Whole-genome resequencing analysis of the medicinal plant Gardenia jasminoides

Introduction

Gardenia jasminoides Ellis (2n = 22) is an evergreen shrub of the genus Gardenia in the Rubiaceae family. It is distributed in many areas of China and is mainly used in the cultivation of medicinal plants and garden landscape. G. jasminoides is known for its special aroma and is an excellent greening tree (Xiao et al., 2017). Gardenia has very high medicinal value, and its fruits, leaves, flowers and roots can be used as medicine. It is an ideal Chinese medicinal plant (Zhao et al., 2017; Lu et al., 2019; Zhang, Bian & Yao, 2022). The chemical components of gardenia fruits are mainly iridoids, including geniposide and gardenic acid, among which geniposide is one of the most active components (Long et al., 2013; Xu et al., 2016; Zhou et al., 2021). In addition, gardenia also contains carotenoid compounds, such as crocin, which can regulate the neural center, improve memory and cognition (Tian et al., 2020; Shen et al., 2022), and phenolic acid compounds, mainly chlorogenic acid, which can lower blood pressure, antioxidant, free radical scavenging and regulate the body’s immunity (Liu et al., 2020). With the continuous research, development and utilization, gardenia has become the first batch of edible and medicinal plants in China. Gardenia yellow can be used as natural food colorant and food coloring agent (Wu et al., 2022). In Thailand, G. jasminoides is used in religious rituals and hair decorations, and in other Asian countries, G. jasminoides is even used in cooking (Chen et al., 2020a; Chen et al., 2020b).

In recent years, whole genome resequencing has been conducted on many plants, such as soybean, rice and sorghum (Habyarimana et al., 2022; Long et al., 2022; Wang et al., 2022). In order to detect various types of genetic variation and improve and domesticate genes better, whole genome sequencing on 292 pieces of pigeonpea were conducted and found several genome regions that might be targets for domestication and breeding. Then, through genome-wide association analysis, it was found that the genes related to flowering time control, seed development and pod cracking had sequence similarities with other plants. Meanwhile, some trait selection areas were consistent with geographical distribution, suggesting that different geographical environments led to the adaptive changes of some traits in pigeonpea (Varshney et al., 2017). On the other hand, some studies have provided a new perspective for the evolution of loquat genome and fruit domestication by resequencing wild loquat (Jing et al., 2023). In addition, resequencing analysis and KEGG enrichment analysis were used to identify differences in nitrite metabolism between ’Linhuang No.1′and ’Muzao’ cultivars (Li et al., 2021). Genomes including heavy single nucleotide polymorphisms (SNPs) and insertions or deletions (indel) through high-throughput were sequenced to detect complex genetic variation, for example, structure variation (SV) and copy number variation (CNV) (Ganal, Altmann & Röder, 2009; Adedze et al., 2021; Ma et al., 2021). A total of 520,260 SNP variations were identified after sequencing the whole genome of 18 flue-cured virginia (FCV) tobaccos (Nicotiana tabacum). A total of 4,849 homozygotes and 28,584 polynucleotide polymorphisms were also detected, which are critical for the development of excellent tobacco breeding lines (Thimmegowda et al., 2018). InDel markers based on whole genome resequencing have attracted more and more attention due to their advantages of wide distribution, high density, stable variation, strong polymorphism and easy detection in the genome. For example, different InDel markers have been developed based on resequencing data in terms of watermelon peel traits and color (Li et al., 2018). InDel markers were used to identify genetic loci related to the regulation of peanut bacterial wilt resistance genes (Zhang et al., 2022). From the development of gardenia genome, it was found that the functional genes involved through tandem gene replication in gardenia fruit (Xu et al., 2020). At present, there are few reports on genetic diversity analysis of germplasm resources of gardenia, and due to the differences or incomplete species of the gardenia population selected by various researchers. There are abundant variation conditions among wild populations, the polymorphism between and within the species is not completely consistent. The previous study results can not totally reflect the genomic characteristics of these genetic resources. At the same time, due to the lack of genomic data, there are few stable and reliable molecular markers, which greatly limits the progress of genetic breeding of gardenia.

In this study, we employed whole-genome resequencing technology to align the genomes of one wild types and one cultivated variety of G. jasminoides to the reference genome of G. jasminoides. Based on this, we conducted differential analysis at both individual and population levels, aiming to explore the differences and structural variations in gene sequences. This analysis included InDel detection and annotation, SV detection and annotation, distribution of all variations on the genome, and analysis of genes involved in the variation of geniposide and crocin production. Functional enrichment analysis was performed to annotate the variant genes, laying the foundation for the genetic study of G. jasminoides.

Materials and Methods

Results

Mutation detection and annotation

According to the mapping results of clean reads in the reference genome, the SNP, InDel, SV and CNV mutations of two samples FD and YP1 were detected. For SNP mutation, 3,087,176 and 3,241,416 high quality SNP sites were obtained respectively. Then the SNP was annotated, in which the sample FD conversion and subversion ratio was 1.87, the sample YP1 conversion and subversion ratio was 1.89, and the heterozygous SNP ratio FD was 49.64% and 67.07%. The heterozygosity of sample FD is lower than that of sample YP1 (Table S2). From the Venn map between samples, it was found that there were 1,737,580 SNP variants in FD and YP1, 1,503,846 SNP variants specific to YP1, and 1,349,606 in FD (Fig. S2). The SNP mutations of the whole genome can be divided into six types, among which the number of T:A > C:G in YP1 is more than that in FD (Fig. S3). The results of SNP annotation showed that there was no significant difference in SNP variation between FD and YP1 (Figs. 1A, 1B). After that, the InDels of small fragments between the sample and the reference genome were detected. The numbers of InDels of FD CDS and genome fragments were 10,270 and 558,766, the YP1 is 10,752 and 540,581. the number of InDels in the coding region of YP1 was more than that of FD (Fig. S4), and the number of homozygous InDel of FD in the whole genome and coding region was higher than that of YP1, and the number of YP1 heterozygous InDel in the whole genome and coding region was higher than that of FD (Table S3, Figs. 2A, 2B). In addition, the genomic SV of the two samples was detected and annotated. There were 11,543 and 7,634 SV variants in FD and YP1. The number of insertion type variation, deletion type variation, inversion type variation, repetition type variation and chromosome translocation type variation in FD were higher than those in YP1. The annotation results of SV showed that the variation of FD in the gene was more than that in the intergenic region, and the insertion and deletion in the intergenic region was much higher than that in the gene, while the reversal and repetition were much lower than those in the gene. In the annotation of CNV, 885 variant genes, 956 lost genes, 840 predicted single copy number, 885 variant genes, 780 deletion genes and 770 predicted single copy number were obtained by FD substitution type and YP1 substitution type, respectively. The number of variation obtained by FD and YP1 was the same, and the number of lost genes was more than YP1, and the number of single copy was also more than that of YP1 (Table S4).

Variation in the CDS region may cause changes in gene function. By looking for genes with non-synonymous mutations in SNP, InDel and SV in the CDS region between the reference genome and the sample, the genes were found that may have functional differences between the sample and the reference genome. Among the differential gene variations in FD, the type with the largest number of mutations was SNP, with a total of 23,338 variations. In YP1, the number of differential genes was also the most in SNP, with 24,631, in which the number of InDel mutants in YP1 was more than that in FD, while the number of SV variations was on the contrary, and SV variations in FD were more than YP1 (Table 3, Fig. 1).

Functional annotation at the DNA level

Variations in the CDS region may cause changes in gene function, the variation genes of FD and YP1 were compared with GO, COG functional database, and these gene annotations were obtained to analyze gene function. In the GO annotation map of the variant genes of FD and YP1 (Figs. 3A, 3B), they are divided into three categories, namely biological process, cellular component and molecular function, in which the variant genes of FD and YP1 are expressed in 16 terms of biological process, of which FD annotated 3,049 GO items, 7,771 genes, YP1 annotated 3,061 items and 7,881 genes. The significantly rich items of FD and YP1 are biological process (GO: 0008150), FD and YP1 have 1,752 gene mutations, cellular process (GO:0009987), FD and YP1 have 1,328 gene variation. The variant genes of FD and YP1 are expressed in 16 terms of cellular component, of which FD annotates 607 GO items, 7,599 genes, YP1 annotates 605 items and 7,788 genes. The most significant rich item of FD and YP1 is integral component of membrane (GO:0016021), which has 6,148 gene mutations. The mutant genes of FD and YP1 were expressed in 10 terms of molecular function, including 1,139 GO entries and 14,339 genes annotated by FD and 1,147 items and 18,585 genes by YP1. The most significant rich item of FD and YP1 was ATP binding (GO:0005524), which had 2,639 gene variations.

Table 3:

Classified statistics of differential genes produced by various variants.

Smaple ID	Genes with non-synonymous SNP	Genes with InDel	Genes with SV
FD	23,338	7,796	4,907
YP1	24,631	8,132	2,881

DOI: 10.7717/peerj.16056/table-3

COG database can reflect the lineal homologous classification of gene products. COG functional classification map analysis showed that FD and YP1 were annotated in 24 categories (Figs. 3C, 3D), of which the richer categories were carbohydrate transport and metabolism, general function prediction only, secondly metabolites biosynthesis and transport and catabolism, and the least annotated category was RNA processing and modification.

Discussion

G. jasminoides is widely distributed in the tropics of Southeast Asia, but it has been cultivated in China for at least 1000 years. Gardenia population is rich in genetic diversity, and there is a close relationship between geographical distance and genetic distance between populations (Han et al., 2007; Yang et al., 2016). At present, it is generally believed that there is genetic variation among different populations of gardenia in China, and it has adapted to different living environments in the long-term process of cultivation and domestication (Mei et al., 2015). However, the differences and diversity between cultivated and wild populations have not been well explored.

In this study, a total of 39,406,575 clean reads were obtained from one wild and cultivated species, and the average depth of each sample was 10X. A large number of variations such as SNP, InDel, SV and CNV were detected. Compared with other mutations, SNP and InDel have more abundant genetic variation and can better represent the genetic information of the whole genome of gardenia. SNP and InDel variation analysis have been widely used in the fields of population genetic structure, genetic diversity analysis and functional gene mining of non-model organisms (Chen et al., 2021; Giudice, Bazakos & Vassiliou, 2022). 3,087,176 SNP variants were detected in cultivar gardenia FD, and 3,241,416 SNP variants were detected in YP1. The same was true of the number of SNP variations in wild species YP1 than in cultivated species FD1, InDel. It can be seen that wild species YP1 has higher genetic diversity, and in different Brassica napus, it is also found that winter rape has higher genetic diversity because the density of SNP and InDel is higher than that of spring rape (Wu et al., 2019). The most abundant type of SNP variation in sample YP1 and FD was the transformation type, and there was almost no difference between the conversion type of FD and the subversion type of SNP and YP1, especially the SNP ratio of heterozygous type of YP1 was much higher than that of FD, which indicated that YP1 had higher heterozygosity than FD and had higher homology with gardenia material with reference genome, while the cultivated species FD had more homozygous SNP than wild species YP1. It is reported in different varieties of Camellia sinensis that the ratio of non-synonymous mutation to synonymous mutation of InDel variation is almost the same as that of long-term interspecific hybridization. It is inferred that to a large extent, these two samples and the reference genome of gardenia do not come from the same region (An et al., 2020). Transition (Ti) and transversion (Tv) are two types of SNP variants, and the proportion between them is related to the evolution of the species (Mursyidin et al., 2022), the proportion of Ti/Tv between the two samples is very close, which does not rule out the common origin and ancestor of cultivated species FD and wild species YP1. The number of the main mutation types of FD was also less than that of YP1, and the variations of the two samples were mainly concentrated in a large number of frame shift mutations in the intergenic region. Generally speaking, the variation of FD of cultivated species is relatively conservative, which has lost a lot of genetic diversity while having more stable quality, which is similar to the reported results of pear, cotton and wheat (Li et al., 2021a; Li et al., 2021b; Hussain et al., 2022).

This study also explored the variation genes of geniposide and crocin, which are the main medicinal substances in gardenia fruit. The mining of variation genes related to medicinal substances is helpful to the improvement and application of gardenia genetic germplasm. The key genes in geniposide synthesis pathway are mainly annotated on five pathways, of which Tyrosine metabolism (ko00350) and Monoterpenoid biosynthesis (ko00902) have been shown to be closely related to geniposide synthesis in plants (Pan et al., 2021). Tyrosine acts as the precursor of many specialized metabolites, which have a variety of physiological functions. Some plant natural products derived from Tyrosine are also used in human medicine and nutrition, and in some Monoterpenoid plant products, they are used as effective drugs for a variety of diseases (Schenck & Maeda, 2018; Nguyen et al., 2022). Not only that fatty acid degradation (ko00071), carbon metabolism (ko00071), glycolysis/gluconeogenesis (ko00010) are also synthetic precursors of many metabolites (Newman et al., 2008; Lin et al., 2019; Toleco et al., 2020). The variant genes with differences in these pathways need to be further explored in the future. One of the important synthetic precursors of crocin is carotenoid. It is reported that carotenoid biosynthesis pathway has a great influence on the synthesis of crocin in gardenia (Shen et al., 2022), and 30 DEGs have been annotated on the carotenoid biosynthesis KEGG pathway of gardenia (Pan et al., 2021). In addition to carotenoid biosynthesis pathway, flavonoid biosynthesis (ko00941) and flavone and flavonol biosynthesis (ko00944) are important secondary metabolites that exist widely in plants, which not only play an important role in plant growth and development, but also have outstanding applications in food and medicine. The KEGG pathways annotated by CCD4a, ALDH2C3, UGT74F8, UGT94E13 of crocin structural genes are all involved in regulating pigment metabolism. This is also the main reason for the color change of gardenia after ripening (Sommano et al., 2020; Liu et al., 2021; Zhang et al., 2023).

With the intensive artificial selection in pursuit of better quality and higher yield, the cultivated germplasm resources have lost a lot of genetic diversity. It has been reported that in order to accelerate the biological research and genetic improvement of tomato, pan-genomic comparison was carried out between cultivated and wild species of tomato to reveal SVs (Li et al., 2023). The number of samples selected in this study is small, and more wild and cultivated species of G. jasminoides have not been analyzed, so the genetic difference between samples is not significant. In the future, more germplasm resources are needed to explore the genetic diversity and origin of wild gardenia species, to establish a molecular marker system for the identification of different gardenia germplasm, and further functional characterization of the variation can better understand the genetic basis of the differences between cultivated species and their wild relatives.

Conclusions

The whole genome of one wild and one cultivated species of gardenia was resequenced, and a large number of SNPs, InDels, CNVs and SVs mutations were detected. SNP and InDel variation are important tools for studying genomic diversity and genome-based breeding, which have extensive application prospects and important practical value. This study have rapidly expanded our understanding of genetic variation of medicinal plants and provided rich resources for genetic research. KEGG significant enrichment analysis showed that the richness of metabolic pathways proved the possible role of differential genes in important biological and metabolic pathways. It was found that the related genes in geniposide and crocin biosynthesis pathway may have mutations, and the corresponding molecular markers can be developed in follow-up research to mine excellent genes for further verification. The extensive variation between wild and cultivated species provided in this study provides a good way to make further use of the genetic diversity of wild gardenia species for gene-based breeding, and also has a certain guiding significance for the evolution and classification of G. jasminoides.

Supplemental Information