The Strigamia maritima karyotype is 2n = 16

A previous unpublished description of the Strigamia karyotype had shown that 2n = 16 in a male specimen (Pawel Woźnicki, personal communication). We confirmed the Strigamia karyotype using mitotic spreads prepared from embryos of unknown sex (Fig 1A). All embryos examined (10/10) reproduced the same karyotype, and none of them showed evidence for a heteromorphic chromosome pair. Therefore, there was no prima facie evidence for a pair of heteromorphic chromosomes that might be the sex chromosomes.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The karyotype of Strigamia maritima.

(A) Inverted image of a DAPI-stained mitotic metaphase prepared from embryonic material, of which the sex was unknown (left panel). The constructed karyotype (right panel) is a representative example of karyotypes prepared from at least 10 other embryos. It shows a large pair of metacentric chromosomes and seven middle to small acrocentric pairs, gradually decreasing in size. All karyotypes give qualitatively the same result and none of them show evidence for a heteromorphic sex chromosome pair. Scale bar = 10 μm. (B) Comparative genomic hybridization (CGH) on a spermatogonial mitotic metaphase prepared from the testes of a sub-adult male (left panel). Chromosomes were counterstained with DAPI (blue). Female-derived genomic probe was labelled with fluorescein-12-dUTP (green) and male-derived genomic probe with Cy3-dUTP (red). Both probes highlighted one arm of the two large metacentric chromosomes (arrows) and the centromeric heterochromatin of all chromosomes (asterisks), but did not differentiate a sex chromosome pair as demonstrated in the constructed karyotype (right panel). Scale bar = 10 μm.

https://doi.org/10.1371/journal.pone.0150292.g001

We also performed comparative genomic hybridization (CGH) on spread preparations from the testes of sub-adult males. This method can differentiate X and Y, or W and Z, sex chromosomes provided that they differ sufficiently in their DNA composition [9]. However, even with CGH, we were unable to identify the sex chromosomes in mitotic spermatogonia (Fig 1B) and pachytene spermatocytes (S1 Fig).

Nevertheless, we had genomic evidence for a population of underrepresented sequences present specifically in males [1]. Therefore we decided to examine the genomic data further, and to identify a set of candidate sex chromosome-derived genomic scaffolds for validation.

Validation of three X-linked scaffolds

We selected for validation three of the longest scaffolds underrepresented in males relative to females (scf7180001248200, scf7180001248049 and scf7180001247190). After confirming that these scaffolds generated a PCR product in both male and female genomic DNA samples, we next tested these scaffolds for differences in copy number between the sexes using quantitative real-time PCR (qPCR). The expectation is that any sequences located on an X chromosome will be present at twice the copy number in females (XX) compared with males (XY) [10].

The qPCR confirmed the expectations from the analysis of scaffold coverage. We find clear evidence that at least these three underrepresented scaffolds are present at half the copy number in males compared with females (Fig 2B; scaffolds also indicated on Fig 2A with arrowheads). This is most likely due to the location of these scaffolds on an X chromosome in an XX/XY sex determination system. A second region on the reference Hox scaffold was tested as a control. This control region was located more than 240 kb away from the region against which the reference primers were designed. As expected, this control region showed no significant difference in abundance between male and female individuals (Fig 2B).

Validation of six male-specific scaffolds

We had an initial set of 24 candidate Y-linked scaffolds to validate. To try and avoid the problem of repetitive elements, we searched for sequences in these scaffolds that were unique within the assembled genome—i.e. regions of sequence that only return hits to their constituent scaffold, and not to other scaffolds in the genome (using a sliding window of 1 kb, see Methods). Only 7 of the 24 scaffolds contained any unique regions at all. The remainder contained repetitive sequences within every 1 kb window. We designed PCR primers within the unique regions of 6 of these 7 scaffolds. For these 6 scaffolds, the primer sets amplified a PCR product only from male genomic DNA (Fig 2C; scaffolds also indicated on Fig 2A with arrows). To control for the quality of the female genomic DNA, we designed PCR primers against the scaffold containing the Hox cluster, which shows no significant difference in coverage between the sexes, and confirmed that these primers amplify a product in both sexes (Fig 2C). Therefore, we have found at least 6 male-specific scaffolds in the Strigamia genome. This is most likely due to the location of these scaffolds on a Y chromosome.

Note that even those few "unique" male-specific fragments identified in the above assay must contain short repeats derived from repetitive regions that are not present in the final genome assembly. When we tried to use these probes for in situ hybridization (see further comment below), they hybridized at many sites in the genome. However, the paired PCR priming site sequences themselves must be present only on the Y chromosome.

Cytological identification of the Strigamia XY chromosome pair

Having found genomic and molecular evidence for a set of X and Y chromosome-derived scaffolds, we wanted to confirm the existence of X and Y chromosomes cytologically. To achieve this, we performed fluorescent in situ hybridization (FISH) against chromosome spreads using X- and Y-chromosome derived DNA probes. We attempted to isolate mitotic spreads from adult centipedes of known sex, but mitoses were rare in the adults available, and difficult to prepare. We therefore used embryos, which contained abundant mitoses, but were of unknown sex, and male testes, which contained meiotic chromosomes.

Attempts to use the short Y-specific sequences as probes were unsuccessful; these were too short to give any clear single copy in situ signal above a dispersed background of hybridization.

We were, though, able to design a set of X-linked probes that contained minimal repeat sequence, so as to minimize hybridization to other, non-sex-linked regions of the genome. Using a sliding window of 10 kb, we searched the three validated X-linked scaffolds for unique regions. Having found long stretches of unique sequence in each X-linked scaffold, we designed multiple primer pairs in these regions in order to generate PCR fragments spanning the unique regions. In total, we successfully amplified five PCR fragments, two on scf7180001248200 and three on scf7180001248049, covering a total of 54.8 kb of non-consecutive sequence. The relative sizes and genomic positions of these fragments are shown in Fig 3A. These PCR fragments were labelled with digoxigenin and used as a cocktail of X-linked probes for FISH. We performed FISH against mitotic metaphase spreads prepared from single embryos (of unknown sex) and against meiotic spreads prepared from the testes of sub-adult males.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Identification of the X chromosome in the Strigamia karyotype by FISH with a set of X-chromosome derived DNA probes.

(A) The relative positions and sizes of the five PCR fragments, distributed on two different X-linked scaffolds, which were used to produce the X-probes. Two of the fragments are located on scf7180001248200 and three on scf7180001248049. The genomic distance between these scaffolds is not known. The black line depicts the scaffold. The numbers above the line indicate the number of bases, starting at 1 on the left hand side. Green boxes represent the length and relative position of the PCR products, numbered arbitrarily from 1 to 5. (B) A mitotic metaphase chromosome spread prepared from a single embryo. Hybridization signals of the X-probes identify a middle-sized element in the Strigamia karyotype as the X chromosome. As there are two chromosomes with the X-probe signals, we infer that this chromosome spread is derived from a female embryo (XX). (C) Two Strigamia karyotypes constructed from the mitotic metaphases of embryonic cells. They are derived from different embryos. Upper panel: karyotype derived from the female metaphase shown in (B). Lower panel: karyotype derived from an inverted image of a DAPI-stained metaphase of unknown sex. It is the same as that shown in Fig 1A. We infer that the pair of sex chromosomes represents the 4^th pair of chromosomes by size (asterisks). (D, E, F) Meiotic chromosome spreads, prepared from sub-adult male testes. (D) Late zygotene complement showing a clump of incompletely paired bivalents. The X-probes label the longer chromosome of a partially paired bivalent, as schematically illustrated in (D’). We thus infer that this is the X chromosome, and that the other shorter chromosome, without hybridization signals, is the Y chromosome. The X and Y chromosomes are only paired at the distal part of the X chromosome, with a large proximal part unpaired. (E) A particularly clear and well-spread XY bivalent at a similar stage to (D). It shows hybridization signals of X-probes on the unpaired proximal part of the X chromosome, while the Y chromosome is completely paired except for the DAPI-highlighted centromere (see schematic drawing below the XY bivalent). (F) Pachytene complement showing 8 bivalents, each with DAPI-highlighted centromeric chromatin. X-probe hybridization signals are visible on the unpaired segment of the longer chromosome, near the centromere (see schematic drawing on the right-hand side). The X and Y chromosomes now appear almost equal in length in the bivalent. Scale bar is equal to 5 μm in (B) and 10 μm in (D, E, F). Chromosomes were counterstained with DAPI (blue). Arrowheads indicate hybridization signals of the digoxigenin-labelled X-probes (green); arrows indicate a pair of the largest chromosomes (B) or the largest bivalent (F).

https://doi.org/10.1371/journal.pone.0150292.g003

We identified a number of mitotic spreads with positive signals on a pair of middle-sized chromosomes (Fig 3B; arrowheads). We infer that this is a pair of X chromosomes, and thus that the spread is derived from a female embryo. We ordered the FISH-labelled chromosomes by size, and compared this with a previously constructed mitotic karyotype, derived from a different embryo (same karyotype as shown in Fig 1A). Through this comparison, we infer that the X chromosomes represent the 4^th pair of chromosomes by size (out of eight pairs of chromosomes).

Meiotic spreads prepared from sub-adult male testes contain late zygotene/pachytene stages, in which the chromosomes are lining up with each other to form homologous chromosome pairs, or bivalents. If Strigamia males possess an X and Y chromosome, the expectation is that only one chromosome of the sex chromosome bivalent, the X chromosome, should be labelled with the X-probes.

The meiotic spreads that we have examined match this expectation. We have identified a number of meiotic prophase I nuclei in which, in one of the eight bivalents, positive signals are located on only one of the two chromosomes (Fig 3D–3F). In some zygotene/early pachytene complements, in which the chromosomes are incompletely paired, a middle-sized bivalent can be seen in which the two chromosomes are of unequal length. The X-probes label an unpaired segment of the longer chromosome, the presumptive X chromosome (Fig 3D and 3E). We infer that the other shorter chromosome, without hybridization signals, is the Y chromosome. A particularly clear example is shown in Fig 3E, where the XY bivalent is well spread and separated from the other bivalents. It can be seen that X and Y chromosomes are only paired along part of the X chromosome. A large part of the X chromosome remains unpaired, while it appears that the Y chromosome is completely paired except for the terminal centromeric heterochromatin.

In presumably later pachytene complements, we find an approximate equalization of the lengths of the two sex chromosomes in the XY bivalent (Fig 3F). In support of our inference that this represents the XY bivalent, we note that similar phenomena—delayed pairing and length equalization between sex chromosomes of different sizes—is known to occur in a number of other species [11, 12].

Note that it is not surprising to find that the X and Y chromosomes are visibly different in length during the pachytene stage of meiosis, but that such size differences are not detectable in the mitotic karyotypes of these very small chromosomes (compare Fig 3B, 3C and 3D, 3E). This is because pachytene chromosomes are significantly de-condensed and can appear much longer than the highly compact mitotic chromosomes [13, 14]. This means that, within the resolution limit of conventional fluorescence microscopy, differences in size are much easier to detect at this stage.

Genomic DNA extraction

Adult centipedes were collected from the field at a stretch of shingle bank beach near Brora, northeastern Scotland (coordinates: 57°59'N, 03°55'W) as described previously [31]. Once returned to the lab, individual adults were sexed according to morphological criteria [8, 32]. The final leg-bearing segment and the genital segments were removed by dissection from female adults. This was to ensure complete removal of the spermatheca from female specimens, to avoid any contamination with male DNA. Adults were snap frozen in liquid nitrogen and stored at -80°C until processing. Genomic DNA (gDNA) was extracted from male and female specimens with the Genomic DNA Buffer Set and Genomic-tip 20/G (Qiagen) according to the manufacturer’s instructions.

Identification of candidate sex-linked scaffolds

The Strigamia maritima reference genome (release Smar_1.0) is available at www.ncbi.nlm.nih.gov/assembly/322118/. The genome was assembled from a mixture of both male and female samples. The reads for the re-sequenced individuals used to identify the candidate sex chromosome-derived scaffolds are stored at the NCBI Sequence Read Archive (females A and B—SRA accessions SRX326837 and SRX326839, respectively—and male J—SRA accession SRX326841). The mapping of reads back to the Strigamia reference genome was completed by the Strigamia genome project [1]. For each re-sequenced individual, the coverage for each scaffold was calculated as the number of reads that mapped back to that scaffold in the reference genome, multiplied by the length of the sequence reads, and divided by total scaffold length. Individuals A and B had sequence reads of 95 bp, and individual J of 125 bp. Finally, for each individual, scaffold coverage was normalized to the mean coverage across all scaffolds. The ratios of scaffold coverage in different individuals were calculated, and the plots generated, in Microsoft Excel. The raw data is provided in S2 Table.

Quantitative real-time PCR

Relative quantification of genomic scaffolds between the sexes was performed on a 480 LightCycler (Roche) using SYBR Green I assays. The abundance of each candidate scaffold was normalized against the abundance of the Hox scaffold (scf7180001247533). The scaffold containing the Hox cluster was selected as the reference scaffold because it is a long scaffold that shows no significant difference in sequence coverage between the sexes, and so is almost certainly located on an autosome. Primers were designed with the Primer3 freeware with annealing temperatures of approximately 60°C and product sizes of about 200 bp [33]. Scaffold IDs and primer sequences are provided in S3 Table. Each reaction was carried out in triplicate, with a final concentration of each primer at 0.5 μM and total input of gDNA at 0.8 ng. The fluorescence crossing points (C_T values) were estimated with the second derivative maximum method using the LightCycler 1.5 software. PCR amplification efficiency was determined with a calibration curve for each primer pair, and the relative quantifications were carried out with amplification efficiency correction. Only PCR assays producing a single product (as verified by melting curve analysis) and with a PCR efficiency between 1.8 and 2.1 were included in subsequent analysis. The raw qPCR data is provided in S2 Table.

Validation of Y-linked scaffolds

Candidate Y-linked scaffolds were validated by PCR on male and female genomic DNA samples, using ThermoPrime Plus DNA polymerase (Thermo Scientific). Double-distilled water was used as a template in the no template control. Primers were designed against unique regions of the candidate scaffold. Primer sequences are provided in S3 Table. Regions of each scaffold that are unique in the assembled genome were identified by a sliding window analysis: windows of 1 kb of sequence, at 0.5 kb offsets, were searched against the Strigamia reference genome using the NCBI program BLAST [34]. Windows that only returned hits to their constituent scaffold were selected for primer design. PCR products were visualized on 1% agarose gel with 0.1 ng/μl ethidium bromide. PCRs were run under the following conditions: an initial denaturation step of 95°C for 5 min; followed by 35 cycles of 95°C for 30 s, 58°C for 30 s, 72°C for 45 s; and finally an extension step of 72°C for 5 min.

Production and digoxigenin labelling of the X-chromosome derived DNA probes

Unique regions on the validated X-linked scaffolds were identified by a sliding window analysis: 10 kb windows were searched against the Strigamia reference genome using the NCBI program BLAST [34]. Windows that only returned hits to their constituent scaffold were selected for primer design. PCR fragments were amplified using the Expand Long Template PCR System (Roche), according to the manufacturer’s instructions. Primer sequences are provided in S3 Table. PCR fragments were labelled by digoxigenin-11-dUTP (Roche) using the Nick Translation Kit (Abbott Molecular). The 25 μL labelling reaction contained 500 ng of purified PCR product, 1x nick translation buffer, 250 μM dATP, dCTP, and dGTP, 90 μM dTTP, 160 μM digoxigenin-11-dUTP and 5 μL of nick translation enzyme mix. The reaction mix was incubated for 7 h at 15°C. Each PCR fragment was labelled separately.

Preparation of mitotic metaphase chromosome spreads from single Strigamia embryos

Individual eggs at embryonic stage 4–5 [for staging, see [31]] were partly dechorionated using fine tweezers in a physiological solution, and transferred into a hypotonic solution (0.065 M KCl) for 15–20 min. The eggs were then fixed in methanol/acetic acid (3:1) for 10 min. During fixation each embryo was removed from the remaining chorion and torn into about four pieces. Each piece of the embryo was transferred into a 1.5 μL drop of 60% acetic acid on a slide and spread on the slide using a heating plate at 55°C. For chromosome counts and karyotype analyses, some preparations were directly stained with 0.5 μg/mL DAPI (49,6-diamidino-2-phenylindole; Sigma-Aldrich) in antifade based on DABCO (1,4-diazabicyclo[2.2.2]octane; Sigma-Aldrich). The other preparations were passed through a graded ethanol series (70%, 80%, and 100%, 30 s each) and stored at −20°C until further use.

Preparation of meiotic chromosome spreads

Meiotic chromosomes were obtained from the testes of sub-adult males as described in [35]. Briefly, after dissection in a physiological solution, the testes were incubated in a hypotonic solution (0.075 M KCl) for 10 min, fixed in Carnoy fixative (ethanol/chloroform/acetic acid, 6:3:1) for 10–15 min, dissociated with tungsten needles in a drop of 60% acetic acid and spread on the slide using a heating plate at 45°C. Then the preparations were passed through a graded ethanol series (see above) and stored at −80°C until further use.

Comparative Genomic Hybridization (CGH)

Genomic probes were prepared from gDNA extracted separately from adult males and females by the standard phenol-chloroform procedure. The probes were labelled using a Nick Translation Kit (Abbott Molecular Inc.); male DNA with Cy3-dUTP (GE Healthcare) and female DNA with fluorescein-12-dUTP (Invitrogen). Unlabelled female gDNA, used as a species-specific competitor, was sonicated using a Sonopuls HD 2070 (Bandelin Electric), with two cycles of five pulses at 70% power. CGH was performed essentially following the procedure described in [9]. Briefly, chromosome preparations were removed from the freezer, dehydrated in an ethanol series, treated and denatured. Then the preparations were hybridized with a denatured probe cocktail containing labelled male and female gDNAs (250 ng each), unlabelled sonicated female gDNA (3 μg), and sonicated salmon sperm DNA (25 μg) for 3 days at 37°C, washed for 5 min at 62°C in 0.1x SSC containing 1% Triton X-100, counterstained with 0.5 μg/mL DAPI and mounted in antifade based on DABCO.

Fluorescence in situ hybridization (FISH) with the X-chromosome derived DNA probes

For FISH, we followed the procedure described in [35] with some modifications. Briefly, slides were digested with 100 μg/mL RNase A to remove potential transcripts of target sequences, incubated in 10 mM HCl for 20 min in 37°C to reduce the amount of cytoplasm and blocked for 30 min in 5x Denhardt solution (1% Ficoll, 1% polyvinylpyrrolidone and 1% bovine serum albumin). Then, the chromosomes were denatured in 70% deionized formamide, 2x SSC for 3 min 30 s at 68°C. After denaturation, the slides were hybridized overnight with a probe cocktail denatured for 5 min at 90°C. The probe cocktail for one slide (10 μL) contained 40 ng of each labelled PCR fragment and 25 μg of sonicated salmon sperm DNA (Sigma-Aldrich) in 50% formamide, 10% dextran sulphate and 2x SSC. Hybridization signals were detected and amplified by a series of three antibodies, mouse anti-digoxigenin (1:100, Roche), sheep anti-mouse Ig conjugated with digoxigenin (1:200, Millipore), and sheep anti-digoxigenin conjugated with fluorescein (1:200, Roche). The preparations were counterstained with 0.5 μg/mL DAPI and mounted in antifade based on DABCO.

Microscopy and Image Processing

Preparations were observed in a Zeiss Axioplan 2 microscope equipped with appropriate fluorescence filter sets. Black-and-white images were taken either with a cooled F-View CCD camera (DAPI- and CGH-stained preparations) or an Olympus CCD monochrome camera XM10, and captured separately for each fluorescent dye with either AnalySIS software, version 3.2 (Soft Imaging System GmbH) or with cellSens 1.9 digital imaging software (Olympus Europa Holding), respectively. The images were pseudo-coloured and merged using Adobe Photoshop CS4 and CS5 (Adobe Systems Inc.).