Cytogenetic studies in Hasemania crenuchoides (Characiformes: Characidae) and molecular investigations into kinship relationships of the genus

ABSTRACT

The genus Hasemania is considered incertae sedis from the family Characidae which possesses nine fish species. Hasemania crenuchoides is endemic to the Federal District of Brazil and is under threat of extinction. The objective of this work was to characterize the Hasemania crenuchoides karyotype and investigate phylogenetic relationships using the 5ʹ region of the COI gene. Our results showed a chromosome number of 2n = 50, with a fundamental number equal to 82 and a karyotype formula of 32m/sm+ 18st/a. C banding evidenced that constitutive heterochromatin is not only present in pericentromeric regions, but also in telomeric and interstitial regions. Ag-NOR marking occurred in the short arm of pair 6. DAPI staining revealed that few bands were rich in A/T. The diploid number observed is in accordance with previous findings in Characidae, including the presence of a long metacentric pair, and is considered a plesiomorphic character. Based on the literature, together with cytogenetic and molecular data, it is suggested that Hasemania is possibly not monophyletic, composed of two groups: (1) a basal group with H. hanseni and H. nana; and (2) a derived group with H. crenuchoides, H. kalunga, and H. uberaba. Therefore, the molecular and cytogenetic data presented may contribute to understanding phylogenetic relationships within the family.

KEYWORDS:

• Cytogenetics
• cytochrome oxidase
• C-banding
• DAPI
• Ag-NOR

Introduction

Characiforms is an order consisting of the Otophysi clade, Siluriforms, Gymnotiforms and Cypriniforms. Characiforms have broad diversity in terms of morphology and membership comprising: 18 families, 270 genera and 1700 species of freshwater fish (Hastings et al. 2015). They are also widely distributed, found in all continents, with the exception of Antarctica (Hastings et al. 2015).

The family Characidae (Characiforms) is the most diverse, with more than 1140 species (Hastings et al. 2015), one of which is Hasemania crenuchoides Zarske and Géry, 1999. In Brazil, there are 98 genera and 571 species of fish in the family Characidae (Menezes 2018). However, 38 species of this group are facing extinction, including the species studied herein (Ministry of Environmental Affairs 2014).

Characidae appear to be the most complex with regards to establishing phylogenetic relationships in comparison with other families in the same order. However, recent studies have contributed to the clarification of the phylogeny of this group (Oliveira et al. 2011; Mariguela et al. 2013; Thomaz et al. 2015). Conversely, the family hosts 88 genera considered incertae sedis which were previously grouped in subfamilies, mainly in Tetragonopterinae (Reis 2003). Regarding these genera, some, namely Astyanax (Baird and Girard, 1854), Hyphessobrycon (Durbin, 1908) and Bryconamericus (Eigenmann, 1907), with phylogenetic analyses concerning mitochondrial and nuclear genes, have been shown not to possess a monophyletic group (Javonillo et al. 2010; Oliveira et al. 2011), as is the case for the genus Hemigrammus Gill, 1858 (Oliveira et al. 2011).

With regards to molecular biology, use of the 5ʹ terminal of the cytochrome c oxidase subunit I (COI) gene, utilized in the current work, has been shown to be a versatile tool in DNA barcoding (Hebert et al. 2003). This is due to: (1) the robustness of the universal primers for this gene; and (2) to the high rate of substitution that occurs within it. In addition, the COI gene can also be used in phylogenetic analysis (Patwardhan et al. 2014). It has been applied for different taxa, including fish, and proven useful for evaluating monophyly at the genus level (Amaral et al. 2013) together with phylogenetic relationships of closely related species.

Regarding karyotyping, various species have already been analyzed in the order Characiforms, with the diploid number ranging from 2n = 22 for Nannostomus unifasciatus (Steindachner, 1876) to 2n = 102 in Potamorhina altamazonica (Cope, 1878) and P. squamoralevis (Braga and Azpelicueta, 1983; Porto et al. 1992). Among the families within this group, the Characidae have one of the greatest karyotype diversities (Porto et al. 1992) and the presence of a longer metacentric pair is a characteristic of the karyotype of many species in the same family (Scheel 1973).

The diploid number in characids varies from 28 in Hemigrammus sp. (Scheel 1973) to 75 chromosomes in a specific population of A. scabripinnis (Jenyns, 1842) and A. eigenmanniorum (Cope, 1894; Arai 2011), with 2n = 50 predominating in incertae sedis (Piscor 2012). However, more species need to be analyzed to understand karyotype evolution within the Characidae group, especially in Hasemania (Ellis, 1911). Of the nine species identified in nature, only three have had their karyotype described: H. nana (Lütken, 1875), with 8m+ 42sm (50m/sm) (Arefjev 1990; Moreira et al. 2007), H. kalunga (Bertaco and Carvalho 2010) with 14m+ 20sm+ 8st+ 8a (Utsunomia et al. 2016) and H. crenuchoides (present study), all with 2n = 50. Therefore, the objective of this work was to characterize the Hasemania crenuchoides karyotype and investigate phylogenetic relation within the genus.

Materials and methods

Samples

For the sequencing of the 5ʹ region of the Hasemania crenuchoides COI gene, specimens were collected from the conservation unit named Reserva Biológica da Contagem (Rebio), in the Alto-Tocantins River basin (15°39ʹ09.42ʺ S; 47°52ʹ03.22ʺ W), with the appropriate permit from the relevant authorities (permit number ICMBIO 36,170–1).

For karyotype analysis, 18 specimens (10 male and eight female) were collected from the Córrego Paranoazinho which flows into the São Bartolomeu River in the Paraná River basin (15°40ʹ35.04ʺS; 47°51ʹ31.83ʺW). The appropriate collection permit was also obtained (permit number ICMBIO 42,573–1) and the specimens stored in the Ichthyological Collection of the University of Brasilia (Identifier: CIUnB1048). Slides were prepared from the aforementioned samples in order to obtain the chromosomes.

DNA amplification and sequencing

The initial 5ʹ region of the COI gene was amplified from mitochondrial DNA (mtDNA) extracted from two specimens using the PureLink Genomic DNA Mini Kit (Invitrogen, Carlsbad, USA), together with FishF2 and FishR2 primers (Ward et al. 2005). The PCR amplifications were conducted with reactions that had a final volume of 25 µl, containing 2.5 µl of buffer (10× PCR buffer), 0.75 µl of MgCl₂ (50 mM), 0.5 µl of dNTP (10 µM), 2 µl of template, 0.25 µl of each primer (10 µM) and 0.15 µl of Taq DNA polymerase (5 U µl^–1) (Invitrogen, Life Technologies, Carlsbad, USA). The program used was run under the following conditions: 2 min of initial denaturation at 95°C, followed by 35 cycles of denaturation at 94°C for 30 s, annealed at 54°C for 30 s and extension at 72°C for 1 min. The final extension was at 72°C for 10 min. The PCR products were confirmed by gel electrophoresis and purified using ExoSAP-IT (USB Corporation, Cleveland, USA). The two strands were sequenced using the ABI3730 platform by Myleus Facility (Life Technologies/Thermo Fisher Scientific, Carlsbad, USA). The nucleotide sequences were deposited in the GenBank under the corresponding accession numbers: MF802784 and MF802785.

Phylogenetic analysis

The other six sequences of the COI gene for phylogenetic analysis were obtained from the GenBank databank on the NCBI site (http://www.ncbi.nlm.nih.gov/) and the Public Data Portal on the BOLD Systems site (http://v4.boldsystems.org/). The aforementioned sequences were considered to be correct. However, for Hasemania uberaba (Serra and Langeani 2015), the name of the DNA sequence deposited does not conform to current taxonomy: the COI sequence of H. crenuchoides (accession number: JN988883) is actually the COI sequence of H. uberaba due to taxonomic revision (Serra and Langeani 2015). Hemigrammus marginatus (Ellis, 1911) and Hemigrammus erythrozonus (Durbin, 1909) were included in the phylogram based on previous data that supports the proximity of these species with Hasemania (Javonillo et al. 2010; Oliveira et al. 2011; Mariguela et al. 2013). They were aligned using the MUSCLE algorithm (Edgar 2004) in the Molecular Evolutionary Genetics Analysis (MEGA) program, version 7.0 (Tamura et al. 2013), with predefined parameters, with the exception of the open gap value, which was increased to correct the alignment errors. The alignment was visually inspected for optimality. The p-distances were generated by the same program using the K2P model.

Maximum likelihood (ML) analysis was performed using the MEGA 7.0 program (Tamura et al. 2013). The model general time-reversible substitution was used with discrete gamma distribution, with five categories, and invariable sites (GTR+ Γ + I) and, for comparison, Kimura two parameter (K2P), with two more parameters (K2 + Γ + I). For the two models, 1000 bootstrap pseudo-replications were performed with the aim of checking the reliability of the phylogram generated. The heuristic method employed for ML was the nearest-neighbor-interchange (NNI), and the initial tree was set up starting from the statistical method of neighbor-joining (NJ) and BioNJ. Missing data were treated complete deletion. Chalceus macrolepidotus was defined as an outgroup based on already published data (Oliveira et al. 2011). NJ analysis was also performed in MEGA 7.0. The distance model used was K2P with discrete gamma distribution, with five categories. A total of 1000 bootstrap pseudo-replications were performed.

Samples and chromosome preparation

Following the procedures described by Bertollo et al. (1986), in order to increase the mitotic index, a solution of biological yeast was first injected at a proportion of 0.01 ml g^–1 in close proximity to the cephalic kidney. After 24 h, the animals were treated with a 0.025% colchicine solution for 45 min, anesthetized and euthanized. The renal tissue was removed, macerated and submitted to a hypotonic solution of 0.075 M potassium chloride (KCl) for 30 min at 37°C, prior to fixing in a solution of methanol-acetic acid (3:1).

Conventional staining

Metaphase chromosomes were visualized by staining the prepared slides with 10% Giemsa in a phosphate buffer at pH 6.8 for 10 min. The measurement and capture of the chromosomal parameters were conducted using the IdeoKar program (Mirzaghaderi and Marzangi 2015) and the ratio of the arms for the chromosome classification, in accordance with Levan et al. (1964).

C-banding

C-banding was performed in accordance with the methodology proposed by Summer (1972), with modifications for the detection of constitutive heterochromatin. The slides were treated in 0.2 N hydrochloric acid (HCl) at room temperature for 15 min and rinsed with distilled water a short time later. The slides were subsequently incubated in 5% barium hydroxide (Ba (OH)₂) at 45°C for 30 s and then briefly immersed in 0.1N HCl at 45°C. Finally, after washing in distilled water, they were treated with sodium citrate (2xSSC) at 60°C for 15 min and stained with 10% Giemsa for 15 min.

Ag-NOR detection

Following the technique proposed by Howell and Black (1980), the detection of the nucleolar organizer region (NOR) was realized by silver nitrate impregnation. First, the slides were treated in 0.2N HCl for 3 min at room temperature. A drop of 2% gelatin solution was subsequently added to each slide, followed by two drops of 50% silver nitrate solution on top. The slides were then incubated at 60°C for 5 min in a moist dark chamber and stained with 10% Giemsa for 1 min.

DAPI (4ʹ,6ʹ-diamidino-2-phenylindole) staining

Regions rich in A/T were marked in accordance with Pieczarka et al. (2006). Initially, the slides were treated with alcohol as follows: 2 min twice in 70%, 2 min twice in 90%, and one 4-min exposure in 100%. Slides were then oven-dried at 65°C for 1 h and immersed in a 70% formamide solution at 65°C for 20 s. Slides were subsequently immersed in chilled 70% alcohol for 4 min and then treated with alcohol as follows: 2 min twice in 70%, 2 min twice in 90%, and one 4-min exposure in 100%. Finally, 20 µl of DAPI+ Antifade (Vectashield, Burlingame, USA) were added to each slide.

Results

Karyotype

All of the Hasemania crenuchoides specimens presented a diploid number of 50 chromosomes, possessing a karyotype composed of: three pairs of metacentric chromosomes (m), 13 pairs of submetacentric chromosomes (sm), eight pairs of subtelocentric chromosomes (st) and one pair of acrocentric chromosomes (a), thus having a KF (karyotype formula) equal to 6m+ 26sm+ 16st+ 2a (32m/sm+ 18st/a). The fundamental number (FN) was equal to 82. No heteromorphism between males and females was observed in the karyotype (Figure 1(a), 1(b)).

Figure 1. Hasemania crenuchoides karyotype. (a) Karyotype of the female (2n = 50) with conventional Giemsa staining. Chromosome pair 6 shown with Ag-NOR marking. (b) Karyotype of the male (2n = 50) with conventional Giemsa staining. (c) Karyotype of the female submitted to C-banding. (d) Karyotype of the female with DAPI staining.

In the chromosome analysis, silver nitrate impregnation marked only one pair of chromosomes at metaphase. The Ag-NOR obtained was located in a proximal position of the short arm of submetacentric pair 6 (Figure 1(a)).

C-banding evidenced that constitutive heterochromatin is not only present in pericentromeric regions, but also in telomeric and interstitial regions. Blocks of constitutive heterochromatin were visualized in the short arms of pairs: 2, 3, 20, and 22, and in the long arms of pairs: 5, 12, 17, 18, and 21. Pericentromeric C-bands were observed in pairs: 2, 3, 4, 11, 14, 15, 16, 19, 20, and 23. Two extensive regions of C-bands were evidenced: one in the short arm of pair 3 and another in the long arm of pair 18 (Figure 1(c)). DAPI staining revealed only few regions were rich in A/T. These marked areas are in agreement with the C-banding, presenting only in pairs: 3, 18, and 20 (Figure 1(d)).

Phylogram

For alignment of the sequences in the 5ʹ region of the COI gene, 467 bp were considered. The analysis showed that the composition of nucleotides is formed by T = 31.88%, C = 26.98%, A = 23.34%, and G = 17.80%, with GC content = 44.78%. There were 309 (66.17%) conserved sites and 158 (33.83%) variants. The estimated value of parameter α (shape) for discrete distribution Γ was 0.6689 using the GTR model and 0.9167 for K2P. The transition/transversion bias (R) was estimated by maximum likelihood, obtaining a value equal to 3.87 for GTR and 3.69 for K2P.

In the ML analysis, the nucleotide substitution model used was GTR+ I + Γ because it possesses the lowest Akaike information criterion corrected (AICc) and a low Bayesian information criterion (BIC) generated by the MEGA 7.0 program (Tamura et al. 2013). For comparison, the K2P model was also used in agreement with the trees generated in the Barcode of Life Data Systems (BOLD) in ML and NJ analysis. The p-distances generated by this model varied from 0 to 0.242 (Table 1). The trees were supported by good bootstrap values in general, both in the ML and in the NJ analysis (Figure 2).

Table 1. Pairwise p-distance in black and standard error in blue for the COI gene.

Hasemania uberaba		0.010	0.010	0.009	0.018	0.017	0.018	0.017
Hasemania crenuchoides 1	0.045		0.000	0.011	0.018	0.017	0.018	0.017
Hasemania crenuchoides 2	0.045	0.000		0.011	0.018	0.017	0.018	0.017
Hasemania kalunga	0.043	0.062	0.062		0.017	0.016	0.018	0.016
Hemigrammus marginatus	0.178	0.193	0.193	0.178		0.017	0.013	0.016
Hasemania hanseni	0.143	0.161	0.161	0.141	0.148		0.018	0.012
Hasemania erythrozonus	0.184	0.199	0.199	0.191	0.077	0.161		0.016
Hasemania nana	0.146	0.152	0.152	0.139	0.137	0.079	0.139

Figure 2. Phylogenetic tree drawn up from analyses of maximum likelihood and neighbor-joining of 467 bp in the 5ʹ region of the COI gene belonging to eight specimens. The values to the left of the nodes correspond to the bootstrap values obtained by ML and those to the right, to the NJ.

Discussion

The diploid number 2n = 50 of Hasemania crenuchoides is the same as that found for the only two species of the genus that have had their karyotype described: H. nana (Arefjev 1990; Moreira et al. 2007) and H. kalunga (Utsunomia et al. 2016). This is consistent with previous predominant findings in the Characiformes order (Arai 2011) and in the incertae sedis group of the family Characidae (Oliveira et al. 2007).

With regards to H. nana karyotype structure, two results have been reported in previous studies. The first study (Arefjev 1990) presented KF of 12m+ 18sm+ 10st+ 10a (30m/sm+ 20st/a) and the second study (50m/sm) reported 8m+ 42sm (Moreira et al. 2007). Moreira et al. (2007) raised the possibility that the six specimens analyzed by Arefjev (1990) may not have been correctly identified as H. nana. However, considered as a single species, it would be the occurrence of many chromosomal rearrangements, as the author states (Moreira et al. 2007). The clarification of this problem is complicated further by the fact that Arefjev (1990) did not report exactly where the collection took place. Regarding the chromosome data included in the phylogeny, we only plotted the data of Moreira et al. (2007). The C-banded karyotype of the species is also described, presenting few chromosomes with bands, which are pericentromeric. This fact contrasts with the finding for H. crenuchoides, which additionally presented terminal C-bands. Single Ag-NOR was described in chromosome 7, adjacent to heterochromatic region (Moreira et al. 2007).

Utsunomia et al. (2016) characterized H. kalunga cytogenetically, defining the diploid number as 50 and KF as 14m+ 20sm+ 8st+ 8a (34m/sm+ 16st/a). They described the interstitial and terminal C-bands in some chromosomes, and pericentromeric in all of them. But at H. crenuchoides there are only 10 pairs with pericentromeric blocks of constitutive heterochromatin. The Ag-NOR appeared single and was found in submetacentric pair 13. H. crenuchoides also presented only one pair with Ag-NOR, located in submetacentric 6.

Regarding the phylogram, previous studies have shown the proximity of Hasemania sp. and Hasemania nana to Hemigrammus marginatus (Oliveira et al. 2011; Mariguela et al. 2013) and Hemigrammus erythrozonus (Javonillo et al. 2010) when several species of Characidae were analyzed, demonstrating that the relationships are well founded in the present phylogram. The present data also sustain these relationships by the high bootstrap values shown for the clade that joins H. nana, H. hanseni (Fowler, 1949), H. marginatus, and H. erythrozonus. Furthermore, when the five species of Hasemania are analyzed (H. crenuchoides, H. kalunga, H. nana, H. hanseni, and H. uberaba), the phylogram shows Hasemania as a non-monophyletic group composed of two groups – a basal group with H. hanseni and H. nana, and a derived group with H. crenuchoides, H. kalunga and H. uberaba (Figure 2).

Previous studies based on morphological data have already reported the proximity between H. crenuchoides, H. kalunga (Bertaco and Carvalho 2010; Serra and Langeani 2015), H. uberaba, and H. piatan (Zanata and Serra, 2010; Serra and Langeani 2015). The presence of a completely cartilaginous rhinosphenoid, larger bodies and a similar color positions H. crenuchoides, H. kalunga, and H. uberaba closer together when compared with other Hasemania species (Serra and Langeani 2015). This proximity led to specimens of Hasemania described by Langeani et al. (2007) being considered as H. crenuchoides, until they were reconsidered as belonging to H. uberaba due to taxonomic revision (Serra and Langeani 2015). This proximity is also sustained by the presented phylogram. Thus, it is probable that H. crenuchoides, H. uberaba, and H. kalunga are close, not only considering the molecular data, but also their morphological similarity and cytogenetic data, which are described in Figure 2.

Hemigrammus erythrozonus and Hemigrammus marginatus presented 10m+ 34sm+ 6a (44m/sm+ 6st/a) (Portela-Castro and Júlio 2002) and 12m+ 36sm+ 2a (48m/sm+ 2st/a) (Muramoto et al. 1968), respectively, both with 2n = 50. Additionally, H. marginatus has a single Ag-NOR located in chromosome 19, adjacent to a heterochromatic region as similar to H. nana. Analyzing the karyotype formula (KF), together with m-sm and st-a chromosomes, it is noted that two clades are differentiated: the one that includes: H. crenuchoides, H. uberaba, and H. kalunga, and the other which includes: Hasemania nana, Hasemania hanseni, Hemigrammus marginatus, and Hemigrammus erythrozonus (Figure 2). The first clade presents m-sm varying between 32 and 34, where the KF of Hasemania crenuchoides is 32m/sm+ 18 st/a, and 34m/sm+ 16st/a for Hasemania kalunga. The second clade, in which H. nana appears, has a high predominance of m-sm, varying among 44, 48 and 50. Indeed, the cytogenetic analysis contributes to what is suggested by the molecular data, in showing that the karyotype of H. crenuchoides has chromosomal evolution events that bring it phylogenetically closer to the species from the clade that it is part of, and distance it from H. nana. Rearrangements of the pericentric inversion type may have been responsible for the predominance of metacentric and submetacentric chromosomes in H. nana and H. hanseni.

This work presents the first time that chromosome, morphological and molecular data of the genus are being reviewed together. These are different approaches that provide consistent evidence that this group of Hasemania is not monophyletic.

Conclusion

The karyotype of Hasemania crenuchoides was described and showed that the diploid number of 2n = 50 is the same as that found for the only two species of the genus that have had their karyotypes described – H. nana and H. kalunga. It is consistent with the predominant findings in the Characiformes order and in the incertae sedis group of the family Characidae. Using the literature, together with cytogenetic and molecular data by COI, it was possible to conclude that Hasemania is a not monophyletic group. Additionally, the data from the present study can be used with other molecular phylogeny tools and morphological analysis for future phylogenetic research in the genus Hasemania, thus increasing the number of possible species for comparison and helping to clarify the chromosomal evolutionary process for this group. In order to understand the phylogenetic relationships within Hasemania more thoroughly, it will be interesting to carry out cytogenetic characterization of other species in the genus and other species within the incertae sedis group in the family Characidae.

Acknowledgments

The authors thank Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) from the Pro-Amazonia Project number 3284/2013 for financial support. We thank Edgar Guimarães Bione, a teacher from the University of Brasília-Ceilândia, for help in collection of Hasemania. We thank the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio) for the collection permits n° 42573-1 and 36170-1/2014.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Additional information

Funding

This work was supported by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior [3284/2013].