logjam Encodes a Predicted EMP24/GP25 Protein That Is Required for Drosophila Oviposition Behavior

Corresponding author: Ginger E. Carney, Georgia Institute of Technology, Atlanta, GA 30332-0230., ginger.carney{at}biology.gatech.edu (E-mail)

Communicating editor: T. SCHÜPBACH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A newly characterized Drosophila melanogaster gene, logjam (loj), functions in female reproduction by modulating oviposition behavior. The locus encodes at least six overlapping transcripts with unique 5' ends. P-element mutants that express very low levels of loj transcripts are unable to oviposit mature eggs. This phenotype can be rescued by the introduction of a transgene expressing the most abundant loj transcript. As for many genes that specify behavioral outputs, loj is present in the adult central nervous system (CNS). Interestingly, it is also observed in vitellogenic egg chambers, suggesting that there may be multiple functions for this gene in egg-laying behavior. loj encodes a predicted protein with homology to the EMP24/GP25 transmembrane components of cytoplasmic vesicles and likely functions in intracellular trafficking.

USING the Drosophila melanogaster model system we have a unique opportunity to clarify our understanding of the genetic and molecular underpinnings of behaviors. D. melanogaster are complex animals with 13,500 predicted genes (ADAMS et al. 2000 ), yet they are amenable to a genetic dissection of the molecular pathways that regulate biological processes. Drosophila exhibit a wonderfully rich repertoire of interesting behaviors such as courting and responding to potential mates, olfactory learning, and foraging for food. All of these behaviors require integration of multiple sensory inputs and generation of various motor outputs and are recognized to be under genetic control (reviewed in HALL 1994 ; DUBNAU and TULLY 1998 ; SOKOLOWSKI 2001 ).

The sexually dimorphic behaviors associated with Drosophila reproductive success are clearly governed by the actions of multiple genes, some of which function as dedicated components of the somatic sex-determination hierarchy, beginning with the initial setting of Sex-lethal (Sxl) activity (reviewed in BURTIS 1993 ; MCKEOWN 1994 ; CLINE and MEYER 1996 ). Two genes, doublesex (dsx) and fruitless (fru), have been shown to be regulated by the upstream components of this signaling cascade (BAKER and RIDGE 1980 ; MCKEOWN et al. 1988 ; NAGOSHI et al. 1988 ; ITO et al. 1996 ; RYNER et al. 1996 ). A third gene, dissatisfaction (dsf), was shown to interact genetically with this pathway but is not regulated directly by known upstream members of the hierarchy (FINLEY et al. 1997 , FINLEY et al. 1998 ). All three genes encode DNA-binding proteins (BURTIS et al. 1991 ; ITO et al. 1996 ; RYNER et al. 1996 ; FINLEY et al. 1998 ), but the identity of genes necessary for reproductive behaviors that function directly downstream of them is not known.

The pathways for regulating female and male sexual behaviors diverge in the requirements for these regulators of downstream functions. dsf and dsx, but not fru, control female sexual behaviors (ITO et al. 1996 ; RYNER et al. 1996 ; FINLEY et al. 1997 ; WATERBURY et al. 1999 ). dsf mutant females have defective uterine muscle innervation, a phenotype that is correlated with an absence of oviposition behavior in these females (FINLEY et al. 1997 ). The major function of dsx is to control sex-specific morphological differentiation in males and females (HILDRETH 1965 ; KEISMAN and BAKER 2001 ; KEISMAN et al. 2001 ; AHMAD and BAKER 2002 ). Misexpression of DSX^F in males transforms them to phenotypic females, which allows them to mate with other males (WATERBURY et al. 1999 ).

All three genes affect male behavior. fru is the major regulator of male sexual behavior but has no demonstrable role in female reproduction (ITO et al. 1996 ; RYNER et al. 1996 ; VILLELLA et al. 1997 ). dsf functions in male courtship object choice (FINLEY et al. 1997 , FINLEY et al. 1998 ), while DSX^M is needed for male sine-song production and general robustness of male courtship (VILLELLA and HALL 1996 ).

The major reproductive behaviors of females are (1) receptivity to courtship followed by copulation and (2) oviposition of eggs, a behavior that is independent of the act of mating and is under voluntary control of the female. Virgin females are refractive to the mating advances of males for the first day of their lives (MANNING 1966 , MANNING 1967 ). During this time the ovaries mature (reviewed in MAHOWALD and KAMBYSELLIS 1980 ; SPRADLING 1993 ), cuticular pheromonal profiles change to make the female more enticing to the male (reviewed in JALLON 1984 ; TOMPKINS 1984 , TOMPKINS 1998 ), and hormonal fluxes foster the development of female sexual receptivity (MANNING 1966 ). Unmated females retain mature eggs but eventually lay unfertilized eggs beginning at approximately the fifth day after eclosion (MAHOWALD and KAMBYSELLIS 1980 ).

Once the female has mated, her behavior alters as she rebuffs further mating advances. Proteins and other compounds in the male ejaculate affect the female's rate of ovulation, oviposition, and her receptivity to male mating overtures (CHEN et al. 1988 ; MONSMA and WOLFNER 1988 ; AIGAKI et al. 1991 ; KUBLI 1992 ; HERNDON and WOLFNER 1995 ; WOLFNER 1997 , WOLFNER 2002 ; WOLFNER et al. 1997 ; HEIFETZ et al. 2000 , HEIFETZ et al. 2001 ; CHAPMAN et al. 2001 ; FLEISCHMANN et al. 2001 ; SAUDAN et al. 2002 ). Mated females have been observed to retain eggs if they do not find oviposition conditions to be ideal (GROSSFIELD 1978 ).

Many identified genes function in one or both of these behaviors, but their relationship to the sex-determination hierarchy is unclear. Some of these genes may function downstream in the sex-determination regulatory cascade, while others may function in convergent pathways that interact with sex-determination genes.

dsf regulates both receptivity and oviposition (FINLEY et al. 1997 ). Other loci involved in female reproductive behaviors may function in one pathway but not the other. Mutations in the ecdysone receptor (EcR; CARNEY and BENDER 2000 ) or Tyramine B-hydroxylase (TBH; MONASTIRIOTI et al. 1996 ) affect oviposition behavior. chaste females are unreceptive to wild-type male courtship (YAMAMOTO et al. 1997 ), while spinster (spin) mutant females are unreceptive and fail to make normal eggs (NAKANO et al. 2001 ), making an unambiguous determination of spin's role in oviposition difficult. Expression and function of these genes is required in the female for manifestation of sexual behaviors, distinguishing them from the gene products present in the male ejaculate that affect female postmating behaviors when they are transferred to females during copulation (CHEN et al. 1988 ; MONSMA and WOLFNER 1988 ; AIGAKI et al. 1991 ; KUBLI 1992 ; HERNDON and WOLFNER 1995 ; WOLFNER 1997 , WOLFNER 2002 ; WOLFNER et al. 1997 ; HEIFETZ et al. 2000 , HEIFETZ et al. 2001 ; CHAPMAN et al. 2001 ; FLEISCHMANN et al. 2001 ; SAUDAN et al. 2002 ).

Clearly, the genes described to date compose an incomplete list of those that are required for female reproductive behaviors in Drosophila. From genetic screens designed to isolate additional loci needed for these sex-specific behaviors, we identified a new gene, logjam (loj), whose expression is required for oviposition. The predicted protein from this locus shares identity with the EMP24/GP25 family of cytoplasmic vesicle membrane proteins. The unexpected finding that loj encodes a member of this vesicle protein family highlights the importance of intracellular trafficking for proper behavioral outputs.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains:
The loj⁰⁰⁸⁹⁸/TM3, ry Sb e P-element strain (originally designated fs(3)00898; KARPEN and SPRADLING 1992 ) was provided by C. Berg. loj⁰⁰⁸⁹⁸ fails to complement the third chromosomal deficiency Df(3L)ZN47 (64C-65C). This deficiency stock and the l(2)04026/TM3, ry Sb Ser P-element insertion line were obtained from the Drosophila stock center in Bloomington, Indiana. l(2)04026 was originally identified as a lethal allele. However, when the stock is grown at 25° under noncrowded conditions, homozygous adults eclose, albeit at low frequency. Here we refer to l(2)04026 as loj⁰⁴⁰²⁶. loj⁰⁴⁰²⁶/loj⁰⁰⁸⁹⁸ and loj⁰⁴⁰²⁶/Df(3L)ZN47 animals eclose in the expected Mendelian ratio, suggesting that a second site on the loj⁰⁴⁰²⁶ chromosome reduces viability of homozygotes.

Both P-element insertions were mobilized by standard methods and a total of 59 ry^- stocks were generated by introduction of the 2-3 transposase source (ROBERTSON et al. 1988 ). Eight of these lines fail to complement a loj P insertion for fertility. By PCR analysis, at least two of these lines appear to have large deletions in the loj region and one is limited to the region upstream of the loj⁰⁰⁸⁹⁸ insertion. The remaining alleles, when heterozyous with a loj P insert, produce females that are fertile. Many of these alleles have been shown to be partial deletions in the loj⁰⁰⁸⁹⁸ P-element sequence, indicating that complete or partial loss of the P insertion is sufficient to rescue the loj phenotype.

Mating assays:
Virgin females were collected and aged individually in food vials for 4–6 days at 25°. Each female was placed into a courtship chamber with a wetted filter paper and an aged Canton-S male and observed for 30 min or until successful mating occurred (FINLEY et al. 1997 ).

RNA hybridization:
For Northern blotting, total RNA was prepared from Canton-S, loj⁰⁰⁸⁹⁸, loj⁰⁰⁸⁹⁸/TM3, loj⁰⁴⁰²⁶, and loj⁰⁴⁰²⁶/TM3 adult flies using TRIzol (Invitrogen, San Diego), and 6 µg of RNA was electrophoresed in each lane of a 1% agarose gel with MOPS/formaldehyde and then blotted to nylon membranes. For in situ analyses, 30-µm cryostat sections of Canton-S and loj⁰⁰⁸⁹⁸ adult females and males were hybridized for detection of RNA expression as described previously (GOODWIN et al. 2000 ). Riboprobes for Northern and in situ analyses were derived from the LD30746 cDNA [Berkeley Drosophila Genome Project (BDGP)] that corresponds to the class I transcript (see Fig 3) and labeled with digoxigenin (DIG) using the DIG RNA-labeling system (Roche, Indianapolis). The DIG luminescent detection kit (Roche) was used for Northern signal detection.

View larger version (63K): In this window In a new window Download PPT slide   Figure 1. Northern blot of RNA from wild-type and loj adults. Riboprobes that recognize all loj transcript classes hybridize to Canton-S (CS) RNA as well as loj heterozygous control RNA (loj00898/TM3 and loj04026/TM3). Signal is absent in loj00898 and loj04026 mutant females. The predicted size of the bands as determined by RNA molecular weight markers is 1150 nucleotides (nt).

View larger version (40K): In this window In a new window Download PPT slide   Figure 2. Multiple transcripts are produced by the loj locus. The structures of the six identified classes of loj transcripts are shown as well as their predicted nucleotide lengths. Exons are boxed. Shaded boxes indicate exons predicted to be shared by all six of the transcript classes. Positions of the loj00898 and loj04026 P-element insertions are shown above the class I message. The class Ia message structure is predicted from a single BDGP cDNA clone (RE44409) derived from embryonic tissue and has a 20-nt 5' extension to the B exon. Arrows indicate PCR primers. a and b are the real-time PCR primers (see Table 2). The forward and reverse primer pair allows transcript-specific amplification of class I messages (see Fig 3). The forward primer spans the A and B exons, and the brackets indicate intronic sequences that were not included in this primer. ATG is the predicted translation start site present in the B exon.

View larger version (25K): In this window In a new window Download PPT slide   Figure 3. loj is expressed at multiple times in development. Class I-specific primers were used to amplify cDNA from sexed, wild-type animals from the wandering third instar larval, pupal, and adult stages. Dissected heads, abdomens, and ovaries from wild-type animals express loj, which is consistent with the enhancer trap patterns shown in Fig 4. M, DNA molecular weight marker; f, female; m, male; abd, abdomen; a-ov, abdomens with ovaries removed; ov, ovary; G, genomic DNA control lane (negative control).

These probes are expected to hybridize to all six loj transcript classes. However, the smaller class IV and class V messages are not detected on the blots, suggesting that they are present at low levels or that there is temporal specificity to their expression. Both hypotheses are supported by the fact that few rapid amplification of cDNA end (RACE) clones that correspond to class IV and V transcripts were isolated from adults (see below).

5' and 3' RACE:
5' RACE products were isolated from adult flies using the Invitrogen 5' RACE system or the First Choice RLM-RACE kit (Ambion, Austin, TX), which is designed to recognize only full-length, capped messages. Adult 3' RACE products were isolated with the Invitrogen 3' RACE system and the First Choice RLM-RACE kit. 5' RACE primers were designed to exons B (5'-GATCAAAAGACAACAAATCAGAGCAATCGGTAGT-3' and 5'-CAAAAGACAACAACAAATCAGA-3'), C (5'-TGATGGTAGCAGTCCTCCTT-3'), and D (5'-AATCGGGAGAACTGGTTGT-3' and 5'-TTGCGGACGGCGAAACCA-3'). 3' RACE primers were designed to exons C (5'-AAGGAGGACTGCTACCATCA-3'), E (5'-CACCGTTGGCACTGTGGAG-3'), and F (5'-TCCAAAAGTCGCATCTAAG-3'). All products were purified and cloned into the pPCR-Script Amp SK(+) cloning vector (Stratagene, La Jolla, CA) for sequencing. Out of 29 5' RACE clones chosen for sequence analysis, a total of 17 class I clones were sequenced (see Fig 3 for clone structures). Three class II, two class III, three class IV, and four class V clones were sequenced as well.

Real-time PCR and standard reverse transcriptase-PCR:
For real-time PCR, whole-body total RNA was extracted from a homogenate of 15–40 adult females of the genotypes Canton-S, loj⁰⁰⁸⁹⁸, loj⁰⁰⁸⁹⁸/Df(3L)ZN47, loj⁰⁰⁸⁹⁸/TM3, loj⁰⁴⁰²⁶, loj⁰⁴⁰²⁶/Df(3L)ZN47, and loj⁰⁴⁰²⁶/TM3. cDNA was prepared with the Superscript first-strand synthesis system for reverse transcriptase (RT)-PCR (Invitrogen). For relative quantitative detection of products (real-time PCR), cDNAs were amplified with primers directed to a region common to all known loj transcripts (forward primer 5'-AAGGAGGACTGCTACCATCA-3' and reverse primer 5'-AATCGGGAGAACTGGT-3') using the SYBR green PCR master mix for RT-PCR (PE Biosystems, Foster City, CA). An ABI PRISM 7700 sequence detection system (PE Biosystems) was used to detect products. Amplification of rp49 (forward primer 5'-TGACCATCCGCCCAGCATACA-3' and reverse primer 5'-TCTCGCCGCAGTAAAC-3') in independent reactions served as the endogenous control for normalizing the amount of cDNA in each reaction. Controls for both sets of reactions included reactions that lacked cDNA template or primers as well as no reverse transcriptase reactions to test for genomic DNA contamination.

Results from each primer set were used to generate a standard curve for statistical comparisons of the samples. Samples were run in quadruplicate and were analyzed for differences in transcript levels and statistical significance using the SAS program (Version 8.01) for ANOVA analysis. The average from multiple, independent runs is reported for each genotype. All values were normalized to loj⁰⁰⁸⁹⁸/TM3 heterozygotes. The slight difference between loj⁰⁰⁸⁹⁸/TM3 and loj⁰⁴⁰²⁶/TM3 females' transcript levels was not significant, but transcript levels in both of these control female genotypes were significantly different from results obtained with all mutant combinations tested ( = 0.05, Tukey's Studentized range). To show that the observed decrease in loj transcript levels in mutant females was specific to this locus, we designed primers (forward primer 5'-GCTACGATGACAACTTCTG-3' and reverse primer 5'-CGGAGTCGGGAAACTT-3') to a nearby open reading frame (ORF; CG10467) that is located 0.5–1.0 kb upstream of the loj⁰⁰⁸⁹⁸ P element. Transcription from this gene is not affected in either loj⁰⁰⁸⁹⁸ or loj⁰⁴⁰²⁶ homozygous females.

For standard RT-PCR, whole-body or tissue-specific total RNA was extracted from a homogenate of 5–25 male or female wandering third instar larvae, pupae, or adult animals or from adult tissues (head, abdomen, abdomen without ovaries, and ovaries). cDNA was generated as described above. The cDNA was amplified with primers specific to the class I transcript (forward primer 5'-CACCTCTAAGCAACAGAGA-3' and reverse primer 5'-TGATGGTAGCAGTCCTCCTT-3') using Taq polymerase.

Immunohistochemical and enhancer trap analyses:
Tissues from loj⁰⁰⁸⁹⁸/TM3, loj⁰⁴⁰²⁶/TM3, loj⁰⁰⁸⁹⁸, and loj⁰⁴⁰²⁶ larvae, pupae, and adult animals were dissected in PBS. For antibody detection of ß-galactosidase (ß-gal) expression, tissues were fixed in 3.7% formaldehyde, rinsed with PBST (PBS, 0.1% Triton X-100), and incubated with rabbit anti-ß-gal (1:10,000; Cappel, Malvern, PA) overnight at 4°. Tissues were rinsed with PBST, incubated with biotinylated anti-rabbit secondary antibody (Vector, Burlingame, CA) at room temperature, and developed with diaminobenzidine (Sigma, St. Louis) in the presence of the ABC reagent (Vector). Signal was not detected in female abdomens (data not shown).

Wild-type, loj⁰⁰⁸⁹⁸, loj⁰⁴⁰²⁶, loj⁰⁰⁸⁹⁸/Df(3L)ZN47, and loj⁰⁴⁰²⁶/Df(3L)ZN47 adult female genital tracts were prepared as above and incubated with rabbit anti-synaptotagmin (1:1000) for visualization of synaptic innervation patterns (FINLEY et al. 1997 ).

For X-gal staining, loj⁰⁰⁸⁹⁸/TM3, loj⁰⁴⁰²⁶/TM3, loj⁰⁰⁸⁹⁸, and loj⁰⁴⁰²⁶ female and male tissues were fixed in 3.7% formaldehyde, rinsed with PBST, and incubated overnight at 37° with reaction buffer [PBS with 2.5 mM MgCl₂, 3 mM K₃Fe(CN)₆, 3 mM K₂Fe(CN)₆] in the presence of 8% X-gal (Sigma) in DMSO.

Heat-shock rescue:
Constructs for heat-shock overexpression of loj were generated by cloning the class I cDNA (clone LD30746) into the pCaSpeR-hs-act plasmid (D'AVINO and THUMMEL 1999 ). This transcript was chosen because it is widely expressed and is expected to encode the same protein as class Ia, II, and III messages. The construct was injected into yw embryos by the CBRC Core Facility at Massachusetts General Hospital. More than 26 independent transgenic lines containing pCaSpeR-loj were established.

Expression from pCaSpeR-loj was sufficient to rescue the oviposition phenotype. Three independent insertions on the X chromosome (H1b, H1a, and C10e) were introduced into the loj⁰⁰⁸⁹⁸ background. Mutant and control females were collected at 22° and placed in food vials at 25° with wild-type males. Eggs were counted at 5–11 days for each genotype. For the heat-shock treatment, females within 12 hr of eclosion were placed at 37° for 1 hr.

Sequence analysis:
LOJ motifs were predicted using the following programs: SignalP (Center for Biological Sequence Analysis, The Technical University of Denmark, Lyngby, Denmark; http://www.cbs.dtu.dk/services/SignalP/; NIELSEN et al. 1997 ) to detect the location of potential signal peptide cleavage sites; COILS (European Molecular Biology Network, Swiss node; http://www.ch.embnet.org/software/COILS_form.html; LUPAS et al. 1991 ) to identify coiled-coil protein interaction motifs; TMpred (European Molecular Biology Network, Swiss node; http://ch.embnet.org/cgi-bin/TMPRED_form_parser; HOFMANN and STOFFEL 1993 ) to predict transmembrane-spanning domains; and PSORT (University of Tokyo, Toyko; http://psort.nibb.ac.jp; NAKAI and HORTON 1999 ) to predict signal peptides, transmembrane domains, topology, coiled-coil domains, and subcellular localization of the LOJ protein.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The loj⁰⁰⁸⁹⁸ and loj⁰⁴⁰²⁶ P-element inserts cause female sterility and an egg-retention phenotype:
Our interest in identifying genes that are required for female reproductive behaviors led us to screen candidate female-sterile mutants for deficits in mating and oviposition. A female-sterile P-insertion strain, designated here as loj⁰⁰⁸⁹⁸, produces homozygous females that do not lay eggs. We examined these females for the retention of mature eggs in the female genital tract, a defect characteristic of females that are mutant for the sex-determination gene dsf (FINLEY et al. 1997 ) and of females mutant for TBH, which encodes a biosynthetic enzyme that produces the neurotransmitter octopamine from tyramine (MONASTIRIOTI et al. 1996 ). In females homozygous for the P insert or trans-heterozygous for loj⁰⁰⁸⁹⁸ and a deficiency (Df) chromosome that uncovers the mutation (loj⁰⁰⁸⁹⁸/Df; see MATERIALS AND METHODS), one or more mature eggs become lodged within the genital tract causing a "logjam" of eggs within the female and preventing the release of eggs (data not shown). While in essentially every case an egg is found in the uterus, eggs are also found within the lateral or common oviducts in many females.

Synaptic connections to the oviduct and uterine muscles are present in loj mutants (data not shown); such synapses were previously shown to be absent from the uterine muscles of dsf females (FINLEY et al. 1997 ). loj females mate with latency and frequency similar to that of control females (Table 1) and store sperm within their sperm storage organs (G. E. CARNEY and B. J. TAYLOR, unpublished results). Therefore, loj⁰⁰⁸⁹⁸ mutations affect egg-laying behavior but not receptivity to mating.

We used the P-element tag in loj⁰⁰⁸⁹⁸ to determine the genomic position of the insert. Inverse PCR (YEO et al. 1995 ) coupled with sequencing of DNA flanking the loj⁰⁰⁸⁹⁸ P element revealed that loj⁰⁰⁸⁹⁸ is inserted within polytene region 65A on the third chromosome. Expressed sequence tags (ESTs) isolated by the BDGP correspond to a predicted gene (CG10733) in the same region.

Knowledge of the position of loj⁰⁰⁸⁹⁸ allowed us to identify in the same region a second P insert (designated loj⁰⁴⁰²⁶) that fails to complement the loj⁰⁰⁸⁹⁸ phenotype, providing two independent mutant alleles that lead to egg retention. loj⁰⁴⁰²⁶ is inserted within the predicted CG10733 ORF (see Fig 2), indicating that the inserts are likely to disrupt its function. The oviposition defect can be reverted by mobilization of either P element (see MATERIALS AND METHODS), providing strong evidence that the insertions are responsible for the observed absence of oviposition behavior.

Both P-element insertions disrupt loj expression and function in adult females:
To determine if loj corresponds to the CG10733 ORF, we first established the wild-type transcript expression pattern from this region in adult animals. Northern analysis of wild-type adult female or male whole-body RNA revealed a diffuse hybridizing band (Fig 1), potentially containing multiple transcripts. The 1.15-kb size of the band(s) in both sexes is consistent with the size of clones from the BDGP EST project (e.g., clone LD30746) corresponding to this region. Northern blots of RNA from adult females homozygous for either P-element insertion do not produce a hybridization signal when probed with labeled sequences derived from the LD30746 cDNA clone (Fig 1).

To confirm that the P elements disrupt expression from the CG10733 ORF in adult females, we designed primers directed to the common region of the predicted gene (see Fig 2) and used a quantitative RT-PCR approach to assess whether any loj transcripts are detectable in loj mutants. Real-time PCR with primers that recognize all transcript classes (see below) reveals that the loj⁰⁰⁸⁹⁸ and loj⁰⁴⁰²⁶ insertions substantially downregulate loj transcripts (Table 2) but do not affect a nearby (within 0.5–1 kb) gene (see MATERIALS AND METHODS). These results indicate that loj corresponds to the predicted CG10733 ORF.

Multiple transcripts are produced from the loj genetic region:
We used 5' and 3' RACE and BDGP database searches to determine the structure of transcripts from females and males, identifying at least six different transcript classes (Fig 2), some of which correspond to cDNAs found in the BDGP database. Class I transcripts are structurally similar to but extended 5' relative to related BDGP cDNA clones (RE61227, RE18835, GM12884, LD30746, and LD38695). These class I messages are found at high levels in adults of both sexes and contain a strong consensus ATG translation start site (CAVENER and RAY 1991 ) in the second exon (Fig 2). Class Ia (isolated by BDGP and a likely splicing variant of class I), class II, and class III transcripts are expected to encode the same protein as class I. Two cDNAs from the BDGP database (SD16982 and GM24669) contain sequences that are common to class II and III RACE clones. A fourth transcript (class IV) has a 5' end that begins in the second exon downstream of the consensus translation start site present in class I, Ia, II, and III transcripts (Fig 2).

Class V messages have 5' ends that begin with the first common exon (exon C; see Fig 2) but contain a 9-bp 5' extension that is not present in other transcripts (Fig 2). There is no consensus translation start, and the first ATG in-frame with loj is present in the second class V exon (corresponding to exon D of class I messages; see Fig 2).

Results from 3' RACE experiments indicate that all transcript classes likely have the same 3' end. These 5' and 3' RACE studies allowed us to map the positions of the P inserts relative to the different transcripts. loj⁰⁰⁸⁹⁸ is inserted into the 5' untranslated region of the class I, Ia, and class III transcripts, while loj⁰⁴⁰²⁶ lies in predicted protein coding sequences for classes I–III (Fig 2). Neither P element is inserted into sequences that are transcribed to class IV or V messages.

loj is expressed in the larva, pupa, and adult:
We used RT-PCR to analyze the expression pattern of the highly expressed class I transcripts in females and males during multiple developmental stages to understand when and where loj is expressed. Our results indicate that loj is expressed in female and male larvae, pupae, and adults (Fig 3). We found that in adults class I transcripts are present in heads, abdomens, abdomens from which the ovaries have been removed, and the ovaries themselves (Fig 3).

To gain further knowledge as to which tissues in the adult express loj we took two approaches: (1) an enhancer trap analysis using the loj⁰⁰⁸⁹⁸ P-element tag to assess expression from the P inserts in larvae, pupae, and adult animals and (2) in situ hybridization with loj riboprobes. Larval expression occurs in the brain and ventral nerve cord (VNC; Fig 4A and Fig B), and the overall patterns of expression from the enhancer trap elements are similar in pupae and adults of both sexes (Fig 4, C–H). ß-Gal activity or expression is limited to the central nervous system (CNS) and ventral epidermis (Fig 4 and data not shown). In larvae signal is present in the CNS (Fig 4A and Fig B) and ventral epidermis, while in pupae signal is observed in the CNS (Fig 4C and Fig D). Other pupal tissues were not tested. In the pupal and adult brain, staining is evident in a small number of cells in the antennal lobe, medullar regions of the optic lobe, and the subesophageal ganglion (Fig 4C, Fig E, and Fig G). Expression is also present in dorsal and lateral cells of the abdominal and thoracic ganglia of the VNC (Fig 4D, Fig F, and Fig H).

View larger version (80K): In this window In a new window Download PPT slide   Figure 4. loj is expressed in the larval, pupal, and adult CNS. Composite photographs of expression patterns from the loj00898 enhancer trap insertion in larval (A and B), pupal (C and D), and adult (E–H) CNS tissues. Br, brain; VNC, ventral nerve cord; al, antennal lobe; m, medulla; sog, subesophageal ganglion; ab, abdominal ganglion; T1, T2, and T3 delineate positions of the thoracic neuromeres of the VNC. Females are shown in A, G, and H. Males are shown in B, E, and F. For the male in E, we observed weak expression in a pattern similar to that of the female in G. Bar, 100 µm.

RNA in situ hybridization to sectioned tissues from wild-type adult females revealed expression only in the ovaries, with strong hybridization to follicle cells and weak hybridization to nurse cells of midstage vitellogenic (yolk-containing) eggs (Fig 5A and Fig B). No hybridization to any tissues was detected in sectioned adult wild-type males (Fig 5E and Fig F). We did not detect RNA hybridization in loj⁰⁰⁸⁹⁸ homozygous females or males (Fig 5C and Fig D and data not shown), providing further evidence that, at least in female ovarian tissue, gene function is disrupted in loj mutants.

View larger version (171K): In this window In a new window Download PPT slide   Figure 5. Vitellogenic egg chambers contain loj mRNA. In situ hybridization to sectioned wild-type adult females shows loj expression in vitellogenic eggs (A and B). Arrows point to the high-level staining found in follicle cells and arrowheads to the low-level staining present in the nurse cell compartment. loj00898 mutant females lack expression in vitellogenic eggs (C and D; see arrows). Wild-type males do not have detectable loj expression in this assay (E and F). Arrows in E and F point to sperm bundles. Bar, 20 µm.

The oviposition defect can be rescued by loj overexpression:
We generated transgenic animals containing a heat-inducible pCaSpeR-loj plasmid (see MATERIALS AND METHODS). Basal expression from each of three independent pCaSpeR-loj constructs is sufficient to partially rescue the egg-laying defect. However, a single heat pulse early in the adult stage provides a greater level of rescue with two of the transgenes (Table 3).

loj encodes a predicted protein with homology to EMP24/GP25 cargo receptor proteins:
loj encodes a predicted protein that shares amino acid homology with the EMP24/GP25 family of putative vesicle cargo receptor proteins (Fig 6). These 24-kD proteins, present in diverse eukaryotic lineages including mustard, yeasts, worm, and mammals, were isolated as components of coat protein (COP)-coated vesicles that shuttle cargo between the endoplasmic reticulum and Golgi membranes for processing (SCHIMMOLLER et al. 1995 ; STAMNES et al. 1995 ) and may have positive as well as negative functions in mediating molecular transport through the secretory pathway (SCHIMMOLLER et al. 1995 ; STAMNES et al. 1995 ; BELDEN and BARLOWE 1996 ; MARZIOCH et al. 1999 ; MUNIZ et al. 2000 ). The predicted LOJ protein shares important hallmarks of this class of vesicle proteins, including an amino-terminal signal sequence, a large luminal domain containing a coiled-coil protein interaction domain, a single transmembrane region, and a small, carboxy-terminal cytoplasmic tail (Fig 7). The coiled-coil domain of EMP24 proteins likely mediates heteromeric contacts between EMP24 proteins (BELDEN and BARLOWE 1996 ; FULLEKRUG et al. 1999 ; MARZIOCH et al. 1999 ; CUIFO and BOYD 2000 ; DENZEL et al. 2000 ).

View larger version (33K): In this window In a new window Download PPT slide   Figure 6. LOJ is a member of the EMP24/GP25 vesicle trafficking protein family. The predicted LOJ protein is closely related to EMP24/GP25 proteins from other taxa. Identical residues are shown in black and have an exclamation point in the consensus line indicating absolute conservation in EMP24/GP25 proteins. Amino acids identical in >65% (four or more of the six) of the sequences are shaded dark gray. Light gray residues are identical or are conserved according to the following scheme: (FYW), (ILVM), (DE), (GA), (ST), (NQ), (RKH). An asterisk in the consensus line indicates identity or conservation in 65% of the sequences. The alignment was performed in ClustalX (THOMPSON et al. 1997 ) using the Gonnet 250 pairwise alignment and Gonnet series multiple alignment parameters. Coloring was performed with MacBoxshade V2.15. Hs, Homo sapiens; Dm, D. melanogaster; Ce, Caenorhabditis elegans; Sc, Saccharomyces cerevisiae.

View larger version (52K): In this window In a new window Download PPT slide   Figure 7. LOJ shares important functional domains with EMP24/GP25 proteins. The sequence of the longest identified class I transcript is shown with the likely translation start and stop sites indicated in boldface type. Predicted amino acids are below the nucleotide sequence. Boldface amino acid residues represent the predicted signal peptide. Residues in boldface italics compose the coiled-coil domain, and boxed residues indicate the transmembrane region.

In total, seven predicted EMP24-related products are encoded by the Drosophila genome (Fig 8). Prior to this study, there were no characterized Drosophila mutations in these putative cargo receptors.

View larger version (62K): In this window In a new window Download PPT slide   Figure 8. Multiple EMP24/GP25 proteins are encoded by the Drosophila genome. Sequence alignment of the seven predicted Drosophila EMP24/GP25 proteins. Shading and generation of the alignment are as described for Fig 6.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A clear understanding of how genetic signals are transduced at the molecular, cellular, and biochemical levels to specify behavioral outcomes is an important goal for behavioral biologists. Drosophila female reproductive behaviors are a good model system for tackling this difficult issue because they provide a strong selective phenotype for mutant screens and because some of the genes that contribute to these behaviors are already known. Using genetic and behavioral screens, we identified a new class of proteins important for oviposition. Our results indicate that loj has an essential role in egg-laying behavior and encodes a member of a conserved family of proteins that function in vesicular trafficking in the endomembrane system.

Molecular and phenotypic characterization of loj in wild-type and mutant animals:
Northern analysis (Fig 1) and RT-PCR (Fig 3) reveal that loj transcripts are present in larvae, pupae, and adults. Enhancer trap and in situ hybridization experiments suggest that the majority of this signal is derived from loj expression in the brain, VNC, and developing eggs (Fig 4 and Fig 5). By in situ hybridization to sectioned adult tissues, loj expression is detectable only in adult female vitellogenic egg chambers (Fig 5A and Fig B) and not detectable in the CNS or epidermis. Lack of recognizable in situ signal in these tissues in either sex is possibly due to the small number of cells that are likely to be labeled in each section. An alternative possibility is that the difference in the Northern and RT-PCR data vs. data from in situ hybridization is that the level of RNA per cell is below detection except in the ovary. In female and male larvae, ß-gal staining is present in the CNS (Fig 4A and Fig B), ventral epidermis, and posterior portion of the proventriculus (data not shown). Midstage pupae also have loj CNS expression (Fig 4C and Fig D).

Females with little loj expression are able to mate but do not lay eggs (Table 1 and Table 2), and the oviposition defect can be rescued by the expression of a transgene containing a class I cDNA (Table 3). While wild-type females lay more eggs than the rescued loj mutants do (data not shown), the numbers of eggs deposited by mutant females containing a pCaSpeR-loj transgene are well above those laid by mutants lacking the transgene and indicate that decreased levels of loj are responsible for the oviposition phenotype. In one instance, a loj mutant female that lacked the transgene laid an egg (Table 3). In our experience this is an extremely rare event because we have observed only a half dozen or so eggs laid by the many hundreds of mutant females that have been tested in egg-laying assays.

mRNA expression in loj⁰⁴⁰²⁶ females is significantly reduced relative to other mutant combinations tested (Table 2). The fact that loj⁰⁴⁰²⁶ females have decreased transcript levels relative to loj⁰⁴⁰²⁶/Df animals may indicate that this insertion chromosome contains another interacting factor that affects loj expression. Two additional Drosophila EMP24 family members are located on the third chromosome. One possibility is that interactions between mutations in one of these two genes and loj⁰⁴⁰²⁶ account for the further reduction in transcript levels observed in the loj⁰⁴⁰²⁶ homozygous mutants. Precedence for a similar effect is found in yeast and mice where loss of one EMP24 protein has been shown to affect levels of interacting EMP24 proteins (MARZIOCH et al. 1999 ; DENZEL et al. 2000 ). While this effect in mice is post-transcriptional (DENZEL et al. 2000 ), it is possible that an alternate regulatory mechanism occurs in multicellular organisms. Despite the variation in transcript expression levels, there are no obvious phenotypic differences between loj⁰⁰⁸⁹⁸, loj⁰⁴⁰²⁶, or loj/Df heterozygous adults.

Interestingly, while loj females have a profound behavioral defect, there is no observable deleterious effect on male fertility from loss of loj activity. The fact that loj mutant adults eclose at the expected Mendelian frequencies and, in the case of males, are able to reproduce suggests that the molecular function supplied by loj can be partially supplied by another locus, possibly one of the remaining six Drosophila EMP24/GP25 family members. A second possibility is that loj does not function in adult males.

Due to the presence of a strong consensus translational start site in exon B (Fig 2), the class I transcripts highly expressed in both sexes are predicted to encode the same protein as the class Ia, II, and III mRNAs. The class IV and V messages lack an obvious translational start site and are not detectable on Northern blots (see MATERIALS AND METHODS). Theoretical translation of these sequence and database searches did not identify homologies (other than EMP24/GP25) to the predicted peptides. One possibility is that the transcripts are not translated to a functional protein, similar to the sex-nonspecific product of tra (BOGGS et al. 1987 ) or the female-specific product of fru (USUI-AOKI et al. 2000 ). Alternatively, translation may occur by internal ribosome entry or via another mechanism, such as recognition of a less-conserved translation start site. If this is the case, each of these two messages potentially encodes a different version of LOJ that has a truncated lumenal domain. It is interesting to speculate that truncated proteins produced from the class IV and class V transcripts may have different specificities for cargo molecules or altered interactions with other Drosophila EMP24/GP25 proteins. The class IV and V proteins should contain the same transmembrane and carboxy-terminal cytoplasmic domains as all loj products but have shortened amino-terminal lumenal domains for interacting with cargo molecules.

Oviposition:
The final deposition of an egg on a substrate is the culmination of a series of events that requires maturation and function of the necessary neural circuitry, ovulation of eggs, and transmittal of the egg through the genital tract to its final release. This passage through the ductwork to the uterus and final release can be defined as "oviposition proper." The genes that regulate all of these processes (excluding those with strictly developmental functions) can be considered "oviposition genes," since lack of proper function at any level of this process will lead to an oviposition phenotype.

Little is known about how the oviposition circuitry in Drosophila is generated. In the case of the grasshopper, a large portion of the oviposition neural circuit is completed by the end of embryonic development, well before it is needed for the behavior (THOMPSON and ROOSEVELT 1998 ). Therefore, some oviposition genes must function early in this process. Once the appropriate circuitry has been established, signaling must occur to initiate and maintain the behavior. The activity of motorneurons that directly synapse on the uterine and oviductal muscles is likely to be controlled by descending inputs from the command interneurons in the brain, including the subesophageal ganglion. Local circuit interneurons in the posterior abdominal ganglion and sensory inputs from neurons in the ovaries and internal reproductive tract also are expected to function in activating and modulating oviposition behavior (THOMPSON 1986A , THOMPSON 1986B ).

Relationship of the loj adult female phenotype to oviposition:
The female response to mating is increased ovulation and oviposition and decreased receptivity, processes that are regulated in part by products in the male ejaculate that are transferred to females upon mating (CHEN et al. 1988 ; MONSMA and WOLFNER 1988 ; AIGAKI et al. 1991; KUBLI 1992 ; HERNDON and WOLFNER 1995 ; WOLFNER 1997 , WOLFNER 2002 ; WOLFNER et al. 1997 ; HEIFETZ et al. 2000 , HEIFETZ et al. 2001 ; CHAPMAN et al. 2001 ; FLEISCHMANN et al. 2001 ; SAUDAN et al. 2002 ). In loj mutants both ovulation and oviposition are affected. We propose that the loss of signals from the CNS as well as the egg are responsible for the lack of oviposition behavior noted for loj mutant females and that disruption of ovulation is a secondary effect of the mutation.

One consequence of the failure of oviposition in loj mutants is that mature eggs are found inside the female genital tract, particularly in the uterus. Interestingly, the genital tracts of loj females are innervated as assayed by detection of synaptotagmin-positive synapses, which suggests that the egg-laying deficit may be due to loss of a required signal rather than simply to loss of motorneuronal input to the genital tract muscles. This hypothesis is supported by the homology that loj shares with components of COP-coated vesicles, which function in the secretory pathway in both anterograde and retrograde cytoplasmic transport of cellular components between the endoplasmic reticulum and Golgi as well as functioning in shuttling neurotransmitters to their release sites (ROTHMAN and WIELAND 1996 ; SCHEKMAN and ORCI 1996 ; LIU et al. 1999 ). The signaling molecule that is missing in loj mutants may need to mature by passage through the secretory system or may need to be loaded into neuronal vesicles. Such a signal could come from the nervous system, the egg itself, or from both tissues. Hence, the observed CNS and egg chamber expression have important implications for how the loj product functions in oviposition.

loj function in the CNS:
The brain and VNC are important centers for processing gustatory, olfactory, and visual inputs and transforming this information into an appropriate behavioral outcome such as oviposition. Previous work using mosaic animals, in which male and female tissue was judged only by external cuticular markers, mapped egg-laying behavior to the thorax (SZABAD and FAJSZI 1982 ), which contains the VNC. Our enhancer trap data suggest that loj is expressed in numerous cells in the thoracic and abdominal ganglia of the VNC (Fig 4).

The brain also has a role in oviposition regulation, since decapitated or anesthetized Drosophila females lay eggs as a reflex response (GROSSFIELD 1978 ); loj females not only do not lay eggs voluntarily, but also do not exhibit this reflex response to anesthesia (G. E. CARNEY and B. J. TAYLOR, unpublished results). While loj expression in the brain seems at odds with the idea that the brain has an inhibitory function on oviposition that must be released, there are possible explanations. One is that the brain provides excitatory as well as inhibitory signals for oviposition, with positive signaling being mediated by loj. However, releasing this inhibition may not be sufficient if a positive loj signal is needed from the VNC as well. Therefore, two positive functions of loj may have been lost in female mutants with the net result that oviposition cannot occur.

loj function in the egg:
A second interesting component of the adult expression pattern is the presence of loj in midstage vitellogenic egg chambers (Fig 5). The majority of this signal is found within the follicle cells that provide nutrients and other components to developing oocytes and produce the outer coverings of the egg, the vitelline membrane and the chorion (reviewed in MAHOWALD and KAMBYSELLIS 1980 ; SPRADLING 1993 ). One intriguing possibility is that these loj-positive egg cells provide cues to the female's genital tract and musculature that aid proper egg release from the ovary and navigation through the genital tract to the uterus. This signaling mechanism would be expected to function prior to formation of the vitelline membrane and chorion or to be a component of these protective coverings of the mature egg. Another possibility is that the observed expression in egg chambers is unrelated to the behavioral phenotype and that loj function is required maternally for proper embryonic development. The second hypothesis is supported by our observation that the eggs from rescued mutant females expressing hs-loj constructs do not hatch (G. E. CARNEY and B. J. TAYLOR, unpublished results). However, this observation does not preclude the possibility that loj has an earlier function in the egg that is rescued by loj overexpression.

As noted above, the process of ovulation is affected in loj females. Since a mature egg is found in the uterus of essentially every mutant female, it appears that initial egg release is not affected. However, we have observed partially ovulated eggs in the upper portions of the lateral oviducts as well as multiple eggs in these portions of the genital tract. In wild-type females we have not observed this phenomenon. We theorize that ovulation initially proceeds normally in loj females but that the presence of an unlaid egg in the uterus disrupts the feedback loop that regulates ovulation. Therefore, loj mutant females have a weak ovulation defect that is a secondary consequence of loss of oviposition behavior. An alternative hypothesis is that loj functions in both of these processes. The ovulation and oviposition phenotypes may be the consequence of the failure of eggs to signal their intentions to the reproductive tract, suggesting that signaling from the egg is an active process required for oviposition.

Relationship of loj to the somatic sex-determination hierarchy:
The genetic members of the somatic sex-determination hierarchy (e.g., tra, fru, dsx, and dsf) control female or male reproductive behaviors in Drosophila. The lodged-egg mutant phenotype and loj CNS expression pattern suggest the possibility that loj is regulated by or interacts with the sex-determination hierarchy to control female egg-laying behavior. The most likely possibilities are that loj functions downstream of dsx or is one of the determinants of oviposition behavior regulated by dsf. Two reproductive behaviors, receptivity and oviposition, are controlled by dsf, while loj affects only oviposition behavior. Additionally, while dsf mutants lack uterine innervation (FINLEY et al. 1997 ), loj uteruses are innervated. Preliminary real-time PCR experiments in which loj expression was assayed in dsf and dsx mutant females indicate that loj expression is not altered in these females at the whole-animal level (G. E. CARNEY and B. J. TAYLOR, unpublished results). This result implies that loj mRNA expression is not controlled directly by these genes and that loj interacts indirectly with the sex-determination hierarchy to regulate female behavior. However, it is possible that loj expression is regulated by sex-determination genes in subsets of cells or tissues and that this regulation is obscured by high levels of loj expression in tissues where loj is not controlled in this manner. An alternative is that certain loj transcripts are regulated by the sex-determination hierarchy while others are not, a possibility that was not tested in our real-time PCR experiments.

Further studies are now necessary to determine how expression and function of loj are related to oviposition behavior. The discovery that a putative cargo receptor is necessary for a behavioral output reveals an important and previously unrecognized use for this class of proteins. The loj mutants provide an opportunity to dissect the functional requirements for loj and other EMP24 proteins in the context of a multicellular organism. Studies of genes such as loj provide a means for understanding the regulation of behaviors in Drosophila as well as in other organisms, since signaling cascades utilized by the fruit fly are likely to be evolutionarily conserved.

	ACKNOWLEDGMENTS

We thank C. Berg for the loj⁰⁰⁸⁹⁸ insertion line. The loj⁰⁴⁰²⁶ and third chromosome deficiency strains were obtained from the Bloomington Drosophila Stock Center. T. J. Littleton supplied the anti-synaptotagmin antibody. N. J. Bowen provided the Drosophila EMP24 sequence alignments. M. Hedengren provided advice on Northern protocols. We thank the Central Services Lab of the Oregon State University Center for Gene Research and Biotechnology for use of the ABI Prism 7700 and Jason Monterecy at Massachusetts General Hospital for injection of transgenic constructs. This work was funded by National Institutes of Health grants GM-20030 (G.E.C.) and GM-56920 (B.J.T.).

Manuscript received November 20, 2002; Accepted for publication January 14, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMS, M. D., S. E. CELNIKER, R. A. HOLT, C. A. EVANS, and J. D. GOCAYNE et al., 2000  The genome sequence of Drosophila melanogaster.. Science 287:2185-2195.[Abstract/Free Full Text]

AHMAD, S. M. and B. S. BAKER, 2002  Sex-specific deployment of FGF signaling in Drosophila recruits mesodermal cells into the male genital imaginal disc. Cell 109:651-661.[Medline]

AIGAKI, T., I. FLEISCHMAN, P. S. CHEN, and E. KUBLI, 1991  Ectopic expression of sex peptide alters reproductive behavior of female D. melanogaster.. Neuron 7:557-563.[Medline]

BAKER, B. S. and K. A. RIDGE, 1980  Sex and the single cell. I. On the action of major loci affecting sex determination in Drosophila melanogaster. Genetics 94:383-423.[Abstract/Free Full Text]

BELDEN, W. J. and C. BARLOWE, 1996  Erv25p, a component of COPII-coated vesicles, forms a complex with Emp24p that is required for efficient endoplasmic reticulum to Golgi transport. J. Biol. Chem. 271:26939-26946.[Abstract/Free Full Text]

BOGGS, R. T., P. GREGOR, S. IDRISS, J. M. BELOTE, and M. MCKEOWN, 1987  Regulation of sexual differentiation in D. melanogaster via alternative splicing of RNA from the transformer gene. Cell 50:739-747.[Medline]

BURTIS, K. C., 1993  The regulation of sex determination and sexually dimorphic differentiation in Drosophila. Curr. Opin. Cell Biol. 5:1006-1014.[Medline]

BURTIS, K. C., K. T. COSCHIGANO, B. S. BAKER, and P. C. WENSINK, 1991  The Doublesex proteins of Drosophila melanogaster bind directly to a sex-specific yolk protein gene enhancer. EMBO J. 10:2577-2582.[Medline]

CARNEY, G. E. and M. BENDER, 2000  The Drosophila ecdysone receptor (EcR) gene is required maternally for normal oogenesis. Genetics 154:1203-1211.[Abstract/Free Full Text]

CAVENER, D. R. and S. C. RAY, 1991  Eukaryotic start and stop translation sites. Nucleic Acids Res. 19:3185-3192.[Abstract/Free Full Text]

CHAPMAN, T., L. A. HERNDON, Y. HEIFETZ, L. PARTRIDGE, and M. F. WOLFNER, 2001  The Acp26Aa seminal fluid protein is a modulator of early egg hatchability in Drosophila melanogaster. Proc. R. Soc. Lond. Ser. B Biol. Sci. 268:1647-1654.[Medline]

CHEN, P. S., E. STUMM-ZOLLINGER, T. AIGAKI, J. BALMER, and M. BIENZ et al., 1988  A male accessory gland peptide that regulates reproductive behavior of female D. melanogaster.. Cell 54:291-298.[Medline]

CLINE, T. W., and B. J. MEYER, 1996 Vive le difference: males vs. females in flies vs. worms, pp. 637–702 in Annual Review of Genetics, edited by A. CAMPBELL, W. ANDERSON and E. W. JONES. Annual Reviews, Palo Alto, CA.

CUIFO, L. F. and A. BOYD, 2000  Identification of a lumenal sequence specifying the assembly of Emp24p into p24 complexes in the yeast secretory pathway. J. Biol. Chem. 275:8382-8388.[Abstract/Free Full Text]

D'AVINO, P. P. and C. S. THUMMEL, 1999  Ectopic expression systems in Drosophila. Methods Enzymol. 306:129-142.[Medline]

DENZEL, A., F. OTTO, A. GIROD, R. PEPPERKOK, and R. WATSON et al., 2000  The p24 family member p23 is required for early embryonic development. Curr. Biol. 10:55-58.[Medline]

DUBNAU, J., and T. TULLY, 1998 Gene discovery in Drosophila: new insights for learning and memory, pp. 407–444 in Annual Review of Neuroscience, edited by W. M. COWAN. Annual Reviews, Palo Alto, CA.

FINLEY, K. D., B. J. TAYLOR, M. MILSTEIN, and M. MCKEOWN, 1997  dissatisfaction, a gene involved in sex-specific behavior and neural development of Drosophila melanogaster.. Proc. Natl. Acad. Sci. USA 94:913-918.[Abstract/Free Full Text]

FINLEY, K. D., B. J. TAYLOR, P. T. EDEEN, E. GROSS, and M. FOSS et al., 1998  dissatisfaction encodes a Tailless-like nuclear receptor expressed in a subset of CNS neurons controlling Drosophila sexual behavior. Neuron 21:1363-1374.[Medline]

FLEISCHMANN, I., B. COTTON, Y. CHOFFAT, M. SPENGLER, and E. KUBLI, 2001  Mushroom bodies and post-mating behaviors of Drosophila melanogaster females. J. Neurogenet. 15:117-144.[Medline]

FULLEKRUG, J., T. SUGANUMA, B. L. TANG, W. HONG, and B. STORRIE et al., 1999  Localization and recycling of gp27 (hp24g3): complex formation with other p24 family members. Mol. Biol. Cell 10:1939-1955.[Abstract/Free Full Text]

GOODWIN, S. F., B. J. TAYLOR, A. VILLELLA, M. FOSS, and L. C. RYNER et al., 2000  Aberrant splicing and altered spatial expression patterns in fruitless mutants of Drosophila melanogaster.. Genetics 154:725-745.[Abstract/Free Full Text]

GROSSFIELD, J., 1978 Non-sexual behavior of Drosophila, pp. 1–126 in The Genetics and Biology of Drosophila, edited by M. ASHBURNER and T. R. F. WRIGHT. Academic Press, London.

HALL, J. C., 1994  The mating of a fly. Science 264:1702-1714.[Abstract/Free Full Text]

HEIFETZ, Y., O. LUNG, E. A. FRONGILLO, and M. F. WOLFNER, 2000  The Drosophila seminal fluid protein Acp26Aa stimulates release of oocytes by the ovary. Curr. Biol. 29:99-102.

HEIFETZ, Y., U. TRAM, and M. F. WOLFNER, 2001  Male contributions to egg production: the role of accessory gland products and sperm in Drosophila melanogaster.. Proc. R. Soc. Lond. B Biol. Sci. 268:175-180.[Medline]

HERNDON, L. A. and M. F. WOLFNER, 1995  A Drosophila seminal fluid protein, Acp26Aa, stimulates egg laying in females for 1 day after mating. Proc. Natl. Acad. Sci. USA 92:10114-10118.[Abstract/Free Full Text]

HILDRETH, P. E., 1965  Doublesex, a recessive gene that transforms both males and females of Drosophila into intersexes. Genetics 51:659-678.[Free Full Text]

HOFMANN, K. and W. STOFFEL, 1993  TMbase—a database of membrane spanning protein segments. Biol. Chem. 374:1666.

ITO, H., K. FUJITANI, K. USAI, K. SHIMIZU-NISHIKAWA, and S. TANAKA et al., 1996  Sexual orientation in Drosophila is altered by the sartori mutation in the sex-determination gene fruitless that encodes a zinc finger protein with a BTB domain. Proc. Natl. Acad. Sci. USA 93:9687-9692.[Abstract/Free Full Text]

JALLON, J.-M., 1984  A few chemical words exchanged by Drosophila during courtship and mating. Behav. Genet. 14:441-478.[Medline]

KARPEN, G. and A. SPRADLING, 1992  Analysis of subtelomeric heterochromatin in the Drosophila minichromosome Dp1187 by single P element insertional mutagenesis. Genetics 132:737-753.[Abstract]