Cytogenetic markers using single-sequence probes reveal chromosomal locations of tandemly repetitive genes in scleractinian coral Acropora pruinosa

The short and similar sized chromosomes of Acropora pose a challenge for karyotyping. Conventional methods, such as staining of heterochromatic regions, provide unclear banding patterns that hamper identification of such chromosomes. In this study, we used short single-sequence probes from tandemly repetitive 5S ribosomal RNA (rRNA) and core histone coding sequences to identify specific chromosomes of Acropora pruinosa. Both the probes produced intense signals in fluorescence in situ hybridization, which distinguished chromosome pairs. The locus of the 5S rDNA probe was on chromosome 5, whereas that of core histone probe was on chromosome 8. The sequence of the 5S rDNA probe was composed largely of U1 and U2 spliceosomal small nuclear RNA (snRNA) genes and their interspacers, flanked by short sequences of the 5S rDNA. This is the first report of a tandemly repetitive linkage of snRNA and 5S rDNA sequences in Cnidaria. Based on the constructed tentative karyogram and whole genome hybridization, the longest chromosome pair (chromosome 1) was heteromorphic. The probes also hybridized effectively with chromosomes of other Acropora species and population, revealing an additional core histone gene locus. We demonstrated the applicability of short-sequence probes as chromosomal markers with potential for use across populations and species of Acropora.


Results
Karyological features and whole genome hybridization. The majority (55%) of the observed metaphase spreads (n = 100) of A. pruinosa had a chromosome number (2n) of 28 (Fig. 1a), followed by 27 (26%). Neither of the two conventional staining techniques (Giemsa and C-banding) provided a unique and clear banding pattern that could distinguish the homologous chromosomes (Fig. 1b,c). In C-banding, not all chromosomes showed a darkly stained centromeric region (Fig. 1c). On the contrary, 4' ,6-diamidino-2-phenylindole (DAPI) staining revealed constricted regions of the centromeres (Fig. 1d). Using the DAPI-stained chromosomes, their average centromere locations and individual lengths were measured, and chromosomes were arranged in order of decreasing lengths (Fig. 1d, Table 1). The centromere indices (0.54-0.57) indicated a centromeric characteristic for all the chromosomes (Table 1). Differences in chromosome lengths were not readily noticeable, in which the shortest chromosome was more than half (64.71% ± 4.3%) the size of the longest chromosome. To determine a heteromorphic pair, the size difference between each putative homologous chromosome was statistically compared (Supplementary Table S1). The size difference of the first homologous pair (chromosome 1) was found to be significantly larger than that of the other homologs (Table 1). This indicates that the first chromosome pair is heteromorphic in A. pruinosa.
To assess the locations of all repetitive loci that are readily detected by FISH, whole genome hybridization (WGH) was conducted using a probe prepared from the whole genome of A. pruinosa sperm. Results showed several faint hybridization signals on some chromosomes, but a broad and intense signal was detected at the telomeric region of the q-arm of a single chromosome (Fig. 2a). The arrangement of chromosomes according to size revealed that the intense hybridization signal was on the longer chromosome of the heteromorphic chromosome 1 (Fig. 2b). This indicates that a long and unique array of sequences was present only on this single chromosome and was absent from other chromosomes, as well as on its homologous pair. Because this hybridization pattern was observed on all metaphase spreads and across different embryos, we eliminated the possibility of allelic variation between the heteromorphic pair of chromosome 1. In addition, the location of the hybridization signal is the portion of the chromosome that is missing in its homologous pair (Fig. 2b), thus suggesting a region that may not have the function and characteristics of a locus.
Probe hybridization and sequence characterization. Hybridization of the At-p5S and At-pH2AB probes revealed readily detected single loci in separate homologous pairs (Fig. 3). The hybridization with At-p5S and At-pH2AB probes manifested as band-like and dot-like signals, respectively. This indicates that the location of At-pH2AB is clustered but may include a relatively long interspersed region between arrays, whereas that of At-p5S is broader and more contiguous. Based on the average relative sizes of the chromosomes where the hybridization signals were detected, the At-p5S loci were located on chromosome 5 and the At-pH2AB loci were on chromosome 8 ( Table 2).
Characterization of the probe sequence revealed that At-p5S is composed of small nuclear spliceosomal RNA genes (U1 and U2 snRNAs) and contains three interspacer regions (Fig. 4a). These regions were flanked by short sequences of the 5S rDNA, arranged in a head-to-tail fashion. The At-pH2AB probe is composed of two histone domains (H2A and H2B), separated by a spacer region (Fig. 4b). The two genes are arranged in a tail-to-tail fashion, which is typical among invertebrates [13][14][15] .
To confirm whether the short 5S rDNA sequence of the At-p5S probe was involved in the hybridization, we blasted the probe sequence (divided into identified regions) against the whole genome of Acropora digitifera  Table S2). Result of the analysis showed that the entire probe's length including the short 5S rDNA sequences on both ends was present and tandemly repeated. The arrangement of 5S-ITS1-U2-ITS2-U1-ITS3-5S was also highly consistent within estimated length of 423,641 bp (supplementary Table S2, highlighted in yellow). The probes prepared from A. pruinosa were tested for the chromosomes of A. muricata and A. pruinosa Kochi. Hybridization signals were effectively detected in these two Acropora chromosomes (Fig. 5). In A. muricata, the hybridization pattern was the same as observed in A. pruinosa (one homologous pair for each probe). In addition, the loci were also observed at roughly the same chromosomal position, near the centromere of the p-arm (Fig. 5a). Conversely, in A. pruinosa Kochi, the hybridization signal for At-pH2AB was detected on two homologous pairs, with additional signal that was less intense than the other (Fig. 5b). This indicates that this locus contains fewer copies of core histone gene repeats than the other. Aside from the differences in signal intensity, the chromosomal positions of the additional At-pH2AB loci slightly departed from the centromere compared with those for the other At-pH2AB loci.

Discussion
The chromosome number (2n = 28) of A. pruinosa agrees with those of other 18 species of Acropora and five other species from other coral genera (Montipora and Fungia) 7 . It is unclear whether the chromosome number 2n = 27 observed in this study was a result of missing one chromosome during mitotic preparations or it is another karyological characteristics in this coral species. Having two chromosome numbers (karyotypic mosaicism) is not uncommon in Acropora 7,9 . Acropora pruinosa Kochi was reported with chromosome numbers, 2n = 28 and 2n = 29, which was confirmed by the presence of an additional and unpaired chromosome in the case of 2n = 29 9 .
Large-scale hybridization signals on a single chromosome were observed using WGH in this study on A. pruinosa (2n = 28) as well as in a previous study on A. pruinosa Kochi (2n = 29) 9 . However, for A. pruinosa with even number of chromosomes, the presence of a unique chromosome with no apparent pair based on length and hybridization pattern might indicate the presence of heteromorphic pairs. In most animals, these heteromorphic pairs are often associated with sex chromosomes. Although the sex-linked loci and genes have been identified in the gonochoric coral Corallium rubrum 16 , the role of heteromorphic chromosomes in the sexual characteristics of scleractinians has not been explored. This investigation is particularly important in Acropora because colonies of some coral species may contain male or female polyps, aside from the well-known co-sexual polyps 17 . The www.nature.com/scientificreports/   www.nature.com/scientificreports/ heteromorphic pairs observed in this study were present in all mitotic cells, and we propose two mechanisms how these cells maintained to carry this unusually long chromosome: (1) After meiotic segregation in the hermaphroditic gonads, either the eggs or the sperms exclusively receive this chromosome, (2) a cycle that involves translocation of the portion of chromosome from the autosomes, causing the chromosome that receives it the longest one. The second mechanism has been demonstrated in other organisms, which involves translocation of the nucleolar organizer region (NOR) containing repetitive tandem arrays of 18S and 28S rRNA genes from autosomes to the telomeric end of sex chromosomes [18][19][20] . This NOR in the sex chromosomes functions in the pairing of X-Y chromosomes during meiosis 21 . This is also supported by the presence of 18/28S rDNA loci at the telomere of one of the longest chromosome pairs in A. pruinosa Kochi 9 . Further work must be conducted   www.nature.com/scientificreports/ to characterize the sequence arrays that constitute this hybridization signal on the longest chromosome and to confirm whether this chromosome is associated with functioning as a sex chromosome. The loci of the U1/U2 snRNA and core histone gene clusters showed intense hybridization signals on separate chromosome pairs. However, because the minimum sequence length of hybridization that can be readily detected in FISH is 6 kbp 15 , which is greater than the length of our probes (Table 2), it is possible that other loci composed of fewer or shorter arrays of the target genes exist. This is supported by the results of the experiment on the presence of several rDNA arrays obtained from subcloning, with shorter size of the target gene (LC557012, LC557015) that showed no hybridization signal. A sequence of similar length, but composed of indels (LC557016), compared with the identified repetitive histone array also showed no hybridization in FISH. Because these sequences were confirmed in the genome of A. pruinosa, we speculate that these arrays were either not repetitive (single-copy locus) or were short enough to be detected by FISH. Nonetheless, this study confirms the existence and chromosomal locations of highly clustered arrays of these genes. Studies have reported that this clustering of highly conserved genes is related to pseudogenes, which are acquired through hybridization of ancestral genes and have lost their coding potential 22,23 . Pseudogenes are implicated in the diversity of the nuclear ribosomal genes in Acropora, but only one rDNA sequence has been implicated to present across several species that are associated with pseudogenes 24 . It has also been reported previously that large clusters of pseudogenes consist of tRNAs and snRNAs on mammalian chromosomes 25,26 . Other identified pseudogenes that have repetitive gene copies in humans are the ribosome biogenesis protein gene (RLP24) and E3 ubiquitin-protein ligase gene (MDM2) 27 . Clustering of pseudogenes was also implicated in a mechanism to disable its function as a result of acquired mutations 28,29 . The arrangement of these genes in these clusters is tandemly repeated and lacks introns, and thus presumably arose from reverse transcription of mRNA, followed by multiple integration to specific regions in the chromosome [29][30][31] .
The linkage of snRNA genes and 5S rDNA sequence and their tandemly repetitive characteristics observed in this study was first reported for mollusks 32 . The same linkage involving U1, U2, and U5 snRNA genes was also found in fish 33 and crustaceans involving only U1 snRNA 34 . Here, we report for the first time a tandemly repetitive linkage of 5S rDNA sequence and snRNA genes in the phylum Cnidaria. Although many FISH studies of single or multiple loci of repetitive 5S rRNA genes [35][36][37] and snRNA genes 38,39 have been reported, it is uncertain whether the loci observed in these studies may involve linkage to one another or to any other gene. We showed that repetitive linkage of snRNA and putative 5S rRNA genes produced a single locus on the chromosomes. Conversely, in fish, the loci of these two repetitive genes were not linked and were located on different chromosomes 40 .
Only the H2A and H2B genes arranged in a typical manner were confirmed to constitute the observed loci. However, in cnidarians, various arrangements of repetitive core histone genes, including H1, H3, and H4, have been documented 12 . In Mytilus edulis, aside from the core histone genes, the sequence of the solitary linker H1 gene is also tandemly repeated 41,42 . The loci of these solitary H1 gene clusters were found to be located on chromosome pairs different from core histone genes 43 . This suggests the possible presence of other repetitive histone loci that can be observed in scleractinian chromosomes. Surprisingly, a unique arrangement of repetitive arrays involving linkage between histone and 5S rRNA genes was observed among crustaceans 44 and fish 45,46 .
The varying hybridization patterns of core histone probes in other Acropora population might suggest chromosomal rearrangements during the evolutionary processes within Acropora. In the genus Mus, locations of clusters of conserved genes are shifted across different chromosomes, providing evidence of genome reshuffling that occurred during its evolution 47 . Variations in the number of histone loci within closely related taxonomic groups have also been observed in other taxa. In bivalves, loci of histone genes are in two chromosome pairs in the mussel, Mytilus galloprovinciali 48 , and in the scallop, Patinopecten yessoensis 49 , but there is only one locus in the mussel species, Perumytilus purpuratus 50 and in three other species of scallops (Argopecten irradians, Chlamys farreri, and C. nobilis) 49 .
We demonstrated that single-sequence probes containing conserved genes produced a readily detectable hybridization signal on the chromosomes of A. pruinosa. These probes also hybridized on chromosomes of other Acropora population and species and thus have a potential for use as chromosomal markers within the taxa. In www.nature.com/scientificreports/ addition, the single-sequence probes revealed the presence of other loci in other species, which revealed the differences in chromosome organization. This study may provide a foundation for discovering the loci of other tandemly repetitive genes, such as 18 and 28S rDNA that can be used as additional chromosomal markers for improved karyotyping of Acropora.

Methods
Sample collection and chromosome preparation. Embryos of A. pruinosa were obtained from artificial fertilization of egg-sperm bundles collected from spawning colonies in Kaiyo-Cho, Tokushima, Japan (33.545°N, 134.315°E) (Fig. 6a) on the night of July 20, 2019. The coral is characterized by indeterminate colony outline (Fig. 6b), with appressed and tubular radial corallites (Fig. 6c) 51 . Embryos were grown in 0.2 µm filtered seawater for 10-14 h and treated with 0.01% (v/v) colchicine followed by the addition of hypotonic solution (seawater: dH 2 0 = 1:1). Other coral embryos used in this study were preserved ones such as Acropora muricata collected also in Kaiyo-Cho and another Acropora pruinosa collected in Otsuki, Kochi, Japan (32.777°N, 132.731°E).
To distinguish A. pruinosa collected in Otsuki, Kochi, Japan, the name A. pruinosa Kochi was used throughout this study. Chromosomes were prepared from the embryos based on the method described by Taguchi et al. 8 , with slight modifications. About 30-50 embryos were collected by centrifugation and 0.5 mL of Carnoy's fixative (absolute methanol:glacial acetic acid = 3:1) was added. Lipids were removed by soaking the embryos in diethyl ether for 4-6 h. Cells were centrifuged at 2000 × g for 2 min and then resuspended in 0.5 mL of Carnoy's fixative. A drop of cell suspension was placed on a slide and then flame-dried. For G-banding, slides were treated with 0.025% trypsin solution for 1 min, and then stained with Giemsa solution diluted with 5% 0.06 M phosphate buffer (pH 6.8). To examine the chromosomal distribution of constitutive heterochromatin, C-banding was performed using the standard barium hydroxide/saline/Giemsa method 52 with slight modifications. Chromosome slides were treated with 0.2 N HCl at 25 °C for 30 min and then with 5% Ba(OH) 2 at 50 °C for 1 min. The slides were then soaked in 2X SSC at 60 °C for 30 min. Experimental research, including the collection of the coral bundles, complied with the relevant institutional, national, and international guidelines and legislation.   Figure S1). The PCR products were ligated into a bacterial plasmid using the pGEM-T Easy Vector Systems (Promega, USA) and transformed into JM109 competent cells (Promega, USA). The cells were then spread plated onto Luria broth (LB) plates containing 100 mg/mL of ampicillin, 40 mg/mL of 5-bromo-4-chloro-3indolyl-β-D-galactoside (X-Gal), and 0.05 mmol/L isopropyl-β-D-thio-galacto-pyranoside (IPTG). The plates were incubated for 15 h, and bacterial colonies were screened for positive inserts using colony PCR followed by gel electrophoresis. Positive colonies were grown in LB medium for 15 h and plasmids were extracted thereafter using Mini Plus Plasmid DNA (Viogene, USA). The inserts that were positive in FISH screening were sequenced with M13 universal primers using the ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit ver.2.0 (PE Biosystems, Japan). Primer walking was conducted for insert sizes greater than 1 kbp. The sequence reads were checked, assembled, and vector sequences were removed manually using MEGA X 54 . DNA sequences were submitted to the DNA Data Bank of Japan (DDBJ) with accession numbers LC557012-LC557016.
Probe preparation and FISH. FISH probes were prepared from the plasmid DNA using the Random Primed DNA Labeling Kit (Roche, USA) according to the manufacturer's protocol. The DNA was fluorescently labeled directly using cyanine-3-dUTP (Cy3-dUTP) (PerkinElmer, USA) or indirectly using digoxigenin-dUTP (DIG-dUTP)/anti-Digoxigenin-FITC (Roche, USA) at 37 °C for 15-18 h. The probe obtained using 5S rDNA sequence as the target was named At-p5S, whereas that obtained from histone was named At-pH2AB. FISH was performed according to the method described by Taguchi et al 9 , with slight modifications. Slides of A. pruinosa chromosomes were denatured in 70% formamide solution at 70 °C for 2 min and then serially submerged in ice-cold 70%, 90%, and 99% EtOH for a total of 6 min. About 1 µL of DNA probes were mixed with 10 µL hybridization solution (H7782, Sigma, Japan) and then denatured at 80 °C for 1 h. For whole genome hybridization experiment, probes were then incubated at 37 °C for 1 h to allow pre-annealing of simple tandem repeats (i.e. G-C repeats). This is to minimize the hybridization signals and reveal clusters composed of high-complexity sequences. The slides with denatured chromosomes were incubated with the probe solution at 37 °C for 12-15 h to allow hybridization. Post hybridization washing was performed with 50% formamide at 43 °C for 20 min and subsequently with 2X SSC at 37 °C for 8 min. The slides were incubated twice in 1X phosphate-buffered detergent (PBD) at 25 °C for 5 min. The chromosomes were then counterstained with DAPI-Vectashield (Vector Laboratories, USA) and viewed under an AxioImager A2 fluorescence microscope equipped with an Axiocam MRm CCD camera (Zeiss, Germany). Images of suitable metaphase spreads from different embryos were captured using the AxioVision software (Zeiss). The FISH images were analyzed by measuring the chromosome lengths and hybridization signal locations using the DRAWID software 55 . Centromere indices (long arm/total length) was computed based on the formula of Lucas and Gray 56 .