Transgenic Inhibitors Identify Two Roles for Protein Kinase A in Drosophila Development

Corresponding author: John A. Kiger, Jr., Molecular and Cellular Biology, University of California, 1 Shields Ave., Davis, CA 95616., jakiger{at}ucdavis.edu (E-mail)

Communicating editor: R. S. HAWLEY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have initiated an analysis of protein kinase A (PKA) in Drosophila using transgenic techniques to modulate PKA activity in specific tissues during development. We have constructed GAL4/UAS-regulated transgenes in active and mutant forms that encode PKAc, the catalytic subunit of PKA, and PKI(1-31), a competitive inhibitor of PKAc. We present evidence that the wild-type transgenes are active and summarize the phenotypes produced by a number of GAL4 enhancer-detector strains. We compare the effects of transgenes encoding PKI(1-31) with those encoding PKAr^*, a mutant regulatory subunit that constitutively inhibits PKAc because of its inability to bind cyclic AMP. Both inhibitors block larval growth, but only PKAr^* alters pattern formation by activating the Hedgehog signaling pathway. Therefore, transgenic PKI(1-31) should provide a tool to investigate the role of PKAc in larval growth regulation without concomitant changes in pattern formation. The different effects of PKI(1-31) and PKAr^* suggest two distinct roles, cytoplasmic and nuclear, for PKAc in Hedgehog signal transduction. Alternatively, PKAr^* may target proteins other than PKAc, suggesting a role for free PKAr in signal transduction, a role inhibited by PKAc in reversal of the classical relationship of these subunits.

CYCLIC AMP and its target protein kinase A (PKA) are central elements of a ubiquitous signaling pathway important in the cell cycle, cellular communication, memory formation, and behavior. In Drosophila, genetic manipulation of cAMP levels is possible through mutations of the dunce gene, which encodes a cAMP-specific phosphodiesterase, and mutations of the rutabaga gene, which encodes a calcium-dependent form of adenylyl cyclase. Zygotes homozygous for dunce null mutations are retarded in growth but develop into morphologically normal flies. These adult flies contain up to six times the normal levels of cAMP (DAVIS and KIGER 1981 ), and they exhibit defects in the nervous system (behaviors and memory formation) and in oogenesis (BYERS et al. 1981 ; BELLEN and KIGER 1987 ; LEVINE et al. 1994 ). Mutant rutabaga flies appear to be completely normal except for defective memory formation. Measurements of cAMP in early embryos produced by dunce and rutabaga double-mutant females demonstrate that dunce plays a major role (modulated by rutabaga) in maternal regulation of embryonic cAMP content (WHITEHOUSE-HILLS et al. 1992 ), which in turn can produce a wide spectrum of developmental defects (BELLEN et al. 1987 ; BELLEN and KIGER 1988 ).

PKA mediates most of the known effects of cAMP in a wide range of eukaryotic species and is thought to be a major target of cAMP in Drosophila. PKA consists of a cAMP-binding regulatory moiety (PKAr) and a catalytic moiety (PKAc). It is generally described as a heterotetrameric complex, R₂C₂, consisting of a dimer of two identical regulatory subunits (R₂ = PKAr), with each subunit bound to a monomeric catalytic subunit (C = PKAc). Upon binding cAMP, the tetramer dissociates to R₂ + 2C, freeing the active site of the catalytic subunit from inhibition (TAYLOR et al. 1990 ). In Drosophila, PKAc is encoded by the DC0 gene. Hypomorphic mutants of DC0 show effects suggesting that PKAc is required at a number of stages for normal growth and development (LANE and KALDERON 1993 ), learning (SKOULAKIS et al. 1993 ), and behavior (LEVINE et al. 1994 ).

A role for PKAc in Hedgehog signaling during development has been inferred from experiments designed to reduce the level, or inhibit the activity, of PKAc. Reduction in PKAc level has been achieved in Drosophila by producing mitotic clones of cells homozygous for lethal alleles of DC0 (JIANG and STRUHL 1995 ; LEPAGE et al. 1995 ; LI et al. 1995 ; PAN and RUBIN 1995 ; STRUTT et al. 1995 ) or by use of homozygous-viable recessive mutations of DC0 (LEPAGE et al. 1995 ). Inhibition of PKAc activity has been achieved by ectopic expression, under GAL4/UAS control, of a mutant Drosophila PKA regulatory subunit type I, PKAr^*, which is defective in its ability to bind cAMP (LI et al. 1995 ). Both techniques elicit abnormal development that mimics that caused by stimulation of the Hedgehog signaling pathway, i.e., induction of decapentaplegic (dpp), wingless (wg), and patched (ptc) expression. In the wing imaginal disc, for example, stimulation of the Hedgehog pathway in the wing margin of the anterior compartment leads to duplications of anterior wing patterns. In interpreting these experiments, it is assumed that free PKAc activity represses the Hedgehog signaling pathway, and that reduction (by DC0 mutation) or inhibition (by PKAr^*) of free PKAc activity stimulates the pathway. How an extracellular Hedgehog signal might normally affect PKAc activity is not clear. Studies carried out in Drosophila cell culture suggest that PKAc repression of Hedgehog signal transduction may occur by direct phosphorylation of the cytoplasmic form of transcription factor Ci, encoded by cubitus interruptus, leading to its conversion by proteolysis to a repressor of Hedgehog target genes (AZA-BLANC et al. 1997 ; CHEN et al. 1998 ). Conversely, inhibition of proteolysis leads to an accumulation of full-size Ci, its translocation to the nucleus, and activation of Hedgehog target genes.

Reciprocally, the effects of ectopic PKAc expression have been studied using a transgene encoding a mutant mouse PKAc (PKAc^*) that is defective in its ability to bind PKAr. This transgene shows no effect on patterning of the wing imaginal disc, but it produces blistered wings in adult flies (JIANG and STRUHL 1995 ; LI et al. 1995 ). In embryos, on the other hand, PKAc^* is reported to induce expression of wg and ptc, genes that are also activated in embryos by PKAr^* (OHLMEYER and KALDERON 1997 ).

To dissect the various roles of the cAMP signaling pathway in Drosophila, we have sought to control PKAc activity using the GAL4/UAS transgenic system developed by BRAND and PERRIMON 1993 . We have created UAS transgenic strains to express synthetic genes encoding the N-terminal domain (residues 1–31) of the rabbit skeletal muscle PKAc-inhibitor protein, PKI, in active [Arg^19,20] and inactive [Gly^19,20] forms (TAYLOR et al. 1990 ). These two Arg residues play an essential role in forming a highly specific pseudosubstrate binding site recognized by PKAc. The N-terminal domain of PKI is a highly specific competitive inhibitor of mammalian PKAc, and it has been shown to inhibit Drosophila PKAc in vitro (DRAIN et al. 1991 ). We have also created UAS transgenic strains to express wild-type Drosophila PKAc and a mutant PKAc in which Arg replaces Trp at position 224. We use these strains to show that transgenic PKI(1-31) expression inhibits both transgenic and endogenous PKAc activities, and we compare its effects with those of PKAr^*. We find that PKI(1-31) and PKAr^* have similar effects on larval growth but differ in their ability to induce ectopic Hedgehog signaling in embryos and imaginal discs, thus identifying distinct roles for PKA in growth and pattern formation.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

PKAc transgene construction:
A DC0 cDNA clone modified to contain EcoRI and KpnI sites immediately upstream of the PKAc initiation codon (LANE and KALDERON 1993 ) was provided by Dr. Daniel Kalderon. The entire coding sequence was excised by first cutting with XbaI (to preserve the overlapping EcoRI site at the termination codon) and then cutting with EcoRI. This fragment was cloned into pUAST (BRAND and PERRIMON 1993 ) cut with EcoRI and XbaI to form plasmid pUAST-DC08 (Figure 1). A gene encoding a FLAG epitope appended to the C terminus of PKAc was constructed from pUAST-DC08 by PCR using an upstream primer overlapping the BglII site [CCGATCCAGATCTATGAG] and a downstream primer that removes the overlapping EcoRI and XbaI sites and adds nucleotides encoding the epitope (underlined), followed by two termination codons and an XbaI site [GCATTCTAGACTATTACTTATCGTCATCGTCCTTGTAATCAAACTCAGCAAACTCCTTGG]. The PCR product was cut with BglII and XbaI and inserted into pUAST-DC08 cut with these enzymes to create pUAST-DC0F1. The integrity of this substitution was verified by sequencing from the BglII site to the new XbaI site. Fly strains carrying the two transgenes were created by embryo injection of pUAST-DC08 and pUAST-DC0F1 plasmids with a helper plasmid, phs2-3 (MISRA and RIO 1990 ).

View larger version (4K): In this window In a new window Download PPT slide   Figure 1. Structure of the DC0 cDNA in plasmid pUAST-DC08. Restriction sites are shown above the line with numbering beginning at the first base of the 5' EcoRI site. Sites of coding interest are shown below the line: the AUG initiation codon, the UAG termination codon, the R224 mutant codon, and its wild-type equivalent.

Automated sequencing of the entire DC0 cDNA in pUAST-DC08 revealed a T-to-C substitution at nucleotide 2793 (KALDERON and RUBIN 1988 ), which produces a Trp-to-Arg substitution at amino acid 224 of PKAc. This position corresponds to amino acid 221 of mammalian PKAc, a highly conserved site. The adjacent Asp²²⁰ is nearly invariant in protein kinases, and it contributes directly to stabilization of the catalytic loop and the PKI peptide binding site (KNIGHTON et al. 1991 ). The wild-type DC0 sequences published by FOSTER et al. 1988 and KALDERON and RUBIN 1988 are in agreement on T at this position, suggesting that the cDNA mutation is a product of reverse transcription.

Vectors with the wild-type DC0 sequence were produced using a genomic XhoI-XbaI fragment of DC0 cloned in plasmid p-XX, which was provided by Dr. F. Rob Jackson. The relevant region of p-XX was sequenced to verify that it had T at position 2793, and the SalI-BglII fragment containing the wild-type sequence was excised. Because the pUAST vector contains SalI sites, it was necessary to subclone the KpnI-XbaI fragment of pUAST-DC08 into Bluescript to give pBS-DC08. The mutant SalI-BglII fragment of pBS-DC08 was excised and replaced with the wild-type SalI-BglII fragment of p-XX to give pBS-DC0A. The XhoI-BglII fragment of pBS-DC0A was then used to replace the mutant XhoI-BglII fragments of pUAST-DC08 and pUAST-DC0F1, giving vectors pUAST-DC0A and pUAST-DC0FA, which were used to produce transgenic strains as described above. The strains used here are designated as UAS-PKAc, UAS-PKAcF, and UAS-PKAcR²²⁴F.

PKIF transgene construction:
Plasmids containing synthetic genes encoding the N-terminal domain of rabbit skeletal muscle protein kinase A inhibitor protein in active and inactive forms, PKI(1-31) and mutant [Gly^18,19]PKI(1-31), were provided by Dr. Joseph Avruch (GROVE et al. 1987 ). Genes encoding PKI(1-31) with an appended C-terminal FLAG epitope were produced by PCR with an upstream primer containing a BamHI site [TGCAGGGATCCCACCATG] and a downstream primer that adds nucleotides encoding the epitope (underlined), followed by two termination codons and an XbaI site [CGTATCTAGACTATTACTTATCGTCATCGTCCTTGTAATCAGAGGCAGAGGAGACC]. The PCR products were cloned into Bluescript to give plasmids pBS-PKIF and pBS-PKIG^19,20F, whose structures were verified by sequencing. BamHI-XbaI fragments from each of these were subcloned into pUAST cut with BglII and XbaI to give vectors pUAST-PKIF and pUAST-PKIG^19,20F, which were used to produce transgenic strains designated UAS-PKIF and UAS-PKIG^19,20F, as described above.

Fly strains and crosses:
The enhancer detector vector pGawB (BRAND and PERRIMON 1993 ), carried in the strain w GAL4 106-1, was jumped by crossing females to males of strain w ; Dr/TMS, Sb 2-3 (LINDSLEY and ZIMM 1992 ). Independent jumps to autosomes of the same strain used for embryo injections were mapped by standard procedures to chromosomes and established in balanced stocks (chromosomes II and III only) by students in the undergraduate genetics laboratory course at UC Davis as part of their instruction in Mendelian genetics. GAL4 stocks are designated with each student's initials and strain number, e.g., GAL4-RK5.

Strains carrying UAS transgenes were created from injected embryos of a y w strain, mapped to a chromosome, and established in balanced stocks by standard procedures, taking care that all insertions recovered on a particular chromosome are independent.

UAS-PKAr^* strains and GAL4-E22C were generously supplied by Dan Kalderon (LI et al. 1995 ; OHLMEYER and KALDERON 1997 ). The UAS-GFP strain is described by YEH et al. 1995 . All other mutations are described in LINDSLEY and ZIMM 1992 . All crosses were carried out at 25°.

Immunostaining and microscopy:
Heads of wandering third instar larvae were everted in modified Robb's medium and fixed for 1 hr at room temperature in 0.05 M sodium phosphate, pH 7.6, 4% formaldehyde, and 1% NP-40, followed by dehydration in absolute methanol and storage at -20°. They were rehydrated by washing three times in PT (PATEL 1994 ) and blocked for 1 hr at room temperature in 0.1 M sodium phosphate, pH 7.6, 0.1% Tween 20, 2% BSA, and 2% serum. Anti-FLAG M2 monoclonal antibody (Eastman Kodak, Rochester, NY) at a final dilution of 1:500 and anti-DC0 rabbit polyclonal antibody (LANE and KALDERON 1993 ) at a final dilution of 1:400 were incubated with larval heads in fresh block for 1–2 hr at room temperature followed by storage at 4° overnight. Antibodies were removed, and heads were washed repeatedly in PBT (PATEL 1994 ) and blocked for 30 min at room temperature. Secondary antibodies labeled with FITC or Texas red (Vector Laboratories, Burlingame, CA) were added directly to the heads at a final dilution of 1:100 and incubated at room temperature for 3–4 hr. Antibodies were removed and heads were washed repeatedly in PBT, rinsed once in PBS, and equilibrated in 70% glycerol, 2.5% DABCO (PATEL 1994 ). Wing discs were dissected and mounted in the same solution and viewed with a Leica TCS-NT confocal microscope. Fluorescence was analyzed quantitatively using the Leica instrumentation and Scion Image software.

Wings were dissected from bodies in 95% ethanol, dehydrated in 100% ethanol, equilibrated in toluene, mounted in cedarwood oil (Sigma, St. Louis), and photographed with a dissecting microscope (Zeiss, Jena, Germany). Eggs were collected on grape juice agar (without added yeast; ASHBURNER 1989 ), and larvae were mounted in Hoyer's medium (ASHBURNER 1989 ), cleared at 60°, and photographed with a Zeiss Axioplan microscope using phase contrast.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Effects of ectopic PKI(1-31):
The activity of UAS-PKIF transgenes has been established by identifying GAL4 strains that produce phenotypic effects in conjunction with the transgene. Approximately 220 new GAL4 insertions on chromosomes II and III were established in balanced stocks and crossed to a strain carrying two transgenes on chromosome I, UAS-PKIF 10-2 and 12-2. Twenty-one GAL4 strains were found to significantly affect the recovery of adult progeny. Table 1 shows the progeny recovered from four such randomly chosen GAL4 strains (heterozygous with balancer chromosomes CyO or TM3) when crossed to four strains with an active PKIF transgene and to two strains with an inactive PKIG^19,20F transgene. Note that progeny carrying a GAL4 chromosome and an inactive transgene are recovered with about the same frequency as sibs carrying a balancer chromosome and an inactive transgene, while progeny carrying a GAL4 chromosome and an active transgene are recovered at much lower frequencies than sibs carrying a balancer chromosome. The high degree of lethality exhibited by the four PKIF transgenes demonstrates that they are expressed and biologically active.

From the results obtained with these and other GAL4 strains, we find that most lethality caused by PKIF occurs at hatching or during the larval stages. PKIF expression driven by GAL4-PP3 causes larvae to die just before or soon after hatching from the egg. Other GAL4 strains permit development to second or third instars before death, most larvae being abnormally small and, near death, misshapen. Some larvae continue to feed for many days after pupation would normally have occurred and never reach normal size. Of the few larvae that eventually pupate, some produce small but morphologically normal adult flies, some as small as half size. These small flies emerge from small pupal cases, indicating that their size is a result of retarded larval development rather than failure to carry out the final mitotic division during imaginal disc development. This growth retardation and failure to pupate resembles that described for hypomorphic DC0 mutants (LANE and KALDERON 1993 ), suggesting that PKIF is inhibiting endogenous PKAc as intended. The particular phenotype produced depends upon the particular GAL4 strain, and must be caused by differences in the tissue specificity, time of expression during development, and strength of expression of either the GAL4 or PKIF transgene.

The GAL4 expression pattern of each of these strains has been examined using a UAS-GFP transgene. Green fluorescent protein (GFP) expression in first instar larvae of GAL4-PP3 and GAL4-JW1 is much more intense than in first instar larvae of GAL4-KO5 and GAL4-RK5. GAL4-PP3 and GAL4-JW1 have in common the expression of GFP in the central and peripheral nervous systems, muscles, tracheae, and proventriculus. GAL4-KO5 and GAL4-RK5 show more restricted expression in first instar larvae, but by third instar, GFP expression in both is evident in the central nervous system, tracheae, proventriculus, and fat body. Further study of particular PKIF transgene expression will be required to establish which tissue is the focus of a particular phenotype.

Effects of ectopic PKAc:
The activities of PKAc transgenes with and without FLAG epitope, as well as the mutant PKAcR²²⁴ with FLAG epitope, have been assayed with the same set of GAL4 strains as shown in Table 2. The UAS-PKAcF and UAS-PKAc transgenes are clearly expressed and biologically active. It would appear that the C-terminal FLAG epitope does not affect PKAc activity. The UAS-PKAcR²²⁴F mutant transgene appears to be completely inactive. The nature and degree of lethality observed in the crosses shown in Table 2 depends both upon the GAL4 strain and the UAS strain used. The particular phenotype must depend upon the factors discussed previously, especially the strength of expression of the UAS transgene. The UAS-PKAcF 1.1 transgene is one of the strongest; combined with each of the four GAL4 strains, all larvae die well before pupation. Slightly weaker transgenes show more heterogeneity in combination with different GAL4 strains. For example, UAS-PKAcF 1.3 is like PKAcF 1.1, except that with GAL4-KO5, many individuals die as blackened pharate adults before eclosion. Two of the weakest transgenes are UAS-PKAcF 5.9 and UAS-PKAc 15.3; combined with GAL4-RK5, some individuals die as blackened pharate adults, but adult progeny also eclose exhibiting a variety of overlapping phenotypes. Some are as small as half size; some have warped or collapsed wings, flecks on the surface of the eye, bristles glued to the thorax, and combinations of these defects; others are normal. Combined with GAL4-KO5, UAS-PKAcF 5.9 produces only dead and blackened pharate adults, while UAS-PKAc 15.3 also produces some eclosed adults with collapsed wings.

Comparison of the effects of ectopic PKI(1-31) and PKAc:
GAL4 strains that produce effects with PKAc transgenes are recovered more frequently than those that produce effects with PKIF transgenes, indicating that the number (or target size) of tissues that can be affected is greater for PKAc than PKI(1-31). A screen of an additional 96 new GAL4 strains identified 37 strains that are lethal in conjunction with PKAcF 5.2 and 5 that produce abnormal adults. As indicated above, only 21 out of 220 new GAL4 strains were lethal with PKIF transgenes. When a group of 28 randomly chosen GAL4 strains were crossed to PKAcF 1.1, 19 were lethal (at larval or pupal stages) or produced reduced numbers of normal and abnormal adult progeny. When these 19 GAL4 strains were crossed to PKIF 5-1, 1 was completely lethal and 7 produced reduced numbers of morphologically normal adult progeny (3 of these produced small but normal adults). No abnormal effects were observed with the remainder of the GAL4 strains.

In several screens of GAL4 strains using PKAcF transgenes, many GAL4 strains have been found that produce larval death, death of pharate adults before eclosion, or living adults exhibiting wing and other defects. In the cases of larval death caused by PKAc, the larvae die without lingering to feed for a long period as do larvae affected by PKI(1-31). Moreover, in contrast to the effects of PKAc, we have yet to find a GAL4 strain that produces morphologically abnormal adult flies as a consequence of PKI(1-31) expression.

Interaction between ectopic PKI(1-31) and PKAc:
Evidence that the lethality produced by ectopic PKI(1-31) is caused by inhibition of endogenous PKAc activity is provided by the failure of PKIG^19,20F to produce any effect (Table 1). Additional evidence for this is provided by coexpression of PKIF and PKAcF transgenes in Table 3. Transgenic PKAcF 5.9 expression rescues some individuals from lethality caused by three different PKIF transgenes expressed by GAL4-RK5. Indeed, PKIF and PKAcF transgenes appear to be mutual suppressors since transgenic PKAcF expression alone (particularly by PKAcF 1.3) can be lethal. In contrast, coexpression of UAS-PKAcR²²⁴F and UAS-PKIF using several different GAL4 strains does not titrate PKIF activity, further substantiating the inactivity of the R²²⁴ mutant protein (data not shown).

Comparison of the effects of ectopic PKI(1-31) and PKAr^*:
As mentioned above, we have yet to find a GAL4 strain that produces morphologically abnormal adult flies as a consequence of PKI(1-31) expression. This striking observation contrasts with effects of UAS-PKAr^* transgenes that mimic ectopic Hedgehog expression (LI et al. 1995 ; WOLFGANG et al. 1996 ). Therefore, we have closely compared the effects of PKAr^* and PKIF transgenes expressed by the same GAL4 strains.

A particularly illuminating strain is GAL4-RZ4. The combination of UAS-PKAr^* BDK 35 and GAL4-RZ4 produces adult flies with 51% having anterior wing duplications (62/122 have at least one wing affected in some degree) characteristic of ectopic Hedgehog signaling. Those flies with normal wings are often reduced in size because of retarded larval growth (Figure 2). The combination of UAS-PKIF 5-1 and GAL4-RZ4, on the other hand, produces adult flies that vary in size as a result of retarded larval growth, but always have normal morphology (Figure 2).

View larger version (49K): In this window In a new window Download PPT slide   Figure 2. Effects on wing morphology and size of PKAr* and PKIF expressed under the control of GAL4-RZ4. (Top) Wings from flies of approximately average size produced by crossing GAL4-RZ4/CyO flies to UAS-PKAr* BDK 35 and UAS-PKIF 5-1 flies. (Bottom) Wings are from late-emerging siblings produced in the same crosses. Left wings are of genotype GAL4-RZ4; UAS-PKAr* BDK 35; right wings are of genotype GAL4-RZ4; UAS-PKIF 5-1.

Figure 3 compares the effects of expressing PKIF 5-1 and PKAr^* BDK 35 (A and C) with the effect of expressing PKAcF 5.9 (B). The anterior margin of the wing is clearly sensitive to ectopic expression of either PKAr^* or PKAcF, producing quite different results, but not to ectopic expression of PKIF. Virtually the same fraction of adult flies that show wing duplications caused by PKAr^* (51%) show a notched anterior wing margin produced by PKAcF (49/94 or 52%). The effect of PKAcF on the wing suggests a reduction of wg expression in the cells forming the wing margin (COUSO et al. 1994 ). Coexpression of PKAcF with either PKIF or PKAr^* (Figure 3D and Figure E) produces normal development, proving that both PKIF and PKAr^* are inhibitors of ectopic PKAcF activity. Furthermore, this demonstrates that PKIF is indeed active in cells producing the anterior wing margin. Nevertheless, expression of PKIF alone has no effect on wing development (Figure 3A), while PKAr^* alone elicits Hedgehog signaling and wing duplication (Figure 3C). Moreover, coexpression of PKAcF and PKAr^* blocks ectopic Hedgehog signaling produced by PKAr^* alone (Figure 3E), as would be expected if both are expressed at comparable levels. We have been unable similarly to titrate PKAr^* activity by coexpression of PKAcR²²⁴F, suggesting that the R²²⁴ mutant protein binds neither PKAr nor PKIF (see above and data not shown).

View larger version (49K): In this window In a new window Download PPT slide   Figure 3. Effects on wing morphology of expressing PKIF, PKAcF, and PKAr*, alone or in combination, under the control of GAL4-RZ4. The UAS transgenes used are PKIF 5-1, PKAr* BDK 35, and PKAcF 5.9. See text for discussion.

We have also compared the effects of PKIF and PKAr^* using GAL4-1J3. With this strain, PKAr^* is severely detrimental; in our hands, most adults fail to emerge from the pupal cases and exhibit severely truncated legs with multiple ectopic bristle columns, as described by LI et al. 1995 . On the other hand, PKIF causes no effect on viability with this GAL4 strain, and adults are normal (data not shown).

To understand the actions of PKIF and PKAr^*, it is important to assess their levels relative to that of endogenous PKAc. We have done this indirectly by quantitating the level of ectopic PKAcF expression that is suppressed by these inhibitors. Ectopic PKAcF can be detected by an antibody against the FLAG epitope, and both endogenous PKAc and ectopic PKAcF can be detected using an antibody against PKAc (LANE and KALDERON 1993 ). Figure 4 shows a confocal section of a third instar wing disc stained with both antibodies. Regions of ectopic PKAcF expression can be identified using the anti-FLAG antibody (green). In Figure 4A, the extent of the tissue is shown by low, nonspecific, background fluorescence, while the region of highest ectopic PKAcF expression (yellow in the merged image in Figure 4B) is overexposed. Total PKAc level throughout the tissue can be assessed using the anti-PKAc antibody (red). In Figure 4C, the fluorescence level has been adjusted so that it is within the linear range of the pixels to obtain a quantitative image. Measures of pixel values show the difference between the two regions marked with arrows to be approximately threefold (196/64, full scale = 256). Because it is not evident at what low level of fluorescence staining becomes nonspecific, threefold is a minimum estimate of the ratio of ectopic PKAcF to endogenous PKAc expression. Both PKIF and PKAr^* inhibit the phenotypic effects of this ectopic PKAcF, indicating that these inhibitory transgenes must be expressed at levels comparable to that of PKAcF 5.9 (assuming that inhibition involves stoichiometric association of either inhibitory protein with PKAcF). Yet, expressed alone in the presence of significantly lower levels of endogenous PKAc, these two inhibitors do not produce equivalent phenotypic effects. PKIF has no effect, whereas PKAr^* activates Hedgehog signaling.

View larger version (31K): In this window In a new window Download PPT slide   Figure 4. Confocal section of a wing imaginal disc from a larva of genotype GAL4-RZ4; UAS-PKAcF 5.9. (A) Ectopic PKAcF is detected in green using an anti-FLAG antibody. (B) Area of highest ectopic expression appears yellow in the merged image. (C) Both ectopic PKAcF and endogenous PKAc are detected in red using an antibody to the DC0 gene product. The bar in A is 50 µm. See text for discussion.

The results of crossing UAS-PKAr^* strains with the GAL4 strains used to characterize UAS-PKIF strains are shown in Table 1. In general, death occurs earlier for larvae expressing PKAr^* than for those expressing PKIF. With GAL4-PP3 and GAL4-JW1, a significant portion of embryos die and turn brown before hatching, while the rest die as young first instars. With GAL4-RK5, small larvae feed and linger for up to 3 wk before dying. With GAL4-KO5 and GAL4-RK5, the adults that emerge are all abnormal: the cuticle of the thorax and abdomen has an abnormal sheen; there is melanization of thorax and wings, as well as warped wings; there are crossed scutellar bristles, an indication of abnormal wing disc development. The surviving adults expressing PKAr^* are not abnormally small like those morphologically normal survivors expressing PKIF.

OHLMEYER and KALDERON 1997 have used GAL4-E22C to express UAS-PKAr^* in the embryonic ectoderm, and they found changes in gene expression and larval morphology consistent with the induction of Hedgehog signaling. We have used this strain to express UAS-PKAr^* BDK 35 and have found that most larvae hatch, and all die, within a few hours. Many of these larvae exhibit an absence of thoracic denticle belts, as well as the change in the pattern of abdominal denticle belts described by OHLMEYER and KALDERON 1997 (Figure 5, left). These pattern alterations are consistent with a mild overexpression of the wg gene (NOORDERMEER et al. 1992 ). In contrast, larvae expressing UAS-PKIF hatch and exhibit normal morphology (Figure 5, right). Depending upon the particular UAS-PKIF strain, the larvae may die soon after hatching, or may linger for many days with little growth before dying (some may molt to second instars), or may grow slowly, with some survivors emerging as normal adults. Coexpression of UAS-PKAr^* BDK 35 and UAS-PKAcF 5.9 restores normal pattern and morphology to abdominal denticle belts and partially restores thoracic denticle belts (Figure 5, middle). Embryos expressing the strongest UAS-PKAc transgenes usually fail to hatch, and the majority develop normal larval denticle belts (data not shown). In contrast, denticle belts produced by UAS-PKAc^* expression are described as similar to those produced by UAS-PKAr^* expression (OHLMEYER and KALDERON 1997 ).

View larger version (87K): In this window In a new window Download PPT slide   Figure 5. Effects of expressing PKAr* and PKIF on larval ventral denticle belts. The normal pattern and morphology is seen at right (GAL4-E22C; UAS-PKIF), where the second (T2) and third (T3) thoracic segment belts and the first (A1) and second (A2) abdominal segment belts are labeled. PKAr* expression removes the thoracic belts and changes the morphology and pattern of the abdominal belts, as seen on the left (GAL4-E22C; UAS-PKAr*). Coexpression of PKAr* and PKAcF suppresses the effect of PKAr* on the pattern and morphology of the abdominal belts and partially restores the thoracic belts, as seen in the middle (GAL4-E22C; UAS-PKAr*, PKAcF). The UAS transgenes used are PKIF 5-1, PKAr* BDK 35, and PKAcF 5.9.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The screens of GAL4 strains presented here suggest that many tissues are sensitive to ectopic PKAc expression, which is not surprising. Significantly, fewer tissues appear to be sensitive to inhibition of endogenous PKAc activity by PKI(1-31), suggesting that a subset of tissues does not require PKAc activity. The only developmental phenotype produced by both PKAc and PKI(1-31) is the small, morphologically normal adult observed in some crosses. This is reminiscent of the identical learning defects exhibited by dunce and rutabaga mutants, where in certain neurons, disturbance, up or down, of ambient cAMP levels produces the same learning defect (FEANY 1990 ). Indeed, the production of small adults appears to mimic the phenotype of NF1 mutants (THE et al. 1997 ). NF1 regulates the rutabaga adenylyl cyclase (GUO et al. 1997 ), suggesting that PKI(1-31) acts downstream of NF1 to inhibit cAMP-activated PKAc kinase activity.

Phenotypes produced by PKI(1-31) and PKAr^* are surprisingly different. The phenotypic effects of PKI(1-31) appear to represent a subset of those of PKAr^*. Both retard or block larval growth. PKAr^* alone affects patterning in embryos and imaginal discs by activating Hedgehog signaling, and it alone causes abnormal differentiation in imaginal discs (which may reflect minor aberrations in patterning). The origin of this difference might reside in some fundamental difference in the biological properties of PKI(1-31) and PKAr^* or perhaps in their relative stabilities in different cell types. However, PKI(1-31) is demonstrably active in wing imaginal discs (Figure 3D) and in other tissues (Table 3) since it is capable of inhibiting ectopic PKAcF. Regardless of the origin of the difference, it would appear that PKI(1-31) specifically targets larval growth.

Newly hatched larvae consist of two cell types: (1) mitotic cells composing the imaginal discs, gonad, and some neuroblasts, and (2) endoreplicating cells making up the exclusively larval tissues. These latter cells do not divide after hatching, but they increase in size as the larva grows, being maintained by cycles of DNA replication without nuclear division. These two cell types are regulated in fundamentally different ways, as demonstrated by their responses to nutritional deprivation (BRITTON and EDGAR 1998 ). It would appear that mitotic cells are not sensitive to expression of PKI(1-31), but only to expression of PKAr*, whereas endoreplicating cells are sensitive to both. It is interesting to note that UAS-PKIF, GAL4-E22C larvae die within a few hours of hatching if kept on grape juice agar lacking amino acids, but they can feed for many days before dying or pupating if kept on regular food. Normal larvae can survive for days without amino acids (BRITTON and EDGAR 1998 ), suggesting that PKI(1-31) sensitizes larvae to starvation.

Both proteins are effective inhibitors of the catalytic site of PKAc, possessing a pseudosubstrate binding site with a pair of adjacent Arg residues that interact with the catalytic site (TAYLOR et al. 1990 ). PKAr, which is larger than PKI(1-31) or full-length PKI(1-77), makes additional contacts with PKAc that make the PKAr:PKAc complex more stable than the PKI(1-77):PKAc complex (HERBERG and TAYLOR 1993 ). PKI(5-24) and PKI(1-77) bind to PKAc with the same affinity (LEW et al. 1997 ), and PKI(1-31) is probably no different. The C terminus of PKI(5-24) is not involved in binding to PKAc, so the FLAG epitope at the C terminus of PKI(1-31) should not interfere with its inhibitory function (KNIGHTON et al. 1991 ). Free PKAc and PKI(1-77) are small enough to enter the nucleus by diffusion (HAROOTUNIAN et al. 1993 ). PKI(1-77) possesses a nuclear export signal (residues 35–49) that is hidden until PKAc is bound, whereupon the PKI(1-77):PKAc complex is extruded from the nucleus (WEN et al. 1995A ). Moreover, the expression and intracellular distribution of PKI(1-77) is regulated during the cell cycle and is necessary for its progression (WEN et al. 1995B ). PKI(1-31) would be expected to inhibit both nuclear and cytoplasmic PKAc. In addition, it could compete with a Drosophila homologue of PKI(1-77) for nuclear PKAc and block its export. PKAr, on the other hand, is cytoplasmic whether or not it is complexed with PKAc because it is too large to enter the nucleus (HAROOTUNIAN et al. 1993 ). In addition, anchoring proteins are known that bind PKAr to the membrane or cytoskeleton (PAWSON and SCOTT 1997 ). PKAr^* would be expected to inhibit cytoplasmic PKAc and to deplete nuclear PKAc by forming a cytoplasmic sink for PKAc that diffuses from the nucleus.

With regard to Hedgehog signaling, a possible target of PKAr^* and PKI(1-31) in the cytoplasm would be the complex responsible for the proteolysis of the transcription factor Ci (AZA-BLANC et al. 1997 ), where they would inhibit phosphorylation of the PKA sites necessary for proteolysis of Ci¹⁵⁵ to the repressor form Ci⁷⁵ (CHEN et al. 1998 ). The ability of PKAr^* to interact with anchoring or other proteins might give it greater access to this complex than PKI(1-31), accounting for the failure of the latter to activate Hedgehog target genes.

Another possible explanation of the different actions of PKAr^* and PKIF(1-31) is that free PKAr^* (and by implication free PKAr) has a target other than PKAc through which it activates Hedgehog signaling. Precedent for such a role exists. In Dictyostelium, free PKAr binds and activates a cAMP-specific phosphodiesterase that is postulated to have functional homology to the cAMP-specific phosphodiesterase encoded by dunce. The Dictyostelium phosphodiesterase is also activated by bovine PKAr¹_a, and a synthetic monomeric form of this regulatory subunit is a more potent activator than the dimeric form (SHAULSKY et al. 1998 ). (The Dictyostelium PKAr protein lacks a dimerization domain, and its PKA exists as a heterodimer.) In this scenario, in the absence of a cAMP signal, PKAc would bind to PKAr, inhibiting this novel activity. Reduction in the level of PKAc, e.g., in a mitotic clone of cells homozygous for a lethal allele of DC0, would lead to free PKAr that would activate Hedgehog signaling.

In an alternative scenario, the effect of PKAr^* on Hedgehog target genes could be caused by its ability to deplete nuclear PKAc, a role that cannot be fulfilled by PKI(1-31). Since the normal role of PKI(1-77) is not only to inhibit, but to export, nuclear PKAc, it is possible that PKAc plays another critical role in the nucleus in addition to its catalytic role in phosphorylation. For example, PKAc might function as a corepressor with Ci⁷⁵ to block transcription of Hedgehog target genes. Consistent with this hypothesis, CHEN et al. 1998 have shown that PKI(1-60) activates Ci-mediated chloramphenicol acetyltransferase transcription from a model Gli enhancer in Drosophila Kc cells, a finding they attribute solely to inhibition of proteolysis of cytoplasmic Ci¹⁵⁵. It may be that PKAc can function as a corepressor even if its catalytic site is occupied by PKI(1-31). Corepression by PKI(1-31):PKAc and Ci⁷⁵ might block transcription of target genes, even in the presence of Ci¹⁵⁵ produced by concommitant inhibition of Ci¹⁵⁵ proteolysis in the cytoplasm. Small changes in the ratio of Ci¹⁵⁵ and Ci⁷⁵ are believed to be critical for activation of Hedgehog target genes (AZA-BLANC et al. 1997 ). In addition, DAY et al. 1989 have pointed out that PKI(1-77) may differ from PKI(1-31) because only the former reduces basal transcription from cAMP-stimulated promoters. If PKAc has such an additional role, then the R²²⁴ mutant must have lost this function, as well as its ability to bind PKAr^* and PKIF, since PKAcR²²⁴F produces no abnormal phenotypes (data not shown) and has no effect on viability (Table 2). On the other hand, the hypothesized nuclear role of PKAc might be catalytic if nuclear PKAc is in some way inaccessible to nuclear PKI(1-31).

These considerations suggest that normal Hedgehog signal transduction may require both inhibition of cytoplasmic PKAc activity and export of nuclear PKAc. A Drosophila homologue of PKI(1-77) would be a good candidate for carrying out these functions. The fact that PKI(1-77) seems to play some role in regulating the cell cycle (WEN et al. 1995B ) may help to explain why PKI(1-31) has different effects on endoreplicating cells and mitotic cells. Resolving the nature of the roles played by PKAc in the cytoplasm and in the nucleus may lead to simultaneous understanding of the effects seen here on pattern formation and on cell growth.

Direct comparisons of the effects of PKI(1-31) and of PKI(1-77) are needed to provide more insight into how different PKAc inhibitors are functioning. PKAc transgenes with specific catalytic site mutations should provide evidence for or against a noncatalytic nuclear role for PKAc. PKAr* transgenes with domain-specific mutations should provide insight into the role of PKAR^* in Hedgehog signaling. Identification of a Drosophila homologue of PKI(1-77) and study of its regulation will be important to achieve a clear understanding of the roles of PKAc. From a practical standpoint, PKI(1-31) transgenes should provide a useful tool for investigating the role of PKA in larval growth regulation, independent of its effects on pattern formation. Mutations that permit larvae to survive the effect of PKI(1-31) and develop to adults should help to identify elements controlling larval growth. Conversely, mutations that sensitize adults or embryos to PKI(1-31) may reveal elements important for pattern formation.

	ACKNOWLEDGMENTS

We are grateful to Dan Kalderon for generously supplying materials and fly strains, and to the University of California–Davis students of MCB 160L for producing many GAL4 strains. J.A.K. thanks Dan Kalderon and Richard Maurer for discussions, and is particularly grateful to Kevin Moffat for much instruction and help at the University of Warwick. The manuscript benefited from the comments of two referees. The initial stages of this work were supported by a Fogarty Senior International Fellowship to J.A.K. from the National Institutes of Health. DNA sequencing in the Genetics Department at Cambridge was supported by the Wellcome Trust. Subsequent work has been supported by funds of the Agricultural Experiment Station at UC Davis.

Manuscript received October 30, 1998; Accepted for publication January 22, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

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

AZA-BLANC, P., F.-A. RAMÍREZ-WEBER, M.-P. LAGET, C. SCHWARTZ, and T. B. KORNBERG, 1997  Proteolysis that is inhibited by Hedgehog targets Cubitus Interruptus protein to the nucleus and converts it to a repressor. Cell 89:1043-1053[Medline].

BELLEN, H. J. and J. A. KIGER, JR., 1987  Sexual hyperactivity and reduced longevity of dunce females of Drosophila melanogaster.. Genetics 115:153-160[Abstract/Free Full Text].

BELLEN, H. J. and J. A. KIGER, JR., 1988  Maternal effects of general and regional specificity on embryos of Drosophila melanogaster caused by dunce and rutabaga mutant combinations. Wilhelm Roux' Arch. Dev. Biol. 197:258-268.

BELLEN, H. J., B. K. GREGORY, C. L. OLSSON, and J. A. KIGER, JR., 1987  Two Drosophila learning mutants, dunce and rutabaga, provide evidence of a maternal role for cAMP on embryogenesis. Dev. Biol. 121:432-444[Medline].

BRAND, A. H. and N. PERRIMON, 1993  Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118:401-415[Abstract].

BRITTON, J. S. and B. A. EDGAR, 1998  Environmental control of the cell cycle in Drosophila: nutrition activates mitotic and endoreplicative cells by distinct mechanisms. Development 125:2149-2158[Abstract].

BYERS, D., R. L. DAVIS, and J. A. KIGER, JR., 1981  Defect in cyclic AMP phosphodiesterase due to the dunce mutation of learning in Drosophila melanogaster.. Nature 289:79-81[Medline].

CHEN, Y., N. GALLAHER, R. H. GOODMAN, and S. M. SMOLIK, 1998  Protein kinase A directly regulates the activity and proteolysis of cubitus interruptus. Proc. Natl. Acad. Sci. USA 95:2349-2354[Abstract/Free Full Text].

COUSO, J. P., S. A. BISHOP, and A. MARTINEZ ARIAS, 1994  The wingless signaling pathway and the patterning of the wing margin in Drosophila.. Development 120:621-636[Abstract].

DAVIS, R. L. and J. A. KIGER, JR., 1981  Dunce mutants of Drosophila melanogaster: mutants defective in cyclic AMP phosphodiesterase enzyme system. J. Cell Biol. 90:101-107[Abstract/Free Full Text].

DAY, R. N., J. A. WALDER, and R. A. MAURER, 1989  A protein kinase inhibitor gene reduces both basal and multihormone-stimulated prolactin gene transcription. J. Biol. Chem. 264:431-436[Abstract/Free Full Text].

DRAIN, P., E. FOLKERS, and W. G. QUINN, 1991  cAMP-dependent protein kinase and the disruption of learning in transgenic flies. Neuron 6:71-82[Medline].

FEANY, M. B., 1990  Rescue of the learning defect in dunce, a Drosophila learning mutant, by an allele of rutabaga, a second learning mutant. Proc. Natl. Acad. Sci. USA 87:2795-2799[Abstract/Free Full Text].

FOSTER, J. L., G. C. HIGGINS, and F. R. JACKSON, 1988  Cloning, sequence and expression of the Drosophila cAMP-dependent protein kinase catalytic subunit gene. J. Biol. Chem. 263:1676-1681[Abstract/Free Full Text].

GROVE, J. R., D. J. PRICE, H. M. GOODMAN, and J. AVRUCH, 1987  Recombinant fragment of protein kinase inhibitor blocks cyclic AMP-dependent gene transcription. Science 238:530-533[Abstract/Free Full Text].

GUO, H.-F., I. THE, F. HANNAN, A. BERNARDS, and Y. ZHONG, 1997  Requirement of Drosophila NF1 for activation of adenylyl cyclase by PACAP38-like neuropeptides. Science 276:795-798[Abstract/Free Full Text].

HAROOTUNIAN, A. T., S. R. ADAMS, W. WEN, J. L. MEINKOTH, and S. S. TAYLOR et al., 1993  Movement of the free catalytic subunit of cAMP-dependent protein kinase into and out of the nucleus can be explained by diffusion. Mol. Biol. Cell 4:993-1002[Abstract].

HERBERG, F. W. and S. S. TAYLOR, 1993  Physiological inhibitors of the catalytic subunit of cAMP-dependent protein kinase: effect of MgATP on protein-protein interactions. Biochemistry 32:14015-14022[Medline].

JIANG, J. and G. STRUHL, 1995  Protein kinase A and Hedgehog signaling in Drosophila limb development. Cell 80:563-572[Medline].

KALDERON, D. and G. M. RUBIN, 1988  Isolation and characterization of Drosophila cAMP-dependent protein kinase genes. Genes Dev. 2:1539-1556[Abstract/Free Full Text].

KNIGHTON, D. R., J. ZHENG, L. F. TEN EYCK, N.-H. XUONG, and S. S. TAYLOR et al., 1991  Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science 253:414-420[Abstract/Free Full Text].

LANE, M. E. and D. KALDERON, 1993  Genetic investigation of cAMP-dependent protein kinase function in Drosophila development. Genes Dev. 7:1229-1243[Abstract/Free Full Text].

LEPAGE, T., S. COHEN, F. J. DIAZ-BENJUMEA, and S. M. PARKHURST, 1995  Signal transduction by cAMP-dependent protein kinase A in Drosophila limb patterning. Nature 373:711-715[Medline].

LEVINE, J. D., C. I. CASEY, D. KALDERON, and F. R. JACKSON, 1994  Altered circadian pacemaker functions and cyclic AMP rhythms in the Drosophila learning mutant dunce.. Neuron 13:967-974[Medline].

LEW, J., N. CORUH, I. TSIGELNY, S. GARROD, and S. S. TAYLOR, 1997  Synergistic binding of nucleotides and inhibitors to cAMP-dependent protein kinase examined by acrylodan fluorescence spectroscopy. J. Biol. Chem. 272:1507-1513[Abstract/Free Full Text].

LI, W., J. T. OHLMEYER, M. E. LANE, and D. KALDERON, 1995  Function of protein kinase A in Hedgehog signal transduction and Drosophila imaginal disc development. Cell 80:553-562[Medline].

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

MISRA, S. and D. C. RIO, 1990  Cytotype control of Drosophila P element transposition: the 66kd protein is a repressor of transposase activity. Cell 62:269-284[Medline].

NOORDERMEER, J., P. JOHNSTON, F. RIJSEWIJK, R. NUSSE, and P. A. LAWRENCE, 1992  The consequences of ubiquitous expression of the wingless gene in the Drosophila embryo. Development 116:711-719[Abstract].

OHLMEYER, J. T. and D. KALDERON, 1997  Dual pathways for induction of wingless expression by protein kinase A and Hedgehog in Drosophila embryos. Genes Dev. 11:2250-2258[Abstract/Free Full Text].

PAN, D. and G. M. RUBIN, 1995  Protein kinase A and hedgehog act antagonistically in regulating decapentaplegic transcription in Drosophila imaginal discs. Cell 80:543-552[Medline].

PATEL, N. H., 1994  Imaging neuronal subsets and other cell types in whole-mount Drosophila embryos and larvae using antibody probes. Methods Cell Biol. 44:445-487[Medline].

PAWSON, T. and J. D. SCOTT, 1997  Signaling through scaffold, anchoring, and adaptor proteins. Science 278:2075-2080[Abstract/Free Full Text].

SHAULSKY, G., D. FULLER, and W. F. LOOMIS, 1998  A cAMP-phosphodiesterase controls PKA-dependent differentiation. Development 125:691-699[Abstract].

SKOULAKIS, E. M. C., D. KALDERON, and R. L. DAVIS, 1993  Preferential expression in mushroom bodies of the catalytic subunit of protein kinase A and its role in learning and memory. Neuron 11:197-208[Medline].

STRUTT, D. I., V. WIERSDORFF, and M. MLODZIK, 1995  Regulation of furrow progression in the Drosophila eye by cAMP-dependent protein kinase A. Nature 373:705-709[Medline].

TAYLOR, S. S., J. A. BUECHLER, and W. YONEMOTO, 1990  cAMP-dependent protein kinase: framework for a diverse family of regulatory enzymes. Annu. Rev. Biochem. 59:971-1005[Medline].

THE, I., G. E. HANNIGAN, G. S. COWLEY, S. REGINALD, and Y. ZHONG et al., 1997  Rescue of a Drosophila NF1 mutant phenotype by protein kinase A. Science 276:791-794[Abstract/Free Full Text].

WEN, W., J. L. MEINKOTH, R. Y. TSIEN, and S. S. TAYLOR, 1995a  Identification of a signal for rapid export of proteins from the nucleus. Cell 82:463-473[Medline].

WEN, W., S. S. TAYLOR, and J. L. MEINKOTH, 1995b  The expression and intracellular distribution of the heat-stable protein kinase inhibitor is cell cycle regulated. J. Biol. Chem. 270:2041-2046[Abstract/Free Full Text].

WHITEHOUSE-HILLS, S., H. J. BELLEN, and J. A. KIGER, JR., 1992  Embryonic cAMP and developmental potential in Drosophila melanogaster. Roux's Arch. Dev. Biol. 201:257-264.

WOLFGANG, W. J., I. J. H. ROBERTS, F. QUAN, C. O'KANE, and M. FORTE, 1996  Activation of protein kinase A-independent pathways by Gs in Drosophila. Proc. Natl. Acad. Sci. USA 93:14542-14547[Abstract/Free Full Text].

YEH, E., K. GUSTAFSON, and G. L. BOULIANNE, 1995  Green fluorescent protein as a vital marker and reporter of gene expression in Drosophila. Proc. Natl. Acad. Sci. USA 92:7036-7040[Abstract/Free Full Text].

This article has been cited by other articles:

				D. Ayaz, M. Leyssen, M. Koch, J. Yan, M. Srahna, V. Sheeba, K. J. Fogle, T. C. Holmes, and B. A. Hassan Axonal Injury and Regeneration in the Adult Brain of Drosophila J. Neurosci., June 4, 2008; 28(23): 6010 - 6021. [Abstract] [Full Text] [PDF]

				A. Guichard, J. M. Park, B. Cruz-Moreno, M. Karin, and E. Bier From The Cover: Anthrax lethal factor and edema factor act on conserved targets in Drosophila PNAS, February 28, 2006; 103(9): 3244 - 3249. [Abstract] [Full Text] [PDF]

				L. S. Collier, K. Suyama, J. H. Anderson, and M. P. Scott Drosophila Costal1 Mutations Are Alleles of Protein Kinase A That Modulate Hedgehog Signaling Genetics, June 1, 2004; 167(2): 783 - 796. [Abstract] [Full Text] [PDF]

				J. M. Shulman and M. B. Feany Genetic Modifiers of Tauopathy in Drosophila Genetics, November 1, 2003; 165(3): 1233 - 1242. [Abstract] [Full Text] [PDF]

				R. A. Baines Postsynaptic Protein Kinase A Reduces Neuronal Excitability in Response to Increased Synaptic Excitation in the Drosophila CNS J. Neurosci., September 24, 2003; 23(25): 8664 - 8672. [Abstract] [Full Text] [PDF]

				A. R. Rodan, J. A. Kiger Jr, and U. Heberlein Functional Dissection of Neuroanatomical Loci Regulating Ethanol Sensitivity in Drosophila J. Neurosci., November 1, 2002; 22(21): 9490 - 9501. [Abstract] [Full Text] [PDF]

				S. M. Jackson and C. A. Berg An A-kinase anchoring protein is required for Protein kinase A regulatory subunit localization and morphology of actin structures during oogenesis in Drosophila Development, January 10, 2002; 129(19): 4423 - 4433. [Abstract] [Full Text] [PDF]

J. A. Kiger Jr., J. E. Natzle, and M. M. Green Hemocytes are essential for wing maturation in Drosophila melanogaster PNAS, August 10, 2001; (2001) 181338998. [Abstract]