Comparative analyses of chloroplast genomes in ‘Red Fuji’ apples: low rate of chloroplast genome mutations

Plant materials

Germplasm of 24 Red Fuji apples was collected from nine regions in China. Two or three accessions were selected from each region for repeated experiments. The sources and quantities of the accessions are listed in Table 1. Fresh leaves of Red Fuji were dried in silica gel and taken to the laboratory.

DNA extraction and sequencing

Total genomic DNA of each material was extracted using the plant genomic DNA secure Kit (DP320) of China Tiangen Biotechnology (Beijing) Co., Ltd. The integrity and purity of the DNA samples were determined by agarose gel electrophoresis. NuoheZhiyuan Technology Co., Ltd. (Beijing, China) used the Illumina Hiseq 2000 sequencing platform to perform genome sequencing of quality DNA samples. The read length was 150 bp. A total of 492.617 G of raw data was generated by sequencing, and the filtered clean data was 491.884 G, with an average of 20987.043M per sample (Table S1).

Plastome assembly and annotation

We chose a reference map for assembly. The obtained clean data were assembled using Novoplasty software (https://github.com/ndierckx/NOVOPlasty), the configuration file config.txt needs to be set before use. Download the published chloroplast genome of Malus yunnanensis (GB: MH394390) from NCBI as the reference file, select its rbc L gene as the seed file, kmer = 39. The preliminary annotation of the assembled plastome was performed using Geseq (https://chlorobox.mpimp-golm.mpg.de/geseq.html). Then, Sequin 16.0 was used to correct the start and stop codons and intron/exon boundaries for the genes with software annotation errors. Finally, the complete cp genome sequence and its annotations were submitted to GenBank. The specific serial numbers are detailed in Table 1.

Characterization and comparative analyses of chloroplast genomes

The total length of the genome, the length of each region (large single-copy regions, small single-copy regions, inverted repeat), gene composition, base composition and GC (AT) content were calculated using Geneious 11.1.2 software, The characteristics of Red Fuji genomes were analyzed.

Boundary regions and comparative analysis

Different plant species have different gene sequences in the four junction regions. The change of cp genome length is the main reason for the contraction and expansion of the IR region (Raubeson et al., 2007). The border regions between large single copy (LSC) and IR regions, and between small single copy (SSC) and IR regions, were compared using IRscope (https://irscope.shinyapps.io/irapp/).

Codon usage analyses

Codon usage bias (CUB) refers to the phenomenon that some codons are used more than other synonymous codons in the process of gene translation between different species or within the same species (Alexandra & Tamir, 2014). CUB is a useful tool for understanding genetic and evolutionary processes. In this study, MEGA7.0 software CodonW was used to analyze codon preference, The results of the analysis are shown in charts.

Repeat sequence and SSRs analyses

The online REPuters (https://bibiserv.cebitec.uni-bielefeld.de/reputer/) software was used to identify repeats which contain forward, reverse, and complex, repeats as well as palindromes of the cp genome of Red Fuji (Frazer et al., 2001). The following parameters were used to identify repeats with REPuter: Hamming distance of 3, Maximum Computed Repeats of 50, and repeat size >30 bp. Simple sequence repeats (SSRs) were checked using MISA (https://webblast.ipk-gatersleben.de/misa/index.php), with motif sizes of 1–6 nucleotides (Thiel et al., 2003). All other parameters were used as default values. The parameters of repetitive units and the minimum number of repetitions were set to 10 repetitions for single nucleotide type, 6 repetitions for dinucleotide type, 5 repetitions for trinucleotide type, 5 repetitions for tetranucleotide type, 5 repetitions for pentanucleotide type, and 5 repetitions for hexanucleotide type.

Genome comparison

The cp genomes of Red Fuji apples from nine regions were compared and visualized by mVISTA (http://genome.lbl.gov/vista/index.shtml) online software. mVISTA is an online tool used for multiple DNA sequence alignments, where sequence similarity can be evaluated by comparing coding regions with non-coding regions, introns, and exons (Frazer et al., 2004). The genome of M. sieversii was selected as the reference, and the input file was the original FASTA format nucleotide sequence file and the gff3 format annotation file. The nucleotide diversity (PI) of Red Fuji was calculated in Dnaspv6 software. SNP and indel detection was carried out using Geneious 11.1.2.

Phylogenetic analysis

A phylogenetic tree was constructed based on the combined cp genomes. Pyrus pyrifolia was used as the outgroup. The species and the accession numbers of their cp genomes in NCBI are listed in Table S2. The sequences were aligned using Mafft (https://mafft.cbrc.jp/alignment/server/). The model is GTRGAMMA+I. Maximum Likelihood (ML) methods were used to construct phylogenetic trees was used to perform the RAxML-HPC BlackBox 8.2.12 in CIPIES (https://www.phylo.org/) (Salichos, Stamatakis & Rokas, 2014) the ML analysis.

Characteristics of Red Fuji cp genomes

The 24 cp genomes of the Red Fuji apples collected were highly conserved in terms of gene content, gene order, and gene intron number. The genomes ranged in length from 160,063 bp to 160,070 bp, and as expected, the cp genomes contained the LSC and SSC, separated by a pair of IR regions (Fig. 1). The length of the IR region was 26,307–26,308 bp, the LSC region was 88,272–88,274bp, and the SSC region was 19,176–19,181 bp. The overall GC content was approximately 36.6%. The GC content in the IR region was 42.7%, which was higher than 34.2% in the LSC region and 30.4% in the SSC region (Tables 2 and 3). These numerical values were consistent among the 24 accessions and did not change.

Each plastome encoded 131 genes, including 84 protein-coding genes (PCGs), 37 transfer RNA (tRNA) genes, and eight ribosomal RNA (rRNA) genes. There were 22 introns in the annotated genes. Of the annotated genes, 18 embodied one intron, and two (ycf3 and clpP) had two introns. Additionally, 62 protein-coding and 22 tRNA genes were located within the LSC; 12 protein-coding genes, 14 tRNA coding genes, and 8 rRNA coding genes were located within IRs, and 12 protein-coding genes and one tRNA gene were located within SSC. The genes in the LSC region accounted for 64.1% of the cp genome, the two IR regions accounted for 26%, and the genes in the SSC region accounted for 9.9%.

IR boundary analysis

First, IR analysis was performed on the materials from each region, and the results showed that the cp genomes of Red Fuji in the same region were highly consistent (Figs. S1–S9). Therefore, the subsequent analysis was based on comparing the nine regions Nine samples were selected from each region to compare the distribution of genes in the boundaries of the four regions of the cp genome. The IR regions of the nine Fuji genomes sequenced were highly conserved (Fig. S10). We compared gene variation at the boundaries of five plastosomes of three Red Fuji, one M. sieversii (MK434920), and one M. sylvestris (MK434924). The results displayed that the boundaries of the four regions were relatively conserved, and the gene types distributed in the boundary region were highly consistent. The genes distributed in the boundary between the LSC and IR region were rps 19 and trn H-GUG. In addition, rp1 2, ycf 1, and ndh F are distributed in the boundary of the SSC/IR region. (Fig. 2). The length of ycf 1 extending to the IR region was 1,074 bp, and that of ndh F was 12 bp. The length of rps 19 extending to the IRb region extended was 69 or 115 bp, depending on the different species. The position of the gene in the boundary region was relatively fixed. There were no differences in the boundaries of the five samples. Hence, these results indicate that the IR region is highly conserved.

Codon usage analyses

Since the base composition and AT/GC content was the same in the 24 genomes of Red Fuji, we chose one of them to calculate the frequency of amino acid and codon usage. The codon usage results of materials from Wunan Town, Wuwei City showed that a total of 20 amino acids (excluding the stop codon) were encoded, and the usage frequency of each amino acid ranged from 1.56% to 9.38%. Leucine, serine, and arginine were the most abundant amino acids. Among the encoded amino acids, except methionine and tryptophan, other amino acids were encoded by two to six synonymous codons (Fig. 3). For example, serine, arginine, and leucine were encoded by six synonymous codons, of which the most frequently used codons were TTA, TCT, and CGT, respectively. Four synonymous codons were used for valine, proline, threonine, alanine, and glycine. The most frequently used codons were GTT, CCT, ACT, GCT, and GGT, respectively. Three synonymous codons were used for isoleucine, of which ATT was the most frequently used codon. Two synonymous codons were used for phenylalanine, tyrosine, histidine, glutamine, asparagine, lysine, aspartic acid, glutamic acid, and cysteine, which had different frequencies of use. The codons with a higher frequency for these nine amino acids were TTT, TAT, CAT, CAA, AAT, AAA, GAT, GAA, and TGT, respectively. The cp genome of Red Fuji apple has 32 codons with RSCU >1, of which 30 codons end with A/T, accounting for 93.75%. Most of the 29 codons with relative synonymous codon usage values (RSCU)<1 ended in G/C, accounting for 96.55%. Our results were similar to those of other angiosperms Qian et al. (2013). The codons of the cp genome of Red Fuji that were used frequently at the third codon positions were A and T.

Repeats and microsatellites analyses

A total of 49 repeat sequences were detected in each material, which consisted of 26 forward, 17 palindromic, four reverse, and two complement repeats (Table S3). Among them, there are 33 repeats with a length of 30–39 bp, 6 repeats with a length of 40–49 bp, and two repeats with more than 50 bp (Fig. 4). The most common type was forward repeats (53.1%), and the least common was complement-type repeats (4.1%). Using the MISA software, we searched for the SSR loci in the 24 Red Fuji apple cp genomes. A total of 1,380 microsatellites, two dinucleotide repeats, 46 to 48 mononucleotides repeats, and 8 or 9 complex polynucleotides were found. Mononucleotides (81.2%), mainly poly-A (polyadenine) and poly-T (polythymine), formed the largest proportion of SSRs in the Red Fuji cp genome. The majority of SSRs were situated in the intergenic region of LSC (Table 4).

Genome comparison and divergence analyses

In this study, the online software mVISTA was used for sequence alignment and variation analysis of cp genome of Red Fuji, using M. sieversii (MK434920) as the reference. The nucleotide sequence similarity of the ten cp genomes was extremely high, suggesting that there was no variation in the cp genome of Red Fuji compared with its ancestral species (Fig. 5). At the same time, it can be found that divergence existed in the highly conserve dregions, and the coding region and IR region were more conserved. The intergenic regions with the highest levels of divergence were those in between psb L-atp A, psb M-psb D, and ndh c-atp E.

The nucleotide polymorphisms were analyzed using the DNAsp software, and the nucleotide diversity (PI) was 0. Geneious was used to detect nine regions’ single nucleotide polymorphisms (SNPs) of M. sieversii, M. sylvestris, and Red Fuji apples in nine regions, Only two indel loci were detected in the rps 16 and ndh A genes. The results suggested that the rate of variation of Red Fuji in different regions was rarely low.

Phylogenetic analysis

Phylogenetic analysis using Maximum likelihood (ML) yielded well-supported tree topologies (Fig. 6). All red Fuji are clustered into one branch, M. sieversii and M. sylvestris, also clustered into a branch, that diverged earlier than Red Fuji, suggesting that they are the ancestors of the cultivated species.