Genomic Organization of Repetitive DNA Elements and Extensive Karyotype Diversity of Silurid Catfishes (Teleostei: Siluriformes): A Comparative Cytogenetic Approach

The catfish family Siluridae contains 107 described species distributed in Asia, but with some distributed in Europe. In this study, karyotypes and other chromosomal characteristics of 15 species from eight genera were examined using conventional and molecular cytogenetic protocols. Our results showed the diploid number (2n) to be highly divergent among species, ranging from 2n = 40 to 92, with the modal frequency comprising 56 to 64 chromosomes. Accordingly, the ratio of uni- and bi-armed chromosomes is also highly variable, thus suggesting extensive chromosomal rearrangements. Only one chromosome pair bearing major rDNA sites occurs in most species, except for Wallago micropogon, Ompok siluroides, and Kryptoterus giminus with two; and Silurichthys phaiosoma with five such pairs. In contrast, chromosomes bearing 5S rDNA sites range from one to as high as nine pairs among the species. Comparative genomic hybridization (CGH) experiments evidenced large genomic divergence, even between congeneric species. As a whole, we conclude that karyotype features and chromosomal diversity of the silurid catfishes are unusually extensive, but parallel some other catfish lineages and primary freshwater fish groups, thus making silurids an important model for investigating the evolutionary dynamics of fish chromosomes.

Siluridae catfishes represent one of the most interesting fish groups from a 2n evolutionary point of view, owing to their wide distribution, unique ecological niche, and known evolutionary trajectory [7]. Siluridae includes one of the largest freshwater fish species-Silurus glanis-which commonly reaches 2 m in size and over 300 kg in weight [14], and is highly valued in the food market [15]. Several other species, such as Micronema cheveyi, Phalaconotus apogon, P. bleekeri, Wallago attu, and W. micropogon, also comprise important food sources [16] or are ornamental fishes, like the glass catfish Kryptopterus bicirrhis [17].
Cytogenetic studies have proven useful to discover and explore cryptic biodiversity in a number of fish groups [18]. Particularly, in complex and ecologically diverse fish groups, cytogenetic studies have made important contributions to elucidate the evolutionary pathways of distinct fish groups, owing to their particular chromosomal and genomic characteristics [19]. In fact, these approaches can reveal a set of characters usually not accessible by other research methods, thus refining evolutionary investigations [19,20]. Particularly, repetitive DNA sequences, which constitute the major component of the eukaryotic genome, have enormous potential for expanding the knowledge of karyotype differentiation [21]. In addition, the recent use of the comparative genomic hybridization (CGH) has allowed deeper analyses of fish genome organization at the chromosomal level by comparing closely related species [22][23][24][25].
Among the silurid catfishes, chromosomal studies are often restricted to conventional protocols to determine the diploid number (2n) and karyotype composition. Molecular cytogenetic approaches (e.g., chromosomal mapping of rDNA sequences) have been done in only two species of the genus Ompok [26]. Up to date, only 24 species from 8 of the 13 recognized genera had been cytogenetically examined ( Table 1). The overall data show that the chromosome number varies from 2n = 28 in Silurus microdorsalis [27] to 2n = 92 in Kryptopterus cryptopterus [28].

Fluorescence In Situ Hybridization (FISH)-Mapping
The 18S rDNA probe hybridized to only one chromosomal pair in most species, namely B. truncates, K. limpok, K. macrocephalus, M. cheveyi, O. fumidus, P. apogon, P. bleekeri, and W. attu. This site is located in the subtelomeric/telomeric region of the short arms of that chromosome pair in all species, except for K. bicirrhis, in which it is located in the telomeric region of the long arms. Exceptions for this frequent pattern are K. geminus, O. siluroides, and W. micropogon, in which two chromosome pairs bear 18S rDNA genes and S. phaiosoma with five chromosome pairs ( . In contrast, the 5S rDNA sites showed a large variation in distribution, ranging from one chromosome pair in K. limpok, O. fumidus, O. siluroides, W. attu, and W. micropogon, up to six pairs in K. geminus, B. truncates, M. cheveyi, and S. phaiosoma. In addition, four other species had remarkably increased numbers of chromosomes displaying such sites, namely K. bicirrhis and K. macrocephalus with eight, and P. apogon and P. bleekeri with nine chromosome pairs ( . Fluorescence in situ hybridization (FISH) using the (TTAGGG)n telomeric probe revealed hybridization signals on telomeres of all chromosomes of S. phaiosoma ( Figure S1).  (9); Phalacronotus bleekeri (10); Silurichthys phaiosoma (11); Wallago attu (12); Wallago micropogon (13); Silurus aristotelis (14); and Silurus glanis (15) arranged following Giemsa-staining. Bar = 5 µm.

Fluorescence In Situ Hybridization (FISH)-Mapping
The 18S rDNA probe hybridized to only one chromosomal pair in most species, namely B. truncates, K. limpok, K. macrocephalus, M. cheveyi, O. fumidus, P. apogon, P. bleekeri, and W. attu. This site is located in the subtelomeric/telomeric region of the short arms of that chromosome pair in all species, except for K. bicirrhis, in which it is located in the telomeric region of the long arms. Exceptions for this frequent pattern are K. geminus, O. siluroides, and W. micropogon, in which two chromosome pairs bear 18S rDNA genes and S. phaiosoma with five chromosome pairs (Figures 3-5).

Comparative Genomic Hybridization (CGH)
CGH experiments employing the gDNA of Kryptopterus (K. geminus x K. limpok) and Wallago (W. attu x W. micropogon) indicated a large genomic divergence between the congeneric species. Specifically, K. geminus and W. attu exhibited many hybridization sites in the centromeric and terminal chromosomal regions when their own gDNA probes were hybridized against their chromosomal background. However, the K. limpok and W. micropogon gDNA probes produced only some weak terminal signals when hybridized against the K. geminus and W. attu chromosomes, respectively. In contrast, the Phalacronotus species (P. bleekeri and P. apogon) showed a significant shared repetitive content ( Figure 6).

Comparative Genomic Hybridization (CGH)
CGH experiments employing the gDNA of Kryptopterus (K. geminus x K. limpok) and Wallago (W. attu x W. micropogon) indicated a large genomic divergence between the congeneric species. Specifically, K. geminus and W. attu exhibited many hybridization sites in the centromeric and terminal chromosomal regions when their own gDNA probes were hybridized against their chromosomal background. However, the K. limpok and W. micropogon gDNA probes produced only some weak terminal signals when hybridized against the K. geminus and W. attu chromosomes, respectively. In contrast, the Phalacronotus species (P. bleekeri and P. apogon) showed a significant shared repetitive content ( Figure 6). Phalacronotus bleekeri (i-l) after comparative genomic hybridization (CGH) procedures. Male-derived genomic probes from K. geminus and K. limpok were hybridized together against male chromosomes of K. geminus (a-d). Male-derived genomic probes from W. attu and W. micropogon were hybridized together against male chromosomes of W. attu (e-h). Male-derived genomic probes from P. bleekeri and P. apogon were hybridized together against male chromosomes of P. bleekeri (i-l). First column (a,e,i): DAPI images (blue). Second column (b,f,j): hybridization pattern using K. geminus (b), W. attu (f), and P. bleekeri (j) gDNA probes (red). Third column (c,g,k): hybridization pattern using K. limpok (c), W. micropogon (g), and P. apogon (k) gDNA probes (green). Fourth column (d,h,l): merged images of both genomic probes and DAPI staining. The shared genomic regions are depicted in yellow. Bar = 5 µm.

Discussion
The comparison of cytogenetic data for silurid species uncovered a large genomic diversification. This has been highlighted by some published data (Table 1), as well as here by highly diversified 2n, karyotype structures, numbers, and positions of ribosomal genes, and likely also by genomic differentiation, as preliminarily demonstrated by CGH experiments. The review of available cytotaxonomic data indicated a remarkable karyotype diversity, where the 2n number ranges from 28 in S. microdorsalis to 92 in Kryptopterus cryptopterus and K. geminus (Table 1). Oliveira and Gosztonyi [68] suggested that 2n = 56 corresponds to the typical number of chromosomes for Phalacronotus bleekeri (i-l) after comparative genomic hybridization (CGH) procedures. Male-derived genomic probes from K. geminus and K. limpok were hybridized together against male chromosomes of K. geminus (a-d). Male-derived genomic probes from W. attu and W. micropogon were hybridized together against male chromosomes of W. attu (e-h). Male-derived genomic probes from P. bleekeri and P. apogon were hybridized together against male chromosomes of P. bleekeri (i-l). First column (a,e,i): DAPI images (blue). Second column (b,f,j): hybridization pattern using K. geminus (b), W. attu (f), and P. bleekeri (j) gDNA probes (red). Third column (c,g,k): hybridization pattern using K. limpok (c), W. micropogon (g), and P. apogon (k) gDNA probes (green). Fourth column (d,h,l): merged images of both genomic probes and DAPI staining. The shared genomic regions are depicted in yellow. Bar = 5 µm.

Discussion
The comparison of cytogenetic data for silurid species uncovered a large genomic diversification. This has been highlighted by some published data (Table 1), as well as here by highly diversified 2n, karyotype structures, numbers, and positions of ribosomal genes, and likely also by genomic differentiation, as preliminarily demonstrated by CGH experiments. The review of available cytotaxonomic data indicated a remarkable karyotype diversity, where the 2n number ranges from 28 in S. microdorsalis to 92 in Kryptopterus cryptopterus and K. geminus (Table 1). Oliveira and Gosztonyi [68] suggested that 2n = 56 corresponds to the typical number of chromosomes for Siluriformes, as this same number is found in Diplomystes, a sister group of all extant Siluriformes [69], in addition to 2n = 54-58 being the most frequent pattern among siluriform catfishes. The extensive numerical variation of the chromosome number, both below and far above the supposedly basal 2n, as well as the rate of bi-armed chromosomes in the karyotypes, indicate that a diversified number of rearrangements including fissions, fusions, and inversions may have acted to give rise to karyotypic diversity noticed in this family.
What could have driven the extensive karyotype diversification among silurid species? It is widely known that karyotype diversification relates to speciation processes [70][71][72], sometimes with repetitive DNAs acting as primary driving forces (reviewed in the work of [73]). The mapping of repetitive sequences, especially ribosomal genes, has proven useful for estimating evolutionary karyotype changes [74]. Although rDNAs represent conservative elements of the eukaryotic genomes, recent studies have shown that the dynamism of the rDNA clusters is strongly related to significant intragenomic diversification [21,[75][76][77][78][79][80][81]. Accordingly, rDNA elements showed remarkable differences among silurid species, especially regarding the high variability in the number and position of the 5S rDNA sites as compared with the more stable pattern of the 18S rDNA sites.
Extensive chromosomal variability of 5S rDNA loci also has been described for several other fish groups ( [21,22,79,82]; for review, see the work of [74]). A question that arises is whether the dispersion of this rDNA class would be a byproduct of genomic/chromosomal changes. However, the absence of a direct correlation between higher 2n numbers and amplification and dispersion of the 5S rDNA clusters is an indication that this rDNA class was not the unique trigger for the chromosomal rearrangements occurring among the respective silurid species. In this sense, an alternative and attractive hypothesis refers to the action of transposable elements. Indeed, in several species, a significant fraction of the rDNA units is interrupted by transposable elements (TEs) highly specialized for insertions [82][83][84][85][86]. In some cases, TEs have been postulated to play a decisive role in spreading rDNA sequences over the genome [22,23,82]. Remarkably, structural changes in the location of rDNAs also could be linked with speciation events. In the sister salmonid species, Coregonus albula and C. fontanae, ecological speciation was directly associated with the spreading of rDNA sites, affecting recombination rates in both genomes [22]. It is known that multiple rDNA insertions in new genomic regions may create "hot spots" that promote chromosome rearrangements, representing a pathway for rapid genome reorganization during speciation (reviewed in the work of [84]). Nevertheless, up to now, we have no data concerning TEs among silurid species, which will be the goal of further investigations to assess this hypothesis. Besides, it is known that a variety of teleost lineages have undergone one or more rounds of independent whole-genome duplications (WGDs), which are among the most important evolutionary events occurring in fish species [18]. Although there is no direct indicative that silurids analyzed here have experienced WDG events, we cannot exclude the potential role of this process in the high genomic/chromosomal divergence observed.
The available data for silurids allow us to recognize three particular patterns in relation to the 2n numbers and karyotype structures that they present: (i) congeneric species that are highly divergent, as observed in Kryptopterus and Wallago species; (ii) congeneric species that share similar features, as represented by the two Phalacronotus species, P. apogon, and P. bleekeri; and (iii) particular species displaying a significantly lower chromosome number compared with the other species, as observed in S. phaiosoma. For the latter, although multiple chromosomal fusions would be expected to be related, no interstitial telomeric sequences (ITS) were observed ( Figure S1). However, this does not definitely refute the possibility that fusion events occurred during karyotypic diversification, as losses of telomeric sequences can occur after such rearrangements, leading to gradual shortening of non-functional telomeric arrays [87][88][89].
Regarding the first two above-mentioned scenarios, we performed CGH experiments in order to assess whether they are linked with the repetitive DNA content. The remarkable chromosomal dynamism in both Kryptopterus and Wallago species corresponds with an extensive variation of their repetitive DNA content, as demonstrated by a range of non-overlapping species-specific signals revealing an advanced stage of sequence divergence among their genomes (Figure 7). In fact, such repetitive DNA differentiations occurred concomitantly with 2n and structural changes in karyotypes. In contrast, no substantial variation of repetitive DNA content was found among the Phalacronotus species, where the hybridization of both gDNAs produced no species-specific signal amplifications ( Figure 6). In these species, karyotypic changes were markedly reduced. As repetitive DNAs are highly abundant in eukaryotic genomes and display faster evolutionary rates [19,90,91], their role as the main factor in promoting karyotype rearrangements has been extensively investigated. Several reports have evidenced huge inter-population variations of this genomic fraction, promoting biodiversity and possibly linked with ongoing speciation and differentiation of sex-specific regions [24,[92][93][94][95].
Other siluriform groups also experienced massive karyotype differentiation. Clarias species (Clariidae), for example, display a large range of 2n number, from 48 to 104 [79,96], in some cases also including polyploidization and interspecific hybridization events. In C. batrachus (2n = 104), a surprising spread of the 5S rDNA sequences over 27 chromosomal pairs occurs, directly linked with multiple centric fissions [79]. In addition, some other siluriform lineages experienced large karyotypic differentiation such as Callichthyidae, Loricariidae, and Trichomycteridae. In contrast, in families Amyblicipitidae, Ictaluridae, and Sisoridae, only few species possess a reduced 2n number (reviewed in the work of [97]). However, it is evident that siluriform catfishes have, in general, much higher karyotypic diversity than their sister lineage Characiformes. With caution, in view of the fact that only about 15% of the siluriform fishes have been cytogenetically examined to date, it is noteworthy that the largest chromosomal diversity was observed for Siluridae.

Individuals and Mitotic Chromosome Preparation
Fifteen silurid species were collected from distinct natural ecosystems of Thailand and Europe ( Figure 7). The numbers and sexes of the individuals are presented in Table 2. The specimens were deposited in the fish collections of the Cytogenetic Laboratory, Department of Biology, Faculty of Science (Khon Kaen University) and National Museum of Natural History, Paris (MNHN 1997-0481, MNHN 1996-1382. Mitotic chromosomes were obtained by the protocol described in the work of [98]. All the experiments followed ethical protocols, and anesthesia was conducted with clove oil prior to the sacrifice of the animals. The process was approved by the Animal Ethics Committee of Khon Kaen University based on the Ethics of Animal Experimentation of the National Research Council of Thailand AEKKU23/2558. Samples of S. glanis and S. aristotelis were obtained under state fisheries permits and research was conducted with approval from the University of Thessaloniki Ethics Committee.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Kryptopterus bicirrhis
To Daeng peat swamp forest (Thailand) (site 2) (07  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed. )

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.
; 06  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed. )

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed. )

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.
; 07  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.
; 05  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Silurichthys phaiosoma
To Daeng peat swamp forest (Thailand) (site 11) (04  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Silurus aristotelis
Trichonida Lake (Greece) (site 14) (03  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.  Sites 1 to 15 correspond to the localization of each collection region shown in Figure 7.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.

Fluorescence In Situ Hybridization (FISH)
FISH was done under high-stringency conditions on metaphase chromosome spreads [99], with specific probes for 5S and 18S rDNA and telomeric sequences. The 5S rDNA probe included the transcriptional segment of the 5S rRNA gene, with 120 base pairs (bp), and the 200-base pair non-transcribed spacer (NTS) [100]. The 18S rDNA probe corresponded to a 1400 base-pair segment of the 18S rDNA gene [101]. Both rDNA probes were directly labeled with the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany) by the fluorescent labels Atto488 (18S rDNA) and Atto550 (5S rDNA), according to the manufacturer's manual. We applied both rDNA probes in all analyzed species, with the exception of S. aristotelis, where only 18S rDNA mapping was performed.
In order to check the presence of ITS (interstitial telomeric sequences), telomeric (TTAGGG) n sequences were mapped in the species with the lowest 2n (S. phaiosoma) using the DAKO Telomere PNA FISH Kit/Cy3 (DAKO, Glostrup, Denmark).

Comparative Genome Hybridization (CGH)
Total genomic DNA (gDNAs) of the K. geminus, K. limpok, P. apogon, P. bleekeri, W. attu, and W. micropogon were extracted from liver tissue by the standard phenol-chloroform-isoamyl alcohol method [102]. As substantial variation in both 2n number and karyotype formula were observed among species of the genus Kryptopterus and Wallago, the gDNA of K. geminus was compared with that of K. limpok in metaphase chromosomes of K. geminus. Similarly, the gDNAs of W. attu and W. micropogon were hybridized in metaphase chromosomes of W. attu. For these purposes, gDNAs of K. geminus and W. attu were directly labeled with Atto550 using the Nick-translation Labeling Kit (Jena Bioscience, Jena, Germany), while the gDNAs of K. limpok and W. micropogon were labeled with Atto488. To block common genomic repetitive sequences, C0t-1 DNA (i.e., a fraction of genomic DNA enriched for highly and moderately repetitive sequences), prepared according to Zwick et al. [103], was used in all experiments. The final hybridization mixture for each experiment was composed of 500 ng labeled DNA of each compared species, plus 15 µg of male-derived C0t-1 DNA from the respective species and the hybridization buffer (50% formamide, 2× SSC, 10% SDSC 10% dextran sulfate and Denhardt's solution, pH 7.0). The gDNA of Phalacronotus apogon (Atto488) was also compared with that of P. bleekeri (Atto550) against metaphase chromosomes of P. apogon. The CGH experiments were performed according to Symonová et al. [22].

Cytogenetic Analyses
At least 30 metaphase spreads per individual were analyzed to confirm the 2n, karyotype structure, and FISH results. Images were captured using an Axioplan II microscope (Carl Zeiss Jena GmbH, Germany) with CoolSNAP and the images were processed using Image Pro Plus 4.1 software (Media Cybernetics, Silver Spring, MD, USA). Chromosomes were classified as metacentric (m), submetacentric (sm), subtelocentric (st), or acrocentric (a), according to the arm length ratios [104].

Conclusions
Chromosomal characteristics, including the mapping of repetitive DNA sequences and CGH procedures, clarified the evolutionary dynamism among silurid species. In this sense, the known extensive diversification of their karyotypic macrostructure could be better characterized. Our data provide evidence for a direct correlation between the genomic repetitive content and the notable karyotypic divergence in silurids. Thus, it is likely that repetitive DNAs played a direct role in promoting the chromosomal differentiation and biodiversity within this fish family.