Abstract

Type 2 diabetes (T2D) is one of the most common human diseases. QTL analysis of the diabetic Otsuka Long-Evans Tokushima Fatty (OLETF) rats has identified numerous hyperglycemic loci. However, molecular characterization and/or gene identification largely remains to be elucidated due mostly to the weak genetic variances contributed by each locus. Here we utilized Drosophila melanogaster as a secondary model organism for functional evaluation of the candidate gene. We demonstrate that the tissue specific knockdown of a homologue of igf2bp2 RNA binding protein leads to increased sugar levels similar to that found in the OLETF rat. In the mutant, the expression of two of the insulin-like peptides encoded in the fly genome, dilp2 and dilp3, were found to be downregulated. Consistent with previous reports of dilp mutants, the imp mutant flies exhibited an extension of life span; in contrast, starvation tolerance was reduced. These results further reinforce the possibility that imp is involved in sugar metabolism by modulating insulin expression.

1. Introduction

The world health organization (WHO) currently estimates that over 300 million individuals worldwide suffer from diabetes, 90% being type 2 diabetes (T2D) [1]. T2D, the primary feature of which is a state of chronic elevation of plasma glucose levels, is a polygenic disease that is caused by a metabolic and hormonal imbalance between insulin secretion from pancreatic β-cells and insulin resistance in peripheral tissues. Much effort has been devoted to the development and characterization of monogenic diabetes animal models, which have led to significant advancements in our understanding of the genetic basis of glucose/lipid metabolisms as well as the molecular pathogenesis of complications [2]. In spite of the progress, the importance of polygenic or spontaneous diabetes models is not diminished because the majority of genetic variations that are causative for a complex disease are not amorphic, but hypomorphic [35]. However, the importance of spontaneous diabetes models has been relatively underestimated owing to the difficulty of positional cloning [6].

The OLETF rat is one of the most studied strains by virtue of its similarity to a particular human population marked by propensity for disorders of glucose metabolism [7]. Traditional genetic analysis in the OLETF has been based on mapping QTL using microsatellite markers, followed by genetic isolation of QTL in congenic strains [8]. Recently there have been several studies reporting the successful positional cloning of QTLs by further extensive fine mapping of congenic strains [3]. However, in these cases the LOD scores of the QTL are relatively high (above 6.0) and consistently the identified mutations led to a more than twofold increase or reduction in expression levels [3, 4]. Thus, the search for genetic factors for polygenic traits remains to be a formidable challenge, especially for those whose LOD scores are not very high. Ideally rodent models should be used for functional probing of the candidate genes, yet the screening of a large number of genes is considered to be intractable with current techniques.

Drosophila melanogaster has progressively been recognized as the most feasible nonmammalian model for metabolic diseases [9, 10]. The Drosophila genome encodes eight insulin-like peptides and the backbone of the insulin/IGF-like signaling (IIS) pathway is highly conserved in comparison to that of vertebrates. Furthermore, physiological roles of the IIS pathway, including growth, lifespan, stress resistance, and metabolism, are also analogous across animal kingdom, making Drosophila a potential alternative agent for functional evaluation of the genes whose candidacy is suggested in other systems. The use of Drosophila as an evaluation method will be useful at least for those genes whose orthologs are encoded by the fly genome. One of the hyperglycemic QTLs identified in our previous studies is intriguing in terms of its association with obesity (see Section 4) and it is worth further investigation.

In the current study, prior to full-scale screening, we chose to focus on another QTL, Niddm22 (Nidd4/of as our original nomenclature), because of the presence of a strong candidate gene. Niddm22 is a region of 35.4 cM, corresponding to a physical distance of 24 Mbp, on the rat chromosome 11 [8]. Several human linkage studies reported metabolic QTLs in its syntenic region [11, 12]. According to the Ensemble database (release 73), 161 genes are annotated in this rat chromosomal segment [13]. Among those, 80 genes have fly orthologues. Here we focused on imp, a homolog of vertebrate igf2bp2, because the association studies identified an SNP within the locus to be linked not only to the diabetic phenotype but also to other diabetes-related traits such as fasting glucose, glucose AUC (area under curve), and Cederholm index [1416].

2. Materials and Methods

2.1. Fly Stocks

All fly stocks were reared at 25°C on a standard yeast (4%, w/v), corn meal (8%, w/v), glucose (10%, w/v), and agar medium, under 12 h:12 h light:dark conditions unless otherwise stated. The following fly stocks were used: UAS-imp-RNAi (v20321) from the Vienna Drosophila RNAi Center, Imp protein trap strain (number 110921) from the Drosophila Genetic Resource Center at Kyoto Institute of Technology. Additional fly stocks were generously provided by the Drosophila community: elav-GAL4 and UAS-Dcr2 from Yasushi Hiromi [1719], dilp2-Gal4 from Takashi Nishimura [20, 21]. Using standard fly genetics, UAS-Dcr2 and UAS-imp-RNAi were intercrossed into one strain in order to enhance the effect of RNA interference. Here UAS-imp-RNAi; UAS-Dcr-2 is referred to as UAS-imp.

2.2. Metabolic Studies

Whole-fly or hemolymph trehalose was measured by a Trehalose Assay Kit (Megazyme, K-TREH). For whole-fly preparation, 10 larvae were collected and briefly rinsed in Ringer’s solution. The larvae were homogenized by vigorous shaking in the presence of Zirconia beads (NIKKATO, 0.8 YTZ Ball). The resultant homogenate was heated at 70°C for 5 min and centrifuged at 12000 rpm for 5 min and the resultant supernatant was used for subsequent measurements. Hemolymph was prepared as previously described [22]. Briefly, 10 third instar larvae were pricked with a tungsten needle and transferred to a microfuge tube which had been pierced in the bottom, which was then piggybacked and centrifuged for 5 min at 4°C, 7,000 rpm. The resultant supernatant or hemolymph was used for subsequent measurements. Protein quantity was determined by Quant-iT Protein Assay Kits (Invitrogen).

2.3. Lifespan Assay

Lifespan studies were performed as previously mentioned with modifications [23]. For both fed and starved samples, three to ten virgin males and virgin females with approximately 1 : 1 ratio were placed in a single plastic vial. For starvation, the vial contained a piece of filter paper moisturized with distilled water. Flies were transferred to fresh medium or moisture vials every four to five days, and deaths were scored three times per week. The number of live individuals was recorded until all flies died.

2.4. q-PCR

Total RNA was extracted from 25 whole larvae in TRIzol reagent (Invitrogen). One microgram of total RNA was used for reverse transcription with iScript Select cDNA Synthesis Kit (Bio-Rad) by using oligo(dT) primer. q-PCR was performed on a MiniOpticon real-time PCR System (Bio-Rad) using iQ SYBR Green Supermix (Bio-Rad). Primers used for Q-RT-PCR are summarized in Table S1 available online at http://dx.doi.org/10.1155/2015/758564 [24, 25].

2.5. Statistical Analyses

For all experiments, error bars represent SEM, and values are the results of ANOVA followed by post hoc analyses using Scheffe’s test.

2.6. Microscopy

Fluorescent and bright field images were taken using an Axio 200 microscope (Zeiss).

3. Results

3.1. CNS Specific imp Knockdown Resulted in Hypertrehalosemia

Previous studies showed that imp is expressed in the central nervous system and pole cells during embryonic development and germ cells in adults [2628]. In order to examine the expression pattern of imp in larvae, we analyzed a protein trap strain, ZCL0310 [28]. In the third instar wandering larvae, the expression was exclusively detected in the central nervous system (CNS) (Figure 1). In contrast, imp is not expressed in other metabolically crucial tissues, including body wall muscle, fat body, gut, and oenocytes (Figure 1). Next we produced an CNS specific imp knockdown strain. The mutant had normal hatching rate and developmental growth. No morphological defect was observed. We confirmed that, in the third instar larvae, imp expression was reduced to about 20% of that of control (Figure 2(a)). Hemolymph was extracted from the third instar larvae that were immersed in the food medium (fed state). We also tested hemolymph from larvae starved for 15 hours (starved state). In both cases, the trehalose levels were significantly higher for the imp mutant compared with either control strain (Figure 2(c)). In contrast, no difference was observed among these strains for protein levels in either fed or starved condition. In order to examine the effect of imp knockdown mutation on total trehalose levels, whole-fly trehalose that is normalized by total protein levels was compared. In the starved state, the total amount of trehalose was higher than the control (Figure 2(b)). Because in our QTL analysis Niddm22 locus was identified as fasting hyperglycemic QTL, the observation of a more prominent effect on the starved state implies gene candidacy.

3.2. dilp Expression Is Downregulated in imp Knockdown Mutant

We examined the expression of a subset of dilp genes that are crucial for carbohydrate metabolism [29, 30]. The expression levels of dilp2 and dilp3, but not dilp5, were significantly reduced in the imp mutant larva (Figure 3). Imp belongs to a family of mRNA-binding proteins that play an important role in RNA localization, stability, and translation. RNA binding is mediated by highly conserved KH domains [31]. One of the most characterized KH domains, KH3 of Nova, recognizes a single UCAY element in the context of a 20-base hairpin RNA [32]. We found 2, 6, and 2 consensus motifs in dilp2, dilp3, and dilp5 mRNA, respectively (Table S2), leading us to the hypothesis that Imp may posttranscriptionally control the translation of dilps by direct binding. To test this, we knocked down the function of imp only in insulin-producing cells (IPCs) in which the three dilp isoforms are most exclusively expressed. However, the levels of dilps (dilp2, dilp3, and dilp5) or imp were unchanged and no hypertrehalosemia was observed (Figure S1). Furthermore, immunostaining revealed no apparent imp expression in IPCs (Figure S2). All of these results suggest that imp influences subtypes of dilp expression in a cell-non-autonomous manner.

3.3. imp Knockdown Resulted in Longer Lifespan and Reduced Starvation Tolerance

There are numerous reports that link the IIS pathway to lifespan or aging [24]. Our results so far suggest that the IIS signaling activity may be chronically lowered in the imp mutant. Consistently the imp knockdown mutant exhibited a significant increase in average and maximum lifespan over that of control flies (Figure 4(a)). Previous studies reported that the IPC-ablated flies were slightly starvation resistant [24, 33]. The extended lifespan is usually considered to be the result of enhanced stress resistance. However, the longevity on starvation of imp mutants is significantly lower than that of the control strains (Figure 4(b)), suggesting that imp may be involved in stress response regulation independent of dilp activity.

It is also widely known that IIS signaling pathway plays an essential role in the control of cell size and growth [30, 34]. The loss of dilp2, but not dilp3 or dilp5, reduces body weight [30]. Body size and weight of the imp knockdown strain are unchanged (Figure S3). This may be because the remaining dilp2 expression is sufficient to maintain normal growth, or other members of dilps that control growth compensate for the effect of dilp2 downregulation.

3.4. Polymorphisms of igf2bp2 Locus and Expression Analysis

Next we sequenced the coding region of igf2bp2, a rat orthologue of imp for the OLETF and F344 rat. There is one SNP in the fourth exon; however this SNP is a synonymous substitution (Table S3). Furthermore, in the tissue examined, our q-PCR analysis failed to detect any difference in the expression levels between the two strains. Further studies will be necessary for establishing the causality of the igf2bp2 in the OLETF rat.

4. Discussion

In the present study we demonstrated that the examination of the homologous gene provides us with a unique opportunity to search for novel metabolic genes. We tested Niddm22 here partly because our aim was to establish the methodology; we now wish to tackle other novel QTLs.

Previously we demonstrated that one of the hyperglycemic QTLs, Niddm20 (Nidd2/of as our original nomenclature), located on chromosome 14 of the OLETF rat, is quite unique for the following reasons: (1) its LOD score (4.07 for 30 min postprandial plasma glucose) is one of the highest among the other QTLs [35, 36]; (2) there is a strong epistasis with other QTLs [37]; (3) most importantly it interacts with the obese condition: the congenic strain exhibits more severe diabetic symptoms when combined with either genetically or nutritionally induced obesity [38, 39]. From the clinical point of view, identification of causative genes in such QTL has to be given higher priority. We further fine-mapped the region to discover that Niddm20 is composed of at least two narrower loci, each of which is localized at proximal and distal ends of the QTL region [40]. The analysis of subcongenic strains of Niddm20 showed that the exclusion of either locus from the original Niddm20 region resulted in the loss of the hyperglycemic phenotype, suggesting an epistatic relationship between the subloci. Within the syntenic region of the human genome, neither diabetic QTL has been reported nor have any of the T2D susceptibility genes been mapped [41]. Therefore, elucidation of the molecular nature of Niddm20 may provide novel opportunity for understanding human T2D. According to the Ensemble database, there are 62 genes annotated within the proximal 10Mb of Niddm20, none of which has been implicated with T2D. Our aim is to utilize Drosophila for functional evaluation of those candidate genes. A similar attempt was recently reported elsewhere [42] and it is hoped that Drosophila as a secondary model will help to find novel diabetic genes.

igf2bp2 has been implicated by genome-wide association studies as a candidate susceptibility gene for T2D [14, 15]. Several association studies correlated igf2bp2-SNPs more with reduced pancreatic β-cell activity than insulin resistance [43, 44]. However, the SNP is found in the second intron and the mechanism by which this susceptibility is engendered is unknown. Dai et al. reported that igf2bp2 mRNA is promoted by phosphorylation of Igf2bp2 by mTOR [45]. Another study indicated that Igf2bp2 directly binds to laminin-β2 mRNA and regulates its translation in a glucose concentration-dependent manner in the podocyte [46]. Several Drosophila studies also investigated the role of imp in the context of mRNA translocalization as well as translational regulation [26, 27, 47]. In the tests imp plays a crucial role in the aging of germ line stem cells (GSC), implicating a possible connection with the extended lifespan observed in the imp knockdown flies [48].

Among the genes annotated in Niddm22, there are other candidate genes that are inferred to be involved in metabolic functions, including somatostatin [49], Ahsg [50], and Adipoq [51]. Adipoq is the only other gene that has been identified as a diabetes candidate gene by GWAS. Recently it was reported that an adiponectin receptor homologue is involved in carbohydrate metabolism in the fly; however its orthologous ligand is not encoded in the fly genome and the authentic ligand remains to be discovered [52].

In general Drosophila offers a convenient resource for providing a rapid, inexpensive in vivo test of gene function. In addition fly genetics could also be useful for understanding molecular mechanisms. Indeed some insulin pathway components have been identified or validated by Drosophila research [53]. Reduced dilp level in the imp mutant is consistent at least partially with mammalian studies. It is, however, important to notice that in this system the following: (1) genes can only be characterized for which there are functional homologues in fly and (2) findings of diabetes-like phenotypes may not be valid for vertebrates or humans. For example, even though many insulin signaling components are conserved between flies and mammals, there are as many as eight insulin-like genes in Drosophila and they are expressed in tissues of various developmental origins, such as dilp2, dilp3, and dilp5 in neurons or dilp6 in the fat body [54]. With that in mind, it is hoped that genetic screening using this strain as a platform might elucidate details of the molecular pathway.

5. Conclusions

In summary, we performed a functional analysis on one of the diabetic candidate genes derived from the OLETF rat. We showed that downregulation of imp led to a hypertrehalosemic condition in Drosophila. Although further studies will be necessary to confirm the causative relationship between imp and diabetes in the OLETF rat, our results indicate that Drosophila is a useful secondary model for examination of the mammalian diabetes model.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

This study was supported in part by a grant from the National Bio Resource Project (NBRP) for Rat in Japan (Kozo Matsumoto) and JSPS KAKENHI Grant number 20300145 (Kozo Matsumoto, Hiroyuki Kose). The authors thank Y. Hiromi, T. Nishimura, the Bloomington Stock Center, Drosophila Genetic Resource Center in Kyoto, and the Vienna Drosophila RNAi Center for fly stocks.

Supplementary Materials

List of primers that were used for q-PCR analysis. Primers for gapdh1 and imp were designed by Primer3 http://bioinfo.ut.ee/primer3-0.4.0/primer3/.

  1. Supplementary Materials