Chromosomal phylogeny of Vampyressine bats (Chiroptera, Phyllostomidae) with description of two new sex chromosome systems

Background The subtribe Vampyressina (sensu Baker et al. 2003) encompasses approximately 43 species and seven genera and is a recent and diversified group of New World leaf-nosed bats specialized in fruit eating. The systematics of this group continues to be debated mainly because of the lack of congruence between topologies generated by molecular and morphological data. We analyzed seven species of all genera of vampyressine bats by multidirectional chromosome painting, using whole-chromosome-painting probes from Carollia brevicauda and Phyllostomus hastatus. Phylogenetic analyses were performed using shared discrete chromosomal segments as characters and the Phylogenetic Analysis Using Parsimony (PAUP) software package, using Desmodontinae as outgroup. We also used the Tree Analysis Using New Technology (TNT) software. Results The result showed a well-supported phylogeny congruent with molecular topologies regarding the sister taxa relationship of Vampyressa and Mesophylla genera, as well as the close relationship between the genus Chiroderma and Vampyriscus. Conclusions Our results supported the hypothesis that all genera of this subtribe have compound sex chromosome systems that originated from an X-autosome translocation, an ancestral condition observed in the Stenodermatinae. Additional rearrangements occurred independently in the genus Vampyressa and Mesophylla yielding the X1X1X2X2/X1X2Y sex chromosome system. This work presents additional data supporting the hypothesis based on molecular studies regarding the polyphyly of the genus Vampyressa and its sister relationship to Mesophylla. Electronic supplementary material The online version of this article (doi:10.1186/s12862-016-0689-x) contains supplementary material, which is available to authorized users.

Most studies based on morphological data support the sister taxa relationship between the genera Ectophylla and Mesophylla [7,10,11]. However, a study based on cranio-dental characters agrees with the molecular consensus about the close affinity of the genera Mesophylla and Vampyressa [12]. On the other hand, a study of restriction site variation of mitochondrial genes ND3 and ND4 supports the morphological view [13]. A reanalysis based on a direct survey on DNA sequences of these genes agrees with the molecular hypothesis [9].
Cytogenetic studies are agreed on the close relationship between Mesophylla and Vampyressa. All were based on gross chromosome morphology and the shared sex chromosome systems [14][15][16][17]. However, no cladistics analysis was made taking into account the different karyotypes of the species within this group.
Chromosome rearrangements are rare genomic changes (sensu [18]) and, because of their Mendelian inheritance, can be used for phylogenetic inferences.
In this paper, we use multidirectional chromosome painting with whole chromosome probes from two phyllostomid bats (Carollia brevicauda and Phyllostomus hastatus), as well as chromosome banding and Fluorescence In Situ Hybridization (FISH) with 18S rDNA and telomeric probes, to establish a genome-wide comparative chromosomal map for all genera of Vampyressine bats. Using chromosomal rearrangements as characters, we built a phylogeny that sheds some new light on the evolutionary relationships among these bats.

Specimens examined
Representative species of the subtribe Vampyressina were collected in natural habitats during field expeditions to different places in the Amazon Basin (Table 1). Specimens were maintained in the laboratory with food and water, free from stress, until their necessary euthanasia. This study was specifically approved by the Animal Ethics Committee (Comitê de Ética Animal) from Universidade Federal do Pará (Permit 68-2015) over and above the use of approved general protocols. Voucher specimens were fixed in 10

Chromosomal preparation and chromosome banding
Metaphase spreads were obtained from bone marrow preparations after colchicine treatment, following [27] and [28], as well as from fibroblasts cultured according to [29]. G-banding was performed using trypsin treatment [30] and subsequent incubation in saline solution (0.5 × SSC) at 60°C, and staining with Wright's solution [31]. C-banding was carried out according to [32] and Ag-NOR staining followed [33].

Data from literature
In previous publications [25,37] the species Phyllostomus hastatus, Carollia brevicauda, Artibeus obscurus, Uroderma bilobatum and U. magnirostrum were analyzed using the same whole-chromosome probes as used here. We added those data to the present work, and included them in our phylogenetic analysis. Sotero-Caio et al. [38] analyzed Diphylla ecaudata, Diaemus youngi and Desmodus rotundus also using the same probes. These species (Desmondontinae) were used as outgroup on phylogenetic analysis because of their basal position in published molecular phylogenies and because this is a monophyletic subfamily with well supported molecular [1] and chromosomal [38] synapomorphies.

Phylogenetic analysis
A data matrix was established where the chromosomal rearrangements are the characters under analysis, and discrete chromosomal syntenic blocks are the characters states that reflect the occurrence of rearrangements (Additional file 1: Table S1). The data matrix also included species previously analyzed [25,37,38], and it is available as a supplementary file. Maximum parsimony (MP) analysis was made using PAUP 4.0b10 [39]. A heuristic search to find most parsimonious tree(s) was performed using Tree Bisection Reconnection (TBR) branch-swapping; the bootstrap posterior probability was obtained with one thousand replicates. The Bremer support or "decay index" [40,41] was calculated to verify the inconsistency of the branches in the consensus tree using the software "Tree Analysis Using New Technology" (TNT) version 1.1 [42], freely distributed by the Willi Hennig Society.

Results
Karyotype description and multidirectional chromosome painting in Chiroderma villosum (CVI) We analyzed two specimens of C. villosum (Table 1). Both samples show 2n = 26 chromosomes and Fundamental Number, FN = 48 ( Fig. 1, CVI). The autosome complement consists solely of biarmed elements (meta, submeta and subtelocentric). The X is a medium-sized subtelocentric and the Y is a small acrocentric chromosome. The constitutive heterochromatin (CH) is restricted to centromeric regions of all chromosomes, including the X (Additional file 2: Figure S1, CVI). FISH using telomeric probes show signals at the distal ends of chromosomes, with dominant strong signals at the centromeric regions of all but pairs 6, 8, 11 and the Y (Additional File 3: Figure S2, CVI). Silver staining (not shown) and FISH with 18S rDNA revealed a single Nucleolar Organizer Region (NOR) in the short arm of pair 7 (Additional File 3: Figure S2, CVI). Multidirectional chromosome painting using wholechromosome probes from Carollia brevicauda (CBR) and Phyllostomus hastatus (PHA) revealed 25 and 23 shared conserved syntenic segments, respectively, with the number of syntenic blocks ranging from one to four on each chromosome of CVI ( Fig. 1, CVI).

Karyotype description and multidirectional chromosome painting in Vampyriscus bidens (VBI)
All five specimens of V. bidens (Table 1) collected at different sites in the Brazilian Amazon have a 2n = 26 and FN = 48, with all chromosomes biarmed. The X is subtelocentric and the Y is acrocentric (Fig. 1, VBI). The CH is located at the centromeric region of chromosomes and in the proximal region of the Y chromosome (Additional file 2: Figure S1, VBI). Hybridization with telomeric probes showed signals at the tips of chromosomes and weaker signals at the centromeric regions of some chromosomes in association with the CH (Additional file 3: Figure S2, VBI). FISH using 18S rDNA probes, and silver staining, showed two NOR located on pairs 8 and 9 (Additional file 3: Figure S2, VBI). Chromosome painting using CBR and PHA probes revealed 25 and 23 conserved segments on V. bidens, respectively. The number of syntenic blocks ranges from one to three in different VBI chromosomes (Fig. 1, VBI).

Karyotype description and multidirectional chromosome painting in Vampyriscus brocki (VBR)
Vampyriscus brocki collected in Amapá State has a 2n = 24 and a FN = 44 ( Fig. 1, VBR). Its karyotype, as in the congeneric VBI, has only biarmed elements (meta, submeta and subtelocentric). The X is subtelocentric and the Y is acrocentric. The CH is located in the pericentromeric regions of chromosomes and the proximal part of Y (Additional file 2: Figure S1, VBR). FISH signals with telomeric probes were detected at the tips of chromosomes (Additional file 3: Figure S2, VBR). Signals from an 18S rDNA probe were detected at the distal portion of the short arm of pairs 5, 7, 8, and 9; in agreement with silver staining results (Additional file 3: Figure S2, VBR).
Chromosome painting using CBR and PHA probes detect 25 and 23 syntenic segments, respectively. We found a maximum of five segments per chromosome with CBR probes and four with PHA ( Fig. 1, VBR).

Karyotype description and multidirectional chromosome painting in Vampyressa thyone (VTH)
Two specimens of V. thyone caught from Serra of Apeú in Roraima state were karyotyped and show a 2n = 24 complement in the female and 2n = 23 in the male, with FN = 20 ( Fig. 1, VTH). The chromosome complement has only uniarmed pairs, and the sex chromosome system is X1X1X2X2/X1X2Y, so there is one chromosome less in the male karyotype. The CH is located in small bands in the pericentromeric regions of chromosomes, with one strong CH block in the interstitial part of ×2 (Additional file 2: Figure S1, VTH). FISH using the telomeric probe showed signals at the tips of chromosomes (Additional file 3: Figure S2, VTH); the 18S rDNA probe revealed two NORs on pairs 5 and 9 (Additional file 3: Figure S2, VTH), confirming previous results from silver nitrate (AgNOR) staining.
Chromosome painting revealed 28 and 29 shared segments with CBR and PHA probes, respectively. Results from each set of probes showed a maximum of four syntenic blocks per chromosome in VTH (Fig. 1, VTH).

Karyotype description and multidirectional chromosome painting in Mesophylla macconnelli (MMA)
We analyzed three specimens of M. macconnelli. The diploid number is 2n = 22 in female and 2n = 21 in males, with the FN = 18 (Fig. 1, MMA). The chromosomal complement has only uniarmed pairs. The sex chromosome system is X1X1X2X2/X1X2Y. The CH is present on small bands in the pericentromeric region of the autosomes and on the proximal part of the Y (Additional file 2: Figure S1, MMA). Probes revealed telomeric signals on the distal ends of all chromosomes (Additional file 3: Figure S2, MMA). Staining with silver nitrate (Ag-NOR) and FISH with 18S rDNA probe showed one signal in the distal portion of pair 2 (Additional file 3: Figure S2, MMA).
Chromosome painting on MMA metaphases revealed 28 syntenic segments using probes from CBR, and 30 syntenic segments using probes from PHA. We found a maximum of four segments per chromosome matching CBR probes, and five matching PHA probes (Fig. 1, MMA).

Karyotype description of Platyrrhinus incarum (PIN) and Vampyrodes caraccioli (VCA)
We analyzed single specimens of P. incarum and V. caraccioli. Both species have a 2n = 30, FN = 56 complement and a Neo-XY sex chromosome system (Additional file 4: Figure S3, PIN and VCA, respectively). The CH is located in the pericentromeric region of all chromosomes for PIN (Additional File 2: Figure S1, PIN) and in the centromeric regions of the chromosomal complement of VCA (Additional file 2: Figure S1, VCA); additional heterochromatic blocks are present at the distal ends of the short arms of chromosomes 5, 7 and 13 in both species. In PIN, FISH with telomeric probes produced signals at the tips of chromosomes and in the centromeric regions of pair 13 for both species. Additionally, minor signals in some pairs co-located with CH for PIN (Additional file 3: Figure S2, PIN). In VCA, telomeric probes detect signals at the tips of chromosomes and the centromeric regions of pairs 1, 5 and 13 (Additional file 3: Figure S2, VCA). Probe signals from 18S rDNA are located in pair 7, confirming the silver staining in PIN (Additional file 3: Figure  S2, PIN), and in chromosome pairs 5 and 7 in VCA.
The banding patterns of these two species (PIN and VCA) are similar to those of Artibeus obscurus (Additional file 4: Figure S3 [37]. The only differences found are a pericentric inversion on pair 5 in PIN, the multiple sex chromosome system in A. obscurus (XX/XY1Y2) and the Neo-XY in PIN and VCA. Because of the similar Gbanded karyotypes in PIN, VCA, and AOB, we infer that the syntenic groups would be the same as in AOB.

Phylogenetic analyses using chromosome as characters
We analyzed all representative genera in Vampyressina. Multidirectional chromosome painting identified chromosome homologies and we used a total of 86 discrete chromosomal characters to build a matrix of their presence or absence (Additional File 1 Table).

Genera Platyrrhinus and Vampyrodes
There are only two recognized species in Vampyrodes [6], whereas Platyrrhinus has at least 21 currently described species [5,43,44]. Both genera have the same 2n = 30 and FN = 56 karyotype, as well as the Neo-XY sex chromosome system (found in all species described thus far [14,15,45]). G-banding patterns have been published only for Platyrrhinus.
In our phylogeny, P. incarum and V. caracciolli share a polytomy along with the branch corresponding to the genera Vampyriscus and Chiroderma. This polytomy of P. incarum and V. caracciolli can be explained by their similarity in G-banding pattern and presence of only some autapomorphies. Since these species have no synapomorphies, their precise positions on the phylogeny  Table 1. The characters "Y/15 acro" and "Y/15 meta" are different chromosomes and resulted from different rearrangements. While in the character 9/13 includes both these chromosomes in full, 9q/13 includes the 13 and the long arm of 9. Figure 3 details the evolution of the sex chromosome systems; note that it matches perfectly with the phylogeny shown here cannot be determined, despite molecular studies suggest a close relationship among Vampyrodes, Platyrrhinus, Vampyriscus and Chiroderma [9]. The chromosomal homeology among these species and Artibeus obscurus supports the hypothesis that the 2n = 30-31 karyotype is probably the ancestral condition for the Stenodermatinae [46]. We found only two characteristics supporting the close relationship of Platyrrhinus and Vampyrodes, the shared NOR location observed on pair 7 and an interstitial telomeric site found on pair 13. These characteristics were not included in our phylogeny due to the variability of these marks in the Stenodermatinae. Because of the retention of several ancestral characters and the NOR position, we conclude that the genera Platyrrhinus and Vampyrodes probably are sister taxa, in accordance with most studies based on either the morphological or molecular data [7,9,43,47].

Inter-and intraspecific variation in the genus Uroderma
Uroderma shows wide karyotypic variation when compared with other species in Stenodermatinae. Three chromosomal races have been described for U. bilobatum (Baker et al. 1972 [48], 1975 [49]). Specimens with 2n = 42 and FN = 50 occur in South America, east of the Andes (2n = 42 race). Individuals with 2n = 38 and FN = 44 are distributed on the Atlantic side of Guatemala, Honduras and Yucatan Peninsula of Mexico (Central America) and along the Pacific coast of Colombia and northern Ecuador (2n = 38 race). Specimens with 2n = 44 and FN = 48 occur in the Pacific coast of Central America in El Salvador, Guatemala, Honduras and México (2n = 44 race). A hybrid zone was found in Honduras, but introgression between the two races is low (Hoffmann et al. 2008 [46]). This hybrid zone has been extensively studied since the second half of the 20 th century and there is major controversy regarding primary and secondary origins of the zone, the role of selective force in its maintenance, and the amount of gene flow across this zone (Baker et al. 1972 [48], Hoffmann et al. 2008 [46]).
Our data support the hypothesis that the 2n = 42 race is distributed throughout South and Central America. Geographic barriers such as the Andean mountains and probably the Gulf of Fonseca (which was under water throughout much of the Pleistocene) separated the race with 42 chromosomes, and gave rise to the 44 and 38 cytotypes (Hoffmann et al. 2008 [46]). There is temporal agreement between the geographic barrier and divergence of the three races that happened around 0.9 to 0.2 Mya (Hoffmann et al. 2008 [46]). Baker et al. (1982 [50]) have proposed the rearrangements that took place to form the cytotypes with 42, 44 and 38 chromosomes. We suggest a different set of events: first a pericentric inversion and a fusion/fission differentiated the races with 42 and 44 chromosomes; second a translocation and a tandem fusion from a karyotype with 2n = 42 would explain the variation observed in races with 38 chromosomes. Thus, we agree with Hoffmann et al. (2008 [46]) that the ancestral chromosomal race on the deepest branch could have 42 chromosomes, and that the hybrid zone arose later.
Our results agree with the phylogenetic analysis of nuclear TSHB-I2 sequences made by Hoofer et al. (2008 [52]), strongly supporting the monophyly of Vampyriscus bidens with V. brocki (and their relationship with Chiroderma), the sister taxon relationship between Mesophylla and Vampyressa, as well as the polyphyly of Vampyressa. As pointed out by those authors, these results are in strong agreement with previous molecular studies by Baker et al. (2003 [1]).
Despite our results showing similar pairwise intergeneric relationships as the molecular studies, we found different branching between these groups. Although our analysis recovered a basal branching for the non-Stenodermatinae outgroups (PHA and CBR), and the monophyly of Stenodermatinae, the position of AOB within the Vampyressina disagrees with molecular and morphological data. Nevertheless, the presented topology is strongly supported by chromosomal data, namely the rearrangements shown in Fig. 2. For instance, both branches show an Y/15 translocation, but the resulting chromosomes have different morphologies, suggesting that they have different origins. Alternatively, they could have undergone further rearrangements since their divergence from their most recent common ancestor or be a case of hemiplasy (Robison et al. 2008 [53]) where the ancestral population could have contained both forms in polymorphic fashion, each cytotype becoming fixed in a different branch. This fusion is an easily recurrent rearrangement, since all the species in both branches have an X/15 fusion. During meiosis the Y and the free autosome (the other 15) will be very close because of pairing with the X/15, thus increasing the possibility that any chromatin break in the Y and free 15 would result in a fusion. Also, Artibeus obscurus has the X/15 fusion only, but not Y/15. This genus is divided into two groups, large and small Artibeus. For some authors the large should be organized in the subgenus Artibeus and small in Dermanura (Redondo et al. 2008 [54]; Wilson & Reader, 2005 [55]) while others consider that these groups are distinct genera   [1]; Solari et al. 2009 [56]). All karyotypes already described for large Artibeus have a XY1Y2 system, but there is variation in Dermanura: toltecus and aztecus have a XY1Y2 system (Baker, 1973 [57]) while watsoni and phaeotis have neo-XY (Baker, 1967 [58]; Hsu et al. 1968 [59]). A. cinereus has a XY1Y2 system in Central America (Baker, 1973 [57]) and neo-XY in South America (Souza & Araujo, 1990 [45]; Noronha et al. 2010 [60]). The neo-XY in Dermanura represents a third fusion between the Y and 15.
We tested many data matrices, using both PAUP and TNT, and all the resulting trees have the same branching pattern, suggesting that Vampyressina could be a polyphyletic subtribe resulting from convergent evolution, in a similar fashion to that observed in the subfamilies of nectarivorous bats (Ribeiro et al. 2003 [61]; Sotero-Caio et al. 2013 [62]). Although some branches have low bootstrap values, the decay index values give them a good support. Whereas the bootstrap is a more quantitative test, the Bremer index is qualitative, as is chromosomal variation.

Sex chromosome evolution in the subtribe Vampyressina
Translocations between an autosome and a sex chromosome are rarely fixed in speciation because of meiotic problems, including inactivation of the homologues fused to the X chromosome (Mattei et al. 1982 [63]). Phyllostomidae is an exception to this rule where there are subfamilies with most or all species having multiple or compound sex chromosome systems (e.g., Stenodermatinae and Carolliinae -Hsu et al. 1968 [59], Tuker 1986 [64]). In this work, we present for the first time a hypothesis explaining the sex chromosome evolution in the Vampyressina and describe the events that gave rise to the sex chromosome systems observed in M. macconnelli and V. thyone.
We suggest that all species in the Vampyressina have a Neo-XY sex chromosome system and that the different sex chromosome systems as seen in Mesophylla and Vampyressa represent a derived character state (Fig. 3). In the genus Mesophylla, the sex chromosome system has evolved from a Neo-XY system after a further translocation of the compound Y with an autosome that led to a Neo-X1X2Y. To our knowledge, it has never been described before in mammals. In V. thyone, the sex chromosomes system evolved after the fission of the Fig. 3 Sex chromosome evolution in the subfamily Stenodermatinae. a The original XY system found in most mammals. b Fusion of one autosome with X, originating the XX/XY1Y2 system found, for instance, in AOB. c Neo-XY system after the fusion of the free homologue (called Y2) with Y (called Y1), found in UMA and UBI and probably the common ancestor of Vampyressa and Mesophylla. d Fission at centromere of Neo-X, freeing the autosome portion once more, originating the X1X1X2X2/X1X2Y system in VTH. e Fusion of a second autosome to Y of Neo-XY. f The resulting Neo-X1X1X2X2/X1X2Y in MMA autosome (corresponding to PHA15) that was translocated to the X chromosome in the ancestral karyotype of Stenodermatinae. As the Y-autosome translocation also involves an autosome, its free homologue would be X2.

Conclusions
Our comparative analysis of all genera of vampyressine using a cytogenomic approach confirmed the molecular phylogeny previously described for this group, being independent evidence supporting that phylogeny. Our results also confirm that all genera of this subtribe have compound sex chromosome systems that originated from an X-autosome translocation, an ancestral condition observed in the Stenodermatinae subfamily. In Mesophylla we found a sex chromosome system never described before in mammals. We were able to trace all the chromosomal sex system evolution in vampyressine and show that it mirrors the phylogeny of this group. This work presents additional data supporting Hoofer & Baker (2006 [9]) hypothesis based on molecular studies regarding the polyphyly of the genus Vampyressa and its sister relationship to Mesophylla.

Additional files
Additional file 1: Table S1.  Availability of data and materials All the data supporting your findings are contained within the manuscript.
Authors' contributions AJBG collected the samples, collaborated in all of the cytogenetic data and phylogenetic analyze, undertook the bibliographic review. CYN collaborated in all cytogenetic procedures and analyzes. LRRR collected part of the samples, collaborated with cytogenetic data and reviewed the paper. TCMB helped conceive the study and participated in developing the laboratory techniques. TFAR participated in the cytogenetic and phylogenetic analyses, and reviewed the manuscript. PCMOB and FY helped on writing the paper and discussed the data. MAFS helped on the development of the probes, participated in the cytogenetic analyses, and reviewed the manuscript. JCP helped conceive the study and participated in developing the laboratory techniques, and coordinated the writing of this paper. All authors read and approved the final manuscript.

Competing interests
The authors declare that they have no competing interests.

Consent to publish
Not applicable.

Ethics approval and consent to participate
The necessary euthanasia occurred in accordance with animal welfare guidelines established by the Animal Ethics Committee (Comitê de Ética Animal) from Universidade Federal do Pará (Permit 68-2015). JCP has a permanent field permit, number 13248 from "Instituto Chico Mendes de Conservação da Biodiversidade". The Cytogenetics Laboratory from UFPa has a special permit number 19/2003 from the Ministry of Environment for samples transport and 52/2003 for using the samples for research.