Molecular cloning and functional characterization of the sex-determination gene doublesex in the sexually dimorphic broad-horned beetle Gnatocerus cornutus (Coleoptera, Tenebrionidae)

Various types of weapon traits found in insect order Coleoptera are known as outstanding examples of sexually selected exaggerated characters. It is known that the sex determination gene doublesex (dsx) plays a significant role in sex-specific expression of weapon traits in various beetles belonging to the superfamily Scarabaeoidea. Although sex-specific weapon traits have evolved independently in various Coleopteran groups, developmental mechanisms of sex-specific expression have not been studied outside of the Scarabaeoidea. In order to test the hypothesis that dsx-dependent sex-specific expression of weapon traits is a general mechanism among the Coleoptera, we have characterized the dsx in the sexually dimorphic broad-horned beetle Gnatocerus cornutus (Tenebrionidea, Tenebirionidae). By using molecular cloning, we identified five splicing variants of Gnatocerus cornutus dsx (Gcdsx), which are predicted to code four different isoforms. We found one male-specific variant (GcDsx-M), two female-specific variants (GcDsx-FL and GcDsx-FS) and two non-sex-specific variants (correspond to a single isoform, GcDsx-C). Knockdown of all Dsx isoforms resulted in intersex phenotype both in male and female. Also, knockdown of all female-specific isoforms transformed females to intersex phenotype, while did not affect male phenotype. Our results clearly illustrate the important function of Gcdsx in determining sex-specific trait expression in both sexes.

Identification of sex-specific genome sequences in G. cornutus via RAPD-PCR. The sex of G. cornutus is indistinguishable by external morphology during larval and prepupal periods, so we developed PCR-dependent sexing methods before the molecular analyses of dsx. To develop PCR-based sexing method in this species we performed RAPD-PCR, using RAPD 10mer Kits (Operon Biotechnology, Tokyo, Japan). Genomic DNA (gDNA) was extracted from whole bodies of unmated adult male and female G. cornutus by dissecting a single leg, washing it in water, homogenizing it in 50 μ l of TES (0.1 M Tris-HCL (pH 9.0), 0.1 M EDTA, 1% SDS) and incubating it at 70 °C for 30 min. Then 7 μ l of 8 M K-Acetate was added and the samples left on ice for 30 min. After centrifugation, the gDNA was ethanol precipitated. The gDNA was used in as a template in PCR performed with AmpliTaq Gold 360 Master Mix (Applied Biosystems, Foster City, CA, USA) according to manufactures' protocol. The PCR program was: 95 °C for 9 min, and 45 cycles of [94 °C for 1 min, 35 °C for 1 min and 72 °Cfor 2 min]. PCR products were amplified using from male (but not female) samples using the A-09 primer (5′ -GGGTAACGCC-3′ ). This male-specific PCR product was isolated by electrophoresis on a 2% agarose gel (MetaPhor Agarose, FMC BioProducts, Rockland, ME, USA) and purified using the MagExtractor PCR & Gel-Clean up kit (TOYOBO, Osaka, Japan). The purified product was subcloned using a TOPO TA Cloning Kit (Invitrogen, Carlsbad, CA, USA), and sequenced using the BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems). From the obtained male-specific genome sequence, we designed a primer-pair for PCR dependent sexing: Gc-Y-05; 5′ -AGTGTTGACGCAAACCTATC-3′ Gc-Y-06; 5′ -AGTTCTGCAGCCATATCAGT-3′ For sexing PCR, DNA was prepared as follows: whole bodies (for larvae and prepupae) or dissected legs (for adults) were washed and homogenized in 100 μ l of 50 mM NaOH and incubated at 95 °C for 15 min. Then, 100 μ l of 200 mM Tris-HCl (pH 8.0) was added and the samples were centrifuged at 15,000 rpm for 10 min

RACE-PCR amplification of full-length Gcdsx transcript variants. For amplification of full-length
of Gcdsx we performed RACE-PCR. Using total RNA isolated as above, we synthesized cDNA for RACE-PCR using the SMART RACE kit (Clontech, Mountain View, CA, USA) according to the manufacture's protocol. We designed four gene-specific primers (primers for initial and nested PCR for both 3′ and 5′ -RACE) from the sequence of the Gcdsx fragment described above ( Expression analyses of Gcdsx variants by PCR. In order to investigate expression pattern of each Gcdsx variant, we performed expression analyses via PCR. We dissected heads of last instar, early and late prepupal period, and early and late pupal period G. cornutus and then preserved them at − 80 °C until use. Prepupal and pupal stages were judged based on external appearance and pigmentation level, respectively. Using the rest of

Functional analyses of Gcdsx variants via RNAi.
For investigating function of Gcdsx, we performed analyses via RNAi. Using primer pairs designed for each region (Table 2), partial sequences of Gcdsx were amplified by PCR and subcloned into TOPO vector (pCR4-TOPO) with the TOPO TA cloning Kit (Invitrogen). Then, insert regions were amplified with the universal primer pair with the T7 sequence (5′ -TAATACGACTCACTATAGGGAGACCACGTCCTGCAGGTTTAAACG-3′ and 5′ -TAATACGACTCACTATAGGGAGACCACCGAATTGAATTTAGCGGC-3′ ). Amplified PCR products were then purified as previously described. dsRNAs were synthesized using the MEGAscript T7 kit (Ambion, Austin, Tx, USA). dsRNA of the DsRed sequence negative control sequence was synthesized by the same method. Injection of dsRNA into larvae was performed using a microinjector (FemtoJet, Eppendolf, Hamburg, Germany) with a glass needle (Natsume-Kogaku, Nagano, Japan). The concentration of the dsRNA solution was 10 μ g/μl and approximately 0.10 to 0.68 μ l was injected into late instar larvae. Injected larvae were reared separately to induce pupation. Eclosed adult phenotypes were observed by binocular microscope and photographed with a VHX-900 digital microscope (Keyence, Osaka, Japan).

Results and Discussion
Development of PCR-based sexing method in G. cornutus. The sex of G. cornutus is indistinguishable during the larval periods, and it is also indistinguishable in individuals that have been phenotypically disrupted by dsx RNAi treatment. Thus, we first developed PCR-based sexing methods. We performed PCR with twenty different 10-mer primers provided with the a RAPD kit (A-01 to A-20) using gDNA of male or female G. cornutus. We thus obtained a 402 bp male-specific amplicon using the A-09 primer (Fig. 2). This sequence did not show significant homology with any previously identified sequence by Blastn using the NCBI database (http:// www.ncbi.nlm.nih.gov/). We then designed a pair of sexing primers (Gc-Y-05 and 06) based on this male-specific sequence, which amplified a 203 bp PCR product only when male-gDNA was used as the template (See Supplementary Fig. 1S). Considering that this species has XY sex determination system 20 , this male-specific sequence might be on Y chromosome. Thus, we used this PCR-based sexing method for identifying the original genetic sexes in phenotypically disrupted dsx RNAi individuals in later experiments (data not shown).
In the expression and functional analyses of sex-determination genes, it is important to know the sex of sample individuals in advance. However, some insects do not show morphological dimorphism, especially in larval and prepupal periods, which are critical periods in investigating development of sexually-selected exaggerated traits in some beetles 2 . Furthermore, when trying to identify the initial sex-determination molecular signal, it is necessary to know the sex of early embryos 21 , which are often difficult to determine morphologically. Our new male-specific primer pair enabled us to distinguish the sample sex in any developmental stage.
This PCR marker is also allows us to determine the original sex of dsx RNAi-treated individuals in the present study and will be used in future investigations of sex-determination mechanisms in this species.
Identification of dsx in G. cornutus . First, by using degenerate PCR, we identified a partial sequence of a putative G. cornutus dsx homolog. We obtained the full-length clone of this gene by subsequent RACE-PCR. The putative protein coded by the identified gene sequence contains amino acids of conserved DNA binding domain (DM domain/OD1 domain) (Fig. 3A) and conserved dimerization domain (OD2 domain). The DM domain is found in all members of the Dmrt gene family, including Dsx, while the OD2 domain is specific to Dsx 22 . Phylogenetic analysis using this conserved DM domain sequence indicated that the identified sequence from G. cornutus was grouped with other coleopteran Dsx sequences (Fig. 3B). These results strongly suggest that the identified gene was the homolog of dsx from G. cornutus. Consequently we named this gene Gcdsx.
We identified five different splicing variants of Gcdsx via RACE-PCR. One variant was isolated only from male, two variants were only from females and two variants were from both sexes, thus we named those variants Gcdsx-exon8-R 5′ -CTTGGAGCCCACTCTGAATC-3′ Table 2. Primers for dsRNA synthesis. Gcdsx-exon1,2-F and R are identical to Gcdsx-denegenerate-F and R.
Gcdsx-M, Gcdsx-FS, Gcdsx-FL, Gcdsx-C1 and Gcdsx-C2 (Fig. 4A). All of the variants have the same 5′ exon (exon1) encoding the DM domain, so all of the variants differed in the 3′ region (Fig. 4A). Exons 3, 5 and 6 were specific to Gcdsx-FS and Gcdsx-FL (Fig. 4A). Gcdsx-FS and Gcdsx-FL were distinguished by Gcdsx-FS specific exon 4 (Fig. 4A). We could not identify any Gcdsx-M specific exons. Exon 8 appeared in the non-sex-specific variants Gcdsx-C1 and Gcdsx-C2, which differed in some non-coding regions, but had identical putative ORFs. The lengths of the putative ORFs (in amino acids) for the Gcdsx variants were: 319 (GcDsx-M), 224 (GcDsx-FS), 249 (GcDsx-FL) and 149 (GcDsx-C1 and GcDsx-C2). Next, to examine the expression of these isoforms, we performed expression analysis between sexes by RT-PCR using the primers in listed in Fig. 4A. These results indicated that the sex-specificity of those variants were unchanged during all stages of postembryonic development (Fig. 4B). On the other hand, the expression level of sex-specific Gcdsx variants seemed to be greater during the prepupal period than in the late larval or late pupal periods in both sexes (Fig. 4B).

Comparison of GcDsx isoforms with other coleopteran Dsx proteins. Alignment of GcDsx iso-
forms with other coleopteran Dsx proteins indicated that three classes of Dsx isoforms are likely to be shared within coleopteran species (Fig. 5). Considering the expression patterns of those isoforms in this species and other species [5][6][7]19 , the three classes can be categorized as female-specific short (Dsx-FS), female-specific long (Dsx-FL) and male-specific (Dsx-M) (Fig. 5). The Dsx-FS class can be distinguished from Dsx-FL by protein size and the presence of four highly conserved residues (RQYG) in the C-terminal end (Fig. 5), while Dsx-FL has longer C-terminal ends (Fig. 5). Conservation of amino acid sequence of Dsx-FL among species is relatively lower than the other classes, and variation in isoforms within single species can be recognized (e.g. O. taurus has four Dsx-FL isoforms with different C-terminal residues) (Fig. 5). Dsx-M has the longest protein sequence in all of the five coleopteran species and is easily distinguishable from Dsx-FL or Dsx-FS by lacking 14 residues in the C-terminus of the OD2 domain (Fig. 5). All Dsx-M sequences have similar protein length and share other structural characteristics, such as a well conserved RP(S/R)SRA sequence at the protein's C-terminus, and especially possession of a conserved region just after the carboxy terminus of the OD2 domain (Fig. 5). This male-specific conserved domain shows weak similarity to the OD2 sequence. For example, the OD2 domain has two highly conserved sequences (WEMMPL) and (LEEAS(R/K)RIDEG), which are partially found in the male-specific domain. Gcdsx-C1 and C2 encode the same protein, GcDsx-C, which possesses a conserved DM (OD1) domain, but lack the OD2 domain. This isoform is truncated by the insertion of a stop codon at 5′ end of exon 8. T. dichotomus has a DsxC1 homolog with a similar size (144 aa) and characteristics (i.e. complete lack of an OD2 domain) as GcDsx-C.
In general, insect Dsx proteins have wide structural diversity on their C-terminal sides, so that it is difficult to align many regions of Dsx proteins from different insect orders. But within the coleopteran lineage, Dsx isoforms are well conserved both in structure and expression patterns [5][6][7]19 . Here, we propose a shared set of Dsx isoforms which may indicate common ancestry within the Coleoptera. That is, males express a single sex-specific isoform (Dsx-M) and females express two different sex-specific isoforms, one long one short (Dsx-FL and Dsx-FS, respectively) (Fig. 5).
Functional analyses of Gcdsx. In order to reveal the function of Gcdsx, we performed RNAi knockdowns of Gcdsx isoforms. By knocking down the function of these isoforms, we clearly demonstrated their critical function in sex-specific trait development (Table 3, Fig. 6). In individuals injected with DsRed dsRNA as a control, none of the sexually dimorphic structures were affected in comparison with wild-type (non-injected) individuals of either sex (Fig. 6A,B). However, when we injected dsRNA against Gcdsx exon 1 and 2, which are shared with all of the Gcdsx variants, injected individuals had phenotypically disrupted morphology in both sexes. In Gcdsx exon1,2 RNAi females, mandibles became slightly longer and genae became wider than DsRed injected control individuals. Additionally a small pair of bumps was formed between the eyes, where a pair of horns normally forms in males (Fig. 6C). On the other hand, in Gcdsx exon1, 2 RNAi males had much smaller sexually dimorphic structures (Fig. 6D). That is, mandibles became much shorter and genae became narrower. A pair of small horns became a pair of faint bumps rather than obvious horns (Fig. 6D). In conclusion, Gcdsx exon1,2 knocked-down males and females showed a similar intersexual phenotype, which is likely to be a developmental default state of this species (Fig. 6C,D). These results indicate critical function of GcDsx in weapon expression in G. cornutus as same as in other previously studied weaponed beetles. Thus, it is suggested that dsx gene has been repeatedly co-opted as a developmental regulator of sexually dimorphic weapon trait formation in the various beetle lineages.
Next, we performed RNAi experiments using dsRNA against isoform-specific regions of Gcdsx. We designed dsRNA corresponding to exon 4 (i.e. Gcdsx-FS specific knockdown), exon 5 (i.e. knockdown of both female-specific Gcdsx-FS and Gcdsx-FL) and exon 8 (i.e. knockdown of both non-sex-specific Gcdsx-C1 and Gcdsx-C2). Knockdown of Gcdsx-FS via injection of dsRNA of exon 4 and Gcdsx-C1/C2 via injection of dsRNA of exon 8 did not affect any morphological traits in either sex (Fig. 6E,F,I,J). In contrast, knockdown of both of female-specific Gcdsx (Gcdsx-FS and Gcdsx-FL) via injection of dsRNA of exon 5 caused an intersexual phenotype in females (Fig. 6G), the same as with the knockdown of all Gcdsx isoforms (Fig. 6C), but did not affect phenotype in males (Fig. 6H). Considering that intersexual phenotypes (short mandibles, narrow genae and a pair of faint bumps) is likely to be a developmental default state, GcDsx has a critical function on sex-specific trait expression in males, and inhibition of male sex-specific traits in females. It is known that insect Dsx protein functions as transcription factor, so that GcDsx-M and GcDsx-FL appear to play a central role in the regulation of downstream genes controlling male differentiation (longer mandibles, wider genae and a pair of horn) and female differentiation (tiny mandibles, absence of genae and horns), respectively.
This and previous studies have demonstrated by isoform specific RNAi that Dsx-M and Dsx-F are essential for expression or inhibition of sex-specific weapon characters in males and females, respectively 6,23 . Structural differences in the C-terminal region of Dsx proteins are critical for sex-differentiation or more specifically, for sex-specific transcriptional regulation of downstream genes.
On the other hand, we could not find evidence for a function of Dsx-FS alone in sex-specific trait development, i.e. Dsx-FS isoform-specific RNAi did not affect sexually dimorphic weapon characters. There are two non-mutually exclusive interpretations of these results. First, Dsx-FS may have functions in other sexual traits such as gonad development or in different developmental stages. To date, all functional studies of dsx in weaponed coleopteran species have focused on postembryonic development, especially on the prepupal period when sexually dimorphic adult structures develop. Thus, although Dsx-FS type isoforms seems to be non-functional for   sex-specific weapon traits via knock-down during prepupal period, it is necessary to investigate Dsx-FS function in gonad development in prepupal periods and early sex-determination including early germ cell differentiation during the embryonic stage. The second possibility is that Dsx-FL can compensate for Dsx-FS function, so that the effects of Dsx-FS knockdown were masked by the Dsx-FL isoform. In Tribolium castaneum, dsxFS RNAi affected ovarian development 19 . This result can be explained by either of those two possibilities. Further studies are necessary to reveal the functions of the two structurally different female isoforms of Dsx in coleopteran insects.