rDNA Loci Evolution in the Genus Glechoma (Lamiaceae)

Glechoma L. (Lamiaceae) is distributed in eastern Asia and Europe. Understanding chromosome evolution in Glechoma has been strongly hampered by its small chromosomes, constant karyotype and polyploidy. Here phylogenetic patterns and chromosomal variation in Glechoma species are considered, using genome sizes, chromosome mapping of 5S and 35S rDNAs by fluorescence in situ hybridisation (FISH), and phylogenetic analyses of internal transcribed spacers (nrITS) of 35S rDNA and 5S rDNA NTS sequences. Species and populations of Glechoma are tetraploid (2n = 36) with base chromosome number of x = 9. Four chromosomes carry pericentric 5S rDNA sites in their short arms in all the species. Two to four of these chromosomes also carry 35S rDNA in subterminal regions of the same arms. Two to four other chromosomes have 35S rDNA sites, all located subterminally within short arms; one individual possessed additional weak pericentric 35S rDNA signals on three other chromosomes. Five types of rDNA locus distribution have been defined on the basis of 35S rDNA variation, but none is species-specific, and most species have more than one type. Glechoma hederacea has four types. Genome size in Glechoma ranges from 0.80 to 0.94 pg (1C), with low levels of intrapopulational variation in all species. Phylogenetic analyses of ITS and NTS sequences distinguish three main clades coinciding with geographical distribution: European (G. hederacea–G. hirsuta), Chinese and Korean (G. longituba), and Japanese (G. grandis). The paper presents the first comparative cytogenetic analyses of Glechoma species including karyotype structure, rDNA location and number, and genome size interpreted in a phylogenetic context. The observed variation suggests that the genus is still in genomic flux. Genome size, but not rDNA loci number and distribution, provides a character for species delimitation which allows better inferences of interspecific relationships to be made, in the absence of well-defined morphological differentiation.


Introduction
Cytological plasticity plays an important role in plant diversification and speciation [1][2][3]. This plasticity manifests itself in changes of chromosome number (aneuploidy/dysploidy or polyploidy), chromosome structure (inversions, translocations, additions, and deletions), genome size, and more subtle changes in sequence composition [4]. Chromosome number changes have long been used to draw evolutionary and systematic inferences, but modern molecular cytogenetic techniques allow us to trace chromosomal structural changes in a phylogenetic context [5][6][7][8]. Molecular cytogenetic mapping of 5S and 35S rDNA loci using FISH (fluorescence in situ hybridisation) has proved useful for identifying the direction of chromosomal change in closely related species, even when chromosome numbers are conserved and chromosomes are constant in morphology [9][10][11][12][13].
Chromosomal evolution in wild plant groups can be analysed in a phylogenetic context using chromosome numbers, localisation of 35S and 5S rDNA loci, and genome size variation [5,6,8,32], which may allow the directionality of chromosomal evolution to be inferred [33][34][35][36]. Comparative molecular cytological analyses are lacking in Glechoma. Here we study patterns of genome evolution in a phylogenetic context in Glechoma species, as well as establishing the dynamics of rDNA and genome size evolution in these closely related, widely distributed, polyploid taxa, using FISH and flow cytometry. The specific aims of this study are (1) to assess phylogenetic relationships of multiple populations of taxa based on ITS and 5S rDNA NTS sequences, (2) to elucidate trends in evolution of 5S and 35S rDNA in the genus, and test whether these correlate with phylogenetic relationships between the polyploids, and (3) to examine patterns of genome size change in Glechoma.

Ethics Statement
The investigated taxa are neither endangered nor protected. All material was collected at public land, where no special permissions are required.

Plant material
A total of 48 individuals, two to six individuals per population, of four species of Glechoma and three individuals of one population of Meehania urticifolia were collected from natural populations and transplanted to the Botanical Garden of the University of Vienna (Table 1).

Chromosome numbers and fluorescence in situ hybridisation (FISH)
Actively growing root-tips were pretreated with 0.002 M solution of 8-hydroxyquinoline for 2.5 h at room temperature and 2.5 h at 4˚C, then fixed in ethanol: acetic acid (3: 1) and stored at -20˚C until use. Glechoma hederacea and G. longituba have a gynodioecious sexual system comprising females and hermaphrodites [37][38]. Only hermaphrodites were examined. Chromosome numbers were initially determined by standard Feulgen staining of root tip cells [6]. To improve resolution of karyotype structure, preparations were also made using enzymatic digestion (described below), stained with 2ng/μl DAPI (4',6-diamidino-2-2phenylindole), mounted in antifade medium Vectashield (Vector Laboratories, Burlingame, CA, USA) and examined under a fluorescence microscope.
Chromosomes were examined with an AxioImager M2 epifluorescent microscope, photographed with a high resolution black-white microscopy camera (Carl Zeiss, Vienna, Austria), and files processed using AxioVision 4.8 (Carl Zeiss, Vienna, Austria) with only those functions that apply to whole images equally. At least 15-20 well-spread metaphases and prometaphases were analysed for each individual. One to three individuals per population, in total 22 individuals of four species of Glechoma and two individuals of Meehania urticifolia, were analysed using FISH (Table 1).

Genome size estimation
Genome sizes of 48 plants of Glechoma and three plants of Meehania urticifolia were determined by flow cytometry with Solanum pseudocapsicum (1C = 1.29 pg/1C [6,40]) as internal standard. Freshly collected leaves from individual plants and the standard were chopped together in Otto's buffer I [41], filtered through a nylon mesh, and incubated with RNase at 37˚C in a water-bath for 30 min. Propidium iodide (PI) staining was performed in Otto's buffer II at 4˚C. Each individual was measured three times. Measurements were performed with a CyFlow flow cytometer (Partec, Germany) equipped with a green laser (532 nm, Cobolt, Sweden). The 1C values were calculated according to the assumed linear fluorescence intensity relationship of both object and standard nuclei [6,40].

DNA sequencing and phylogenetic analyses
Total genomic DNA was extracted from silica gel-dried leaf material of selected analysed individuals (Table 1), as described in Jang et al. [6]. The internal transcribed spacer (ITS) region of nuclear ribosomal DNA was amplified and sequenced using universal ITS primers [6; ITS_18s_F: The cycling conditions included initial denaturation at 96˚C for 1 min, followed by 35 cycles each of 1 min at 96˚C, 30 sec at 55˚C, 45 sec at 72˚C, and final elongation at 72˚C for 10 min. Amplified fragments were separated on a 1.5% agarose gel, and PCR products were purified from the gel using Invisorb 1 Spin DNA Extraction Kit (Invitek GmbH, Berlin, Germany). PCR amplification products of the 5S rDNA non-transcribed spacer DNA were cloned as described in Emadzade et al. [42] with three clones sequenced per individual and one individual per species (Table 1).
Sequences were assembled in SeqManII 5.05 (Lasergene, Madison, WI) and manually aligned using BioEdit 7.0.5.3 [43]. Model testing and inference of phylogenetic relationships were performed using maximum likelihood (ML) analysis as implemented in IQ-TREE 0.9.5 [44]. Gaps were treated as unknown characters. The best model was chosen according to the Bayesian Information Criterion and used for the final ML analysis. Nodal support was determined using non-parametric bootstrapping [45] with 1,000 replicates. All sequences were deposited in GenBank (Table 1). Trees obtained from ITS sequences were rooted using Meehania urticifolia as outgroup, those obtained from NTS sequences were rooted using paralogy

Chromosome numbers
All 48 plants of four Glechoma species from nine populations, at least two individuals per species, were tetraploids with 2n = 36. Three plants of Meehania urticifolia were diploid with 2n = 18 (Table 1; Fig 1).

Numbers and distribution patterns of ribosomal DNA loci
Twenty four plants, representing one to three individuals per population of each species of Glechoma and Meehania, were examined by FISH.
In M. urticifolia, one 5S rDNA locus and two 35S rDNA loci were found. The 5S rDNA locus was in the same chromosomal arm as, and proximal to, one of the 35S rDNA loci. The second 35S rDNA locus was in the subterminal region of the short arm of a different chromosome (Table 1; Fig 2A).
Glechoma consistently possessed four 5S rDNA signals. The number of 35S rDNA sites varied from six to eight, but in one individual of G. hirsuta three further very weak signals were seen on three different chromosomes (Table 1; Fig 2F). 5S rDNA sites were always located on short arms and in more proximal positions than the 35S rDNA loci (Fig 3). 35S rDNA occupied subterminal regions of the short arms (Fig 2) and occasionally could be seen as distinct secondary constrictions (Fig 2F and 2G).  2. three chromosomes with 35S and 5S rDNA subterminal on short arms, one chromosome with a short-arm 5S rDNA locus, and four chromosomes with subterminal short-arm 35S rDNA sites (Figs 2D, 2G and 3). This pattern was seen in a G. hederacea plant from Poland and two G. hirsuta plants from an Austrian population.
3. two chromosomes with terminal short-arm 35S and 5S rDNA, two chromosomes with a short-arm 5S rDNA site, and four chromosomes with subterminal short-arm 35S rDNA loci (Figs 2B and 3). Two plants of G. grandis from Tokyo showed this pattern. 4. four chromosomes with short-arm subterminal 35S and 5S rDNA signals, with three other chromosomes carrying short-arm 35S rDNA sites (Figs 2F and 3). One plant of G. hederacea and two of G. hirsuta, all from Perchtoldsdorf in Austria (Table 1), had this pattern, while a single G. hirsuta plant showing this pattern had additional, weak, short-arm pericentric 35S rDNA signals on three different chromosomes (Fig 2F).
5. four chromosomes with short-arm 35S and 5S rDNA signals, with two other chromosomes carrying short-arm subterminal 35S rDNA signals (Figs 2E and 3). This was observed in two G. hederacea plants from Vienna.

Genome sizes
The 1C DNA content of Glechoma differs between European G. hederacea and G. hirsuta (0.82 pg and 0.83 pg) and the Asian species G. grandis and G. longituba (0.92 pg and 0.91 pg). The genome size of Meehania urticifolia was 0.58 pg (Table 1).

Phylogenetic relationships in Glechoma
ITS sequences of 19 plants of one to three individuals per population of each Glechoma species were analysed. The length of the ITS region ranged from 597 to 607 bp. The total alignment was 608 bp long, with 357 variable characters, of which 153 were parsimony-informative. Glechoma grandis and G. longituba formed monophyletic well-supported clades (bootstrap support, BS 100%; Fig 4A). Within G. longituba, two subclades were identified-Chinese (BS 68%), and Korean (BS 59%). Glechoma longituba was recovered as sister group (BS 58%) to the clade comprising G. hederacea and G. hirsuta (BS 98%). Relationships between the latter two species, however, remain unresolved (Fig 4A). In addition, the non-transcribed spacer region (NTS) of 5S rDNA was cloned in a subset of individuals (one to three per species) for which ITS sequences were available. For each individual three clones were sequenced (Table 1). In most individuals two distinct NTS sequence types were identified, as expected after a duplication of 5S rDNA loci following polyploidization. Fragment length ranged from 453 to 513 bp with NTS alignment spanning 517 bp. 118 characters were variable, of which 52 were parsimony-informative. The maximum likelihood tree includes a split into two groups (BS 84%) corresponding to the two monomer types, each containing clones of all species (Fig 4B). Although the number of sequences per species differed between the two groups, inferred relationships among species within each group were congruent. Thus, G. grandis and G. longituba were sister taxa within a clade (BS 83-100%) that was itself a sister group to a clade comprising G. hederacea and G. hirsuta (BS 56-84%; Fig 4B).

Discussion
Interpreting variation in rDNA loci number and localisation in closely related species within a phylogenetic framework and in an evolutionary context has provided valuable information about relationships between diploids and polyploids when genetic knowledge is limited (e.g., Prospero [6]; Aeschynomene [7]; Hepatica [9]; Heliophila [34]; Nicotiana [46]). The Glechoma species investigated here show micro-and macromorphological similarity, a uniform basic chromosome number and the same level of polyploidy [15-16, 21, 23]. However, the species exhibit variation in the number and distribution of rDNA loci and in their genome sizes.

rDNA evolution and phylogenetic relationships in Glechoma
Most frequently in Glechoma there are eight 35S rDNA sites and four 5S rDNA sites, all in subterminal positions on short arms, as observed in G. grandis, G. hederacea and G. longituba (Fig  3). This is additive with respect to diploid Meehania urticifolia (Fig 3) which may represent the ancestral condition of Glechoma. A further four rDNA distribution patterns in Glechoma can be derived by loss of one of the 35S rDNA loci in Meehania, although never of both (Fig 3). A similar phenomenon has been reported in Paphiopedilum section Parvisepalum in Orchidaceae [12]. One individual of G. hirsuta had three additional, albeit very weak, signals of 35S rDNA in pericentric positions on chromosomes without other rDNA signals. Interestingly, all Glechoma species, except G. longituba (seven individuals) exhibit between-individual variation in their 35S rDNA patterns. By contrast, the number and the distribution of 5S rDNA signals is constant. Greater polymorphism of 35S rDNA in comparison to 5S rDNA is common in plants [9,[55][56][57][58][59]. The close proximity of these two genes within the same chromosome, or even chromosome arm, also characterises many other plant groups [60].
Phylogenetic relationships inferred from ITS and 5S rDNA NTS sequences (Fig 4) are congruent with previous molecular phylogenetic results reported for the genus [21,61] in that the distinctness of Asian G. grandis and G. longituba is well supported, while the European G. hirsuta and G. hederacea are intermixed (Fig 4) reflecting the high degree of morphological similarity and taxonomic uncertainty in this species pair. Glechoma is a relatively young genus (3-6 mya [61][62]) and contains possible hybrids and several ploidy levels [48]. Coincident low levels of rDNA locus variation, particularly of 35S rDNA, coupled with low amounts of morphological differentiation have been reported in other polyploid complexes [7,55].
Phylogenetic analyses of the 5S rDNA NTS regions suggest the presence of two types of monomers in every Glechoma species (Fig 4B), a pattern congruent with an allopolyploid origin. The alternative scenario of an autopolyploid origin is less likely, as it would require divergent evolution of homologous 5S rDNA loci, which has not been reported for plants so far. However, to establish what role hybridisation has played in the origin and evolution of the genus Glechoma needs other, more variable, sequences to be analysed, coupled with more extensive sampling. Polyploidisation has frequently been implicated in diversification and the colonisation of new habitats and areas [63][64]. Thus, colonisation and spread across Eurasia by the genus Glechoma might have been facilitated polyploidy [48].

Conclusions
This is the first molecular cytogenetic analysis of karyotype structure and genome size variation in the tetraploid genus Glechoma. An evolutionary interpretation in a phylogenetic context has been possible. 35S rDNA locus number is variable within and between Glechoma species, with several patterns shared by different species and with nearly every species exhibiting more than one pattern. The ancestral karyotype of Glechoma remains unidentified. However, plants carrying four 5S rDNA and eight 35S rDNA sites represent a doubling of the pattern seen in the closely related diploid Meehania urticifolia which may be a candidate for the structure of the ancestral karyotype. The variability of 35S rDNA sites suggests that the polyploid genomes in Glechoma are still in genomic flux [65]. Genome sizes and rDNA locus variation might provide characters to support species delimitation within Glechoma, since morphological diagnostic features are particularly weak [15][16][21][22][23].