The PIKE homolog Centaurin gamma regulates developmental timing in Drosophila

Abstract

Phosphoinositide-3-kinase enhancer (PIKE) proteins encoded by the PIKE/CENTG1 gene are members of the gamma subgroup of the Centaurin superfamily of small GTPases. They are characterized by their chimeric protein domain architecture consisting of a pleckstrin homology (PH) domain, a GTPase-activating (GAP) domain, Ankyrin repeats as well as an intrinsic GTPase domain. In mammals, three PIKE isoforms with variations in protein structure and subcellular localization are encoded by the PIKE locus. PIKE inactivation in mice results in a broad range of defects, including neuronal cell death during brain development and misregulation of mammary gland development. PIKE -/- mutant mice are smaller, contain less white adipose tissue, and show insulin resistance due to misregulation of AMP-activated protein kinase (AMPK) and insulin receptor/Akt signaling. here, we have studied the role of PIKE proteins in metabolic regulation in the fly. We show that the Drosophila PIKE homolog, ceng1A, encodes functional GTPases whose internal GAP domains catalyze their GTPase activity. To elucidate the biological function of ceng1A in flies, we introduced a deletion in the ceng1A gene by homologous recombination that removes all predicted functional PIKE domains. We found that homozygous ceng1A mutant animals survive to adulthood. In contrast to PIKE -/- mouse mutants, genetic ablation of Drosophila ceng1A does not result in growth defects or weight reduction. Although metabolic pathways such as insulin signaling, sensitivity towards starvation and mobilization of lipids under high fed conditions are not perturbed in ceng1A mutants, homozygous ceng1A mutants show a prolonged development in second instar larval stage, leading to a late onset of pupariation. In line with these results we found that expression of ecdysone inducible genes is reduced in ceng1A mutants. Together, we propose a novel role for Drosophila Ceng1A in regulating ecdysone signaling-dependent second to third instar larval transition.

Figure 1. Drosophila Ceng1A is a functional GTPase with a catalytic internal GAP domain.

(A) Schematic representation of the predicted domain structure of the three murine PIKE isoforms (PIKE-A, -L and –S) and the homologous Drosophila Ceng1A isoforms (Ceng1A-PA, -PB and -PC). ceng1A codes for a N-terminal GTPase domain, a PH domain, a GAP domain and an ankyrine motif. Predicted domains of the three Drosophila Ceng1A proteins with PIKE domains show a high degree of conservation. Numbers indicate percentage of identity [similarity] on the amino acid level. (B) The GTPase and GAP domains of Ceng1A were used for a colorimetric in vitro GTPase assay. Two constructs were cloned and expressed in E. coli for the analysis: A 6xHis tagged construct containing the C-terminus of Ceng1A including the GAP domain and a 6xHis tagged construct containing the GTPase domain. (B') The graph depicts absorption at 635 nm versus protein concentration in µM. Addition of the GAP domain increases GTP hydrolysis of the GTPase domain. (B’’) Relative amount of hydrolyzed GTP. The GTPase domain alone shows modest GTPase activity, but activity was increased 1.5-fold by including the GAP domain. (C) Gene locus organization and generation of ceng1A knock-out mutants. Exon/intron structure of the ceng1A locus is depicted. Start sites of the three transcripts (ceng1A -RA, -RB and –RC) are indicated. A loss-of-function mutation for ceng1A was generated by ends-out gene targeting . The targeting construct for the homologous recombination was designed to delete exons 5–10, which encode all functional domains. (C') Real-time RT-PCR analysis of ceng1A expression in the generated ceng1A mutants.

Figure 2. ceng1A mutants do not differ from control animals in size and weight in larval, pupal or adult stages.

For all experiments zygotic and maternal ceng1A mutant animals were used. (A – A’’) Comparison of ceng1A mutant adults, larvae and pupa with wildtypic counterparts does not reveal any morphological defects. (B) Length of third instar ceng1A mutant pupae is not altered relative to controls. (C) Total weight (in mg) as well as area of the wings (D) does not differ in adult ceng1A mutants compared to controls. n = 50; error bars indicate SEM; p>0,5 (t-test).

Figure 3

. Ceng1A is not involved in metabolic control. (A) ceng1A mutant larvae feed normally, since quantitative analysis of Brilliant Blue FCF-colored yeast uptake shows no difference between control and ceng1A mutants. n = 10, each n equals 10 larvae; error bars indicate SEM. Real-time RT-PCR analysis of early third instar larvae (B) and dissected CNS (C) shows that transcription of InR, 4EBP, lip3, dilp2, 3 and 5 as well as PTTH is not changed in ceng1A mutants. n = 10; error bars indicate SEM; n.s.  = p>0,5 (t-test). (D) Immunoblots using anti-pAMPK and anti-Akt antibodies indicate that there is no difference in AMPK and Akt phosphorylation between control and ceng1A mutant larvae in starvation and fed conditions. Anti-actin served as a loading control. Quantification (D') of western blots of control and ceng1A mutant larvae stained for pAMPK relative to loading control. n = 3; error bars indicate SEM. (E) Survival curve of ceng1A mutant and w- control flies under starvation conditions. The mean survival time of ceng1A mutants (53,6 h) is not significantly changed compared to w- control flies (50,5h). p>0.5 (log rank test) (F) Thin layer chromatography to determine triacylglyceride (TAG) levels of control and ceng1A mutant flies upon starvation. Height of TAGs is shown in the right lane. Samples were taken at the start of the experiment (0h) and after 20 and 28 hours of starvation. (F') Body TAG levels are not changed in ceng1A mutants. Graph represents the relative absorbance of TAG bands quantified by photo densitometry. n = 3; error bars indicate SEM (G) Oil Red O staining of third instar fat bodies. Larvae were grown on either standard food (NFD) or high-sugar food (HSD). Fat body storage lipid load of ceng1A mutants does not differ from control larvae. (G') Analysis of relative lipid droplet size reveals no differences between control and ceng1A mutant larvae under both conditions. n = 50; error bars indicate SEM; p>0,5 (t-test).

Figure 4. ceng1A mutants show delayed second instar larval stage and reduced ecdysone signaling.

Relative amount of control (A) and ceng1A mutants (B) in L1, L2, L3 and pupal stage was determined from egg lay (0) to 140 hours after egg deposition. n = 3; each n corresponds to 50 larvae; error bars indicate SEM. ceng1A mutants are delayed in second instar larval stage compared to controls. (C – E) Onset of pupariation ceng1A mutant or wildtypic larvae on standard food (C, NFD), high fat diet (D, HFD) and high sugar diet (E, HSD) was analyzed. Delay in development of ceng1A mutant larvae is nutrition independent. n = 3; each n equals 50 larvae; error bars indicate SEM (F) Scheme of average stage length in control and ceng1A mutant larvae (derived from A and B). Growth rate was assessed as an increase of weight over time from egg deposition to pupariation. The main delay of growth is between 45 and 80 hours after egg lay (corresponding to second instar larval stage). Afterwards, growth rate of ceng1A mutants increases in parallel to the control growth rate, only with the in L2 stage accumulated time delay. (G) From 48 hours after egg deposition to pupariation, expression of the ecdysone target genes E75B was analyzed via real-time RT-PCR. Peaks of E75B expression coincide in control and ceng1A mutant animals. However, E75B induction levels are reduced in ceng1A mutants. Transparent yellow bars correlate peaks of E75B expression with the growth rate at that time point. n = 5 for all experiments; error bars indicate SEM.