Novel insights into chromosome evolution of Charadriiformes: extensive genomic reshuffling in the wattled jacana (Jacana jacana, Charadriiformes, Jacanidae)

Abstract The order Charadriiformes comprises three major clades: Lari and Scolopaci as sister group to Charadrii. Until now, only three Charadriiformes species have been studied by chromosome painting: Larus argentatus (Lari), Burhinus oedicnemus and Vanellus chilensis (Charadrii). Hence, there is a lack of information concerning the third clade, Scolapaci. Based on this, and to gain a better understanding of karyotype evolution in the order Charadriiformes, we applied conventional and molecular cytogenetic approaches in a species belonging to clade Scolopaci - the wattled jacana (Jacana jacana) - using Gallus gallus and Zenaida auriculata chromosome-specific probes. Cross-species evaluation of J. jacana chromosomes shows extensive genomic reshuffling within macrochromosomes during evolution, with multiple fission and fusion events, although the diploid number remains at high level (2n=82). Interestingly, this species does not have the GGA7-8 fusion, which was found in two representatives of Charadrii clade, reinforcing the idea that this fusion may be exclusive to the Charadrii clade. In addition, it is shown that the chromosome evolution in Charadriiformes is complex and resulted in species with typical and atypical karyotypes. The karyotypic features of Scolopaci are very different from those of Charadrii and Lari, indicating that after divergence, each suborder has undergone different chromosome rearrangements.

Comparative chromosome painting has contributed to the reconstruction of the evolutionary chromosomal history of birds Furo et al., 2015;Kretschmer et al., 2015aKretschmer et al., , 2018a. Since the first production of chromosome-specific probes for Gallus gallus (GGA; Galliformes) in 1999 (Griffin et al., 1999), other species have been chosen for the same purpose, such as Burhinus oedicnemus (Charadriiformes) (Nie et al., 2009), Leucopternis albicollis (Accipitriformes) (de Oliveira et al., 2010), and more recently, Zenaida auriculata (ZAU) (Columbiformes) (Kretschmer et al., 2018b). ZAU has the same organization of macrochromosomes as proposed for the avian putative ancestral karyotype (PAK)  and is similar to Gallus gallus (the only difference being the fusion of PAK4 with PAK10, forming GGA4p and GGA4q, respectively). Furthermore, ZAU probes have the advantage over GGA probes in cross-species hybridization as they tend to produce stronger hybridization signals than Gallus gallus probes in Neognathae (Kretschmer et al., 2018b), probably because the divergence between ZAU and other Neognathae species is more recent than chicken.
The wattled jacana (Jacana jacana) belongs to clade Scolopaci, and its karyotype has not been reported yet. It is an interesting species with a polyandrous mating system, in which a single female defends a harem of up to four males by aggressively excluding other females from their territory; males provide nearly all parental care (Osborne et al., 1977;Emlen and Wrege, 2004). The purpose of this study was to perform comparative chromosome painting using whole chromosome probes of Gallus gallus and Zenaida auriculata in order to test: i) if a species of the clade Scolopaci has a chromosomal organization similar or different to the species of the clade Charadrii or Lari, and ii) if the karyotype of the Jacana jacana is reorganized by interchromosomal rearrangements, despite maintaining a conserved chromosomal morphology and diploid number.

Material and Methods
Sampling Two males and two females of the Jacana jacana were sampled from São Gabriel, Rio Grande do Sul/Brazil. The collection and analyses were developed in agreement with SISBIO 44173-1 and Comissão de Ética no Uso de Animais-CEUA 018/2014 authorization.

Acquisition of mitotic cells
Mitotic cells were obtained from fibroblast culture according to Sasaki et al. (1968). Briefly, a small skin sample of each specimen was collected and incubated in 2 ml of collagenase type IV (0.5%) for one hour at 37 ºC. The resulting cell suspension was washed in 5 mL of Dulbecco's Modified Eagle Medium (DMEM) and centrifuged for 10 min at 800 rpm. Afterwards, the supernatant was removed and the cells were resuspended in 5 ml of DMEM, supplemented with 10% fetal calf serum (FCS) and 1% Penicillin-Streptomycin (10,000 U/mL), and incubated at 37 o C. To arrest cells in metaphase, cultures were treated with 0.016% colchicine for 1 hour. After hypotonic treatment in 0.075 M KCL (15 minutes), cells were fixed in methanol:glacial acetic acid (3:1) and dropped onto clean slides.

Chromosomal analysis
To determine the 2n and chromosome morphology, chromosome preparations were stained with 5% Giemsa solution in 0.07 M phosphate buffer (pH 6.8), and 40 metaphase spreads were analyzed for each specimen. The chromosome measurements for macrochromosomes (14 pairs) and sex chromosomes Z and W were made using ImageJ software (Rueden et al., 2017). The remaining chromosomes were not measured because of their very small size, being considered microchromosomes. Karyotypes were arranged according to chromosome size and centromere position. Heterochromatic regions and W chromosome were identified by CBG banding performed according to Sumner (1972).

Fluorescence in situ hybridization (FISH)
FISH experiments were performed using whole chromosome probes of Zenaida auriculata (ZAU1-5, which are homologous to GGA1-3,4q and 5, respectively) (Kretschmer et al., 2018b) and G. gallus (GGA4, 6-16). Chromosome probes GGA15 and GGA16 were present in the same pool (GGA R5). Both sets of probes were labeled with biotin-dUTP or digoxigenin-dUTP, and the hybridizations were performed according to de Oliveira et al. (2010). After 3 days of hybridization at 37°C, biotin-labeled probes were visualized using a layer of Cy3-streptavidin and digoxigenin-labeled probes with sheep anti-digoxigenin FITC coupled antibodes. After detection, chromosomes were counterstained with DAPI and examined by fluorescence microscopy.
Although we used ZAU1-5 probes, all karyotype comparisons were made with their homology with G. gallus, since most efforts to map the genome of birds have concentrated on the chicken, including the homology to the putative ancestral avian karyotype.

Karyotype description
Jacana jacana has 2n=82 chromosomes (Fig. 1). The karyotype comprises 14 pairs of biarmed macrochromosomes (including the sex chromosomes), the first pair being markedly larger than the other autosomes (Table 1). Pairs 2 to 8 are similar in size, as are the ninth to the eleventh. The latter (11th) can be distinguished from the others by a secondary constriction, containing the rDNA sequen-ces. The Z chromosome is submetacentric and its size is equivalent to the autosomal pair 2. The W chromosome is acrocentric and it is smaller than Z.

CBG banding analysis
Sequential analysis of the same metaphase with Giemsa staining and C-banding confirms the identification of the W chromosome (Figure 2). The W chromosome is almost completely heterochromatic, whereas the autosomes do not show heterochromatic blocks.

Comparative chromosome painting
Cross-species chromosome painting results show extensive genomic rearrangement of the ancestral chromosomes in Jacana jacana karyotype (JJA, Fig. 3 and 4). The fissions of GGA2 homologous chromosomes (4 pairs Figure 4 shows the homology between the GGA1-16 and JJA.

Discussion
Given that approximately 63% of birds have 2n=74-86 (Christidis, 1990), the chromosome organization    of Jacana jacana resembles the typical avian karyotype at a first glimpse, with 2n=82 and comprising few macrochromosomes pairs (15 pairs, including Z and W chromosomes) and numerous microchromosomes (26 pairs). However, the chromosome morphology and size observed (Fig. 1) is different from other bird species, especially paleognathous birds, and even from species of Charadriiformes (Nishida-Umehara et al., 2007;Hansmann et al., 2009;Nie et al., 2009;Kretschmer et al., 2015b). The main difference is in the large number of medium biarmed chromosomes observed, which may be a result of chromosomal fusions or intrachromosomal rearrangements (centromere repositioning or inversions).
Chromosome painting indicates that J. jacana karyotype underwent extensive reorganization during its evolution from the PAK, mainly involving macrochromosomes. Hence, although pair 1 was entirely homologous to GGA1, pairs 2 to 6 (GGA2-6) show many fissions, sometimes in more than one segment. For instance, the ancestral chromosome 2 (GGA2) is split between 4 pairs and the ancestral 3 (GGA3) between three pairs, while ancestral pairs 4-6 (GGA4q-6) are split between two pairs each. However, despite the high frequency of chromosome fission in this species, the diploid number has been maintained close to the ancestral karyotype (2n=80) , due to several chromosomal fusions (Fig 1, 3 and Table 2). The associations between parts of ancestral chromosome 2 (GGA2) and 8 (GGA8) in JJA6 and between parts of ancestral 3 (GGA3) and 7 (GGA7) in JJA7 have not been described previously in bird species (reviewed in Kretschmer et al., 2018a). These results highlight the value of chromosome painting in evolutionary cytogenetics, since a conserved diploid number does not always represent a conserved karyotype. It would be of interest to analyze if Charadriiformes species with high diploid numbers, but without fusions, have the same breakpoints as J. jacana, since some studies indicate that breakpoints recur in different avian lineages (Volker et al., 2010;Warren et al., 2010;Skinner and Griffin, 2012;Kretschmer et al., 2018b).
Most of the microchromosomes are fully conserved in J. jacana. For instance, the use of the GGA1-16 chromosome probes showed the existence of only one gap in the JJA8 chromosome. Considering that 22 pairs of chicken microchromosomes (GGA17-38) were not studied, and 20 pairs of microchromosomes were left unhybridized by any of the probes used in J. jacana, it seem possible that at least two pairs of microchromosomes are involved in fusion events, one of them occurring in JJA8, which would correspond to GGA5 fused to a microchromosome.
This study on J. jacana is the first in which a representative of the clade Scolopaci has been analyzed by chromosome painting (Figure 3 and 4). In Charadriiformes, comparative chromosome mapping data are available for B. oedicnemus (Nie et al., 2009) and V. chilensis (Kretschmer et al., 2015b), both belonging to the Charadrii, and L. argentatus, belonging to the Lari (Baker et al., 2007). A comparison of the chromosome painting data obtained to date shows that each species has a different genome organization (Table 2). Based on this, and considering the diploid number variation from 2n=42 in B. oedicnemus (Nie et al., 2009) to 2n=98 in Gallinago gallinago (Hammar, 1970), we can assume that the order Charadriiformes underwent a unique chromosomal evolution. For instance, V. chilensis (2=78) can be considered a species with a more basal karyotype, since it has a relatively conserved diploid number, differing by only a fusion of the ancestral chromosomes 7 (GGA7) and 8 (GGA8) (Kretschmer et al., 2015b). On the other hand, L. argentatus has a lower diploid num- ber (2n=70) (Hansmann et al., 2009), mainly due to the occurrence of fusions of macrochromosomes (GGA5-9) with microchromosomes. In contrast, B. oedicnemus shows a diploid number considered to be extremely low for the class Aves (2n=42). According to Nie et al. (2009), this divergence occurred through multiple fusions involving both microchromosome-microchromosome and microchromosome-macrochromosome fusion events, such as in GGA9, 2, 4p, 6 ( Table 2) and macrochromosome-macrochromosome fusions without any identified chromosome fission (Nie et al., 2009). In J. jacana, we can see a unique karyotype, resulting from fissions of the macrochromosomes, not previously reported for the order Charadriiformes, and fusions between the segments resulting from these fissions. Unfortunately, there are no chromosome-painting data for other species belonging to the Scolopaci, but it is possible that the associations observed in J. jacana are exclusive to the genus or even to the Scolopaci, since they were not observed in B. oedicnemus, V. chilensis (Charadrii), or L. argentatus (Lari). In addition, since J. jacana and L. argentatus do not have the fusion between ancestral chromosomes 7 and 8, the hypothesis that this fusion is an exclusive characteristic of the clade Charadrii is reinforced (Kretschmer et al., 2015b).
Considering that Vanellus chilensis, a member of clade Charadrii, which is placed in a basal position in the Charadriiformes phylogeny (Baker et al., 2007), has a conserved karyotype (Kretschmer et al., 2015b), we propose that the last common ancestor for the three suborders (Lari, Scolopaci and Charadrii) had a conserved karyotype, and after divergence, each suborder underwent different chromosomal rearrangements ( Fig. 5 and Table 2). For instance, considering available results of chromosome painting, we can assume that genome organization in the suborder Lari has involved mainly fusion between microchromosomes and macrochromosomes, while in the suborder Scolopaci the main rearrangements were fissions and fusions between macrochromosomes. On the other hand, in the suborder Charadrii we see many fusions between microchromosomes, as observed in B. oedicnemus, or conserved karyotypes, such as observed in V. chilensis. Despite the fact that cytogenetic data in this group are still fragmental, they do not corroborate the proposal based on mitochondrial and nuclear sequences, which places the genus Burhinus, which has a highly diverged karyotype in a basal position, as sister group to the rest of the Charadrii, including Vanellus (Baker et al., 2007) with a conserved karyotype.
In conclusion, despite the existence of chromosomepainting data for only a few species of the order Chara- 6 Kretschmer et al.