Identification of Informative Microsatellite Markers in the Avena Chloroplast Genome Provides New Insights into Oat Phylogeny

1 Introduction

Avena L. is a genus within the Poaceae family that comprises 26 cultivated and wild species exhibiting different ploidy levels (diploid, tetraploid, and hexaploid) [1]. Oat is an important food and forage crop known since ancient times. Recently, the crop has come under increasing scrutiny as a source of functional nutrition due to its numerous health benefits [2,3,4,5,6]. Oat’s grains have an interesting nutritional profile that includes high-quality protein, unsaturated fats, soluble fiber, polyphenolic compounds, and micronutrients [7]. Additionally, oats (Avena sativa L.) are used as therapeutic plants, particularly in dermatology [8]. Because of that, they serve as subjects of multifarious studies [9,10,11,12,13,14,15]. Significant efforts have been applied to oat breeding research [16,17,18].

Development of novel crop varieties requires accurate understanding of phylogenetic relationships as well as the introduction of wild relatives to breeding programs as donors of new traits. Therefore, the investigation of wild species and Avena phylogeny has not only theoretical but also important applied concerns. Avena genus species possess four primary genomes: A, B, C, and D. Currently, known oat species include diploids with A- and C-genomes, tetraploids with AB- and AC-genomes, and hexaploids with ACD-genome [19]. Recently, the role of genomic approaches in the investigation of Avena species has increased [20,21,22]. Nevertheless, the phylogenetic relationships among oat species remain unresolved. Therefore, the search for informative molecular markers and the analysis of variability within Avena species are still important.

Chloroplast loci are widely used in plant evolutionary studies; yet identifying informative plastid markers suitable for detecting polymorphism at low taxonomic levels can be challenging [23,24,25,26]. Recent research highlights a growing interest in the development of highly informative chloroplast microsatellite systems for various plant taxa, including woody perennials and ecologically specialized species. For example, Guo et al. (2022) developed and characterized cpSSR markers for tree peony, underscoring their utility in population-level analyses. Similarly, Hoang et al. (2025) successfully generated chloroplast microsatellite markers for Bruguiera hainesii, demonstrating the applicability of cpSSR approaches in non-model tropical plant species [27,28].

Chloroplast microsatellites, however, can effectively address this issue and often exhibit high levels of variability across different plant groups [24,29,30,31]. Moreover, chloroplast simple sequence repeat (cpSSR) markers possess several important characteristics, including haploidy, lack of recombination, and uniparental inheritance [32]. These features make cpSSRs promising markers for studying groups of closely related species, including analyses of allopolyploid origins. Their effectiveness in resolving fine-scale population structure has been demonstrated in several recent works; in particular, Xiong et al. (2022) applied chloroplast DNA sequences and cpSSR markers to elucidate phylogeographic patterns and intraspecific divergence in Elymus sibiricus, confirming the value of cpSSR markers for studies at both inter- and intrapopulation levels [33].

Due to the overall low substitution rate in the chloroplast genome, cpSSR markers can often be transferred across related species. Typically, they are transferable among species of the same genus or closely related genera [34,35,36]. However, there are examples of successful application of chloroplast microsatellite markers to more distantly related taxa. For instance, Diekmann et al. [37] developed cpSSR markers based on chloroplast genome sequences from multiple grass species (Gramineae) representing different subfamiliesand successfully applied these markers to the “core pooids” within the Pooideae subfamily. Similar cross-taxon transferability of cpSSR loci has been reported in other plant groups, further supporting the potential of plastid microsatellites for broad phylogenetic applications [27].

Therefore, in the present study, we evaluated the possibility of applying microsatellite markers previously developed for the chloroplast genome analysis in wheat [38] to investigate intraspecific variability and clarify phylogenetic relationships within the genus Avena.

2 Materials and Methods

A total of 70 samples representing 25 species of the genus Avena were selected for analysis (Table 1). Only the narrowly distributed narrow-range species A. bruhnsiana, which is closely relatedto A. ventricosa and shares the same genome type (Cv), was not included in the study. The samples were obtained from the collection of the N.I. Vavilov All-Russian Institute of Plant Genetic Resources (VIR, St. Petersburg) and the Genbank of Plant Resources, Canada (PGRC, Saskatoon). Four accessions of different Avena species representing different genome types—A. damascena (Ad), A. hirtula (As), A. sativa (ACD), and A. clauda (Cp)—were used to test the primer pairs.

Total DNA was extracted from individual plants using a modified CTAB protocol [39]. A total of 18 cpSSR markers from the study by Ishii et al. [38] were used for the analysis (Table 2). PCR analysis was performedllowing the approach of Ishii et al. [38] with minor modifications. Each PCR reaction (15 μL) contained 100 ng of template DNA, 1× reaction buffer, 1.5 mM MgCl2, 0.2 mM dNTPs, 0.2 μM of each primer, and 0.3 U of Taq polymerase (Dialat LTD). The thermal cycling conditions were as follows: initial denaturation at 94°C for 5 min; 35 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 1 min, and elongation at 72°C for 2 min; followed by a final extension at 72°C for 7 min.

Amplified products were separated on 6% denaturing polyacrylamide gels using a 38 × 50 cm Bio-Rad SequiGen GT cell according to the manufacturer’s instructions. The Invitrogen™ 10 bp DNA Ladder (Thermo Fisher Scientific Inc.) was used to estimate fragment sizes. DNA fragments were visualized by silver staining [40]. Analysis of samples representing different alleles for each marker was performed with replications to confirm allele scoring. Nei’s genetic diversity index (H) was calculated for each locus [41]. Multidimensional scaling (MDS) based on the Dice pairwise genetic similarity coefficient was performed using PAST software [42].

3 Results

3.1 Diversity of Chloroplast SSR Loci

Initially, 18 primer pairs developed by T. Ishii and colleagues [38] were tested using DNA templates from four oat species representing different genome types. PCR amplification with 12 of the 18 primer pairs revealed polymorphism among the oat samples analyzed (Table 2). These twelve primer pairs were subsequently used to assess chloroplast genome variability in 70 samples representing 25 Avena species.

Across the 12 analyzed loci, the number of alleles identified among the oat samples ranged from 2 to 9 per locus. The lowest allele numbers were detected for the intron sequences of atpF (WCt_10), and inf170ycf3 (WCt_12), which exhibited 2 and 3 alleles, respectively. The highest number of alleles (9) was observed at three loci: matK–5′trnK (WCt_1), petN–trnC (WCt_5), and ndhF–rpl32 (WCt_24). The gene diversity index (H) for individual loci ranged from 0.136 to 0.738, with an average value of 0.479. The lowest H value was recorded for the psbC–trnS locus (WCt_3), while the highest was found at the ndhF–rpl32 locus (WCt_24) (Table 3).

The Avena species differed markedly in their levels of genetic diversity. For A. prostrata and A. macrostachya, only one sample per species was available; therefore, the gene diversity index was not calculated. The highest diversity was observed in the C-genome diploid species A. ventricosa and A. eriantha (H = 0.33). In contrast, another C-genome species, A. clauda, displayed considerably lower gene diversity (H = 0.04) (Table 4).

Among A-genome species, A. atlantica (As-genome type) exhibited the highest gene diversity (H = 0.32). The lowest diversity values were detected in A. damascena (Adgenome), for which all three samples shared the same haplotype, and in A. strigosa and A. wiestii (each H = 0.04; As-genome). In A. canariensis (As-genome), the H value was relatively low (0.07), whereas in A. hirtula (As-genome) it reached 0.19.

All AB-genome species were substantially less polymorphic than A. agadiriana (AaBa-genome). The genetic diversity index for A. agadiriana was 0.30, while A. barbata and A. vaviloviana (both AB-genome) showed values of 0.07 and 0.04, respectively. No polymorphism was detected among the analyzed samples of A. abyssinica (AB-genome).

For A. insularis (AC-genome), the H value was 0.22, whereas A. magna and A. murphyi, which share the same genome type, exhibited lower values of 0.11 and 0.08, respectively. Among hexaploids, A. sterilis showed the lowest diversity (H = 0.07); in the remaining species, H ranged from 0.15 to 0.19 (Table 4).

Based on allele length variation, a total of 45 haplotypes were identified among the 70 samples based on allele length variation. Most samples were therefore characterized by unique haplotypes (Table 1). Identical haplotypes were detected in several cases: among samples of the same species (e.g., A. damascena, A. longiglumis, A. magna, A. clauda, A. ventricosa, A. strigosa); among species sharing the same genome type (haplotype 8 in A. wiestii and A. hirtula, both As-genome; haplotypes 29, 33, 34 in hexaploid ACD-genome species); and between polyploid species and their diploid progenitors (haplotypes 8, 9, 11 shared by diploid As-genome species and AB-genome tetraploids).

3.2 Genetic Relationship Inferred from cpSSR Markers

Multidimensional scaling (MDS) was performed, and a scatter plot was constructed (Fig. 1). The resulting clustering corresponded well to the genomic composition of the species. Three major groups were distinguishable:1—a single sample of A. macrostachya (CmCm-genome); 2—a group of diploid C-genome species; 3—a group comprising diploid species of various A-genome types, tetraploid species with AB- and AC-genomes, and hexaploid species with the ACDgenome.

Figure 1: A multidimensional scaling plot of Avena cpSRR data. Symbols: red plus—A. macrostachya accession (CmCmgenome); filled gray diamonds—Cp-genome species accessions; gray rectangles—accessions of A. ventricosa (Cv-genome); blue triangles—accessions of A. longiglumis (Al-genome); black dots—accessions of A. damascena (Ad-genome); blue square—A. prostrata (Ap-genome) accession; filled pink squares—accessions of A. canariensis (Ac-genome); green Xs—As-genome species accessions; blue stars—AB-genome species accessions; violet rings—accessions of A. agadiriana (AaBa-genome); light green diamonds—AC-genome species accessions; filled yellow triangles—ACD-genome species accessions.

For a more detailed assessment of “the third group”, it was analyzed separately. As shown in Fig. 2, samples of A. agadiriana (AaBa-genome) (Group 3-1) form the most distinct cluster. The remaining samples separate into two groups. The first includes diploid species with the As-genome and tetraploid species A. barbata, A. vaviloviana, and A. abyssinica with the AB-genome (Group 3-2).

The last group—Group 3-3-comprises diploid species A. prostrata (Ap-genome), A. damascena (Ad-genome), A. longiglumis (Al-genome), A. canariensis (Ac-genome), as well as tetraploid species with the AC-genome and hexaploid species with the ACD-genome (Fig. 2).

Figure 2: Scatter plot of Avena accessions with A-, AB-, AC- and ACD-genomes based on multidimensional scaling of cpSRR data. Symbols: violet rings—A. agadiriana accessions (AaBa-genome); green Xs—As-genome species accessions; blue stars—AB-genome species accessions; blue triangles—accessions of A. longiglumis (Al-genome); black dots—A. damascene accessions (Ad-genome); blue square—A. prostrata accession (Ap-genome); filled pink squares—accessions of A. canariensis (Ac-genome); light green diamonds—AC-genome species accessions; filled yellow triangles—accessions of ACD-genome species.

In the intergenic region ndhF-rpl32, an extended deletion was identified that was typical for all samples of the tetraploid species A. magna and A. murphyi with AC-genome and all samples of hexaploid species with ACD-genome (Fig. 3). In contrast, samples of another tetraploid species A. insularis with an AC-genome, did not possess this deletion.

Figure 3: Length polymorphism of ndhF-rpl32 region in Avena accessions. The first and the last tracks (L)—10 bp DNA ladder. a—diploid Avena accessions with C-genome and A. macrostachya accession with CmCm-genome; b—hexaploid Avena accessions with ACD-genome; c—A. insularis accessions with AC-genome; d—A. magna, A. murphyi accessions with AC-genome; e—diploid Avena accessions with A-genome and tetraploid Avena accessions with AB- and AaBa-genome.

4 Discussion

Eighteen primer pairs developed by T. Ishii and colleagues [38] to study microsatellite variability in wheat were tested for their applicability to various Avena species. Although wheat and oats belong to different supertribes within the Pooideae subfamily of Poaceae, and their most recent common ancestor (MRCA) is estimated at 33.5 million years ago [43], twelve of the eighteen molecular markers proved suitable for use in Avena. Further analysis identified the most informative loci: matK–5′trnK, petN–trnC, and ndhF–rpl32, which can be employed to evaluate chloroplast genome polymorphism across large sets of Avena samples.

Considerable differences in gene diversity were observed among the studied oat species. The most polymorphic species were the C-genome diploids A. eriantha and A. ventricosa, the As-genome diploid A. atlantica, and the tetraploids A. agadiriana (AaBa-genome) and A. insularis (AC-genome).

Interestingly, A. atlantica has a relatively narrow distribution, being endemic to the Moroccan coast [44]. Despite its limited area, its gene diversity (H = 0.32) exceeded that of more widely distributed A-genome diploids such as A. longiglumis, A. wiestii, and A. hirtula. For A. canariensis, which is also narrowly distributed (Spain, Canary Islands, Fuenteventura), had a H value of only 0.07, while A. damascena (Syria, Damascus region and Morocco) exhibited no diversity (H = 0).

Among AB-genome tetraploids, A. abyssinica showed no variation among samples, reflecting its restricted distribution in Ethiopia. In contrast, A. barbata, which has a much wider range [44], exhibited the highest polymorphism among AB-genome species (H = 0.07). Another narrow-range AB-genome species, A. vaviloviana, had an intermediate diversity value (H = 0.04). Overall, AB-genome species displayed low polymorphism with very small differences among them. Earlier cytogenetic and molecular studies also reported distinct similarity among A. abyssinica, A. barbata, and A. vaviloviana [45,46].

The autotetraploid A. macrostachya appeared as the most isolated species on the scattergram, consistent with its pronounced morphophysiological differences from other oats. It is the only perennial and cross-pollinated Avena species and has been placed in a separate subgenus, Avenastrum [1,44]. Previous studies have also highlighted considerable genomic differences between A. macrostachya and other Avena species. On one hand, it possesses metacentric chromosomes similar to A-genome species [19,47,48]; on the other hand, meiosis in interspecific hybrids, banding patterns, and comparative analyses of ITS1 and ITS2 sequences indicate closer affinity to C-genome species [47,49,50,51,52,53].

Our results demonstrate marked differences between the chloroplast genomes of C-genome diploids and those of oat species with other genome types. Samples of diploid A-genome species, tetraploids with AB-, AaBa-, and AC-genomes, and hexaploids with the ACD-genome cluster together on the scattergram, whereas C-genome diploid species are distinctly separated. These findings are consistent with previous reports indicating different cytoplasm types in A- and C-genome diploids and support the hypothesis that the cytoplasmic genome of polyploid Avena species was derived from A-genome diploids [54,55,56,57,58].

Within the group of samples with A-type cytoplasm, those of A. agadiriana (AaBa-genome) were the most isolated. Notably, our findings do not support Fu’s [54] observation of a closer relationship between the chloroplast genomes of A. agadiriana and A. longiglumis (AlAl). In the scattergram based on chloroplast microsatellite loci, these species form separate clusters. The origin of the A. agadiriana genome remains unclear. Previous comparisons of its chromosomes with the karyotypes of diploid Avena species bearing different A-genome variants suggested that one of its ancestors could be A. damascena [45]. However, in our study, the haplotypes of A. agadiriana differed markedly from those of A. damascena and other species. As noted above, A. agadiriana also exhibited high polymorphism. The observed remarkable intraspecific diversity in its chloroplast genome sequences is consistent with previous reports of high nuclear genome polymorphism in this species [59,60,61,62]. Considering the relative conservatism of the chloroplast genome, the pronounced differences of A. agadiriana from other oat species, and its considerable intraspecific variation, it is possible that this species originated in ancient times, involving a currently unknown primitive A-genomic diploid as the cytoplasm donor.

Based on karyotype differences and analyses of chromosome pairing in F1 hybrids, genome A has been subdivided into several types, denoted by the letter “A” with the following indices: As, Ap, Al, Ac, and Ad [19,63,64,65,66]. According to our data, the chloroplast genomes of species with different A-genome types also differ considerably. Diploid A-genome species are divided into two groups. The first group includes diploid species A. wiestii, A. hirtula, A. atlantica, and A. strigosa (As-genome), together with tetraploid species A. barbata, A. vaviloviana, and A. abyssinica (AB-genome). The second group includes diploid species A. prostrata (Ap-genome), A. damascena (Ad-genome), A. longiglumis (Al-genome), and A. canariensis (Ac-genome), as well as tetraploid species with AC-genomes and hexaploid species with the ACD-genome.

The observed similarity of haplotypes among the AB-genome tetraploid species A. abyssinica, A. barbata, and A. vaviloviana with those of diploid As-genome species supports the hypothesis that the ancestor of these tetraploids was a diploid species from the As-genome group. Previous studies [19,67,68,69,70] also support the origin of A. abyssinica, A. barbata, and A. vaviloviana from As-genome diploids. Notably, seven of the nine analyzed AB-genome samples shared haplotype 8, which was also present in two of three A. wiestii samples and in the A. hirtula sample—both diploid As-genome species. Sample v522 of A. vaviloviana had haplotype 9, also detected in sample ws550 of A. wiestii, and sample br591 of A. barbata carried haplotype 11, also found in sample hr594 of A. hirtula. Consequently, no species-specific haplotypes were identified for AB-genome tetraploids; all analyzed samples possessed haplotypes shared with the diploid As-genome species A. wiestii and A. hirtula, which are likely the maternal progenitors of these tetraploids. The absence of species-specific haplotypes also suggests a relatively recent origin for this tetraploid group.

The similarity of haplotypes between tetraploid and hexaploid species with AC and ACDgenomes and diploid species with Ap-, Al-, Ac-, and Ad-genomes aligns with previous findings, supporting the derivation of the cytoplasm of ACgenome tetraploids from A-genome diploids and the cytoplasmic genome of hexaploid oats from ACgenome tetraploids [56,57,58,68,71,72]. However, the direct ancestor and exact cytoplasm donor of hexaploid oat species remain unknown. Following the description of Avena magna in 1968, it was considered the most likely ancestor of cultivated hexaploid oats [73]. After the discovery of A. insularis in 1998, chromosome pairing and hybridization studies with A. sativa as well as FISH-analysis suggested that A. insularis was closer to hexaploids than any other known tetraploid [74,75].

Our results, however, do not support this assumption. In the intergenic region ndhF–rpl32, an extended deletion of approximately 50 bp was detected, which is typical for all samples of the tetraploid species A. magna and A. murphyi (AC-genome) and all hexaploid ACDgenome species (Fig. 3). In contrast, A. insularis samples, as well as all diploid A-genome species, lacked this deletion. This deletion does not appear to be a microsatellite variant but rather another type of indel mutation. Because such a deletion represents a rare evolutionary event, it is highly unlikely to have occurred independently in different species. Therefore, A. insularis is likely the most primitive among AC-genome tetraploids and cannot be the direct maternal ancestor of hexaploid species. A. magna and A. murphyi likely share a common ancestor, and hexaploid oat species may have descended from one of these tetraploids or their shared ancestor. The antiquity of A. insularis among AC-genome tetraploids is further supported by its high genetic diversity.

It is also noteworthy that the chloroplast microsatellite markers used in this study were more informative than those used by Li and colleagues [76], with an average of 5.9 alleles per locus compared to 3.2 in their study. Our findings do not confirm Li et al.’s hypothesis that different diploid species could serve as cytoplasmic donors for various AC- and ACD-genome species.

5 Conclusion

This study demonstrates the possibility of transferring of chloroplast microsatellite markers between species belonging to different “core pooids” supertribes within the Pooideae subfamily (Poaceae). The most polymorphic species were the C-genome diploids A. eriantha and A. ventricosa, the As-genome diploid A. atlantica, and the tetraploids A. insularis (AC-genome) and A. agadiriana (AaBa-genome). Although A. insularis is often considered the species most closely related to hexaploid oats, it is likely the most primitive among tetraploid species with an AC genome and cannot be the direct maternal ancestor of hexaploid species. The absence of species-specific haplotypes in tetraploid species with an AB- genome suggest a recent origin for this group. The most possible maternal donors of these tetraploids are the the As-genome diploid species A. wiestii and A. hirtula. In contrast, A. agadiriana with the AaBa-genome, differed markedly from other oat species, this, together with the notably high level of intraspecific variation observed, may indicate the ancient origin for this species. Overall, this study identified new informative chloroplast markers for the genus Avena and provided further insights into the phylogenetic relationships among oat species.