Journal of Genetics (2022)101:15 Ó Indian Academy of Sciences

https://doi.org/10.1007/s12041-021-01356-5 (0123456789().,-volV)
(0123456789().,-volV)

RESEARCH ARTICLE

Examination of Niddm20 candidate genes of OLETF rats in

Drosophila melanogaster using inducible GeneSwitch GAL4 system

LAN ENLI1, YUTO MORONUKI1,3, TAKAHISA YAMADA2 and HIROYUKI KOSE1*

1Department of Natural Sciences, International Christian University, Mitaka, Tokyo 181-8585, Japan
2Faculty of Agriculture, Department of Agrobiology, Niigata University, Nishi-ku, Niigata 950-2181, Japan
3Present address: Laboratory of Bio-Resource Regulation, Department of Integrated BioSciences, Graduate School of

Frontier Sciences, University of Tokyo, Kashiwa, Chiba 277-0882, Japan

*For correspondence. E-mail: kose@icu.ac.jp.

Received 6 July 2021; revised 24 November 2021; accepted 25 November 2021

Abstract. Type 2 diabetes mellitus (T2D) is a metabolic disorder caused by a complex interplay between genetic and environmental
factors. While remarkable progress have been made in our understanding of the genetic components that lead to disease expression, the
‘diabetes genes’ identified to date are inadequate for assessing disease risk, which suggests that many other genes remain to be discovered.
Here, we used Drosophila to examine the protein-coding genes annotated in the hyperglycaemic locus of Otsuka Long-Evans Tokushima
Fatty (OLETF) rats, a well-characterized model of T2D. We identified lilliputian (lilli), a fly homolog of AF4/FMR2 family member 2, as a
novel candidate gene for T2D. Lilli knockdown adult females had significantly higher haemolymph trehalose and glucose levels, while
dilp2, which plays a major role in sugar metabolism, was downregulated. Tissue-specific knockdown strain revealed that lilli plays a crucial
function in oenocytes and the fat body, which together are homologous organs to the liver. Together, these findings demonstrate the
importance of using fly models for investigating polygenic diseases such as T2D.

Keywords. type 2 diabetes; lilli; AF4/FMR2; insulin/IGF signalling; Drosophila.

Introduction obesity loci, and these results were subsequently con-

firmed by construction of congenic strains (Kose et al.
Type 2 diabetes (T2D) is one of the most pressing global 2002). However, the identification of causative genes or
health issues. It is a polygenic disease caused by multiple polymorphism remains limited.
genetic and environmental factors (Kahn et al. 2006; Drosophila melanogaster has been extensively utilized in
Roden and Shulman 2019). These two kinds of factors the areas of physiology and metabolism. The fly genome
interact with each other in complex ways, making it dif- encodes most of the insulin / insulin-like growth factor (IGF)
ficult to understand disease expression (O’Rahilly 2009; signalling (IIS) pathway components, and a functional sim-
Langenberg and Lotta 2018). Despite many genomewide ilarity to that of mammals has been demonstrated (Alfa and
association studies having been conducted, a comprehen- Kim 2016). Genomewide screening of obesity genes was
sive picture of the genetic architecture of the disease has conducted with flies (Pospisilik et al. 2010). Functional
yet to be elucidated (Stančáková and Markku 2016). evaluations of T2D candidate genes implicated in human
Originally established in 1980, the Otsuka Long-Evans and rat studies have also been reported (Pendse et al. 2013;
Tokushima fatty (OLETF) rat is a well-characterized Kawasaki et al. 2015). This work has established that Dro-
model animal for T2D (Kawano et al. 1994) because sophila provides a unique opportunity to investigate the
(i) T2D is polygenic (Moralejo et al. 1998); (ii) the onset molecular events underlying complex traits such as haemo-
of the disease is obesity dependent (Ogino et al. 2000); lymph sugar, obesity, longevity, learning and mating
and (iii) the OLETF rat shows similar complications to behaviour.
human patients (Kawano et al. 1999). Quantitative trait In this study, we focussed on one of the hyperglycaemia
loci (QTL) analysis revealed 14 hyperglycaemia and six loci, Niddm20, which is on rat chromosome 14 and is
15 Page 2 of 8 Lan Enli et al.

demarcated by two microsatellite markers, D14Rat23 and Materials and methods

D14Rat12 (figure 1, Nidd2/of as our original nomenclature).
A subsequent evaluation using the congenic strain demon- Fly strains
strated that the locus does contain polymorphisms causative
of carbohydrate regulation (Kose et al. 2002). Intriguingly, Unless otherwise stated, all fly stocks were reared at 25°C on a
the Niddm20 congenic rat showed overt hyperglycaemia standard yeast, cornmeal and agar medium under 12 h : 12 h
when fed with a high-fat diet or when a leptin signalling light:dark conditions. We utilized the GeneSwitch (GS) system
deficiency was introduced (Kose et al. 2007). This suggests to temporally induce Gal4 transcription (Nicholson et al. 2008).
that the causative genes may provide a clue to the molecular The following Gal4 driver strains were used: Act5C-GS (BDSC
nature that links the obesity-causing stimuli and disease #9431) (Rogulja and Irvine 2005); elav-GS (BDSC #43642l);
expression. The Niddm20 locus was later found to possess FB-GS (BDSC #59939); OE-GS (BDSC #62089) (Nicholson
multiple causative genes which are localized at both ends of et al. 2008); w;P{w[?mW.hs]=GawB}lilli[NP2790]/CyO
the region. They were designated proximal Niddm20 and (DGRC #104326); UAS-lilli-RNAi (VDRC #106142); lilliA17-
2
distal Niddm20 (figure 1). It was demonstrated that neither /CyO (DGRC #108389) and UAS-EGFP clc3 (Gift from
the proximal or distal region alone was sufficient to achieve H. Tsujimura). All the RNAi strains were obtained from VDRC
statistically significant hyperglycaemia (Akhi et al. 2005). and NIG stock center (table 1).
Proximal Niddm20 is defined by D14Rat23 and D14Wox1,
while the distal Niddm20 is defined by D14Rat10 and
D14Rat12 (figure 1). In this study, we focussed on the RU486 administration
proximal sub-region for convenience since it has a fewer
annotated genes than the distal one. A stock solution of 10 mg/mL RU486 (Funakoshi) in
We identified lilli, the gene encoding for transcription ethanol was added to yeast paste (33% w/v) to yield a final
elongation factor as a novel candidate gene. Although lilli concentration of 5 lg/mL. Forty adult females of three to
has been reported to modulate insulin signalling pathways in five days old were collected. Prior to exposure to the
the context of cell growth, it has not been implicated in the RU486-containing yeast paste, flies were fasted for 3 h in a
context of sugar metabolism (Wittwer et al. 2001). This chamber with a wet tissue and subsequently transferred to a
study demonstrates the potential of using Drosophila to vial containing the RU486-containing yeast paste. They
unravel parts of the molecular network of diabetes that have were allowed to feed on the medium for 24 h before hae-
eluded analysis in the mammalian system. molymph extraction.

Haemolymph extraction, trehalose/glucose assay

Females pierced with a tungsten needle in the thorax were

placed in a small Eppendorf tube with a small hole at the
bottom. The tube was piggy-bagged on a large Eppendorf
tube and centrifuged at 5000 rpm for 5 min at 4°C. The
haemolymph was collected from the large tube with a
microcapillary tube (1 microlitre) and incubated at 70°C to
inactivate enzymes in the haemolymph. This was followed
immediately by chilling on ice. The haemolymph trehalose
and glucose were respectively measured by Trehalose Assay
kit (Megazyme, K-TREH) and LabAssay Glucose kit
(WAKO), according to the manufacturer’s protocols.

qPCR

Total RNA was extracted from the whole larvae body using
ISOGEN II (NIPPON GENE), and subsequently was reverse-
transcribed into cDNA using FastGeneTM Scriptase II Ready
Figure 1. Proximal Niddm20, Niddm20, the hyperglycaemic locus Mix (NIPPON GENE). qPCR was performed using AriaMx
of the OLETF rat. The locus is defined by microsatellite markers real-time PCR system (Agilent Technologies) with KAPA
D14Rat23 and D14Rat12 (Moralejo et al. 1998). The subsequent
analysis showed that the region consists of two subregions (shown
SYBR Fast qPCR Kit (NIPPON Genetics). All qRT-PCR
in gray), both contain hyperglycaemic genes. The proximal peak, products were normalized to dGAPDH gene expression. The
which is about 8 Mbp was investigated in this study. following gene-specific primers were used: dGAPDH:
 Examination of Niddm20 candidate genes of OLETF rats Page 3 of 8 15

Table 1. Protein-coding genes in the proximal Nidd2/of. and fat body were dissected from third instar larva and fixed for
20 min in a fixative solution (4% formaldehyde (polyscience),
Stock 0.1% Tween-20(WAKO) in PBS (150 mM NaCl, 1.4 mM
RDG symbol Fly gene CG centre Stock ID NaH2PO4, 8.7 mM Na2HPO4, pH 7.6)). The tissues were rinsed
Pcgf3 l(3)73Ah CG4195 NIG 4195R-1 thrice for 10 min each with PBST (0.1% Triton X-100 in PBS).
Pde6b Pde11 CG34341 NIG HMS01286 Blocking was performed in PBST containing 5% normal goat
Pigg PIG-G CG2144 VDRC 107722 serum for 1 h. The primary antibody reaction was allowed
Dr1-201 NC2beta CG4185 n/a overnight at 4°C. After washing with PBST thrice for 10 min
Tmed5 p24-1 CG1967 NIG 1967R-2 each, the tissues were incubated in the secondary antibody
Mtf2 Pcl CG5109 NIG 5109R-1
Dipk1a CG12038 CG12038 VDRC 107527 solution containing DAPI (WAKO, 1 lg/mL) overnight at 4°C.
Rpl5 RpL5 CG17489 VDRC 109986 The primary and secondary antibody were rabbit anti-GFP
Ube2d4 CG2924 CG2924 NIG 2924R-3 (BioAcademia) and Alexa-conjugated Goat anti-rabbit antibody
Gfi1 sens CG32120 n/a (Jackson ImmunoResearch), respectively. Images were captured
Glmn CG30496 CG30496 VDRC 110548 by FV3000 confocal microscope (Olympus).
Lpcat2b LPCAT CG32699 VDRC 104570
Ephx4 CG5704 CG5704 NIG 5704R-2
Brdt Gcn5 CG4107 n/a
Cdc7 Cdc7 CG32742 VDRC 104902 Statistics
Barhl2 B-H1 CG5529 VDRC 104681
Gbp6 atl CG6668 NIG HMS01627
Abcg3l3 CG31689 CG31689 NIG 31689R-1 Differences between the experimental and control conditions
LOC679894 tho2 CG31671 NIG 31671R-2 were analysed by Student t-test. A P value of \0.05 was
Pkd2 Pkd2 CG6504 VDRC 110681 considered to indicate statistical significance.
Sparcl1 SPARC CG6378 VDRC 100566
Aff1 lilli CG8817 VDRC 106142
AABR07014298.1 ND-B14 CG13240 NIG 2LG-0040
Mapk10 CG8565 n/a Results
Cds1 cds CG7962 n/a
Nkx6-1 NK7.1 CG8524 NIG HMJ22342
Mrps18c n/a CG31450 VDRC 103657 Examination of hyperglycaemic candidate genes in proximal
sub-region of Niddm20
List of genes annotated in the proximal Nidd2/of, which have
Drosophila orthologus. The region demarked by two microsatellite We first listed genes annotated between the two markers
markers spans ca. 8 Mbp; n/a, not applicable. (D14Rat23 and D14Wox1) on Ensembl Database (release 95)
and found 66 protein-coding genes, of which 28 have orthologs
(dGAPDH1.751-) ATCGTCGAGGGTCTGATGAC, reported in the Drosophila genome. We excluded RNA-coding
(dGAPDH1.-994) CGGACGGTAAGATCCACAAC; lilli: genes and pseudogenes. We tested those genes whose RNAi
(lilli 7262-) AAGAAAGGATTGTCGGGCCA, (lilli-7398) driver strains are available from the fly stock centre (table 1).
TATTCGAACCGGAGCCCTGC; dilp2: (forward) ATGAG- Each RNAi strain was crossed with the ubiquitous Gal4 strain,
CAAGCCTTTGTCCTTC, (reverse) GACCACGGAGCAG- but all of the F1 progeny was found to be lethal in larva or pupae
TACTCCC (S. J. Broughton et al. 2005); dilp3: (forward): stage (data not shown). Therefore, we utilized the GS system to
AGAGAACTTTGGACCCCGTGA, (reverse) TGAACC- temporarily downregulate the target genes after hatching to the
GAACTATCACTCAACAGTCT; dilp5: (forward) GAGG- adult stage. Intriguingly, we found variances in baseline tre-
CACCTTGGGCCTATTC, (reverse) halose levels (e.g., before gene-knockdown induction, one-way
CATGTGGTGAGATTCGGAGCTA (Broughton et al. 2005); ANOVA P\0.0001) among the strains tested, implying that the
InR: (forward) TGAGCATGTGGAGCACATCAAGATG, difference in the genetic backgrounds of the laboratory strains
(reverse) CGTAGGAGATTTTCTCGTTTGGCTG (Okamoto can make a significant difference in the metabolic function. We
et al. 2013); and 4E-BP: (forward) TGATCACCAG- found that downregulation of two of the genes, lilli and tho2,
GAAGGTTGTCATCTC, (reverse) GAGCCACGGA- significantly elevated haemolymph trehalose levels (figure 2).
GATTCTTCAGAAAG (Okamoto et al. 2013). Independent Since lilli showed higher statistical significance and is reported
PCR reactions for all samples were performed thrice or more. to function in the insulin signalling pathway (see below), we
The uniqueness of each amplicon was verified using dissocia- decided to further characterize lilli.
tion curves.

Function of lilli is required in oenocytes and fat body

Immunohistochemistry
Lilli is a homologue of a transcription elongation factor AF4/
Immunofluorescence staining was performed using the standard FMR2, which is associated with human diseases such as
procedure as follows. Epithelial tissue with oenocyte clusters mental retardation and acute lymphoblastic leukaemia (Gu
15 Page 4 of 8 Lan Enli et al.

Figure 2. Lilli and tho2 knockdown mutants showed hypertrehalosaemia. The ubiquitous knockdown of the target genes with the GS
system, with quadruplicate independent samples, revealed two strains that significantly elevated hemolymph trehalose levels. Error bars
represent the standard error of the mean (SEM) (t-test, *P \ 0.05, **P \ 0.001). All the unlabelled comparisons are not significant.

and Nelson 2003). The gene is not known to play an important (Yuva-Aydemir et al. 2019). We then examined levels of
role in basic metabolism, but lilli is critical for normal pho- glucose, which is another type of sugar in circulation (Klow-
toreceptor neurons in Drosophila via the IIS pathway (Wittwer den 2008). Hemolymph glucose was elevated in oenocyte-
et al. 2001). qPCR analysis confirmed that, in the knockdown specific and fat-body-specific lilli knockdown flies (figure 3d).
flies, the expression levels of lilli were downregulated in a In a study where screening was performed to identify genes
drug-dependent manner (figure 3a). To further confirm this, involved in glucose and trehalose regulation using mid-third
we examined heterozygote lilli mutant, F1 progeny between a instar larva, glucose and trehalose appeared to be under
wild-type OR and lilliA17-2/CyO and found elevated haemo- independent regulatory mechanisms (Ugrankar et al. 2015).
lymph trehalose levels (figure 3b). The lilliA17-2 allele was Although this screening did not identify lilli, our results imply
generated by ethyl methanesulphonate mutagenesis and is a that lilli may be a part of the basis of carbohydrate regulation.
loss-of-function allele, although the exact molecular nature To examine the lilli expression, we examined the Gal4
has yet to be determined (Melicharek et al. 2008). Given the enhancer trap strain in which the transgene containing Gal4
common caveat of off-target issues of RNAi and unknown under the minimum promoter is inserted into the lilli locus
modifier mutations, the functional significance of lilli will (Hayashi et al. 2002). We found that the reporter gene was
need to be conclusively validated by rescue experiments. expressed in both oenocytes and fat body, consistent with the
Next, to determine which tissues or organs require lilli, we notion that lilli is active in these tissues (figure 3, e–h).
performed tissue-specific knockdown and found that down-
regulation of lilli in either oenocytes or the fat body elevated
haemolymph trehalose levels (figure 3c). The oenocytes are a Lilli functions via IIS pathway
cluster of cells on the lateral surface of each segment and,
together with the fat body, are considered to subsume the There are eight insulin homologs encoded in the fly genome,
functions of mammalian liver (Gutierrez et al. 2007). We did and three of them (dilp2, 3 and 5) are highly expressed in the
not observe any change in neuron-specific knockdown stain. insulin-producing cells of the brain (Rulifson et al. 2002;
This is interesting because in addition to its role in growth Grönke et al. 2010; Colombani et al. 2012). The secretion of
control during development, lilli is critical for proper loco- Dilp peptides are tightly controlled by nutritional demands
motive activity and the longevity of post-mitotic neurons (Ikeya et al. 2002). We found that mRNA expression of dilp2,
 Examination of Niddm20 candidate genes of OLETF rats Page 5 of 8 15

Figure 3. Lilli is required in oenocytes and fat body for sugar metabolism. (a) qPCR confirmed the reduced expression of lilli. (b) The
heterozygote of lilli also showed elevated hemolymph trehalose, suggesting that half dosage reduction of lilli is sufficient for
hypertrehalosaemia. Tissue-specific GS knockdown showed that the lilli is needed in both oenocytes and the fat body, but not CNS (c). lilli
knockdown resulted in hyperglycaemia (d). Error bars represent the standard error of the mean (SEM) (t-test; *P\0.05; ns, not significant).
(e–h) Lilli expression was detected in the oenocyte and fat body of the lilli enhancer trap strain, w;P{w[?mW.hs]=GawB}lilli[NP2790]/
CyO. NP2790 enhancer trap line was crossed with UAS-GFP. (e, f) GFP fluorescent signal in the oenocyte of OR strain, control (e) and lilli
enhancer trap strain (f). (g, h) Lilli expression was also observed in the fat body of the enhancer trap strain (h) but not in the control strain
(g). (e0 –h0 ) Corresponding bright field images. Scale bar is 20 lm.

which is believed to play a key role in controlling glucose/ The latest rat-genome database describes 110 genes in this
trehalose in circulation, was lowered in both oenocyte-specific region, some of which were not examined in this study,
and fat-body-specific lilli knockdown flies (figure 4, a&b). On including lncRNAs, pseudogenes, and those whose ortho-
the other hand, dilp3 expression was upregulated, which may logs are not found in the fly genome. Thus, the current study
be a result of compensatory response (Grönke et al. 2010). is by no means a comprehensive screening. However, three
Finally, we examined the whole-body expression of InR aspects need to be pointed out. First, screening was per-
and 4E-BP, which are direct targets of dFOXO (Puig and Tjian formed on adult flies rather than third instar larvae, which
2005; Bai et al. 2013). Since dFOXO, a well-conserved main are a far more favoured developmental stage in similar
down-stream transcription factor of the IIS pathway, is metabolic studies of Drosophila, partly due to the feasibility
exported to the cytoplasm when the signalling is activated, the of haemolymph extraction. However, common diagnostic
upregulation of these genes is considered to indicate the state assays in mammals, such as insulin injection and OGTT,
of insulin resistance. We found that InR expression was have been developed for adult flies (Haselton et al. 2010;
slightly repressed only in the oenocyte-specific lilli knock- Haselton and Fridell 2011). Further, as is widely recognized,
down flies (figure 4c), but no difference was observed in either the IIS pathway has been linked to longevity (Broughton
gene of fat-body specific knockdown strains (figure 4d). et al. 2005, 2008). Thus, understanding the relatively less
explored adult-stage metabolism will help to identify con-
nections between metabolism and longevity. Second, we
Discussion addressed the function of oenocytes in the context of sugar
metabolism. The liver plays a pivotal role in metabolic
Candidate gene search on adult flies homeostasis. Insulin resistance in the liver is often a com-
mon denominator of many metabolic deficiencies, including
In this study, we examined 21 protein-coding genes anno- elevated fasting glucose. In insects, two morphologically
tated in one of the hyperglycaemic loci in the OLETF rat. independent tissues, oenocytes and the fat body, have
15 Page 6 of 8 Lan Enli et al.

Figure 4. IIS pathway components are deregulated in lilli knockdown flies. The whole body lilli mRNA expression was confirmed to be
downregulated in the tissue-specific knockdown females by qPCR. (a) oenocytes specific knockdown strain (b) the fat body specific
knockdown strain. For both tissue specific knockdown strains, the expression of dilp2 was downregulated, whereas the level of dilp3 was
elevated. Two of the IIS pathway marker genes were examined (c&d). One of the dFOXO targets, InR, was reduced in its expression when
lilli was knocked-down in oenocytes (c). Neither gene showed difference expression in the fat-body specific knockdown strain (d). Error
bars represent the standard error of the mean (SEM) (t-test; *P \ 0.05; ns, not significant).

equivalent functions to the mammalian liver (Gutierrez et al. (Gu and Nelson 2003; Melko et al. 2011). In vertebrates,
2007). Several studies have addressed the metabolic roles of there are four palalogues of AF4/FMR2: FMR2, MLLT2,
the fat body in larvae (Sekine et al. 2010; Bai et al. 2012; LAF4 and AP5q31. FMR2 is an ortholog of lilli. In verte-
Ugrankar et al. 2015), but the oenocytes have gained less brates, metabolic roles may be played by one or more of
attention (Chatterjee et al. 2014). Third, we used the GS other palalogs. In Drosophila, null mutation of lilli is
Gal4 system in an attempt to examine the effect of temporal embryonic lethal with defects in germinal band extension,
gene repression as opposed to life-long chronicle gene cytoskeletal movement and cell growth (Tang et al. 2001). It
knockdown. Indeed, the progeny of most of the RNAi was reported that Lilli controls cell growth via the IIS
strains used in this study were lethal when mated with the pathway and is associated with PTEN in a cell-autonomous
ubiquitous Gal4 driver (data not shown). In a large-scale manner (Wittwer et al. 2001). Since the IIS pathway is
screening to look for genes in sugar metabolism, either known to be activated in both oenocytes and the fat body,
ubiquitous or tissue-specific GAL4 driver strains were uti- lilli may be involved in metabolic regulation by cell-au-
lized. Neither study identified lilli as being involved in tonomously modulating the ISS pathway (Hwangbo et al.
metabolism (Pendse et al. 2013; Ugrankar et al. 2015). 2004; Chatterjee et al. 2014). lilli has also shown to be a part
of the regulatory network of TGF-b signalling (Su et al.
2001). Although the role of TGF-b in the pathogenesis of
Lilli’s function in metabolism diabetes is not fully understood, several studies have found
that the signalling is critical for b-cell differentiation (Lee
In human and mammalian model animal studies, AF4/FMR2 et al. 2021), the development of complications (Ma et al.
has been linked to fragile XE syndrome (FRAXE), which 2020) and adiposity (Yadav et al. 2011).
shows mild to borderline intellectual disabilities. However, To identify and better understand the genetic founda-
so far, the gene has not been linked to metabolic diseases tions of polygenic diseases, numerous rodent disease
 Examination of Niddm20 candidate genes of OLETF rats Page 7 of 8 15

models have been established, most of which were Gutierrez E., Wiggins D., Fielding B. and Gould A. P. 2007
developed either by selective breeding or fortuitous find- Specialized hepatocyte-like cells regulate Drosophila lipid
metabolism. Nature 445, 275–280.
ing in a colony. However, cloning of QTL causative genes
Gu Y. and Nelson D. L. 2003 FMR2 Function: insight from a
remains limited, mostly because the effect of each cau- mouse knockout model. Cytogenet. Genome Res. 100,
sative polymorphism or mutation is often below or close 129–139.
to the detection limit and also because epistasis plays a Haselton A., Sharmin E., Schrader J., Sah M., Poon P. and Fridell
major role in dictating quantitative traits. Even in geneti- Y. W. 2010 Partial ablation of adult Drosophila insulin-producing
neurons modulates glucose homeostasis and extends life span
cally homogenous rodent models, the overall genetic
without insulin resistance. Cell Cycle 9, 3063–3071.
architecture of individual QTL has been intractable (Deng Haselton A. T. and Fridell Y. W. 2011 Insulin injection and
2015), which suggests that a simpler system like Droso- hemolymph extraction to measure insulin sensitivity in adult
phila can play an important role in finding novel genes. Drosophila melanogaster. J. vis. Exp. 52, 2–5.
For lilli mutants, the immediate question that needs to be Hayashi S., Ito K., Sado Y., Taniguchi M., Akimoto A., Takeuchi
H. et al. 2002 GETDB, a database compiling expression patterns
addressed is whether the mutant is susceptible to nutri-
and molecular locations of a collection of Gal4 enhancer traps.
tional stress, such as a high-fat diet. Even if lilli is not a Genesis 34, 58–61.
causative gene for T2D in the OLETF rat, the establish- Hwangbo D. S., Garsham B., Tu M. P., Palmer M. and Tatar M.
ment of lilli as one of the components for metabolic 2004 Drosophila dFOXO controls lifespan and regulates insulin
homeostasis should shed light on the mechanism of T2D signalling in brain and fat body. Nature 429, 562–566.
Ikeya T., Galic M., Belawat P., Nairz K. and Hafen E. 2002
development.
Nutrient-dependent expression of insulin-like peptides from
neuroendocrine cells in the CNS contributes to growth regulation
Acknowledgements in Drosophila. Curr. Biol. 12, 1293–1300.
Kawano K., Hirashima T., Mori S. and Natori T. 1994 OLETF
(Otsuka Long-Evans Tokushima Fatty) rat: a new NIDDM rat
The authors thank KYOTO Stock Center (DGRC) at Kyoto Insti-
strain. Diabetes Res. Clin. Pract. 24, S317–S320.
tute of Technology, the Vienna Drosophila RNAi Center, and the
Kawano K., Mori S., Hirashima T., Man Z. W. and Natori T. 1999
National Institute of Genetics (NIG) for fly stocks.
Examination of the pathogenesis of diabetic nephropathy in
OLETF rats. J. Vet. Med. Sci. 61, 1219–1228.
Kawasaki K. Yamada S., Ogata K., Saito Y., Takahama A., Yamada
T. et al. 2015 Use of Drosophila as an evaluation method reveals
References imp as a candidate gene for type 2 diabetes in rat locus Niddm22.
J. Diabetes. Res. (https://doi.org/10.1155/2015/758564).
Akhi M., Kose H. and Matsumoto K. 2005 Fine mapping of the Klowden M. J. 2008 Physiological systems in insects, Academic
hyperglycemic and obesity QTL by congenic strains suggests Press, London.
multiple loci on rat chromosome 14. J. Med. Invest. 52, 109–113. Kose H., Band Y., Izumi K., Yamada T. and Matsumoto K. 2007
Alfa R. W. and Kim S. K. 2016 Using Drosophila to discover Epistasis between hyperglycemic QTLs revealed in a double
mechanisms underlying type 2 diabetes. Dis. Model. Mech. 9, congenic of the OLETF rat. Mamm. Genome 18, 609–615.
365–376. Kose H., Moralejo D. H., Ogino T., Mizuno A., Yamada T. and
Bai H., Ping K. and Marc T. 2012 Drosophila insulin-like peptide-6 Matsumoto K. 2002 Examination of OLETF-derived non-
(dilp6) expression from fat body extends lifespan and represses insulin-dependent diabetes mellitus QTL by construction of a
secretion of Drosophila insulin-like peptide-2 from the brain. series of congenic rats. Mamm. Genome 13, 558–562.
Aging Cell 11, 978–985. Lee J. H., Lee J. H. and Rane S. G. R. 2021 TGF-b signaling in
Broughton S., Alic N., Slack C., Bass T., Ikeya T., Vinti G. et al. pancreatic islet b cell development and function. Endocrinology
2008 Reduction of DILP2 in Drosophila triages a metabolic 162, 1–10.
phenotype from lifespan revealing redundancy and compensation Ma X., Cui Z., Du Z. and Lin H. 2020 Transforming growth factor-
among DILPs. PLoS One 3, 3–11. b signaling, a potential mechanism associated with diabetes
Broughton S. J., Piper M. D., Ikeya T., Bass T. M., Jacobson J., mellitus and pancreatic cancer? J. Cell. Physiol. 235, 5882–5892.
Driege Y. et al. 2005 Longer lifespan, altered metabolism, and Melicharek D., Shah A., DiStefano G., Gangemi A. J., Orapallo A.,
stress resistance in Drosophila from ablation of cells making Vrailas-Mortimer A. D. et al. 2008 Identification of novel
insulin-like ligands. Proc. Natl. Acad. Sci. USA 102, 3105–3110. regulators of atonal expression in the developing Drosophila
Chatterjee D., Katewa S. D., Qi Y., Jackson S. A., Kapahi P. and retina. Genetics 180, 2095–2110.
Jasper H. 2014 Control of metabolic adaptation to fasting by Melko M., Douguet D., Bensaid M., Zongaro S., Gecz J. and
dILP6-induced insulin signaling in Drosophila oenocytes. Proc. Bardoni B. 2011 Functional characterization of the AFF (AF4/
Natl. Acad. Sci. USA 111, 17959–17964. FMR2) family of RNA-binding proteins: insights into the
Colombani J., Andersen D. S. and Léopol P. 2012 Secreted peptide molecular pathology of FRAXE intellectual disability. Hum.
dilp8 coordinates Drosophila tissue growth with developmental Mol. Genet. 20, 1873–1885.
timing. Science 336, 582–585. Moralejo D. H., Wei S., Wei K., Weksler-Zangen S., Koike G.,
Deng A. Y. 2015 Genetic mechanisms of polygenic hypertension: Jacob H. J. et al. 1998 Identification of quantitative trait loci for
fundamental insights from experimental models. J. Hypertens 33, non-insulin-dependent diabetes mellitus that interact with body
669–680. weight in the Otsuka Long-Evans Tokushima Fatty rat. Proc.
Grönke S., Clarke D. F., Broughton S., Andrews T. D. and Assoc. Am. Physicians 110, 545–558.
Partridge L. 2010 Molecular evolution and functional character- Nicholson L., Singh G. K., Osterwalder T., Roman G. W., Davis R.
ization of Drosophila insulin-like peptides. PLoS Genet. 6 L. and Keshishian H. 2008 Spatial and temporal control of gene
(https://doi.org/10.1371/journal.pgen.1000857). expression in Drosophila using the inducible GeneSwitch GAL4
15 Page 8 of 8 Lan Enli et al.

System. I. Screen for larval nervous system drivers. Genetics Stančáková A. and Markku L. 2016 Genetics of type 2 diabetes.
178, 215–234. Endocr. Dev. 31, 203–220.
Ogino T., Wei S., Wei K., Moralejo D. H., Kose H., Mizuno A. Su M. A., Wisotzkey R. G. and Newfeld S. J. 2001 A screen for
et al. 2000 Genetic evidence for obesity loci involved in the modifiers of decapentaplegic mutant phenotypes identifies
regulation of body fat distribution in obese type 2 diabetes rat, lilliputian, the only member of the Fragile-X/Burkitt’s Lym-
OLETF. BMC Genom. 70, 19–25. phoma family of transcription factors in Drosophila melanoga-
Okamoto N., Nakamori R., Mura T., Yamauch Y., Masuda A. and ster. Genetics 157, 717–725.
Nishimura T. 2013 A secreted decoy of InR antagonizes insulin/IGF Tang A. H., Neufeld T. P., Rubin G. M. and Müller H. J. 2001
signaling to restrict body growth in Drosophila. Genes Dev 27, 87–97. Transcriptional regulation of cytoskeletal functions and seg-
Pendse J., Ramachandran P. V., Na J., Narisu N., Fink J. L., mentation by a novel maternal pair-rule gene, lilliputian.
Cagan R. L. et al. 2013 A Drosophila functional evaluation of Development 813, 801–813.
candidates from human genome-wide association studies of Ugrankar R., Berglund E., Akdemir F., Tran C., Kim M. S., Noh J.
type 2 diabetes and related metabolic traits identifies tissue- et al. 2015 Drosophila glucome screening identifies Ck1alpha as
specific roles for dHHEX. BMC Genomics 14 (https://doi.org/ a regulator of mammalian glucose metabolism. Nat. Commun. 6,
10.1186/1471-2164-14-136). 1–10.
Pospisilik J. A., Schramek D., Schnidar H., Croni S. J., Nehme Wittwer F., van der Straten A., Keleman K., Dickson B. J. and
N. T., Zhang X. et al. 2010 Drosophila genome-wide obesity Hafen E. 2001 Lilliputian: An AF4/FMR2-related protein that
screen reveals hedgehog as a determinant of brown versus controls cell identity and cell growth. Development 128,
white adipose cell fate. Cell 140, 148–160. 791–800.
Rogulja D. and Irvine K. D. 2005 Regulation of cell proliferation Yadav H., Quijano C., Kamaraju A. K., Gavrilova O., Malek
by a morphogen gradient. Cell 123, 449–461. R., Chen W. et al. 2011 Protection from obesity and diabetes
Rulifson E. J., Kim S. K. and Nusse R. 2002 Ablation of insulin- by blockade of TGF-b/Smad3 signaling. Cell Metab. 14,
producing neurons in files: Growth and diabetic phenotypes. 67–79.
Science 296, 1118–1120. Yuva-Aydemir Y., Almeida S., Krishnan G., Gendron T. F. and
Sekine O., Love D. C., Rubenstein D. S. and Hanover J. A. 2010 Gao F. B. 2019 Transcription elongation factor AFF2/FMR2
Blocking O-linked GlcNAc cycling in Drosophila insulin- regulates expression of expanded GGGGCC repeat-containing
producing cells perturbs glucose-insulin homeostasis. J. Biol. C9ORF72 allele in ALS/FTD. Nat. Commun. 10 (https://doi.
Chem. 285, 38684–38691. org/10.1038/s41467-019-13477-8).

Corresponding editor: ROHIT JOSHI