Identification of Hybrids in Potamogeton: Incongruence between Plastid and ITS Regions Solved by a Novel Barcoding Marker PHYB

Potamogeton is one of the most difficult groups to clarify in aquatic plants, which has an extensive range of interspecific morphological and ecological diversity. Internal transcribed spacer (ITS) is prevalent for phylogenetic analysis in plants. However, most researches demonstrate that ITS has a high percentage of homoplasy in phylogenetic datasets. In this study, eighteen materials were collected in Potamogeton from China and incongruence was shown between the rbcL and ITS phylogenies. To solve the discrepancy, we employed a novel barcode PHYB to improve resolution and accuracy of the phylogenetic relationships. The PHYB phylogeny successfully resolved the incongruence between the rbcL and ITS phylogenies. In addition, six hybrids were confirmed using PHYB, including P. compressus × P. pusillus, P. octandrus × P. oxyphyllus, P. gramineus × P. lucens, P. distinctus × P. natans, P. distinctus × P. wrightii, and S. pectinata × S. amblyophylla. Whereas, only one hybrid was identified (P. compressus × P. pusillus) by ITS, indicating that ITS homoplasy was present in Potamogeton and ITS was completely homogenized to one parental lineage. Thus, ITS might have limited utility for phylogenetic relationships in Potamogeton. It is recommended that a three-locus combination of chloroplast DNA gene, ITS and PHYB is potential to effectively reveal more robust phylogenetic relationships and species identification.


Introduction
Potamogeton is a cosmopolitan group of aquatic herbs with submersed or floating leaves. Traditionally, species in this genus can been divided into two subgenera Potamogeton and Coleogeton [1,2]. However, substantial researches suggest that subgenus Coleogeton should be elevated to the generic level and named it Stuckenia [3,4]. Moreover, two morphological characteristics have been recognized: linear-leaved and broad-leaved groups [1,5]. On the other hand, the Potamogeton species are also classified into heterophyllous and homophyllous types.
Potamogeton (including Stuckenia) is a typically notorious group in taxonomy due to a wide range of morphological and ecological diversity [6][7][8]. Furthermore, interspecific hybridization is frequent in this genus, owing to coexisting in the same ecological niche [9]. There are 69 species and more than 50 hybrids worldwide in Potamogeton, and there are about 28 Potamogeton species distributed in China [6,10]. But only a few hybrids have been identified in this genus in China [4,5,11,12]. Given several morphological characteristics for the hybrids are similar to their parents, it is always not conclusive to identify the hybrids solely based on morphology [13,14]. Furthermore, chromosomes for the Potamogeton species are so small that it is difficult to apply accurate cytological assays to identify the species, including chromosome numbers counting and in-situ hybridization [4,9,15].
Fortunately, barcoding markers have been successfully carried out to resolve substantial mysteries for phylogenetic relationships. Internal transcribed spacer (ITS) is prominent for phylogenetic analysis in plants [16]. ITS has an advantage for phylogenetic reconstruction and species identification, including universality, simplicity, intragenomic uniformity, intergenomic variability, and low functional constraint [16]. Thus, ITS is a dominate maker in phylogenetic analysis, containing more than one third phylogenetic researches [16]. In addition, ITS is considered to be the best performing DNA barcode in Potamogeton compared with rbcL, matK, and trnH-psbA [12,17]. Du et al. confirmed six putative hybrids using a combination of ITS and rbcL markers [11,12]. Wang et al. dissected phylogenetic relationships and hybrid origin of Potamogeton species in China using ITS [4]. However, ITS exhibits a high percentage of homoplasy, owning to compensatory base change, paralogy, pseudogene, sequencing error, alignment problem, incomplete concerted evolution, and a combination of these phenomena [16]. Therefore, homoplasy markedly reduces its reliability for phylogenetic reconstruction and species identification.
An alternative routine is to utilize low-copy nuclear genes to improve resolution and accuracy for phylogenetic analysis [18,19]. Low-copy genes are biparentally inherited and have lower homoplasy than ITS [16,20]. Moreover, these genes contain codons to facilitate homologous alignments and limit alignment ambiguity. Substantial nuclear genes have been developed for phylogenetic analysis, such as argenine decarboxylase coding sequence [20], phytochrome [21][22][23], alcohol dehydrogenase [24,25], LEAFY [26], and CHS [27]. PHYB is a group of phytochrome family and it has been used for phylogenetic constructions within Poaceae, Celastraceae and so on [22,28]. In the present study, we selected eighteen materials in Potamogeton from China. We initially used a combination of chloroplast DNA (cpDNA) gene rbcL and nuclear ribosomal DNA (nrDNA) region ITS to investigate whether incongruence was occurred between the rbcL and ITS phylogenies. Then a nuclear gene PHYB was employed to effectively explore more robust phylogenetic relationships. This study might lay a foundation for future studies in Potamogeton.

Plant materials and DNA extraction
Initially, we confirmed eighteen materials in Potamogeton based on taxonomy described by Wiegleb and Kaplan [6]. Geographic coordinates of all collected samples are list in Table 1. In total, ten species were identified solely according to the taxon, including eight species in Potamogeton and two species in Stuckenia ( Table 1). None of the collected samples are endangered in China and no permits are required. In addition, six putative materials were also collected in this study and we named them according to the ITS phylogenic tree except that the identified hybrids were based on parental lineage. Voucher species were deposited in the Wuhan University herbarium. We also downloaded substantial sequences from GenBank to elevate the discrimination power (S1 Table). Ruppia maritime was used as an outgroup in the phylogenetic analysis, which is close relative to Potamogeton [29].
All sampled materials were collected from fresh preserved leaf tissue and stored with silica gel. Total DNA from the dried materials (40 mg) was isolated using the Plant Genomic DNA Kit (TianGen, Beijing, China) according to the manufacturer's instruction.

PCR amplification and sequencing
The cpDNA gene rbcL was amplified using the primer 1375 and primer 26 described by Iida et al. [30]. The nuclear region ITS was amplified using ITS1 and ITS4 [31,32]. Initially, the primers B-up and B-down were used to amplify PHYB according to Mathews et al. [22]. Then specific primers were designed to amplify partial PHYB sequences (PF: ATGTGACA-CAGTTGTGGACCA; PR: CATCATCCTTGTCTTCAGGGT). The PCR reaction mixtures were 50 μL, containing 10-35 ng total DNA, 2.5 mM dNTP, 1.5 mM MgCl 2 , 50 mM KCl, 5 μM forward and reverse primers, and 2 units ExTaq DNA polymerase (Takara, Dalian, China). PCR amplification profiles consisted of 5 min at 94˚C for initial denaturation; 35 cycles of 1 min at 94˚C, annealing for 1 min at 55˚C, 1 min at 72˚C for extension; and 10 min at 72˚C for a final extension step. All PCR products were further purified by the High Pure PCR Product Purification Kit (Roche) according to the manufacturer's recommendation. The purified PCR products for rbcL were directly sequenced in both directions using the specific primers. For the two nuclear barcodes, the purified PCR products were cloned into the PMD19-T vector following the manufacturer's instruction (Takara, Dalian, China). Then five positive clones were sequenced in both directions. All sequencings were performed on an ABI 3730 DNA Sequencer using BigDye Terminator version 3.1 (Applied Biosystems). For identical sequences from the cloned PCR products, only one sequence was contained in the dataset. All obtained sequences in this study were deposited in GenBank (Table 1).

Sequence analysis
Two complementary sequences were assembled using ContigExpress in Vector NTI Suite 2.0 v5.5.1 [33]. All assembled sequences were further aligned using Clustal X v2.0 [34]. Then the aligned sequences were used to construct the phylogenetic trees using maximum-parsimony (MP) and neighbor-joining (NJ) methods. MP analysis was performed using the program PAUP Ã v. 4.0b10 [35]. Heuristic search strategy was conducted using 1000 replicates with random taxon-addition sequences, in combination with tree-bisection-reconnection (TBR) branch swapping, and with the options MULPARS in effect and STEEPEST DESCENT off [36]. Gaps and random addition replicates were considered as missing bases. A strict consensus tree was generated by setting maxtrees at 20000. The NJ tree was constructed using the program Mega 6.0 with 1000 bootstrap replicates based on the a Kimura-2 parameter distance matrix [37]. Nodes with bootstrap values less than 50 were collapsed.

ITS
The boundary of ITS was identified by compared with previous published sequences [4,11,17]. Length of the ITS sequences for the eight Potamogeton species was identical with 630 bp. Length of the two accessions for Stuckenia amblyophylla was 654 bp and that for Stuckenia pectinata was 645 bp. In addition, existing DNA sequences for twelve Potamogeton species were downloaded from GenBank to further evaluate the discriminatory power (S1 Table). Two unique sequences for each material were obtained from five clones. Topology characters of the MP and NJ trees were similar with minor variations in bootstrap values. The MP analysis was set as the heuristic search and limited to 20000 trees. Thus, a strict consensus tree was obtained with the tree length of 446, consistency index (CI) of 0.922, and retention index (RI) of 0.966 (Fig 1). The phylogenetic tree revealed that the two clones of P. compressus1 were separately clustered together with the species P. compressus and P. pusillus, which were incompatible with the morphological identification. Nevertheless, identification of the other species using ITS was consistent with morphological features.

rbcL
Length of the marker rbcL was 881 bp and no sequence variation was observed among multiple sequencings for per sample. For phylogenetic analysis, MP and NJ tree showed similar topology features except minor variations in bootstrap values. The MP analysis yielded a strict consensus tree with the tree length of 91, CI of 0.901, and RI of 0.953. The rbcL phylogeny revealed that the maternal parent of P. compressus1 was P. pusillus (Fig 2). However, identification of the other materials using rbcL was consistent with the morphological identification and ITS phylogenetic analysis. For example, the maternal parent of P. lucens1 was P. lucens; the maternal parent of P. wrightii1 was P. wrightii; the maternal parent of P. natans1 was P. natans; the maternal parent of P. oxyphyllus1 was P. oxyphyllus; and the maternal parent of S. pecti-nata1 was S. pectinata.

PHYB
Considering the rbcL and ITS phylogenies were incompatible, it is necessary to use another marker to elevate resolution and accuracy. In this study, we used a nuclear gene PHYB to further improve the discriminatory power. Two accession numbers were obtained for each material and sequence of R. maritime (AB508058.1) was used as an outgroup in the phylogenetic analysis. Length of PHYB was identical with 930 bp. Topology characters of the MP and NJ trees were similar with minor variations in bootstrap values. The MP analysis was set as heuristic search and limited to 20000 trees. The MP analysis yielded a strict consensus tree with the tree length of 447, CI of 0.747, and RI of 0.841. The phylogenetic tree showed that the one clone of P. compressus1 was clustered with P. compressus, and the other clone probably was clustered with P. pusillus (Fig 3). Nonetheless, identification of the materials P. oxyphyllus1, P. lucens1, P. natans1, P. wrightii1, and S. pectinata1 was incongruent with the ITS phylogeny.
The PHYB phylogeny showed that two clones of P. oxyphyllus1 were separately assembled with P. octandrus and P. oxyphyllus; two clones of P. lucens1 were separately assembled with P. gramineus and P. lucens; two clones of P. wrightii1 were separately clustered with P. distinctus and P. wrightii; and two clones of S. pectinata1 were separately assembled with S. pectinata and S.  amblyophylla; and one clone of P. natans1 was clustered with P. distinctus. Nevertheless, two clones of the remaining materials were clustered together, indicating that these materials were pure species, which was consistent with the ITS phylogeny (Fig 3).

Discussion
Potamogeton is a notoriously group in taxonomy due to high morphological and ecological diversity [6,7]. Moreover, hybridization and polyploidization are prevalent in Potamogeton [6,9]. Traditionally, cytological experiments are used to distinguish the interpretation of hybridization and polyploidization [15]. However, chromosomes for the Potamogeton species are extremely small, and it is difficult to use cytological assays accurately to identify them, such as counting chromosome numbers, in-situ hybridization, and karyological characters. In general, molecular barcodes are popular to investigate the phylogenetic relationships in Potamogeton [4,11,12,17]. ITS is considered to be one of the most prominent markers in phylogenetic analysis [16]. In this study, the ITS phylogeny demonstrated the material P. compressus1 was a hybrid between P. compressus and P. pusillus; and the rbcL phylogeny showed that the maternal parent of P. compressus1 was P. pusillus. The incompatible results contributed to uniparental origin of the cpDNA gene [16,38]. Low-copy nuclear genes are increasingly applied to obtain a better understanding of phylogenetic analysis, such as LEAFY [26,38], PHYB [21,22], and CHS [27]. The low-copy genes are potential to improve the robustness of phylogenetic reconstruction at taxonomy [19]. They are particularly effective to resolve closely interspecific relationships in plants. Moreover, they are helpful to generate strong phylogenetic relationships where universal barcoding markers, such as nrDNA and cpDNA genes, are ambiguous to reveal the internal relationships. PHYB is a group of phytochrome family, including PBYA, PHYB, PHYC, PHYD and PHYE [39,40]. These proteins act as photoreceptors for red and far-red light in green algae and land plants [41,42]. Although PHYB-related subgroups show an expansion in Arabidopsis that PHYD and PHYE are closely related to PHYB, there is no evidence for concerted evolution for PHYB [40]. Moreover, PHYB has been used for many phylogenetic studies [28,43]. Southern blot analysis and PCR supply show that PHYB is a single-copy gene in grasses except maize [28,44].
In this study, the first exon of PHYB was amplified to identify the Potamogeton species. If members of a low-copy gene family are used in the phylogenetic analysis, paralogs should be easily distinguishable. Then we can design specific primers and provide relatively accurate phylogenetic characters [20]. PHYB was initially amplified according to the prior research, and then the specific primers were designed in Potamogeton [22]. Furthermore, the first exon of PHYB is useful to infer the intraspecific relationships with no inferred gaps, unambiguous sequence alignment, and many parsimony informative positions [43,45].
The PHYB phylogeny in this study was the first time to address comprehensively phylogenetic relationships in Potamogeton. The ITS phylogeny indicated that P. compressus1 was a hybrid between P. pusillus and P. compressus. The maternal parent of P. compressus1 was P. pusillus based on the rbcL phylogeny. Combination analysis of the PHYB and rbcL phylogenies for P. compressus1 revealed the same conclusion with the ITS phylogeny. These results revealed that species identification solely based on cpDNA genes was controversial and demonstrated that phylogenetic analysis using PHYB was receivable [38]. Given limited samples in this study, it is problematic to identify the material P. natans1 solely based on one marker. A combination of rbcL, ITS and PHYB indicated that P. natans1 was a hybrid between P. distinctus and P. natans. Additionally, only one hybrid (P. compressus1) was identified in the ITS phylogeny. However, the PHYB phylogenetic tree revealed that P. oxyphyllus1 was a hybrid between P. octandrus and P. oxyphyllus; P. lucens1 confirmed the origin from hybridization between P. gramineus and P. lucens; P. wrightii1 was derived from hybridization between P. distinctus and P. wrightii; and S. pectinata1 confirmed the origin from hybridization between S. pectinata and S. amblyophylla. Previous researches showed that ITS has a high level of homoplasy [16]. ITS homoplasy is prevalent in phylogenetic analysis, owning to compensatory base changes, pseudogene, paralogy, lack of complete concerted evolution, and alignment or sequencing problem [16]. These unpredictable evolutionary events and complex behaviors obviously attenuated its power for phylogenetic analysis. Several researches demonstrated that concerted evolution within different tandem repeats of ITS regions can quickly eliminate one parental pattern and completely homogenized to the other parental lineage [46,47]. Additionally, ITS probably forms a chimeric mixture of ITS types or maintain biparental types [16,48]. The incongruence between the ITS and PHYB phylogenies revealed that ITS homology was also existed in Potamogeton. ITS is homogenized to the maternal lineage for the five hybrids, including P. octandrus × P. oxyphyllus, P. gramineus × P. lucens, P. distinctus × P. natans, P. distinctus × P. wrightii, and S. pectinata × S. amblyophylla. However, identification of most materials was consistent between the PHYB and ITS phylogenies, indicating that the PHYB phylogeny was reliable and ITS was helpful for phylogenetic analysis to a certain extent. Thus, a combination of ITS and cpDNA gene is unable to accurately provide evidence for hybrid origin in Potamogeton. A three-locus combination of cpDNA, ITS and PHYB probably is a potential choice to identify the Potamogeton species and further reveal phylogenetic relationships in Potamogeton.
In summary, PHYB was used to identify the Potamogeton species and six hybrids were confirmed for the PHYB phylogeny (P. compressus × P. pusillus, P. octandrus × P. oxyphyllus, P. gramineus × P. lucens, P. distinctus × P. natans, P. distinctus × P. wrightii, and S. pectinata × S. amblyophylla).Whereas, only one hybrid (P. compressus × P. pusillus) was identified for the ITS phylogeny. The data indicated that ITS homoplasy was present in Potamogeton and ITS was completely homogenized towards one parental lineage. Thus, ITS did not effectively reveal the phylogenetic relationship in Potamogeton. A three-locus combination of cpDNA gene, ITS and PHYB probably can obtain more robust insights into the evolutionary and phylogenetic relationships in Potamogeton.
Supporting Information S1