Functional Characterization of Two Elongases of Very Long-Chain Fatty Acid from Tenebrio molitor L. (Coleoptera: Tenebrionidae)

The elongases of very long chain fatty acid (ELOVL or ELO) are essential in the biosynthesis of fatty acids longer than C14. Here, two ELO full-length cDNAs (TmELO1, TmELO2) from the yellow mealworm (Tenebrio molitor L.) were isolated and the functions were characterized. The open reading frame (ORF) lengths of TmELO1 and TmELO2 were 1005 bp and 972 bp, respectively and the corresponding peptide sequences each contained several conserved motifs including the histidine-box motif HXXHH. Phylogenetic analysis demonstrated high similarity with the ELO of Tribolium castaneum and Drosophila melanogaster. Both TmELO genes were expressed at various levels in eggs, 1st and 2nd instar larvae, mature larvae, pupae, male and female adults. Injection of dsTmELO1 but not dsTmELO2 RNA into mature larvae significantly increased mortality although RNAi did not produce any obvious changes in the fatty acid composition in the survivors. Heterologous expression of TmELO genes in yeast revealed that TmELO1 and TmELO2 function to synthesize long chain and very long chain fatty acids.

Fatty acids (FAs) are molecules with a variety of biological functions including acting as energy sources and serving as components of cellular lipids and other molecules including eicosanoid hormones such as prostaglandins and leukotriences 1 . FAs are precursors of sphingolipids, glycerolipids 2 , hydrocarbons 3,4 , fatty alcohols and wax esters 5 , which can participate in many cell biological processes such as reproduction, growth, migration, differentiation, and apoptosis and are components of pheromones in various arthropod species [6][7][8] .
Structurally, FAs are composed of long hydrocarbon chains that end in a carboxyl group and are classified based on the chain length and the number of double bonds. Fatty acids are roughly classified by length into the following groups: (1) Short Chain Fatty Acids (SCFA) which have five or fewer carbons, (2) Medium Chain Fatty Acids (MCFAs) which contain 6-12 carbons, (3) Long-Chain Fatty Acids (LCFAs), which contain more than 12 carbon atoms, and (4) Very Long-Chain Fatty Acids (VLCFAs), which contain 22 or more carbon atoms. Each class is associated with unique functions. For example, in mammals LCFAs act as ligands for peroxisome proliferator-activated receptors (PPARs) and regulate energy metabolism 9 while VLCFAs have important anti-inflammatory roles 10 .
Fatty acid synthesis consists of a four-step cycle that includes condensation, reduction, dehydration and reduction steps and occurs primarily in the endoplasmic reticulum (ER) 11 . Each cycle extends an initial acetyl-CoA by two carbons and can be repeated up to seven times to form palmitic acid (C16:0) 2 . Further growth requires elongases of long or very long chain fatty acids (ELOVL or ELO) which can elongate C ≥ 14 fatty acids through a fatty acid condensation reaction. Elongases of very long chain fatty acids have been isolated from various organisms, including yeast, mammals, plants and other species [12][13][14] .
While ELOVLs have been relatively well characterized in vertebrates, little is known about these enzymes in insects, even though LCFAs and VLFAs are widespread among insect taxa [15][16][17][18][19][20] . Although some ELOVLs were isolated and characterized in a handful of arthropod species including Drosophila melanogaster and Aedes albopictus 21 , the ELOs and ELOVLs in insects are poorly understood and require further investigation 4,22,23 . The yellow mealworm beetle, Tenebrio molitor, is a species that has abundant fatty acids including VLCFAs, yet the T. molitor elongases associated with VLCFA synthesis have not been identified. In this study, we identified and characterized two elongases (TmELO1 and TmELO2) of Very Long-Chain Fatty Acids of T. molitor. These two ELOs were chosen out of the 20 putative ELOs identified from a transcriptome analysis due to the availability of full length sequences. Functions of TmELO1 and TmELO2 were analyzed using both an in vitro expression system in yeast and in vivo RNA interference in mature T. molitor larvae.
Multi-sequencing alignment of the TmELOs with ELOs from other species showed that the proteins had some similar motifs such as KXXEXXDT, HXXMYXYY, TXXQXXQ and HXXHH, a histidine-box motif that is conserved in all elongases 28 (Fig. 1). The TmELO1 and TmELO2 had the highest identity at 92.51% with Tribolium castaneum LOC660197 and 88.62% with T. castaneum LOC660257 respectively, two uncharacterized predicted elognases. Compared with D. melanogaster elongases, TmELO had identity at 60.17% with CG31523, an uncharacterized, predicted elongase in D. melanogaster (Genbank accession no. NP_649474.1). The TmELO sequences shared 17.18-43.79% identity with Homo sapiens with HsELOVL7 being the most similar. The TmELO sequences had low identity to S. cerevisiae with the highest identity being only 18.67%.
A phylogenetic tree was constructed comparing the amino acid sequences of T. molitor elongases 1 and 2 with elongases from other organisms (Fig. 2). The phylogenetic analysis shows that TmELO1 clustered with T. castaneum LOC660197 and D. melanogaster CG31523 while TmELO2 clustered with T. castaneum LOC660257 and D. melanogaster CG2781, another predicted member of the ELO family.
Relative transcript level at developmental stages. Expression profiles of TmELO1 and TmELO2 at various developmental stages were generated using qRT-PCR (Fig. 3). Expression of both elongases was detected throughout all developmental stages examined. Expression of TmELO1 peaked in 1 st instar larvae and pupae while TmELO2 peaked in embryos and was somewhat reduced throughout the remaining developmental stages. During embryonic development, expression of TmELO2 was significantly higher than TmELO1. Expression of TmELO1 was significantly higher than TmELO2 throughout all other stages examined.  Table 2. The major fatty acids of T. molitor were C14:0, C16:0, C18:0, C18:1 and C18:2. Saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs) and polyunsaturaed fatty acids (PUFAs) were 25.55%, 32.26% and 41.25%, respectively. The fatty acids (C > 16) were 81.2%.
Double-stranded RNA (dsRNA) fragments were generated for gene silencing using RNA interference. The dsRNA fragment lengths of TmELO1 were 361 bp (dsELO1-1) and 298 bp (dsELO1-2), and TmELO2 were 401 bp (dsELO2-1) and 342 bp (dsELO2-2), respectively. The relative transcript levels of TmELO1 and TmELO2 in dsRNAs-injected mature larvae decreased significantly at 1d, 2d, 3d and 4d after injection (Fig. 4). One day after injection, the transcript level of TmELO1 was significantly reduced by 76.5-81.3% and TmELO2 was significantly reduced by 75.8-86.7%. Both TmELO genes were largely suppressed after 1d, 2d, 3d and 4d, which showed that the dsRNA had a long-lasting effects on the TmELO1 and TmELO2 expression. Fatty acid compositions in the survivors after RNAi injections were not significantly different than controls (data not shown), however, dsRNA injection was associated with increased larval mortality (Fig. 4). Compared to the control, mortality after TmELO1 RNAi treatment significantly increased to 50.8%, while, the mortality following TmELO2 RNAi treatment was slightly, but not significantly higher than controls. In-vitro functional characterization of TmELOs. The ORFs of TmELO1 and TmELO2 cDNA were isolated from T. molitor, cloned into a pYES2 yeast expression vector, and the recombinant plasmids were introduced into yeast strain INVSc1. Transcripts of two TmELOs were detected by RT-PCR in the TmELO-transformed yeast (Fig. 5). The fatty acid analysis of transformed yeast showed TmELO1 expression increased relative amounts of C14:0, C16:0, C16:1 and C 20:0, and C24:0 fatty acids and reduced C18:0 fatty acids when compared with controls ( Table 3). Expression of TmELO2 in yeast increased relative amounts of C14:0, C14:1, C16:0, C16:1 fatty acids and reduced C18:1 fatty acids. In general, our observations suggest that in yeast, TmELO1 could produce C20:0 and elongate SFA to C24; TmELO2 primarily increases the percentage of C16:0 and C16:1. (Table 3).

Discussion
Tenebrio molitor larvae contain a large proportion of C ≥ 16 fatty acids including LCFAs and VLCFAs and here we characterize two elongases, TmELO1 and TmELO2 that are involved in their synthesis. Both TmELO1 and TmELO2 are expressed throughout the lifetime of T. molitor and do not exhibit sex-specific expression. During embryonic stages, TmELO2 appears to be the predominant form while TmELO1 is expressed at higher levels throughout the remaining developmental stages. Our results are consistent with findings in mammals that demonstrate developmental regulation of different elongases, and that expression is regulated not only by developmental signals but also by diet 29 . The effect of diet on T. molitor ELOVLs during development is intriguing but is outside the scope of the current study.
In the present study, expression of T. molitor elongases in yeast cells demonstrates differences in the fatty acids produced by TmELO1 and TmELO2 ( Table 3). Expression of TmELO2 led to significant increases of fatty acids up to 16 carbons in length, suggesting that TmELO2 function is limited to long-chain fatty acid synthesis. Meanwhile, TmELO1 expression led to similar increases in long-chain fatty acids but also produced a significant increase in C20 fatty acids and slight, but not significant increases in C22, and C24 fatty acids, suggesting that TmELO1 may possess an additional role in the synthesis of very long-chain fatty acids. Surprisingly, RNAi-mediated knockdown of TmELO1 and TmELO2 in T. molitor did not significantly alter the fatty acid composition (data not shown). The lack of significant changes suggests that residual TmELO1 and TmELO2 transcript remained following RNAi treatment. It is also likely that T. molitor elongases have redundant roles and can maintain fatty acid levels in the absence of TmELO1 or TmELO2 activity. In support of the idea of redundant roles of elongases, ELOs in other species also have broad and somewhat overlapping functions. For example, in humans, HsELOVL1 can elongate C20-C26 SFAs 30 , while HsELOVL3 can elongate FAs from C16 to C22 2 . Our observation that the two TmELOs generate FAs of different lengths is not unprecedented. In mammals, ELOVL1-7 play different roles in the elongation cycle of fatty acids 2, 31, 32 . ELOVL1 produces C20 to C26 SFAs and MUFAs 30 . ELOVL2 can elongate C20 to C22 PUFAs 33 . ELOVL3 can elongate FAs from C16 to C22, with the highest activity toward C18-CoAs 2 . ELOVL4 is specialized to elongate ultra long-chain fatty acids (ULCFAs) which are longer than C26 34,35 . ELOVL5 elongates FAs from C18 to C20 36 . ELOVL6 can elongate C12:0 FAs to C16:0 37 . Finally, ELOVL7 has highest activity toward C18:3n-3 FA and C18:3n-6 FA 38,39 .
Three ELOs of S. cerevisiae also showed variable functions. ScELO1 can elongate C14 FAs to C16 FAs 12 while ScELO2 is involved in the synthesis of SFAs and MUFAs to C24 and ScELO3 is essential for the synthesis of C24:0 to C26:0 and can also elongate a wide range of SFAs and MUFAs 13 . Similar results were also observed in Drosophila. Chertemps et al. (2005) reported an elongase gene named elo68α in Drosophila males which can elongate myristoleic and palmitoleic acids (C14:1n-9 and C16:1n-9) in-vtiro expression in yeast 22 . Chertemps et al. (2007) found that eloF cDNA of D. melanogaster expressed in yeast could elongate medium-chain saturated and unsaturated fatty acids up to C30 4 .
In T. molitor, silencing of TmELO1 via RNAi resulted in an increased mortality rate indicating that TmELO1 is essential for mealworm survival. Similar requirements for elongases in organism survival have been reported in other species. For example, in Drosophila, RNAi to CG6660, which encodes a predicted elongase, induced a

Materials and Methods
Insects. T. molitor were obtained from Shandong Agricultural University and have been reared in our laboratory at Zhejang A&F University for two years. The artificial climate chambers were maintained at 26 ± 1 °C, 65 ± 5% relative humidity and a 8:16 (L:D) photoperiod. The mealworm were reared on wheat bran mixed with cabbage.
Cloning of TmELO cDNAs. Total RNA was extracted from different stages of T. molitor using RNAiso Plus (TAKARA, Dalian, China), per the manufacturer's protocol and resulting RNA was stored at −80 °C. The concentration and quality of total RNA was measured with a UV/VIS spectrophotometer (BioDrop μLite, Cambridge, UK) and agarose gel electrophoresis was used to verify integrity of the RNA. The 500ng total RNA was used in cDNA synthesis reactions using the PrimeScript TM 1st strand cDNA Synthesis kit (TAKARA, Dalian, China) and resulting cDNA was stored at −20 °C. De novo transcriptome sequencing was conducted at Tianke High-Tech Development Co., Ltd. (Zhengjiang, China) and the transcriptome analysis of T. molitor was performed in our laboratory (data not shown). The specific primers used to amplify two putative ELO genes are listed in Table 4. The PCR conditions for all these amplicons were 35 cycles of 94 °C for 15 s, 55 °C for 30 s and 72 °C for 1 min. A 3′A-overhang was generated by Taq DNA polymerase (TAKARA), and amplified products were gel purified with an E.Z.N.A gel extraction kit (Omega bio-tek, Norcross, GA) and linked with a plasmid pMD19-T vector (TAKARA) to form a recombinant plasmid, which was then transformed into DH5α (E. coli) for 10 hrs cultivation in LB solid medium with 50 μg/ml ampicillin. The plasmids were sequenced with sequence-specific primer M13 and maintained as glycerol stocks at −80 °C.
Sequence characterization of TmELOs. Full-length cDNA, ELO nucleotide sequences were translated into amino acid sequences and chemical and physical characteristics of TmELO genes were determined by the Expert Protein Analysis System program (http://web.expasy.org/protparam/) 24 . Trans-membrane structures were calculated using the TMHMM Server v.2.0 (http://www.cbs.dtu.dk/services/TMHMM/). Subcellular locations were predicted by Euk-mPLoc 2.0 (http://www.csbio.sjtu.edu.cn/bioinf/euk-multi-2/) [25][26][27] . SOMPA 44 was used for the prediction of protein secondary structure. Amino acid sequences of other species ELOs were obtained from NCBI (http://www.ncbi.nlm.nih.gov/) and a phylogenetic tree was constructed and analyzed with respect to its homolog using the Neighbor Joining method. A bootstrap consensus tree of 1000 replicates was used to evaluate branch strength for analysis using MEGA 6.06.

Preparation of dsRNA and injection.
The dsRNA fragments of two ELO genes which were used to silence the corresponding mRNA transcripts were synthesized using the in vitro Transcription T7 Kit (TAKARA) 45 and the specific primers used are listed in Table 4. Plasmid pMD19-T vectors linked with amplified TmELO1 and TmELO2 were the templates. The quality and length of the dsRNA fragments were measured by BioDrop μLite (BioDrop, Cambridge, England) and electrophoresis. The fragments were purified and diluted into a final concentration of 1 μg/μl. All dsRNAs were stored at −20 °C. Each of the two ELO dsRNA were injected separately into mature larvae of T. molitor and EGFP dsRNA was used as the control. The larvae were collected 1d, 2d, 3d and 4d after injection, flash-frozen in liquid nitrogen and stored at −80 °C. Animals were observed daily and the death rate was recorded through the pupal and adult stages till two weeks after injection. Every sample contained 20 larvae (n = 3) which were injected with 5 μg dsRNA fragments.
Quantitative real-time PCR. Total RNA extraction from T. molitor and cDNA synthesis was conducted using the same procedures as Cloning of TmELO cDNAs. The cDNA products were diluted to 5 ng/μl. Primers (  at 80 °C for 2 h to prepare mealworm fatty acid methyl ester (FAME)(n = 4). Total lipid was extracted from yeast cells by using acidified glass beads to break cell wall for 10 min, and using chloroform:methanol (2:1, v-v) and 1% H 2 SO 4 in methanol (v:v) at 80 °C for 2 h to prepare yeast FAME (n = 4). Mealworms and yeast FAME were added to 2 mL 0.9% NaCl and extracted twice with 2 ml of hexane. Hexane was subsequently removed by nitrogen gas and the total FAME was resuspended in 300 ml of hexane. After mealworms or yeast were homogenized, 100 μg C17:0 was added as an internal control. FAME were then extracted and analyzed by GC.   Table 4. Sequences of primers used in DNA cloning, dsRNA synthesis and qRT-PCR. Xho I site (GGATCC), BamH I site (CTCGAG) and T7 promoter sequences (GATCACTAATACGACTCACTATAGGG) are underlined.