Molecular Cytogenetic Characterization of C-Band-Positive Heterochromatin of the Greater Long-Tailed Hamster (Tscherskia triton, Cricetinae)

The greater long-tailed hamster (Tscherskia triton, Cricetinae) has a unique karyotype (2n = 28), containing 11 pairs of acrocentric chromosomes with large C-band-positive centromeric heterochromatin blocks. To understand the origin and evolutionary process of heterochromatin in this species, we isolated novel families of chromosome site-specific highly repetitive DNA sequences from TaqI-digested genomic DNA and then characterized them by chromosome in situ and filter hybridization. The TaqI-families of repetitive sequences were classified into 2 types by their genome organization and chromosomal distribution: the 110-bp repeated sequence organized in large tandem arrays (as satellite DNA), localized to centromeric C-positive heterochromatin of acrocentric autosomes (chromosomes 1–11) and submetacentric X chromosome, and the 405-bp repeated sequence that was composed of 30–32-bp internal repeats, distributed in the pericentromeric region on the short arms of X and Y chromosomes. The repetitive sequences did not cross-hybridize with genomic DNA of any genera of Cricetinae (Mesocricetus, Cricetulus, and Phodopus). These results suggest that the 110-bp and 405-bp repeats rapidly diverged in the lineage of T. triton, evolving in a concerted manner among autosomes and X chromosome and within X and Y chromosomes, respectively. The 110-bp centromeric repeat contained a 17-bp motif in which 9 bases are essential for binding with the centromere-associated protein CENP-B, suggesting the possibility that the 110-bp major satellite DNA carrying the 17-bp motif may have a role in the formation of specified structure and/or function of centromeres in T. triton.


Introduction
Highly repetitive DNA sequences, including tandemly repeated sequences known as satellite DNA, constitute a significant part of eukaryotic genomes [Singer, 1982;Kalitis and Choo, 1997]. Satellite DNA is a major compo-This article is licensed under the Creative Commons Attribution 4.0 International License (CC BY) (http://www.karger.com/Services/ OpenAccessLicense). Usage, derivative works and distribution are permitted provided that proper credit is given to the author and the original publisher. Kamimura nent of C-band-positive constitutive heterochromatin that is organized in large tandem arrays of noncoding short sequences. It is most abundant in the centromeric/ pericentromeric region of autosomes and sex chromosomes and can be located at subtelomeric chromosomal positions. Besides being the structural component of chromosomes, it is involved in various biological functions, such as centromere formation, chromosomal organization, sister chromatid pairing, chromosome association with the mitotic spindle during mitosis, chromosome segregation, and chromosome stability, among others [John and Miklos, 1979;Singer, 1982;Plohl et al., 2008Plohl et al., , 2012Plohl et al., , 2014Louzada et al., 2020;Thakur et al., 2021]. Satellite DNA is the evolutionarily most labile element in the genome because of the large differences in nucleotide sequences, chromosomal distribution, and copy number between species.
The greater long-tailed hamster (Tscherskia triton) inhabits northeastern Asia, such as eastern Siberia, northeastern China, and Korea. Molecular phylogenetic analysis using mitochondrial and nuclear genes demonstrated that T. triton is phylogenetically closer to Cricetulus, Cricetus, and Allocricetulus, compared with other genera of Cricetinae [Neumann et al., 2006;Ding et al., 2020]. However, the cytogenetic analyses including C-and G-banding of T. triton revealed that this species has a unique karyotype (2n = 28, FNa = 30), containing 11 pairs of large-and middle-sized acrocentric chromosomes with large C-positive heterochromatin blocks in the centromeric region, which is remarkably different from those of other hamster species [Tsuchiya and Won, 1976;Fujimoto et al., 1997]. Moreover, some individuals of this species have 1 or 2 B chromosomes [Wang et al., 1999;Borisov, 2012]. Such a separate position of T. triton within the Cricetus/Cricetulus group was also demonstrated based on chromosomal homologies of hamster species by Romanenko et al. [2007], who indicated that Cricetinae is classified into 3 sister groups: Cricetus/Cricetulus group including Allocricetulus and Tscherskia, Mesocricetus group, and Phodopus group. As the karyotype of T. triton differs essentially from those of other genera in the Cricetus/Cricetulus group, the authors suggested that chromosome rearrangements may have occurred more frequently in the lineage of T. triton from the ancestral karyotype of Cricetinae, compared with other hamster species. Thus, the molecular cytogenetic characterization of Cpositive heterochromatin in T. triton provides us with important information for understanding the evolutionary process of chromosome site-specific repetitive sequences and karyotype reorganization in this lineage. In this study, we isolated novel families of highly repetitive sequences constituting C-positive heterochromatin from T. triton, characterized them by chromosome in situ and filter hybridization, and examined their nucleotide sequence conservation in Cricetinae species.

Specimens, Cell Culture, Chromosome Preparation, and Karyotype Analysis
The lymphocytes isolated from the spleen of male individuals, provided from Tokyo University of Agriculture by Dr. Kimiyuki Tsuchiya, were cultured for chromosome preparations following the methods described previously [Matsuda et al., 1992;Matsuda and Chapman, 1995]. For replicating R-banding for fluorescence in situ hybridization (FISH) analysis, the cells were cultured with BrdU (25 μg/mL) for 4 h, including 30 min colcemid treatment, before harvesting (total culture time is 64 h). After staining the slides with Hoechst 33258 (1 μg/mL) for 10 min, they were heated at 75°C for 3 min and then exposed to a 20 W black light at a distance of 1.5 cm for additional 7 min. The slides were washed in distilled water, dried, and kept at −80°C until use.
For G-and C-banding analysis, chromosome preparations were made without BrdU treatment. Chromosome G-and Cbanding were performed with the ASG (Acetic/Saline/Giemsa) method [Sumner et al., 1971] and the BSG (Barium hydroxide/ Saline/Giemsa) method [Sumner, 1972], respectively. The metaphase spreads were captured on film, and chromosomes cut with scissors were arranged in pairs.
The nucleotide sequences of all the fragments were searched for homology with all DNA sequences in the NCBI non-redundant sequence database (http://blast.ncbi.nlm.nih.gov). Dot matrix analysis of the nucleotide sequences was performed with MAFFT version 7 (http://mafft.cbrc.jp/alignment/server/).

Fluorescence in situ Hydridization
FISH on replication-banded chromosome preparations was performed as described previously [Matsuda et al., 1992;Matsuda and Chapman, 1995]. The DNA fragments were labeled with biotin-16-dUTP using a nick translation kit (Roche Diagnostics, Basel, Switzerland) and ethanol-precipitated with sonicated salmon sperm DNA and E. coli tRNA. After hybridization, the slides were incubated with FITC-avidin (Roche Diagnostics) and stained with propidium iodide. The digital FISH images were captured using a Leica DMRA microscope (Leica Microsystems, Wetzlar, Germany).

Southern Blot and Slot Blot Hybridization
For Southern blot hybridization, 10 μg of genomic DNA of T. triton was digested with 8 restriction endonucleases (BglII, EcoRI, HaeIII, HinfI, HpaII, MspI, SmaI, and TaqI). Digested DNA was resolved by electrophoresis on 3% agarose gel, after which DNA fragments were transferred onto Hybond N+ nylon membranes (Cytiva, Tokyo, Japan). For slot blot hybridization, we used genomic DNA of male T. triton extracted in this study and that of male individuals of 7 Cricetinae species (Cricetidae), 2 Muridae species, and 1 Calomyscidae species obtained in our previous study [Yamada et al., 2006], namely M. auratus, M. brandti, C. griseus, Cricetulus migratorius, Phodopus sungorus, Phodopus campbelli, and Phodopus roborovskii of Cricetidae; mouse (Mus musculus) and rat (Rattus norvegicus) of Muridae; and Calomyscus mystax of Calomyscidae. After denaturing genomic DNA with 0.4 N NaOH for 10 min and then neutralizing with 1 M sodium acetate, 500 ng DNA was blotted onto Hybond N+ nylon membranes using BIO-DOT SF blotting equipment (Bio-Rad, Hercules, CA, USA). DNA probes were labeled with digoxigenin (Dig) using PCR Dig Labeling Mix (Roche Diagnostics), and hybridized to the membranes at 42°C in DIG Easy Hyb solution (Roche Diagnostics) following the manufacturer's instructions. After hybridization, the membranes were washed at 42°C in 2× SSC, 1× SSC, 0.5× SSC, and 0.1× SSC for 15 min each and exposed to Bio Max Ms Autoradiography Film (Kodak, Rochester, NY, USA).

G-and C-Banded Karyotypes of T. triton
G-and C-banded karyotypes of male T. triton are shown in online supplementary Figure 1 (see www. karger.com/doi/10.1159/000527478 for all online suppl. material). Each chromosome was numbered and arranged following the description by Fujimoto et al. [1997]. The diploid chromosome number was 2n = 28, consisting of 11 pairs of large-and middle-sized acrocentric chromosomes (chromosomes 1 to 11), with gradual size decrease among them, 2 pairs of small-sized metacentric chromosomes (chromosomes 12 and 13), and middlesized subtelocentric X and metacentric Y chromosomes. Large C-positive heterochromatin blocks were observed in the centromeric region of 11 pairs of acrocentric autosomes and the pericentromeric region of X and Y chromosomes; the whole Y long arm was heterochromatic. These results were consistent with those in previous studies [Tsuchiya and Won, 1976;Fujimoto et al., 1997;Wang et al., 1999;Borisov, 2012], whereas no B chromosomes were observed in this male individual.

Chromosomal Distribution of Repetitive Sequences
Chromosome site-specific fluorescence signals were observed for the following clones: Cen-TaqI-L from the 0.4-kb band, Cen-TaqI-M from the 0.25-kb band, Cen-TaqI-S1 from the 0.14-kb band, and 8 Cen-TaqI-S2 clones (Cen-TaqI-S2-1-S2-8) from the 0.11-kb band showed the hybridization signals in the centromeric region of 11 pairs of acrocentric chromosomes (chromosome 1 to 11) and subtelocentric X chromosome, but no signals were found on chromosomes 12 and 13 and the Y chromosome (Fig. 1). Other 2 clones isolated from the 0.4-kb band (XY-TaqI-1 and XY-TaqI-2) were sex chromosome-spe-cific, which showed hybridization signals in the pericentromeric region of the short arms of X and Y chromosomes (Fig. 2).

Nucleotide Sequences
The sizes, GC content, and chromosomal distribution of repeated sequence fragments isolated from T. triton, and nucleotide sequence identities between the fragments within the same sequence group are summarized in Table 1. Nucleotide sequences of 11 centromeric repeated sequence fragments and their sequence alignment are shown in Figure 3a. Eight Cen-TaqI-S2 fragments belonged to the same sequence family; the sizes were 110 or 111 bp, and their GC content was 40.7% on average. The similarity of nucleotide sequences between the fragments was 97.4% on average, ranging from 92.7% to 100%. Cen-TaqI-S1 had 140-bp sequence, which contained the abovementioned 110-bp sequence, and its GC content was 39.3%. The nucleotide length and GC content of Cen-TaqI-L and Cen-TaqI-M were 401 bp and 40.4% and 256 bp and 40.6%, respectively. Cen-TaqI-L and Cen-TaqI-M contained 3 and 2 copies of the 110-bp monomer unit, respectively. Cen-TaqI-M, Cen-TaqI-S1, and all Cen-TaqI-S2 fragments contained the 17-bp motif with all 9 bases which were previously shown to be essential for binding with centromere protein B (CENP-B) in M. musculus and Mus caroli of Murinae (Fig. 3) [Kipling et al., 1995].
Two 405-bp sex chromosome-specific XY-TaqI-1 and XY-TaqI-2 fragments had the same nucleotide sequence, with GC content of 40.3%. Dot-matrix analysis revealed that the fragments contained 12 internal repeat units of size 30-32 bp, whose identities ranged from 54.5 to 96.8% (73.9% on average) (Fig. 4).   a Number of clones with the same sequence group/isolated clones. Cen-TaqI-L and XY-TaqI fragments were isolated from the same prominent band. b Range of GC content is given in parentheses. c Range of nucleotide sequence identities between fragments is given in parentheses.

Fig. 3.
Nucleotide sequences of the centromere-specific Cen-TaqI sequence fragments. a Comparison of nucleotide sequences of 11 Cen-TaqI fragments of 4 different sizes. The Cen-TaqI-S2 fragments (Cen-TaqI-S2-1 to S2-8) were composed of 110-bp monomer units. The Cen-TaqI-L, Cen-TaqI-M, and Cen-TaqI-S1 fragments contained 3, 2, and one 110-bp monomer units, respectively. The consensus sequence of 110-bp monomer units is shown at the top of the fragments. Dots indicate the same bases as those of the consensus sequence. Hyphens indicate the gap of bases. Underlining indicates HinfI restriction sites (GANTC). The framed consensus sequences indicate a 17-bp motif that is homologous to mouse CENP-B box. b Alignment of the 17-bp CENP-B box motif, partial nucleotide sequences at position 5-21 of Cen-TaqI consensus sequence of the Cen-TaqI-L, Cen-TaqI-M, and Cen-TaqI-S2-1 to S2-8 fragments, position 35-51 of Cen-TaqI-S1, position 151-167 of the Cen-TaqI-L and Cen-TaqI-M fragments, and position 295-311 of the Cen-TaqI-L fragment. The framed 9 nucleotides are essential for binding with CENP-B as previously shown in humans and rodents [Kipling et al., 1995;Masumoto et al., 2004]. The Cen-TaqI-M, Cen-TaqI-S1, and Cen-TaqI-S2-1 to S2-8 fragments contained the 17-bp CENP-B motif. 328 DOI: 10.1159/000527478 No significant homology with any sequences was found for these repetitive sequences by homology search with NCBI database.

Organization of Repetitive Sequences in the Genome
To study the organization of repetitive sequences in the genome, Southern blot hybridization was performed using the Cen-TaqI-S2-6 and XY-TaqI-1 fragments as probes (Fig. 5). In the TaqI-digested genomic DNA probed with Cen-TaqI-S2-6 ( Fig. 5a), polymeric ladder signals of tandem repeats of 110-bp and 140-bp monomer unit bands, which were identical with the size of the Cen-TaqI-S2 and Cen-TaqI-S1 fragments, respectively, were observed. Polymeric ladder signals of the 140-bp monomer unit that contained the 110-bp Cen-TaqI-S2 monomer unit (Fig. 3a) were also observed in the HinfI-digest, likely derived from HinfI restriction sites included in the tandem array of the Cen-TaqI-S1 fragment. Signals of 110-bp and 140-bp monomer units were present in the greatest abundance, with progressively decreasing copy number of each higher order, indicating that TaqI restriction sites were highly conserved in these repeated sequences. In addition, faint bands of less than 110 bp were detected in the HinfI-digest, considered to be derived from 2 HinfI restriction sites included in these sequences. The EcoRI-and MspI-digests produced higher intensity of hybridization bands with increasing size of the multimers. HpaII and MspI are known to recognize identical sequences (CCGG); however, HpaII does not cleave when the cytosine is methylated at CG sites. The polymeric ladder bands were observed in the MspI-digest, whereas no ladder bands were found in the HpaII-digest. The Cen-TaqI-S family sequence was highly methylated in liver cells.
In hybridization probed with XY-TaqI-1 (Fig. 5b), weak hybridization bands of the 405-bp monomer unit were observed in the TaqI-digest, indicating that the TaqI site was not highly conserved in the tandem array of this repeated sequence. On the other hand, signals of sizes larger than 405 bp were observed in the BglII-digest. One BglII restriction site was present in the XY-TaqI-1 fragment (Fig. 4b), indicating the presence of larger-sized repeat units that contain the 405-bp monomer unit in the genome. At least 3 hybridization bands with sizes smaller than 405 bp were found in the HinfI-digest, probably derived from internal HinfI restriction sites. Methylation

Nucleotide Sequence Conservation of Repetitive Sequences between Different Species
Slot blot hybridization was performed to examine nucleotide sequence conservation of 2 repeated sequences (Cen-TaqI-S2 and XY-TaqI) in 7 species from 3 genera of Cricetinae (Mesocricetus, Cricetulus, and Phodopus), 1 Calomyscus species of Calomyscidae, and mouse and rat species of Muridae (Fig. 6). Both the sequences strongly cross-hybridized with genomic DNA of T. triton, but no hybridization signals were found in the other 10 species.

Discussion
Two novel families of highly repetitive sequences constituting heterochromatin were isolated from T. triton and were classified by nucleotide sequences and chromosomal distribution as 110-bp centromeric repeat and 405bp sex chromosome-specific pericentromeric repeat. Both the repetitive sequences were T. triton-specific and  did not hybridize to the genome of other Cricetinae, Muridae, and Calomyscidae species. In Microtus and Arvicola of Arvicolinae, nucleotide sequences of several satellite DNAs are conserved between many species whose karyotypes are different [Modi et al., 2003;Acosta et al., 2010;Rovatsos et al., 2021]. However, our results indicate that in T. triton the 110-bp centromere-specific and 405bp sex chromosome-specific pericentromeric repeats occurred and amplified after a greater degree of chromosome rearrangements in T. triton after this lineage separated from other genera of the Cricetus/Cricetulus group around 5 [Neumann et al., 2006] or 12 (http://www.timetree.org/) [Kumar et al., 2017] million years ago.
The 110-bp repeat occurred as a long stretch of tandem array, which was largely amplified in the centromeric region of 11 pairs of acrocentric autosomes and subtelocentric X chromosome. It then may have evolved in a concerted manner as a consequence of a 2-level process called molecular drive, involving sequence homogenization via intra-and interchromosomal crossover events followed by gene conversion and fixation as a result of a random assortment of genetic material in meiosis and chromosome segregation [Smith, 1976;Dover, 1982Dover, , 1986Liao, 1999;Plohl et al., 2012]. Such concerted homogenization of largely amplified centromeric repetitive sequences among acrocentric chromosomes has also been reported in mouse (M. musculus) [Matsuda and Chapman, 1991], small Japanese field mouse (Apodemus speciosus) [Fukushi et al., 2001], and a medaka species (Oryzias hubbsi) [Uno et al., 2013], and is a common characteristic in species whose karyotypes are mainly composed of acrocentric chromosomes with large centromeric heterochromatin. The 405-bp sex chromosome-specific repeat was independently amplified in the pericentromeric region of the sex chromosomes and only homogenized between the short arms of X and Y chromosomes.
Although both the repeats were AT-rich, no short ATrich stretch or AT motif (A-or T-boxes), which mediates the attachment of nuclear DNA to nuclear matrices or scaffolds as target sites of the high-mobility group nucleosomal proteins that are involved in centromere organization, such as stable curvature of centromeric DNA and chromatin folding, was found [Radic et al., 1992;Ouspenski and Brinkley, 1993;Slama-Schwok et al., 2000;Jagannathan et al., 2018]. However, the 110-bp centromeric repeat contained a motif that is essential for binding with the centromereassociated protein CENP-B [Masumoto et al., 1989]. CENP-B has a DNA-binding domain at its N-terminus for a 17-bp motif (YTTCGTTGGAARCGGGA) called the CENP-B box, in which the 9 underlined nucleotides are essential for CENP-B box function [Masumoto et al., 2004]. The CENP-B box was originally found in centromeric satellite DNA of humans and mice (M. musculus) and also in great apes and New World monkeys [Haaf et al., 1995;Thongchum et al., 2019] and wallaby [Bulazel et al., 2006]. Furthermore, it was also found in the centromeric satellite DNA of acrocentric chromosomes of Rodentia, such as M. caroli of Muridae [Kipling et al., 1995] and T. triton of Cricetidae [present study], indicating that this motif also existed in the common ancestor of muroid rodents. This result provides the possibility that the 110-bp centromeric satellite DNA may have played a role in the formation of specified structure and/or function of centromeres by providing an array of binding sites for CENP-B in T. triton. However, the CENP-B protein and its associated CENP-B box DNA motif exist at both active and inactive centromeres on mitotically stable dicentric human chromosomes [Earnshaw et al., 1989;Sullivan and Schwartz, 1995], but they are not detectable at centromeres of the Y chromosomes of humans and mouse [Earnshaw et al., 1987;Masumoto et al., 1989] nor on neocentromeres [Choo, 1997;Depinet et al., 1997;du Sart et al., 1997], and CENP-B null mice are viable [Hudson et al., 1998]. These results indicate that CENP-B may not be essential for kinetochore formation and that CENP-B may be involved in some other aspect of centromere function, such as chromosome movement or DNA packaging. Therefore, the absence of the 110-bp satellite DNA in chromosomes 12 and 13 and the Y chromosome in T. triton does not rule out the possibility that other sequences different from the 110bp centromeric satellite DNA mediate centromere function in the chromosomes. Further molecular-level analyses including immunostaining of this hamster's cells and ChIP-qPCR analyses using an antibody against CENP-B need to determine if the 110-bp centromeric satellite DNA functions as a CENP-B binding sequence.
Next-generation sequencing (NGS) technologies allied to newly developed bioinformatics tools capable of searching repetitive DNA sequences in unassembled data allow for the de novo identification of novel satellite DNA sequences in not only model organisms but also a wide range of wild species lacking reference genomes [Melters et al., 2013;Ruiz-Ruano et al., 2016;Garrido-Ramos, 2017;Lower et al., 2018]. In hamster species, the extent of satellite DNAs has not been previously uncovered because they have been cloned from only a part of species. However, recently, Ivanova et al. [2022] comprehensively uncovered novel satellite DNA sequence families from C. griseus, which were not previously identified by usual molecular biology methods, by exploring NGS data using the RepeatExplorer pipeline and inferred the karyotype evo-Cytogenet Genome Res 2022;162:323-333 DOI: 10.1159/000527478 lution in related species from chromosomal distribution of the sequences. The combination of NGS and highthroughput in silico analysis on the repetitive content of genomes using bioinformatic pipelines contributes to improvement of our knowledge about the relationship between the evolution of repetitive sequences and karyotype evolution in hamster species.