Genetic Analysis of the Drosophila Gs Gene

Corresponding author: Michael Forte, Vollum Institute, L474, Oregon Health Sciences University, 3181 S.W. Sam Jackson Park Rd., Portland, OR 97201., forte{at}ohsu.edu (E-mail)

Communicating editor: T. C. KAUFMAN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

One of the best understood signal transduction pathways activated by receptors containing seven transmembrane domains involves activation of heterotrimeric G-protein complexes containing Gs, the subsequent stimulation of adenylyl cyclase, production of cAMP, activation of protein kinase A (PKA), and the phosphorylation of substrates that control a wide variety of cellular responses. Here, we report the identification of "loss-of-function" mutations in the Drosophila Gs gene (dgs). Seven mutants have been identified that are either complemented by transgenes representing the wild-type dgs gene or contain nucleotide sequence changes resulting in the production of altered Gs protein. Examination of mutant alleles representing loss-of-Gs function indicates that the phenotypes generated do not mimic those created by mutational elimination of PKA. These results are consistent with the conclusion reached in previous studies that activation of PKA, at least in these developmental contexts, does not depend on receptor-mediated increases in intracellular cAMP, in contrast to the predictions of models developed primarily on the basis of studies in cultured cells.

AN expanding array of extracellular signals mediates cellular responses through a family of receptors possessing seven transmembrane domains (7TMR). Essential components of the intracellular signal transduction cascade initiated by activation of these receptors are intermediary heterotrimeric G-proteins composed of -, ß-, and -subunits that couple receptors to appropriate intracellular effectors (for recent reviews see HAMM 1998 ; MORRIS and MALBON 1999 ). In this pathway, ligand binding by a receptor drives a conformational change in the -subunit of the heterotrimeric G-protein, catalyzing the exchange of bound GDP for GTP. The activated, GTP-bound form of the -subunit then dissociates from the ß-subunits. Each released component can then activate a variety of downstream effectors. The signal is terminated by hydrolysis of the GTP to GDP by a GTPase activity intrinsic to the -subunit, allowing its reassociation with the ß-dimer, thereby reforming the initial complex. In the case of some G-proteins, the GTPase activity of the -subunit can be stimulated by a family of proteins containing a conserved RGS domain (DE VRIES and GIST-FARQUHAR 1999 ). This signaling pathway plays a key role in triggering physiological responses to a wide variety of hormones, neurotransmitters, and sensory stimuli and is one of the most evolutionarily ancient forms of transmembrane signal transduction known in eukaryotes. In this scheme, the -subunit plays a key role since it is responsible for coordinating the coupling of receptors to appropriate effectors (NEER 1995 ; HAMM 1998 ). There are hundreds of receptors that mediate their intracellular effects through this pathway, yet only 16 genes are known to encode heterotrimeric G-protein -subunits in mammals (NEER 1995 ). Thus, the -subunits can be considered a "bottleneck" in this transduction cascade since many 7TMR may couple to the same -subunit to mediate cellular responses in different contexts.

Perhaps the best understood example of this signal transduction pathway involves 7TMR coupling to heterotrimeric G-protein complexes containing -subunits in the Gs family, one of the first proteins in this family to be identified (GILMAN 1989 ). When activated, Gs stimulates membrane-bound, metazoan adenylyl cyclases (AC), resulting in the elevation of intracellular levels of the key second messenger, cAMP (TESMER and SPRANG 1998 ; HURLEY 1999 ; SIMONDS 1999 ). Although adenylyl cyclases, as coincidence detectors, can also be activated by a number of other mechanisms, receptor-dependent elevation of cAMP is mediated primarily through Gs (BOURNE and NICOLL 1993 ; SMIT and IYENGAR 1998 ). The primary role of intracellular cAMP has been extensively studied and is believed to involve the activation of protein kinase A (PKA; INSEL et al. 1975 ; COFFINO et al. 1976 ; GOTTESMAN 1980 , GOTTESMAN 1985 ; GOTTESMAN et al. 1980 ). Binding of cAMP by inhibitory regulatory subunits of PKA mediates their dissociation from catalytic subunits, thereby leading to kinase activation (FRANCIS and CORBIN 1994 ). PKA-mediated phosphorylation of a variety of substrates in turn controls diverse cellular phenomena such as metabolism, development, cell proliferation, gene transcription, and learning and memory. In addition, cell-specific responses to activation of Gs pathways, which may be based on selective expression of receptor and AC isoforms (COOPER et al. 1995 ), the selective activation of specific intracellular pools of PKA (COLLEDGE and SCOTT 1999 ; EDWARDS and SCOTT 2000 ; SKALHEGG and TASKEN 2000 ), or activation of other effectors of cAMP such as cyclic nucleotide-gated channels (BROILLET and FIRESTEIN 1999 ; BIEL et al. 1999 ), have also been documented. However, the exact contribution of any of these factors to the cAMP-mediated responses in specific cell types has yet to be precisely defined.

The identification of a number of alterations in the human Gs protein as the cause of a variety of human diseases, as well as genetic analysis of the gene encoding Gs in mice and Caenorhabditis, has emphasized the critical role of this pathway in metazoan development and homeostasis. Albright's hereditary osteodystrophy, which is characterized by multihormone resistance, is associated with heterozygosity for loss-of-function mutations in the single human Gs gene, although the spectrum of defects present depends on whether the mutant gene is inherited maternally or paternally (for reviews see SPIEGEL 1997A , SPIEGEL 1997B ; FARFEL et al. 1999 ). Homozygous null individuals have not been identified and may not survive, presumably due to major developmental defects resulting in embryonic lethality. Consistent with this conclusion, mice in which the expression of Gs has been eliminated die at an early embryonic stage (E10.5; YU et al. 1998 ). In addition, loss-of-function mutations in the Caenorhabditis gene encoding Gs, gsa-1, result in embryonic lethality (KORSWAGEN et al. 1997 ). Conversely, mutations that cripple the GTPase activity of the Gs protein, thereby leading to constitutive activation of the pathway or "gain-of-function" phenotypes, are found in a variety of pituitary and thyroid malignancies and in a number of endocrine disorders such as McCune-Albrights syndrome (SPIEGEL 1997A ; FARFEL et al. 1999 ). Interestingly, disorders resulting from gain-of-function mutations in Gs arise somatically and are not heritable, indicating that gain-of-function, as well as loss-of-function, mutations lead to major developmental abnormalities. In addition, expression of mutationally activated forms of Gs in Caenorhabditis leads to neuronal degeneration (KORSWAGEN et al. 1997 , KORSWAGEN et al. 1998 ; BERGER et al. 1998 ).

In all mammalian systems, the ability to examine in detail the developmental consequences of genetic alteration of the pathway defined by Gs are limited by the intractability of the embryo and the difficulty of genetic manipulations. By contrast, the ease with which genetic and embryological manipulations are carried out in Drosophila makes it a valuable system in which to examine mutations in the Gs pathway and their consequences for development. Sequence analysis of the Drosophila genome indicates that over 200 proteins that contain the seven transmembrane domains typically found in 7TMR are encoded (BRODY and CRAVCHIK 2000 ). Previous studies have identified and characterized a number of fly receptors (primarily for biogenic amines) that activate mammalian Gs in cultured cell systems, resulting in the stimulation of adenylyl cyclase and increases in cAMP (WITZ et al. 1990 ; SUGAMORI et al. 1995 ; HAN et al. 1996 , HAN et al. 1998 ). In addition, six genes in Drosophila encode G-protein -subunits and three encode ß- and -subunits (ADAMS et al. 2000 ). The Drosophila genome also encodes at least seven adenylyl cyclase isoforms (CANN and LEVIN 1998 ; BRODY and CRAVCHIK 2000 ), one of which (that encoded by the rutabaga gene) has been shown to be activated by GTP-bound Gs, as are all known isoforms of mammalian adenylyl cyclase (LEVIN et al. 1992 ; COOPER et al. 1995 ; SIMONDS 1999 ). In addition, phenotypes generated by mutations in the dunce gene, encoding a cAMP-specific phosphodiesterase responsible for the degradation of cAMP, have been extensively studied at a number of levels (DAVIS and DAUWALDER 1991 ; NIGHORN et al. 1994 ; DAVIS 1996 ). These results, combined with the presence in Drosophila of a number of PKA isoforms (KALDERON and RUBIN 1988 ; LANE and KALDERON 1993 ; MELENDEZ et al. 1995 ), suggest that genetic manipulation in Drosophila may provide an important system in which to define the role of this pathway in the development and function of specific cell types and to examine how signals generated on activation of this pathway are integrated with other signaling events.

Past work has shown that a single Drosophila gene is responsible for encoding a Gs protein that is highly homologous to mammalian orthologs (71% identity; QUAN et al. 1989 ; QUAN and FORTE 1990 ). Drosophila Gs has been demonstrated to functionally complement the absence of mammalian Gs in specific somatic cell line mutants, demonstrating that this protein is the functional, as well as structural, homolog of mammalian Gs (QUAN et al. 1991 ). Although ubiquitously expressed in all cells, highest levels of Gs protein are found in the Drosophila nervous system (WOLFGANG et al. 1990 , WOLFGANG et al. 1991 ). In addition, the phenotypic consequences of constitutive activation of Gs pathways have been examined by the regulated expression of mutant Drosophila Gs proteins in which the intrinsic GTPase activity of the protein was eliminated (WOLFGANG et al. 1996 ; CONNOLLY et al. 1996 ; CHYB et al. 1999 ). The gain-of-function phenotypes generated in these studies depended on the precise temporal and spacial pattern of mutant Gs expression, indicating that the role of Gs-mediated signaling depends in large part on the cellular context. Furthermore, we found that, in one case, the phenotype produced by activation of Gs pathways (alterations in cellular adhesion of wing epithelia) did not require the presence of PKA (WOLFGANG et al. 1996 ). This observation suggests the existence of additional primary effectors of Gs activation, and perhaps cAMP, other than PKA in at least this cell type, if not others.

To complement these gain-of-function studies, in this article we report the results of an F₂ lethal screen aimed at the identification of loss-of-function mutations in the Drosophila Gs gene (dgs). Seven mutants that are either complemented by transgenes representing the wild-type dgs gene or contain nucleotide sequence changes resulting in the production of altered Gs protein have been identified. Examination of the phenotypes generated by mutations representing loss of Gs function suggest that the activity of PKA in these contexts does not depend on signaling through Gs, since loss of Gs does not result in phenotypes generated on mutational elimination of PKA, consistent with conclusions reached in previous examination of gain-of-function phenotypes (WOLFGANG et al. 1996 ). Thus, our results are consistent with the view that receptor-dependent activation of adenylyl cyclase through Gs in the production of cAMP is not necessarily coupled in a dependent fashion to the activation of PKA.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks, EMS mutagenesis, and dgs rescue constructs:
Stocks were obtained either from the Bloomington Stock Center or colleagues (S. Smolik and D. Kalderon; Table 1). Flies were mutagenized by first starving 2- to 4-day-old cn;ry⁵⁰⁶, e^s males for 6 hr. Starved flies were then allowed access to 24 mM ethyl methanesulfonate (EMS; Sigma, St. Louis) in 5% sucrose for 21 hr. Treated males were then crossed en masse to CyO/Gla females. Subsequent crosses were as described in RESULTS. The Drosophila gene encoding Gs [dgs; also known as G-salpha60A (ADAMS et al. 2000 )] maps to position 60A12 on the second chromosome (QUAN et al. 1989 ; QUAN and FORTE 1990 ). To remove irrelevant recessive lethal mutations introduced during EMS mutagenesis, we recombined four alleles (dgs^B19, dgs^R19, dgs^R60, and dgs^R79) onto a cn¹,bw¹,sp¹ chromosome by selecting for the absence of sp¹ and acquisition of bw¹ on chromosomes containing dgs mutant alleles, thereby leaving only a small amount of the originally mutated 2R distal to 59E on recombinant chromosomes. We generated three to five recombinant lines for each allele and each of the recombinants except dgs^R19 could be rescued as homozygotes by introduction of a dgs transgene (see below), demonstrating that these lines had no second site recessive lethal mutations. Since dgs^R19 could not be rescued by introduction of the dgs transgene, the chromosome carrying this allele is likely to contain a linked, second site recessive lethal mutation, presumably in the region distal to 59E derived from the original mutagenized chromosome.

A strain carrying a dgs transgene rescue construct (Gs27) was created by subcloning a 10-kb BamHI fragment containing the entire dgs gene and genomic sequences representing roughly 2 kb 5' to the transcriptional initiation site and 2 kb 3' to the translational stop site into pCASPR4 (PIRROTTA 1988 ; QUAN and FORTE 1990 ). Transformants were generated by standard procedures. The Gs27 transgene used in this study is located on the X chromosome.

Sequence analysis of mutants:
Eggs were collected from mutant stocks balanced over a SM6 balancer chromosome containing the eve promoter driving ß-galactosidase (ß-gal). Detection of ß-gal activity was performed by dechorionating embryos with 50% bleach, then permeabilizing by washing with isopropanol, n-hexane, Drosophila Ringer solution and modified Drosophila Ringer's solution (MDR; 9 mM MgCl₂, 10 mM MgSO₄, 3 mM NaH₂PO4, 68 mM glutamic acid, 67 mM glycine, 4 mM malic acid, 0.2 mM Na acetate, 5 mM CaCl₂, and 0.1% bovine serum albumin). Embryos were stained for ß-gal activity by resuspension in 1 ml of 0.5 MDR containing 60 µl of 20 mg/ml X-gal dissolved in dimethyl formamide, 40 µl 0.1 M K₄Fe(CN)₆, and 40 µl 0.1 M K₃Fe(CN)₆. Several homozygous mutant embryos identified by the absence of ß-gal staining were removed and stored in MDR at -80°. DNA from mutant embryos was extracted by homogenizing embryos in 10 µl of extraction buffer (GLOOR et al. 1991 ).

The dgs gene contains nine exons (QUAN and FORTE 1990 ) including alternate initial exons encoding 5' untranslated sequences. PCR amplification of the dgs gene coding sequence in mutant embryos employed three sets of primers generating products that represent exons 2–4 (5' CGGAATTCCTGGAGCAGAACGGAGAAAGCAC 3' and 5' CGGGATCCGTTTGGCTCAACACATCGATGGC 3'), exons 5–7 (5' CGGAATTCGCTTGATTAAGGCACTGAAACCG 3' and 5' CGGGATCCCGAATTTCTTGGCCAGATCTGAG 3'), and exons 8 and 9 (5' CGGAATTCCTCAGATCTGGCCAAGAAATTCGC 3' and 5' CGGGATCCGGTCATACCAGATCTAAATGCAGCG 3').

In each case, amplification was done in 40-µl reactions containing 2 µl DNA prepared from homozygous mutant embryos (35 cycles consisting of 94° for 30 sec, 55° for 30 sec, and 72° for 30 sec). The products were purified and cloned into pBluescript (Stratagene, La Jolla, CA) following digestion with EcoRI and BamHI, and the inserts were sequenced. The position of individual mutations was confirmed by sequence analysis of DNA prepared from three independent batches of mutant embryos.

Lethal phase analysis:
To examine embryonic lethality from zygotic mutations, balanced dgs mutant flies were outcrossed to wild-type (Canton-S) flies to remove lethality associated with the balancer. The unbalanced mutants were then crossed inter se and 25% of the fertile eggs assumed to be homozygous mutant. To assess embryonic lethality of progeny from mothers with germline dgs clones, progeny were crossed to males carrying dgs mutations over a CyO balancer containing a green fluorescent protein (GFP) expression construct. Nonfluorescent eggs that contained unbalanced maternally and paternally mutant embryos were collected and the percentage of dead embryos determined. To determine the percentage of dead dgs mutant embryos in each case, eggs were collected on egg plates and the number of unhatched eggs determined 24 hr after removal of the adults. Unhatched eggs were aged another 3 to 5 days and eggs that turned brown after this incubation were scored as dead embryos, and white eggs were scored as unfertilized; unfertilized eggs were eliminated from further analysis.

Generation of clones:
Mosaic females carrying dgs germline clones were generated using the FLP-dominant female sterile (DFS) system (CHOU and PERRIMON 1992 , CHOU and PERRIMON 1996 ). Mutant alleles were recombined on to P(w⁺, FRT)^42B chromosomes and yw P{w = hs-flp}122; P {w⁺ FRT)^42B bw dgs*/CyO females were crossed to P(w⁺ FRT)42B P (w⁺ ovo^D1)^32X9/CyO males. Progeny from this cross were heat shocked by incubation at 38° for 1 hr, 46 to 60 hr after egg laying. Resulting dgs*/ovo^D1 virgins were then collected and crossed to balanced dgs*/CyO-GFP males, and nonfluorescent mutant progeny were examined.

Histology:
For whole mount embryo preparations, embryos were dechorionated in bleach and fixed for 20 min with vigorous shaking in a 1:1 mixture of heptane and 4% formaldehyde (Fluka) in phosphate-buffered saline (PBS). Prior to devitellinization, the embryos were collected and washed in PBS containing 0.3% Triton X-100 (PBT; Sigma) and stained for ß-gal as described above with the addition of 0.3% Triton X-100. Once a strong ß-gal reaction was observed, embryos were devitellinized by vortexing in a 1:1 mixture of heptane and methanol. The devitellinized embryos were hydrated and fixed for 20 min in 4% formaldehyde in PBS and then washed for 1 hr with three changes of PBT containing 10% horse serum. Embryos were then incubated overnight at 4° in a 1:7500-fold dilution of an antisera generated to a unique peptide found at the C terminus of all Gs proteins (WOLFGANG et al. 1990 , WOLFGANG et al. 1991 ). The antibody binding was then detected using the ABC elite HRP kit (Vector, Burlingame, CA). Embryos were dehydrated, transferred to methylsalycylate (Sigma), and mounted in a 1:4 mixture of methylsalycylate and Permount.

Larval locomotion assay:
Larval crawling activity was measured in a manner similar to that described in PEREIRA et al. 1995 . Third instar feeding-stage larvae were collected from staged, uncrowded vials. Single larvae were placed in the center of an egg plate coated with a thin, uniform layer of yeast. After 5 min at 21°, the path made by larvae through the yeast was traced onto the lid of the petri dish. The tracings were scanned and the path length was measured using National Institutes of Health IMAGE.

Measurement of cAMP:
To assess cAMP levels in dgs mutants, nonfluorescent first instar larvae from stocks of individual dgs mutations maintained over a CyO balancer chromosome carrying a GFP expression construct were collected by hand and stored at -80°. Larvae were resuspended in ice-cold 5% trichloroacetic acid and homogenized by sonication. Extracts were then spun at 4° at 4000 x g for 15 min. Resulting pellets were then assayed for protein content and supernatants extracted two times with acidified ether. Following extraction, supernatants were dried and residue dissolved in cold 0.5 ml 0.1 M Tris-HCl ph 7.5, 0.01 M EDTA. Levels of cAMP were determined by competitive binding to cAMP binding protein using a commerical kit (Diagnostic Products).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of mutations in the Drosophila Gs gene:
Since loss-of-function mutations in genes encoding Gs in both mice and Caenorhabditis result in lethality, we performed an F₂ lethal screen to recover mutations in the gene (dgs) encoding the Drosophila Gs protein. The screen was based on the observation that immunoblots of extracts prepared from Df(2R)or^BR11, cn, bw, sp/SM6, eve-lacz, Roi flies probed with Gs-specific antibodies had approximately half the amount of immunoreactive protein compared to controls (not shown), suggesting that this deficiency eliminates the dgs gene. Thus, male cn;ry⁵⁰⁶, e^s flies were mutagenized with EMS and crossed en masse to CyO/Gla females. Single F₁ CyO or Gla males were then mated to virgin Df(2R)or^BR11, cn, bw, sp/SM6, eve-lacz, Roi females. From roughly 4000 such crosses, we recovered 120 balanced mutations that were lethal over this deficiency. Of these, 47 were excluded from further consideration since they were lethal over In(2LR) lt^G10 /Cy Roi, an overlapping deficiency that does not remove dgs as assessed by immunoblot analysis (not shown). The 73 remaining balanced lethal mutations were crossed to deficiency stocks containing a dgs rescue transgene Gs27; df(2R) or^BR11, cn, bw, sp/CyO (MATERIALS AND METHODS). Seven of the lethals were shown to be in the same complementation group and all but dgs^R19 (which carries a linked, second site recessive lethal mutation; see MATERIALS AND METHODS) were rescued by the Gs27 transgene. These are denoted as the B19, R19, R33, R60, R65, R67, and R79 alleles of the dgs gene.

To further confirm that these alleles represent mutations in the gene encoding Drosophila Gs, the nucleotide sequence of several of the dgs alleles (dgs^R60, dgs^R19, and dgs^B19) was determined. The dgs gene contains nine exons that include alternate initial exons encoding 5' untranslated sequences (QUAN and FORTE 1990 ). Sequence analysis of the coding region (exons 2–9) of three independent preparations of genomic DNA from homozygous mutant embryos for each allele demonstrated that dgs^R60 is generated by the change of nucleotide 723 from a T to an A, resulting in the change of residue 241 from a Tyr to a stop codon; dgs^R19 by the change of nucleotide 795 from a G to an A, resulting in the change of residue 265 from a Trp to a stop codon; and dgs^B19 by the change of nucleotide 1117 from an A to a T, resulting in the change of residue 373 from an Ile to a Phe. These results demonstrate that the mutants identified in this screen represent mutations in the gene encoding the Drosophila Gs protein.

Lethal phase analysis:
Initial characterization of all alleles except dgs^B19 indicated that each results in lethality at late embryonic or early first instar stages. To quantify the extent of embryonic lethality, we first outcrossed balanced mutant lines to Canton-S flies to remove lethality associated with the chromosomes not carrying the dgs gene. Unbalanced offspring were crossed in heteroallelic or homoallelic combinations and to Df(2R)or^BR11, cn, bw, sp/SM6a. In such crosses, 25% of all progeny contain two mutant alleles or, in crosses to the deficiency, progeny were hemizygous for the individual alleles. Table 2 shows that in heteroallelic combinations, the percentage of mutant embryos that die is low, ranging from 5 to 14%, while embryonic lethality was generally higher when the four alleles were tested as homozygotes (6 to 61%) or hemizygotes (10 to 34%). In all allelic combinations and in hemizygotes (except those containing dgs^B19), larvae died either as they hatched or shortly after hatching. First instar larvae that successfully hatched displayed little or no movement and did not grow.

In contrast, hatched dgs^B19 mutant larvae, as homozygotes, trans-heterozygotes, or hemizygotes, survived for varying lengths of time, a very few becoming pharate adults that never eclosed. dgs^B19 mutant larvae were lethargic, grew more slowly, and were thinner and more transparent due to reduced amounts of fat body compared to their heterozygous siblings (Fig 1). In uncrowded vials, pupation by homozygous dgs^B19 larvae was delayed at least 1 day relative to balanced siblings. dgs^B19 homozygous pupae were deformed due to incomplete shortening of the body during pupariation. In addition, the larval mouth hooks were often not withdrawn into the pupal case. Of the pupae examined (n = 885), 30% of dgs^B19 larvae pupate while pupation by larvae hemizygous for dgs^B19 occurred at a lower rate (11%; 573 pupae examined). On rare occasions, homozygous dgs^B19 pharate adults were found in vials. These had a normal external adult morphology but when removed from the pupal case were immobile and, when left undisturbed, never eclosed. Thus, substantial numbers of dgs^B19 homozygote and hemizygote larvae pupated, but subsequently died at all pupal stages before adult eclosion.

View larger version (84K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Comparison of (A) control dgsB19/ CyO-GFP to (B) dgsB19/ dgsB19 homozygote mutant, wandering stage third instar larvae. Note that dgsB19 mutant larvae are similar in length but thinner and contain less fat body. A similar phenotype is seen for dgsB19 hemizygotes.

To assess whether maternal Gs contributed to embryonic survival, we examined embryos from females with mutant germlines. A number of the dgs mutants (dgs^R19, dgs^R60, dgs^R79, and dgs^B19) were first recombined onto a cn, bw, sp chromosome (MATERIALS AND METHODS) and then recombined onto a chromosome containing an FRT site at 42B. Using the FRT-ovoD system (CHOU and PERRIMON 1992 , CHOU and PERRIMON 1996 ), we examined embryos that were maternally as well as zygotically mutant for each allele. The presence of a wild-type or mutant paternal allele was indicated by the presence or absence, respectively, of GFP expressed from an insert on the paternal balancer chromosome (MATERIALS AND METHODS). Ovaries in females carrying dgs germline mutations were apparently normal in all respects (W. J. WOLFGANG and M. FORTE, unpublished observations). Table 3 shows that, for all alleles except dgs^B19 and dgs^R60, the percentage embryonic lethality increased in mothers with germline clones compared to zygotic mutants of the same genotype. However, in most cases (except dgs^R19 homozygotes and all dgs^B19 combinations), from 40 to 70% of the mutant embryos still hatch despite the absence of the germline as well as the zygotic dgs gene. The higher rate of embryonic lethality in dgs^R19 homozygotes may be due to a second site recessive mutation carried on this chromosome since this was the only allelic combination that could not be rescued with the dgs transgene. As noted above, for all alleles except for dgs^B19, mutant embryos that did hatch showed little movement, no growth, and died shortly after hatching. Similar to the results obtained in the case of zygotic mutant embryos, larvae that hatched from mothers with dgs^B19 maternal clones died at varying times throughout all postembryonic stages. A few developed to pharate adults that never eclosed. Finally, all maternally mutant embryos were rescued by a paternally contributed dgs gene to produce fertile adults.

Measurement of cAMP levels:
A primary mechanism for establishing intracellular cAMP concentrations is the control of its synthesis by receptor regulation of adenylyl cyclase through Gs. To determine the contribution of signaling through Gs to the total level of cAMP in Drosophila, the effect of null (dgs^R60 and dgs^R79) and hypomorphic (dgs^B19) dgs mutations on basal cAMP levels was determined in homozygous first instar larvae and compared to control Canton-S larvae. Larvae homozygous for the dgs^R19 mutation were excluded from this analysis due to low hatching rates and the potential influence of a linked, second site, lethal mutation (MATERIALS AND METHODS). In each case, the cAMP levels in mutant larvae were significantly lower than observed in controls (P < 0.001). The mean cAMP concentration in Canton-S larvae was 59.3 ± 1.3 pmol/mg protein (mean ± SD; n = 3). Null dgs mutations reduced this level roughly four- to fivefold (dgs^R60 = 14.7 ± 0.76 pmol/mg protein, n = 3; dgs^R79 = 12.8 ± 0.75 pmol/mg protein, n = 3). In contrast, the hypomorphic dgs^B19 mutation resulted in cAMP levels that were reduced 40% (37.7 ± 1.5 pmol/mg protein, n = 3) when compared to control larvae. These results demonstrate that cellular levels of cAMP in Drosophila are determined in large part by signaling through Gs.

Expression of Gs-protein in mutant embryos:
Immunolocalization studies have shown that Gs protein is expressed in all embryonic cell types, with high levels present in the forming embryonic neuropil (WOLFGANG et al. 1990 , WOLFGANG et al. 1991 ). To determine whether individual dgs alleles eliminate Gs immunoreactivity, we examined embryonic Gs expression in progeny from the crosses described above in which maternally contributed Gs protein had been eliminated. Fig 2 shows stage 15–16 embryos double stained for ß-gal activity and Gs protein. In embryos that received a paternal dgs gene as indicated by the blue ß-gal product (Fig 2B, Fig D, Fig F, and Fig H), there was elevated Gs staining in the neuropil with lower, uniform levels of staining throughout the rest of the embryo, as was observed in wild-type embryos (WOLFGANG et al. 1991 ). Embryos maternally and zygotically mutant for dgs^B19 showed the same level and pattern of Gs staining as heterozygous siblings (compare A and B in Fig 2). In contrast, embryos generated from maternal clones homozygous for the other alleles examined (dgs^R19, dgs^R60, and dgs^R79) showed a complete absence of Gs staining compared to their heterozygous sibs. We also examined the expression of Gs in homozygous dgs^R33, dgs^R65, and dgs^R67 mutant embryos generated from heterozygous flies. Although these embryos contained low, uniform levels of Gs protein, presumably maternal in origin, significant neuropil staining due to zygotic expression was present only in dgs^R33 embryos, whereas dgs^R65 and dgs^R67 contained no detectable Gs staining in the central nervous system (data not shown).

View larger version (125K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Immunolocalization of Gs in stage 15 and 16 embryos. (A, C, E, and G) Females with germline mutations for dgs alleles dgsB19, dgsR19, dgsR60, and dgsR79, respectively, in trans to a different null allele (either dgsR19 or dgsR60). (B, D, F, and H) Balanced heterozygotes from females with germline mutations for dgs alleles dgsB19, dgsR19, dgsR60, and dgsR79, respectively. Note that for alleles dgsR19, dgsR60, and dgsR79 there is no detectable Gs staining in the embryos (compare C, E, and G to D, F, and H). In contrast, dgsB19/dgsB19 embryos show wild-type levels of Gs (compare A and B). In addition, mutant embryos appear morphologically normal at the resolution of these images. The blue ß-gal staining indicates the presence of the CyO balancer chromosome.

Embryonic patterning in dgs mutants:
Embryos deficient in PKA show a variety of morphological defects, including alterations in cuticular patterning (LANE and KALDERON 1993 ; OHLMEYER and KALDERON 1997 ). To compare the effect of individual dgs mutations on embryonic pattern formation we examined cuticles from late stage mutant embryos or early first instar larvae generated from germline clones as well as those zygotically mutant. We chose to examine the dgs mutations in heteroallelic combinations to eliminate the potential effect of second site, recessive mutations on patterning. All dgs mutants that hatched had normal cuticle patterns. In addition, most dead embryos also had normal cuticle patterns (Table 4 and Fig 3), with the exception of dgs^B19 mutant embryos generated from germline clones. In this case, of the mutant embryos that did not hatch, 30% showed defects in the telson formation, 30% exhibited telson defects combined with posterior abdominal segment defects, and 30% were wild type. However, since only 4% of the total dgs^B19/dgs^R19 mutant population failed to hatch, the actual percentage of dgs^B19 mutant embryos and first instar larvae showing patterning defects is only 3%, similar to that observed for other alleles (1–4% of the total mutant population). Thus, 95–99% of all mutant embryos show normal cuticular patterning.

View larger version (45K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Cuticle from a dgsR60/dgsR19 mutant embryo in which the mother carried a germline mutant clone for dgsR60. The cuticle pattern is normal, as is generally the case in all mutant backgrounds (Table 4).

Analysis of the behavior of dgs^B19 larvae:
Observation of dgs^B19 mutant larvae on egg plates and in bottles indicated that they had sluggish, uncoordinated movements. To quantify larval mobility and activity, we assayed the behavior of homozygous dgs^B19 third instar larvae using the "rover" assay (PEREIRA et al. 1995 ). Individual larvae were placed on an egg plate uniformly coated with yeast and, after 5 min, the length of the track left in the yeast by the larvae was measured. These data were then represented as histograms showing the number of larvae (y-axis) that crawled a given distance (x-axis). Although there is some individual overlap, Fig 4 clearly demonstrates that, as a population, larvae homozygous (Fig 4A) or hemizygous (Fig 4B) for dgs^B19 crawl shorter distances than heterozygous controls (Fig 4C). (Fig 4A, 2.3 ± 0.28 cm mean ± SE, P = 0.0001; Fig 4B, 1.1 ± 0.17 cm mean ± SE, P = 0.0001; Fig 4C, 9.4 ± 0.63 cm mean ± SE; P values compare mutants to controls using an unpaired Student's t-test). There appears to be no dominant effect of the mutation on activity since dgs^B19/CyO-GFP individuals showed crawling activity similar to a variety of controls (data not shown). The reduced crawling observed for dgs^B19/dgs^B19 homozygotes was rescued by introduction of the dgs transgene (Fig 4D; 9.8 ± 0.465 cm mean ± SE). During this analysis, it was noted that some dgs^B19 mutant larvae often crawled in continuous circles, backward, or on their backs for extended periods of time, behaviors that are not observed in heterozygous or nonmutant larvae. It was also noted that dgs^B19 mutants appear to not be attracted to yeast granules on the egg plates, suggesting sensory-motor deficits. These results indicate that dgs^B19 mutant larvae have severe defects in neural and/or muscle physiology.

View larger version (27K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Histograms showing the number of feeding third instar larvae (y-axis) that crawled a given distance in 5 min (rounded up to the nearest centimeter; x-axis). For dgsB19 homozygotes (A, dgsB19/ dgsB19) and hemizygotes (B, dgsB19/ Df), the histograms are clearly shifted to the left when compared to heterozygous controls (C, dgsB19/ CyO-GFP). Gs27 transgene can rescue the dgsB19 crawling phenotype (D, Gs27; dgsB19/ dgsB19). Means ± SE in centimeters are as follows: A = 2.3 ± 0.28, P < 0.0001; B = 1.1 ± 0.17, P < 0.0001; C = 9.4 ± 0.63; D = 9.8 ± 0.465. P values determined by unpaired t-test comparing the means of A and B to either C or D or C and D are not significantly different; P > 0.6.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

One of the most evolutionarily ancient forms of transmembrane signal transduction known in eukaryotes utilizes a receptor that modulates the activity of intracellular second-messenger systems through the activation of intermediary heterotrimeric G-proteins (MORRIS and MALBON 1999 ). Our goal has been to genetically manipulate the activity of a specific G-protein -subunit, Gs, which has been traditionally viewed as a key component of the pathway that results in receptor-mediated activation of adenylyl cyclase and, thereby, of increases in intracellular cAMP concentration. In earlier studies, we examined the consequence of constitutive activation of Gs (and presumably the entire pathway downstream of Gs) through the restricted expression of site-directed mutants of Gs in which the GTPase activity of the protein was crippled, resulting in gain-of-function phenotypes (CONNOLLY et al. 1996 ; WOLFGANG et al. 1996 ). In this study, we now report the identification and characterization of mutations in the Drosophila Gs gene, resulting in loss-of-function phenotypes.

In a standard F₂ lethal screen, seven recessive mutations defining a single complementation group were identified. These mutations were defined as representing mutations in the gene encoding the Drosophila Gs protein, dgs, on the basis of two criteria. First, six of the seven mutations could be completely rescued as homozygotes by a transgene representing the entire dgs gene. Second, nucleotide sequence analysis of three of the mutations indicated that the mutant phenotypes were due to specific nucleotide changes that resulted in the expression of either truncated proteins (dgs^R60 and dgs^R19) or proteins in which a conserved amino acid had been altered (dgs^B19; Fig 5). It is reasonable to assume that dgs^R60 and dgs^R19 represent null alleles, since each would generate Gs proteins missing the conserved G4 and G5 domains required for recognition of the guanine ring of bound nucleotides, as well as conserved C-terminal domains (SPRANG 1997 ). These truncations produced no discernable dominant phenotype and, therefore, the resultant mutant proteins are unlikely to possess a novel activity that might generate neomorphic phenotypes. In addition, immunocytochemical analysis suggests that dgs^R79 also represents a null allele since no Gs protein can be detected in embryos generated from germline clones and this allele generates phenotypes similar to that observed for dgs^R60 and dgs^R19. The dgs^B19 mutation can be considered a hypomorphic allele, since homozygous, trans-heterozygous, and hemizygous mutants could survive up to the pharate adult stage; the null mutations described above die late in embryogenesis or shortly after hatching. In addition, the amino acid change generated by this mutation, I373F, alters a C-terminal residue conserved in all Gs proteins. Indeed, the C-terminal 41 amino acids in Gs isoforms identified in Caenorhabditis, Drosophila, and mammals are completely conserved (HARRIS et al. 1985 ; QUAN et al. 1989 ; JANSEN et al. 1999 ). Numerous studies have identified this domain as essential not only for interaction with receptors but also for determining G-protein receptor specificity (e.g., RASENICK et al. 1994 ; CONKLIN et al. 1996 ; AKTER et al. 1998 ; GILCHRIST et al. 1998 , GILCHRIST et al. 1999 ). Thus, it is reasonable to speculate that this substitution results in a reduction in receptor/Gs interaction, leading to the hypomorphic phenotype. Consistent with this interpretation, maternally mutant dgs^B19 embryos show levels and patterns of Gs staining similar to heterozygous siblings, suggesting that the dgs^B19 phenotype is generated by expression of a protein with reduced function. Careful biochemical analysis will be required to confirm this interpretation.

View larger version (58K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Crystal structure of Mg·GTP (indicated) Gs showing the location of mutations present in dgsR60 (Y241Stop), dgsR19 (W265Stop), and dgsB19 (I373F). Based on SUNAHARA et al. 1997 .

Loss-of-function mutations in Gs have been described in Caenorhabditis (KORSWAGEN et al. 1997 ), mice (YU et al. 1998 ), and humans (FARFEL et al. 1999 ). In mice, homozygous mutations are associated with early lethality; embryos undergo implantation but stop developing normally before day E10.5. In addition, mice heterozygous for Gs mutations show a variety of distinct phenotypes depending on whether the chromosome carrying the mutations is inherited maternally or paternally, reflecting the fact that the Gs gene in mammals is imprinted in a tissue-specific fashion (YU et al. 1998 ; WROE et al. 2000 ). Consistent with this observation, humans carrying heterozygous mutations in the Gs gene also show a variety of distinct, hormone-unresponsive phenotypes reflecting imprinting (FARFEL et al. 1999 ). Phenotypes associated with loss-of-function mutations in Drosophila are most similar to those obtained in Caenorhabditis (KORSWAGEN et al. 1997 ). In both species, animals homozygous for either zygotic or maternal mutations in genes encoding Gs are morphologically normal at all levels, although development becomes arrested at the first stage of larval development. Thus, the Gs protein and the pathway it defines do not play an important role in patterning of the Drosophila embryo or in the elaboration of any specific embryonic tissue. Consequently, the larval lethal phenotype in both organisms is probably the result of an essential, vital function of Gs during the initial stages of larval development rather than a requirement for this protein, or the signaling pathway it modulates, in any particular step during embryogenesis. Consistent with this interpretation, a few individuals homozygous for the dgs hypomorphic allele, dgs^B19, survived until pharate adult stage but grew more slowly and were thinner and more transparent than heterozygous siblings due to reduced amounts of body fat. Finally, loss of maternal dgs resulted in a modest increase in embryonic lethality but had no apparent impact on female fecundity, indicating that maternal dgs contributes in a limited manner to embryonic survival but is not required for oogensis in the germline. However, differences between the phenotypes generated by manipulation of Gs activity in each organism do exist. For example, expression of constitutively active Gs protein in Caenorhabditis results in massive neurodegeneration (KORSWAGEN et al. 1997 , KORSWAGEN et al. 1998 ; BERGER et al. 1998 ) while expression of activated forms of Gs in the nervous system of mosaic humans with McCune-Albrights syndrome is apparently without consequence (FARFEL et al. 1999 ). In Drosophila, restricted expression of constitutively active forms of Gs does not result in neurodegeneration or alter developmental processes underlying the Drosophila nervous system formation (W. J. WOLFGANG and M. FORTE, unpublished results) but does result in profound learning and memory deficits in adult flies (CONNOLLY et al. 1996 ). These results suggest that in Drosophila and humans compensatory pathways may be present that can balance some of the consequences of constitutive activation of Gs signaling pathways. Compensatory responses in mice have been demonstrated in response to expression-activated forms of Gs in specific cell types (MA et al. 1994 ). Alternatively, neurons in Caenorhabditis may be more sensitive to the expression of these mutant proteins.

Larvae homozygous or hemizygous for the dgs hypomorphic allele, dgs^B19, were sluggish (Fig 4), showed uncoordinated movements, and were not attracted to yeast granules, indicating deficits in sensory-motor processes. Similarly, Caenorhabditis mutants lacking Gs showed little pharyngeal and body-wall muscle activity (KORSWAGEN et al. 1997 ). These observations suggest that Gs mutants have defects in neural and/or muscle physiology. A large number of studies have demonstrated that mutations that alter cAMP homeostasis in Drosophila (e.g., rutabaga and dunce) and pharmacological modulation of cAMP levels lead to abnormalities in channel function and nerve excitability (ZHONG and WU 1993 ; ZHAO and WU 1997 ; DELGADO et al. 1998 ), synaptic transmission and plasticity (ZHONG and WU 1991 ; ENGEL and WU 1996 ; YOSHIHARA et al. 1999 ; KUROMI and KIDOKORO 2000 ; RENGER et al. 2000 ; YAO et al. 2000 ), growth cone motility (KIM and WU 1996 ), and nerve arborization (ZHONG et al. 1991 ). Further, cAMP has been shown to modulate synaptic transmission in many other species (e.g., DIXON and ATWOOD 1989 ; KANDEL and ABEL 1995 ; SALIN et al. 1996 ; CHEN and REGEHR 1997 ). Immunocytochemical analysis has demonstrated that Gs is concentrated in the forming nervous system of Drosophila embryos (WOLFGANG et al. 1991 ) and in differentiating synapses of the Drosophila neuromuscular junction (W. J. WOLFGANG and M. FORTE, unpublished results), suggesting that Gs-dependent modulation of cAMP levels can play an important role during these developmental events. These results, coupled with the fact that the alteration of several components of the cAMP pathway [e.g., dunce, rutabaga, expression of activated Gs (CONNOLLY et al. 1996 )] affect more complex neuronal processes such as learning and memory, support the idea that the cAMP-dependent signaling plays a critical role in a wide variety of neuronal processes throughout development. In future studies, careful analysis of neuronal development, synapse formation, and synapse function following manipulation of Gs activity will help define the precise role of this protein and pathway in these developmental events.

The phenotypes generated by mutations in the dgs gene in Drosophila stand in direct contrast to those generated by mutations in the DC0 gene, encoding a presumed downstream effector of this pathway, PKA. For example, females carrying germline clones for DC0 mutations fail to lay eggs (LANE and KALDERON 1993 , LANE and KALDERON 1995 ). In various heteroallelic combinations, PKA mutant females show defects in oogenesis due to the disruption of microtubule distribution and the localization of RNAs encoding key determinants (e.g., bicoid and oskar) along the anteroposterior axis (LANE and KALDERON 1994 ). In contrast, adult females whose germlines carry dgs null alleles (e.g., dgs^R60 and dgs^R19) lay morphologically normal eggs that develop to late embryonic stages with 50% hatching. Furthermore, embryos deficient in PKA also show a variety of morphological defects including preblastoderm arrest and alterations in cuticular patterning (LANE and KALDERON 1993 ; OHLMEYER and KALDERON 1997 ). Patterning defects arise due to the role of PKA as a modulator of the conversion of cubitus interruptus (ci), the transcription factor responsible for transducing signals mediated by the morphogen hedgehog, from a transcriptional activator to a transcriptional repressor; in PKA-deficient embryos, ci is not processed to the repressor form, resulting in phenotypes resembling ectopic hedgehog expression (OHLMEYER and KALDERON 1998 ; PRICE and KALDERON 1999 ; CHEN et al. 1999 ). In contrast, embryos maternally or zygotically deficient for Gs do not show patterning defects consistent with alterations in the hedgehog signaling pathway. In addition, direct measurement of cAMP levels in larvae homozygous for dgs null mutations (dgs^R60 and dgs^R79) has shown that signaling through Gs plays a major role in establishing basal levels of this second messenger. Since dgs mutations do not generate the embryonic patterning defects observed in DC0 mutants, basal PKA activity is likely not to depend on pathways activated by Gs that contribute to basal levels of cAMP. Interestingly, alterations observed on a physiological level following mutational and pharmacological manipulation of cAMP levels cannot be mimicked, or are not observed to the same extent, following partial, mutational inactivation of PKA (RENGER et al. 2000 ). These observations suggest that PKA activation by cAMP in Drosophila, at least in these developmental contexts, does not proceed by receptor-mediated modulation of the activity of adenylyl cyclase and increased intracellular cAMP. This interpretation is also consistent with two other observations. First, phenotypes generated by expression of constitutively active forms of Gs could not be suppressed by genetic and biochemical elimination of PKA activity (WOLFGANG et al. 1996 ). Second, expression of a cAMP-independent form of PKA in both embryos and imaginal discs is able to rescue phenotypes generated by elimination of PKA activity (LI et al. 1995 ; JIANG and STRUHL 1995 ; OHLMEYER and KALDERON 1997 ). These results are consistent with the conclusion that the activity of PKA, again at least in these developmental contexts, does not depend on receptor-mediated increases in intracellular cAMP, in contrast to the predictions of models developed primarily on the basis of studies in cultured cells. Since PKA is activated by cAMP, these observations leave open the question of how activation of PKA is mediated or of the role of cAMP in these contexts. However, these observations do not address the role of PKA in the generation of phenotypes present in dgs mutants.

The results presented here further support the notion that Gs-mediated activation of adenylyl cyclase in the production of cAMP and the activation of PKA are not necessarily coupled in a linear or dependent fashion. Three general alternatives can be invoked as the underlying basis for the differential response to the elimination of Gs vs. PKA. First, these genetic studies may point to the existence of a novel cAMP-independent signal transduction pathway activated directly by Gs. Indeed, activation of cAMP-independent pathways by Gs has been proposed in a variety of mammalian cell systems (YATANI et al. 1988 ; MATTERA et al. 1989 ; SHUBA et al. 1990 , SHUBA et al. 1991 ; BOMSEL and MOSTOV 1993 ; PIMPLIKAR and SIMONS 1993 ; HANSEN and CASANOVA 1994 ; WANG and MALBON 1996 ; MA et al. 2000 ). Alternatively, since expression of a constitutively activated form of Gs in cultured mammalian cells (QUAN et al. 1991 ) and fly eyes (CHYB et al. 1999 ) results in increased intracellular cAMP, it may be that the primary mediator of the effects of cAMP in these cellular contexts is a molecule other than PKA, such as cyclic nucleotide-gated channels. In addition, since ß-subunits are potent modulators of a number of biochemical processes, it is possible that free ß, generated as a consequence of the absence of Gs, may in fact be responsible for some phenotypes associated with dgs mutations. Clearly, a goal of future studies will be to differentiate between these formal alternatives.

	FOOTNOTES

¹ Present address: School of Animal and Microbial Sciences, University of Reading, Whiteknights, P.O. Box 228, Reading RG6 6AJ, England.

	ACKNOWLEDGMENTS

The authors thank Jacqueline Parker for excellent technical assistance, Christine Fenner for outstanding secretarial support, and Philip Copenhaver and Sarah Smolik for comments on the manuscript. This work was supported by grants from the National Institutes of Health awarded to M.F.

Manuscript received November 22, 2000; Accepted for publication May 1, 2001.

	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].

AKTER, S. A., L. M. LUTTRELL, H. A. ROCKMAN, G. IACCARINO, and R. J. LEFKOWITZ et al., 1998  Targeting the receptor-Gq interface to inhibit in vivo pressure overload myocardial hypertrophy. Science 280:574-577[Abstract/Free Full Text].

BERGER, A. J., A. C. HART, and J. M. KAPLAN, 1998  Gs-induced neurodegeneration in Caenorhabditis elegans. J. Neurosci. 18:2871-2880[Abstract/Free Full Text].

BIEL, M., X. ZONG, and F. HOFMANN, 1999  Cyclic nucleotide gated channels. Adv. Second Messenger Phosphoprotein Res. 33:231-250[Medline].

BOMSEL, M. and K. E. MOSTOV, 1993  Possible role of both the and ß subunits of the heterotrimeric G protein, Gs, in transcytosis of the polymeric immunoglobulin receptor. J. Biol. Chem. 268:25824-25835[Abstract/Free Full Text].

BOURNE, H. R. and R. NICOLL, 1993  Molecular machines integrate coincident synaptic signals. Cell 72(Suppl):65-75.

BRODY, T. and A. CRAVCHIK, 2000  Drosophila melanogaster G protein-coupled receptors. J. Cell Biol. 150:83-88.

BROILLET, M. C. and S. FIRESTEIN, 1999  Cyclic nucleotide-gated channels. Molecular mechanisms of activation. Ann. NY Acad. Sci. 868:730-740[Abstract/Free Full Text].

CANN, M. J. and L. R. LEVIN, 1998  Genetic characterization of adenylyl cyclase function. Adv. Second Messenger Phosphoprotein Res. 32:121-135[Medline].

CHEN, C. and W. G. REGEHR, 1997  The mechanism of cAMP-mediated enhancement at a cerebellar synapse. J. Neurosci. 17:8687-8694[Abstract/Free Full Text].

CHEN, Y., J. R. CARDINAUX, R. H. GOODMAN, and S. M. SMOLIK, 1999  Mutants of cubitus interruptus that are independent of PKA regulation are independent of hedgehog signaling. Development 126:3607-3616[Abstract].

CHOU, T. B. and N. PERRIMON, 1992  Use of a yeast site-specific recombinase to produce female germline chimeras in Drosophila. Genetics 131:643-653[Abstract].

CHOU, T. B. and N. PERRIMON, 1996  The autosomal FLP-DFS technique for generating germline mosaics in Drosophila melanogaster. Genetics 144:1673-1679[Abstract].

CHYB, S., W. HEVERS, M. FORTE, W. J. WOLFGANG, and Z. SELINGER et al., 1999  Modulation of the light response by cAMP in Drosophila photoreceptors. J. Neurosci. 19:8799-8807[Abstract/Free Full Text].

COFFINO, P., H. R. BOURNE, U. FRIEDRICH, J. HOCHMAN, and P. A. INSEL et al., 1976  Molecular mechanisms of cyclic AMP action: a genetic approach. Recent Prog. Horm. Res. 32:669-684.

COLLEDGE, M. and J. D. SCOTT, 1999  AKAPs: from structure to function. Trends Cell Biol. 9:216-221[Medline].

CONKLIN, B. R., P. HERZMARK, S. ISHIDA, T. A. VOYNO-YASENETSKAYA, and Y. SUN et al., 1996  Carboxy-terminal mutations of Gq alpha and Gs alpha that alter the fidelity of receptor activation. Mol. Pharmacol. 50:885-890[Abstract].

CONNOLLY, J. B., I. J. ROBERTS, J. D. ARMSTRONG, K. KAISER, and M. FORTE et al., 1996  Associative learning disrupted by impaired Gs signaling in Drosophila mushroom bodies. Science 274:2104-2107[Abstract/Free Full Text].

COOPER, D. M., N. MONS, and J. W. KARPEN, 1995  Adenylyl cyclases and the interaction between calcium and cAMP signalling. Nature 374:421-424[Medline].

DAVIS, R. L., 1996  Physiology and biochemistry of Drosophila learning mutants. Physiol. Rev. 76:299-317[Abstract/Free Full Text].

DAVIS, R. L. and B. DAUWALDER, 1991  The Drosophila dunce locus: learning and memory genes in the fly. Trends Genet. 7:224-229[Medline].

D