Patterns of Gene Expression in Drosophila InsP3 Receptor Mutant Larvae Reveal a Role for InsP3 Signaling in Carbohydrate and Energy Metabolism

Background The Inositol 1,4,5-trisphosphate receptor (InsP3R) is an InsP3 gated intracellular Ca2+-release channel. Characterization of Drosophila mutants for the InsP3R has demonstrated that InsP3-mediated Ca2+ release is required in Drosophila larvae for growth and viability. Methodology/Principal Findings To understand the molecular basis of these growth defects a genome wide microarray analysis has been carried out with larval RNA obtained from a strong InsP3R mutant combination in which 1504 independent genes were differentially regulated with a log2 of fold change of 1 or more and P<0.05. This was followed by similar transcript analyses from InsP3R mutants where growth defects were either suppressed by introduction of a dominant suppressor or rescued by ectopic expression of an InsP3R transgene in the Drosophila insulin like peptide-2 (Dilp2) producing cells. Conclusions/Significance These studies show that expression of transcripts related to carbohydrate and amine metabolism is altered in InsP3 receptor mutant larvae. Moreover, from a comparative analysis of genes that are regulated in the suppressed and rescued conditions with the mutant condition, it appears that the organism could use different combinations of pathways to restore a ‘normal’ growth state.


Introduction
The efficient survival of multicellular organisms requires physiological co-ordination between cells, tissues and various organs. This co-ordination is achieved through signaling pathways some of which evolved in parallel with multicellular complexity. Release of calcium (Ca 2+ ) from intracellular stores appears to be such a signaling pathway that co-evolved with metazoan life forms suggesting that it might be an important modulator of basic physiological pathways [1]. Our understanding of how intracellular Ca 2+ signaling modulates systemic physiology however remains cursory. Intracellular Ca 2+ -release in response to inositol 1, 4, 5-trisphosphate (InsP 3 ) signals occurs through a ligand-gated Ca 2+ channel, the InsP 3 receptor (InsP 3 R) present on the membranes of intracellular Ca 2+ stores, primarily in the endoplasmic reticulum (ER). Vertebrate genomes have three genes for the InsP 3 R for which knock-outs have been generated in the mouse model [2,3]. Among these, the InsP 3 R1 knock-out animals are lethal to a large extent. Analysis of the few survivors obtained showed growth defects and ataxia [3]. In the InsP 3 R2 and 3 knock-outs exocrine secretion is defective resulting in feeding defects [2].
The Drosophila genome contains a single gene for the InsP 3 R (itpr), making it an attractive genetic model for understanding systemic roles for InsP 3 -mediated Ca 2+ release. There are several well characterized itpr mutant alleles of which the stronger alleles are larval lethal and exhibit severe growth defects prior to lethality [4,5]. In order to understand the molecular basis of these growth defects we have carried out a series of transcriptional profiling experiments from itpr mutant animals in the absence and presence of either an extragenic suppressor or a rescuing transgene. The suppressor is the Ca-P60A Kum170ts mutant allele for the Drosophila sarco-endoplasmic reticular Ca 2+ -ATPase (SERCA), referred to as Kum 170 throughout this paper [6]. Kum 170 reduces the rate of Ca 2+ uptake by the ER at 25uC in Drosophila neurons [7] and thus suppresses itpr mutant alleles by altering the dynamics of intracellular Ca 2+ signals. The rescued condition has been described earlier and consists of expression of an itpr + cDNA transgene in insulin-producing cells (IPCs) of the brain [4]. In vertebrates changing levels of growth hormones and growth factors like the Insulin-like Growth Factors (IGFs) strongly affect body and organ size. In Drosophila there is no functional separation between insulin-like growth factors and insulin signaling. Thus a single insulin/IGF system manages growth and energy metabo-lism. The cellular and molecular basis of rescue of itpr mutants by expression of the itpr + cDNA in Drosophila IPCs remains to be understood.
Here, we characterized the transcriptional profile of itpr mutant larvae just prior to the manifestation of growth deficits, followed by a comparison of these changes in transcripts from suppressed and rescued mutant animals. Our data show that growth deficits are preceded by significant changes in gene expression and support a link between intracellular calcium signaling and energy metabolism. Reversal of the growth deficit either by suppression or rescue has helped identify candidate pathways and genes that might function downstream of intracellular Ca 2+ release.

Materials and Methods
Drosophila melanogaster strains itpr sv35/ug3 is a larval lethal heteroallelic combination of single point mutants in the itpr gene generated in an ethyl methanesulfonate (EMS) screen. Detailed molecular information on these alleles has been published [5]. The embryonic wild-type itpr cDNA (UASitpr + ) [8] was used for rescue experiments and Ca-P60A Kum170ts [6,9] was used as a suppressor. The dOrai 20119 allele (referred to as dOrai 2 ) was procured from the Bloomington Drosophila Stock Center [10]. Dilp2GAL4 used for the rescue experiments was from Dr. E. Rulifson [11]. The wild-type Drosophila strain used in all experiments is Canton-S (CS). Fly strains were generated by standard genetic methods using individual mutant and transgenic fly lines described above. Flies were grown at 25uC in standard cornmeal medium containing agar, corn flour, sucrose, yeast extract along with anti-bacterial and anti-fungal agents.

Larval staging
Staging experiments for obtaining molting profiles of heteroallelic mutant larvae were carried out as described previously with minor modifications [5]. Briefly, flies were allowed to lay eggs for a period of 8 hrs at 25uC. Embryos were allowed to hatch and grow further at this temperature. Larvae of the desired genotype were selected from these cultures at 56-64 h after egg laying (AEL) and transferred into vials of standard cornmeal medium lacking agar (agar less medium). Larvae were grown in agar less medium at 25uC and screened at the indicated time points for number of survivors and their phase of growth and development. For each time interval, a minimum of 75 larvae were screened in batches of 25 larvae each. Computation of means, SEM, and t-tests were performed using Origin 7.5 software (Origin Lab, Northampton, MA, USA).

RNA isolation for microarray and qPCR
For isolation of total RNA, larvae of the requisite genotypes were selected at 58-62 hr AEL and snap frozen in liquid nitrogen. Total RNA was isolated using TRI Reagent (Sigma, St. Louis, USA) according to manufacturer's specifications. After quantification, RNA with a ratio of OD 260 /OD 280 .1.8 was taken for further experiments. Integrity of the isolated RNA was confirmed by the presence of full length rRNA bands on a 1.2% formaldehyde agarose gel. For microarrays three independent sets of RNA were isolated from control (CS), mutant (itpr sv35/ug3 ) and rescued larvae (UASitpr + /+; Dilp2GAL4/+ /ug3 ; referred to as rescue) and two sets of RNA from Kum 170ts suppressed larvae (Kum 170ts /+; itpr sv35/ug3 ; referred to as suppressor). For qPCR three independent sets of RNA were isolated from all genotypes tested.

Microarrays
Further quality control of isolated RNA, labeling of RNA probes with either cyanine 3-dUTP (Cy3) or cyanine 5-dUTP (Cy5), their hybridization to one Drosophila 8*60 K array AMADID: 27326 (Agilent Technologies Inc., Santa Clara, CA, USA) consisting of 8 replicate 60 K microarrays, scanning of signal intensities and analysis plus normalization of signal intensities were carried out by Genotypic Technology, Bangalore, India. Details of the numbers of probes present on the 60 K microarray are given in Table 1. Approximately 6000 EST probes could not be matched to existing CG IDs in FlyBase [10]. The experiments consisted of three arrays hybridized with control and mutant RNA (mutant), three arrays hybridized with mutant and Dilp rescue (rescue) and two arrays hybridized with mutant and Kum suppressor (suppressor) conditions. Independent sets of larvae were collected and used to isolate RNA for each of the hybridizations. The biological replicates were tested for reproducibility using the Pearson's correlation protocol from Microsoft Office Excel 2003. Correlation coefficients for biological replicates were.0.6 in all cases ( Table 2). All data obtained is MIAME compliant and raw hybridization data from the microarray have been submitted to the Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo) under series GSE29736.

Data analysis
Data analysis was performed using GeneSpring GX7.3.1 software (Silicon Genetics, Redwood City, CA, USA). Differentially regulated genes were ranked on the basis of signal intensity, normalized ratio, flag value and variance across biological replicate experiments. Top ranked genes had a higher intensity; high-normalized ratio for up and low for down; they were unflagged and showed very low variance or standard deviation. The normalized signal intensity of all filtered genes was taken for calculating log2 of the fold change in all experiments. The default multiple test correction in GeneSpring GX is the Benjamini and Hochberg False Discovery Rate protocol which provides a good balance between discovery of statistically significant genes and limitations of false positive occurrences. This, combined with a ttest performed at the 0.05% significant level for each gene, was used for calculating the final P-values. With this protocol the genes identified by chance (false discovery rate) is 5% of genes that are considered statistically significant.
Annotations of biological processes, molecular function and cellular localization were obtained using publicly available Gene Ontology information (The Gene Ontology Consortium) [12] uploaded to GeneSpring GX7.3.1 software and the publicly available DAVID database [13,14] and FlyBase [10]. For comparison with published microarrays, lists of regulated genes with CG IDs were downloaded. Numbers of common genes were obtained by comparing the CG IDs from published data with our data using GeneSpring GX.   Table 2. Correlation analysis among the biological replicates of microarray hybridizations for RNA obtained from larvae of the genotypes itpr sv35/ug3 (mutant), UASitpr + ; Dilp2GAL4/+; itpr sv35/ug3 (rescue) and Kum 170/+ ; itpr sv35/ug3 (suppressor).    Each qPCR experiment was repeated three times with independently isolated RNA samples. Cycling parameters were 95uC for 5 min, 40 cycles of 95uC for 15 s and 60uC for 30 s, 72uC for 30 s, then 1 cycle of 72uC for 5 min. The fluorescent signal produced from the amplicon was acquired at the end of each polymerization step at 72uC. A melt curve was performed after the assay to check for specificity of the reaction. The fold change of gene expression in the mutant relative to wild-type was determined by the comparative DDC t method [15]. In this method the fold change = 2 2DDCt where DDC t = (C t(target gene) -C t(rp49) ) mutant 2 (C t(target gene) 2 C t(rp49) ) Wild type .

Results
Growth defects in strong InsP 3 R mutants can be suppressed by reducing Ca 2+ uptake into the endoplasmic reticulum Ca 2+ store Strong hypomorphic alleles of the Drosophila itpr gene exhibit severe growth defects prior to lethality [4]. To understand the molecular changes and Ca 2+ signals underlying these growth defects a dominant loss-of-function allele for the Sarco-endoplasmic reticulum Ca 2+ ATPase (referred to as Kum 170 ) was introduced in the lethal itpr heteroallelic mutant combination of itpr sv35/ug3 . Previous studies have shown that the Kum 170 mutant allele can partially suppress phenotypes of weak itpr mutant alleles that arise during pupal development [9]. The cellular basis of this suppression is apparently elevated basal cytosolic Ca 2+ , which was measured in primary cultured neurons of the Kum 170 /+ allele [7]. The elevation of basal Ca 2+ observed is most likely due to a Table 3. List of individual genes whose expression level is restored significantly in mutants by both rescue and suppression. reduced rate of Ca 2+ uptake from the cytosol to the ER store. In itpr mutant animals with one copy of the Kum 170 allele (Kum 170 /+; itpr sv35/ug3 ), a significant increase in larval viability was observed as compared with itpr mutants on their own ( Figure 1A). The viable larvae grew to a size comparable to wild-type and a few pupated and eclosed as adults ( Figure 1B). In contrast to Kum 170 , there was no significant suppression of growth and viability of itpr sv35/ug3 animals by the allele dOrai 2/+ either on its own or in combination with Kum 170/+ ( Figure 1A and B). The dOrai 2/+ allele is a gain-offunction allele for the Drosophila store-operated Ca 2+ channel and suppresses itpr mutant phenotypes that arise during pupal development [7]. Since the Kum 170 allele was a more robust suppressor of growth and viability in itpr mutant larvae it was used in subsequent experiments.

Transcriptional profiling of itpr sv35/ug3
To understand the molecular changes underlying growth defects in itpr sv35/ug3 and their suppression a series of microarray experiments were performed. Initially, we obtained the transcriptional profile of RNA isolated from whole itpr sv35/ug3 larvae, just before growth defects are evident (at 60 hrs after egg laying or AEL), and compared this with larvae of the wild-type Canton-S strain, in which genetic background the itpr mutants were originally generated. This experiment was repeated with three biological replicates and genes that were up and down-regulated (log 2 fold change$1; P,0.05) in itpr sv35/ug3 , represented as red dots, are shown in the volcano plot in Figure 2A. Yellow dots represent genes that were not altered significantly by the above criteria. Amongst the 1504 independent genes that crossed this stringency threshold, 768 were up-regulated (right) and 736 were down-regulated (left). A complete list of these genes with individual hybridization intensity values and fold changes is given in the supplementary data (Table S1).
To identify specific pathways affected by reduced InsP 3 receptor function genes with significantly altered transcription profiles were classified into established Gene Ontology (GO) categories. Groups of genes under a similar category whose numbers were significant (P,0.05) are depicted in the bar graphs in Figure 2B and C. Categories with higher significance (P,0.01) are marked with asterisks. In the mutant UP state 358 genes could be classified in various GO categories for biological processes (Table S2, BP), while in the mutant DOWN state 369 genes could be classified similarly (Table S2). From this analysis itpr sv35/ug3 larvae show the greatest perturbation in genes affecting carbohydrate and amine metabolism. Genes affecting these two biological processes are . The Y-axis represents the log 2 of fold changes which were calculated by the DDC t method in which the C t values of each gene were normalized to the level of a housekeeping gene (rp49) in control RNA from CS larvae. Each value is the mean 6 SEM of three independent experiments, obtained from three independent RNA samples. The rescued and suppressed values were tested for significant difference from the mutants by Students t-tests followed by a Bonferroni correction for multiple tests. Except for CG2650 and unc119 all other genes were significantly altered (P,0.01). doi:10.1371/journal.pone.0024105.g004 both up-regulated and down-regulated and hence were analyzed further by placing them in metabolic pathways using the KEGG database. Down-regulated genes that classified under carbohydrate metabolism included Aconitase (TCA cycle), Enolase (Glycolysis), Zwischenferment (Pentose phosphate pathway) and several genes in galactose metabolism, indicating a reduction in the energy flux in itpr mutants. Up-regulated genes in carbohydrate metabolism were primarily in starch and sucrose pathways (e.g. amylase, maltase) and in amino sugar metabolism (chitinase). The chitinases overlap with genes up-regulated in amine metabolism. This group also includes enzymes that fall in pathways of tryptophan, arginine and proline metabolism. Amine metabolism genes that are down-regulated include several t-RNA synthetases and two enzymes affecting the one carbon pool by folate, another pathway that feeds into energy metabolism. Amongst other classes of up-regulated genes, of interest are a set affecting circadian rhythms, which include clock, cycle, cry, dopamine transporter (DAT), serotonin receptor 1A (5HT1A) and norpA. Some of these fall under other GO categories listed. For example clock and cycle are part of the list under positive regulation of transcriptional regulation, while norpA (encoding phospholipase Cb) also comes under phospholipid metabolism. Genes encoding proteins regulating secretion and exocytosis such as Synoptagmin 1, Synoptagmin 7, Synapsin, and Calcium activated protein for secretion (Caps) are also upregulated. In the set of genes significantly down-regulated under the translation category are several ribosomal proteins including mitochondrial ribosomal proteins. As expected from earlier findings genes belonging to the molting cycle are also downregulated [16]. Next, genes differentially regulated in itpr mutant larvae were re-analyzed with GO terms for cellular compartments (CC) and molecular functions (MF) (Fig 2C, Table S2). The largest number of down-regulated genes localized to mitochondria supporting an early effect of reduced InsP 3 receptor function on mitochondrial function and energy metabolism.

Biological classification of differentially regulated genes in itpr mutants by transcriptional profiling after suppression of growth defects
To identify pathways and genes likely to impact growth and viability most significantly, we measured whole genome transcript levels from RNA of itpr mutant larvae with a suppressor allele (Kum 170 /+; itpr sv35/ug3 ) and compared them to transcript levels in the itpr mutant (itpr sv35/ug3 ; Figure 3). To make this differential analysis more rigorous in parallel we carried out a comparison of RNA obtained from itpr mutant larvae rescued by expression of a UASitpr + transgene in Drosophila insulin-like peptide producing cells with Dilp2GAL4. These larvae exhibit normal growth similar to Kum 170 /+; itpr sv35/ug3 animals [4]. Both rescue and suppressed conditions restored expression levels of a significant number of genes altered in itpr sv35/ug3 (Figure 3; log2 fold change $1 and P,0.05). The heat maps show differential expression levels of genes in the itpr mutant when compared to wild-type (first column in Figure 3A and C). This was followed by comparison of expression levels in itpr sv35/ug3 with either the Dilp2GAL4 rescue (second column) or suppression by Kum 170 (third column) conditions. A complete list of these genes with individual hybridization intensity values and fold changes is available in the supplementary information for Kum 170 (Table S3) and for Dilp2GAL4 rescue (Table S4). The numbers of genes whose expression levels returned towards the wild-type in the rescue or suppressed condition are depicted in the Venn diagrams (Fig 3B  and D). A complete list of these genes with individual hybridization intensity values and fold changes is available in the supplementary information (Table S5).
To understand which biological processes are responsive to rescue/suppression transcriptionally altered genes in these conditions were classified into functional categories. As shown in Figure 3E the suppressed condition significantly (P,0.05) upregulated expression level of genes for lipid and pyruvate metabolism, defense and stress responses. In the rescue condition different biological classes were up-regulated of which only defense response is common with the suppressed condition. Similarly for the down-regulated genes in suppressed and rescued conditions very few common processes were discovered. A similar analysis for GO categories that were suppressed and rescued both under cellular components and molecular functions also gave very few common components or functions ( Figure S1). The complete lists of the genes in these GO categories are listed in Table S6 and S7. Taken together these observations suggest that suppression and rescue effect different processes to restore ''normal'' function.
Next we identified genes that were further up or down-regulated in the suppressed and rescued conditions. Among the up-regulated genes in mutants, 153 in the rescued condition and 85 in the suppressed condition were further up-regulated (Table S5). A total of 69 genes were up-regulated further in BOTH rescue and suppressed conditions (Table S5). The number of down-regulated genes in mutants that were further down-regulated in either suppressor or rescue was considerably fewer (Table S5). In this class there were 14 genes in rescue and 3 in the suppressor. None of these were common between the two conditions. A possible interpretation of this observation is that transcriptional upregulation in itpr mutants is a compensatory mechanism which is further enhanced for some genes in the rescue and suppressed conditions. Genes that are down-regulated may be causative. Taken together these analyses suggest that rescued and suppressed conditions do not necessarily help the animal revert to its ''normal'' wild-type state.
The analyses so far helped to identify pathways that are differentially regulated in the various conditions tested. Next, we identified individual genes that may or may not cluster significantly in a GO pathway but whose expression is restored towards wild-type levels by a log 2 fold change of 1 or more and P,0.05 in both rescue and suppressed conditions. Of the 15 genes identified in the UP mutant and DOWN rescue and suppressed states, 5 are predicted to effect oxidation-reduction processes and metabolism. There is one transcription factor (ribbon) known to effect development and up-regulated in larvae fed on a protein deficient diet of sugar only [17]; an alkaline phosphatase encoding gene (CG8147) which has earlier been identified as regulated by starvation and circadian rhythms [18,19]; a circadian clock controlled gene CG2650, which is also regulated by altered mitochondrial function [20]. In the remaining genes, 5 were of unknown function, while the functions predicted for others are microtubule based movement (CG10859) and proteolysis with a possible role in immune function (CG18180; Table 3). CG1659, the Drosophila homolog of the C.elegans gene unc119 has no known function but is highly expressed in the embryonic, larval and adult nervous system [10]. Among the genes of unknown function three (CG18180, CG17974 and CG12934) are up-regulated when larvae are nutrient deprived [17]; in addition CG12934 is up-regulated under starvation, while CG18180 is up-regulated by altered mitochondrial function [20]. Genes that were down-regulated in the mutants and whose expression was restored in the rescue and suppressed condition are shown in Table 3. Of the 16 genes identified in the DOWN mutant and UP rescue and suppressed states, two (CG32601 and pangolin) are transcriptional regulators of which pangolin is a repressor of Wnt signaling [21], l(2)dtl has homology with a ubiquitin ligase binding protein and is predicted to function in cell cycle regulation [22,23] and insulin like peptide 5 is one of the Drosophila insulin peptides that function in metabolic control of energy utilization [24]. Two genes code for serine-type endopeptidases (CG33459 and CG12385). Of these CG33459 is upregulated by JAK/STAT activation [25] and CG12385 is downregulated by starvation and stress [17,26]. From the predicted functions of three other genes (CG3752, CG13325, CG8525) it is likely that these affect different aspects of metabolism. Previous microarray data from related conditions has shown that they are down-regulated as follows; CG3725 under starvation [17], CG13325 in larvae fed on sucrose only [17] and CG8525 in an insulin signaling mutant (dFOXO; 26). Seven genes in this group are of unknown function. Among these CG8317 is also down-regulated in sucrose fed larvae, though it is up-regulated under starvation [17].

Validation of selected target genes identified by the microarray screens and their normal tissue specific expression
Quantitative real-time PCR (qPCR; Figure 4) was carried out to validate the altered expression level of selected genes that underwent significant changes in the mutant. In these experiments we tested candidates that were differentially regulated in rescued and suppressed conditions, as well as other genes that were of potential interest amongst the genes identified in the mutant condition only. Unlike the microarray where rescue and suppressed conditions were compared with mutant, here we quantified the expression levels of all conditions with RNA extracted from wild-type. Amongst the up-regulated genes we validated six genes in mutant, rescue and suppressor conditions, each of which was a candidate from a different biological process. In the rescue and suppressed conditions the expression level of two genes (CG2650 and CG1659) was down-regulated when compared with the mutant, but remained high when compared with RNA from wild-type. The remaining four up-regulated genes validated in mutants were all down-regulated compared to wild-type. These included Cyp12d1-d (CG33503) and DAT (dopamine transporter with a neuronal function; CG8380; Fumin). In the microarray, DAT is regulated similarly but was filtered out due its higher P value. Amongst genes that were down-regulated in the mutant condition we tested l(2)dtl (CG11295), Idgf4 (CG1780), pangolin (CG34403) and CG3752, all of which were validated as down-regulated in mutant and up-regulated in suppressor and rescue conditions by qPCR ( Figure 4). Due to the filtering cut offs Idgf4 does not appear in Table 3. To begin understanding the functional significance of these transcriptional changes the expression level of a subset of genes was also assessed in the brain and fat bodies of third instar larvae ( Figure 5). Both these tissues have a central role in the control of energy metabolism in Drosophila and are known to express itpr transcripts at moderate levels [8,10]. The genes tested were l(2)dtl and DAT which were found to be brain enriched, cyp12d and ng2 which were enriched in the fat body and Idgf3 and Idgf4 which were expressed in both tissues. ng2 and Idgf3 were included since both are down-regulated significantly in mutants and up-regulated in the rescue condition. They were filtered out from the genes in Table 3 since they do not show a significant change in the suppressor.

Discussion
Signaling through the InsP 3 receptor and intracellular Ca 2+ release are thought to affect multiple physiological conditions in mammals [27]. However, the underlying molecular mechanisms regulated or controlled by InsP 3 -mediated Ca 2+ release in the context of specific physiological conditions still need elucidation. The genome-wide microarray analysis reported here with existing well-characterized InsP 3 R mutants has allowed us to investigate this question in the context of growth. From the gene ontology and pathway analyses of gene expression changes in the strong itpr mutant studied here, it is clear that metabolic genes particularly those related to carbohydrate and amine metabolism are significantly altered. The underlying cause(s) of these metabolic changes is of interest. One possibility, partially supported by our data as well as recent evidence from other groups [28] is a change in mitochondrial bioenergetics leading to subsequent effects on metabolic pathways. In the itpr mutant condition 50 downregulated genes cluster in the mitochondrion (Fig 2; CC) and a highly significant number (P = 3.2E25) have the molecular function (MF) of oxido-reduction. A set of genes classified as oxidoreductases are also up-regulated. However, these are primarily for detoxification and xenobiotic responses and are presumably up-regulated as a compensatory stress mechanism.
Microarray studies in Drosophila with nutritionally altered, stress and mitochondrial mutant (tko) conditions have been published [17,20,26]. Therefore, we compared the list of up-regulated and down-regulated genes in our study with published gene lists for microarrays in what appear to be related conditions. In this analysis the maximum overlap obtained was between genes upregulated in itpr sv35/ug3 larvae with larvae undergoing starvation, followed by larvae grown on sugar alone (i.e. protein deficient), followed by the mitochondrial mutant tko ( Figure 6A). The correlation with starvation conditions is not surprising since one of the earliest phenotypes of itpr sv35/ug3 larvae is reduced feeding [4]. The change in circadian rhythm genes observed in the mutant condition may also be related to the feeding changes since circadian cycles are thought to integrate feeding and metabolic functions [29]. Even so the response of itpr sv35/ug3 larvae does not correlate directly with either starvation or protein deficiency conditions or mitochondrial dysfunction, indicating that the transcriptional response to reduced intracellular Ca 2+ release is complex ( Figure 6B). Interestingly, the response of groups of genes that overlap with starvation, protein deficient and mitochondrial mutant conditions revert in the rescued and suppressed conditions ( Figure 7). As suggested from Table S7, however, the extent of reversion differs in the two conditions, supporting the idea that the physiological states of rescued and suppressed larvae are different. In the rescued condition the itpr cDNA is expressed in a limited set of neuronal cells, while in the suppressed condition the dominant mutant for Ca-P60A Kum170ts affects the whole organism. Since the rescuing neurons are known to secrete several insulin-like peptides, which regulate cellular physiology through a pathway that acts via the transcription factor dFOXO, we also compared our data with microarray data from a Foxo mutant combination [30]. We do not observe a correlation with dFOXO regulated genes. Thus the transcriptional changes observed in itpr sv35/ug3 must involve other gene regulatory factor(s) that are either directly responsive to changes in intracellular Ca 2+ or the altered metabolic state of the organism. We analyzed our gene lists for possible candidates and identified ribbon (CG7230), pangolin (CG34403) and l(2)dtl (CG11295) as potential regulators of gene expression. None of these molecules appear to have a motif that would make them directly responsive to Ca 2+ changes, suggesting that their regulation is indirect. The precise mechanisms of their regulation in itpr mutants need further study. Interestingly, ribbon is upregulated in protein-deficient larvae, suggesting that in addition to its established role in regulation of development it could be a metabolic regulator [17]. No data are available at this stage to support a role for pangolin and l(2)dtl in the regulation of energy metabolism. Since pangolin is best known as a negative regulator of wingless signaling, we searched among the genes regulated in itpr sv35/ug3 mutant for other components of the wingless pathway (Table S1). wnt5 was up-regulated by a log2 fold change of 0.6 to 0.9 and marginally down-regulated in both suppressor and rescue conditions. In vertebrates, Wnt5 has been implicated in InsP 3 R activation during development [31]. l(2)dtl mutants in Drosophila are lethal as embryos and l(2)dtl transcripts are up-regulated under heat-shock [32]. The molecular function of l(2)dtl has not been investigated in Drosophila. In mouse the l(2)dtl homolog is known as CTD2. Several independent studies have shown that it plays an important role in cell proliferation through regulation of a G2/M checkpoint [22]. A functional analysis of these genes needs to be carried out in Drosophila itpr mutants. This will help validate their potential regulation by InsP 3 mediated intracellular Ca 2+ signaling and also in understanding how developmental pathways impact growth and metabolism. Figure S1 Gene Ontology classification for cellular components and molecular function (GO, CC and MF) of up-regulated and down-regulated genes in suppressed and rescued larvae. The X-axis represents the number of genes in the suppressed and rescued condition in the marked category. Right panels indicate up-regulated genes and left panels indicate down-regulated genes. Number of genes in the categories shown had a P value#0.05. * indicates functional categories with P,0.01. Complete lists of genes for each category are in Tables  S6 and S7.  (TIF)   Table S1 A list of significantly up-regulated and downregulated genes (log 2 fold change .1; P,0.05) in mutant larvae (itpr sv35/ug3 ).  Table S3 A list of up-regulated and down-regulated genes with log 2 fold change $1 and P,0.05 in suppressed larvae (Kum 170/+ ; itpr sv35/ug3 ).

(XLS)
Table S5 A list of filtered (P,0.05) up-regulated and down-regulated genes that are common between mutant, rescued and suppressed larvae with fold changes and GO terms.