The P. atropurpura leaves were collected from the campus of Guizhou University (26°25^′39.62^″N, 106°40^′5.81^″E). Healthy and tender leaves were selected, washed 3–5 times with distilled water, dried, and stored in a −80°C refrigerator for later use.

The leaves were rapidly frozen in liquid nitrogen and ground into powder, then their genomic DNA was extracted using the modified CTAB method [15]. Preparation of 350 bp DNA fragments was done using a Covaris ultrasonic crusher, followed by end repair and tail addition. The construction of the sequencing library was completed and its sequencing was performed on the Illumina HiSeq X Ten platform using a Paired-End (PE) PE150 sequencing strategy. We obtained 6.75 G of raw data with high throughput sequencing, removed joints and low-quality data regions, and accumulated 27,130,560 clean reads. Using NOVOPlasty was to splice chloroplast genomes with default parameters [16].

The annotation of the P. atropurpura chloroplast genome was performed using the Plastid Genome Annotator (https://github.com/quxiaojian/PGA) (accessed on 13 January 2025) [17], and manually adjusted; tRNA annotation was made using tRNAscan-SE (https://trna.ucsc.edu/tRNAscan-SE/) (accessed on 13 January 2025) and ARAGORN tools (https://packages.debian.org/bullseye/aragorn) (accessed on 13 January 2025); and a circle diagram was generated using Organellar Genome DRAW [18].

The CodonW1.4.2 software [19] was used to statistically analyze the codon preferences of the P. atropurpura chloroplast genome. Using the Reputer software [20], scattered repeat sequences were detected, with a minimum 30 bp repeat length, a minimum permutation value of 50, and a maximum base mismatch of 3. MISA software [21] was made to analyze simple repetitive sequences in the P. atropurpura chloroplast genome, the minimum repeat value for single nucleotides was set to 10, for dinucleotides to five, for trinucleotides to four, and tetra-, penta-, and hexanucleotides to three.

The chloroplast sequences of seven representative Chinese medicinal plants from the family Ericaceae, including Pyrola atropurpurea (PP473790), Pyrola rotundifolia (KU833271.1), Chimaphila japonica (MG461316.1), Gaultheria sinensis (OM048872.1), Agapetes malipoensis (NC_058759.1), Rhododendron simsii (MW030509.1), and Vaccinium bracteatum (LC521967.1) were downloaded from GenBank. The contraction and expansion of LSC, SSC, and IR region boundaries were visualized in the Ericaceae chloroplast genomes using IRscope software [22]. The chloroplast genome rearrangement and collinearity in Ericaceae species were detected using the Mauve multiplex genome alignment method in Geneous10.2.2 software [23]. The Ka/Ks values were calculated with Ka/Ks Calculator v2.0 (https://sourceforge.net/projects/kakscalculator2/) (accessed on 13 January 2025). The Pi value of chloroplast protein-coding genes (PCGs) was analyzed using DnaSP software [24].

The phylogenetic relationship was conducted on all 41 chloroplast genomes in the Ericaceae family, using Magnolia officinalis (NC_020316.1) as an outgroup. PCGs alignment was performed using the MAFFT website (https://mafft.cbrc.jp/alignment/server/index.html) (accessed on 13 January 2025) [25], all missing sites were filtered out using TBtools software [26], the optimal model was calculated using ModelTest NG software [27], and this model was determined on the Akaike information criteria. A phylogenetic tree was built using RAXML-n [28] based on the maximum likelihood (ML) method, and the tree-building model was GTR+I+G4.

The total length of the P. atropurpura chloroplast genome (GenBank accession number: PP473790) is 172,535 bp, with a GC value of 34.95%. The genome has a typical tetrad structure, that is, the entire circular genome can be partitioned into a large single copy (LSC), a small single copy (SSC), and two inverted repeats (IR) regions (Fig. 1). The LSC, SSC, and IR region lengths are 105,081, 11,700, and 27,877 bp, with GC values of 34.72%, 27.99%, and 38.74%, respectively. In total, 132 genes were annotated comprising 43 tRNA, 8 rRNA, 76 PCGs, and 5 pseudogenes (Table S1). In tRNA, trnA-UGC, trnC-GCA, trnH-GUG, trnI-CAU, trnI-GAU, trnL-CAA, trnL-UAG, trnN-GUU, trnR-ACG, and trnV-GAC each have two copies. And trnA-UGC, trnG-GCC, trnI-GAU, trnK-UUU, trnL-UAA and trnV-UAC each own one intron. There are four types of rRNA, each with two copies, situated in the IR region. In the encoded protein genes, there are two copies each of ribosomal protein subunit rpl32, and rps7, and unknown functional protein ycf15, atpF, clpP, rpoC1, ndhB, petB, petD, rpl16, rpl2, rps12, and rps16 all own one intron, while ycf3 has two introns (Table S2). In the deeply explored chloroplast genome, all ndh genes (ndhK, ndhD, ndhG, ndhA, and ndhH) except ndhB, are pseudogenes.

In total, 4223 scattered repeat sequences were predicted in the P. atropurpura chloroplast genome (Fig. S1). Among them, there were 2452 forward repeats (58.06%), 1763 palindromic repeats (41.75%), and eight reverse repeat (0.19%), and no complementary repeats were found. And 358 SSRs were checked in the P. atropurpura chloroplast genome (Fig. 2), including 181 single nucleotide repeats, 14 dinucleotide repeats, 139 trinucleotide repeats, 13 tetranucleotide repeats, 3 pentanucleotide repeats, and 8 hexanucleotide repeats. Most SSRs are positioned in the LSC region (244), with only a few were distributed in the SSC (78) and IR region (36) (Fig. S2). In addition, the majority of SSRs are located in 260 intergenic regions (72.62%), followed by 14 in gene coding regions (3.91%) and 84 in intron regions (23.46%), indicating that SSRs are mainly distributed in intergenic regions (Fig. S3). These SSRs are mainly single base repeats composed of A or T, indicating that the SSRs of this genome have a strong preference for A and T.

The relative usage of synonymous codons (RSCU) analysis is based on P. atropurpura chloroplast genome, 71 coding DNA sequences (CDS) larger than 200 bp were obtained. The research results indicate that the chloroplast genome contains a total of 16,561 codons. Among them, there are 1721 codons encoding leucine (Leu), accounting for the highest proportion of 10.39%. The codon encoding cysteine (Cys) is 188, accounting for the smallest proportion at 1.14% (Table 1). At the same time, 31 codons with RSCU greater than 1 were detected, with all codons ending in A/U except UUG and AUG. The RSCU value of codon AUG encoding methionine was the highest, at 6.7991 (Table 1 and Fig. S4).

The nucleotide polymorphism (Pi) value of chloroplast protein-coding gene sequences of seven Pyrola species was analyzed using DnaSP software (Fig. 3). The complete length of the aligned sequences was 57,658 bp, and a total of 5213 polymorphic sites were discriminated. The Pi value ranged from 0 to 0.3019, with an average of 0.0479. Among the seven highly variable hotspots that were identified (Pi value > 0.11), two genes (trnT-UGU, and trnF-GAA) are situated in the LSC region, and five genes (ccsA, psaC, ndhE, ndhI, and rps15) are seated in the SSC region. No nucleotide polymorphism sites were detected in the IR region, demonstrating that the nucleotide polymorphism in the LSC and SSC regions is significantly higher than in the IR region.

To compare the chloroplast genome differences between P. atropurpura and the representative group of Ericaceae species, we calculated the basic information of chloroplast genomes (Table 2). The P. atropurpura chloroplast genomes and its six closely related species had significant differences, with a total length range of 151,656 bp (Chimaphila japonica) to 176,632 bp (Gaultheria sinensis), all of which are typical tetrad structures. The length compositions of LSC, SSC, and IR in the above-mentioned genome are 78,997 to 109,173 bp, 2979 to 11,946 bp, and 1717 to 33,332 bp, respectively. The total number of genes is 102 to 146, PCGs are 63 to 94, tRNA genes are 4 to 8, and rRNA numbers are 30 to 43. In the variation of GC content, the GC values ranged from 35.0% to 36.8%. Boundary analysis showed obvious differences in the transition regions of the four boundary zones (Fig. 4), which further demonstrated the significant differences in Ericaceae chloroplast genome sequences.

The seven Ericaceae chloroplast genomes were compared with the Mauve software (https://github.com/MauveSoftware/novu) (accessed on 13 January 2025). It was found that the chloroplast genome sequences had high variation, with higher variation in non-coding regions than in coding regions, and higher variation in LSC and SSC regions than in IR regions. Compared to other Ericaceae species, the sequence variation of Chimaphila japonica and Rhododendron simsii is relatively high. The gene number and order of the chloroplast genome in Ericaceae are variable, and an obvious gene rearrangement phenomenon was observed (Fig. 5).

To investigate the evolutionary characteristics of the Ericaceae family in the chloroplast genome, Ka/Ks calculations were performed on 68 common PCGs of the P. atropurpura chloroplast genome and its six closely related species (Fig. 6). The genes without numerical values in Fig. 6 indicate that the Ka/Ks value is zero. Most genes related to the photosystem system, such as petA, psbA, atpA, atpB, atpE, atpI, rbcL, psbE, psbB, psbC, psbD, psbF, ndhB, ndhJ, etc., have Ka/Ks values less than 1, showing that these key genes have been purified and selected. The above genes play a very important role in the functioning of the chloroplast genome and are therefore relatively conserved in evolution. Moreover, the Ka/Ks values of rpl33 and rps16 genes were higher among the five species, demonstrating that these genes were positively selected.

Based on the PCGs of chloroplast genome, a phylogenetic tree of 41 Ericaceae species was built using the ML method of the IQ-TREE software (Fig. 7). Magnolia officinalis (NC_020316.1) was set as the out-group. And the Ericaceae species were mainly classified into four cluster groups, namely Cluster Groups I, II, III, and IV. The three species of the genera Pyrola and Chimaphila are distributed in Cluster I. Cluster Group II is composed of thirteen species in the genus Rhododendron. The thirteen species of the genus Gaultheria are located in Cluster III. And Cluster Group IV is mainly made up of the species in the genus Vaccinium. The phylogenetic result indicated that P. atropura, Pyrola rotundifolia and Chimaphila japonica are clustered together with a 100% support rate, and they have the closest genetic relationship.