Comparative Analysis Revealed Intrageneric and Intraspecific Genomic Variation in Chloroplast Genomes of Actinidia spp. (Actinidiaceae, Viridiplantae)

Abstract

About ten species of the genus Actinidia Lindl. are known as cultivated plants—alongside the most known A. chinensis var. deliciosa, there are many others, including Far-Eastern cold-hardy kiwifruits such as A. arguta, A. kolomikta, and A. polygama. Unlike most plant species, in which the chloroplast genome is maternally inherited, the family Actinidiaceae possesses a complex system of plastid inheritance with possible transmission through both maternal and paternal lines. The main aim of this work was the assembly of the plastid genomes of three species of Actinidia, their comparison with already-available sequences from databases, and evolutionary analyses. We discovered that the gene composition and gene sequences are conserved; the studied species are either subject to purifying selection or not subject to selection at all (with some exceptions, such as the ycf2 gene). However, the chloroplast chromosomes of some Actinidia species have undergone significant structural rearrangements, leading to the persistence of two main forms, both on an intrageneric and intraspecific level. These results expand our understanding of plastid genomics and genetic diversity within the genus Actinidia, providing a basis for future research in molecular marker development, phylogenetic analysis, and population studies.

1. Introduction

The genus Actinidia Lindl., together with the genera Saurauia and Clematoclethra, belongs to the family Actinidiaceae [1,2] and includes more than 60 species [3]. More than ten species of the genus Actinidia bear edible fruits with a pleasant taste and high nutritional value; they possess a unique set of biologically active substances and contain a significant amount of ascorbic acid, though they have a low acidity level in terms of acid-sugar balance [4]. The most common and well-known species throughout the world is Actinidia chinensis var. deliciosa [5]. However, recently, other species of Actinidia have become a relatively new type of commercially grown fruit. In most regions of Russia, it is possible to cultivate three Far Eastern species of the genus Actinidia [6], perennial cold-hardy vines with smaller (berry-sized) kiwifruits with thin edible skin: Actinidia arguta (Siebold ex Zucc.) Planch. ex Miq., A. kolomikta (Rupr. ex Maxim.) Maxim., and A. polygama (Siebold ex Zucc.) Maxim. Based on the morphological characteristics of species, the genus Actinidia was classified into four intrageneric sections: Leiocarpae, Maculatae, Stellatae, and Strigosae [3]. Such classification systems do not always reflect the actual relationships among species [7]; a more accurate approach, which can help solve complex issues, is to take into account molecular data [8,9].

For phylogenetic studies of plants, both individual genes and complete genomes of chloroplasts are often used due to the relatively small plastome size and the presence of conserved regions [10,11,12]. The Actinidia spp. possess a rather conserved genome structure with four independent parts, including an LSC (large single-copy) region, an SSC (small single-copy) region, and two separated inverted repeat regions (IRa and IRb) between the LSC and SSC. According to the NCBI database, chloroplast DNA (cpDNA) among representatives of the genus Actinidia is represented by one circular chromosome with a size of about 156 kbp and 130 genes, of which 83 are protein-coding. One of the most notable features of Actinidia cpDNA is the loss of part of the clpP gene compared to some other members of Ericales [13]. Moreover, despite the high conservation of the plastome in terrestrial plants, sometimes massive structural changes are found in some families of angiosperms, such as Geraniaceae [14,15], Fabaceae [16,17,18], and Campanulaceae [19,20].

In most plant species, the chloroplast genome is maternally inherited. However, some species are known to be able to transmit their plastid genomes through the paternal line. Such contrasting options of organelle inheritance have been found in some gymnosperms [21], as well as in some angiosperms, for example, Musa [22] and Cucumis [23]. Actinidiaceae show a complex system of plastid inheritance with possible transmission through both maternal and paternal lines [24]. Since some of the restructured angiosperm plastomes occur in lineages with biparental plastid inheritance [25], it can be hypothesized that the nature of inheritance may influence the stability of the chloroplast genome. There are cases in the literature [26] where biparental inheritance in combination with the fusion of paternal and maternal plastids can lead to homologous recombination between different copies of the cpDNA, which ultimately entails a change in the structure of the genome [27]. Restructuring of plastomes is often associated with the presence of small, dispersed repeats (SDRs), although the exact mechanism is still unknown [28,29,30,31].

Due to the unusual form of inheritance of chloroplasts and putative heteroplasmy in Actinidia species, the main aim of this work was the de novo chloroplast assembly of three Actinidia species (A. arguta, A. kolomikta, and A. polygama) from the collection of the N.V. Tsitsin Main Botanical Garden, Russian Academy of Sciences, comparison with already available sequences from databases, and the evolutionary characterization of these species in terms of positive selective processes.

2. Materials and Methods

2.1. Sampling and Sequencing

Young leaves of 3 Actinidia species (A. arguta, A. kolomikta, and A. polygama) were obtained from the collection of Far Eastern Species of the N.V. Tsitsin Main Botanical Garden, Russian Academy of Sciences. Total genomic DNA was extracted according to the method of Lo Piccolo [32] with minor modifications: leaf tissues were disrupted by the TissueLyser II system (Qiagen, Hilden, Germany), centrifugation before isopropanol precipitation was performed at room temperature, and precipitated DNA pellets were dissolved in 30 μL of Milli-Q water. In the case of preparing genomic DNA for sequencing on a GridION device, purification was additionally performed on Genomic Tip 20/G columns (Qiagen, Hilden, Germany) according to the manufacturer’s standard protocol. The quality and quantity of DNA were assessed spectrophotometrically in a Nanodrop 1000 device (Thermo Fisher Scientific, Waltham, MA, USA) and using a Qubit fluorometer (Invitrogen, Waltham, MA, USA) using the Qubit ™ dsDNA BR Assay Kit.

To create DNA libraries, the NEBNext^® Ultra™ II DNA Library Prep Kit for Illumina^® (New England Biolabs, Ipswich, MA, USA) was used according to the manufacturer’s protocol. Paired-end sequencing (2 × 150  bp) of the obtained libraries was performed on the high-performance sequencer NovaSeq 6000 (Illumina, San Diego, CA, USA). To carry out the hybrid assembly, sequencing was carried out on a GridION device (Oxford Nanopore Technologies, Oxford, UK) with a Ligation Sequencing Kit according to the manufacturer’s recommendations.

2.2. Chloroplast Genome Assembly and Comparative Analyses

Obtained raw reads were trimmed using the fastp program (v0.23.4) [33], and the resulting data were then mapped to the A. arguta chloroplast genome (NC_034913.1) from the Refseq database [34] using bowtie2 (version 2.4.4) [35] with the local and minimap2 (v2.24-r1122) options [36] for long reads. The mapped reads were then used for the hybrid assembly of chloroplast genomes. The Unicycler (v0.5.0) program [37] with default parameters was used to generate contigs de novo, order them, and concatenate them into supercontigs. Annotation of the three chloroplast genomes was performed using the PGA program (release: 29 October 2020) [38]. The tRNA genes were predicted with tRNAscan-SE (v2.0.12) [39]. Physical maps were generated using the OrganellarGenomeDRAW (OGDRAW) tool (v1.3.1) [40], followed by manual modification. The assembled and annotated complete chloroplast genome sequences of the three species of Actidinia were uploaded to the NCBI database with the following project number: PRJNA1011506.

For comparative analysis, 19 chloroplast genome sequences of representatives of the genus Actinidia, two species of Clematoclethra scandens and Saurauia tristyla from other genera belonging to the family Actinidiaceae, and three species of the genus Vaccinium from the order Ericales were added from the GenBank database. The boundary regions of the chloroplast genome in each species were identified and visualized using IRscope (v.0.1) [41], illustrating their contraction or expansion. To assess the conservation of individual genes, we used mVISTA (online version) [42] and visualized the results through two alignment programs: LAGAN [43], which generates multiple alignments regardless of inversions, and Shuffle-LAGAN, which can identify sequence rearrangements and inversions. This allowed for a comprehensive comparison of gene conservation across species.

We also conducted a selective pressure analysis, comparing 22 species of Actinidia and 6 species from other genera through pairwise comparisons. To identify positive selection, all protein-coding sequences (CDSs) from each of the 28 species were utilized. Alignment of the CDSs for each gene was performed using mafft (v7.427) [44] for each pair. The Ka/Ks ratio was then estimated separately for each of the 79 genes within the Actinidia group. Additionally, a species vs. species Ka/Ks ratio was calculated. The Ka/Ks ratios were determined using the KaKs calculator (v2.0) [45].

3. Results

3.1. Organization of the Chloroplast Genomes of de Novo Assembled Actinidia Species

For A. arguta, about 39,439,070 raw paired Illumina reads and 2,188,589 Nanopore reads were obtained; for A. kolomikta, 68,208,431 paired Illumina reads and 905,244 Nanopore reads; and for A. polygama, 71,186,382 and 873,928 Nanopore reads (full statistics for Nanopore reads are provided in Supplementary Table S1). Complete chloroplast genomes of 3 species of the genus Actinidia (A. arguta, A. polygama, and A. kolomikta) were assembled de novo. They had a 4-part structure and consisted of large (LSC) and small (SSC) single-copy regions separated by a pair of identical inverted repeats (IR) (Supplementary Table S2). The size of the plastome was ∼157,000 bp. 113 genes were annotated, of which 79 were protein-coding genes, 30 tRNA-coding genes (10 of which were represented by two copies), and 4 rRNA genes, each of which was represented by two copies. Maps of the chloroplast genomes are presented in Figure 1 (for A. arguta) and Supplementary Figure S1 (for A. kolomikta and A. polygama).

A comparison was also made for gene composition in the studied samples (Supplementary Table S3). As can be seen from Supplementary Table S3, most of the genes are present in the genomes of chloroplasts of all studied species and cultivars (102 genes in sum). The presence of some genes missing from the initial annotation in A. arguta (psbZ, pafI, rps7, rps11, accD, rps14, and ndhD) was further verified using BLAST. Validation showed that these genes are still present in A. arguta samples, but for various reasons they are not annotated automatically; in other cases, genes may not be recognized by open reading frames (ORF).

3.2. Boundary Regions and Comparative Analysis

By comparing the chloroplast genomes of Actinidia species, the close outgroups Clematoclethra and Saurauia, and another genus in the order Ericales, Vaccinium, we found that gene order near junctions of regions (LSC, SSC, Ira, and IRb) varied greatly, not only between genera, but also within genera and even within species (Supplementary Figure S2). Common characteristic for almost all Ericales species studied are ndfH at the IRb/SSC boundary, trnH at the LSC/IRb boundary, and matK at the IRa/LSC boundary. Actinidia, Clematoclethra, and Saurauia are mostly similar to each other; the main differences are the copy positions of the ycf2 gene and the sequence of several genes in the LSC region, starting with rps3 and ending with rpl23. For comparison, for the genus Vaccinium, ndhB and rps7, trnV on the LSC side, trnH, and rps16 on the IRb side at the LSC/IRb boundary, and rps16, trnH, and psbA at the IRa/LSC boundary.

In the first comparison of boundary regions (Supplementary Figure S2) of the chloroplast genomes of A. arguta, A. kolomikta, and A. polygama, we used only de novo assembled sequences. Separately, we compared the boundary regions for de novo sequences and sequences for the same species from the databases (Figure 2).

The comparison shows that even within the same species, gene order can vary—the sequences from the databases and the de novo assembled A. polygama are similar to most other Actinidia species, as well as Clematoclethra and Saurauia, while the de novo assembled sequences of A. arguta and A. kolomikta are similar to A. zhejiangensis.

LAGAN (Supplementary Figure S3) and Shuffle-LAGAN (Supplementary Figure S4) show that plastid genomes within the genus Actinidia may differ more from each other than from the outgroups Clematoclethra and Saurauia. The highest levels of divergence were found for noncoding regions throughout the sequence, with the most variable being the noncoding regions between rbcL–accD, rpl20–psbB, ndhF–rpl32, and rps15–trn-GUU. The coding regions are largely conserved, with some point variations; the most conserved sequences are tRNAs. High variability was noted for the petD, rpl16, ycf2, ndhA, and rpoA genes. A comparative analysis of the chloroplast genome of some Actinidia species shows that its gene composition is highly conserved. Variations are mainly concentrated in non-coding regions, while coding regions show higher conservation.

3.3. Selective Pressure Analysis

We calculated Ka/Ks ratios, that is, the ratio of nonsynonymous substitutions (Ka) to the rate of synonymous substitutions (Ks) at the species level, by combining all protein-coding genes into a supermatrix (Supplementary Table S4). In Actinidia species and other Ericales studied, the absolute Ka/Ks ratio was approximately ≤ 1 in most cases (range 0.41 to 0.9). This result indicates that, at the level of protein-coding sequences of the entire chloroplast, the studied species are either subject to purifying selection or not subject to selection at all. The exceptions are comparisons of A. chinensis–A. lijiangensis (Ka/Ks = 2.2), A. chinensis–A. deliciosa (Ka/Ks = 2), and A. valvata–A. polygama (de novo assembly) (Ka/Ks = 1.6). However, in these cases, the p-value exceeded 0.05.

Ka/Ks ratios were also calculated for some individual protein-coding genes (Supplementary Table S5). In most cases, the absolute value of Ka/Ks was <1, which may indicate purifying selection or the absence of selection in general if the value was equal to or close to one (concise results are provided in

Table 1. Ka/Ks ratio for chloroplast genes.

Gene	Ka/Ks Ratio
atpA	<1, with some exceptions (e.g., for some Clematoclethra scandens comparisons)
atpB	<1
atpE	<1
atpF	<1, with some exceptions (e.g., for Actinidia tetramera comparisons)
atpH	<1
atpI	<1
ccsA	<1, with some exceptions (e.g., for some Clematoclethra scandens comparisons)
matK	<1, with some exceptions
ndhA	<1
ndhB	<1
ndhC	<1
ndhD	<1
ndhE	<1
ndhF	<1
ndhG	<1
ndhH	<1
ndhI	<1
ndhJ	<1
ndhK	<1
petA	<1
petB	<1
petD	<1
petL	<1
psaA	<1
psaI	<1
psbB	<1
psbC	<1
psbD	<1
psbE	<1
psbH	<1
psbJ	<1, with some exceptions (e.g., for some Saurauia tristyla comparisons with de novo assembled Actinidia cp genomes)
psbL	<1
psbM	<1
psbT	<1
psbZ	<1
rbcL	<1
rpl2	<1
rpl14	<1
rpl16	<1
rpl32	<1
rpl33	<1
rpl36	<1
rpoA	<1, with some exceptions (e.g., for some de novo assembled Actinidia arguta comparisons)
rpoB	<1
rpoC2	<1, with some exceptions (e.g., for some Actinidia rufa comparisons)
rps2	<1, with some exceptions (e.g., for some Actinidia polygama and de novo assembled Actinidia kolomikta comparisons)
rps3	<1, with some exceptions (e.g., for some de novo assembled Actinidia kolomikta comparisons)
rps4	<1, with some exceptions (e.g., for some Actinidia rubus and Actinidia callosa var. henryi comparisons)
rps7	<1, with some exceptions
rps8	<1
rps11	<1
rps12	<1
rps14	<1
rps15	<1
rps18	<1, with some exceptions (e.g., for some Vaccinium corymbosum comparisons)
rps19	<1, with some exceptions
ycf2	>1, with some exceptions
ycf1	<1, with some exceptions

).

As indicated in Table 1, the ycf2 gene had a Ka/Ks value > 1 (with exceptions), which suggests possible positive selection. Nevertheless, in the majority of cases with Ka/Ks < 1, the values also fluctuated around 1, indicating the weakness of selective pressure or even the absence of it. A high Ka/Ks value (>3) was observed for a low number of comparisons, which included the rps4 and ycf2 genes.

4. Discussion

In this study, we sequenced and annotated the cp genomes of three Actinidia species and compared the genomic features of Actinidia cpDNA. A comparison of the cp genomes of samples from the collection of the Main Botanical Garden with sequences of Actinidia, Clematoclethra, Saurauia, and Vaccinium species from the NCBI database revealed not only differences within the order and genus, but also within species.

The de novo assembled cp genomes of A. arguta, A. kolomikta, and A. polygama have a typical circular structure maintaining LSC, SSC, and two IR regions; when compared with genomes from databases, they are conserved in terms of genome length and gene composition. Low Ka/Ks ratios at the chloroplast genome level in Actinidia species indicate that most genes have either been subjected to purifying selection to maintain conserved functions or are not under selective pressure at all.

Most protein-coding genes are also conserved, with a few exceptions. Among these exceptions is the ycf2 gene, which not only exhibits high variability, but also undergoes positive selection (driving change) and changes its position and proximity to the boundary regions on the plastid genome, which is unusual for usually conserved chloroplast genes. The ycf1 and ycf2 genes are the largest chloroplast genes in angiosperms and are essential for the functioning of the plant cell [46]. The presumed function of ycf2 is to provide ATP for the transfer of nuclear preproteins into the chloroplast, and the function of ycf1 is related to the TIC complex that facilitates the transfer of proteins in and out through the chloroplast membrane [47]. Ericales is not the only taxon in which ycf2 undergoes a positive selection process [48,49,50,51] and even at the species level [52]. However, the reason for this and the benefit of this strategy compared to gene transfer to the nucleus remain unclear.

Major differences among the chloroplast genomes we obtained affect intergenic noncoding regions, gene order, and their position relative to the boundaries of LSC, SSC, and IRs. Structural rearrangements in chloroplast genomes occur at different taxon levels, for example, at the genus level [53]. It was previously shown that at the species level, rearrangements do not appear to such a strong degree [54,55]—the main layout remains the same, and only the proximity/distance to the boundary regions changes. Rearrangements are in many cases associated with small, dispersed repeats (SDRs), which are thought to contribute to the double-strand break-induced repair mechanism. SDRs often contribute significantly to repeat space in genomes with highly rearranged gene order and add to structural polymorphism even in closely related lineages [26].

Our study shows that large differences in gene order can exist within the same angiosperm species: in our de novo samples of A. arguta and A. kolomikta, the overall gene order differs from that in the sequences of the same species from the NCBI databases. The general layout of genes near boundary regions of de novo assembled A. arguta and A. kolomikta is similar not to A. arguta and A. kolomikta from NCBI, but to A. zhejiangensis. However, de novo assembled A. polygama is similar to A. polygama from NCBI.

There are certain patterns in the arrangement of genes relative to the boundaries of regions in the family Actinidiaceae (Actinidia, Saurauia, and Clematoclethra) that support a division of cp genomes into two groups (Figure 3, images near the round bracket).

Between the genus Vaccinium and the family Actinidiaceae, there is an “intermediate” form—the cp genome of A. lijiangensis. Other studied sequences can be divided into two main types according to the position of genes near the LSC/IRb boundary.

The discovery of two main types of cp genomes among the genus Actinidia, together with the presence of these two types in representatives of the same species, suggests that at the level of the genus or even the family Actinidiaceae, the presence of several types of cp genomes is consolidated, which can be rearranged further by SDR. Presumable mechanisms for the fixation of several types of cp genomes may be the heteroplasmy of Actinidia (due to the supposed mixed inheritance of chloroplasts) and the rapid elimination of one of the types during selection or even at the stage of development of the organism. The identification of potential intermediate forms of Actinidia with heteroplasmy, the assembly and annotation of cp genomes of other genera of the family Actinidiaceae, and the mechanisms of maintaining the presence of two types of cp genomes require further study.

Such large structural changes can greatly complicate phylogenetic analysis, and additional barcode or sequencing studies are required to confirm the exact cpDNA sequence. This study shows that in the case of the genus Actinidia, assembly and annotation of the cp genome of a single plant may not be sufficient to characterize the cp genome of an entire species—multiple biological replicates from different spatial ranges must be examined to obtain complete genomic data about species.

5. Conclusions

In this study, we provide an assembly and annotation of the cp genomes of three Far Eastern Actinidia species (A. arguta, A. kolomikta, and A. polygama) and also compare the cp genomes of different Actinidia species and some other representatives of the order Ericales. The assembled genomes have a typical quadripartite structure and are conserved in the number of genes, but not in their order—we show that in angiosperms, the order of genes in the cp chromosome can differ even at the intraspecific level. In addition, we identified a highly variable protein-coding gene (ycf2) that undergoes a positive selection process in the order Ericales. These results expand our understanding of cp genomics and genetic diversity within the genus Actinidia, providing a basis for future research in molecular marker development, phylogenetic analysis, and population studies.