New karyotype records for the genus Proechimys (Rodentia: Echimyidae) from Brazilian Amazonia

Abstract We present new karyotype records for six Proechimys species from the Brazilian Amazon. P. echinothrix from the region of Purus River had 2n = 32 chromosomes and a FN = 58, while P. cuvieri from the region of the Japurá River presented 2n = 28 and FN = 46. All individuals presented hybridization with an 18S rDNA probe in a single chromosome pair, with the exception of P. cuvieri from the Japurá region, which presented a third signal in one of the homologs of pair 1. No ITS were found in any of the individuals. Our data supports the hypothesis that the P. cuvieri population from the Japurá Basin and P. echinothrix from the lower Purus are new taxonomic entities. Our data expand the geographic distribution of the cytotype (2n = 40, FN = 54) described for P. gardneri from the Madeira River, and the cytotype (2n = 46, FN = 50), described for P. guyannensis, as well as the recently-described cytotype of P. goeldii (2n = 16, FN = 14). No clear pattern of chromosomal evolution has yet been defined in Proechimys, despite the considerable karyotypic diversity of the genus.


Introduction
The Neotropical rodents of the family Echimyidae are considered the most diverse of the infraorder Hystricognathi, not only in terms of their taxonomy, but also their ecology, morphology, and adaptations (Araújo et al., 2014). The Echimyidae, which includes approximately 90 species in 19 genera (Woods and Kilpatrick, 2005;Patton and Leite, 2015), is a prime example of major adaptive radiations in the Hystricognathi (Fabre et al., 2012). The echimyid genera include Proechimys, which is the most species-diverse, with 22 species (sensu Patton and Leite, 2015), distributed primarily in Amazonia (da Silva et al., 2001;Iack-Ximenes et al., 2005;Patton and Leite, 2015). However, Proechimys is taxonomically problematic, due to the phenotypic similarities among its species and its considerable intraspecific variation. The cytogenetics of Proechimys is also complex, with diploid numbers ranging from 14 to 62 chromosomes and at least 62 known karyotypes Amaral et al., 2013).
A number of studies have shown that non-codifying repetitive DNA sequences play a fundamental role in cell maintenance, and are involved in regulatory mechanisms that determine gene expression. These sequences may result in phenotypic changes, as well as being involved in the speciation process through the evolution of the host genome (Kazazian, 2004;Martins, 2007;Vitte et al., 2014;Zeigler, 2014). Given this, the physical-chromosomal mapping of repetitive sequences may provide important insights into the structure of the genome and generate chromosomal markers, which may be extremely valuable for evolutionary studies, including the identification of specific chromosomes and rearrangements, and in particular the identification of sex chromosomes (Martins, 2007).
Given the persistent taxonomic uncertainties in Proechimys, the existence of the same diploid number in different species, different diploid numbers in the same species, the occurrence of sympatry in many species, and the relatively recent diversification of the genus (which together make it an excellent model for evolutionary studies in the echimyids), the objective of the present study was to provide new karyotype data for the genus. The findings include the location of the 18S rDNA sequences and the telomeric DNA in six Proechimys species, based on individuals collected from different sites located throughout the Brazilian Amazon region. These data were compiled in an attempt to decipher the chromosomal mechanisms involved in the evolution of the genus.

Material and Methods
In the present cytogenetic study, we analyzed 56 Proechimys individuals representing six species (Table 1) Cell suspensions were obtained from bone marrow using the "air-drying" method of Ford and Harmerton (1956) with modifications. Cell division was blocked in vivo using colchicine at a concentration of 0.0125% (defined by the best distension of the chromosomes), at a proportion of 1 mL per 100 g of body weight, for 20 min. The bone marrow was removed from the femur using jets of hypotonic solution (0.075 M KCl) and maintained in this solution for 30 min at 37 ºC. The samples were then washed three times (10 min each time) in Carnoy fixative after pre-fixation. The cell suspensions were dripped onto glass slides and stained with 5% Giemsa for 10 min. The constitutive heterochromatin was stained by C-banding according to Sumner (1972). The Nucleolus Organizer Region (NOR) was located using the procedure described by Howell and Black (1980). Repetitive regions of the 18S rDNA and telomeres were mapped by fluorescence in situ hybridization (FISH) (Pinkel et al., 1986) with adaptations, at a stringency of 77%. The probes were obtained using standard primers for mammals. For the 18S rDNA probe, the primers were F 5'-CCG CTT TGG TGA CTC TTG AT-3' and R 5'-CCG AGG ACC TCA CTA AAC CA-3' (Gross et al., 2010), while for the telomeric region (Ijdo et al., 1991), they were 2 Eler et al.

Results
We extended the known geographic distributions of P. gardneri, P. guyannensis and P. echinothrix by survey-ing areas not previously sampled for these species. We also recorded new data in the chromosome complements of P. cuvieri, P. goeldii, and P. echinothrix. Data on the localization of 18S rDNA and telomeric sequences are presented for all studied species.
The P. gardneri individuals collected from the Abufari Biological Reserve, on the right margin of the Purus River (locality 5) had 2n = 40 (12m + 4sm + 22a + XX/XY) and FNa = 54, with acrocentric X and Y chromosomes of Karyotype records for the genus Proechimys 3 medium and small size, respectively (data not shown). The simple NOR was located interstitially on the long arm of the second submetacentric pair (8). Blocks of heterochromatin were found on four metacentric pairs (3-6), on all the acrocentric chromosomes, and on the X chromosome.
The P. guyannensis individuals from the Saracá-Taquera National Forest (locality 9, right margin of the Trombetas River) and the Trombetas Biological Reserve (locality 9, left margin) had a 2n = 46 (4m + 2sm + 38a + XX/XY) and FNa = 50, with acrocentric pairs 4, 5, and 6 being significantly larger than all the others of the complement (data not shown). The X and Y chromosomes are acrocentric, and the Y chromosome is approximately half the size of the X. The NOR is simple, located interstitially on the long arm of pair 3 (sm). Heterochromatin blocks were observed in the centromeric region of 10 autosomal pairs and the X chromosome, while the Y chromosome is totally heterochromatic.
The P. goeldii individual, from the Jacinto Stream, on the right margin of the Purus River (locality 4) had 2n = 16 (14a + XX/XY) and FNa = 14. The X chromosome is the largest submetacentric ( Figure 2a). As the only specimen of this species is a female, we have no data on the Y chromosome. The blocks of constitutive heterochromatin are located in the centromeric region of all the autosomal chromosomes, while in the X chromosome, in addition to the centromeric region, the short arms are completely heterochromatic ( Figure 2b). The NOR is simple, located interstitially on the long arms of pair 6 (Figure 2c).
In P. echinothrix from the Canutama Extractive Reserve (left margin of the Purus River; locality 3) and the Jacinto stream (right margin of the Purus river ; locality 4), we found 2n = 32 (16m + 8sm + 4st + 2a + XX/XY) and FNa = 58. The sex chromosomes are metacentric, and the X is practically twice the size of the Y (Figure 2d). Pairs 9 and 10 (sm), and 13 (st) are significantly larger than all the others of the complement. The blocks of constitutive heterochromatin are located in the centromeric region of all the autosomes and the X chromosome. The short arm of pair 12 is completely heterochromatic (Figure 2e). The simple NOR is located interstitially on the long arms of pair 11 (sm), coinciding with the secondary constriction ( Figure  2f).
The P. cuvieri individuals collected at the Taboca community, on the Japurá (locality 1), had 2n = 28 (18m + 2st + 6a + XX/XY) and FNa = 46. The X and Y chromosomes are acrocentric, and the X is significantly larger than the Y chromosome ( Figure 2g). The constitutive heterochromatin is located in pericentromeric blocks in chromosome pairs 1, 2, 4, 5, 6 (m), 10 (st), and 12 (a), and also in the centromere of the X chromosome. The Y chromosome is completely heterochromatic (Figure 2h). The NOR is simple, located in the interstitial region of the long arms of pair 7 (Figure 2i).
The P. cuvieri individuals collected from the region of the Jatapu (locality 8), where the species occurs in sympatry with P. guyannesis, and from locality 9 have 2n = 28 (14m + 4sm + 2st + 6a + XX/XY) and FNa = 46 (data not shown). The X and Y chromosomes are acrocentric, and the X is larger than the Y. The NOR is simple, located in the interstitial region of the long arms of pair 7. The constitutive heterochromatin is located in pericentromeric blocks in pairs 5, 6, 7, 9, 11, 12, and 13, and in the X chromosome.
Similarly, all the individuals analyzed in the present study hybridized the 18S rDNA probe in a single chromosome pair (the NOR-bearing pair), with the exception of the five P. cuvieri individuals from locality 1, which presented a third signal in the pericentromeric region and heterochromatin in one of the homologs of pair 1 in all the cells (Figure 3). All the individuals analyzed here (Table 1) presented telomeric signals in the terminal regions of all the chromosomes, but we found no evidence of the presence of interstitial telomeric sequences (ITSs) in any of the individuals (Figure 4).

Discussion
The phylogeny of the echimyid genus Proechimys is poorly resolved, hampering the understanding of the relationships among its species (Patton et al., 2000). Recent cytogenetic studies have provided increasingly valuable markers for the understanding of the evolution of the genus, given that, while the diploid number does not vary, there are meaningful differences in the chromosome morphology, which are reflected in the fundamental number (FN), and thus in the position of the other markers in the karyotype, as in the location of the nucleolus organizer region (NOR). Cytotaxonomic differences have also been found among species in the sex chromosomes, which have resulted from rearrangements, such as inversions, translocations, and the addition or deletion of heterochromatin. In addition to the fact that the taxonomy of the echimyids is still uncertain, a number of other factors limit the understanding of the true diversity of these rodents, such as the paucity of distributional data and the probable existence of cryptic species. In most cases, in addition, the cytogenetic data are limited to fundamental and diploid numbers (Nagamachi et al., 2015).
In the present study, we have expanded the available chromosomal data for a number of Proechimys species, and have added markers for others. In the case of P. gardneri, for example, the karyotype recorded from the Abufari Biological Reserve is identical to that recorded in individuals collected on the left margin of the Madeira River  Figure 5), including the position of the NOR and distribution of the heterochromatin. This allows us to confirm that this cytotype is distributed from the left margin of the Madeira as far west as the left margin of the Purus, and possibly as far as the Juruá River, which would be consistent with the probable distribution of the species, and would imply sympatry with the cytotype described by da Silva (1998) for the Purus-Juruá interfluve.
In the case of the cytotype of P. guyannensis, which has 2n = 46, the karyotype observed in the individuals from the Saracá-Taquera National Forest and the Trombetas Biological Reserve (present study) is the same as that recorded in the P. guyannensis individuals collected from islands of the Balbina hydroelectric reservoir on the Uatumã River Karyotype records for the genus Proechimys 5  Figure 2). The distribution of this cytotype is thus extended approximately 330 km to the east of its previous limit, reaching the margins of the Trombetas River.
Seven cytotypes have been described for the guyannensis group (sensu Patton, 1987), with diploid numbers of 30, 38, 40, and 46 chromosomes (Machado et al., 2005;Eler et al., 2012;Silva et al., 2012). The 2n = 46 cytotype is found in central Brazilian Amazonia, on the left margin of the Amazon River, in the region between the Uatumã River, in Amazonas state, São João da Baliza, in Roraima, and the lower Trombetas River, in Pará ( Figure 5), where it is sympatric with P. cuvieri. Given the ample geographic distribution of the species of the guyannensis group, and their considerable diversity of cytotypes, we believe that this group may contain a number of different species that have yet to be described formally.
The three known P. cuvieri karyotypes (2n = 28 and FN = 46, FN = 48, and FN = 50) include three pairs of chromosomes that are significantly larger than all the others of the complement, together with at least one submetacentric pair (Maia and Langguth, 1993;Patton et al., 2000;Silva et al., 2012). The individuals from the Japurá River analyzed here (2n = 28, FN = 46) are distinct 6 Eler et al.  from all other representatives of this species in having only two pairs (1 and 10) of large chromosomes and no submetacentric chromosomes, with a relatively larger number of acrocentrics, in comparison with the karyotypes described previously. The distribution of the heterochromatin in the individuals analyzed here is similar to that observed in the individuals from the Uatumã River, with signals being found invariably in the centromeric region of most chromosome pairs (including the X chromosome), with a completely heterochromatic Y chromosome. However, the position of the NOR in the complement varied due to the rearrangements, which altered the chromosomal morphology. Clearly, rearrangements of the pericentric inversion type have occurred in this species, given the alterations in the morphology of the chromosomes, even though the diploid number remained constant.
Given the karyotypic differences found in P. cuvieri from the Japurá basin, we believe that this population may represent a new Proechimys species. By contrast, the karyotype of P. cuvieri from the Jatapú basin is similar to that recorded for this species from the Balbina reservoir on the Uatumã River (Maia and Langutth, 1993), Manaus and the Cuieiras basin , the Jaú basin (Patton et al., 2000), and Vale do Jari , although in this latter case, a small difference was found in the morphology of the sex chromosomes. Up to now, the geographic distribution of this cytotype has been restricted to the north of the Solimões-Amazon channel, from the Jaú National Park eastward virtually as far as the mouth of the Amazon River (Vale do Jari). However, the exact geographic limits are still unclear between the three known P. cuvieri cytotypes -2n = 28 and FN = 46, with two distinct arrangements (Patton et al., 2000;Eler et al., 2012;present study), 2n = 28 and FN = 48 (Reig et al., 1979;Patton et al., 2000) (Figure 6).
In the case of P. echinothrix from the Canutama Extractive Reserve and the lower Jacinto Stream, we recorded a karyotype that is different in a number of aspects from that described by da Silva (1998), i.e., 2n = 32, FN = 60, 20m-sm + 10st + XY, with small acrocentrics. The new cytotype described here for P. echinothrix varies principally in the morphology of the sex chromosomes, in addition to the presence of a medium-sized acrocentric pair, and a reduction in the number of subtelocentric chromosomes, Karyotype records for the genus Proechimys 7 Figure 5 -Geographic distribution (dark area) of the 2n = 46 P. guyannensis cytotype, showing the collecting localities: 1 -Saracá-Taquera National Forest and Trombetas Biological Reserve, in Pará state (present study); 2 -São João da Baliza, Amapá (Bonvicino et al., 2005); 3 -Balbina hydroelectric reservoir, on the Uatumã River, in Amazonas state . from 10 to 4 ( Figure 2d). One other diagnostic chromosome marker is the presence of completely heterochromatic short arms in the autosomal pair 12 (Figure 2e), a feature recorded for the first time in the genus Proechimys.
The maintenance of the diploid number in P. echinothrix, together with the variation in the fundamental number (even considering potential divergences in the classification of the chromosome morphology) points to the role of pericentric inversions in the rearrangement of these karyotypes. It is interesting to note that the geographic distribution proposed by da Silva (1998) for this species restricts its occurrence to the left margin of the Juruá as far as the east of the upper Urucu basin. In this case, the results of the present study extend this distribution to both margins of the Purus River (Figure 7). Amaral et al. (2013) described a very different Proechimys karyotype, in terms of both the diploid number, karyotype formula and multiple sex chromosome (2n = 16/17 e FN = 14), from individuals collected in southern Amazonia, in the Brazilian state of Mato Grosso, which they classified as P. cf. longicaudatus. The authors also proposed that P. cf. goeldii from the northern Mato Grosso state, which has the same karyotypic characteristics (studied by Machado et al., 2005) should be reclassified to P. cf. longicaudatus. Rodrigues da Costa et al. (2016) studied two populations of P. cf. longicaudatus from the western Pará state showing the same karyotypic characteristics found by Amaral et al. (2013), and based phylogenetic analyses using the cyt b gene, have suggested that this new cytotype (including data from Amaral et al., 2013) represents a distinct species unrelated to P. longicaudatus or any representative of the group of P. goeldii, diverging from the propositions by Machado et al. (2005) and Amaral et al. (2013), and remaining with uncertain status in the species groups proposed by Patton (1987). In the present study, we recorded the same cytotype in a female from the lower Purus basin, which almost certainly belongs to this same species. However, unpublished molecular data (MNF Silva, personal communication) link this individual from the Purus River to P. goeldii. This extends the distribution of 8 Eler et al.
this cytotype (2n = 16, FN = 14) to the left margin of the lower Purus. The 18S rDNA probe confirmed the presence of a single nucleolar pair in all six species, as shown by the silver staining. As this sequence plays an important role in transcription, the NOR is invariably preserved, despite the structural rearrangements in the autosomal complement of Proechimys (which include the NOR-bearing pair, as in the P. goeldii individuals analyzed here). This is essential to guarantee the functionality of this sequence in the genome. The NOR-bearing pair is considered to be homeologous, not only among Proechimys species, but in the echimyids as a whole (Yonenaga-Yassuda et al., 1985;Eler et al., 2012). We nevertheless recorded a third pericentromeric 18S rDNA signal in P. cuvieri from the Jatapú basin ( Figure  4b), which coincided with a region of non-active heterochromatin. This signal may represent a fragment of the ribosomal sequence that was left over during rearrangements or transferred from a different region of the genome by transposable elements.
The presence of additional 18S rDNA signals, while uncommon in most vertebrates, is also related to the natural origin and disappearance of repetitive sequences in the ge-nome (Rooney and Ward, 2005). In their study of the variation in the 18S rDNA signal in 19 species of the rodent genus Mus, Cazaux et al. (2011) verified the association of these sequences with centromeric regions, and suggested that these regions represent a hotspot of chromosomal breakage. However, these authors concluded that the enormous diversity of karyotypes found in Mus cannot be accounted for solely by rearrangements, but must also have been determined by other factors, such as epigenetic modifications of the DNA. Here, we present the first evidence of multiple 18S rDNA signals in Proechimys, which was not associated systematically with the karyotypic diversity observed in this genus. Even if the region is transcriptionally inactive (not confirmed by the Ag-NOR) and not associated with chromosomal breakage events, it can be considered to be a cytological marker of the P. cuvieri population from the region of the Japurá River.
Despite the many rearrangements identified in Proechimys, interstitial telomeric sequences (ITSs) have not been found in any of the Proechimys analyzed up to now. Telomeric signals have been observed only in the distal portions of the chromosomes. The lack of ITSs in this genus, in spite of its ample variation in diploid numbers, indi-Karyotype records for the genus Proechimys 9 Figure 7 -Geographic distribution of P. echinothrix. Dark area: known distribution of the species, extended to the region of the Jaú National Park (3) (Patton et al., 2000). Hatched area: proposed extension to the Canutama Extractivist Reserve (2) and the lower Jacinto River (3).
cates that either (i) interstitial telomeric sequences may be present in Proechimys species, but have not been detected by FISH, due to the small number of repetitions or the inadequate sensitivity of this analytical technique (Ruiz-Herrera et al., 2008) and/or (ii) the interstitial telomeric sequences have been eroded or substituted by heterocromatic or satellite DNA sequences following rearrangements, given that they generate fragile sites in the chromosome, and are thus eliminated to avoid breakage. Given this, while it is still not possible to define precisely the patterns of chromosomal evolution in the genus Proechimys, the expansion of both chromosome studies and the number of localities sampled has contributed further to the understanding of the diversity found in this genus. The use of chromosomal painting seems to be a promising way to understand the chromosomal evolution in Proechimys, as demonstrated by Oliveira da Silva et al. (2019), explaining the origin of the multiple system of sex chromosomes in P. cf. goeldii, discussing the fixation of chromosomes rearrangements and the sympatric and allopatric models of speciation for the genus. This evidence provides important insights, including new taxonomic markers, and alternative interpretations of the systematics of the group, which emphasize the need for an integrated approach to the understanding of speciation patterns and evolutionary processes, not only in this genus, but in echimyids in general.