Drosophila phosphoinositide-dependent kinase-1 regulates apoptosis and growth via the phosphoinositide 3-kinase-dependent signaling pathway

Abstract

Phosphoinositide-dependent kinase-1 (PDK-1) is a central mediator of the cell signaling between phosphoinositide 3-kinase (PI3K) and various intracellular serine/threonine kinases including Akt/protein kinase B (PKB), p70 S6 kinases, and protein kinase C. Recent studies with cell transfection experiments have implied that PDK-1 may be involved in various cell functions including cell growth and apoptosis. However, despite its pivotal role in cellular signalings, the in vivo functions of PDK-1 in a multicellular system have rarely been investigated. Here, we have isolated Drosophila PDK-1 (dPDK-1) mutants and characterized the in vivo roles of the kinase. Drosophila deficient in the dPDK-1 gene exhibited lethality and an apoptotic phenotype in the embryonic stage. Conversely, overexpression of dPDK-1 increased cell and organ size in a Drosophila PI3K-dependent manner. dPDK-1 not only could activate Drosophila Akt/PKB (Dakt1), but also substitute the in vivo functions of its mammalian ortholog to activate Akt/PKB. This functional interaction between dPDK-1 and Dakt1 was further confirmed through genetic analyses in Drosophila. On the other hand, cAMP-dependent protein kinase, which has been proposed as a possible target of dPDK-1, did not interact with dPDK-1. In conclusion, our findings provide direct evidence that dPDK-1 regulates cell growth and apoptosis during Drosophila development via the PI3K-dependent signaling pathway and demonstrate our Drosophila system to be a powerful tool for elucidating the in vivo functions and targets of PDK-1.

Phosphoinositide-dependent kinase-1 (PDK-1) originally was identified as an upstream regulatory kinase of Akt/protein kinase B (PKB) (1–3). Consequently, the in vivo roles of PDK-1 have been inferred mainly from those of Akt/PKB. Akt/PKB is a growth factor-regulated serine/threonine kinase that contains a pleckstrin homology domain, as does PDK-1. It acts downstream of phosphoinositide 3-kinase (PI3K) to regulate various cellular activities, including glucose metabolism, transcription, and protein translation (4). Akt/PKB also negatively regulates apoptosis in various ways (5–10). To exert its antiapoptotic effects, Akt/PKB either inhibits the activities of proapoptotic proteins, such as BAD (11, 12) and caspase-9 (13), or induces antiapoptotic signals via the NF-κB- and forkhead transcription factor-dependent pathways (14–18). Recent transgenic studies in Drosophila revealed an unexpected function of Akt/PKB and the PI3K signaling pathway: the pathway plays an essential role in the control of cell size. When the activities of one or multiple components of the pathway, including PI3K (19), Drosophila akt1 (Dakt1) (20), and Drosophila p70 S6 kinase (21), were down-modulated, cell size as well as body size were dramatically reduced in a cell-autonomous manner.

Recent studies also suggest that PDK-1 is involved in the activation of members of the AGC superfamily of serine/threonine protein kinases (22), through phosphorylation of their activation loop in response to extracellular stimulations induced by peptide growth factors and hormones. A number of important kinases in this family, including Akt/PKB (1–3), p70 S6 kinase (23, 24), various protein kinase Cs (25, 26), protein kinase C-related kinases (27), and cAMP-dependent protein kinase (PKA) (28), have been proposed as either in vivo or in vitro targets of PDK-1. These results implicate that PDK-1 may play the role of a “master kinase” in regulating these downstream kinases. However, further investigation is required to determine the actual in vivo targets of PDK-1, as it was revealed that some of the AGC family kinases are not directly phosphorylated by PDK-1 in vivo, despite possessing a putative PDK-1 phosphorylation site at the activation loop and being phosphorylated by PDK-1 in vitro (29, 30). In addition, although PDK-1 is regarded as a regulator of at least some of these important kinases, the physiological role of the kinase in a multicellular system has not yet been defined at all. Such gaps in the current knowledge of PDK-1 have prompted us to establish the Drosophila model system for PDK-1 as described in this study.

The PDK-1 Drosophila homolog, dPDK-1 [also known as Drosophila serine/threonine protein kinase 61 (2); referred to as dPDK-1] is 54% identical to its human counterpart in the catalytic domain and is also highly homologous in the noncatalytic carboxyl terminus. In the present study, we have confirmed dPDK-1 to be the Drosophila ortholog of PDK-1 and further characterized the in vivo roles of the kinase through various genetic experiments. dPDK-1 could substitute mammalian PDK-1 in activating Akt/PKB in the cell and also genetically interact with Dakt1. Impairment of dPDK-1 caused ectopic apoptosis and a developmental arrest of the mutant fly at the embryonic stage, as previously demonstrated in the mutants of Dakt1, which is a well-known target of PDK-1. dPDK-1 also modulated cell size in a PI3K-dependent manner. These results provide clear evidence that dPDK-1 regulates Drosophila growth and development and demonstrate the important role of dPDK1 in the PI3K-dependent signaling pathway.

Materials and Methods

Drosophila Strains.

The GAL4 drivers [Patched (ptc)-GAL4, glass multimer reporter (gmr)-GAL4, apterous (ap)-GAL4, and heat-shock (hs)-GAL4] and green fluorescent protein (GFP) balancer chromosome (TM3, P{ActGFP}JMR2, Ser¹) were obtained from the Bloomington Drosophila Stock Center, Bloomington, IN. Upstream Activation Sequence (UAS)-Dp110 and UAS-PI3KDN were gifts from Sally Leevers, Ludwig Institute for Cancer Research, London, UK (19). UAS-Dakt1 was provided by Morris J. Birnbaum, University of Pennsylvania, Philadelphia (20). UAS-PKAc and UAS-PKAr were gifts from John A. Kiger, Jr., University of California, Davis (31).

Drosophila Embryo Analyses.

To isolate dPDK-1 (1) homozygotes, GFP balancer chromosomes (TM3, P{ActGFP}JMR2, Ser¹) were used. Because zygotic expression of GFP is initiated at about 12 h after egg laying, GFP-negative embryos were selected as dPDK-1¹ homozygotes under the fluorescent microscope at 14–16 h after egg laying (stage 16) at 25°C. For the analysis of cuticles, embryos were dechorinated with bleach and fixed in glycerol/acetic acid (1:4) at 60°C for 1 h. Then the cuticles were cleared in Hoyer's medium before microscopic analyses. Terminal deoxynucleotide transfer-mediated dUTP nick end labeling (TUNEL) analyses were performed as described (32), using the In Situ Cell Death Detection Kit (Roche Molecular Biochemicals). For the epistasis experiments, UAS-Dakt1, dPDK-1¹/TM3, Act-GFP, Ser¹ females were crossed to hs-GAL4, dPDK-1¹/TM3, Act-GFP, Ser¹ males, and the progeny were collected onto nylon sieves. GFP-negative embryos then were staged (stages 10–12) and heat-shocked at 37°C for 8 min. Embryos were allowed to recover at 25°C, and TUNEL analyses were performed as described above. Whole-mount immunostaining was carried out essentially as described (33). Anti-hemagglutinin (HA) antibody (Roche Molecular Biochemicals) was used to detect HA-tagged Dakt1 in the heat-shocked embryos.

RNA in Situ Hybridizations.

In situ hybridization experiments were performed with a digoxigenin-labeled RNA probe (34) of reverse transcriptase–PCR (RT-PCR) product of dPDK-1. For RT-PCR, the 5′ portion of the dPDK-1 sequences were used as PCR primers (5′-AACATACTGCGGCGGTTTCATAATTGTGC-3′ for 5′ primer and 5′-ATGCTTTCTCCTTGGCCATAGCCGGCATTG-3′ for 3′ primer).

Cloning and Mutagenesis of dPDK-1.

Full-length dPDK-1 (GenBank accession no. Y07908) was generated by PCR using a Drosophila adult head library as a template and subcloned into the pJ3M expression vector. This was achieved by using the 5′ primer 5′-CGGAATTCGCCAAGGAGAAAGCATCAGC-3′ and the 3′ primer 5′-GGGGTACCTTACTTAGACGCCGTCTTCT-3′. The resulting amplified PCR fragment was subcloned into the EcoRI–KpnI sites of pJ3M vector, and the nucleotide sequences were confirmed by DNA sequencing. We also generated a kinase-dead mutant of dPDK-1 by site-directed mutagenesis using the QuickChange Kit (Stratagene) according to the manufacturer's instructions.

Ectopic Gene Expression Using GAL4 System.

To examine the phenotypes generated by overexpression of dPDK-1, PI3K, PI3K dominant negative (PI3KDN), or Dakt1, we used the UAS-GAL4 system. In this system, the promoter (or enhancer) drives expression of the yeast transcriptional activator GAL4, which in turn activates the target gene containing the UAS, which holds GAL4 binding sites (35). The GAL4 gene is placed near a tissue-specific promoter/enhancer, allowing ectopic expression of the target gene in the desired tissue. For example, Ptc-GAL4, gmr-GAL4, and ap-GAL4 direct target gene expression in the border of anterior/posterior compartment of wing, the developing eye, and the dorsal compartment of wing, respectively. In addition, we have used in our study the hs-GAL4 driver which allows heat shock-dependent ectopic expression of target genes in the whole body.

EP lines developed for modular misexpression screening in Drosophila for detecting tissue-specific phenotypes (36) were used in our study, to ectopically express dPDK-1. In EP screening, the GAL4 line that express the GAL4 in a tissue of interest is crossed to an extensive panel of target lines, each carrying a single target P-element inserted at a unique, random position in the genome. The target element contains a Gal4-responsive enhancer at one end. In progeny containing both elements, an endogenous gene adjacent to the target element should be induced in cells expressing Gal4. Phenotypes due to overexpression or misexpression of the gene adjacent to a target element then can be scored directly, or as suppression or enhancement of a preexisting mutation.

Mammalian Cell Transfection and Akt/PKB Phosphotransferase Assays.

COS cells were maintained in DMEM supplemented with 10% FBS (Life Technologies, Grand Island, NY) at 37°C in a humidified atmosphere with 5% CO₂. Transient transfections in COS cells were performed at 60% confluency by a DEAE-dextran method as described in the manufacturer's manual (Promega). For the Dakt1 activity assay, 300 adult heads were cut and collected separately from gmr-GAL4, gmr-GAL4; UAS-Dakt1, or gmr-GAL4; UAS-Dakt1/EP(3)837. The heads were homogenized in buffer A (20 mM Tris⋅HCl, pH 7.5/0.1% Nonidet P-40/1 mM EDTA/5 mM EGTA/10 mM MgCl₂/50 mM β-glycerophosphate/1 mM sodium orthovanadate/2 mM DTT/40 μg/ml PMSF/10 μg/ml leupeptin). The lysates were clarified by centrifugation at 14,000 rpm for 15 min at 4°C. Akt/PKB activity assays were performed as described (37).

Results

dPDK-1 Plays an Important Role in Drosophila Embryogenesis.

To understand the roles of PDK-1 in a multicellular system, we have isolated flies containing mutations in the dPDK-1 locus for genetic analyses. With the help of the Berkeley Drosophila Genome Project, we found three P-element insertion mutants, EP(3)837, EP(3)3553, and EP(3)3091, containing P-element insertions in either the 5′ or intron region of the dPDK-1 gene (Fig. 1A). In detail, the inserted positions of the P-element in EP(3)837 and EP(3)3553, which have been determined by inverse PCR, are located at 179 bp and 144 bp upstream of the dPDK-1 transcription start site, respectively (Fig. 1A). The insertion sites and directions of the P-elements are oriented to induce gene expression and imply that these mutants can be used to study the gain of function of dPDK-1 (36). As expected, strong expression of dPDK-1 was observed in the eye imaginal disk of EP(3)837 line carrying the gmr-GAL4 driver, which directs expression of the gene in the developing eye (Fig. 2J). Another EP line, EP(3)3091, has a P-element in the fourth intron of dPDK-1 (Fig. 1A). The insertion site of EP(3)3091 predicts that the transcription of dPDK-1 is disrupted by the insertion of the transposon. Indeed, EP(3)3091 displayed a complete lethal phenotype. In addition, we have generated another PDK-1-deficient lethal line, ΔdPDK5, by an imprecise excision of P-element in EP(3)837. From genomic Southern analyses, we found that this line contained about a 10-kb deletion that includes the first exon of dPDK-1 (data not shown). This mutant failed to complement the lethality of EP(3)3091 (data not shown), suggesting that both lines are alleles of dPDK-1 mutants. Thus, EP(3)3091 and ΔdPDK5 are hereafter referred to as dPDK-1¹ and dPDK-1², respectively (Fig. 1A).

Figure 1.

Characterization of dPDK-1 mutants. (A) The insertion sites of the P-element in dPDK-1 mutants. The triangle represents the P-element. Two EP lines, EP(3)837 and EP(3)3553, have P-elements in the 5′ flanking region of dPDK-1 gene, and dPDK-1¹ has the P-element in the fourth intron. dPDK-1² was generated by an imprecise excision of P-element from EP(3)837. The deletion site of dPDK-1² is presented by a straight line, which includes the first exon of the kinase. (B) A heterozygous embryo of dPDK-1¹ (dPDK-1¹/TM3, Act-GFP, Ser¹) at stage 16. (D and F) Wild-type embryos at stage 16. (C, E, and G) dPDK-1¹ homozygous embryos (dPDK-1¹/dPDK-1¹) at stage 16. (B and C) Fluorescent confocal views of heterozygous (B) and homozygous (C) embryos. GFP was expressed only in the heterozygous embryo (B) in the midgut and salivary glands. We could first detect the expression of GFP about 12 h after egg laying (stage 15), when the zygotic gene expression was initiated. (D and E) The phenotype of the wild-type embryo (D) and embryo lacking the zygotic dPDK-1 activity (E). The mutant embryos showed a complete loss of cuticles. (F–H) TUNEL signals in stage 16 of wild type (F), the embryo lacking zygotic dPDK-1 activity (G), and the dPDK-1¹/dPDK-1¹ mutant embryo ectopically expressing Dakt1 (H). (I) Whole-mount immunostaining analyses of heat-shocked (hs+) or control (hs−) hs-GAL4, dPDK-1¹/UAS-Dakt1, dPDK-1¹ embryos were carried out with anti-HA antibody. HA-tagged Dakt1 was strongly induced in the heat-shocked embryo. (Magnifications: ×100.)

Figure 2.

Effects of dPDK-1 on cell size control. Scanning electron micrographs of the external eyes (A–F) and their tangential sections (G–I) representing the genotypes indicated below are shown. (A, D, and G) gmr-GAL4/+. (B, E, and H) EP(3)837/+. (C, F, and I) gmr-GAL4/+;EP(3)837/+. (J) Induction of dPDK-1 mRNA expression in the eye imaginal disk by gmr-GAL4. Samples were prepared from EP(3)837/EP(3)837 (Upper) or gmr-GAL4/+;EP(3)837/+ (Lower). Overexpressed dPDK-1 mRNA was detected in the imaginal disk of gmr-GAL4/+;EP(3)837/+ by in situ hybridization as described in Materials and Methods. (K) Comparison of wing phenotypes of ap-GAL4/+ (Left) and ap-GAL4/+;EP(3)3553/+ (Right). (Inset) The basal view of the right wing of ap-GAL4/+;EP(3)3553/+. (Magnifications: A–C, ×200; D–I, ×1,000; J, ×100; K, ×20.)

None of the homozygous dPDK-1¹ and dPDK-1² flies emerged as larva, and both displayed an embryonic lethality. To isolate dPDK-1¹ homozygous individuals, we used a GFP balancer chromosome as described in Materials and Methods. As shown in Fig. 1 B and C, the GFP-negative embryos were selected as dPDK-1 homozygotes. All of the hatched larvae from dPDK-1¹ or dPDK-1²/TM3, GFP, Ser females showed GFP expression. As shown in Fig. 1E, dPDK-1¹ homozygous embryos produced no ventral cuticles (compare with the same-aged GFP expressing heterozygotes in Fig. 1D), and they did not develop into the larval stage. These results are similar to those seen in the mutation of Dakt1 (38), whose mammalian homologs are well-known targets of PDK-1. Briefly, absence of maternal and zygotic Dakt1 activity also resulted in an embryonic lethality, along with defective cuticle formation.

Next, we tested whether dPDK-1 also is involved in the cell survival-signaling pathway. TUNEL assays were performed with dPDK-1¹ homozygous embryos to examine the involvement of the kinase in apoptosis, as described in Materials and Methods. As shown in Fig. 1G, apoptotic activity was dramatically induced in the dPDK-1 zygotic loss-of-function mutant. Compared with the extensive and widespread TUNEL signals in the dPDK-1¹ embryos, we hardly detected any apoptotic activities in the wild-type embryos of the same age (Fig. 1F). Furthermore, the induced apoptosis in dPDK-1¹ mutant embryos was extensively suppressed by expression of Dakt1 using the hs-GAL4-UAS system (Fig. 1H). An induced expression of Dakt1 by heat shock in the embryos used for Fig. 1H was further confirmed by immunostaining analyses as described in Materials and Methods (Fig. 1I). Collectively, these results strongly suggest that PDK-1 plays an important role in Drosophila embryonic development and apoptosis.

dPDK-1 Regulates Cell Growth and Proliferation.

As described previously, a series of components in the PI3K pathway including Dakt1 and Drosophila p70 S6 kinase modulate cell size in a cell-autonomous manner. Thus, we examined whether overexpression of dPDK-1 affects cell size using the GAL4-UAS system in Drosophila. dPDK-1 was overexpressed under the control of gmr-GAL4, which directs expression of the gene in the developing eye. This ectopic overexpression of dPDK-1 caused an increase in ommatidia size (≈1.33-fold bigger than controls, compare Fig. 2F with D and E or Fig. 2I with G and H). In addition, we examined the effect of overexpression of dPDK-1 in a specific compartment of the wing disk. The wing disk is composed of two compartments (dorsal and ventral), which fold and generate the flattened wing blade. When we ectopically overexpressed dPDK-1 in the dorsal compartment with ap-GAL4 driver, the wing of EP(3)837 was convex toward the dorsal side as shown in Fig. 2K. This is likely the result of an increase in the size of the cells in the dorsal compartment. Indeed, a similar situation was observed in the UAS-Drosophila p70 S6 kinase flies (21). These results suggest that dPDK-1 regulates cell and organ size.

dPDK-1 Functions Downstream of the Drosophila PI3K.

Despite the fact that there is no clear evidence on how the intrinsic kinase activity of PDK-1 is regulated, the kinase has been found to act downstream of PI3K (1–3). Thus, we examined whether dPDK-1 and PI3K can genetically interact in the fly lines, in which dPDK-1 was coexpressed with PI3K or a dominant negative Dp110 (PI3KDN). A previous study showed that overexpression of the PI3K catalytic subunit, Dp110, increased cell size, whereas overexpression of a PI3KDN results in the opposite phenotype (19). This change in cell size results in the change of organ and body size. Overexpression of PI3KDN under ptc-GAL4 [the driver induces GAL4 expression throughout the anterior compartment with a stripe of maximal intensity along the border of anterior/posterior compartment extending into the posterior compartment (39)] resulted in reduction of the distance between L3 and L4 veins (Fig. 3A). However, this phenotype was strongly suppressed by coexpression of dPDK-1 with PI3KDN (Fig. 3B), suggesting that dPDK-1 acts as a vital downstream effector of PI3K in cell and compartment size control. Conversely, overexpression of the PI3K wild-type caused an increase in the distance between L3 and L4 veins (Fig. 3C), and coexpression of dPDK-1 and PI3K further increased the distance in a synergistic manner (Fig. 3D). These results provide strong in vivo evidence that dPDK-1 functions downstream of Drosophila PI3K in the control of cell and compartment size.

Figure 3.

Genetic interactions between dPDK-1 and Drosophila PI3K in cell size control. Microscopic views of the wings expressing ptc-GAL4/UAS-PI3KDN (A), ptc-GAL4/UAS-PI3KDN; EP(3)3553/+ (B), ptc-GAL4/UAS-PI3K (C), and ptc-GAL4/UAS-PI3K; EP(3)3553/+ (D). The relative distances between two arrows are ptc-GAL4/UAS-PI3KDN (58%), ptc-GAL4/UAS-PI3KDN; EP(3)3553/+ (100%), ptc-GAL4/UAS-PI3K (125%), and ptc-GAL4/UAS-PI3K; EP(3)3553/+ (142%), respectively. Increased activities of the PI3K signaling pathway induced a loss of cross-veins (B and D), as reported (45).

dPDK-1 Is a Dakt1 Upstream Kinase.

To determine whether dPDK-1 functions in a manner similar to its mammalian counterpart, myc-tagged dPDK-1, myc-tagged human PDK-1 (hPDK-1), and/or HA-tagged human Akt/PKB were transiently expressed in COS cells (Fig. 4A). As expected, dPDK-1 strongly induced human Akt/PKB activity, to levels comparable to those induced by hPDK-1. Conversely, coexpression of a dominant negative hPDK-1 or a dominant negative dPDK-1 strongly inhibited the epidermal growth factor-induced activation of human Akt/PKB. These results indicate that the Drosophila ortholog of PDK-1 can properly function and substitute its mammalian counterpart to relay the growth factor-induced activation signals to a mammalian Akt/PKB.

Figure 4.

dPDK-1 activates both mammalian and Drosophila Akt/PKB. (A) Activation of mammalian Akt/PKB by Drosophila PDK-1 in COS cells. The cells were transiently transfected with pJ3H-human Akt/PKB (Akt) alone, pJ3H-Akt plus wild type (WT) or kinase dead (KD; K111I) pJ3 M-human PDK-1 (hPDK-1), or pJ3H-Akt plus wild type (WT) or kinase dead (KD; K191I) pJ3 M-dPDK-1. Quiescent cells were stimulated with (+) or without (−) epidermal growth factor (EGF). Cell lysates were subjected to either immune complex kinase assays for HA-tagged Akt/PKB (second from Top) or immunoblot analyses for Myc-tagged PDK-1 (third from Top) and HA-tagged Akt/PKB (Bottom). The Akt/PKB activity was quantified and shown as a bar graph (Top). (B) Activation of Dakt1 by dPDK-1 in the raw tissue of Drosophila. Head extracts from gmr-GAL4/±, gmr-GAL4/+; UAS-HA-Dakt1/+, or gmr-GAL4/+; UAS-HA-Dakt1/EP(3)3553 were subjected to immune complex kinase assays for HA-tagged Dakt1 (Middle), and Dakt1 activity was quantitated and shown as a bar graph (Top). The same tissue lysates were used for immunoblot analyses for HA-tagged Dakt1 (Bottom).

Next, we examined whether dPDK-1 can activate Dakt1 in Drosophila. To test this, dPDK-1 and HA-tagged Dakt1 were coexpressed in the Drosophila eye using the gmr-GAL4 driver, and the phosphotransferase activities of Dakt1 were measured from the head extracts of gmr-GAL4, gmr-GAL4; UAS-HA-Dakt1, or gmr-GAL4; UAS-HA-Dakt1/EP(3)837. As shown in Fig. 4B, Dakt1 activity was strongly increased in the flies coexpressing dPDK-1. Consistent with this increased activity, an electrophoretically retarded Akt/PKB band, corresponding to a highly phosphorylated and activated form, was observed (Fig. 4B, lane 3, Lower). This biochemical evidence strongly supports that Dakt1 is indeed a physiological target of dPDK-1.

To further confirm the in vivo roles of dPDK-1, we examined genetic interactions between Dakt1 and dPDK-1 in Drosophila. Overexpression of Dakt1 in the Drosophila eye increased eye size and generated a bulging eye with enlarged ommatidia (Fig. 5 B, F, and J) compared with the control (Fig. 5 A, E, and I), as reported (20). In addition to this change in size, the ommatidia array became irregular, and eye bristles were enlarged with a frequent loss of number. When dPDK-1 was coexpressed with Dakt1 in the eye, it displayed a severely crumpled morphology (Fig. 5C). The eye bristles were enlarged even more severely, and the boundaries of all ommatidia and photoreceptor cells disappeared (Fig. 5 G and K). These dPDK-1/Dakt1 phenotypes were further enhanced by an increased dose of gmr-GAL4 driver (Fig. 5 D, H, and L). These findings, taken together, clearly demonstrate the functional and genetic interactions between dPDK-1 and Dakt1.

Figure 5.

dPDK-1 genetically interacts with Drosophila Akt, but not with Drosophila PKA. Scanning electron micrographs of the external eyes (A–H and M–P) and their tangential sections (I–L). The genotypes of the samples were (A, E, and I) UAS-Dakt1/UAS-Dakt1, (B, F, and J) gmr-GAL4/+; UAS-Dakt1/+, (C, G, and K) gmr-GAL4/+; UAS-Dakt1/EP(3)837, (D, H, and L) gmr-GAL4/gmr-GAL4; UAS-Dakt1/EP(3)837; (M) gmr-GAL4/+; UAS-dPKAc/+; (N) gmr-GAL4/+; UAS-dPKAc/EP(3)837; (O) gmr-GAL4/+; UAS-dPKAr/+; (P) gmr-GAL4/+; UAS-dPKAr/EP(3)837. (Magnifications: A–D and M–P, ×200; E–L, ×1,000.)

dPDK-1 Does Not Interact with Drosophila PKA in Vivo.

We also examined the genetic interactions between dPDK-1 and Drosophila PKA. As described previously, although PKA has been proposed to be a putative substrate of PDK-1 (28), the in vivo relevance of this has not been clearly determined. When the catalytic subunit of Drosophila PKA (dPKAc) was overexpressed in the developing eye of Drosophila, the eye was discolored and swelled up with wrinkles (Fig. 5M). Scanning electron microscopic views of the eye showed that the boundaries of all ommatidia and photoreceptor cells disappeared. However, unlike Akt/PKB, coexpression of dPDK-1 did not affect these phenotypes of dPKAc (Fig. 5N). Furthermore, the regulatory subunit of Drosophila PKA (dPKAr) also did not interact with dPDK-1 (Fig. 5 O and P). These results support that PKA is not regulated by PDK-1 in Drosophila, which is highly consistent with recent results that PKA is phosphorylated and activated normally in a PDK-1-deficient cell line (30). These results strongly support that our Drosophila system is a physiologically relevant tool for determining the actual in vivo targets of PDK-1.

Discussion

In the present study, we have isolated and characterized Drosophila PDK-1 mutants. The dPDK-1-deficient mutants in our study displayed an embryonic lethality. As shown in Fig. 1, the zygotic loss-of-function mutants of dPDK-1 failed to grow into larva, and they did not form proper cuticles. This finding corresponds with previous reports, in which disruptions of PDK-1 homologs resulted in lethal phenotypes in both budding and fission yeast (40–42). Furthermore, loss-of-function mutations of PDK-1 in Caenorhabditis elegans also cause a developmental arrest at the dauer larval stage (43). These results support the idea that PDK-1 activity is critical for the normal development of eukaryotes.

In metazoan development, apoptosis is an essential process that regulates tissue homeostasis. Especially, Akt/PKB has been implicated in antagonizing this critical process via several mechanisms, including phosphorylation and inactivation of proapoptotic proteins, such as BAD and caspase-9 (11–13). In Drosophila, Dakt1-deficient mutants displayed an induction of ectopic apoptosis and defective cuticle formation during embryogenesis (38). Interestingly, these defects in the Dakt1 mutant were rescued by ectopic expression of p35 caspase inhibitor, indicating the involvement of caspases in Dakt1-mediated apoptosis. In the present study, the dPDK-1-deficient embryos showed extensive apoptosis throughout the entire body (Fig. 1G), similar to the Dakt1 mutant (38). In addition, the normal, nonapoptotic phenotype was rescued by overexpressing Dakt1 in the dPDK-1 mutants. This finding demonstrates PDK-1's role in apoptosis related to an organism's development and also strongly suggests that PDK-1 acts upstream of Akt/PKB in the modulation of apoptosis during Drosophila development, as shown in Fig. 1H. Consistent with this possibility, dPDK-1 promoted the activation of Dakt1 in cultured cells or in raw tissues. Furthermore, dPDK-1 enhanced Dakt1 phenotypes during eye morphogenesis, including photoreceptor cell degeneration and bristle enlargement.

Our results also strongly support that dPDK-1 modulates cell size. Overexpression of dPDK-1 not only increased cell and compartment size, but also suppressed the cell and compartment reducing phenotypes of PI3KDN. These results are highly consistent with previous results showing other PI3K signaling molecules, such as Akt/PKB and p70 S6 kinase, to be involved in cell or organ size control. Thus, we could conclude that dPDK-1 functions between Drosophila PI3K and Dakt1 to relay extracellular signals to cell growth promotion.

Although it has been firmly established in both in vitro and in vivo systems that Akt/PKB (1–3), p70 S6 kinase (23, 24), and protein kinase C (25, 26) are downstream targets of PDK-1, recent in vitro studies have suggested that three other AGC family kinases, PKA (28), RSK (44), and protein kinase C-related kinase (27), are regulated by PDK-1. However, further investigation is required to determine whether these are actual in vivo targets of PDK-1, as experiments with mammalian cell lines demonstrated that the activity of these three kinases was not affected by the absence of PDK-1 (29, 30). Interestingly, our results demonstrated that Drosophila PKA is not regulated by dPDK-1 in an in vivo multicellular system, as coexpression with dPDK-1 in Drosophila did not have an effect on the PKA overexpression phenotype. Actually, this disparity between the in vitro and in vivo activities of PDK-1 was not unexpected, as the activity of the kinase highly depends on its localization and the activities of its upstream regulators in vivo, which is not accounted for in in vitro experiments. Thus, the Drosophila system we have developed appears to be a relevant and powerful tool for determining the in vivo validity of proposed targets of PDK-1 and may be used in the future to examine the validity of RSK and protein kinase C-related kinase as PDK-1 targets.

In summary, we have provided genetic evidence that PDK-1 controls apoptosis and cell growth through Akt and in a PI3K-dependent manner. In addition, we have demonstrated a powerful method for elucidating the in vivo functions and targets of PDK-1 in a multicellular system.

Abbreviations

PDK-1

phosphoinositide-dependent kinase-1

PI3K

phosphoinositide 3-kinase

PKB

protein kinase B

dPDK-1

Drosophila phosphoinositide-dependent kinase-1

Dakt1

Drosophila akt1

PKA

cAMP-dependent protein kinase

GFP

green fluorescent protein

UAS

upstream activation sequence

TUNEL

terminal deoxynucleotide transfer-mediated dUTP nick end labeling

HA

hemagglutinin

PI3KDN

PI3K dominant negative