SSR marker development for analysis of the genetic diversity and identification of species and infraspecific ranks in the genus Phyllostachys

Abstract

Bamboo plants possess important ecological, economic, and cultural values. However, it is difficult to accurately identify them on the basis of their morphological traits alone. Here, based on the whole-genome data of moso bamboo (Phyllostachys edulis) and its 20 forms, we conducted preliminary identification and comparative analyses of simple sequence repeats (SSRs) to develop molecular markers. In total, 3,835,632 SSR loci were identified from 31,537.81 Mb of genomic sequences, among which dinucleotide SSRs were the most abundant. Most SSRs were located in intergenic regions, whereas relatively fewer were in genic regions. In addition, we found that SSR-containing genes involved in plant hormone signal transduction may be associated with the morphogenesis of moso bamboo, which was speculated to be related to differential gene expression patterns among different forms. Furthermore, 206 SSR primer pairs with polymorphisms were obtained to analyse the genetic diversity of moso bamboo and its forms, which exhibited moderate polymorphism. The proportion of genetic variation among species within the genus Phyllostachys was 58%, while that within species was 42%. Moso bamboo and its 20 forms had relatively close genetic relationships and low genetic differentiation, while 20 species of the genus Phyllostachys were clustered into three groups with distinct levels of genetic diversity. Finally, DNA fingerprints and molecular identity cards were constructed for 20 moso bamboo forms and 20 species of the genus Phyllostachys using core SSR markers. These results provide novel SSR markers for bamboo identification, germplasm conservation, and molecular marker-assisted breeding.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09292-1.

Introduction

Bamboos belong to the grass family Poaceae, subfamily Bambusoideae, and are characterized by rapid growth, a short maturation cycle, and sustainable utilization potential [1]. There are 837 bamboo species belonging to 39 genera in China, accounting for 51% of the world’s total bamboo species [2], providing abundant resources for germplasm innovation and new variety breeding. Moso bamboo (Phyllostachys edulis) is a representative large-sized evergreen monopodium bamboo that has ornamental, edible, economic, and ecological value and has great development potential and application prospects [3, 4]. Moso bamboo forests cover 5.28 million ha, accounting for 69.78% of the national total bamboo forest area in China [2]. During the long evolutionary process, abundant morphological variations have emerged within moso bamboo, with particularly significant differences in morphological traits such as culm shape and culm colour [5, 6]. As ornamental bamboos, the forms of moso bamboo significantly increase the ornamental value of landscapes [7].

However, bamboo relies primarily on asexual reproduction, resulting in a high degree of genetic similarity among its infraspecific ranks [8, 9]. Variety identification relies mainly on morphological and physiological indicators. For instance, the phenotypic and physiological characteristics of six different types of moso bamboo were investigated, and the results revealed that they exhibited differences in culm morphology, size, and colour, with significant variations in physiological indicators [10]. Nine morphological traits were identified as markers for colour and shape [11]. However, because the morphological characteristics of the infraspecific ranks of moso bamboo highly overlap, it is difficult to distinguish varieties with close origins or precisely elucidate the extent of genetic differentiation and genetic relatedness among infraspecific ranks, leading to phenomena such as homonymy and synonymy [12, 13]. Furthermore, constrained by the biological characteristics of bamboo species, progress in bamboo breeding is quite slow, with fewer than 50 new varieties authorized to date in China—none of which have undergone Distinctness, Uniformity, and Stability (DUS) testing. Therefore, the development of appropriate molecular markers will provide robust technical support for the DUS testing of new varieties of bamboo.

Currently, a variety of molecular markers have been developed and widely used in the construction of genetic maps, variety identification, gene localization, and population genetic analysis, among which simple sequence repeat (SSR) molecular markers can reveal genetic variations in functional genes of different organisms via analysis of DNA fragments containing unique genetic information [14–16]. SSR markers have become indispensable tools in genetic research in recent decades because of their high degree of polymorphism, codominance, reliability and other core advantages. Even in the era of rapid development of high-throughput sequencing, SSR markers still have important value and status in many application scenarios (such as variety-specific protection, forensic identification, and routine analysis in small- and medium-sized laboratories) [17–19]. For plants that have been maintained in a state of vegetative propagation for a long time, such as bamboos, SSR markers exhibit superior performance in terms of germplasm identification because of their inherent technical advantages and their suitability for the biological characteristics of asexual reproduction. Hence, SSR markers remain valuable.

Although a variety of molecular marker techniques have been applied to bamboo, such as random amplified polymorphic DNA (RAPD) [20], amplified fragment length polymorphism (AFLP), and intersimple sequence repeat (ISSR) [21], SSR molecular marker technology still has a place in bamboo. For instance, the genetic diversity of 10 cultivated forms of moso bamboo was investigated using 43 pairs of SSR primers, which revealed that there was no SSR genetic diversity among the cultivated forms [22]. This might be due to the small number of selected SSR loci with insufficient coverage or the influence of other mechanisms, which may have resulted in a failure to detect differences. Therefore, when studying moso bamboo forms or cultivated varieties with extremely similar genetic backgrounds, it is necessary to develop an adequate number of molecular markers with high resolution to achieve precise differentiation. With the continuous development of genome resequencing technology, the development of a large number of SSR loci has become more convenient. There have been reports on the development and application of SSR markers based on whole-genome resequencing of bamboo [23, 24], but reports on SSR markers at the infraspecific level are lacking.

In this study, on the basis of the published whole-genome sequence of moso bamboo and the sequences of its 20 forms newly obtained in our laboratory, a preliminary comparative analysis of SSR loci and their characteristics was conducted, after which functional enrichment analysis was performed on genes containing SSR loci, whose expression patterns were analysed using newly obtained transcriptome data to explore their potential functions. Furthermore, molecular markers were developed for polymorphic SSR loci located in genic regions, which were used to evaluate the genetic diversity of the infraspecific ranks of moso bamboo and bamboo species within the genus Phyllostachys. Finally, on the basis of the screened core SSR markers, DNA fingerprints and molecular identification cards were constructed for the 20 moso bamboo forms and 20 species of the genus Phyllostachys. The results provide technical support for subsequent studies on bamboo, such as interspecific and infraspecific rank germplasm identification, germplasm conservation, marker-assisted selection, and resource management and utilization, thereby providing a foundation for the sustainable development of the bamboo industry in the future.

Materials and methods

Plant materials and DNA extraction

A total of 20 leaf samples of moso bamboo forms were collected from the Jiangxi Academy of Forestry (Nanchang, China), the Zhejiang Academy of Forestry Sciences (Hangzhou, China), China Bamboo Expo Park (Huzhou, China) and the Taiping Experimental Base of the International Centre for Bamboo and Rattan (Huangshan, China). The samples were used for genome resequencing [6] and SSR loci identification and included 3 samples from Nanchang, 3 from Hangzhou, 3 from Huzhou and 11 from Huangshan (Table S1). In addition, a total of 40 bamboo germplasms, including 20 moso bamboo forms and 20 different species in the genus Phyllostachys (Table S2), were used for genetic diversity analysis. The genomic DNA (gDNA) of the leaves was extracted using a modified CTAB method [25]. The quality and concentration of the DNA were measured using a NanoDrop 2000. Standardization of DNA sample concentrations was performed for all samples.

Identification and characterization of SSR loci

All SSR loci were identified from the genomic sequences of moso bamboo (Ph. edulis data from BambooGDB) [26] and its 20 forms (PRJNA1065936) using the microsatellite identification tool (MISA) [27] with the following parameters: a minimum number of 12 repeats for the mononucleotide motif; 6 repeats for the dinucleotide motif; 5 repeats for the trinucleotide motif; 4 repeats for the tetranucleotide and pentanucleotide motifs; and 3 repeats for the hexa-, hepta-, octa-, nona- and decanucleotide motifs. These SSR sequences were subsequently compared with those of moso bamboo to identify polymorphic SSR loci.

Primer design for SSR markers

A set of primers was designed for all SSR candidate loci using Primer3 [28] to amplify 100–300-bp products with 40–60% GC content. The primer lengths ranged from 18 to 27 bp, with an optimal length of 22 bp. The melting temperature™ range was 55–65 °C (Table S3).

Functional annotation and enrichment analysis

To further understand the functions of the genes containing SSRs, Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was performed for all genes containing SSRs. Furthermore, the functional annotation of selected genes involved in plant hormone signal transduction pathways was conducted using the localized Pfam database [29].

RNA-Seq analyses

Shoot samples of Ph. edulis f. tubaeformis (SY) and moso bamboo (Ph. edulis, PE) were collected at the Lingfengsi forest farm, An’ji County, in Zhejiang Province. The 18th internodes of SY and PE shoots at different heights (1.8 m and 2.8 m) on the same rhizome were selected and harvested in April 2025 according to previous studies [30, 31]. Each 18th internode was divided into two parts: the upper part of the 18th internode (SYU and PEU) and the lower part of the 18th internode (SYD and PED). Afterwards, the samples were immediately frozen in liquid nitrogen and used for RNA-Seq and RT‒qPCR. Moreover, the RNA-seq datasets (PRJNA1065934) of Ph. edulis ‘Kikko-chiku’ (GJ) obtained by our laboratory [6] were used for expression analysis of the genes containing SSR loci in the samples, including the basal internode (the first variation internode, GJ1), the end variation internode (GJ3), the end variation internode forwards of the fifth internode (variation internode, GJ2), the end variation internode backwards of the fifth internode (normal internode GJ4), and the corresponding top, middle, and basal parts of the shoots without internode variation derived from the same rhizome of the selected GJ, named PET, PEM, and PEB, respectively. Each sample had three biological replicates.

Total RNA was extracted from the collected samples using a Genstone Biotech Kit (Beijing, China). A total of 500 ng of RNA from each sample was pretreated with DNase I and used for reverse transcription. In total, 33 sets of transcriptome data were obtained using the Illumina NovaSeq 6000 platform. HISAT2 (v 2.2.0) was used to align the paired-end reads to the PE reference genome. Transcript quantification was performed in terms of fragments per kilobase of transcript per million mapped reads (FPKM). Differential expression analysis was performed using DESeq2 (v 1.38.0), and genes with a fold change (FC) ≥ 2 and a false discovery rate (FDR) < 0.05 were considered to be differentially expressed.

Real-time quantitative PCR (RT‒qPCR)

Premier 5.0 software was used to design the specific primers according to the screened gene sequences (Table S4). RT‒qPCR was performed with Roche LightCycler 480 SYBR Green I Master Mix (Roche, 04887352001, Germany) in a qTOWER 2.2 system (Analytik Jena, Germany) with three independent experiments, using PeGAPDH [9] as the internal control. The relative abundance of each gene was calculated from the 2^−∆∆CT values relative to the reference gene [32].

Determination of endogenous phytohormone contents

The endogenous phytohormone contents of variant internodes (SYD and SYU) and normal internodes (PED and PEU) were measured for three biological replicates using a high-performance liquid chromatography‒mass spectrometry method described by Simura [33]. A total of four types of plant hormones were tested, namely, auxins, cytokinins (CKs), brassinosteroids (BRs) and gibberellins (GAs). In addition, the endogenous phytohormone contents of GJ and PE were previously obtained by our laboratory [6].

SSR polymorphism assessment

The genomic DNA (gDNA) of PE and SY was selected as a template to validate the availability of SSR primers. For standard PCR, the reaction mixture was 10 µL in volume and consisted of 5 µL of 2× Taq Plus Master Mix II (Dye Plus) from Vazyme Biotech Co., Ltd. (Nanjing, China), 1 µL of gDNA (40–50 ng µL^− 1), 0.5 µL (10 mmol L^− 1) each of the forwards and reverse SSR primers and 3 µL of ddH₂O. PCR amplification was conducted in a thermocycler according to the following protocol: 95 °C for 3 min, followed by 34 cycles of 95 °C for 25 s, the appropriate annealing temperature for 25 s, 72 °C for 18 s, and a final extension at 72 °C for 3 min, after which the PCR products were analysed by electrophoresis on a 2.0% agarose gel.

Clear and single bands of PCR products with the expected size range were selected for bamboo genetic diversity analysis. The TP-M13-SSR PCR method [34] was used to test the polymorphisms of the selected SSR primers among moso bamboo and its 20 forms. M13 primers fluorescently labelled with 5-carboxy-x-rhodamine, 5-carboxy-fluorescein, tetramethyl-5-carboxyrhodamine, or 5-hexachlorofluorescein were employed to label and detect polymorphisms. The touchdown PCR program was as follows: initial denaturation at 95 °C for 3 min; 10 touchdown cycles of 95 °C for 28 s; annealing starting at 65 °C for 30 s with a decrease of 1 °C per cycle and 72 °C for 30 s, followed by 34 cycles of 95 °C for 30 s, annealing at the optimal Tm for 30 s, 72 °C for 30 s; and a final extension at 72 °C for 5 min. The final products were subjected to SSR analysis using capillary electrophoresis detection on an ABI 3730 DNA Analyzer (Rui Biotech Inc., Beijing, China). In addition, the polymorphic SSR loci used for the genetic diversity analysis were verified through independent assays.

Polymorphism of SSR loci was analysed using capillary electrophoresis data. Using PE as the reference, we first determined the position of the major peak for each locus in the control sample, which corresponded to the fragment length of the amplified repeat sequence. The position of the major peak for each locus in the other samples was determined on the basis of the peak profile and its relative position and subsequently compared with that of the control sample. If the peak profiles were identical but the number of differential bases does not correspond to a multiple of the repeat unit, the locus was considered to be nonpolymorphic.

Data analysis and construction of molecular identity (ID) cards

For correlation analysis, the Benjamini–Hochberg false discovery rate (FDR) method was employed to correct for multiple testing, and only correlations with an adjusted p < 0.05 were considered to be statistically significant. The results of capillary electrophoresis were analysed using GeneMarker ver. 2.2.0 software [35]. Hardy–Weinberg equilibrium (HWE) tests were conducted for 206 SSR loci using Genepop ver. 4.8. Exact p-values were estimated using Markov Chain Monte Carlo (MCMC) simulations with the following parameters: dememorization = 10,000, batch number = 20, and batch length = 5,000. The standard error (S.E.) for each locus was controlled to be less than 0.01. PowerMarker ver. 3.25 and PopGene32 ver. 1.32 were used to calculate genetic diversity data and perform unweighted pair group method with arithmetic mean (UPGMA) cluster analysis for bamboo accessions [36, 37]. Afterwards, a genetic distance heatmap was drawn using R software, and the phylogenetic trees were further edited and visualized using iTOL [38]. Moreover, principal coordinate analysis (PCoA) and analysis of molecular variance (AMOVA) were performed on genotype data from SSR markers using GenAlEx ver. 6.501 [39]. AMOVA was used to determine the variance within and among species using 9,999 random repetitions. On the basis of the polymorphism amplified by SSR primers in each sample, the presence and absence of bands were converted into a 0/1 matrix. The data of each sample across core primers were combined to generate DNA fingerprints and a numeric molecular ID of digits for each sample. Characterization diagrams were generated using TBtools and a QR code generator (https://www.hlcode.cn/).

Results

Identification of SSR loci at the whole-genome level

Genome size indirectly determines the application strategy of SSRs in molecular breeding by affecting the abundance, types, and evolutionary rates of SSRs. Genome analysis of 20 moso bamboo forms revealed that the number of assembled sequences of HGZ (Ph. edulis ‘Mira’) was the greatest, and that of LC (Ph. edulis f. bicolor) was the lowest. The average sequence length of LC was the longest, and that of HC (Ph. edulis f. luteosulcata) was the shortest. Among the 20 moso bamboo forms, the maximum sequence length occurred in HGZ, whereas the minimum sequence length occurred in HC. In simple genome assembly, a minimum sequence-length threshold (500 bp) was set to filter out invalid fragments. These results indicate that there are certain differences in the number, density, and types of SSR loci associated with different moso bamboo forms (Table S5). The genome size of HGZ was the greatest, followed by that of HPH(Ph. edulis f. huamozhu), and that of HC was the smallest (Table S6).

A total of 3,835,632 SSR loci containing mono- to decanucleotide repeats were identified from the 20 moso bamboo forms. The number of SSR loci in HGZ was the greatest, followed by that in Ph. edulis f. ventricosa(FD), while that in HC was the lowest. The SSR density of HP (Ph. edulis f. holochrysa) was the maximum, and that of HGZ was the minimum (Fig. S1 and Table S6). In the genomes of moso bamboo and its 20 forms, SSRs with mono- to hexanucleotide repeat motifs accounted for more than 95% of the total. Among them, the number of SSRs with dinucleotide repeat motifs was the greatest, followed by those with mononucleotide and trinucleotide repeat motifs (Fig. 1a). Among the 21 bamboo samples, PE had the highest percentage of SSRs with dinucleotide repeat motifs (5.2%), whereas HGZ had the lowest percentage (4.3%) (Fig. 1b).

Fig. 1.

Statistical analysis of SSR types across moso bamboo and its 20 forms. a Proportion of ten SSR types (mono- to decanucleotide repeat motifs). Each colour represents one type of repeat motif. X-axis denotes the name abbreviations of moso bamboo and its 20 forms, Y-axis indicates the proportion. b The proportion of dinucleotide SSR across the 21 samples. AJJ: Ph. edulis f. anjiensis; AJZ: Ph. edulis f. purpureoculmis; BMZ: Ph. edulis f. porphyrosticta; DMZ: Ph. edulis f. abbreviata; FD: Ph. edulis f. ventricosa; FMZ: Ph. edulis f. obtusangula; GJ: Ph. edulis ‘Kikko-chiku’; HB: Ph. edulis f. pachyloen; HC: Ph. edulis f. luteosulcata; HGJS: Ph. edulis f. venusta; HGZ: Ph. edulis ‘Mira’; HP: Ph. edulis f. holochrysa; HPH: Ph. edulis f. huamozhu; JS: Ph. edulis f. gracilis; LC: Ph. edulis f. bicolor; LPH: Ph. edulis f. nabeshimana; MYZ: Ph. edulis f. exaurita; PE: Ph. edulis; QZ: Ph. edulis f. obliquinoda; SY: Ph. edulis f. tubaeformis; YMZ: Ph. edulis f. epruinosa

Characterization of SSRs in 20 moso bamboo forms

Using PE SSR loci as the reference, a comparative analysis of SSR loci revealed that FD (Ph. edulis f. ventricose) had the greatest number of polymorphic SSR loci (4,777), whereas the number of polymorphic SSR loci was the lowest in HC (2,556) (Fig. 2a). On the basis of the genome annotation data, polymorphic SSR loci were categorized into two primary regions: intergenic regions and genic regions. SSR loci located in intergenic regions were the most abundant, accounting for 97.3%, whereas only 2.7% of SSR loci were in genic regions. A total of 760 polymorphic SSR loci were distributed in the exons, introns, 5’UTRs, and 3’UTRs of the genetic regions of 738 genes (Table S7), among which those in introns and exons each accounted for only 1.4% and 17.4%, respectively, while those in the UTRs accounted for the majority (81.2%) (Fig. 2b). The genes containing SSR loci were predominantly distributed across scaffold 1 to scaffold 24 (Fig. 2c). Statistical classification of SSR loci in genetic regions revealed mononucleotide, dinucleotide and trinucleotide repeats as the predominant motifs. Moreover, dinucleotide repeats were the most abundant (328), whereas heptanucleotide repeats were rarely detected in these SSR loci (only 2) (Fig. 2d). Octanucleotide, nanonucleotide and decanucleotide repeat motifs were not detected within these genic regions.

Fig. 2.

Comparison of SSR characteristics among 20 moso bamboo forms. a The SSR polymorphism of each form. X-axis denotes the name abbreviations of 20 moso bamboo forms, Y-axis indicates the number of polymorphic SSR loci. AJJ: Ph. edulis f. anjiensis; AJZ: Ph. edulis f. purpureoculmis; BMZ: Ph. edulis f. porphyrosticta; DMZ: Ph. edulis f. abbreviata; FD: Ph. edulis f. ventricosa; FMZ: Ph. edulis f. obtusangula; GJ: Ph. edulis ‘Kikko-chiku’; HB: Ph. edulis f. pachyloen; HC: Ph. edulis f. luteosulcata; HGJS: Ph. edulis f. venusta; HGZ: Ph. edulis ‘Mira’; HP: Ph. edulis f. holochrysa; HPH: Ph. edulis f. huamozhu; JS: Ph. edulis f. gracilis; LC: Ph. edulis f. bicolor; LPH: Ph. edulis f. nabeshimana; MYZ: Ph. edulis f. exaurita; QZ: Ph. edulis f. obliquinoda; SY: Ph. edulis f. tubaeformis; YMZ: Ph. edulis f. epruinosa. b The distribution of SSR loci in different regions of moso bamboo genome. c The distribution of 760 SSR loci on the scaffolds of moso bamboo. The blue blocks represent scaffold length; The purple blocks represent the distribution of genes containing SSR loci; The green blocks represent gene density on each scaffold. d Statistics of different SSR types in genic regions. X-axis denotes SSR numbers, Y-axis indicates repeat motif types. Mono: Mononucleotide; Di: Dinucleotide; Tri: Trinucleotide; Tetra: Tetranucleotide; Pena: Pentanucleotide; Hexa: Hexanucleotide; Hepta: Heptanucleotide

Functional enrichment analysis of genes containing SSR loci

To explore the pathways associated with genes containing SSR loci, we conducted KEGG enrichment analysis of 738 SSR-related genes. A total of 119 pathways were enriched in these processes (Table S8). Furthermore, we selected the top 15 enriched KEGG terms for graphical presentation, among which the most significantly enriched pathway was plant hormone signal transduction, followed by the MAPK signalling pathway and lysine biosynthesis. In addition, we also observed that plant‒pathogen interactions and spliceosome pathways were enriched (Fig. 3a; Table S8). To further investigate the function of genes containing SSR loci, we conducted functional annotation on 53 genes involved in plant hormone signal transduction using the Pfam database. The most significantly enriched term was protein kinase domain, followed by basic helix-loop-helix (bHLH) DNA-binding domain and protein tyrosine and serine/threonine kinase (Fig. 3b and Table S9). The protein kinase domain is the core functional region of protein kinases, which are crucial components in cellular signalling pathways and are responsible for protein phosphorylation; this domain is termed the “power switch” of cellular activity [40]. Notably, among the most significantly enriched functional terms, PeBRI1 (PH02Gene12910) and PeBSK3 (PH02Gene32060) are homologous to rice OsBRI1 and OsBSK3, which are related to brassinosteroid (BR) signal transduction and have been confirmed to play important roles in the formation of plant shape in rice [41, 42]. In addition, PePIL13 (PH02Gene13739) is homologous to rice OsPIL1/13 and is a key regulator of internode elongation (Table S9) [43]. We speculate that these genes may play important roles in the morphogenesis of moso bamboo forms.

Fig. 3.

Functional enrichment analysis of genes containing SSR loci. a The top 15 enriched terms of 738 genes. The size of dots indicates the number of enriched genes in the related pathway, and the dot colour corresponds to different p-value ranges. b Pfam annotation terms of 53 genes in plant hormone signal transduction pathways. c Venn diagram showing overlap between DEGs and 738 genes containing SSR loci. d The top 20 enriched terms of DEGs containing SSR loci (The DEGs are in the dashed-line regions of Fig. 3c)

To further investigate whether genes containing SSR loci exhibited differential expression between normal and variant internodes, the transcriptome data of two representative forms of SY and GJ were used. A total of 23,669 genes were identified as differentially expressed genes (DEGs), which were significantly enriched in processes such as the biosynthesis of secondary metabolites (1665), plant hormone signal transduction (725), and ribosomes (429) (Fig. S2b). Specifically, there were 18,257 DEGs and 16,972 DEGs in the comparisons of SY vs. PE and GJ vs. PE (Fig. S2a; Fig. 3c), respectively, in which 375 and 341 of these DEGs containing SSR loci were identified, respectively (Fig. 3c). A total of 469 SSR-related DEGs were identified in the transcriptome data of SY and GJ. Moreover, KEGG enrichment analysis revealed that these SSR-related DEGs were significantly enriched in plant hormone signal transduction pathways (Fig. 4d).

Fig. 4.

Expression analysis of DEGs containing SSR loci in SY. a The expression patterns of DEGs related to BR, GA, auxin, CK, ABA and JA based on RNA-Seq data. b Expression levels of five DEGs in SY and PE shoot samples. Columns and lines represent RT-qPCR and RNA-Seq data, respectively

Differential expression analysis of genes containing SSR loci

To explore the DEGs with SSR loci between variant internodes and nonvariant internodes, the DEGs screened from these two transcriptome datasets were intersected with 53 SSR-containing genes in the plant hormone signal transduction pathways, yielding 23 and 17 genes from each intersection, respectively. Among these genes, the greatest number were enriched in the BR and GA signal transduction pathways, followed by those in the auxin and CK signal transduction pathways. Interestingly, in these two transcriptome datasets, a relatively large number of genes were identified, including brassinosteroid receptor protein genes (BRI1 and BAK1) in the BR signal transduction pathway and bHLH transcription factor genes in the GA signal transduction pathway (Fig. 4a; Fig. S3a). It is speculated that the expression of these DEGs may be associated with internode formation, with such an association potentially related to differences in SSR loci within these genes. Therefore, we performed PCR amplification and comparative product analysis of these DEGs with SSR loci in the variant internodes and nonvariant internodes. There were 17 SSR loci that all exhibited polymorphisms in the comparisons of SY vs. PE and GJ vs. PE, among which 12 SSR loci had repeat sequences of two or more bases (Table S10).

To verify the expression profiles of the DEGs, five DEGs related to plant hormone signal transduction were randomly selected from each group for RT‒qPCR validation using different bamboo internode samples. The RT‒qPCR results revealed that the expression patterns of the DEGs in most samples were consistent with those obtained from high-throughput sequencing (Fig. 4b and Fig. S3b). These findings indicate that the transcriptome data are reliable, indicating a certain correlation between SSR loci polymorphisms and gene expression.

Correlation analysis between hormone contents and hormone signalling-related gene expression levels

Kendall correlation analysis and Mantel tests were employed to assess the correlations between different hormone contents and the expression levels of DEGs of plant hormone signal transduction pathways in SY and PE. The results indicated that the contents of auxin, BR, and GA were primarily positively correlated with the expression levels of the DEGs. In particular, the auxin content was strongly positively correlated with the DEGs related to the BR, GA, CK, and auxin signalling pathways (Fig. 5), among which the correlation coefficients of three genes, BAK1 (PH02Gene13505), B_ARR (PH02Gene39580), and AUX/IAA (PH02Gene33862), were relatively high (r ≥ 0.9, p < 0.01). In addition to a strong positive correlation with the brassinosteroid receptor gene BRI1 (PH02Gene12910) (0.8 ≤ r < 0.9, p < 0.01), there was a significant positive correlation between the BR content and AUX/IAA (0.8 ≤ r < 0.9, p < 0.01). In addition, a positive correlation was observed between the GA content and DELLA (PH02Gene03190) and bHLH (PH02Gene31426) expression (Fig. 5). Similarly, correlation analysis in GJ revealed a strong correlation between the expression of BAK1 (PH02Gene00446) and the auxin content and between the expression of GID1 (PH02Gene33346) and the GA content (Fig. S4). It is speculated that the expression of hormone signalling-related DEGs leads to differences in hormone contents, ultimately resulting in phenotypic differences among internodes.

Fig. 5.

Correlation analysis between hormone contents and expression levels of genes related to BR, GA, auxin, and CK in SY. The circle colour from bule to red represents the correlation of each pair genes from negative to positive. The line width is proportional to the correlation coefficient. The orange line and red line represent the positive correlation at levels of p < 0.01 and 0.01 < p < 0.05. The blue line and purple line represent the negative correlation at levels of p < 0.01 and 0.01 < p < 0.05 respectively, whereas p ≥ 0.05 is represented by gray lines

Notably, correlation analysis of SY revealed that the expression of genes related to BR signalling (BAK1, BRI1, BSK, and BIN2) was significantly correlated with that of genes related to GA (|r| > 0.7, p < 0.01) (Fig. 5), and this correlation was also observed in GJ. In the two sets of correlation analyses, DEGs containing SSR loci exhibited differences in amplification between variant and nonvariant internodes (Table S10). On the basis of a series of correlation results, we speculate that differences in SSR loci may affect gene expression levels and ultimately influence the phenotypic traits of bamboo in terms of hormone contents. We subsequently utilized these SSR loci for genetic diversity analysis and construction of DNA fingerprinting profiles.

Polymorphism analysis of SSR markers in 22 bamboo samples

A total of 26,169 primers were successfully designed after 27,801 polymorphic SSR loci sequences were predicted (Fig. S5a). SSR loci located in genic regions and having highly polymorphic repeat sequences of two or more bases were selected for primer synthesis. To validate the polymorphism of SSR markers in 22 bamboo samples (moso bamboo and its 20 forms) and an outgroup species, Ph. bambusoides), and 332 pairs of SSR primers were selected and synthesized (Table S3). Following PCR amplification of the bamboo samples and agarose gel electrophoresis of the amplicons, 247 pairs of SSR markers were selected according to their clear band patterns and amplification fragment lengths (Fig. S5b–c), which were subsequently used to perform PCR amplification and capillary electrophoresis on 22 bamboo samples. The results revealed that 222 pairs of SSR primers displayed polymorphisms with clear peaks. To ensure the accuracy and reliability of the genetic diversity analysis, null allele frequencies were estimated using FreeNA, excluding those loci with a null allele frequency > 0.2. A total of 206 SSR markers were selected and used for the subsequent genetic diversity analyses (Table S11). For example, SSR251 exhibited the greatest number of genotypes, including seven types (Fig. S6), followed by SSR82, SSR201, SSR239 and SSR271, each with six genotypes (Fig. S7–Fig. S10). The result of HWE tests revealed that most SSR loci deviated significant or extremely significant from HWE. The inbreeding coefficient F_IS was predominantly negative, indicating an overabundance of heterozygotes with the population (Table S12).

The average numbers of different alleles (Na) and effective alleles (Ne) of the 206 pairs of SSR primers were 4.005 and 2.033, respectively. The Shannon’s information index (I) values ranged from 0.109 to 1.356, with an average of 0.810. The average Ho and He values detected for each marker were 0.634 and 0.469, respectively. There were 145 pairs of SSR primers with He < Ho, accounting for more than half of the total primer pairs, indicating an excess of heterozygotes in the tested bamboo samples. The polymorphic information content (PIC) value ranged from 0.043 to 0.621, with an average of 0.392, indicating moderate polymorphism (0.25 < PIC < 0.5) (Table S11). A total of 51 primer pairs had high PIC values (PIC > 0.5) (Table 1), and 79.1% of the primers had a PIC ≥ 0.25 (Table S11). These results indicate that these SSR primers are suitable for genetic diversity studies.

Table 1.

Genetic diversity index of 51 SSR markers in 22 bamboo samples (PIC > 0.5)

Primer name	Na	Ne	I	Ho	He	PIC	Genotype number
SSR10	7	2.916	1.313	0.909	0.672	0.595	5
SSR12	6	2.89	1.262	0.864	0.669	0.59	4
SSR46	5	2.856	1.205	0.864	0.665	0.583	4
SSR47	6	2.521	1.12	1	0.617	0.523	4
SSR48	6	2.89	1.262	0.955	0.669	0.59	4
SSR70	8	2.933	1.356	1	0.674	0.599	4
SSR77	4	2.501	1.045	0.909	0.614	0.518	4
SSR78	4	2.464	1.03	0.895	0.61	0.51	3
SSR79	4	2.602	1.083	0.818	0.63	0.539	4
SSR84	6	2.638	1.177	0.955	0.635	0.548	4
SSR85	6	2.508	1.117	0.909	0.615	0.522	5
SSR87	5	2.514	1.088	0.909	0.616	0.522	4
SSR93	4	2.488	1.042	0.818	0.612	0.517	5
SSR98	4	2.602	1.083	0.818	0.63	0.539	4
SSR100	6	2.766	1.224	0.909	0.653	0.571	5
SSR110	6	2.508	1.117	0.909	0.615	0.522	5
SSR113	7	2.652	1.21	0.955	0.637	0.55	5
SSR114	6	2.89	1.262	0.864	0.669	0.59	4
SSR115	4	2.602	1.083	0.818	0.63	0.539	4
SSR116	8	2.933	1.356	1	0.674	0.599	5
SSR119	5	2.756	1.161	0.905	0.653	0.566	4
SSR121	5	2.727	1.17	0.818	0.648	0.564	5
SSR123	4	2.774	1.11	0.864	0.654	0.566	3
SSR124	6	2.864	1.236	0.909	0.666	0.585	4
SSR127	6	2.774	1.236	0.909	0.654	0.572	5
SSR131	5	2.758	1.192	0.864	0.652	0.569	5
SSR133	5	2.881	1.23	0.818	0.668	0.589	4
SSR134	4	2.488	1.025	0.864	0.612	0.514	4
SSR138	7	2.782	1.267	0.955	0.655	0.574	5
SSR139	4	2.488	1.042	0.955	0.612	0.517	4
SSR147	6	2.911	1.282	0.905	0.673	0.597	4
SSR148	5	2.63	1.146	0.909	0.634	0.546	4
SSR149	8	2.933	1.356	1	0.674	0.599	4
SSR159	5	2.719	1.147	0.864	0.647	0.56	5
SSR161	6	2.638	1.177	1	0.635	0.548	5
SSR166	4	2.488	1.025	0.864	0.612	0.514	4
SSR171	6	2.907	1.281	0.864	0.671	0.594	5
SSR175	5	2.742	1.173	0.864	0.65	0.565	4
SSR186	3	2.719	1.041	0.818	0.647	0.554	2
SSR205	5	2.673	1.161	0.952	0.641	0.553	3
SSR211	4	2.602	1.083	0.818	0.63	0.539	4
SSR218	6	3.113	1.318	0.864	0.695	0.621	4
SSR228	5	2.758	1.192	0.864	0.652	0.569	5
SSR230	5	2.495	1.074	0.909	0.613	0.518	5
SSR239	7	2.89	1.32	0.864	0.669	0.594	6
SSR241	5	2.856	1.205	0.864	0.665	0.583	4
SSR246	5	2.881	1.23	0.909	0.668	0.589	3
SSR255	5	2.495	1.074	1	0.613	0.518	3
SSR260	5	2.881	1.23	0.909	0.668	0.589	3
SSR271	7	2.89	1.32	0.955	0.669	0.594	6
SSR281	5	2.609	1.114	0.864	0.631	0.541	4

Genetic distance-based clustering and PCoA

To verify the polymorphism of these primers, genotyping was performed on moso bamboo and its forms, with Ph. bambusoides (GZ) used as the outgroup. Genetic distances among 22 bamboo samples were calculated using PowerMarker V3.25 to construct a genetic distance matrix, and a heatmap of genetic distances was generated using R software (Fig. 6a). The genetic distances ranged from 0 to 0.85 (on a 0–1 scale). Compared with the other 21 bamboo samples, GZ had the greatest genetic distance, indicating its distant genetic relationship. PE and AJZ (Ph. edulis f. purpureoculmis) showed relatively smaller genetic distances than the other forms of moso bamboo did, suggesting a closer genetic relationship. The genetic distances of JS (Ph. edulis f. gracilis) and GZ was greater than that of JS and the other 20 samples. Among these, the genetic distances between JS, PE, and AJZ were smaller. The other 18 forms of moso bamboo displayed extremely low genetic distances but remained distinguishable from one another. Cluster analysis was performed to further assess the genetic relationships among the 22 bamboo accessions. A phylogenetic tree was constructed on the basis of similarity coefficients, in which the 22 samples were classified into two major clusters. GZ was separately classified into Group I, while the remaining samples were assigned to Group II. Additionally, Group II was further subdivided into three subgroups: JS was grouped into Subgroup II-1; PE and AJZ were in Subgroup II-2; and the other forms were in Subgroup II-3 (Fig. 6b).

Fig. 6.

Cluster analysis and PCoA based on genetic distances for 22 bamboo samples. a Heat map of genetic distances. The colour from pink to purple represents the genetic relationship from close to far. b Cluster analysis. c Principal coordinate analysis. Axis 1 accounts for 60.21% of the variation, while axis 2 explains 24.89% of the variation. The rhombuses represent 22 bamboo samples. AJJ: Ph. edulis f. anjiensis; AJZ: Ph. edulis f. purpureoculmis; BMZ: Ph. edulis f. porphyrosticta; DMZ: Ph. edulis f. abbreviata; FMZ: Ph. edulis f. obtusangula; GJ: Ph. edulis ‘Kikko-chiku’; GZ: Ph. bambusoides; HB: Ph. edulis f. pachyloen; HC: Ph. edulis f. luteosulcata; HGJS: Ph. edulis f. venusta; HGZ: Ph. edulis ‘Mira’; HP: Ph. edulis f. holochrysa; HPH: Ph. edulis f. huamozhu; JS: Ph. edulis f. gracilis; LC: Ph. edulis f. bicolor; LPH: Ph. edulis f. nabeshimana; MYZ: Ph. edulis f. exaurita; PE: Ph. edulis; QLZ: Ph. edulis ‘Curviculmis’ QZ: Ph. edulis f. obliquinoda; SY: Ph. edulis f. tubaeformis; YMZ: Ph. edulis f. epruinosa

To validate the above cluster analysis results, PCoA was performed on those 22 bamboo samples. The results revealed that the samples were also distinctly clustered into two groups (Fig. 6c). GZ comprised Group I, away from the others, and moso bamboo and its forms were congregated into Group II, which reflected the degree of difference in their genetic backgrounds. The samples in Group II were further divided into three subgroups. Similarly, JS was alone in Subgroup 1, away from other forms. PE and AJZ were comprised Subgroup 2, and the other 18 moso bamboo forms were included in Subgroup 3. The results of the PCoA and cluster analysis were consistent with those of traditional taxonomy [5].

Construction of DNA fingerprints and molecular IDs for 20 moso bamboo forms

A total of 24 core SSR primers were screened from 206 primer pairs to construct fingerprinting maps (Fig. 7a), enabling the complete discrimination of 20 bamboo forms. Capillary electrophoresis was performed individually for each of the 206 SSR markers in 20 moso bamboo forms to screen for SSR markers that exhibited unique peak values (corresponding to the size of the DNA products) for a specific moso bamboo form. For example, the peak value of AJZ in SSR251 was distinct, indicating that SSR251 could be used to distinguish AJZ from other moso bamboo forms. However, discriminating moso bamboo forms with closely related genetic backgrounds requires the combination of multiple primer pairs. For instance, the peak values of HPH (Ph. edulis f. huamozhu) and LPH (Ph. edulis f. nabeshimana) in SSR147 were identical, although both differed from the peak values of the other forms. Therefore, the SSR2 marker was employed to discriminate HPH from LPH (Fig. 7a). Through comparative analysis, we ultimately revealed that only 24 SSR primer pairs were needed to clearly discriminate the 20 moso bamboo forms.

Fig. 7.

DNA fingerprints and molecular IDs of moso bamboo forms. a DNA fingerprints of 20 moso bamboo forms. The abbreviations on the left of the figure denote the name of 20 moso bamboo forms. The symbols at the top of the figure represent different primer names of SSR markers, and the numbers at the bottom of the figure represent the size of DNA products. The orange and gray circle dots indicate the presence and absence of specific bands in PCR products of each primer pair. b Morphological characteristics and the molecular identity cards of five representative moso bamboo forms. AJJ: Ph. edulis f. anjiensis; AJZ: Ph. edulis f. purpureoculmis; BMZ: Ph. edulis f. porphyrosticta; DMZ: Ph. edulis f. abbreviata; FMZ: Ph. edulis f. obtusangula; GJ: Ph. edulis ‘Kikko-chiku’; HB: Ph. edulis f. pachyloen; HC: Ph. edulis f. luteosulcata; HGJS: Ph. edulis f. venusta; HGZ: Ph. edulis ‘Mira’; HP: Ph. edulis f. holochrysa; HPH: Ph. edulis f. huamozhu; JS: Ph. edulis f. gracilis; LC: Ph. edulis f. bicolor; LPH: Ph. edulis f. nabeshimana; MYZ: Ph. edulis f. exaurita; QLZ: Ph. edulis ‘Curviculmis’ QZ: Ph. edulis f. obliquinoda; SY: Ph. edulis f. tubaeformis; YMZ: Ph. edulis f. epruinosa

Furthermore, on the basis of the higher PIC values of SSR primers and specific primers for distinguishing one moso bamboo form, we selected SSR primer combinations to construct DNA fingerprints and molecular IDs. Compared with the other moso bamboo forms, JS and AJZ were more easily distinguishable, as they could be distinguished by more than half of the 24 core SSR primer pairs. However, HPH, DMZ (Ph. edulis f. abbreviate), LPH, LC (Ph. edulis f. bicolor), HB (Ph. edulis f. pachyloen), and MYZ (Ph. edulis f. exaurita) required more than two primer pairs in combination for differentiation, indicating that their genetic similarity was significantly greater than that of JS and AJZ. Finally, we constructed molecular IDs of 20 moso bamboo forms on the basis of their genotype data from 24 SSR primer pairs. The morphological characteristics and molecular IDs of five representative moso bamboo forms, namely, GJ, HC, HGZ, HPH, and SY, are shown in Fig. 7b. Those for the other 15 moso bamboo forms are shown in Fig. S11.

Molecular ID construction for the species of the genus Phyllostachys

On the basis of 30 randomly screened SSR primer pairs, capillary electrophoresis analysis was performed on 20 species of the genus Phyllostachys (Table S2). Ultimately, 22 out of the 30 SSR primer pairs with good peak patterns were selected for genetic analysis. Genetic diversity was evaluated using PopGene32 software (Table S13), followed by the construction of an UPGMA dendrogram based on Nei's genetic distances. The results revealed that 20 bamboo species of the genus Phyllostachys were successfully distinguished. Among them, PE, ZZ (Ph. nigra), and JMZ (Ph. kwangsiensis) were clustered into Group I. GZ and its forms, including BZ (Ph. bambusoides f. lacrima-deae), HCB (Ph. bambusoides f. mixta), DHZ (Ph. bambusoides ‘Duihuazhu’), and RMZ (Ph. aurea) were clustered together in Group II. The other bamboo species of the genus Phyllostachys were clustered into Group III and were further classified into different subgroups. For instance, JXYZ (Ph. aureosulcata ‘Spectabilis’), HGJZ (Ph. aureosulcata ‘Aureocarlis’), and HCZ (Ph. aureosulcata) were clustered together and belonged to Ph. aureosulcata. Furthermore, BBJZ (Ph. dulcis), WBJZ (Ph. vivax), and HGWBJ (Ph. vivax f. aureocaulis) were clustered together (Fig. S12a). A subset of core SSR primer pairs (SSR2, SSR8, SSR22, SSR49, SSR116, and SSR170) was selected to construct fingerprinting maps (Fig. S12b) and molecular IDs for 20 species of the genus Phyllostachys (Fig. S13). The results of these core SSR primer pairs exhibited high resolution, enabling effective differentiation of not only intraspecific varieties and forms but also species within the genus Phyllostachys. Thus, the products of these SSR primer pairs can serve as molecular markers for genotyping bamboo samples.

AMOVA of the bamboo samples

AMOVA is employed to assess variance components at different levels, such as individuals, species, and populations. On the basis of three groups of dendrograms for 20 species of the genus Phyllostachys, PE, GZ, HCZ, and their forms were used for AMOVA. The genetic variation rate among species in the genus Phyllostachys was 58%, whereas that within species (intraspecific forms) of moso bamboo was 42% (Fig. S14 and Table S14). These results revealed significant differences within and among populations (p = 0.001). The genetic variation among species was greater than that within species, which was consistent with previous research results [44]. These results indicate that the genetic variation of bamboo samples within the genus Phyllostachys is primarily derived from interspecific differences.

Discussion

SSR molecular markers serve as promising candidate markers for germplasm identification and genetic diversity in plants because of their widespread genomic distribution, high degree of polymorphism, and robust stability. In recent years, advances in whole-genome sequencing technology have enabled increasing studies on the development of genome-wide SSR molecular markers with cost-effective approaches [45–47].

Comparison of SSR distribution in bamboo with that in other grasses

In this study, SSR identification and comparative analysis were performed on moso bamboo and its 20 forms. We found that the largest genome, HGZ (2,035.558 Mb), had the lowest SSR density (108.496 SSRs/Mb), while the smallest genome, HC, had a higher SSR density (122.331 SSRs/Mb) but a lower SSR density than those of the other moso bamboo forms did. The SSR density of the 20 moso bamboo forms was much greater than that of D. brandisii (82.4 SSRs/Mb) and Zea mays (52.1 SSRs/Mb) but much lower than that of O. sativa (165.5 SSRs/Mb) [24, 48]. These findings indicate that there was a large difference in SSR distribution frequency between different species, which may be related to external factors such as different conditions for SSR search and genome assembly quality or internal factors, including species and varieties. In addition, the distribution of SSR motifs was relatively similar in moso bamboo and its 20 forms, with dinucleotide repeats being the most abundant. This finding is consistent with those of previous studies of five different bamboo species [49]. However, we only preliminarily assessed the differences in the length, number, and characteristics of SSR loci among the 20 moso bamboo forms in this study, and comparisons with other species may be biased. In future work, we will assemble high-quality genomes to address these limitations.

Potential effects of differences in SSR loci on gene expression

Variations in and physical locations of SSR sequences can affect gene expression and protein function. In particular, SSRs located in genic regions can not only affect gene transcription and posttranscriptional regulation but also lead to the gain or loss of gene function [50]. Here, we focused on analysis of SSR loci located in the genic regions and further performed functional annotation of SSR-related genes. The results revealed that the greatest number of genes were enriched in terms of protein kinases and bHLH DNA-binding domains. Notably, the bHLH family not only contributes to the regulation of plant morphology but also regulates anthocyanin biosynthesis [51]. Anthocyanins are key pigments responsible for plant colouration. For instance, the variation in the sheath colour of Chimonobambusa utilis has been confirmed to be associated with CubHLH17 [52]. Therefore, protein kinase genes and bHLH and MYB TFs enriched in the plant signal transduction pathway may play key roles in the variations in culm type and culm colour among moso bamboo types.

In the regulation of bamboo growth and development, plant hormones play particularly critical roles [53]. It has been reported that exogenous application of GA and BR significantly increased the internode length of moso bamboo, but neither could rescue the shortened internode trait of Ph. edulis f. tubaeformis [54]. We speculated that the shortened internodes of Ph. edulis f. tubaeformis may not be caused by insufficient hormone synthesis but by the impairment of plant hormone signal transduction pathways, which in turn hinder the elongation of internode cells. The DEGs in the internodes of the two forms compared with those in the PE group were significantly enriched in plant hormone signal transduction pathways. Among these DEGs, those harbouring polymorphic SSR loci were enriched mainly in the BR and GA signal transduction pathways, followed by the auxin signal transduction pathway. These findings indicate that the shortened internode trait of SY and GJ may be regulated by the synergistic involvement of multiple hormone signals. It is hypothesized that variations in the SSR loci of these genes may affect their expression levels. In this study, the hormone correlation analyses have notable statistical limitations, such as the small sample size, which reduce the statistical power of correlation tests. To overcome these limitations, we used Kendall’s rank correlation, which is suitable for small datasets, and applied FDR correction for multiple testing. However, these measures cannot fully eliminate the biases caused by the limited sample size. Therefore, the correlation results should be regarded as exploratory references only.

SSR polymorphisms in genic regions contribute to bamboo genetic diversity and germplasm identification

Genetic diversity describes the amount of genetic variation that exists within a biological population, which is the basis for biological evolution and adaptation to environmental change. The development of SSR markers derived from functional genes will likely offer an effective tool for genetic diversity analysis. In the present study, a relatively high level of genetic diversity (I = 0.81) was detected among 22 bamboo samples when 206 newly developed SSR primer pairs were used. This genetic diversity is more pronounced than that detected by other molecular markers previously reported [55–57]. The main reason for this difference may be that compared with the AFLP, RAPD, and EST-SSR markers, the selected SSR markers are more polymorphic. Moreover, the high resolution of capillary electrophoresis technology greatly enhances the detection of allelic polymorphisms. Additionally, the inclusion of GZ further increased the number of SSR primer polymorphisms. The values of Ho and He also reveal that there is a detectable basis of genetic variation within moso bamboo and its forms, indicating that they are a population without a highly uniform genetic background. These findings further confirm the existence of genetic differentiation among moso bamboo forms at the genomic level. In this study, we observed that the majority of loci deviated from HWE, which is associated with the biological characteristics of long-term asexual reproduction in bamboo, which fundamentally breaks the random mating prerequisite for HWE. In addition, the small size of samples and selective pressure may jointly contribute to the significant deviation from HWE.

In addition to being used for genetic diversity analysis, SSR markers play a crucial role in genetic relationship analysis and germplasm identification. In this study, 206 newly developed SSR markers were used to further confirm that there is a close genetic relationship between moso bamboo and its 20 forms. Despite their high similarity in genetic background, the selected primers can distinguish each form from another.

Practical implications of SSR markers for DUS testing and bamboo breeding

Currently, molecular markers are considered cost-effective tools for DUS testing for plant variety protection by the International Union for the Protection of New Varieties of Plants (UPOV), which is explained in the document TGP/15 “Guidance on the Use of Biochemical and Molecular Markers in the Examination of Distinctness, Uniformity and Stability (DUS)” [58]. The morphological traits used in DUS testing are vulnerable to environmental and seasonal influences, but SSR markers can overcome these limitations, optimize the DUS evaluation system for new plant varieties, and improve identification accuracy [59]. For example, species-specific SSR primers developed for the genus Pleioblastus, combined with DUS guidelines, enable accurate germplasm identification in this genus [60]. Moreover, the 19 selected SSR markers, in conjunction with DUS traits, can effectively discriminate the 113 Dioscorea polystachya varieties [61]. Therefore, SSR markers are conducive to promoting the standardization and precision of DUS testing. Although the number of bamboo cultivars and new varieties has increased annually, formal DUS testing for new bamboo varieties has not been fully established, resulting in unclear species boundaries and ambiguous relationships among closely related genera and species [62].

Even though hybrid bamboo breeding is extremely difficult, breeders still obtain some interspecific hybrids. Furthermore, the developed SSR markers are suitable for the hybrid identification of bamboo. Thirty SSR primer pairs were randomly selected from the published literature and used to identify the authenticity of hybrids generated from Bambusa multiplex × Dendrocalamus latiflorus and B. multiplex × B. chungii. The results indicated that the 34 hybrid progenies derived from B. multiplex × D. latiflorus were true hybrids, whereas there were five false hybrids among the 59 hybrid progenies from B. multiplex × B. chungii [63]. These results provide a valuable reference for parental selection and hybrid combination design in bamboo breeding and will facilitate the breeding of new varieties with elite and distinctive traits.

Limitations of SSR vs. SNP markers

Although SSR markers have many advantages and still have significant value in genetic breeding studies of small populations, they also have limitations, such as complex procedures, time-consuming data scoring, and population range [64, 65]. To ensure the quality of SSR markers, we implemented multiple targeted measures, including optimized parameters for SSR identification, stringent primer screening, improved amplification protocols, and defined genotyping thresholds, etc. These measures effectively reduced the false positive rate and the genotyping error rate. Nevertheless, strand slippage may still occur during PCR amplification, which is also an inherent limitation of all conventional SSR genotyping techniques based on capillary electrophoresis. In addition, high GC content, base mismatches, and insertion–deletion errors during sequencing may affect the accuracy of the screening of polymorphic SSR loci. A limitation of this study is that we selected only GZ, HCZ, PE, and some of their forms to conduct AMOVA among species and within species, which resulted in a slightly lower genetic variation rate among species of the genus Phyllostachys and a slightly higher one in intraspecific ranks. As the number of samples is increased, the results of AMOVA will become more accurate. Moreover, the clustering of some bamboo species in the genus Phyllostachys is not highly consistent with traditional taxonomy [5]. For example, JMZ and ZZ in Group I belong to Sect. Phyllostachys and Sect. Heterocladae, respectively, which may be due to an insufficient number of SSR markers used in the clustering analysis. We will conduct additional research to solve this problem in the future.

In contrast to single-nucleotide polymorphism (SNP) markers, SSR markers cannot enable automated, rapid and high-throughput genotyping. Owing to their ultrahigh abundance in the genome, high throughput, and ease of data analysis, SNPs have become the mainstream choice for large-scale, high-precision genomic research. With the advancement of sequencing technology, SNPs have been widely used in large-scale population genetic and breeding studies [65, 66]. We identified 5.45 million SNPs from the whole-genome resequencing of 427 individuals of moso bamboo covering 15 representative geographic areas, which were used for analyses of population structure, origin and evolution, and association studies of property-related traits [67]. Furthermore, we identified 3.23 million high-quality SNPs from 20 moso bamboo types and revealed that low genetic diversity and high genotype heterozygosity are associated with the formation of moso bamboo forms [6]. However, owing to the unique biological characteristics of bamboo, which remains in the vegetative growth stage for a long time, it lacks true populations. SNP markers for variety identification have not yet been developed in bamboo at this stage. We believe that SNP markers will become powerful tools for bamboo germplasm identification and DUS testing in the future.

Conclusion

In the present study, a total of 3,835,632 SSR loci were identified from the genome sequences of 20 moso bamboo forms. Among them, dinucleotide SSRs were the most predominant type, and only 2.7% of the SSR loci were located in genic regions. The plant hormone signal transduction pathway was significantly enriched when all the identified SSR-containing genes were compared with those annotated by KEGG, in which the functional terms of protein kinases and bHLH TFs were abundant. A total of 206 polymorphic SSR primer pairs were developed and used to differentiate the infraspecific ranks of moso bamboo and the species in the genus Phyllostachys, which confirmed the diversity levels and genetic relationships among these bamboo samples. Furthermore, specific molecular IDs of 20 moso bamboo forms and 20 species of the genus Phyllostachys were constructed using core SSR primers. Overall, this study provides important information for future germplasm identification, classification and molecular marker-assisted selection studies on the superior quality traits of bamboo.

Supplementary Information

12870_2026_9292_MOESM1_ESM.doc ^{(121.6MB, doc)}

Additional file 1: Fig. S1 Genome sizes, SSR densities and numbers of 20 moso bamboo forms. Fig. S2 Identification of differentially expressed genes (DEGs) and their KEGG enrichment analysis. Fig. S3 Expression analysis of DEGs containing SSR loci. Fig. S4 Correlation analysis between hormone contents and expression levels of genes related to auxin, CK, and GA. Fig. S5 The effectiveness of SSR primers based on agarose gel electrophoresis. Fig. S6 Capillary electropherograms for the amplification of 7 forms with primer SSR251. Fig. S7 Capillary electropherograms for the amplification of 6 forms with primer SSR82. Fig. S8 Capillary electropherograms for the amplification of 6 forms with primer SSR201. Fig. S9 Capillary electropherograms for the amplification of 6 forms with primer SSR239. Fig. S10 Capillary electropherograms for the amplification of 6 forms with primer SSR271. Fig. S11 Morphological characteristics and molecular identity cards of 15 moso bamboo forms. Fig. S12 The dendrogram and DNA fingerprints of bamboo species in the genus Phyllostachys. Fig. S13 Morphological characteristics and molecular identity cards of 20 species in the genus Phyllostachys. Fig. S14 Analysis of molecular variance (AMOVA) among species and within species.

12870_2026_9292_MOESM2_ESM.xls ^{(301.5KB, xls)}

Additional file 2: Table S1 The list of 20 moso bamboo forms. Table S2 The list of 20 moso bamboo forms and 20 species of the genus Phyllostachys used for genetic analysis. Table S3 332 pairs of SSR primers used in this study. Table S4 Primers used for RT-qPCR analysis. Table S5 Comparative analysis of the genome sequence lengths of 20 moso bamboo forms. Table S6 The information of SSR identification across 20 moso bamboo forms.Table S7 The detail information of 760 SSR loci in genic regions. Table S8 KEGG enrichment analysis of 738 genes. Table S9 The information of Pfam annotation terms of 53 genes. Table S10 DEGs with polymorphic SSR loci between variant internodes and non-variant internodes. Table S11 Genetic diversity index of the 206 SSR primer pairs in 22 bamboo samples. Table S12 The Hardy-Weinberg Equilibrium exact tests of 206 SSR loci Table S13 Genetic diversity index of the 22 SSR primer pairs in 20 species of the genus Phyllostachys. Table S14 Analysis of molecular variance (AMOVA) among species and within species.

Abbreviations

SSR

Simple sequence repeat

DUS

Distinctness, Uniformity, and Stability

RAPD

Random amplified polymorphic DNA

AFLP

Amplified fragment length polymorphism

ISSR

Inter-simple sequence repeat

MISA

Microsatellite identification tool

KEGG

Kyoto Encyclopedia of Genes and Genomes

CKs

Cytokinins

BRs

Brassinosteroids

GAs

Gibberellins

gDNA

Genomic DNA

ID

Identity card

UPGMA

Unweighted Pair Group Method with Arithmetic Mean

PCoA

Principal coordinate analysis

AMOVA

Analysis of molecular variance

bHLH

Basic helix-loop-helix

TFs

transcription factors

DEGs

Differentially expressed genes

Na

Number of observed alleles

Ne

Number of effective alleles

Ho

Observed heterozygosity

He

Expected heterozygosity

I

Shannon’s information index

PIC

Polymorphic information content

Authors’ contributions

**Dandan Kou: ** Writing – original draft, Data curation, Formal analysis. **Yan Liu: ** Data curation **Hui Li: ** Methodology, Software. **Xihong Wan: ** Methodology, Software. **Chenlei Zhu: ** Data curation. **Xianghua Yue** : Sample collection. **Zhimin Gao: ** Writing – review, Design, and Supervision.

Funding

This research was supported by the National Natural Science Foundation of China (Grant No. 32494791).

Data availability

The data that supports the findings of this study are available in the supplementary material of this article. Raw reads from the sequencing data are available at the NCBI Sequence Read Archive (; Project numbers: PRJNA1348092).

Declaration

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.