Introduction

The Bamboo Phylogeny Group (BPG 2006, 2012) suggested that a taxonomic reorganization of bamboos by means of a reclassification strongly supported by estimate of phylogeny was necessary, which can be addressed with plastid genome (plastome) markers (Kelchner and Group 2013). The advancement of next-generation sequencing technologies has enabled the rapid acquisition of whole plastome sequences at low cost when compared with traditional sequencing approaches. Thus, the increasing number of plastome sequences in the public databases is now an efficient option for increasing phylogenetic resolution and evolutionary studies between and within different plant groups, families, genus and species (Zhang et al. 2011; Wysocki et al. 2015; Rogalski et al. 2015).

Besides phylogenetic studies, the low evolving rate, and the predominantly nonrecombinant, uniparentally inherited nature of plastome may greatly facilitate the use of plastid DNA markers in plant population genetic studies (Powell et al. 1995; Provan et al. 2001; Rogalski et al. 2015). By comparing nuclear and plastid markers, inferences about the relative contributions of seed and pollen flow to the genetic structure of natural populations are possible (Provan et al. 2001; Delplancke et al. 2012; Roullier et al. 2011; Khadivi-Khub et al. 2013). Powell et al. (1995) reported the presence of simple sequence repeats in plastids (cpSSR), consisting of DNA sequences in tandemly repeated motifs of 1–6 base pairs (bp). This cpSSR became widely studied due to their ability to generate highly informative DNA markers. The development and application of these plastid molecular markers was demonstrated by several studies (see review by Rogalski et al. 2015). Plastid genome sequences are also useful tools utilized to study plastid gene transfer to nucleus (Huang et al. 2003, 2005; Stegemann et al. 2003; Bock 2006; Stegemann and Bock 2006), and recently, evidence for horizontal transfer of mitochondrial DNA to the plastid genome in Pariana Aubl., a bamboo genus was found (Ma et al. 2015).

Bamboos are indigenous of all continents, except Antarctica and Europe, and occupy a broad range of habitat types, especially forests, from temperate to tropical climate zones (Clark et al. 2015). Brazil holds a high diversity of bamboo species, of which we can highlight Merostachys Spreng. (43 native species—41 endemic species), Chusquea Kunth (45 native species—41 endemic species) as the most common genera, and Guadua Kunth (18 native species—5 endemic species) as the highest important for use in construction, furniture, and handcraft (Greco et al. 2015; List of Species of the Brazilian Flora 2015).

Merostachys s.l. is the largest genus of the subtribe Arthrostylidiinae Soderstr. and R.P. Ellis (Lizarazu et al. 2011). This genus spreads from Mexico to Southern Brazil in forest and forest margin habitats, but it is particularly diverse in Brazil, where remains many undescribed species (Judziewicz et al. 1999; Lizarazu et al. 2011; Santos-Gonçalves et al. 2012; Viana et al. 2013; Viana and Filgueiras 2014; Greco et al. 2015). Its taxonomic identification has been demonstrated to be very complicated due to their long periods of vegetative development, ca. 25–30 years (Lizarazu et al. 2011).

Guadua s.l. is one of the five genera placed in the subtribe Guaduinae Soderstr. and R.P. Ellis. This genus occurs throughout tropical America, from Mexico to Brazil and northern Argentina (Londoño and Peterson 1992). Guadua chacoensis (Rojas) Londoño & P.M. Peterson is a native species from Brazil, but also occurs in northern Argentina, southeastern Bolivia, and southern Paraguay, and is one of the three southeasternmost species of the genus (Londoño and Peterson 1992).

Here, we sequenced and analyzed the complete plastome sequences of two Brazilian native species of tribe Bambuseae: G. chacoensis and Merostachys sp. We also performed a full plastome phylogeny using 20 Bambuseae species with 2 newly sequenced and 18 existing plastomes, and characterized the occurrence, type, and distribution of SRRs in the Bambuseae. Merostachys sp. is the first species of subtribe Arthrostylidiinae to have its plastome sequenced.

Materials and methods

Sequencing and assembly of the bamboo plastomes

Fresh leaf material of Merostachys sp. (voucher Th. Greco 18—FLOR_16/09/2011) were collected from a natural population, and Guadua chacoensis (voucher Th. Greco 159—FLOR_15/02/2013) were collected from cultivated plants in Florianópolis, Santa Catarina—Brazil (27°35′49″S48°32′56″W). Both species were located on a private property and were collected with the owner permission. The chloroplast isolation and plastid DNA extraction were carried out using leaves from a single plant following the protocol described by Vieira et al. (2014). A total of 1 ng of plastid DNA was used to prepare sequencing libraries with Nextera XT DNA Sample Prep Kit (Illumina Inc., San Diego, California, USA) according to the manufacturer’s instructions. Libraries were sequenced using MiSeq Reagent Kit v3 (600 cycles) on Illumina MiSeq Sequencer (Illumina Inc., San Diego, California, USA).

The obtained paired-end reads (2 × 300 bp) were used for de novo assembly performed by CLC Genomics Workbench 8.0v. Initial annotation of the obtained plastome sequences was performed using Dual Organellar GenoMe Annotator (DOGMA) (Wyman et al. 2004). From this initial annotation, putative starts, stops, and intron positions were determined based on comparisons to homologous genes in other plastomes. The tRNA genes were further verified by using tRNAscan-SE (Schattner et al. 2005). The physical map of the circular plastomes was drawn using OrganellarGenomeDRAW (OGDRAW) (Lohse et al. 2013). REPuter (Kurtz et al. 2001) was used to identify the IRs in both plastomes sequenced in this work and those used for phylogeny estimation by forward versus reverse complement (palindromic) alignment. The minimal repeat size was set to 30 bp and the identity of repeats ≥90 %. The complete nucleotide sequence of Merostachys sp. and G. chacoensis plastome were deposited in the GenBank database under accession number KT373815 and KT373814, respectively.

Phylogeny estimation

The phylogeny estimation was done according to Wysocki et al. (2015). The IRA was omitted to prevent over representation of the IR sequences. G. chacoensis and Merostachys sp. plastomes were aligned using MAFFT program (Katoh and Standley 2013) along with 18 previously published Bambuseae plastomes (Table 1) and Lolium perenne (Poaceae: Pooideae; NC_009950) plastome was used as outgroup. Nucleotide positions that contained one or more gaps introduced by the alignments were omitted from the matrix. The substitution model was selected using jModeltest (number of substitution schemes = 5). The General Time Reversible model of substitution, incorporating invariant sites and a gamma distribution (GTR + I + G), was used in subsequent plastome analyses. Maximum likelihood (ML) analysis was performed using the RAxML v 7.2.8 (Stamatakis 2006) with 1000 non-parametric bootstrap replicates. MrBayes 3.2.2 (Ronquist and Huelsenbeck 2003) was used to perform a Bayesian inference (BI) analysis. The Markov chain Monte Carlo (MCMC) analysis was run for 2,000,000 generations. Average standard deviation of split frequencies remained below 0.001 after the 25 % burn-in. Resulting trees were represented and edited using FigTree v1.4.1.

Table 1 Comparison of plastid genome of tribe Bambuseae (Poaceae) analyzed in this study
Full size table

Simple sequence repeats analysis

Simple sequence repeats (SSRs) were detected using MISA perl script, available at (http://pgrc.ipk-gatersleben.de/misa/), with threshold of eight repeat units for mononucleotide SSRs, four repeat units for di- and tri-nucleotide SSRs, and three repeat units for tetra-, penta- and hexa-nucleotide SSRs. For the SSRs identification, we used the plastid genome sequence of all species described in Table 1 with one IR region removed, and in their coding sequences (CDS). Additionally, we compared the identified SSRs in order to find a potential set of microsatellite markers identified for all Bambusea species described in Table 1. Primers were designed for 16 interspecific polymorphic SSR with the software PRIMER3 (http://bioinfo.ut.ee/primer3-0.4.0/) by setting product size ranges from 100 to 300 bp, primer size from 18 to 25 bp, and GC content from 40 to 60, with 1 °C as the maximum difference between the melting temperatures of the left and right primers.

Results

Plastome assembly and content

The Illumina MiSeq reads obtained and submitted to de novo assembly resulted in high genome coverage for both G. chacoensis (~140×) and Merostachys sp. (~460×) plastomes (Table 2). The determined complete plastome sequence of Merostachys sp. and G. chacoensis is 136,334 and 135,403 bp in size, respectively, with a GC content determined of 38.81 % for Merostachys sp. and 38.76 % for G. chacoensis. The Guadua genus presents the smaller plastome size in comparison with other species belonging to tribe Bambuseae (135,324–135,403 bp), ~1000 bp smaller than Merostachys sp., but with an identical gene content (Tables 1, 3). Both plastomes exhibit the general quadripartite structure typical of angiosperms, consisting of a pair of IRs (20,258–19,773 bp) separated by the LSC (82,859–82,877 bp) and SSC (12,960–12,980 bp) regions (Table 1). The IR/SSC boundary is just within the coding sequence of ndhH, creating a short ndhH fragment in IRA. They encode an identical set of 132 genes and 4 pseudogenes with the same gene order and gene clusters, of which 90 genes were single copy and 21 genes were duplicated (Fig. 1; Table 3). The following genes were identified and are listed in Fig. 1 and Table 3: 4 duplicated ribosomal RNA genes, 21 unique and 9 duplicated transfer RNA genes, 15 unique and 6 duplicated genes encoding large and small ribosomal subunits, 1 translational initiation factor, 4 genes encoding DNA-dependent RNA polymerases, 45 unique and 1 duplicated genes encoding photosynthesis-related proteins, 4 unique and 1 duplicated genes encoding other proteins, including the unknown function gene ycf68.

Table 2 Guadua chacoensis and Merostachys sp. plastid genome sequencing and assembly data
Full size table
Table 3 List of genes identified in Guadua chacoensis and Merostachys sp. plastid genome. The gene content of these species is strictly the same
Full size table
Fig. 1
figure 1

Gene map of Guadua chacoensis plastid genome. Genes drawn inside the circle are transcribed clockwise, and genes drawn outside are counterclockwise. Genes belonging to different functional groups are color-coded. The darker gray in the inner circle corresponds to GC content, and the lighter gray to AT content. Gene map of Merostachys sp. is equal to G. chacoensis in gene order and content

Full size image

Phylogeny estimation

The matrix of the 21 completely aligned Bambusoideae plastomes was 122,950 nucleotide positions in length, and the exclusion of gaps reduced this matrix to 105,270 sites. ML (−lnL = −206179.38) and BI (−lnL = −206212.43) analyses produced phylogenomic trees that were identical in topology, which is represented by Fig. 2. The ML bootstrap values between 68 and 98 % for 4 nodes and 100 % for the rest, and the BI posterior probability values between 0.95 and 0.99 for 3 nodes and 1.0 for the rest (Fig. 2). These trees supported monophyly of Paleotropical and Neotropical Bamboos clades. The Neotropical bamboos segregated into three well-supported lineages, Chusqueinae, Guaduinae, and Arthrostylidiinae, with Guaduinae forming a well-supported sister relationship with Arthrostylidiinae. All Neotropical bamboos genera (Chusquea, Guadua, Olmeca, Otatea, Merostachys) were resolved as monophyletic with 100 % ML bootstrap support (Fig. 2). The Paleotropical bamboos segregated into two well-supported lineages, Hickeliinae and Bambusinae + Melocanninae. The genus Bambusa was resolved as monophyletic with 100 % ML bootstrap support with very short branches and two internal nodes with 81 and 98 % ML bootstrap support (Fig. 2). Dendrocalamus latiflorus (Bambusinae) and Neohouzeaua sp. (Melocanninae) formed a well-supported node, with 100 % ML bootstrap support. These two species are nowadays classified as belonging to different subtribes (Clark et al. 2015). Greslania sp. and Neololeba atra form a well-supported sister relationship.

Fig. 2
figure 2

Bayesian phylogeny of 20 complete plastomes sequences of Bambuseae tribe species and the outgroup Lolium perenne (Poaceae: Pooideae). The numbers above the branches are maximum likelihood bootstrap values/Bayesian posterior probabilities. The branch length is proportional to the inferred divergence level, the scale bar indicating the number of inferred nucleic acids substitutions per site

Full size image

Simple sequence repeats analysis

We analyzed the occurrence, type, and distribution of SRRs in the Bambuseae species listed in Table 1. The number and types of SSR identified for all 20 Bambuseae species were similar (Fig. 3). In the complete plastome sequence, the average of the total identified SSR in all Bambuseae species were 141.8 (1 SSR unit every 823.14 bp), while for coding sequences a lower value was found (38.15; 1 SSR unit every 1466.60 bp). Among them, mono- and di-nucleotide repeats were the most common, whereas tri- and tetra-nucleotide repeats occurred with lower frequency (Fig. 3). Penta- and hexa-nucleotide repeats occurred only in complete plastome sequence, no penta- or hexa-nucleotide repeats were identified in coding sequences for any analyzed species.

Fig. 3
figure 3

SSR unit size distribution in Bambuseae tribe complete plastome sequences and coding sequences. For this analysis, we considered as complete plastome sequence the SSC, LSC and only one copy of the IR sequence

Full size image

Although we observed that the number and type of SSR identified in Bambuseae species were quite similar, the same was not true for their presence across all species. Less than half of SSR (di-, tri-, tetra-, and penta-nucleotide) were common across all species (Table 4). In addition, all these SSR cited in Table 4 were assessed for interspecific polymorphism, and except for (TTCTA)6–7, common to G. weberbaueri and G. chacoensis, respectively, no other SSR showed interspecific polymorphism. For mononucleotide repeats, it was possible to identify 16 polymorphic SSR loci with conserved down- and up-stream regions (Table 5). For these regions specific primers were designed, and the polymorphisms for these 16 loci ranged from 3 to 10 alleles, 15 in intergenic regions and intron 1 (Table 5).

Table 4 SSR unit type distribution for di-, tri-, tetra-, and penta-nucleotide repeats for all 20 Bambuseae species cited in Table 1
Full size table
Table 5 Set of 16 polymorphic SSR loci common across all 20 Bambuseae species cited in Table 1, including primer sequence (F: forward and R: reverse), expected size of fragment, repeat motif, number of alleles across the evaluated species, and SSR loci location (IGS: intergenic sequence)
Full size table

The most mononucleotide repeats identified were constituted by A/T sequences (96.61 %), and for the dipolymers, 56.85 % were also constituted by multiple A and T bases, no dinucleotide repeats constituted by multiple C and G bases were found (Fig. 3). Similarly, for trinucleotide repeats, 31.92 % were constituted by multiple A and T bases (AAT/ATT), 48.17 % by AAG/CTT and 19.92 % by AGC/CTG. We also identified 11 tetranucleotide repeats, with AAAT/ATTT being the most common (33.86 %). Interestingly, we also did not identified any tri- of tetra-nucleotide unit size constituted by only C or G bases.

Discussion

Plastome assembly and content

The plastome size in Neotropical bamboos was typically shorter than in Paleotropical bamboos, with determined plastomes measuring between 135,324–138,257 and 138,276–139,606 bp, respectively (Table 1). This increase in Paleotropical plastome sizes seems to be distributed by LSC and IR regions, which ranges between 80,743–82,859 and 81,925–83,145 bp for LSC regions in Netropical and Paleotropical bamboos, respectively, and between 19,773–21,804 and 21,755–21,852 bp for IR regions, also respectively. Differently, for the SSC regions, little length variation was found (12,671–12,980 bp for Neotropical and 12,743–12,980 bp for Paleotropical bamboos), suggesting that this region is more conserved in size than LSC and IR regions for Bambuseae. The IR/SSC boundary was conserved in Bambuseae tribe, in G. chacoensis, Merostachys sp. and all Bambuseae plastomes analyzed in this study (Table 1) the IR/SSC boundary was just within the coding sequence of ndhH, with a short ndhH fragment in IRA (Wu et al. 2009; Zhang et al. 2011; Burke et al. 2012). This feature is similar to the results found in other Poaceae species (Ogihara et al. 2002; Saski et al. 2007), but differ from the Olyreae species, which were shown to present the IR/SSC boundary within ndhF coding sequence, an suggested synapomorphy for the tribe (Burke et al. 2012, 2014). The gene content and gene order of the G. chacoensis and Merostachys sp. plastomes were identical to other sequenced Bambuseae plastomes, including the loss of introns in clpP and rpoC1, the loss of accD, ycf1 and ycf2 genes, and the IR expansion to include rps19 (Wu et al. 2009; Zhang et al. 2011; Burke et al. 2012).

Phylogeny estimation

Plastid genomes are very evolutionarily conserved, presenting rates of substitutions extremely low as compared to nuclear genomes (Palmer 1985). In addition, different plastome regions evolve at different rates, allowing measuring evolutionary distance at many taxonomy levels (Palmer, 1985; Shaw et al. 2005, 2007). Phylogenetic inferences based on whole plastome sequences have been used to address evolutionary features in Bambusoideae (Wu et al. 2009; Zhang et al. 2011; Burke et al. 2012, 2014; Wysocki et al. 2015). These full plastome analyzes provide enough information to resolve difficult interspecific relationships, an issue to woody bamboos that generally hybridize readily and exhibit very long generation times (Wysocki et al. 2015).

In Bambuseae, the well supported monophyletic lineages that represent Neotropical and Paleotropical woody bamboos retrieved here were previously reported in phylogenetic studies using combined analysis of plastid DNA regions (Sungkaew et al. 2009) and complete plastome sequences (Wysocki et al. 2015). In Wysocki et al. (2015), the segregation of Neotropical bamboos into two well-supported lineages, Chusqueinae and Guaduinae, was reported.

In this work, we present for the first time a phylogenetic tree based on complete plastome sequence including an Arthrostylidiinae species, providing the phylogenetic tree with well supported monophyly of subtribes and with a sister relationship between Guaduinae and Arthrostylidiinae. Previous analyses based on matK and/or ndhF sequence data also reported this sister relationship (Guala et al. 2000; Zhang and Clark 2000; Soreng et al. 2015), and a study based on morphology and rpl16 intron sequence data indicate that Guaduinae may be derived from within Arthrostylidiinae (Clark et al. 2007). In Chusqueinae, especially Chusquea spectabilis, there was a high substitution rate in both ML and BI analyses. This feature has been reported by Wysocki et al. (2015) and associated with the extremely different flower intervals in Chusqueinae species, and the shorter flowering intervals observed in C. spectabilis. In addition, C. spectabilis long branches may be associated with the fact that this species belongs to subgenus Magnifoliae, one of the two earliest-diverging clades in genus Chusquea, and C. circinata (subgenus Rettbergia) is known to be a sister of the large Euchusquea clade, represented in this analysis by C. liebmannii (Fisher et al. 2014).

Low support (82 % ML bootstrap) between the Paleotropical bamboos subtribes, Hickeliinae and Bambusinae were reported by Wysocki et al. (2015), corroborating our results with low support between Hickeliinae and Bambusinae + Melocanninae (74 % ML bootstrap support). Wysocki et al. (2015) also reported the well supported clade Neololeba atra + Greslania sp., as well as Bambusa spp. monophyly. However, the inclusion of Neohouzeaua sp. (Melocanninae), resulted in its unexpected sister relationship to D. latiflorus (Bambusinae), and consequently, the non-monophyly of Bambusinae.

Simple sequence repeats analysis

Powell et al. (1995) introduced the cpSSR as a readily usable chloroplast marker, exhibiting length variation and polymorphism. Even though in plastomes the occurrence of di-, tri-, tetra-, penta-, and hexa-nucleotide repeats is less common (George et al. 2015), it became a widely used molecular marker, which have aroused considerable interest due to their ability to generate highly informative DNA markers (Provan et al. 2001). These regions may be used for both intraspecific and interspecific variability analyses, with practical value for monitoring gene flow, population differentiation and cytoplasmic diversity (Powell et al. 1995), in both basic plant sciences and applied agricultural. In native species, cpSSRs have been used most frequently in population genetics studies, but also for understanding uniparental genetic structure (e.g. seed or pollen dispersal) and in studies of hybridization (see review by Wheeler et al. 2014). Given that not all chloroplast loci are likely to be equally diverse across plants, it is important to consider a more targeted approach to developing cpSSR loci specific to a study system rather than relying on universal primers (Wheeler et al. 2014).

The cpSSRs may be identified in completely sequenced plant chloroplast genomes by simple database searches, followed by primers designed to screen for polymorphism even in non-model species, and is becoming a common pathway for developing variable markers (Wheeler et al. 2012; Li et al. 2014; Nazareno et al. 2015). George et al. (2015) analyzed the abundance and distribution of simple and compound SSR in 164 sequenced plastomes from wide range of plants. Corroborating our results in Bambuseae, George et al. (2015) described that mononucleotide repeats occurs in higher number as compared to di-, tri-, tetra-, penta-, and hexa-nucleotide repeats, longer SSRs are excluded from coding regions, AT/TA followed by CT/TC was the most common dinucleotide repeat motif observed, and GC/CG was rarely found and even absent from few plastomes.

In the present study, we determined the complete plastome sequence of Merostachys sp. and Guadua chacoensis, which were identical in gene content and order, also with identical gene content as compared to other Bambuseae plastomes. These results enabled the report for the first time of a phylogenetic tree based on complete plastome sequences including an Arthrostylidiinae species, the Merostachys sp. The ML and BI trees supported monophyly of Paleotropical and Neotropical Bamboos clades, with all Neotropical bamboos genera resolved as monophyletic. The Paleotropical bamboos segregated into two well-supported lineages, Hickeliinae and Bambusinae + Melocanninae. We also identified several SSR loci, whose distribution and types are highly similar between sequenced plastomes. Among them, we identified 16 polymorphic SSR loci suitable for interspecific population genetic analysis, with number of alleles varying from 3 to 10. These 16 polymorphic cpSSR loci in Bambuseae plastome can be assessed for the intraspecific level of polymorphism, leading to innovative highly sensitive phylogeographic and population genetics studies for this tribe. Merostachys sp. and G. chacoensis are both native species from the Atlantic Forest Biome, which is considered one of the 25 biodiversity hotspots of the world (Myers et al. 2000), increasing the interest in population genetics studies for these species.

Availability of supporting data

All nucleotide sequences were deposited in the NCBI Genbank repository. Genbank accessions numbers can be found in Table 1.