Comparative study of four Mystus species (Bagridae, Siluriformes) from Thailand: insights into their karyotypic diversity

Abstract Karyotypes of four catfishes of the genus Mystus Scopoli, 1777 (family Bagridae), M. atrifasciatus Fowler, 1937, M. mysticetus Roberts, 1992, M. singaringan (Bleeker, 1846) and M. wolffii (Bleeker, 1851), were analysed by conventional and Ag-NOR banding as well as fluorescence in situ hybridization (FISH) techniques. Microsatellite d(GC)15, d(CAA)10, d(CAT)10 and d(GAA)10 repeat probes were applied in FISH. The obtained data revealed that the four studied species have different chromosome complements. The diploid chromosome numbers (2n) and the fundamental numbers (NF) range between 52 and 102, 54 and 104, 56 and 98, or 58 and 108 in M. mysticetus, M. atrifasciatus, M. singaringan or M. wolffii, respectively. Karyotype formulae of M. mysticetus, M. atrifasciatus, M. singaringan and M. wolffii are 24m+26sm+4a, 26m+24sm+2a, 24m+18sm+14a and 30m+22sm+6a, respectively. A single pair of NORs was identified adjacent to the telomeres of the short arm of chromosome pairs 3 (metacentric) in M. atrifasciatus, 20 (submetacentric) in M. mysticetus, 15 (submetacentric) in M. singaringan, and 5 (metacentric) in M. wolffii. The d(GC)15, d(CAA)10, d(CAT)10 and d(GAA)10 repeats were abundantly distributed in species-specific patterns. Overall, we present a comparison of cytogenetic and molecular cytogenetic patterns of four species from genus Mystus providing insights into their karyotype diversity in the genus.


Introduction
Bagridae are the largest family of Thai catfishes, with six genera (Bagrichthys Bleeker, 1857, Batasio Blyth, 1860, Hemibagrus Bleeker, 1862, Mystus Scopoli, 1777, Pseudomystus Jayaram, 1968, and Sperata Holly, 1939 and 28 species in Thailand. They play an important role in the national economic value of the country, as they are kept in aquaria and contribute heavily to the aquaculture industry. Most species of the genus Mystus are booming in aquaculture, with some of them being kept in aquaria (Vidthayanon 2005). However, several species in this family are rather morphologically similar especially during the juvenile stage that may pose difficulties for their identification. Mystus is a poorly diagnosed group, and they are morphologically similar and diagnostic characteristics are usually subtle (Ng 2003;Ferdous 2013).
Conventional cytogenetics may be sufficient to identify intra-and interspecific variations and is an inexpensive approach. However, it has restrictions, and accordingly the use of molecular cytogenetic analyses plays an increasing role for more precise characterization of the structure of genomes, including that of fishes. Especially, fluorescence in situ hybridization (FISH) for mapping of repetitive DNA sequences provided important contributions to the characterization of biodiversity and evolution in divergent fish groups (Cioffi and Bertollo 2012), especially as some microsatellite repeats are species-specific (Cioffi et al. 2015). To date, there are only three studies within Bagridae using such FISH techniques, all performed by our group (Supiwong et al. 2013a(Supiwong et al. , 2014a. In the present study, chromosomal structures and genetic markers for Thai populations of M. atrifasciatus Fowler, 1937, M. mysticetus, M. singaringan (Bleeker, 1846 and M. wolffii (Bleeker, 1851) ( Fig. 1A-D) were for the first time analysed by cytogenetics and molecular cytogenetics.

Material and methods
Ten males and ten females of each species were collected from the Chi (Maha Sarakham Province), Songkhram (Bueng Kan Province), Chao Phraya (Sing Buri Province) and Pak Phanang Basins (Nakhon Sri Thammarat Province), Thailand from 2016-2018. The procedures followed ethical protocols as approved by the Institutional Animal Care and Use Committee of Khon Kaen University, based on the Ethics of Animal Experimentation of the National Research Council of Thailand ACUC-KKU-15/2559. Preparation of fish chromosomes from kidney cells was done as previously reported (Supiwong et al. 2012;Pinthong et al. 2015). The chromosomes were stained with Giemsa solution for 10 minutes. Ag-NOR banding was performed by applying two drops of 2% gelatin to the chromosomes, followed by four drops of 50% silver nitrate (Howell and Black 1980). Metaphases were evaluated according to the chromosome classification of Levan et al. (1964). Chromosomes were classified as metacentric (m), submetacentric (sm), subtelocentric (st) or acrocentric (a). Fundamental number, NF (number of chromosome arm) was obtained by assigning a value of two to metacentric and submetacentric chromosomes and one to subtelocentric and acrocentric chromosomes. The chromosome sizes were calculated applying the method of Tanomtong (2011).
Microsatellites d(GC) 15 , d(CAA) 10 , d(CAT) 10 and d(GAA) 10 repeat probes (Kubat et al. 2008) were directly labeled by Cy3 at 5´ ends during synthesis (Sigma, St. Louis, MO, USA). FISH under high stringency conditions on mitotic chromosome spreads (Pinkel et al. 1986) was performed as previously reported (Supiwong et al. 2017b;Yano et al. 2017). The evaluation was done on an epifluorescence microscope Olympus BX50 (Olympus Corporation, Ishikawa, Japan).  Table 1). Differentiated sex chromosomes between male and female specimens could not be identified in all analyzed species.   and sizes among the four analyzed species is shown in Fig. 2 and Table 2

Diploid chromosome numbers, fundamental numbers and karyotypes of M. atrifasciatus, M. mysticetus, M. singaringan and M. wolffii
The diploid chromosome numbers (2n) in all analyzed species confirmed previous cytogenetic studies (Donsakul 2000(Donsakul , 2001Magtoon and Donsakul 2009;Supiwong et al. 2014a, b), except for M. mysticetus with 2n=50 reported in a previous study (Donsakul 2002) and 52 in the present one. In agreement with the literature, 2n in the genus Mystus ranges between 50 and 58 chromosomes (Arai 2011; Table 1). The possible mechanisms that promoted intra-and interspecific karyotype diversification are biogeographic barriers, small population, limited gene flow (Galetti Jr et al. 2000). Although all studied species except M. mysticetus, had the same 2n as previous studies, the karyotypes were different, probably because of different sampling sites should be considered (Fig. 4). The predominant 2n in this genus is 56 chromosomes (five from 13 species) and may represent an ancestral character in this family (Sharma and Tripathi 1986). This is consistent with the hypothesis of Oliveira and Gosztonyi (2000) that 2n=56 could be a plesiomorphic character in the order Siluriformes. However, NF and karyotypes found in the present study differ from all previous reports (Donsakul 2000(Donsakul , 2001(Donsakul , 2002Magtoon and Donsakul 2009;Supiwong et al. 2014a, b). These differences may be species-specific variations within populations, and/or misidentification of species, or different species in presumed species complexes. NF in Mystus vary from 64 to 110. Ghigliotti et al. (2007) suggested that species with a higher NF value are more advanced in evolutionary terms than such with lower one. That hypothesis can be described that primitive karyotype of fish possesses many acrocentric chromosomes (mono-arm chromosomes). During evolution, the mono-arm chromosomes changed to bi-arm chromosomes. The NF would be unaltered, but the 2n would decrease. Changes in NF appear to be related to the occurrence of pericentric inversions, which play a major role for karyotypic rearrangement in fishes and other vertebrates (King 1993;Galetti Jr et al. 2000;Wang et al. 2010). Accordingly, from comparative analysis among the here studied four Mystus species, NF data and analyses of karyotypic complements indicate for that M. singaringan has the most primitive karyotype while M. wolffii has the most derivative karyotype. As often seen in fishes of this family, no heteromorphic sex chromosomes for males and females could be identified. Nonetheless it must be mentioned, that there are two species, M. gulio (Hamilton, 1822) and M. tengara (Hamilton, 1822), which have differentiated sex chromosome systems as XX/XY and ZZ/ZW, respectively (Arai 2011). Accordingly, differentiated sex chromosome system in this fish group seems to be a quite rare phenomenon. Karyotypes of the genus Mystus in Thailand showed high diversification (Table 1). Seven species have been cytogenetically studied. The 2n ranged between 50 chromosomes in M. mysticetus (Donsakul 2002) and 58 chromosomes in M. wolffii (Donsakul 2000;present study). The predominant 2n is 56 chromosomes found in M. albolineatus (NF = 108, 28m+6sm+12st+10a) (Donsakul 2000), M. bocourti (NF = 100, 22m+22sm+12st/a; NF = 104, 24m+18sm+6st+8a) (Donsakul 2000;Supiwong et al. 2013aSupiwong et al. , 2014a and M. singaringan (NF = 94, 24m+14sm+10st+8a; NF = 98, 24m+18sm+14a) (Donsakul 2001;present study). Our results showed differences among NFs and karyotypes in the studied species. Interestingly, M. mysticetus had two variants, 2n = 50 chromosomes (NF=92, 28m+14sm+8a), found in Ayutthaya Province, Central Thailand (Donsakul 2002), and 52 chromosomes (NF = 100, 26m+22sm+4st/a; NF=102, 26m+24sm+2a) found in Maha Sarakham and Bueng Kan Provinces, Northeast of Thailand (Supiwong et al. 2014a, b; present study) (Fig. 4). This variation may be caused by a rearrangement of chromosomes by centric fusion and pericentric inversion during chromosomal evolution in groups of populations separated by a geographic barrier.

Nucleolus organizer regions (NORs)
The localization of nucleolus organizer regions (NORs) is a simple method to determine chromosomal marker. NORs are specific positions on the chromosome that consist of tandemly repeated sequences of ribosomal genes (rRNA). In eukaryotes, each unit is composed of three genes coding for 18S, 5.8S and 28S ribosomal RNA (Sharma et al. 2002). Generally, most fishes have one pair of small NORs (single NOR) on chromosomes. However, some species of fishes have more than two NORs which may be caused by the translocation between some part of the chromosome with NORs and another chromosome (Sharma et al. 2002). Interspecific and intraspecific NOR polymorphism in the number of NORs per genome, in the chromosomal location of NOR sites, in the relative sizes of individual NORs, and in the number of active NOR sites per cell are commonly observed in fish, where the rDNA loci have been shown to be highly dynamic (Milhomem et al. 2013). Changes in chromosome number and structure can alter the number and structure of NOR as well. The pattern of NORs may be specific to populations, species and subspecies. Robertsonian translocations may cause losses of NOR. Species, which have limited gene exchange due to geographical isolation, have elevated karyotype numbers and NOR variation. (Yüksel and Gaffaroğlu 2008). The NOR is frequently used to compare variations as well as to identify and explain specifications. Therefore, it can be used as taxonomic and systematic characters in order to infer phylogenetic hypotheses of species relationships (Gold 1984;Amemiya and Gold 1990).

Patterns of microsatellite repeats on the genomes of Mystus atrifasciatus, M. mysticetus, M. singaringan and M. wolffii
Repetitive DNAs like microsatellites can be used to spot genomic evolution as previously been reported for different fish groups (Cioffi et al. 2010;Cioffi and Bertollo 2012;Terencio et al. 2013;Yano et al. 2014;Cioffi et al. 2015;Moraes et al. 2017Moraes et al. , 2019Sassi et al. 2019). It is known from fossil records that there is a major evolutionary diversification in Siluriformes fishes; this has in parts already also been verified at chromosomal level.
Here, four bi-and tri-nucleotide microsatellite sequences were mapped on chromosomes of four Mystus species. The patterns of microsatellites d(GC) 15 and d(CAA) 10 repeats in three species in the present study (M. atrifasciatus, M. mysticetus, M. singaringan) are similar to those found in Channa micropeltes (Cuvier, 1831) (Cioffi et al. 2015). On the other hand, they are differences known for C. gachua (Hamilton, 1822), C. lucius (Cuvier, 1831), C. striata (Bloch, 1793) (Cioffi et al. 2015), Toxotes chatareus (Hamilton, 1822) (Supiwong et al. 2017b) and Asian swamp eel, Monopterus albus (Zuiew, 1793) (Supiwong et al. 2019). The pattern of microsatellite d(GC) 15 repeats in M. wolffii is similar to that of C. lucius (Cioffi et al. 2015) and T. chatareus (Supiwong et al. 2017a). Interestingly, the patterns of microsatellite d(CAT) 10 repeats in M. atrifasciatus, M. mysticetus and M. singaringan are similar to the patterns of the (CA) 15 repeats on chromosomes of other species in the family Bagridae (Supiwong et al. 2013a(Supiwong et al. , 2014b. Comparative study on four species showed that not only there are differences of 2n, NF and karyotype, but the patterns of microsatellite repeat on chromosomes also have difference among them. Thus, the cytogenetic data may be a tool for classification of fish species that there is similar morphology as the stripe Mystus (M. atrifasciatus and M. mysticetus).
From previous reports, it may be carefully deduced that most heterochromatin in fish genomes consist of microsatellites (Cioffi and Bertollo 2012). However, microsatellites have also been found in non-centromeric regions, many of them were located either near or within genes (Rao et al. 2010;Getlekha et al. 2016). Indeed, GC rich motifs are common in exons of all vertebrates (Chistiakov et al. 2006). Since higher re-combination rates can be found near the telomeric region (Jensen-Seaman et al. 2004), it is possible that the physical proximity of microsatellite and rDNA repeats could favor the evolutionary spreading of both sequences together, despite the possibility of spreading some errors, too. Repetitive DNA sequences could act as primary driving forces in speciation (Biémont and Vieira 2006). These sequences are closely associated with heterochromatic regions, thus contributing to gene activation and structural maintenance of chromosomes (Dernburg et al. 1996). Therefore, great variations in the amount and position of these sequences could create fertility barriers by fostering the occurrence of chromosomal rearrangements (Cioffi and Bertollo 2012).
Indeed, the distribution of microsatellite motifs in fish genomes could be biased to some specific noncoding regions, as found in the Asian swamp eel, M. albus (Li et al. 2017). Finally, closely related fish species involved in recent speciation events could present a differential pattern in the distribution and quantity of microsatellite sequences on chromosomes, as demonstrated for naked catfishes (Supiwong et al. 2014b), channid fishes (Cioffi et al. 2015) and four Mystus in the present study.

Conclusions
The present research is the first report on NOR and microsatellites d(GC) 15 , d(CAA) 10 , d(CAT) 10 and d(GAA) 10 mapping in M. atrifasciatus, M. mysticetus, M. singaringan and M. wolffii. There are differences in the diploid chromosome number, the fundamental numbers, karyotypes, pairs having NORs, and patterns of microsatellite distributions on chromosomes. These results indicated that (molecular) cytogenetic data can be used for classification in related fish species and to explain karyotype diversification.