stress sensitive B Encodes an Adenine Nucleotide Translocase in Drosophila melanogaster

Corresponding author: Michael Ashburner, Department of Genetics, University of Cambridge, Downing St., Cambridge CB2 3EH, England., ma11{at}gen.cam.ac.uk (E-mail)

Communicating editor: V. G. FINNERTY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Adenine nucleotide translocases (ANT) are required for the exchange of ADP and ATP across the inner mitochondrial membrane. They are essential for life, and most eukaryotes have at least two different Ant genes. Only one gene had been described from Drosophila, and this had not been characterized genetically. We show that mutations in this gene correspond to the previously described loci, sesB and l(1)9Ed. Immediately adjacent to this gene is another encoding a second ANT protein, which has 78% identity to that encoded by sesB/l(1)9Ed. These two genes are transcribed from a common promoter, and their mRNAs are produced by differential splicing. Hutter and Karch suggested that the sesB ANT gene corresponded to Hmr, a gene identified by an allele that rescues otherwise inviable interspecific hybrids between Drosophila melanogaster and its sibling species. This hypothesis is not supported by our study of the ANT genes of D. melanogaster.

HYBRIDS between Drosophila melanogaster and its sibling species, D. simulans, D. mauritiana, and D. sechellia, are either sterile or inviable. Male hybrids from crosses between D. melanogaster females and sibling species males die as late larvae or early pupae, while female hybrids from the reciprocal crosses die as embryos. Mutations or strains of D. melanogaster and D. simulans that rescue either the embryonic or the larval/pupal lethality of interspecific hybrids have been characterized. One of these is Hybrid male rescue (Hmr), a mutation of D. melanogaster that rescues to viability otherwise inviable male hybrids from the cross of D. melanogaster females to sibling species males (HUTTER and ASHBURNER 1987 ; HUTTER et al. 1990 ; SAWAMURA et al. 1993 ; HUTTER 1997 ). Hmr was mapped to 9D1-9E4 on the polytene chromosomes (HUTTER et al. 1990 ), close to the distal breakpoint of In(1)AB, an inversion that also rescues interspecific hybrid inviability (HUTTER et al. 1990 ). HUTTER and KARCH 1994 mapped this In(1)AB breakpoint just distal to a gene encoding a protein very similar to adenine nucleotide translocases (ANT, ADP/ATP translocase) of other organisms, and they considered this gene to be a candidate for Hmr/In(1)AB. We have restudied the relationship between the adenine nucleotide translocase gene, Hmr, and In(1)AB. We cannot confirm any effect of either Hmr or In(1)AB on ANT. Moreover, we show that mutations in two previously characterized genes, sesB and l(1)9Ed, map to the adenine nucleotide translocase, and neither these nor transformants carrying extra copies of the wild-type sesB gene have any effect on the viability of interspecific hybrids.

sesB is one of a series of stress-sensitive mutants (sesA–sesH) characterized by Homyk (HOMYK and SHEPPARD 1977 ; HOMYK et al. 1980 ); other bang-sensitive mutants had been recovered previously in Benzer's laboratory (BENZER 1971 ). Some of these have already been cloned (ROYDEN et al. 1987 ; PAVLIDIS et al. 1994 ; SCHUBIGER et al. 1994 ; ANDREWS et al. 1996 ), but the mechanisms underlying their phenotypes are largely unknown (PAVLIDIS and TANOUYE 1995 ). The cloning of sesB demonstrates that a bang-sensitive phenotype can result from a defect in energy metabolism. The Drosophila adenine nucleotide translocase genes are duplicated tandemly. They are 72% identical in nucleotide sequence and 78% identical in amino acid sequence. Remarkably, these two proteins are produced by alternative splicing of a dicistronic primary transcript; these mRNAs share a 5' untranslated exon. It is clear that the adenine nucleotide translocase gene is not Hmr, which remains to be clarified molecularly, but is sesB.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Stocks and chromosomes:
sesB¹ was obtained from T. Homyk and l(1)DC701 (sesB^9Ed-6) from N. Perrimon. l(1)A17 (sesB^9Ed-1), l(1)G16 (sesB^9Ed-2), l(1)H22 (sesB^9Ed-3), l(1)M27 (sesB^9Ed-4), l(1)Q10 (sesB^9Ed-5), l(1)H6 (ras^l17), l(1)J5 (ras^l19), l(1)E20 (l(1)9Ec²), and l(1)G4 (l(1)9Ec³), as well as the chromosome aberrations Df(1)ras59 and Df(1)RJ7, were from the Edmonton collection. Both the sesB^9Ed-1 and sesB^9Ed-2 chromosomes carried extraneous X-linked lethal mutations; that from sesB^9Ed-2 was removed by recombination. ZHIMULEV et al. 1987 showed that l(1)S12 (LEFEVRE 1971 ) is an allele of sesB. The only allele of "l(1)9Ee" available to us, Q21, was lethal with all sesB alleles; we do not know whether this is because of a stock mix-up at some stage or some other reason, but we note that JANCA et al. 1986 recovered only one allele of l(1)9Ee and were unable to separate it from alleles of l(1)9Ed by deletion mapping. The chromosome aberrations In(1)AB, In(1)AC2, Df(1)v64f, Df(1)HC133, Dp(1;2)v^+75d, Dp(1;3)v^+74c, and Dp(1;2)v⁺⁶³ⁱ were obtained from the Bloomington Stock Center or were in the Cambridge collection. All lethal alleles and aberrations are described in FLYBASE 1999 . The synthetic duplication Dp(1;1)AB^LAC2^R (1A-9E|13E-9D|13B-20F) and its reciprocal deletion, Df(1)AC2^LAB^R (1A-9D|13B-9E|13E-20F), were made as recombinants between In(1)AC2 and In(1)AB, and were verified by polytene chromosome cytology. The ras^- deletions ras-TM3 (new order: 1-9A3.5|52F3.9-21; 60-52F3.9|13D3.4-9E6|13E-20), ras-TM5 (new order: 1-9E2.3|13B3.4-9E7.8|13E1.2-20), and ras-TM6 (new order: 1-9C1|13C3-9E7.8|13E1.2-20) were induced by X rays on In(1)AB (D. COULSON, personal communication). The deletion Df(1)B13 (Df(1)9D1-2;9E4-10) was identified by D. COULSON (personal communication) in the l(1)B13 stock of D. Nash.

A new sesB allele, sesB⁶, was isolated from a P-element screen for lethal mutations on In(1)AB in which P{ry^+t7.2=HBDelta-23}9E (from the Bloomington Stock Center) was mobilized; the chromosomes were screened for lethals against Df(1)HC133.

Crosses:
All crosses were done at 25°, unless stated otherwise, and flies were reared on standard yeast-cornmeal medium. Care was taken to ensure that all eclosed flies were scored when scoring lethal complementation tests.

Cytology:
Polytene chromosome cytology was performed on propionic carmine-orcein squash preparations of salivary gland chromosomes, which were interpreted using the revised maps of Bridges (see LEFEVRE 1976 ).

Bang-sensitivity assay:
Bang sensitivity was assayed essentially according to GANETZKY and WU 1982 . Individual flies were transferred to an empty vial and immediately vortexed for 10 sec using a Vortex-Genie 2 (Scientific Instruments Ltd.) at its highest setting. The time for each fly to right itself after vortexing was recorded.

Clones:
Cosmid 152F6 was obtained from the European Drosophila Mapping Consortium (MADUENO et al. 1995 ). A sesB cDNA (HE50) and other clones were the gift of P. Hutter.

DNA sequencing:
The sequence of the wild-type ANT genes was determined from cosmid 152F6. Mutant alleles were sequenced from adult genomic DNA PCR products amplified with Vent DNA polymerase (New England Biolabs, Beverly, MA). For the sequencing of sesB, PCR reactions were done with primers S5 and S3, corresponding to the nucleotides 5168–5187 and 6270–6291, respectively, of the sequence EMBL:Y10618 (all sequence coordinates are with respect to this record). The PCR products were cloned into pBluescript by PCR-introduced HindIII and XhoI sites in S5 and S3, respectively. Commercial forward and reverse primers for pBluescript and a synthesized internal primer, SM (nucleotides 5464–5484), were used for sequencing the full-length coding region of sesB.

To sequence the gene immediately distal to sesB (Ant2), two primers flanking the coding region were designed, SU (nucleotides 7291–7308) and SD (nucleotides 8646–8664). PCR products were cloned in pBluescript by primer-introduced EcoRI and XbaI sites. Two internal primers, S (nucleotides 7797–7817) and 2.8A (nucleotides 7823–7842), were used to obtain the full sequence of the 1.4-kb EcoRI-XbaI fragment.

To sequence the P-element insertion associated with sesB⁶, we used a protocol and primers designed for the PZ vector (REHM 1998 ). Total DNA was digested with CfoI, ligated, amplified with the primers Plac4/Plac1 and sequenced with primer SP1 (for 5'-end rescue), and amplified with primers Pry1/Pry4 and sequenced with primer Spry3 (for 3'-end rescue).

The breakpoints of two chromosome aberrations were sequenced by inverse PCR. To sequence the distal breakpoint of In(1)AB, DNA from a homozygous stock was digested with PstI, ligated, and amplified with primers corresponding to (1) the reverse complement of Y10681:7514–7533 (AB1) and (2) bases 8830–8849 of this sequence (AB2). The 0.5-kb PCR fragment was purified and sequenced. DNA was prepared from Df(1)v64f/FM7 females, digested with EcoRI, ligated, and amplified with primer AB1 and one corresponding to bases 10250–10270 of Y10168. The expected 0.8-kb band was gel purified and sequenced (amplification from the balancer chromosome gave the expected 1.7-kb product).

Sequencing was performed by using an ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer, Norwalk, CT) and an ABI 377 sequencer (Applied Biosystems, Foster City, CA).

Sequence analysis:
General sequence analyses were done with the GCG package running on a Silicon Graphics workstation on the University of Cambridge Molecular Biology server. Sequence alignments were done using CLUSTAL W (version 1.74) from the EBI server (http://www2.ebi.ac.uk/clustalw/), and gene trees were built using the PHYLIP package (version 3.5c; FELSENSTEIN 1988 ) on the Cambridge Molecular Biology server.

Rapid amplification of cDNA ends:
Drosophila cDNA libraries were the gift of N. Brown. Adenine nucleotide translocase cDNA ends were amplified with one primer homologous to the vector pNB40 (GGTGACACTATAGAATACAAG for 5'-end amplification, TTAATGCAGCTGGCTTATCG for 3'-end amplification) and another homologous to an internal sequence of the adenine nucleotide translocase cDNA. For the amplification of the 5' ends of sesB, a vector primer and primer Rb (complementary to the nucleotides 5442–5460) were used. The 3' ends of sesB were amplified with a vector primer and primer Ra (nucleotides 5191–5210), introducing a BamHI site just before the initiation codon ATG. For the amplification of the 3' ends of Ant2, primer 2.8A (described above) was used. For the amplification of the 5' end of Ant2, primers corresponding to nucleotides 7393–7413 and 7427–7444 were used.

Germline transformation:
The genomic transformation construct was made as follows: The 10.3-kb EcoRI-HindIII fragment (see Fig 2) was cloned into pP{CaSper-hs}. This plasmid, pP{Ant2⁺ sesB⁺10.3} (at 1 µg/µl), was coinjected with plasmid pHS2-3 (at 0.5 µg/µl) in Spradling buffer into embryos from a y w strain.

View larger version (23K): In this window In a new window Download PPT slide   Figure 1. Genetic map of the sesB region of the X chromosome. The genetic extents of duplications and deficiencies, as determined by complementation analysis, are shown with their cytological extents [data from FLYBASE 1999 and this study].

View larger version (12K): In this window In a new window Download PPT slide   Figure 2. The organization and structure of adenine nucleotide translocase genes in D. melanogaster. A restriction map of the 10.3-kb region that was sequenced and used for transformation is shown at the bottom of the figure. Abbreviations for restriction enzymes are as follows: B, BglII; E, EcoRI; H, HindIII, P, PstI; X, XbaI. The structures of the two tandemly arrayed adenine nucleotide translocase genes are shown above the restriction map; the chromosomal-distal gene is Ant2, and the chromosomal-proximal gene is sesB. The two arrows above the 5' untranslated region (UTR) exons indicate the direction of transcription from the two promoters. The solid boxes are coding regions of exons, and the shaded boxes represent the UTR regions of exons; the two different 3' UTRs of sesB are indicated by a difference in shading. The positions of the distal breakpoints of In(1)AB and Df(1)v64f are indicated, as determined by sequencing. The positions of the four mutations of sesB that we have sequenced are indicated. The sequence of this DNA is under EMBL accession no. Y10618. The HUTTER and KARCH 1994 sequence (EMBL:S71762) is of the distal sesB transcript, with the more distal 3' end. It starts at base 1076 (relative to the right-hand HindIII site) and ends at base 6988; the LOUVI and TSITILOU 1992 cDNA sequence is (after removal of the terminal EcoRI recognition site sequences from its end) of the proximal sesB transcript, with the more proximal 3' end; it starts at base 3585 and ends at base 6827. The BDGP LD14425.5' sequence is of the distal sesB transcript, starting at base 1086 and ending (prematurely, we presume) at base 6749 within the 3' UTR. The BDGP GM14781.5' sequence is of the proximal sesB transcript; it starts at base 3529 and ends (prematurely) within the third coding exon at base 5800. The BDGP GM13259.5' sequence is of Ant2; it starts at base 1100 and ends (prematurely) at base 8336, within the third coding exon of Ant2. The PCR product of Ant2 amplified from an imaginal disc cDNA library (see MATERIALS AND METHODS) corresponds to bases 1076–1174 and 7338–7392 of the genomic sequence with three mismatches, presumably errors introduced by the PCR reaction.

General molecular methods:
General methods for handling nucleic acids were according to protocols in ASHBURNER 1989 and SAMBROOK et al. 1989 . Southern and Northern filters were washed at low stringency (2x SSC, 0.5% SDS at 65°).

Sequence data:
The genomic sequence is deposited in the EMBL Nucleic Acid Sequence Data Library under accession no. Y10618. The sequences across the breakpoints of Df(1)v64f and In(1)AB are deposited as EMBL:AJ236836 and EMBL:AJ236835, respectively. We also used the two existing cDNA sequences of sesB, those of LOUVI and TSITILOU 1992 (EMBL:S43651) and HUTTER and KARCH 1994 (EMBL:S71762).

The expressed sequence tag (EST) sequence data of the Berkeley Drosophila Genome Project (BDGP) have been very useful. These partial cDNA clone sequences have been computationally clustered into "clots" on the basis of their sequence overlaps. These clots are available from http://www.fruitfly.org/EST/. The sequence of any one clot is not necessarily stable; it may change as new EST sequences are analyzed. Moreover, at the moment, there is not any stable way to refer to a particular clot sequence. We refer to them by the name of one of their constituent clones, and this will ensure that the latest version of any particular clot can be recovered by this clone name. We have also archived the clot sequences recovered from the Berkeley server for analysis on April 11, 1998, and these are available on request from M.A. For the ANT genes, there were three EST clots on this date. These are consensus sequence 1 (represented by clone GM13259.5', EMBL:AA803475), a 700-bp sequence of only one clone; consensus sequence 2 (represented by GM14781.5', EMBL:AA821173), a 550-bp consensus of four clones; and consensus sequence 3 (represented by LD14425.5', EMBL:AA439770), a 1500-bp consensus of 59 different EST clones.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

This study began with the working hypothesis that the two mutations known to rescue the lethality of interspecific hybrids between D. melanogaster and its sibling species, Hmr and In(1)AB, were allelic (see HUTTER et al. 1990 ) and that the corresponding gene was that encoding an adenine nucleotide translocase (HUTTER and KARCH 1994 ). Hmr had been mapped genetically to 1-31.84, approximately corresponding to region 9D1-9E4 on the polytene chromosome map (HUTTER et al. 1990 ). The distal breakpoint of In(1)AB maps to 9E7-8. A gene encoding an adenine nucleotide translocase had been characterized from D. melanogaster by LOUVI and TSITILOU 1992 and mapped to 9E by in situ hybridization to polytene chromosomes. HUTTER and KARCH 1994 showed that the distal breakpoint of In(1)AB was just distal (i.e., 3') to this gene.

Hmr maps distal to the distal In(1)AB breakpoint:
HUTTER and ASHBURNER 1987 originally mapped Hmr to 1-31.84, relative to lz and v. A small-scale experiment recovered Hmr-ras recombinants (HUTTER et al. 1990 ), although none were recovered in a second experiment reported by HUTTER and KARCH 1994 . To resolve the issue, we scored 2687 rescued male hybrids from the cross of D. melanogaster y Hmr v/ras dy females to D. mauritiana males (at 18°). Six phenotypically ras males (crossovers in the Hmr-ras interval) and 14 males that were neither ras nor v (crossovers in the ras-v interval) were recovered, placing Hmr at 1-32.0, unambiguously distal to ras (at 1-32.35 in 9E4) and, therefore, distal to the distal In(1)AB breakpoint (at 9E7-8).

sesB maps proximal to the distal breakpoint of In(1)AB and corresponds to l(1)9Ed:
There have been several screens for lethal mutations mapping to region 9-10 of the X chromosome in addition to several mutations recovered by Lefevre (LEFEVRE 1981 ; LEFEVRE and WATKINS 1986 ) from more general screens for X-linked lethals. JANCA et al. 1986 defined eight lethal complementation groups within bands 9E1-9F13. We have obtained all extant mutations and aberrations in the 9D-9F chromosome region and have mapped these by complementation (Fig 1). For our present purposes, the critical data are (i) that the stress-sensitive allele, sesB¹ (HOMYK and SHEPPARD 1977 ) is allelic to mutations that define l(1)9Ed of LINDSLEY and ZIMM 1992 , which is group B of JANCA et al. 1986 (Table 1) and (ii) that this locus (which we call sesB) is proximal to the distal breakpoint of In(1)AB. The last fact is most convincingly shown by the complementation of sesB alleles by the synthetic deletion Df(1)AC2^LAB^R (Table 1) and by the fact that both ras^- deletions that we have induced with X rays on In(1)AB are sesB⁺ (Fig 1).

Cytologically, the distal breakpoint of In(1)AB maps to 9E7-8. We place sesB at the same cytogenetic position, not only from the molecular mapping of this inversion breakpoint (HUTTER and KARCH 1994 ; see below), but also because sesB is included in both Df(1)v64f (Df(1)9E7-8;10A1-2) and Df(1)ras217 (Df(1)9A;9E7-8), (data of JANCA et al. 1986 ).

The sesB phenotypes:
A total of 13 mutant alleles of sesB have been described (data from FlyBase), although we have been able to find only 7 now remaining in stock. Of the original 13, 11 were recovered as X-linked lethal mutations and 2 as viables with a stress-sensitive phenotype (HOMYK and SHEPPARD 1977 ; HOMYK et al. 1980 ). The original behavioral mutation sesB¹ was recovered on the basis of its reduced flight ability (SHEPPARD 1974 ) and was then shown to be stress sensitive, i.e., reversibly paralyzed when subjected to a mechanical shock (HOMYK 1977 ; HOMYK and SHEPPARD 1977 ). Even in the absence of any external shock, sesB¹ flies may knock themselves out if they fall down (HOMYK 1977 ). After "vortexing" for 10 sec, 1- to 2-day-old sesB¹ adult males are paralyzed for 38 ± 51 sec (n = 29); for 6-day-old males, paralysis lasts >100 sec; and 6-day-old homozygous adult females are paralyzed for 71 ± 53 sec (n = 20). After recovery from paralysis, the flies are not hyperactive, as is seen in some "bang-sensitive" mutations (PAVLIDIS and TANOUYE 1995 ), but they are refractory to further paralysis for a period of 5–10 min, as observed in other bang-sensitive mutations (GANETZKY and WU 1982 ). From our observations on sesB¹, the flies are generally rather hypoactive and have delayed development. sesB^9Ed-4 flies are very hypoactive and, as a result, are difficult to assay for stress sensitivity by observing their recovery from paralysis. However, we think that this allele is stress sensitive, since females heterozygous for sesB^9Ed-4 and Df(1)HC133 are extremely sensitive to shock and are paralyzed for >100 sec after vortexing. Neither of the two other male-viable alleles (sesB^9Ed-2 and sesB^9Ed-5) is stress sensitive.

We have examined eight sesB alleles. Three of these, sesB^9Ed-2 (G16), sesB^9Ed-4 (M27), and sesB^9Ed-5 (Q10), were described by NASH and JANCA 1983 as "haplo-specific," i.e., lethal as hemizygous females but viable as homozygous females; these were interpreted by NASH and JANCA 1983 as hypomorphic alleles. We have confirmed these data and have shown in Table 1 that sesB¹ is itself in this class, and when hemizygous with Df(1)HC133, it is semilethal. The viability data for homozygous females and hemizygous males (Table 1) suggest an allelic series with sesB^9Ed-2 being the weakest of the "lethal" alleles and sesB^9Ed-1, sesB^9Ed-3, and sesB^9Ed-6 being the strongest. When hemizygous with the deficiency Df(1)HC133, most sesB alleles are lethal in females, with only sesB¹ and sesB^9Ed-4 showing any escapers (Table 1).

sesB encodes an adenine nucleotide translocase:
The facts that the sesB locus maps proximal to the distal breakpoint of In(1)AB and that this breakpoint maps just distal to the 3' end of genes encoding an adenine nucleotide translocase (see below) led us to examine whether or not the sesB mutations mapped to the Ant translocase genes. This was done both by sequencing the gene from four mutant alleles and by transformation rescue of the mutant phenotypes.

To avoid the introduction of errors during PCR reactions, multiple PCR products amplified with the high-fidelity polymerase Vent were cloned independently, and several clones were sequenced. Three EMS-induced alleles sequenced showed substitutions in sesB. For sesB¹, all six sequenced clones (from three independent PCR reactions) showed a missense mutation in codon 289, CTT to TTT, which would result in a leucine-to-phenylalanine substitution in the putative sixth transmembrane domain of the protein. L289 is conserved in 30/35 available adenine nucleotide translocase sequences; the exceptional five are all from yeasts. The lethal allele sesB^9Ed-1 has an alanine-to-threonine missense mutation in codon 144, GCT to ACT (seven clones from two independent PCR reactions were sequenced, and four of them showed the mutation; as the DNA amplified was from a balanced stock, this is expected). A144 is conserved in 34/35 available adenine nucleotide translocase sequences [curiously, the exception is the protein from another insect, Anopheles gambiae (BEARD et al. 1994 )]. The lethal allele sesB^9Ed-6 has a nonsense mutation at codon 112, TGG to TGA (6 of 12 clones from 3 independent PCR reactions). These two lethal alleles, sesB^9Ed-1 and sesB^9Ed-6, were recovered from the same y cv v f progenitor stock after EMS mutagenesis (NASH and JANCA 1983 ).

The insertion site of the P element in the newly induced P-element allele, sesB⁶, was also sequenced (see MATERIALS AND METHODS). The P element is inserted in the first intron (between nucleotides 3917 and 3919) of sesB/Ant2, with the element in the opposite orientation to sesB/Ant2. In addition to the deletion of nucleotide 3918 of sesB, this insertion is associated with 47 bp of unidentified sequence between the sesB sequence and the 3' end of the P element, as well as 58 bp of unidentified sequence between sesB and the 5' end of the P element.

Transformation rescue:
A 10.3-kb EcoRI-HindIII genomic DNA fragment (cloned in pP{CaSper-hs}) that contains the entire adenine nucleotide translocase gene as well as its distal relative (Fig 2; see below) was transformed. Two independent insertions, P{Ant2⁺ sesB⁺10.3}42A and P{Ant2⁺ sesB⁺10.3}102E, were recovered. Both rescue the male lethality of lethal alleles of sesB (Table 2) and the stress-sensitive and hypoactive phenotypes of sesB¹ and sesB^9Ed-4.

Taken together, the sequencing and transformation rescue data unambiguously show that mutations in the adenine nucleotide translocase are responsible for both the behavioral and lethal sesB phenotypes.

Transformants for sesB do not affect interspecific hybrid viability:
HUTTER et al. 1990 showed that Dp(1;2)v^+75d suppressed hybrid rescue by Hmr, presumably because this duplication carries Hmr⁺. To determine whether an extra copy of sesB⁺ would suppress the rescue of D. melanogaster x sibling hybrid males by either Hmr or In(1)AB, we compared the viability of rescued males with or without P{Ant2⁺sesB⁺ 10.3}42A. No differences were found (Table 3); in contrast, Dp(1;2)v^+75d (which carries Hmr⁺) has been shown to completely suppress the rescue of hybrid males by Hmr (HUTTER et al. 1990 ; J. ROOTE and M. ASHBURNER, unpublished observations). In addition, we crossed D. melanogaster females carrying both P{Ant2⁺sesB⁺ 10.3}42A and P{Ant2⁺sesB⁺ 10.3}102E to D. mauritiana males (at 18°); no hybrid male adults were recovered, although the experiment was small (64 females). These data indicate that neither sesB nor Ant2 interact with the genetic factors that influence interspecific male viability.

A second adenine nucleotide translocase-like gene is contiguous to sesB:
Both LOUVI and TSITILOU 1992 and HUTTER and KARCH 1994 had suggested, on the basis of Southern DNA blots, that there was more than one gene encoding adenine nucleotide translocase in D. melanogaster. If so, then the in situ hybridization data indicate that these must both be in region 9E, since neither these studies nor our own showed any indication of other sites of hybridization with ANT cDNA probes. Indeed, on the basis of Southern blot data, HUTTER and KARCH 1994 considered the possibility of tandemly duplicated ANT genes in Drosophila. We sequenced both the 4.4- and 2.5-kb EcoRI fragments that include sesB (Fig 2) and thereby discovered a second open reading frame 1104 bp distal (that is, 3') to sesB. This ORF is similar in sequence (and in intron/exon structure, see below) to sesB. The coding regions of the two genes show 74% nucleic acid sequence identity and 78% protein sequence identity (Fig 3). Our evidence (below) is that this new gene is functional, and we named it Ant2 (adenine nucleotide translocase 2). By hybridization of whole genomic DNA with probes to these genes, there is no indication of further genes of similar sequence in the genome of D. melanogaster.

View larger version (72K): In this window In a new window Download PPT slide   Figure 3. Sequence alignment of the protein-coding regions and internal introns of sesB and Ant2. The exons are in capital letters, the introns in lowercase; gaps are indicated by periods, stop codons by asterisks. The translation of the ANT2 protein is shown. The single-letter amino-acid code is used here and in Fig 6. There are seven amino-acid differences between this sesB sequence and those of LOUVI and TSITILOU 1992 (EMBL:S43651) and HUTTER and KARCH 1994 (EMBL:S71768); in five of these cases, the two latter sequences are in agreement. In addition to checking our own data, we have studied the full-length cDNA sequence assembled from 59 independent EST sequences by the BDGP (LS14425.5', see MATERIALS AND METHODS). In all cases except one, our sequence agrees with this assembled cDNA sequence; the exception is residue 270, Gly in our sequence and Asp in the BDGP sequence. Residue 270 is one of two (the other is residue 202) that are absent from both S43651 and S71768.

The relationship between Ant2 and sesB:
Despite their sequence similarity, sesB and Ant2 appear to be functionally distinct. The two genes cannot, of course, be fully functionally redundant; otherwise, the frequent recovery of lethal mutations by EMS in sesB would not be possible. We have sequenced Ant2 in four sesB alleles (sesB¹, sesB^9Ed-1, sesB^9Ed-3, and sesB^9Ed-5) and found no amino-acid substitutions in these.

Df(1)v64f ends within the 3'-untranslated region of sesB, removing the 5' untranslated exon of Ant2 but not affecting the Ant2-coding region. The viabilities of sesB alleles, when heterozygous with this deletion, are similar to those seen when they are heterozygous for the longer deletion Df(1)HC133, which removes both sesB and Ant2 in their entirety (Table 1).

The distal breakpoint of In(1)AB was also mapped by Southern blot hybridization; it is in the distal part of the 1.8-kb XbaI-PstI fragment (see Fig 2). The breakpoint was sequenced by inverse PCR and was found to be 314 bp 3' to the end of the Ant2 transcript (at position 9288 on EMBL:Y10618). In(1)AB is fully viable when heterozygous with deletions of this chromosome region or when homozygous. The synthetic deletion Df(1)AC2^LAB^R, which extends distally from the In(1)AB distal breakpoint, is completely viable when heterozygous with lethal alleles of sesB (Table 1). The distal breakpoint of Df(1)v64f was also determined by inverse PCR; it is at position 7033 on the sequence EMBL:Y10618, i.e., in the short "intergenic" region between sesB and Ant2.

The structure of the two ANT genes and their expression:
Comparison of cDNA and genomic sequences show that sesB has two alternative transcription starts, and that each transcript is interrupted by an intron in its 5' UTR (Fig 2). The coding region of sesB is interrupted by small introns within codons 99 and 142. The genomic structure of the coding region of Ant2 is similar to that of sesB, with two introns interrupting at positions identical to those within the coding sequence of sesB. The cDNAs previously sequenced by LOUVI and TSITILOU 1992 and HUTTER and KARCH 1994 are clearly derived from sesB, although there are some differences between these and our sequences (see legends to Fig 2 and Fig 3). The HUTTER and KARCH 1994 cDNA and BDGP LD14425.5' sequences are of the distal transcript, and the LOUVI and TSITILOU 1992 cDNA and GM14781.5' sequences are of the proximal transcript.

No cDNA of Ant2 was recovered in our screens of cDNA libraries. The BDGP EST collection, however, includes a partial 5' cDNA of this gene, GM13259.5' (the fact that there was only one cDNA of Ant2 vs. 63 for sesB gives some clue as to the relative abundance of Ant2 and sesB transcripts). Bases 76–700 of GM13259.5' match the corresponding region of the predicted Ant2 cDNA sequence with only five base mismatches. We were puzzled to find that the first 75 bp of the GM13259.5' sequence matched no sequence in the 1.1-kb sesB-Ant2 intergenic region, where the Ant2 promoter might reasonably be expected to be found. Remarkably, a sequence with 100% identity to these first 75 bp is found in the 5' untranslated exon sequence of the distal sesB transcript (see Fig 2). We conclude that sesB and Ant2 are transcribed from the same promoter and share a common 5' untranslated exon. The dicistronic structure of the 5' end of this transcript was confirmed by PCR amplification of the predicted sequence from an imaginal disc cDNA library (see MATERIALS AND METHODS and legend to Fig 2).

Using a 4.4-kb genomic EcoRI fragment containing Ant2 and the 3' end of sesB, two major transcripts of 1.6 and 1.2 kb are seen in the Northern blots (see also LOUVI and TSITILOU 1992 ). Both are seen at all developmental stages assayed, although there are clear variations in their levels. These two transcripts are also seen using a sesB cDNA probe from HUTTER and KARCH 1994 . With a probe corresponding to the 3'-untranslated region of Ant2, a single transcript of 1.6 kb is seen; its abundance in adult flies is considerably less than that of the 1.6-kb sesB transcript (Fig 4). We have not detected any sign of a 2.8-kb transcript, as seen in the wild type by HUTTER and KARCH 1994 . It was the absence of this transcript in In(1)AB and its reduced level in Hmr that led these authors to conclude that In(1)AB was mutant for the adenine nucleotide translocase.

View larger version (123K): In this window In a new window Download PPT slide   Figure 4. Northern blot hybridization of total RNA from adult flies with probes to sesB (HE50 cDNA), Ant2 (3' UTR), and RpL3 (loading control; CHAN et al. 1998 ). The lanes are as follows: (1) Canton-S, (2) Hmr, and (3) In(1)AB. The approximate sizes of the transcripts are indicated in kilobases on the right; the RNA size markers (Promega) are indicated on the left.

The two transcripts seen by the Northern blot analysis result, at least in part, from the alternative 3' ends of sesB, as shown by RACE analysis of sesB cDNAs. Only a single 450-bp product was found from the amplification of the 5' end of sesB cDNA, but two products of 1.1 and 1.6 kb were found from amplification of the 3' end (Fig 5). All three RACE bands were sequenced, producing the expected results. When used as a probe to a Northern blot, the 1.6-kb RACE product only hybridizes to the 1.6-kb transcript.

View larger version (30K): In this window In a new window Download PPT slide   Figure 5. RACE analysis of a sesB cDNA isolated from an imaginal disc cDNA library. Lanes 1 and 2 represent 5' and 3' products, respectively. Sizes are marked by arrows. The primers used for 5'- and 3'-end amplification are described in MATERIALS AND METHODS. Lane 3 is a 1-kb DNA ladder from GIBCO BRL (Gaithersburg, MD).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The relationship between adenine nucleotide translocase and interspecific hybrid lethality rescue:
This study began with the working hypothesis (HUTTER and KARCH 1994 ) that some mutations that rescue the lethality of hybrids between D. melanogaster and its sibling species may be the result of changes in an adenine nucleotide translocase gene. Several lines of evidence are now consistent in suggesting that is not the case: Hmr itself maps distal to both ANT genes and to the In(1)AB breakpoint. Furthermore, we cannot confirm any effect of the distal In(1)AB breakpoint or of Hmr on ANT gene expression, although the In(1)AB breakpoint is indeed very close to, or perhaps even within, the 3' end of Ant2. Finally, neither sesB mutant alleles nor transformants carrying extra copies of sesB and Ant2 have any effect on interspecific viability. The molecular nature of these hybrid-lethality-rescuing mutations remains to be discovered.

ANT genes in D. melanogaster:
We show that there are two genes in D. melanogaster encoding proteins very similar in sequence to the adenine nucleotide translocases of other metazoa. These genes are in a tandem array, and the terminator codon of sesB and the initiator ATG codon of Ant2 are separated by only 1104 bp. They show similarities not only in sequence, but also in intron/exon structure, suggesting that they arose from a single gene by an event such as unequal exchange. There is no evidence, e.g., from Southern blot hybridization or in situ hybridization to polytene chromosomes, of any more genes similar in sequence in the genome of D. melanogaster (see LOUVI and TSITILOU 1992 ). Tandem pairs of related genes are quite common in D. melanogaster; other examples include Adh and Adhr and pairs of polyhomeotic, gooseberry, engrailed, sloppy-paired, zen, and knirps genes (see FLYBASE 1999 ). These vary in the degree of similarity between members of a pair and the extent to which they are functionally similar. This variation presumably reflects both the time of duplication and functional constraints.

In the majority of examples of tandemly duplicated genes, each gene is transcribed independently. This is not the case for Adh and Adhr, which are transcribed as a functional dicistronic mRNA, with internal translation initiation required for the synthesis of the ADHR protein (BROGNA and ASHBURNER 1997 ). sesB and Ant2 would also appear to be dicistronically transcribed, but here, alternate splicing results in separate mRNAs for each protein. These mRNAs share a 5' untranslated exon. The circumstantial evidence is that the Ant2 mRNA is much rarer than the sesB mRNA—indeed, this conclusion (made from the relative abundance of their ESTs in the BDGP collection) is supported by Northern blots. This situation is a remarkable variant on the quite familiar situation of alternative splicing that results in two or more protein isoforms. In the case of sesB and Ant2, the two proteins share no coding sequences at all, and they differ in their primary sequence by >20%. The most similar case of which we are aware, in Drosophila, is that of the gene encoding the UB80 protein (a ubiquitin-fusion protein) and a conserved protein of unknown function called IP259 (MOTTUS et al. 1997 ). The mRNAs for these proteins share a 5' untranslated exon and are derived by alternative splicing. The single exon encoding IP259 is wholly contained within the first intron of Ub80. The Su(var)3-9 gene of Drosophila has a structure that is similar, but with the difference that the shared 5' exons are coding with respect to two different proteins (TSCHIERSCH et al. 1994 ). Mutations of this gene are dominant suppressors of position-effect variegation. Su(var)3-9 encodes two different mRNAs by alternative splicing. One of these encodes a protein with chromo- and SET-domains; the other is quite unrelated in sequence and is of an unknown functional class. These two proteins share their N-terminal 80 amino acids, encoded by two exons.

There are two genes known in Caenorhabditis elegans with an organization similar to that of sesB and Ant2. One is the unc-60 locus, which encodes two actin-binding proteins that are 38% identical in amino-acid sequence. Both genes, unc-60A and unc-60B, share a common 5' untranslated exon, and the former gene is entirely contained within the first intron of the latter (MCKIM et al. 1994 ). Even more remarkable is the organization of the genes coding choline acetyl transferase (Cha) and the vesicular acetyl choline transporter (VAChT). In C. elegans, mammals, and Drosophila, the VAChT gene is contained within the first intron of Cha, and the two share a 5' exon. In C. elegans (ALFONSO et al. 1994 ) and mammals (ERICKSON et al. 1994 ; BERRARD et al. 1995 ), this exon is noncoding; in Drosophila (KITAMOTO et al. 1998 ), the 5' common exon codes for the N terminus of the CHA and probably the VACHT proteins. Another example of this type of gene organization of which we are aware is that for the human and murine leptin receptor and leptin receptor gene-related protein (BAILLEUL et al. 1997 ). Here, the mRNAs encoding these two unrelated proteins share two 5' untranslated exons.

Fig 6 is an alignment of the Drosophila ANT proteins with those sequenced from a variety of animals (those of algae and plants are very clearly related, but they form a distinct sequence subfamily). Many other organisms, such as D. melanogaster, have two (or more) ANT genes. A gene tree of these, constructed by DRAWTREE, is shown in Fig 7. In mammals, there are usually three different functional ANT genes encoding proteins with 88–89% amino-acid sequence identity. These show differential tissue expression (POWELL et al. 1989 ; STEPIEN et al. 1992 ). The sesB protein is equally similar to all three mammalian proteins (78% amino-acid sequence identity); similarly, the Ant2 protein is equally similar to the three mammalian proteins (72% amino-acid sequence identity). The predicted ANT2 protein has, relative to mammalian ANT proteins, a rather hydrophilic 10-amino-acid insertion immediately before the first predicted transmembrane domain.

View larger version (54K): In this window In a new window Download PPT slide   Figure 6. An alignment of adenine nucleotide translocase proteins from protozoa and animals. The sites of three point mutations in sesB are highlighted in bold. In sesB9Ed-6, W112 is changed to a stop codon; in sesB9Ed-1, there is the change A144T, and in sesB1 the change L289F. All sequences are from either the SWISSPROT or SPTREMBL database, and are those available on a search for adenine nucleotide translocase through the EBI SRS server (http://srs.ebi.ac.uk/). PFAL1:Q26006 and PFAL2:Q25692 (Plasmodium falciparum), CELE1:P91410 and CELE2:Q17407 (C. elegans), ANOP:Q27238 (A. gambiae), SESB:P91614 (D. melanogaster sesB), ANT2:O62526 (D. melanogaster Ant2), RANA:Q91336 (Rana sylvatica), BTAU1:P02722 (Bos taurus ANT1), BTAU3:P32007 (B. taurus ANT3), MMUS1:P48962 (Mus musculus ANT1), MMUS2:P51881 (M. musculus ANT2), MMUS3:Q61311 (M. musculus ANT3); RNOR1:Q05962 (Rattus norvegicus ANT1), RNOR2:Q09073 (R. norvegicus ANT2), HSAP1:P12235 (Homo sapiens ANT1), HSAP2:P05141 (H. sapiens ANT2), and HSAP3:P12236 (H. sapiens ANT3). The sequence of the Rana protein (Q91336) was corrected for the obvious errors in annotation of its nucleic acid sequence record and for a missed base (clearly seen if this nucleic acid sequence is aligned with those from mammals), which resulted in a frame-shifted protein sequence.

View larger version (10K): In this window In a new window Download PPT slide   Figure 7. Relationships of adenine nucleotide translocases from Drosophila, C. elegans, and human (as typical of mammals). The tree was constructed with DRAWTREE (PHYLIP), with sequences aligned by CLUSTAL W.

Adenine nucleotide translocases are members of a larger family of mitochondrial carrier proteins that are located in the inner mitochondrial membrane (WALKER 1992 ; KUAN and SAIER 1993 ; WALKER and RUNSWICK 1993 ). These transport a variety of substances across this membrane, including organic acids and phosphate. Mitochondrial carrier proteins are characterized by three ~100-amino-acid domains, each with two -helical transmembrane regions (SARASTE and WALKER 1982 ). [Nonhomologous proteins of similar function to adenine nucleotide translocases are found in the intracellular prokaryotic symbiont Rickettsia prowazekii (WINKLER 1976 ; PLANO and WINKLER 1991 ; ANDERSSON et al. 1998 )]. The genome of Saccharomyces cerevisiae includes >20 genes encoding members of this family (data from PROSITE Release 14.0; see also YEAST PROTEOME 1998 ), and there are many in C. elegans (e.g., RUNSWICK et al. 1994 ). The sesB and Ant2 proteins of Drosophila are very distant (<25% amino-acid sequence identity) to any known members of this protein family other than the ANT proteins. We know of only one other member of the mitochondrial carrier protein family in Drosophila. This is colt, which is required for tracheal function (HARTENSTEIN et al. 1997 ). The colt protein is similar to the C. elegans dif-1 protein (AHRINGER 1995 ); the solute carried by these proteins is not known. There is also a clear colt homolog in D. grimshawi (EMBL:U87812; SPTREMBL:P92028).

Adenine nucleotide translocase is the most abundant integral protein of the inner mitochondrial membrane (KLINGENBERG 1985 ). It is an electrogenic protein and catalyzes the exchange of cytoplasmic ADP for ATP across the membrane in a 1:1 stoichiometry. Because this protein links the processes of ADP production to those of ATP utilization, it can exert a control on the rate of oxidative phosphorylation, as well as on the rates of energy-consuming processes. So far, no phenotypes have been associated with mutation of any of the three adenine nucleotide translocase genes in humans, though several myopathies with abnormal mitochondria remain to be characterized at the molecular level. A knockout mutation of mouse ANT1, the heart/muscle specific protein, is not lethal, but it does cause severe mitochondrial hypertrophy in these tissues and obvious disease in adult mice (GRAHAM et al. 1997 ). The features of Luft's disease, the first myopathy described in humans, are abnormally large mitochondria containing packed cristae in muscle fibers and defective respiratory control with normal phosphorylation ("loose coupling," DIMAURO et al. 1976 ). Defective adenine nucleotide translocase might lead to this type of loose coupling. It would be interesting to study the ultrastructure of the mitochondria in sesB mutant alleles. This could indicate whether or not these mutations would be a useful model for the study of some human mitochondrial diseases.

The sesB phenotypes and ANT:
Stress- or bang-sensitive phenotypes are quite common in D. melanogaster, since they are easily recovered in screens, especially if X-linked; at least 21 loci with alleles showing this phenotype have been described (BENZER 1971 ; HOMYK and SHEPPARD 1977 ; GANETZKY and WU 1982 ; FLYBASE 1999 ). It would seem probable that there are at least 100 genes in D. melanogaster that can mutate to this general phenotype.

Four other bang-sensitive mutants have already been identified molecularly, tko, eas, stn, and Atp. technical knockout (tko) encodes a protein homologous to the prokaryotic ribosome protein S12; this is presumably a mitochondrial ribosomal protein in Drosophila (ROYDEN et al. 1987 ; SHAH et al. 1997 ). Most mutant alleles of tko are lethal, but one, tko^25t, is male viable (but hemizygous female lethal) and results in a bang-sensitive phenotype. This is a single-amino-acid substitution in a highly conserved residue (SHAH et al. 1997 ). easily shocked (eas) encodes an ethanolamine kinase involved in membrane phospholipid synthesis. It seems that eas is associated with an increased membrane excitability caused by an altered membrane phospholipid composition (PAVLIDIS et al. 1994 ). Interpretation of the molecular basis for the stoned (stn) phenotype is complicated by the discovery that this locus encodes two proteins (from a dicistronic transcript); one is novel and the other has similarity to AP50 family clathrin assembly proteins and to the UNC-41 protein of C. elegans (ANDREWS et al. 1996 ; BLUMENTHAL 1998 ). Finally, some hypomorphic mutations of the sodium/potassium-transporting ATPase gene (Atp) have a bang-sensitive phenotype (SCHUBIGER et al. 1994 ).

It is interesting that two genes with bang-sensitive mutant phenotypes, tko and sesB, encode mitochondrial proteins. The bang-sensitive alleles show other similarities, e.g., delayed development, but they differ behaviorally, e.g., tko^25t, but not sesB¹, shows hyperexcitability during recovery from stress-induced paralysis. For both of these genes and for stn (originally recovered as sesC), viable bang-sensitive mutations are relatively rare, and most alleles are lethal.

	FOOTNOTES

¹ Present address: Department of Biology, University of Utah, 257 South 1400 East, Salt Lake City, UT 84112.
² Present address: Webster University, Science Department, 470 E. Lockwood Ave., St. Louis, MO 63119-3185.

	ACKNOWLEDGMENTS

We thank Pierre Hutter for his encouragement to start this work and for clones. We also thank Igor Zhimulev, Norbert Perrimon, and Kathy Matthews of the Bloomington Drosophila Stock Center for stocks. We thank Carol McKimmie for making the transformants, and our colleagues Darin Coulson, Tim Fergestad, Glynnis Johnson, Terri Morley, Esther Betran, Simon Collier, Siegrun Herrmann, and Steve Russell for both practical help and discussion. We are most grateful to Suzanna Lewis of the Berkeley Drosophila Genome Project for her help, as well as Howy Jacobs and Julie Ahringer for stimulating discussions. This work was supported by a Biotechnology and Biological Science Research Council Project Grant to M.A. S.B. was also supported by grants from the European Union, D.A.B. was supported by a National Science Foundation/Sloan Foundation Fellowship and Y.Q.Z. by a grant to M.A., D. M. Glover, and R. D. C. Saunders from the Medical Research Council. D.N. was supported by the National Science and Engineering Research Council (Canada, grant no. A3269).

Manuscript received March 31, 1999; Accepted for publication July 2, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AHRINGER, J., 1995  Embryonic tissue differentiation in Caenorhabditis elegans requires dif-1, a gene homologous to mitochondrial solute carriers. EMBO J. 14:2307-2316[Medline].

ALFONSO, A., K. GRUNDAHL, J. R. MCMANUS, J. M. ASBURY, and J. B. RAND, 1994  Alternative splicing leads to two cholinergic proteins in Caenorhabditis elegans.. J. Mol. Biol. 241:627-630[Medline].

ANDERSSON, S. G. E., A. ZOMORODIPOUR, J. O. ANDERSSON, T. SICHERITZ-PONTEN, and U. C. M. ALSMARK et al., 1998  The genome sequence of Rickettsia prowazekii and the origin of mitochondria. Nature 396:133-140[Medline].

ANDREWS, J., M. SMITH, J. MERAKOVSKY, M. COULSON, and F. HANNAN et al., 1996  The stoned locus of Drosophila melanogaster produces a dicistronic transcript and encodes two distinct polypeptides. Genetics 143:1699-1711[Abstract].

ASHBURNER, M., 1989 Drosophila. A Laboratory Handbook. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

BEARD, C. B., A. E. CREWS-OYEN, V. K. KUMAR, and F. H. COLLINS, 1994  A cDNA encoding an ADP/ATP carrier from the mosquito Anopheles gambiae.. Insect Mol. Biol. 3:35-40[Medline].

BAILLEUL, B., I. AKERBLOM, and A. D. STROSBERG, 1997  The leptin receptor promoter controls expression of a second distinct protein. Nucleic Acids Res. 25:2752-2758[Abstract/Free Full Text].

BENZER, S., 1971  From the gene to behavior. J. Am. Med. Assoc. 218:1015-1022[Medline].

BERRARD, S., H. VAROQUI, R. CERVINI, M. ISRAËL, and J. MALLET et al., 1995  Coregulation of two embedded gene products, choline acetyltransferase and the vesicular acetylcholine transporter. J. Neurochem. 65:939-942[Medline].

BLUMENTHAL, T., 1998  Gene clusters and polycistronic transcription in eukaryotes. Bioessays 20:480-487[Medline].

BROGNA, S. and M. ASHBURNER, 1997  The Adh-related gene of Drosophila melanogaster is expressed as a functional dicistronic messenger RNA: multigenic transcription in higher organisms. EMBO J. 16:2023-2031[Medline].

CHAN, H. Y. E., Y. ZHANG, J. D. HOHEISEL, and C. J. O'KANE, 1998  Identification and characterisation of the gene for Drosophila L3 ribosomal protein. Gene 212:119-125[Medline].

DIMAURO, S., E. BONILLA, C. P. LEE, D. L. SCHOTLAND, and A. SCARPA et al., 1976  Luft's disease: further biochemical and ultrastructural studies of skeletal muscle in the second case. J. Neurol. Sci. 27:217-232[Medline].

ERICKSON, J. D., H. VAROQUI, M. K.-H. SCHÄFER, W. MODI, and M.-F. DIEBLER et al., 1994  Functional identification of a vesicular acetylcholine transporter and its expression from a "cholinergic" gene locus. J. Biol. Chem. 269:21929-21932[Abstract/Free Full Text].

FELSENSTEIN, J., 1988  Phylogenies from molecular sequences: inference and reliability. Annu. Rev. Genet. 22:521-565[Medline].

FLYBASE, 1999 http://flybase.bio.indiana.edu/

GANETZKY, B. and C. F. WU, 1982  Indirect suppression involving behavioral mutants with altered nerve excitability in Drosophila melanogaster.. Genetics 100:597-614[Abstract/Free Full Text].

GRAHAM, B. H., K. G. WAYMIRE, B. COTTRELL, I. A. TROUNCE, and G. R. MACGREGOR et al., 1997  The mouse model for mitochondrial myopathy and cardiomyopathy resulting from a deficiency in the heart/muscle isoform of the adenine nucleotide translocator. Nat. Genet. 16:226-234[Medline].

HARTENSTEIN, K., P. SINHA, A. MISHRA, H. SCHENKEL, and I. TOROK et al., 1997  The congested-like tracheae gene of Drosophila melanogaster encodes a member of the mitochondrial carrier family required for gas-filling of the tracheal system and expansion of the wings after eclosion. Genetics 147:1755-1768[Abstract].

HOMYK, T., JR., 1977  Behavioral mutations of Drosophila melanogaster. II. Behavioral analysis and focus mapping. Genetics 87:105-128[Abstract/Free Full Text].

HOMYK, T., JR. and D. E. SHEPPARD, 1977  Behavioral mutations of Drosophila melanogaster. I. Isolation and mapping of mutations which decrease flight ability. Genetics 87:95-104[Abstract/Free Full Text].

HOMYK, T., JR., J. SZIDONYA, and D. T. SUZUKI, 1980  Behavioral mutants of Drosophila melanogaster. III. Isolation and mapping of mutations by direct visual observations of behavioral phenotypes. Mol. Gen. Genet. 177:553-565[Medline].

HUTTER, P., 1997  Genetics of hybrid inviability in Drosophila.. Adv. Genet. 36:157-185[Medline].

HUTTER, P. and M. ASHBURNER, 1987  Genetic rescue of inviable hybrids between Drosophila melanogaster and its sibling species. Nature 327:331-333[Medline].

HUTTER, P. and F. KARCH, 1994  Molecular analysis of a candidate gene for the reproductive isolation between sibling species of Drosophila.. Experientia 50:749-762[Medline].

HUTTER, P., J. ROOTE, and M. ASHBURNER, 1990  A genetic basis for the inviability of hybrids between sibling species of Drosophila.. Genetics 124:909-920[Abstract].

JANCA, F. C., E. P. WOLOSHYN, and D. NASH, 1986  Heterogeneity of lethals in a "simple" lethal complementation group. Genetics 112:43-64[Abstract/Free Full Text].

KITAMOTO, T., W. WANG, and P. M. SALVATERRA, 1998  Structure and organization of the Drosophila cholinergic locus. J. Biol. Chem. 273:2706-2713[Abstract/Free Full Text].

KLINGENBERG, M., 1985 The ADP/ATP carrier in mitochondrial membranes, The Enzymes of Biological Membranes, Chap. 60, Vol. 4, Ed. 2, edited by A. N. MARTONOSI. Plenum Press, New York.

KUAN, J. and M. H. SAIER, 1993  The mitochondrial carrier family of transport proteins: structural, functional, and evolutionary relationships. Crit. Rev. Biochem. Mol. Biol. 28:209-233[Medline].

LEFEVRE, G., 1971  Salivary chromosome bands and the frequency of crossing over in Drosophila melanogaster.. Genetics 67:497-513[Free Full Text].

LEFEVRE, G., 1976 A photographic representation and interpretation of the polytene chromosomes of Drosophila melanogaster salivary glands, pp. 31–66 in The Genetics and Biology of Drosophila, Vol. 1A, edited by M. ASHBURNER and E. NOVITSKI. Academic Press, London.

LEFEVRE, G., 1981  The distribution of randomly recovered X-ray induced sex-linked genetic effects in Drosophila melanogaster.. Genetics 99:461-480[Abstract/Free Full Text].

LEFEVRE, G. and W. S. WATKINS, 1986  The question of the total gene number in Drosophila melanogaster.. Genetics 113:869-895[Abstract/Free Full Text].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego.

LOUVI, A. and S. G. TSITILOU, 1992  A cDNA clone encoding the adenine nucleotide translocase of Drosophila melanogaster shows a high degree of similarity with the mammalian adenine nucleotide translocase. J. Mol. Evol. 35:44-50[Medline].

MADUEÑO, E., G. PAPAGIANNAKIS, G. RIMMINGTON, R. D. C. SAUNDERS, and C. SAVAKIS et al., 1995  A physical map of the X chromosome of Drosophila melanogaster: cosmid contigs and sequence tagged sites. Genetics 139:1631-1647[Abstract].

MCKIM, K. S., C. MATHESON, M. A. MARRA, M. F. WAKARCHUK, and D. L. BAILLIE, 1994  The Caenorhabditis elegans unc-60 gene encodes proteins homologous to a family of actin-binding proteins. Mol. Gen. Genet. 242:346-357[Medline].

MOTTUS, R. C., I. P. WHITEHEAD, M. O'GRADY, R. E. SOBEL, and R. H. L. BURR et al., 1997  Unique gene organization: alternative splicing in Drosophila produces two structurally unrelated proteins. Gene 198:229-236[Medline].

NASH, D. and F. C. JANCA, 1983  Hypomorphic lethal mutations and their implications for the interpretation of lethal complementation studies in Drosophila.. Genetics 105:957-968[Abstract/Free Full Text].

PAVLIDIS, P., M. RAMASWANI, and M. A. TANOUYE, 1994  The Drosophila