A novel domain duplication SlitFAR3 involved in sex pheromone biosynthesis in Spodoptera litura

Fatty Acyl reductases (FARs) are key enzymes that participate in sex pheromone biosynthesis by reducing fatty acids to fatty alcohols. Lepidoptera typically harbor numerous FAR gene family members. While the FAR gene of moth is involved in the biosynthesis of sex pheromone, the key FAR gene of Spodoptera litura is still unclear. In this work, we predicted 30 FAR genes from S. litura genome. A special domain duplication was found with gene SlitFAR3, which exhibited high and preferential expression in the sexually mature female pheromone glands, and its expression pattern was rhythmic during the scotophase of sex pheromone production. Functional expression in yeast cells combined with comprehensive gas chromatography (GC) indicates that SlitFAR3 gene was able to catalyze four methyl ester precursors into corresponding fatty alcohol products. The domain duplication FAR genes were further found to exist in 11 species of the other 19 Lepidoptera insect. The phylogenetic tree showed SlitFAR3 was grouped with the other six FAR with domain duplication under the pgFAR subfamily clade, which is important candidate FAR genes participating in sex pheromone biosynthesis, and the other eight FAR domain duplication genes of 7 species were scattered in different clades. Domain duplications may facilitate the evolutionary diversi�cation of protein sequences, which played diverse roles. This study is the �rst to focus on the special phenomenon of FAR domain duplication, which will advance understanding the biosynthesis-related genes from the perspective of evolutionary biology.


Introduction
and the components of ether lipids abundant in cell membranes (Nagan and Zoeller 2001). In insects, fatty alcohols are vital precursors of wax esters, cuticular hydrocarbons and sex pheromones. FAR gene function has been extensively studied in the PGs of moths. The substrate speci city of Bombyx mori pgFAR determines the exact proportion of components in the mixture of sex pheromones (Moto et al. 2003). A single FAR gene was found in Yponomeuta spp., which played a pivotal role in the production of noctuid multi-component pheromones (Lienard et al. 2010). The step of fatty alcohol production catalyzed by FAR is considered to be a potential reason for the differences of pheromones in Ostrinia spp. (Lassance et al. 2013). Therefore, FAR plays an essential role in the biosynthesis of moth sex pheromone.
FAR genes belong to a multigene family, which has experienced the evolutionary model of gene birth-anddeath in the process of evolution. The random occurrence of gene replication and loss will lead to the change of gene copy numbers or homologous family members in the genome (Eirín-López et al. 2012).
FAR gene loss was found in the lineage of nematodes, which only involves the loss of a single gene. In contrast, most species had a dynamic gain pattern with FAR genes. The FAR gene was very few in fungi, vertebrates and non-insect invertebrates, while plants and insects have a large number of FAR gene family members in the process of divergence (Eirín-López et al. 2012). FAR genes in insects have been widely screened. For example, there are 17 FAR genes in Nilaparvata lugens (Hemiptera) (Li et al. 2020), 12 FAR genes in Bombus lapidarius, 35 FAR genes in B. terrestris and 16 FAR genes in B. lucorum (Hymenoptera), respectively (Tupec et al. 2019), A total of 206 FAR genes have been identi ed in 12 Drosophila species (Diptera), ranging from 14 to 21 (Finet et al. 2019). However, direct experimental characterization for the function of FAR gene family is still limited.
Over a long period of evolution, each insect has formed its own speci c chemical communication system (Pasteels and Daloze 2003;Smadja and Butlin 2009). In Lepidoptera, female moths release sex pheromones to attract males of the conspeci c to mate in the scotophase. Spodoptera litura (Lepidoptera, Noctuidae) is a notorious crop pest, which is widely distributed throughout the world, In this study, we investigated the FAR genes family from the genome of S. litura with analysis of transcriptome from tissues of antennae, legs, proboscis and genitalia. A special domain duplication was found with gene SlitFAR3, and expression pro les of SlitFAR3 was clari ed by real-time quantitative PCR on pheromone glands (PGs) as well. A heterologous gene expression study in yeast combined with comprehensive gas chromatography (GC) analysis demonstrated that SlitFAR3 gene can transform precursors to produce fatty alcohols and plays an important role in the process of sex pheromone biosynthesis. We further analyzed domain duplication FAR genes in Lepidoptera by a schematic phylogenetic tree. Our ndings will contribute to elucidate the sex pheromone biosynthesis mechanism and understand the evolutionary diversi cation of protein sequences with domain duplication.

Insect rearing
S. litura pupae of both sexes were obtained from Henan Jiyuan Baiyun Industrial Co., Ltd, China. The pupae were sexed in cages (30 cm × 30 cm × 30 cm) under controlled conditions at 25℃ with 70% RH and a photoperiod of 16h: 8h (L:D). Adults were provided with 10% honey solution that was renewed daily. Adults on the rst day of eclosion were designated as 0-day-old individuals. Pheromone glands (PGs) of the 2-day-old female were removed and ash-frozen in liquid nitrogen and stored at -80℃ until used to isolate RNA. Additionally, the abdomen without PG and PGs of female virgin S. litura were used as reference material.

Prediction and Expression analysis of FAR genes
Transcriptome data of antennae, legs, proboscis and genitalia of female and male S. litura were downloaded from NCBI (accession numbers were enumerated in Supplementary Table S2) (Cheng et al. 2017). HISAT2 was used to compare the transcriptome data to the genome of S. litura. The reads count of each gene was calculated by featurecount, and the transcripts per kilobase per million mapped reads (TPM) value of each gene was calculated based on the length of each gene by TBtools ). The gene expression levels were estimated using TPM. Heatmap analysis is based on TPM values. The TPM values and relative expression values were represented as means ± standard error of the mean (SEM) based on three biological replicates.

RNA isolation and cDNA synthesis
The frozen tissues of S. litura were ground to a ne powder using a pestle and mortar chilled with liquid nitrogen before use. RNA was separately extracted from different tissues (8 PGs, 1 abdomen without PG). The total RNA was extracted using TRIzol Reagent (Invitrogen, California, United States) according to the manufacturer's protocol. The quality and concentration of RNA samples were checked with a NanoDrop spectrophotometer (ThermoFisher Scienti c, Wilmington, DE, USA). All samples were tested in three biological replicates. The rst-strand cDNA was synthesized using the PrimeScript™ RT reagent Kit (TAKARA, Japan) with gDNA Eraser (perfect Real Time) according to provided protocol. The synthesized cDNA was stored at -20°C until use.

Speci c Expression of FAR Genes of S. litura
Relative expression levels of the FAR genes in female PGs of 30 FARs was checked by qRT-PCR. And speci c expression pattern of SlitFAR3 in 2-day-old PGs during different times (L, light; D, dark) and different day ages were assessed by qRT-PCR as well. qRT-PCR was performed on ABI PRISM 7500 (Applied Biosystems, United States). The reaction consisted of 10 µL of GoTaq® qPCR Master Mix (Promega), 0.8 µL of primer (10 mM), 2 µL of sample cDNA, and 7.2 µL nuclease-free water. The cycling conditions were 95 ºC for 30 s, followed by 40 cycles of 95 ºC for 5 s, and 60 ºC for 34 s. Then a melting curve was conducted starting at 95 ºC for 15 s, 60 ºC for 60 s, 95 ºC for 15 s to determine the speci city of PCR products. Primers were listed in Supplementary Table S3. The housekeeping genes, actin and gapdh, were used as reference genes for quantifying the transcription level of the FAR genes in different tissues of S. litura.
Each sample was subjected to three biological replicates and three technical replicates, and qPCR data were analyzed by the 2 -∆∆Ct method (Schmittgen and Livak 2008). The differences in the transcript levels of FARs were compared by One-way ANOVA (SPSS 19.0, Chicago, IL, United States), followed by Tukey's test. The relative gene expression values were visualized using GraphPad Prism 8 (GraphPad Sofware Inc., San Diego, CA; https://www.graphpad.com) 2.6 Gene cloning SlitFAR3 gene speci c primers were designed as forward primer (5'-ATGGTTGTGTTGACTTCCAAAG-3') and reverse primer (5'-TTAATACATTTTTCTAGGCTTCAAATATT-3) based on the SlitFAR3 gene sequence (No. LOC111348489). The PCR reaction with a total volume of 50 µL containing: 0.5 µL of TaKaRa LA Taq (China, Beijing), 5 µL of 10 × LA Taq buffer , 8 µL of dNTP mixture, 2 µL of cDNA, 4 µL of primer, 30.5 µL of nuclease-free water. PCR thermal cycling conditions consisted of 94°C for 5 min, 35 cycles at 94°C for 30 s, 48°C for 30 s, 72°C for 2 min 40 s, and 72°C for 10 min. PCR products were separated on a 1% agarose gel.

Plasmid construction and yeast transformation
The open reading frame of SlitFAR3 was ligated into the GAL1 promoter of pESC-URA plasmid through BamH and Xho . The recombinant plasmid was named pESC-URA-SlitFAR3. The pESC-URA-SlitFAR3 recombinant plasmid was transformed into Saccharomyces cerevisiae WAT11 strain with a yeast plasmid transformation kit (Yeastmaker™ Yeast Transformation System 2, Takara) and cultured in SD-URA solid medium at 30 ℃ for 4 days. The monoclonal colonies were inoculated in SD-URA liquid medium at 30 ℃, 200 rpm overnight. The plasmid was extracted with a plasmid extraction kit (Tiangen, Beijing, China) and veri ed by PCR. PCR was performed by using Premix Taq™ enzyme (TaKaRa Taq™ Version 2.0 plus dye, China, Beijing) with a total volume of 50 µL containing: 25 µL of Premix Taq™ enzyme, 5 µL of template (monoclonal colonies were lysed in a metal bath at 95 ℃ for 10 min as a template), 4 µL of primer, 16 µL of nuclease-free water, the conditions of which included 94°C for 5 min, 34 cycles at 94°C for 30 s, 55°C for 30 s, and 72°C for 2 min 40 s, and 72°C for 10 min. 1 mL of veri ed correct bacterial solution was inoculated in 10 mL SC-URA medium at 30 ℃, 200 rpm for 24 h, then diluted to 100 mL SC-URA medium (2% ra nose) in the ratio of 1:20, and continue grew to OD 600 = 1.0.
In order to study the function of SlitFAR3 in vitro, the 100 mL SC-URA medium of recombinant protein pESC-URA-SlitFAR3 was centrifuged at 1100 g and resuspended with 100 mL fresh SC-URA medium (2% galactose) at 30 ℃, 200 rpm for 24 h. The culture medium was diluted in a ratio of 1:5 to 5 ml of SC-URA medium containing 2% galactose, 1% tergitol (Nonidet P-40, sigma) and precursor compounds. Take 5 µL precursor compound into 5 ml ( nal concentration: 0.5 mM) culture medium, 24 h, 30 ℃, 200 rpm. The cells were collected by centrifugation at 1100g, washed with sterile water for three times, and 200 µL hexane was extracted at 21 ℃, 200 rpm for 1 h, and stored at -20 ℃ was subjected to GC-MS. Z9E12-14: OAc as internal standard.

Functional assay
Gas Chromatography/Mass Spectrometry (GC/MS) was used for determine the products. One microliter yeast extract was analyzed on an Agilent 7890A series gas chromatograph coupled to a mass-selective detector Agilent 5975C (Agilent Technologies, California, United States). The GC was equipped with either a HP-5MS capillary column (30 m×250 µm × 0.25 µm lm thickness). Helium was the carrier gas (velocity 20 mL/min) and the injector was con gured in splitless mode and maintained at 250°C injector temperature, 1 mL/min column ow. In order to obtain good separation of the different isomers, the oven temperature was held at 50°C for 2 min and rose at a rate of 5°C/min up to 150°C, 15°C/min up to 250°C with a nal held for 3 min.

The FAR gene family in S. litura
From a bioinformatic screening, 30 candidate genes were identi ed as putative FAR genes in S. litura, which is similar to that reported in the related insect species. The "30 new genes" were named after a four-letter code ( rst letter of the genus followed by the rst three letters of the species name) + FAR + number according to their homology with the related species. 29 sequences were full length FAR genes that were characteristic of typical insect FARs because of their homologic sequences with other known FARs, while the N-terminal sequence of SlitFAR26 gene was incomplete. The sequence characteristics of all FAR genes indicated that the numbers of amino acid residues ranged from 428 to 868, with molecular weights ranging from 49.43 kDa to 99.13 kDa and isoelectric points of 6.21-9.39 (Supplementary Table  S4). Protein family domain architecture (Pfam) revealed that FAR had a typical conserved domain NADB_Rossmann site at the N-terminal and a FAR-C domain at the C-terminal. Among these FAR genes, the SlitFAR3 gene had special domain duplication, containing two NADB_Rossmann domains and two FAR-C domains (Fig. 1).

Tissue expression speci city of FAR gene family in S. litura
To determine the tissue expression level of FAR genes, an intuitive heatmap was illustrated for the transcriptome data of antennae, legs, proboscis and genitalia of female and male S. litura. Subsequently, the FAR genes with similar expression patterns were clustered together. The heatmap results presented the SlitFAR3 was highly expressed in female genitalia, while SlitFAR21 was highly expressed in female and male antennae, with almost no expression in other tissues. The expression levels of the SlitFAR3, SlitFAR6, SlitFAR11, SlitFAR16, SlitFAR23 and SlitFAR30 genes in female genitalia were higher than those in other tissues, and the other 10 FAR genes were hardly expressed in these tissues (Fig. 2).
By applying qRT-PCR on Pheromone glands (PGs) of the 2-day-old female, we further explored the vital FAR genes involved in the biosynthesis of sex pheromone in S. litura. SlitFAR1,3-8, [10][11][12][13]16,[25][26]30 were expressed in female PGs, and the expression level of SlitFAR3 gene was the highest, which was 50 times higher than SlitFAR1 gene. Both SlitFAR8 and SlitFAR12 genes were expressed at more than 5 times higher than SlitFAR1. The expression levels of the other 15 FAR genes were nearly zero (Fig. 3). These results suggested that SlitFAR3 was one of the important genes related to sex pheromone biosynthesis.

Analysis of Expression characteristic of SlitFAR3 gene
We performed a detailed test for expression levels of SlitFAR3 gene in 2-day-old PGs during different times (L, light; D, dark) and different day ages. The results showed SlitFAR3 gene was signi cantly expressed in PGs during scotophase. And the expression level of SlitFAR3 gene peaked at 3 h and 6 h (Fig. 4A), which displayed a similar rhythm with the release of sex pheromone (Sun et al. 2002). In addition, the expression of SlitFAR3 gene expression was higher in 2-day-old female PGs than that of 0day-old female PGs (Fig. 4B).

Cloning and Enzyme Function of SlitFAR3
The full length of SlitFAR3 gene was cloned by PCR from PG cDNA, and the length of the product was 2604 bp (Supplementary Figure S1). To test whether the PG-biosynthetic SlitFAR3 gene of S. litura has the fatty acyl reductase activity, we cloned the coding region of SlitFAR3 gene into Saccharomyces cerevisiae expression plasmid, heterologously expressed this SlitFAR3 in yeast, and analyzed fatty alcohols by GC. In the yeast cells transformed with recombinant plasmids, the functional research was carried out by  (Fig. 5A), which are main precursor compounds for the formation of sex pheromones in S. litura.
To ensure that the produced fatty alcohol was caused by the expression of SlitFAR3 protein, the transformed empty vector pESC-URA was established as negative control yeast strains. And the function of SlittoFAR gene of S. littoralis was investigated as positive control group, which is reported previously in same method as in this study (Supplementary Figure S3). No alcohol products were obtained in the negative control group, which yeast transformed by empty plasmid, and only substrate can be detected (Fig. 5B). The corresponding fatty alcohol products could be detected in the positive control group after adding the substrate (Supplementary Figure S3). In general, SlitFAR3 determined in yeast has substrate reducibility.

Domain duplication FAR gene family in Lepidoptera
SlitFAR3 gene had domain duplication, which has never been reported before. To gain insight into this gene works in lepidopteran insects, we further predicted and analyzed 361 FAR genes from the genomes of 19 Lepidoptera insect species. The number of FAR genes in different species ranges from 11 to 35, and 11 insect species had FAR genes with duplication domains. One FAR gene with duplication domains was found with H. erato, B. anynana, P. napi, L. accius, H. virescens, S. frugiperda, M. sexta, P. xylostella, and two FAR genes with duplication domains was found with P. machaon, P. Xuthus, P. polytes (Fig. 6). We reconstructed a schematic phylogenetic tree of FAR genes from 20 Lepidoptera insect species. The genetic relationship between species in the species tree was consistent with previous report (Simon et al. 2021). These results suggested that the domain duplication of FAR gene was widespread in Lepidoptera species.
We also reconstructed a FAR gene tree using 391 FAR protein sequences of 20 Lepidoptera species including S. litura. As illustrated in Supplementary Figure S4, SlitFAR3 from S. litura under the Lepidopteran pheromone gland speci c FAR (pgFAR) subfamily clade and was grouped with the other six FAR domain duplication from H. erato, H. virescens, S. frugiperda, M. sexta, P. machaon, P. xuthus, which is important candidate FAR genes participating in sex pheromone biosynthesis. The other eight FAR domain duplication genes of 7 species were scattered in different clades of the phylogenetic tree, including one FAR in P. machaon, one FAR in P. xuthus and two FAR genes in P. polytes. (Supplementary Figure S4). , which is mainly involved in the biosynthesis of sex pheromone. Previously, FAR genes of S. litura has been preliminarily screened from transcriptome, but further gene function has not been studied. Based on the genome data and annotation les of S. litura, we predicted 30 FAR genes comprehensively. Combining transcriptome and genomic data of S. litura, we obtained 17 new FAR genes. Normally, FAR gene is around 1500 bp, while the SlitFAR3 gene (NO. LOC111348489) has 2604 bp in this study, which is almost twice the 1365 bp of the previously reported SlitFAR3 gene (NO. KT261697) (Zhang et al. 2015). That means the result of transcriptome splicing could be limited without genome data and annotation les. We cloned its fulllength sequence by PCR, and sequencing indicated it is a completely novel FAR gene with domain duplication (Supplementary Figure S1-S2).

Discussion
To understand whether the FAR genes with special domain duplication also has the typical function of producing fatty alcohol, we analyzed the expression pattern of FAR genes in transcriptome of different tissues and the expression level in PGs. From the transcriptome of different tissues, SlitFAR3 was highly expressed in female genitalia. Furthermore, we revealed that 14 FAR genes were expressed in PGs by qPCR, and the expression level of SlitFAR3 gene was the highest. Therefore, we hypothesized that SlitFAR3 gene is related to the biosynthesis of sex pheromones. It has been measured that the sex pheromone content of female S. litura in the scotophase (Sun et al. 2002). The sex pheromone of S. litura was mainly released in the scotophase, and rested in the light phase (Li et al. 2012). There were two mating peaks, one at the beginning of the scotophase and one at 6-7 h after the scotophase (Wu et al. 2018). We discovered that the expression level of SlitFAR3 gene exhibited a uctuated in the scotophase by qPCR, which was consistent with the release rhythm of sex pheromone. In Y. evonymellus, the pgFAR gene related to sex pheromone biosynthesis was highly expressed in sexually mature PGs and showed a 24 h cyclic uctuation during the pheromone production period (Lienard et al. 2010). The key gene pgFAR of sex pheromone biosynthesis in B. mori is speci cally expressed in PGs, and no detectable signal in other tissues (Moto et al. 2003). Hence, we supposed that SlitFAR3 was related to sex pheromone biosynthesis of S. litura and the gene function was performed through eukaryotic expression system in vitro.
Regarding expression and function of FAR genes involved in the biosynthetic pathway of sex pheromones has received a lot attention specially in Lepidoptera (Tillman et al. 1999 The results displayed that SlitFAR3 has universal catalytic ability for these substrates. The enzymatic catalytic ability of SlitFAR3 in this work clearly illustrations that the sex pheromone component of S. litura is produced through the biosynthetic pathway involving SlitFAR3 (Fig. 7). Likewise, as a positive gene in this expression system, SlittoFAR gene also converts the substrate into the corresponding fatty alcohol product. Our work is the rst research on FAR gene related to sex pheromone biosynthesis with domain duplication. The formation of FAR multiple family is driven by gene loss and gain model, whether the production of new types of domains duplication is the driving force to promote the generation of new genes remains to be explored.
Based on the large-scale study of genomic data, gene duplications and lineage-speci c gene family expansion are considered to be an important mechanism for the evolution of new genes and new biochemical functions (Jordan et al. 2001;Lespinet et al. 2002). Protein domains represent a fundamental evolutionary unit forming proteins (Rossmann et al. 1974), and two proteins with common domains are likely to be evolutionarily related, originating from processes such as duplication and/or shu ing of whole genes or exons encoding domains (Doolittle 1995). Domain duplication and/or shu ing are probably the most important forces driving protein evolution as well as proteome complexity. While duplication of entire genes as well as exons encoding domains increases the abundance of domains in the proteome, domain shu ing increases versatility, the number of different contexts in which domains can occur (Vogel et al. 2005). In the FAR multiple gene family of S. litura, different from the typical FAR gene domain, SlitFAR3 has a special domain duplication, which has not been described before. Consequently, we predicted the FAR genes family of more species in Lepidoptera to determine whether this domain duplication phenomenon is accidental or common in evolution.
From the genomic data of 20 Lepidoptera species, 391 FAR genes were predicted, of which 12 species had 15 domain duplication FAR genes. All predicted FAR genes were reconstructed and the phylogenetic tree exhibited that 47% domain duplication FAR encoded protein belonged to an identi ed Lepidopteraspeci c pgFAR gene subfamily, pgFAR is a unique clade related to sex pheromone biosynthesis in Lepidoptera (Lassance et al. 2013). We speculated that the other FAR genes with domain duplication grouped in pgFAR subfamily clade may also participate in the biosynthesis of sex pheromones. FAR genes scattered in other clades may have diverse functions other than sex pheromone biosynthesis.
In summary, we predicted 30 fatty acyl reductase genes from the genome level of S. litura, and a novel domain duplication was found with SlitFAR3 gene. SlitFAR3 gene exhibited high and preferential expression in mature female PGs, its expression pattern in scotophase was consistent with the release rhythm of sex pheromone. The function of SlitFAR3 reductase to produce fatty alcohol was further con rmed by in vitro yeast express system combined with GC-MS. The results highlight the importance of SlitFAR3 with domain duplication in biosynthesis of sex pheromone, and provide a solid foundation for understand the evolutionary diversi cation of protein sequences with domain duplication.

Declarations
Author Contributions CL and FQL conceived and designed research. BYZ and YJF conducted experiments. CQ contributed new reagents and/or analytical tools. BYZ and FQL analyzed data. BYZ wrote the initial draft. CL and YJF revised this manuscript. All authors read and approved the manuscript.
Ethics approval and consent to participate Not applicable Consent for publication All authors are in accord with the submission.

Availability of data and materials
The datasets generated during and/or analyzed during the current study are available on reasonable request.

Competing interests
The authors declare that no competing interests exist.