases from other organisms. Some of the C. elegans genes are clearly closely related to other AADCs, whereas others are more divergent (Fig 5, Table 1). All contain the core conserved domain (PFAM 00282) defining this group of PLP-dependent decarboxylases. None of the AADC or GAD predicted proteins in C. elegans appears to have a signal sequence.

The protein predicted from K01C8.3 is now believed to encode a tyrosine decarboxylase (tdc-1) used for tyramine and octopamine synthesis, which both appear to be used as neurotransmitters in C. elegans 4849. The best match to K01C8.3/tdc-1 is a predicted Drosophila AADC-homologous protein of unknown function (G30446). Interestingly, K01C8.3/tdc-1 shows a stronger match to known DDCs than any of the other C. elegans AADCs, including C05D2.4 (Table 1), although it is equally similar to known histidine decarboxylases (HisDCs). The strongest match of C05D2.4/bas-1 (outside of nematodes) is to insect and mammalian DDCs, but again the match is only slightly better than to HisDCs. The predicted genes C05D2.3, F12A10.3 and ZK829.2 also have about the same level of identity and similarity to known AADCs and HisDCs. The ZK829.2 predicted protein, however, is much larger (830 AA) than a typical AADC, having extended N- and C-terminal domains not found in other PLP-dependent DCs. Most of ZK829.2 predicted coding sequence is confirmed by cDNA sequences, suggesting that the predicted protein 'extensions' likely are real.

The predicted gene C09G9.4 is the most divergent from known AADC's with only 20 – 24% amino acid identity; it is even more divergent than C.e. GAD/unc-25. It also appears to lack the absolutely conserved Lys of PLP-dependent decarboxylases, although it otherwise retains considerable homology with the conserved domain of this family of proteins. There are no similar proteins among other organisms to provide clues about a possible function for this gene; C09G9.4 is a truly novel member of the group II PLP-dependent enzyme family. Proteins with a similar level of divergence with AADC (~20% identity over a few hundred amino acids) include other group II PLP-dependent enzymes such as sphingosine-1-phosphate lyase and cysteine sulfinic acid decarboxylase. C09G9.4, however, has very little or no significant similarity to these other enzymes. The C. elegans GAD/unc-25 predicted protein has a strong match to identified Drosophila and mammalian GADs (Table 1), and is found as a single copy. There are no other GAD-like genes in C. elegans such as cysteine sulfinate decarboxylase, which is the rate-limiting enzyme in taurine synthesis, and the closest non-AADC relative to GAD in the vertebrates 50.

We performed BLAST searches of a C. briggsae whole genome shotgun assembly using predicted protein sequences of all six C. elegans AADC genes and the unc-25/GAD gene. We found five orthologous genes in C. briggsae – four AADC homologs and one GAD homolog (Table 2). All of these matches included 100% or near 100% identity over extended regions of aligned predicted amino acid sequences, and were paired with high confidence in phylogenies (Fig. 6). Using a core AADC sequence for alignments and tree building, we found that the bas-1 orthologs have evolved more quickly than some of the other AADC's. The C. elegans gene K01C8.3 and its ortholog, for example, are 98% identical in this core region (vs. 91% identity for bas-1 orthologs). Most of the divergence between K01C8.3 and its ortholog is in N- and C-terminal extensions that are not found in other AADC's. The C. elegans C09G9.4 and C. briggsae ortholog are even less similar to one another than are the bas-1 orthologs.

Our most striking observation is that C. briggsae appears to lack orthologs for the C. elegans predicted genes C05D2.3 and F12A10.3. This suggests that gene duplications giving rise to these two genes, which are most closely related to bas-1/C05D2.4, occurred either in the C. elegans lineage after its split with the C. briggsae lineage, or that C. briggsae lost both C05D2.3 and F12A10.3 orthologs (or their common ancestor) following the split. Using phylogenetic analysis of aligned amino acid and nucleotide sequences, we found that C05D2.3 and F12A10.3 share a common ancestor and that the gene duplication giving rise to bas-1 and C05D2.3/F12A10.3 likely occurred prior to the C. elegans/C. briggsae divergence. This is also suggested by the pattern of introns in the genes. [We have confirmed the splicing pattern of C. briggsae bas-1 by isolating a cDNA (S. DePaul & C. Loer, unpublished data).] The C. elegans and C. briggsae bas-1 genes have identical genomic structure which differs from that of C05D2.3 and F12A10.3, which are more similar to one another (Fig. 7). Therefore C05D2.3 and F12A10.3 (or their common ancestor) were retained in the line leading to C. elegans but lost in the C. briggsae line. The original duplication event giving rise to the tandem copies of C05D2.4 and C05D2.3 on chromosome III probably occurred via an unequal crossing-over or similar event. The duplication creating F12A10.3, which is found on chromosome II, presumably occurred subsequently. We noted no homology of other predicted genes downstream of F12A10.3 and C05D2.3 that might suggest an event duplicating more than the AADC gene.

The retention of the genes and their expression in C. elegans suggests that they may have acquired a new function that is under selection, retain a subfunction of the AADC, or instead that they are still in the process of being lost. After sequencing F12A10.3 cDNAs (courtesy of the ORFeome Project), we found that current splicing predictions for the gene were incorrect. We sequenced six F12A10.3 clones and found two slightly different splicing patterns, both different from Genefinder and Intronerator predictions. The two types of clones differed only in whether a final intron was removed or not. We found 4 clones with 9 exons, and 2 clones with 8 exons. The failure of the gene prediction programs in this case is likely to be due to their preference for creating functional transcripts. All F12A10.3 cDNAs instead appear to be non-functional: they have frameshifts relative to AADC-homologous reading frames. The first frameshift occurs in the second exon and quickly leads to a premature stop codon. At best F12A10.3 transcripts would result in a 158 amino acid protein that could not function as an AADC. F12A10.3 therefore appears to be an expressed pseudogene. DNA microarray experiments and representation in cDNA sequencing projects suggest that F12A10.3, like C05D2.3, is likely expressed at a low level (Table 2).

In order to assess whether the bas-1-like genes C05D2.3 and F12A10.3 might be under reduced selection pressure, we calculated the ratio of non-synonymous to synonymous substitutions (K_A/K_S) comparing the bas-1 orthologs and bas-1-like genes. We also calculated these values for the other AADC ortholog pairs from C. elegans and C. briggsae (Table 3). K_A/K_S < 1 indicates purifying (negative) selection, K_A/K_S = 1 indicates no selection (as in true pseudogenes), and K_A/K_S < 1 indicates Darwinian (positive) selection. K_A/K_S for 8179 C. elegans and C. briggsae ortholog pairs had a mean value of 0.06, indicating most genes are under purifying selection 37. We found that the bas-1 genes are under purifying selection (K_A/K_S = 0.039), but the bas-1-like genes appear to be under reduced selective pressure; the average K_A/K_S for comparisons with bas-1-like genes was 0.148, more than three times the value of the bas-1 ortholog comparison. The proportion of observed to potential non-synonymous substitutions (pN) among the bas-1-like gene comparisons was similarly much higher than for the bas-1 orthologs.

Two other AADC ortholog pairs showed strong purifying selection at work, with values like that calculated for the bas-1 orthologs (Table 3), but a value could not readily be calculated for the tdc-1 orthologs. In all the AADC ortholog comparisons, the proportion of observed to potential synonymous substitutions (pS) was near mutational saturation (pS > 0.75); K_A/K_S cannot be calculated when pS > 0.75. Thus, a value could not be calculated either for full-length tdc-1 alignments, or using a tdc-1 core sequence. We were able to calculate a value by aligning the N- and C-terminals sequence of the tdc-1 orthologs (Table 3). These regions of the protein are under levels of selection like the other AADCs, whereas the core has a very low rate of non-synonymous substitution, consistent with the high level of amino acid conservation in this region of the protein.

We compared the protein sequences of other predicted AADCs in C. elegans with those of other organisms in order to guess about their possible functions. This is particularly relevant because all bas-1 mutants retain weak, residual serotonin immunoreactivity (13; C. Loer, unpublished) suggesting that other enzymes may be able to carry out the same reaction. This would not be surprising since animal AADCs tend to have broad specificity 3. Based purely on sequence homology, it seems that predicted genes K01C8.3 and ZK829.2 could act as AADCs or as HisDCs. In fact, the predicted gene K01C8.3 is now believed to be a tyrosine decarboxylase and has been named tdc-1 48. If correct, then its best match in Drosophila (G30446), an uncharacterized AADC homolog, is likely to encode the fly's octopamine-synthetic tyrosine decarboxylase. It has long been known that a separate gene encoded this enzymatic activity in flies, since the activity is still detectable in Ddc deletion mutants 4. It will be interesting to see whether such tyrosine decarboxylases in animals have more restricted substrate specificity, such as the tyrosine and tryptophan decarboxylases in plants 54, or are more similar to typical animal AADCs with a broad specificity. Tighter substrate specificity of a tdc-1 protein could be reflected in the much slower rate of amino acid substitution seen in its C. elegans & briggsae orthologs than in the bas-1 orthologs which encode more 'promiscuous' enzymes.

Whether C. elegans or other nematodes make the neurotransmitter histamine, and therefore need a HisDC enzyme, is unclear. Although histamine has been reportedly isolated from C. elegans 55, this observation is unique among nematodes, and has not subsequently been confirmed. There is no particularly good candidate for a HisDC in C. elegans. The ZK829.2 predicted protein may be most closely related to tdc-1 in its core sequence, although its long N- and C-terminal extensions are perhaps suggestive of a new function. Unfortunately, transgenics with reporter fusions of this gene to date have shown no expression, the pattern of which might suggest a function (C. Loer, unpublished; M. Alkema, personal communication). As with tdc-1, C. elegans ZK829.2 and its C. briggsae ortholog have also evolved more slowly than the bas-1 orthologs. A recent analysis of eukaryotic AADC sequences that includes the C. elegans ZK829.2 and its C. briggsae ortholog as the only nematode representatives clearly demonstrates that AADC genes can evolve at very different rates, and that a constant "molecular clock" cannot be assumed in phylogenetic analyses 56.

Finally, since the C09G9.4 predicted protein is so highly divergent from the typical AADC, and lacks a critical lysine residue that binds the PLP cofactor, it is unlikely to be an AADC enzyme. It has a similar level of divergence from genuine AADCs as do other group II PLP-dependent enzymes such as cysteine sulfinic acid decarboxylase, to which it has little or no similarity. Whatever the function of a C09G9.4-encoded protein, it appears to represent a new PLP-DC-related protein; sequencing of more genomes may yet reveal additional members.

We found that the closest relatives of C05D2.4/bas-1 in C. elegans, the genes C05D2.3 and F12A10.3, are missing from C. briggsae. Furthermore, phylogenetic analysis indicates the two extra genes did not arise in the C. elegans line, but were present (or their commmon ancestor was present) in the species that gave rise to both the C. elegans and C. briggsae lines. Finally, careful examination of the cDNAs and predicted protein sequences of C05D2.3 and F12A10.3 reveals that neither is likely to be functional as an AADC: the former lacks critical amino acids and the latter can encode only a truncated protein. Both are expressed, based on the presence of cDNAs, but probably at a very low level, which is not above background in microarray experiments. It is possible that the duplicate genes are functionally 'lost' in C. elegans as well.

The features of C05D2.3 and F12A10.3 raise a number of interesting questions about the fate of duplicate genes, and the true nature of many 'predicted genes' in C. elegans. Taking a random sampling of predicted genes and generating transgenics with reporter fusion constructs (in order to determine a pattern of expression), Mounsey and colleagues 57 found that a much higher percentage of recently duplicated genes than conserved or unique genes failed to show expression. Assuming that failure of expression was no more likely among recently duplicated genes for technical reasons, this meant that many more of these are in reality not expressed. The numbers suggested that up to 20% of annotated, predicted genes in C. elegans may be pseudogenes. In fact, careful inspection of recently duplicated genes showed that many were actually pseudogenes, like we found to be the case for F12A10.3. Overall, close inspection of predicted genes revealed at least 4% were pseudogenes.

So, why are C05D2.3 and F12A10.3 still present in C. elegans if they lack a function? C. briggsae and C. elegans may have diverged 80 – 110 million years ago 3738. Since the bas-1-like gene or genes were likely present in the common ancestor of C. elegans and C. briggsae, then there seems to have been ample time for loss in the C. elegans line. Under a simple model of gene loss following duplication, only a few million generations would be the mean time to fix a null allele of the gene duplicate 58. In Caenorhabditis, a million generations could be completed in 10,000 years or less. This seems to suggest that the downstream duplicate of bas-1 (ancestor of C05D2.3) may have continued to function for a considerable time after the duplication, perhaps by gene conversion which might have continued until sufficient divergence from bas-1 59. Loss of the critical six amino acids occurred after the second duplication giving rise to the ancestor of F12A10.3, since the appropriate sequence is still present there (although frame-shifted). It is also possible that the C05D2.3 gene retains some function. The gene still encodes a respectable protein, albeit one that seems unable to function as an AADC. It has diverged considerably from bas-1, but has not accumulated stop codons and frameshifts expected for a pseudogene. Walsh 60 has proposed that fixation of an allele with an advantageous new function, vs. becoming a pseudogene, may be the fate of many duplicate genes even when such mutations are rare, given a population that is sufficiently large.

C05D2.3 and F12A10. 3 seem to have been retained longer than expected. Lynch and Force 61 proposed that the unexpectedly high rate of gene duplicate retention in eukaryotic genomes is due to 'subfunctionalization' – the retention of a portion of the original single gene's function by each of the duplicates, which then complement one another. Although this was suggested to occur primarily by regulatory mutations that partition expression of the genes spatially, other forms of subfunctionalization could also occur. Another possible reason for retaining such genes is the presence of non-coding regulatory functions associated with transcription and splicing of these sub-functional transcripts that affect the transcription of other nearby genes, although a bas-1::GFP construct is expressed well without such sequences in cis.

Our analysis of synonymous vs. non-synonymous substitutions indicates that the bas-1-like genes C05D2.3 and F12A10.3 are under relaxed selection relative to bas-1 and other AADCs. It should be noted that precise quantitative comparisons cannot be made with the results presented in the C. briggsae whole genome analysis 37, since we used a different method of calculating K_A/K_S; however our calculations indicate that bas-1 and the other AADC's, like most genes in the Caenorhabditis genomes, are under strong purifying selection. Even if both C05D2.3 and F12A10.3 are now pseudogenes, some significant period of time during which they functioned and were under purifying selection could act to obscure this fact in an analysis of K_A/K_S. Even if C05D2.3 has acquired a new, adaptive function, such a new function might result from changes in only a few sites in the protein, and so again this could be obscured by a majority of sites under purifying selection. With the sequencing of three related Caenorhabditis species it will be interesting to learn of the fates of bas-1 and the bas-1-like genes in other lines.

Mutations were identified by PCR-amplifying small regions from genomic DNA, then sequencing purified PCR product. We isolated genomic DNA (Puregene Kit, Gentra Systems) from four bas-1 mutant strains (alleles ad446, n2948, n3008, pa4) and then PCR-amplified small segments of C05D2.4 and C05D2.3 protein-coding sequence. Five pairs of primers were used to survey the C05D2.4 gene (A+B, E+F, G+H, P+Q, R+S; see below for primer sequences) and 4 pairs for C05D2.3 (C+D, K+L, M+N, P+Q). Bands of the predicted size were excised from 2% agarose gels and purified with GeneClean (Bio101/Qbiogene), then sequenced. All mutations were confirmed by sequencing both strands for two independent PCR reactions. All PCR and sequencing primers were designed using the program Primer3 (http://frodo.wi.mit.edu; 65). Primer sequences were as follows. Within C05D2.4: C05D2-A: gaggaaactcaaggcgacac; C05D2-B: tgttgatggaaccaagtgga; C05D2-E: cgtccttttctctttgcgac; C05D2-F: tggctccgacttgattctct ; C05D2-G: ttacaattaggccgcaaacc; C05D2-H: ccacctgaactgtggtgatg; C05D2-P: ggactcacatgtttccgattg; C05D2-R: ttagacgttggttgcacgag; C05D2-S: attggcgagcagtcaaagtt ; C05D2-T: tcttatgggattaccagaac; C05D2-U: ctacataaagctggaatggt; C05D2-V: gtttcctaaaaatccacgtg; C05D2-W: atgatcgattgatagctgag. Within C05D2.3: C05D2-C: ctaggtgcctttgccttctg; C05D2-D: caagagacgctcgttgtcag; C05D2-K: gccatctaatcctccaacca; C05D2-L: acattgctcccttttcaacg [note that primer L can also prime within C05D2.4]; C05D2-M: ccatcaactttccaatggct; C05D2-N: tctcgacgcccatatttctc; C05D2-Q: ccaattccagcggagaagta.

All DNAs for microinjection were purified with Qiagen tip20 columns. Experimental DNAs were co-injected with the dominant marker plasmid pRF4 containing the mutant gene rol-6(su1006) 66 into N2 (wildtype), CB1490, or bas-1 mutant worms. Progeny of injectees that express the rol-6 dominant plasmid have the easy-to-identify Roller phenotype which results in worms with a helically-twisted body along the anterior-posterior axis. Roller transgenic worms typically carry co-injected DNAs. Transgenic Roller progeny were isolated and propagated; rescue of bas-1 mutants was scored by staining with serotonin antiserum as described previously1667.

The rescuing plasmid C05D2XN contains a 15.8 kbp genomic DNA insert (XhoI to NheI) derived from the cosmid C05D2, and contains both C05D2.4 and C05D2.3 predicted genes. A number of deletions of C05D2XN were made to test for rescue of bas-1 mutants. Clones were analyzed by restriction digests. We also made C05D2XN derivatives in which either C05D2.4 and C05D2.3 was mutated to introduce a premature stop in coding sequence. Clones were sequenced to determine the nature of the introduced mutation. Clone pCL6991 had a 4 bp deletion in the second exon of C05D2.4; pCL7991 had a 2 bp insertion in the first exon of C05D2.3; each causing a frameshift mutation. In the clone pCL7003, derived from the rescuing plasmid pCL3001, almost the entire C05D2.3 coding sequence is deleted. The bas-1 GFP reporter construct (within transgenics) was kindly provided by Ian Hope, and consists of a PCR-generated fragment of C05D2 with 4595 bp upstream of the predicted translation start site and 403 bp protein coding region to make an in-frame protein fusion in the 2nd exon. The construct was created by a multi-site recombination reaction with the C05D2 fragment, PCR-generated GFP, and vector using the Invitrogen Gateway cloning system (I. Hope, personal communication) as continuation of a project to determine expression patterns for C. elegans genes through reporter gene technology 57.

We isolated RNA for RT-PCR from mixed stage CB1490 worms. (The CB1490 strain has ~30% males, whereas wildtype N2 has ~0.2% males. Since adult males have 13 more serotonergic neurons and 6 more dopaminergic neurons than hermaphrodites, we reasoned that these worms might express more bas-1 mRNA.) Worms were isolated from six 100 mm NGM plates, washed several times with M9 buffer, and pooled to form a ~100 μl pellet of worms. The pellet was mixed with 175 μl RNA lysis buffer (SV Total RNA Isolation System, Promega), frozen and ground to a fine powder with a pestle and mortar cooled with liquid nitrogen. The powder was recovered in a fresh microfuge tube, mixed with 350 μl SV RNA dilution buffer and centrifuged to remove debris. The cleared lysate was transferred to a new tube and precipitated with 200 μl 95% ethanol and applied to the SV spin column assembly. The remaining RNA purification was performed exactly as described for the SV System "RNA Purification by Centrifugation." Purified RNA was eluted from the spin column with 100 μl nuclease-free water. RNA was converted to cDNA in a 80 μl reaction containing 800 Units of M-MLV Reverse Transcriptase (Life Technologies/BRL), 16 μl 5X RT buffer, 25 mM dNTPs, 80 Units RNAsin (Promega) 2.0 μg random hexamer primers (Life Tech), and 50 μl purified RNA (from above). Bas-1(C05D2.4) and C05D2.3 cDNAs were amplified by PCR from him-5 cDNA produced as described above, typically using Failsafe PCR mixes (Epicentre Technologies). Conditions for PCR were 94°C (1 min.), 50°C (1 min.), 72°C (3 min.) for 40 cycles, then 72°C (10 min.). Primers used to amplify C05D2.4 cDNAs were SL1-B + C05D2-L and SL1-B + C05D2-B; a partial C05D2.3 cDNA was amplified with primers C05D2-M and -N. Spliced leader primer sequences were as follows: SL1-B: AAAGGATCCTTTAATTACCCAAGTTTGAG; SL2-B: AAAGGATCCTTTTAACCCAGTTACTCAAG. Appropriately-sized bands were isolated with GeneClean or Ultrafree-DA (Millipore), then cloned directly into the pCRII-TOPO vector using the TOPO-TA Cloning kit (Invitrogen). From seven of our RT-PCR derived clones we sequenced, we found two different splice variants different form the Genefinder prediction (see results).

We also obtained cDNA clones from the ORFeome project 40 and the C. elegans EST project (courtesy Yuki Kohara). DNA we received from the ORFeome project is purified from a pool of transformants derived from ligation and transformation of their original RT-PCR, thereby allowing the isolation of internal splice variants from the mix. We transformed with this DNA and isolated several clones. We sequenced two ORFeome clones completely; six more clones were partially sequenced. Interestingly, each of the 8 clones appeared to have at least one mutation (compared to known genomic sequence). Since we found one splice variant containing the 27 bp microexon from among the eight clones that we sequenced, we analyzed 29 additional ORFeome project-derived clones by PCR from single isolated bacterial colonies. Amplification of the region between primers C05D2-T and -L allowed us to distinguish between clones with or without the 27 bp microexon based on product size (464 bp versus 437 bp). Seven of 29 clones were the larger size, so likely contained the 27 bp microexon. None of the ORFeome clones analyzed by sequence or PCR appeared to use the alternative exon 3 splice acceptor noted above. Finally, the two 'YK' cDNA clones from the C. elegans EST project that we examined are from a 'full-length' capped library, and each has an SL1 leader and a poly-A tail. We found that each bas-1 'YK' clone, however, contained a different internal deletion (overall abnormality of these clones is reported at ~5% – J. Theirry-Meig, personal communication). Each of the deletions was adjacent to a short repeated sequence.

C. elegans genomic and predicted cDNA sequences were retrieved from ACeDB, WormBase, and/or GenBank. Blast searches and Blast2 comparisons were performed using the NCBI Blast server. C. briggsae genomic sequence was searched using TBLASTN of the 7/12/02 shotgun assembly http://www.sanger.ac.uk/Projects/C_briggsae/blast_server.shtml. Assembly and consensus sequence determination of our own cDNA sequences was done using the program SeqMan (DNAStar, Madison, WI). Some information about ORFeome clones was retrieved from WorfDB (http://worfdb.dfci.harvard.edu; 68). Phosphorylation predictions were made with NetPhos 2.0 http://www.cbs.dtu.dk/services/NetPhos/ and PhosphoBase http://www.cbs.dtu.dk/databases/PhosphoBase/predict/predict.html. Signal sequence predictions were performed with SignalP v.1.1 (http://www.cbs.dtu.dk/services/SignalP/, 69). Multiple sequence alignments were performed primarily with CLUSTALW http://www.ebi.ac.uk/clustalw/, with some manual adjustments. Phylogenetic analyses were performed with PAUP* version 4.0b10 (Sinauer Associates, 70). Some alignments of cDNAs with genomic DNA to determine intron locations were performed with SIM4 71. Estimates of rates of synonymous and non-synonymous substitutions were made with SNAP (Synonymous/Non-synonymous Analysis Program – http://www.hiv.lanl.gov/content/hiv-db/SNAP/WEBSNAP/SNAP.html using pairwise or multiple sequence nucleotide alignments generated with CLUSTALW and adjusted manually. This program uses the method of Nei and Gojobori 72.