Chromosome Translocations as a Driver of Diversification in Mole Voles Ellobius (Rodentia, Mammalia)

The involvement of chromosome changes in the initial steps of speciation is controversial. Here we examine diversification trends within the mole voles Ellobius, a group of subterranean rodents. The first description of their chromosome variability was published almost 40 years ago. Studying the G-band structure of chromosomes in numerous individuals revealed subsequent homologous, step-by-step, Robertsonian translocations, which changed diploid numbers from 54 to 30. Here we used a molecular cytogenetic strategy which demonstrates that chromosomal translocations are not always homologous; consequently, karyotypes with the same diploid number can carry different combinations of metacentrics. We further showed that at least three chromosomal forms with 2n = 34 and distinct metacentrics inhabit the Pamir-Alay mountains. Each of these forms independently hybridized with E. tancrei, 2n = 54, forming separate hybrid zones. The chromosomal variations correlate slightly with geographic barriers. Additionally, we confirmed that the emergence of partial or monobrachial homology appeared to be a strong barrier for hybridization in nature, in contradistinction to experiments which we reported earlier. We discuss the possibility of whole arm reciprocal translocations for mole voles. Our findings suggest that chromosomal translocations lead to diversification and speciation.


Introduction
The role of chromosome changes in speciation has been discussed for many years [1][2][3][4][5][6]. Chromosome rearrangements disrupt the spatial positioning of chromosomes in the nuclei, and Robertsonian translocations can change intranuclear organization dramatically and alter interactions among chromosomal domains [7][8][9]. In morphologically indistinguishable species, the so-called cryptic species, morphological similarity exists alongside significant differences in genetic markers, such as chromosome structures or specific genes. There seem to be no radical changes in the development of morphological traits, but reproductive isolation appears, which makes such species a good prospective model for studying evolution [10][11][12][13].
Rodents of genus Ellobius, or mole voles, are very similar morphologically, due to their adaptation to a specific subterranean style of life. Also, the genus is an excellent example of diversification of sex chromosomes: XX♀-XY♂in E. fuscocapillus, X0♀-X0♂in E. lutescens, and XX♀-XX♂in E. talpinus, E. tancrei, and E. alaicus [14][15][16][17]. Even before studying chromosomes, some paleontologists considered E. tancrei and E. talpinus to be distinct species, whose ancestral lines diverged in the Pleistocene [18]. Then, although the species status was confirmed by cytogenetic data [19], until the 1980s E. tancrei was referred to as conspecific with E. talpinus [20,21]. E. talpinus and E. tancrei share the same diploid number, 2n = 54, but differ by the fundamental number, NF = 54 and 56, respectively. The centromere reposition which changed the fundamental number probably led to their reproductive isolation [22]. Till now, no chromosomal variability has been shown for E. talpinus (2n = NF = 54) [23]. However, E. tancrei has one of the most variable karyotypes among mammalian species with diploid numbers varying from 54 to 30 [16,20,[24][25][26]. Using data from comparative chromosomal painting, the ancestral acrocentrical karyotype of Ellobius was reconstructed [27]. White [2] hypothesized that the evolution of karyotypes in mammals was directed towards a reduction of chromosome number, which Robertsonian translocations can affect.
Three hypotheses were outlined for an explanation of the wide chromosome variability in E. tancrei [20]: (i) a low-chromosomal form originated from the ancestral 2n = 54 karyotype by chain mutation processes and intermediate forms kept in distinct populations; (ii) a single sudden emergence of a low-chromosomal form was followed by its hybridization with the original form and subsequent stabilization of chromosomally balanced forms; (iii) multiple and independent translocations of acrocentrics in different populations occurred, so metacentrics can be homologous or non-homologous, unlike in first two cases where the Robertsonian metacentrics must be homologous.
This study aims to describe the karyotype structure of all currently known chromosomal forms of E. tancrei from the Pamir-Alay, the zone of greatest chromosomal variation, using chromosome painting and G-banding comparison to determine their paths of evolution.

Results
The karyotype structure was determined for 379 mole voles from 80 localities across the Pamir-Alay mountains, mostly the Surkhob and Vakhsh rivers valleys (Figures 1 and 2). Among the samples, we found 29 animals from 8 localities with the typical for E. tancrei 2n = 54. We compared chromosome painting data for nine specimens (collection numbers 24914, 24915, 24919, 24920, 25613, 25614, 25617, 25620, 25621, see Figure 3 for selected examples) with previously published karyotypes [26,28,29]. Samples demonstrated distinct Robertsonian translocations (Rbs), which could be distinguished by G-banding or chromosome painting. We revealed the existence of three allopatric forms with basic 2n = 34 and specific Rbs. Moreover, scanning G-banded karyotypes of mole voles from the Surkhob and Vakhsh rivers valleys, we revealed numerous hybrids of each of these forms with the original 54-chromosomal E. tancrei.

Chromosome Form I
Chromosome form I is located mostly at the northern bank of the Surkhob River from the left bank of the Sorbog River (approximately 70 • 11 E longitude) at the West to 71 • 21 at the East. Unexpectedly, this form crossed the rough wide mountain river and settled at the southern bank also (Figure 1). We collected data on karyotype structure for 160 animals from 31 localities (Figure 2a). Diploid numbers varied from 2n = 53 to 2n = 30 due to the presence of 1 to 24 Rbs in different combinations alongside with stable NF = 56: Rb(1.8), Rb(2.11), Rb(3.18), Rb(4.14), Rb(5.9), Rb (10.19), Rb(6.12), Rb(13.23), Rb (15.22), Rb (16.20), Rb (17.25), and Rb (24.26). All Rb translocations detected in this form by chromosome painting were homologous to Rbs in mole voles with 2n = 30 [26].

Chromosome Form II
Chromosome form II with 2n = 34, NF = 56 and specific Rbs were located at the confluence of the Surkhob and the Obikhingou rivers, up to approximately 70 • 30 E longitude, and in the adjacent mountains to the south, close to the Khozar-Chasma settlement (Figure 1). A total of 96 specimens from 20 localities were karyotyped (Figure 2b). Compared to chromosome form I, only two specific Rbs were revealed by chromosomal painting: Rb(2.18) and Rb(3.11) (Figure 6a). These translocations appeared to be whole arm reciprocal translocations (WARTs) with Rb(2.11) and Rb(3.18), as found in chromosome form I. Other Rbs were identical to metacentrics of the chromosome form I.

Chromosome Form III
Chromosome form III with 2n = 32-34 was discovered on the right bank of the Vakhsh River, at the area from the right bank of the Sangikar River up 38 • 47 N 69 • 51 E, and at the opposite bank of the Vakhsh River, close to the Mienodara settlement ( Figure 1). In total, 94 animals from 21 localities were studied (Figure 2c). In karyotypes the animals studied, one Rb was the same as in chromosome form II-Rb(2.18), two others were characteristic for this form only (Figure 7a,b). Chromosome 3 underwent translocation to chromosome 24, forming Rb(3.24), and chromosome 11 translocated to 25, creating Rb(11.25). Other Rbs were shared with I and II chromosome forms. Unlike other forms, chromosome form III formed a hybrid zone, bordered by rivers. In the confluence of the Sangikar and Sorbog Rivers, 17 specimens from two localities with 2n = 35-41 were studied, with a different combination of all Rbs. High heterozygosity supported the presumption of hybridization between the low-chromosomal form III, 2n = 34 with the original form, 2n = 54 (Figure 7c). Animals were numerous in 1982, but after the earthquakes and powerful mudslides in 2002 along with anthropogenic pressure, the population decreased dramatically.
The right bank of the Vakhsh was inhabited by mole voles with 2n = 32 and some 2n = 33 and 34, which were found from the confluence of the Surkhob and Obikhingou Rivers and over 25 km to the south. On the opposite bank of the river, we discovered hybrids (26 specimens from five sites) with 2n = 33-35, which were located close to man-made bridges.
The three chromosome forms shared partial homology in Rbs: 2.11-2.18-3.18-11.25-17.25. Such monobrachial homology could arise by the independent translocations in different forms or because of specific mutations (whole arm reciprocal translocations, WARTs) lead to reciprocal translocations occurring between two bi-armed non-homologous chromosomes. After the term WART was introduced by Winking [31], this type of mutation was rarely reported as the mechanism is elusive. A single case was described for the spontaneous occurrence of the WART in a wild-derived house mouse [32].
Possible WARTs have been reported in Sorex araneus (for review see [33]), different species of Mus [34][35][36][37] and some other species. The problem of explaining the occurrence of these translocations has not yet been solved; nevertheless, the presence of monobrachial homology in some cases is easier to explain by WART. In Mus and Sorex rapid chromosomal evolution is evident due to the existence of numerous chromosomal races. We presumed that the previously described E. tancrei with 2n = 48 and two variants of 2n = 50 with monobrachial homology demonstrate an initial stage of speciation due to the meiotic disturbances in hybrids [22,29,38]. In forms with 2n = 34 with numerous Rbs and several non-homological translocations, the meiotic failure is even more predictable. The absence of hybrids even in contact areas indicates the emergence of reproductive barriers for the three chromosomal forms in Pamir-Alay.
We suggest that speciation is currently underway in mole voles. A specific Rb (3.10) was fixed in E. alaicus in 30 years [30]. In E. tancrei some small chromosomes are not involved in Rb translocations: chromosome 17 in forms II and III and chromosomes 21 and 26 in form I. However, these small chromosomes have a potential for further translocations, as demonstrated for autosomes 24 and 26 in karyotype with 2n = 30 (chromosome form I) [26].
Chromosomal variability of E. tancrei is confined to the Pamir-Alay, unlike Sorex and, especially, Mus, different forms of which are distributed worldwide. The reason given for this by Vorontsov and Lyapunova [39] was explosive chromosomal speciation in seismic active regions. High seismic activity in the Pamir-Alay [40] could lead to sudden rearrangements of the mountain topography; this might isolate small populations with specific mutations and promote diversification [41]. Variability of diploid numbers might have emerged by hybridization inside this restricted mountain area. Distinct hybrid zones, which we described here, may have originated due to contacts of forms with partially homologous metacentrics with the original E. tancrei, 2n = 54. Migration for subterranean mole voles is possible in river valleys; the map demonstrates possible geographical connections for forms with 2n = 54 and distinct 2n = 34 ( Figure 1). Fast evolutionary changes might be further accelerated or could disappear, because of seismic activity and anthropogenic pressure. In 2008 we were unable to find animals in the confluence of the Sorbog and Sangikar rivers. In 2018, a valley, where the 30-chromosomal form was detected, was built up and plowed and the number of animals decreased dramatically.
Chromosome painting allowed us to reveal the variability and verify the hypothesis for the independent and repeated chromosome translocations for mole voles in the Pamir-Alay. The origin of monobrachial homology is a route to speciation [42]. The emergence of such a homology leads to fertility disorders or sterility of hybrids, which provides isolation of gene pools. Such cases are rare, and E. tancrei appears to be one of the most valuable models for studying chromosomal speciation.

Materials and Methods
Chromosome sets for 379 specimens of E. tancrei mole voles from 80 localities from the Pamir-Alay mountains (Figures 1 and 2) were analyzed. Chromosome slides and suspension, prepared from bone marrow (Ford and Hamerton, 1956), were collected in 1981,1982,1985,2008, and 2010 for the cytogenetic collection (a part of the joint collection of wildlife tissues for fundamental, applied, and environmental researches of the Koltzov Institute of Developmental Biology RAS, state registration number AAAA-A16-116120810085-1, Core Centrum of the Koltzov Institute of Developmental Biology RAS, state registration number 6868145). For cross-species chromosome painting (9 specimens from 5 localities) we used cell lines from the IMCB SB RAS cell bank ("The general collection of cell cultures" №0310-2016-0002) established following previously published protocols [43,44] and sets of paints derived from the flow-sorted chromosomes of the field vole Microtus agrestis [45]. Cross-species chromosome painting was performed according to protocols [46,47]. G-banding was carried out by trypsin digestion [48]. Karyotypes were checked following the nomenclature of E. tancrei chromosomes [22,26].
Images were captured with VideoTesT-FISH 2.0, VideoTesT-Karyo 3.1. (VideoTesT, St. Petersburg, Russia) and Case Data Manager 6.0 (Applied Spectral Imaging Inc., ASI, Carlsbad, CA, USA) software with ProgRes CCD (Jenoptik, Jena, Germany) or ASI CCD camera on an Axioskop 2 plus (Zeiss, Oberkochen, Germany) microscope with filters for DAPI, FITC, and rhodamine. Hybridization signals were allocated to specific chromosome regions defined by G-banding patterns captured with the CCD camera. Giemsa-stained and G-banded metaphase chromosomes were studied with an Axioskop 40 (Zeiss) microscope and captured with a CMOS camera. Images were processed using Paint Shop Photo Pro X3 (Corel, Ottawa, ON, Canada).
All applicable international, national, and/or institutional guidelines for the care and use of animals were followed. All experiments were approved by the Ethics Committee on Animal and Human Research of the Institute of Molecular and Cellular Biology, Siberian Branch of the Russian Academy of Sciences, Russia (order No. 32 of 5 May 2017). This article does not contain any studies with human participants performed by any of the authors. Acknowledgments: Core Centrum of the Koltzov Institute of Developmental Biology RAS. The "Molecular and Cellular Biology" core facility of the IMCB SB RAS (state registration number 507496).

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.