Genome resequencing revealed genetic differentiation and promoter variation in wax secretion genes among geographic populations of Ericerus pela

Abstract

Background

Ericerus pela is an important source of insect wax with significant economic value. Despite its adaptability to diverse climates, variations in growth and wax layer thickness existed across populations. In white wax production, E. pela sourced from the Yunnan-Guizhou Plateau served as seed insects, transported to wax-producing regions (Sichuan and Hunan) to enhance white wax secretion.

Results

Genome of males was sequenced. Genome resequencing of samples from main distribution areas and some dispersal areas found 7,012,876 SNPs and 1,114,988 InDels that predominantly distributed in intergenic region and intronic region. Phylogenetics revealed three distinct groups. Group A were populations mostly from traditional insect-producing areas; Group B were mainly populations from the wax-producing areas and other dispersal areas; Group C were dispersal populations. Evolution analyses showed group A exhibited a single origin, and maintained low Ne, slow LD decay. While group B exhibited mixed ancestral origin and showed higher Ne, rapid LD decay. ZFst values and ZHp combined analyses identified some regions on tentative pseudochromosomes 3, 6, and 7 had values exceeding threshold, and the juvenile hormone esterase, cytochrome P450 301a1 were identified. Regions on the tentative pseudochromosomes 2, 3, 5 and 6 showed significantly high XP-CLR values, and elovl, fas genes were identified in these regions. Notably, certain genomic regions in three far genes had Pi and Theta values within the top 5% values of the genome-wide distribution. Analysis of their promoters revealed 24 SNPs loci and 7 InDels, and some distributed in the core promoter, CCAAT-boxes and TATA-box in the open chromatin region of far3 gene at the stage that begin to secret wax.

Conclusions

Considerable genetic variation was observed in gene intergenic regions. Notable genetic divergence between group A and group B of E. pela was detected based on population evolution analyses. It provided a molecular basis for selective breeding of E. pela.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-026-12684-0.

Introduction

The Chinese white wax scale insects (Ericerus pela Chavannes), belongs to the Hemiptera order and the Coccidae family, is a resource insect of significant economic value in China. The E. pela has been reared in China for over thousands of years. The white wax is secreted by male nymphs, and each male can secrete wax that almost exceeds its own body weight [1]. The white wax was originally used to make candles and lost wax casting. The white wax has stable chemical properties, moisture-proof lubrication and light, making it an important material in various products include medicine, textile printing, chemical engineering, and machinery [2].

In traditional white wax production, the practice of “producing insects in high mountains and wax in low mountains” has always been adopted. The E. pela eggs (Fig. 1A) from Yunnan and Guizhou in mountainous areas with an altitude of 1,500–2,000 m were selected and produced; the eggs were transported to Sichuan province or Hunan province in where white wax (Fig. 1B) was produced by the male offsprings [3]. The Yunnan-Guizhou Plateau, as the main breeding area of the E. pela, might be the origin of the E. pela. However, after many years of raising and transportation, the population structure and genetic variation of the E. pela remain unclear.

Fig. 1.

Phenotypic characteristics of the E. pela, genomic distribution of SNPs and InDels, and clustering of different E. pela populations. A Female adults for insect-producing. B White wax secreted by male insects. C Statistical result of SNP distribution in E. pela genome. D Statistical result of InDels distribution in E. pela genome. E Phylogenetic tree of E. pela populations established using SNPs based on the neighbor-joining algorithm. Populations in group A were shown in blue, populations in group B were shown in red, and populations in group C were shown in green

The E. pela have strong adaptability to different climatic environments and ecological conditions, and is distributed across various climatic zones and altitudes in China [3]. Although the amount of wax secretion is affected by environmental factors in low mountains, studies have shown that E. pela from different geographic populations raised in the same area showed differences in thickness of wax layer [4, 5].

Wax ester synthesis pathway had been studies in Arabidopsis thaliana, Simmondsia chinensis, Mus musculus, Acinetobacter calcoaceticus, and Euglena gracilis [6–11]. Fatty acyl-CoA reductase (FAR) and wax synthase (WS) are enzymes responsible for catalyzing wax ester synthesis. Transcription factors (TFs) regulating specific wax synthesis genes have only been identified in a limited number of species, such as in Arabidopsis thaliana, Boehmeria nivea, and the microalga Dunaliella parva [12–15].

In our previous study, far genes involved in wax secretion of the E. pela were identified [2, 16–18]. The open chromatin region of the far gene was between − 1000 and − 670 bp at the stage that begin to secret wax, and three possible wax secretion-related regulatory factors were screened out by using the yeast one-hybrid system [16]. Given some genetic differences especially the wax secretion among different geographical populations, it was speculated that there might be differences in the regulatory regions of genes involved in wax secretion.

The white wax production using E. pela has been established. Genomic analysis will provide information about genetic diversity of different E.pela populations. In this study, the genome of male E. pela was sequenced, and the genome of male E. pela of different populations were resequenced. The study will reveal the population evolution of E. pela, and provide a theoretical basis for selective breeding of E. pela.

Materials and methods

PacBio sequencing

The heads with antennae were detached form adult male E. pela individuals. Only heads and antennae were selected to prevent potential endophytic microbial contamination. Given the small size of E. pela, about five hundred heads and antennae were pooled for DNA isolation. A total of 18 µg DNA were obtained for PacBio HiFi seq II sequencing. For PacBio sequencing, DNA was broken by the Megaruptor (Diagenode, Megaruptor 3). The fragments from 13 kb to 16 kb were selected by Sage ELF (Sage Science, SageELF), and formed a Dumbbell shaped structure after adaptor ligation. The SMRTbell library was constructed. The sequencing was carried out by the HiFi model on the PacBio Sequel II.

Hi-C sequencing

Fifty-five E. pela adult males were dissected to collect antennae, heads, and thorax for Hi-C library preparation. The interacted DNA were crosslinked by formaldehyde. The DNA was cut by restricted enzyme. The ends were filled and labelled with biotin. The ends of DNA under the interaction of the same protein were ligated and formed circles. Then the DNA were purified and sheared. The labeled DNA fragments were captured. The suitable DNA fragments were selected for library construction. The library was sequenced by DNBSEQ.

Genome assembly of E. pela

The genome was assembled by using Juicer +3d-dna software [19, 20]. Small scaffolds were filtered out, and the remaining scaffolds were evaluated for consistency in locus interaction frequencies. The obtained sequences were anchored, sorted, and oriented by interaction intensity to create initially reliable chromosome-length scaffolds. Genome overlapping regions were identified using sequence homology and similar long-distance interaction patterns. Scaffolds and contigs were merged via these regions to get the final chromosome-length scaffolds. The genome underwent pseudochromosome-level assembly using Hi-C data. Chromosomal heatmaps were generated from these interaction data.

Genome resequencing of different population of E. pela

The sample collection areas included three towns from Qiaojia County, two villages from Yongshan County (both in Yunnan Province), and two towns from Jinkouhe District of Leshan City (Sichuan Province). One town from each of the following counties or cities: Xide County, Huili County, Mianning County, Jiajiang County, Emei City, Chengdu City (all in Sichuan Province); Tianshui City (Gansu Province); Yancheng City (Jiangsu Province); Changchun City (Jilin Province); Kaifeng City (Henan Province). The regions covered the insect-producing areas, wax-producing areas of the E. pela, and dispersal areas.

Genome DNA was randomly fragmented using a high-performance ultrasonic sample processing system (Covaris, 520104). After fragment selection, fragments of approximately 350 bp were obtained. After adding adapters, the adapter-ligated library underwent single-strand isolation and circularization. The circularized library was subjected to rolling circle amplification (RCA) to generate DNA nanoballs (DNBs). After passing quality control, the DNBs were loaded onto the DNBSEQ T7 (MGI Tech Co., Ltd., Shenzhen, China) for high-throughput sequencing.

The clean reads were aligned to the reference genome assembled above. The unaligned, low-quality (MQ < 4), and duplicate reads were removed. The whole genome resequencing data of 17 E. pela populations were aligned with the reference genome by the genomic alignment software BWA [21].

Genome SNP variation detection and annotation

All potential SNP loci in the whole genome of 17 E. pela populations were detected by using the HaplotypeCaller mode and joint calling method of the variation analysis software GATK [22] based on the alignment to the reference genome sequences. Further filtering and screening were carried out according to criteria included depth, quality value, and strand bias. Finally, a high-confidence SNP dataset was obtained. The ANNOVAR software [23] and the existing genomic annotation files were used to perform corresponding annotations on the SNPs. The PLINK (v1.9) software was used for SNP quality control.

Genome InDel variation detection and annotation

All potential polymorphic InDel (Insertion and Deletion) loci in E. pela genome were extracted by using the HaplotypeCaller mode and joint calling method of the mutation analysis software GATK [22] based on the alignment to E. pela genome sequence. Further filtering and screening were carried out according to the criteria. The ANNOVAR [23] software and the existing genomic annotation files (gff/gtf) were used to find positions of InDels in E. pela genome.

Population structure and diversity analyses

Based on the SNPs detected in the whole genome, the software GCTA was used for principal component analysis. The Neighbour-Joining (NJ) method of the TreeBeST software (v1.9.2) was used to construct a phylogenetic tree. The Admixture software [24] was used for population structure analysis. The script PopLDdey [25] was used to calculate the average coefficient r² of linkage disequilibrium (LD) and draw an LD decay plot. The Variscan software [26] was used to analyze various common polymorphism parameters. Polymorphism parameters were calculated by the Variscan (v2.0.3). The Vcftools (v0.1.6) [27] was used to calculate Fst and Pi (π) values. Fst values were Z-transformed to obtain ZFst, and regions with ZFst values exceeding the threshold corresponding to the top 5% of sites were identified as candidate selective regions. Hp values were calculated using a Perl script and then Z-transformed to derive ZHp, which was combined with ZFst. In addition, XP-CLR (v1.1.2) was used to compute XP-CLR values for detecting regions under selection. The effective population size (Ne) of each E. pela group was calculated using PSMC model (https://github.com/lh3/psmc).

Promoter region analyses of far genes

Three far gene locations on genome were identified according to the annotation results. Putative transcription start sites (TSS) and core promoter regions were identified using Promoter 2.0 (https://services.healthtech.dtu.dk/services/Promoter-2.0/). Regulatory elements were predicted through JASPAR ( https://jaspar.elixir.no/scan/) analysis.

Results

E. pela genome sequencing and assembly results

A total of 2,753,081 HiFi CCS reads (34.64 Gb) were obtained. N50 of the CCS sequences were 12,408 bp. 125.16 Gb of filtered Hi-C sequencing data was obtained after filtering. After removing duplicate data, 44.115Gb of final valid Hi-C data was acquired for subsequent auxiliary assembly.

After genome assembly, N50 of scaffolds was 82.8 M, and N50 of contigs was 2.0 M. By Hi-C assembly a total of 775 Mb of genomic sequences were initially anchored to 9 data-supported tentative pseudochromosomes with sparse contact signals in some regions, which accounted for 91.6% of all assembled sequences. The complete BUSCOs accounted for 95.30% (Table 1).

Table 1.

The BUSCO assessment results of assembled E. pela genome

	Insecta odb10		Arthropoda odb10
	Number	Percentage%	Number	Percentage%
Total BUSCO groups searched	1367	100.00	1013	100.00
Complete BUSCOs (C)	1276	93.30	966	95.30
Complete and single-copy BUSCOs (S)	1199	87.70	903	89.10
Complete and duplicated BUSCOs (D)	77	5.60	63	6.20
Fragmented BUSCOs (F)	28	2.00	14	1.50
Missing BUSCOs (M)	63	4.70	33	3.30

Repeated sequences accounted for 63.04% of genome. In total, 14,831 genes were identified, 81.11% of them were annotated by eight databases (Supplemental Table S1).

Genome resequencing showed distribution of SNP and InDel in E. pela genome

The coverage rate of the 17 resequencing populations ranged from 97.35% to 99.41%. A total of 7,012,876 high-quality SNP loci were identified. The intronic region (3,475,360 SNPs; 49.56%) and intergenic region (3,113,195 SNPs; 44.39%) had the most SNPs. While the UTR3, exonic region, and upstream region only had SNPs less than 5% (Fig. 1C).

A total of 1,114,988 InDel loci were identified. Similar with SNPs distribution, the intronic region had the most InDels (590,875 InDels; 52.99%), followed by the intergenic region (441,411 InDels; 39.59%). Othe region (UTR5, UTR3, exonic, splicing, upstream and downstream) only had 7.16% InDels (Fig. 1D).

Population phylogenetic tree construction and population structure analyses of E. pela

In the phylogenetic tree, the 17 E. pela populations were clustered into three major groups. Populations in group A were from Qiaojia County (A2, A8, A9) and Yongshan County (A15) of Zhaotong City, Yunnan Province, Huili County (A3), Xide County (A13), and Mianning County (A14) of Sichuan Province, and Tianshui City (A11) of Gansu Province. Most of them were traditional insect-producing areas, including Zhaotong in Yunnan Province and Xide in Sichuan Province. Populations in group B were mainly from Emeishan City (B18), Leshan City (B7), Chengdu City (B4), and Jiajiang County (B12) of Sichuan Province, Yongshan County (B17) of Zhaotong City, Yunnan Province, and Yancheng City (B10), Jiangsu Province. They were distributed in the wax-producing areas and other dispersal areas, except the B10 of Yongshan County. Group C were dispersal populations from Leshan City, Sichuan Province (C6), Changchun City, Jilin Province (C20), and Kaifeng City, Henan Province (C21) (Fig. 1E).

When population coefficient K = 2 and K = 3, the population structure indicated populations in group A shared a single ancestral origin and the populations in group C also exhibited a single ancestral origin. The populations in group B displayed mixed ancestral origins. When K = 4, the evidence of mixed ancestral origins in group B became more pronounced. This indicated that both groups A and C exhibit single ancestral origins, while group B shows significantly higher ancestral diversity (Fig. 2A).

Fig. 2.

Population structure analyses of E. pela. A Population structure of 17 E. pela populations. The horizontal axis displayed 17 E. pela populations. From top to bottom were the genetic structure diagrams when K = 2–4. B LD decay. The physical distance was plotted on the horizontal axis, the vertical axis displayed the average coefficient r² of LD corresponding to this physical distance. Curves of different colors represented different groups. C Effective population size of group A. Time was plotted on the horizontal axi and effective population size on the vertical axis. D Effective population size of group B. Time was plotted on the horizontal axis and effective population size on the vertical axis

Linkage disequilibrium and effective population analyses

Compared with group A and C, group B exhibited the fastest LD decay and the shortest LD maintenance distance (Fig. 2B).

Group A exhibited a sharp decline in effective population size from a previously high level around one million years ago, followed by multiple stepwise reductions, and eventually stabilizing at a consistently low level (Fig. 2C). In contrast, the effective population size of group B showed an overall declining trend but with a final effective population size higher than that of group A (Fig. 2D).

Genetic differences based on Fst analyses

The Fst values in a lot of regions between groups B and A were significantly higher than 0.25 (Fig. 3A), indicating obvious genetic differentiation between groups B and A in these regions.

Fig. 3.

Population evolution analyses of E. pela. A Fst values of group B/A distribution on the 9 data-supported tentative pseudochromosomes. The tentative pseudochromosome was plotted on the horizontal axis, the vertical axis represented the Fst value, and the horizontal line represented the threshold of top 5% Fst values. B Scatter plot of E. pela population selection region identification. The blue dots represented the selected areas of the control population (group A), and the red dots represented the selected areas of the analysis population (group B). C ZFst and ZHp values of group B/A distribution on the 9 tentative pseudochromosomes. The horizontal axis displayed tentative pseudochromosomes, the vertical axis displayed ZFst and ZHp values. The black dotted line represented threshold of top 5% values. D XP-CLR values of group B/A distribution on the 9 tentative pseudochromosomes. The horizontal axis displayed 9 tentative pseudochromosomes, the vertical axis displayed XPCLR values. The black dotted line represented threshold of top 5% values

For far3 region, Fst values varied between 0.01 and o.21. For far2 region, Fst values varied between 0 and 0.18. For far1 region, Fst values varied between 0.07 and 0.2. There were 24 regions in the three far genes showed Fst value above 0.15 (Supplemental Table S2).

Selected regions in E. pela genome identified by combined analyses

In these regions identified by ZFst and Pi combined analyses, a total of 254 selected regions were identified that ranked within the top 5% values of the genome-wide distribution (Fig. 3B). 167 protein-coding genes were identified in these regions. Gene annotation revealed probable cytochrome P450 301a1 and facilitated trehalose transporter Tret1.

Some regions on tentative pseudochromosomes 3, 6, and 7 showing significantly elevated ZFst. ZHp values decreased on the tentative pseudochromosomes 3, 5, 6, 7, and 8 (Fig. 3C). A total of 295 regions in group B/ A were identified with ZFst and ZHp values ranking in the top 5%. There were 334 protein-coding genes were identified. Elongation of very long chain fatty acids protein (elovl), facilitated trehalose transporter Tret1, and hormone-sensitive lipase were found.

The XP-CLR values of regions on the tentative pseudochromosomes 2, 3, 5, and 6 significantly higher than 20 (higher than top 5% value of the genome-wide distribution) (Fig. 3D), among which key genes such as elovl, fatty acid synthase (fas), and acyl-CoA synthetase (acs) were found.

Variations in promoter regions of the three far genes

The three far genes involved in wax secretion were located in different regions on the tentative pseudochromosome 6 and 9 (Fig. 4). There were 34 SNP loci and 7 InDel loci in far3 promoter, 14 SNP loci and 6 InDel loci present in far2 promoter, 16 SNP loci and 8 InDel loci existed in far1 promoter.

Fig. 4.

Gene structure of the three far genes involved in wax secretion. Gene regions and intergenic regions, promoters, and open chromatin domains were shown in different colors

For the open chromatin region (at the stage that begin to secret wax [16]) of far3, 11 SNPs variant were identified. Among these, two SNPs were located in the core promoter region, one of which resided within a TATA-box. Additionally, two other SNPs were found within CCAAT-boxes. The remaining 7 SNPs were not situated within the promoter or open chromatin regions. For the open chromatin region of far2, one SNP was found, located within the core promoter region. For open chromatin region of far1, one SNP was identified in the non-promoter region.

For the three far genes, certain genomic regions in groups A and B exhibited the values of Pi, Theta within the top 5% of the genome-wide distribution (Supplemental Table S3).

Discussion

Genome resequencing revealed almost half SNPs and InDels in the E. pela genome distributed within the intergenic regions. E. pela exhibits remarkable adaptability to diverse environmental conditions, which potentially driven by modifications in gene expression regulation. A study had demonstrated that SNP variations altered interactions among regulatory factors, and resulted in improving salt tolerance of the maize (Zea mays) [28]. Similarly, in Drosophila, an intronic SNP regulated the expression of the prominin gene, which modulated its adaptation to divergent microclimatic conditions [29]. Some researches had systematically elucidated how promoter and open chromatin region variations mechanistically regulate PCSK9 and ARID5B gene expression [30, 31]. It was hypothesized that the regulatory regions in the E. pela genome harbored abundant genetic variations that will facilitate environmental adaptation across distinct geographical habitats.

Group A which mainly distributed in the insect-producing areas maintained a low Ne accompanied by slow LD decay, and single ancestral origin, indicative of reduced genetic diversity and potential genetic bottleneck events, consistent with a population from a single, restricted origin. Group B demonstrated relatively higher Ne, rapid LD decay and mixed ancestral origin, reflecting enhanced genetic diversity. It was hypothesized to stem from the traditional wax production way of low-altitude regions for wax production in the past thousand years. The translocation and introduction of E. pela from different places to wax producing regions elevated genetic diversity in this group.

The declined Ne is often associated with population bottlenecks, reflected the differences in selection pressures that the population encountered during the evolutionary process [32]. The stabilization of Ne at a low level in group A suggests that its genetic structure may have been shaped by selection pressures, consistent with features often observed in populations with a high degree of domestication. Similarly, in well-established domestic populations such as chickens, pigs, and dogs, Ne has been maintained at low levels over the long term [33–35].

The Fst value exhibits a positive correlation with the level of genetic differentiation [36, 37]. The Fst values of many regions between groups B and group A of E. pela were significantly higher than 0.25, indicating intense selection pressure, especially the 24 regions of three far genes had Fst value above 0.15. Several regions of the three far genes displayed values of Pi, Theta within the top 5% values. The analysis of top 5% XP-CLR regions between groups A and B revealed elovl1 and fas genes. This showed these regions may be subject to selection or local adaptive differentiation. Similar observations have been found in apple [38] domestication which have high genetic diversity by gene flow in different groups and directional selection. The joint analysis of ZFst and Pi also identified juvenile hormone esterase and cytochrome P450 301a1. In Apis mellifera and Bombyx mandarina, juvenile hormone esterase and cytochrome P450 enzymes were also found under selection in domestication [39].

The variations in the regulatory regions of genes involved in wax secretion could potentially influence gene expression by modulating interactions between regulatory elements and chromatin. CCAAT-boxes which were predicted binding sites for NF-YB transcription factors, harbored two identified SNPs, potentially affect transcription of the gene. Additionally, one SNP localized in a TATA-box could affect the assembly of the transcription initiation complex. There was one SNP in the core promoter region of far2. Similar study showed promoter region variations in PTGS1 and PTGS2 genes can change the binding site to regulate the expression of PTGS1 and PTGS2 gene [40].

Conclusions

This study revealed abundant variations in intergenic region and intronic region. Group A displays low Ne, slow LD decay, and single ancestral origin, pointing to a history of genetic bottleneck and artificial selection. The high Fst values between group A and group B showed selective pressures have driven genetic differentiation. Combined analysis of ZFst, ZHp, Pi, and XP-CLR revealed significant genetic diversity across multiple chromosomal regions of E. pela populations between group A and group B. These regions included segments of far, fas, elvol1, and juvenile hormone esterase genes. SNPs located in the CCAAT and TATA box regions, as well as in the core promoter, may affect binding of transcription factor. These results provide molecular insights into the variation in wax secretion and population genetic differentiation of the E. pela.

Authors’ contributions

Conceptualization, P.Y.; methodology, P.Y.; formal analysis, X.-Y.N.; writing of original draft, X.-Y.N. and P.Y.; writing of review & editing, X.-Y.N. and P.Y.

Funding

Research was supported by the Chinese Academy of Forestry Outstanding Youth Support Program [grant number CAFYBB2023QB007], the General Program of National Natural Science Foundation of China [grant number 32570607], and the Young- and Middle-Aged Academic and Technical Leaders Reserve Talent Project of Yunnan Province [grant number 202105AC160031].

Data availability

Part of the data was deposited in SRA of NCBI with the access number PRJNA1283529.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent of publication

Not applicable.

Competing interests

The authors declare no competing interests.