Cytogenetics of the Hybridogenetic Frog Pelophylax grafi and Its Parental Species Pelophylax perezi

Abstract Hybrid taxa from the genus Pelophylax can propagate themselves in a modified way of sexual reproduction called hybridogenesis ensuring the formation of clonal gametes containing the genome of only one parental (host) species. Pelophylax grafi from South-Western Europe is a hybrid composed of P. ridibundus and P. perezi genomes and it lives with a host species P. perezi (P-G system). Yet it is unknown, whether non-Mendelian inheritance is fully maintained in such populations. In this study, we characterize P. perezi and P. grafi somatic karyotypes by using comparative genomic hybridization, genomic in situ hybridization, fluorescent in situ hybridization, and actinomycin D-DAPI. Here, we show the homeology of P. perezi and P. grafi somatic karyotypes to other Pelophylax taxa with 2n = 26 and equal contribution of ridibundus and perezi chromosomes in P. grafi which supports F1 hybrid genome constitution as well as a hemiclonal genome inheritance. We show that ridibundus chromosomes have larger regions of interstitial (TTAGGG)n repeats flanking the nucleolus organizing region on chromosome no. 10 and a high quantity of AT pairs in the centromeric regions. In P. perezi, we found species-specific sequences in metaphase chromosomes and marker structures in lampbrush chromosomes. Pericentromeric RrS1 repeat sequence was present in perezi and ridibundus chromosomes, but the blocks were stronger in ridibundus. Various cytogenetic techniques applied to the P-G system provide genome discrimination between ridibundus and perezi chromosomal sets. They could be used in studies of germ-line cells to explain patterns of clonal gametogenesis in P. grafi and broaden the knowledge about reproductive strategies in hybrid animals.


Significance
In the genus Pelophylax, there are interspecies hybrids able to propagate, but their reproductive strategy was studied only in one taxon, the edible frog P. esculentus, composed of P. ridibundus and P. lessonae genomes.To elucidate the inheritance pattern in another hybrid P. grafi, containing mixed genomes of P. ridibundus and P. perezi, we analyzed the chromosomal sets of P. grafi and P. perezi.Although the chromosome number was consistent with other Pelophylax taxa (2n = 26), we found differences in genomic arrangements between ridibundus and perezi chromosomes.This will contribute to novel studies of P. grafi gametogenesis, particularly to genomic composition of germ-line cells, which will broaden the knowledge about reproductive strategies in hybridogenetic water frogs.

Introduction
The Western Palearctic water frogs of the genus Pelophylax include three main lineages, the perezi lineage, the lessonae lineage, and the ridibundus/bedriagae lineage (Lymberakis et al. 2007;Akın et al. 2010).Some of these taxa do hybridize naturally, sometimes resulting in nuclear and mitochondrial DNA introgressions (Kolenda et al. 2017) and giving rise to various mixed populations composed usually of one parental (or host) species and the hybrid with hybridogenetic reproduction, which is also categorized as asexual and non-Mendelian that is without recombination (Plötner et al. 2008;Hoffmann et al. 2015;Dufresnes et al. 2017;Papežík et al. 2021).Hybridogenetic P. esculentus, which is characterized by a widespread continental distribution, is originally a hybrid between P. lessonae and P. ridibundus, living together either with parental P. lessonae (the L-E system) or parental P. ridibundus (the R-E system) (Berger 1973;Uzzell and Berger 1975;Graf and Polls-Pelaz 1989;Plötner 2005).Peninsular Italy and Sicily are inhabited by hybridogenetic P. hispanicus, a hybrid with genomes from P. bergeri and P. ridibundus, forming a mixed population with parental P. bergeri (the I-RI system) (Graf and Polls-Pelaz 1989).Finally, P. grafi is considered another hybridogenetic taxon composed of P. ridibundus and P. perezi genomes.Its distribution ranges from South-Western France to North-Eastern Spain, where it lives with a host species P. perezi (the P-G system) (Graf and Polls-Pelaz 1989;Crochet et al. 1995;Sánchez-Montes et al. 2016).Mechanisms of inheritance pathways in the P-G system, presumably analogous to the L-E system (Graf and Polls-Pelaz 1989), remain poorly understood.
Hybrids form clonal gametes in two main steps through a process called hybridogenesis.The first step includes eliminating one of the parental genomes, followed by duplication of the second, non-eliminated parental genome.Elimination of one of the genomes from the germ-line cells of P. esculentus hybrids occurs in both sexes before meiosis during very early gametogenesis in specialized gamete ancestor cells that is gonocytes.The elimination is a unique process that takes place during gonocyte interphase when the chromatin of one set of chromosomes is specifically marked by an unknown molecular mechanism and is discarded from the nucleus in the form of micronuclei (Ogielska 1994;Chmielewska et al. 2018).The second mechanism of genome elimination involves micronuclei formation from lagging chromosomes during gonocytes division (Dedukh et al. 2020).Chromatin in micronuclei is highly condensed and becomes transcriptionally inactive and degraded via autophagy (Chmielewska et al. 2018).Using genomic in situ hybridization (GISH), fluorescent in situ hybridization (FISH), or other species-specific cytogenetic markers (Dedukh et al. 2020;Chmielewska et al. 2022), we can trace which genome is eliminated.The remaining chromosomal set is duplicated and chromosomes undergo meiosis, resulting in genetically non-recombined gametes (reviewed by Graf and Polls-Pelaz 1989;Ogielska 2009).
Clonal transmission of non-eliminated genomes from generation to generation requires breeding between a hybrid and a parental species whose genome has been eliminated.Therefore, hybrids typically have to coexist with one of the parental species (Graf and Polls-Pelaz 1989;Plötner 2005).Studies on populations that included P. esculentus revealed a high diversity of gametogenic pathways in this hybrid (Dedukh et al. 2015(Dedukh et al. , 2019;;Doležálková-Kaštánková et al. 2021;Chmielewska et al. 2022;Pustovalova et al. 2022), whereas P. grafi hybrids are currently thought to form only one population type with unknown pathways of gamete formation.
Gametogenesis and genome elimination in P. esculentus have generated a lot of scientific interest, as exceptions to the rules are often excellent opportunities to learn about the functioning of near-universal mechanisms such as gametogenesis in vertebrates.In this context, having another model of genome exclusion during gametogenesis would allow testing for the generality of the results obtained in P. esculentus.Comparative analysis of karyotypes from two hybridogenetic systems, P. esculentus and P. grafi, may reveal the strength of the interspecies meiotic barrier preventing genomic admixture, whether karyotypes are conserved or not in the Pelophylax genus, and whether the mechanism of genome inheritance in grafi are similar to other Pelophylax.In addition, because P. grafi is currently assumed to persist in a single system (the P-G system where it reproduces with P. perezi), one may assume that the hybridogenesis process may be less variable in this organism compared to the widely studied P. esculentus.Last, very little is known about ploidy variation in populations of the P-G systems.Schmeller et al. (2001) reported the occurrence of triploids in grafi but do not provide the corresponding data.As this has important implications for the biology of the P-G system, assessing ploidy in individuals of the hybrid taxon of the P-G system would provide crucial information for its ecology and conservation.
To investigate the mechanisms underlying genome elimination during hybrid gametogenesis, it is crucial to properly identify parental chromosomes in germ cells.Parental species of the well-studied P. esculentus, P. ridibundus, and P. lessonae (2n = 26), do have five large and eight small chromosomes, by which large chromosomes were considered metacentric or submetacentric, while some small chromosomes were also acrocentric (Ebendal 1977;Koref-Santibáñez and Günther 1980;Zales ńa et al. 2011) (table 1).Distinguishing the chromosomes of the parental species in karyotypes of P. esculentus hybrids became possible only when Heppich et al. (1982) discovered the difference in the content of centromeric AT repeats.After actinomycin D-DAPI (AMD-DAPI) staining the strong fluorescent signals were detected in centromeres of chromosomes belonging to P. ridibundus but not in those of P. lessonae (Heppich et al. 1982;Tunner and Heppich-Tunner 1991;Ogielska et al. 2004).Species within ridibundus and lessonae lineages differ in the number of pericentromeric RrS1 repeat, which is abundant in chromosomes of the species within the ridibundus lineage while less represented in P. lessonae chromosomes (Ragghianti et al. 1995;Marracci et al. 2011).The FISH with the probe against RrS1 repeat allows identifying five large and one small (Ragghianti et al. 1995(Ragghianti et al. , 2007;;Marracci et al. 2011;Chmielewska et al. 2022) or all P. ridibundus chromosomes (Dedukh et al. 2019(Dedukh et al. , 2020)).DNA sequence homologous to RrS1 in P. lessonae is les177.1 (Marracci et al. 2011).A new FISH marker specific for P. lessonae genome is the minisatellite sequence PlesSat01-48 (44 bp), which is present on pericentromeric regions of two chromosome pairs, the acrocentric chromosome 8 and the chromosome 10 (Choleva et al. 2023, preprint).Variability of the interstitial telomeric repeats (ITS) which are flanking the nucleolus organizing region (NOR) on chromosome 10 allows for the recognition of P. ridibundus lampbrush and mitotic chromosomes with two ITS on both sides of the NOR from P. lessonae chromosomes with one ITS on the distal side of the NOR (Dedukh et al. 2013(Dedukh et al. , 2019)).Comparative genomic hybridization (CGH) and GISH successfully identified both parental genomes in P. esculentus hybrid cells (Zales ńa et al. 2011;Doležálková et al. 2016).The latter two cytogenetic methods can additionally discriminate large-scale chromosomal rearrangements in hybrids (Bi et al. 2007).Finally, genome composition in oocytes can be identified through the investigation of lampbrush chromosomes isolated from diplotene oocyte nuclei due to their enormous size and distinct morphology (Zlotina et al. 2017).Variation in morphology allows the identification of individual chromosomes and has been successfully used for the discrimination of genomes transmitted by P. esculentus females (Bucci et al. 1990;Dedukh et al. 2013Dedukh et al. , 2015Dedukh et al. , 2017)).
In this study, we focused on the P-G system involving the parental species P. perezi and P. grafi, a taxon of hybrid origin carrying one set of ridibundus and one set of perezi chromosomes (Crochet et al. 1995;Sánchez-Montes et al. 2016).The origin of the P-G system is currently unknown as P. perezi and P. ridibundus are considered to have been allopatric before current human-induced species translocations; a hypothesis is that the P-G system emerged from crosses between P. perezi and P. esculentus in Western France where they are sympatric (Dubey and Dufresnes 2017;Dufresnes et al. 2017;Dufresnes and Mazepa 2020).As such, P. ridibundus individuals are absent from this system and its genome is passed down from one generation of hybrids to another only through gametes produced by P. grafi.One of the difficulties in studies of the P-G system is that P. ridibundus and P. perezi have very similar morphology, rendering the identification of the hybrid taxon grafi currently impossible based on morphology.Since P. ridibundus has recently invaded large parts of the range of the P-G system (Demay et al. 2023), safe identification of these three taxa relies on genetic analyses (Cuevas et al. 2022); this complicates considerably targeted sampling of grafi and perezi individuals in natural populations.
Particularly, we aimed to characterize P. perezi karyotypes in both somatic cells and meiotic diplotene oocytes to assess their relationship with other taxa of the Pelophylax genus.We also wanted to explore whether chromosomal complements of P. grafi hybrids are similar to other Pelophylax karyotypes and maintain the integrity of the parental genomes, which may support the hypothesis of clonal gamete production by this hybrid.Finally, we performed a comparison of P. perezi and P. ridibundus chromosomal features in the somatic karyotypes of P. grafi using cytogenetic techniques previously successfully applied on the P. esculentus chromosomes.Knowledge of chromosomal features within the P-G system will help us understand inheritance patterns in this system, facilitating the examination of DNA elimination processes in germ-line cells.

Pelophylax perezi Karyotype
Somatic tissue.The diploid karyotype of P. perezi assessed in two individuals contained 26 chromosomes and was consistent in the number of chromosomes and in their morphology.The karyotype included five pairs of large and eight pairs of small chromosomes, all pairs gradually decreasing in length (fig.1A).The pairs of chromosomes were sorted according to centromere position and divided into the following classes: metacentric (nos. 1, 5, 6, 7), submetacentric (nos. 2, 3, 4, 10, 11), and acrocentric (nos. 8, 9, 12, 13) (fig.1B).Chromosome pairs nos.8 and 9, as well as 12 and 13 were similar in centromere position but different in length.The NOR was located on the long arm of each chromosome no.10 (fig. 1

-black arrow).
Lampbrush chromosomes.We isolated complete 24 lampbrush chromosomal sets from two P. perezi females.The lampbrush karyotype of P. perezi included 13 fully paired bivalents with clearly distinguishable five large and eight small bivalents (supplementary fig.S1A, Supplementary Material online).Centromeric regions were indistinguishable neither under the phase contrast microscope nor after DAPI staining.The chromosome arrangement was performed after FISH with the RrS1 probe detecting pericentromeric heterochromatin (supplementary fig.S1B, Supplementary Material online).Using FISH with an oligonucleotide probe specific to (TTAGGG) n repeat, we detected interstitial telomeric sites (ITS) (supplementary fig.S2, Supplementary Material online).Additionally, we evaluated the distribution of marker structures along the chromosomal axis, including chromosomes-associated spheres and nucleoli, and loops with unusual morphology and an intense accumulation of ribonucleoprotein matrix, distinguishing them from other lateral loops.However, the precise recognition of certain marker loops was challenging due to the condensed structure of lateral loops.Still, it allowed us to construct the cytological maps of lampbrush chromosomes where the relative position of the most prominent marker structures was indicated (supplementary fig.S3, Supplementary Material online).Lampbrush chromosomes were arranged by size and numbered with the letters (A-M) as they may differ in comparative length with somatic metaphase chromosomes due to various levels of chromatin condensation.
Further, we identified the relative position of all prominent marker structures on P. perezi lampbrush chromosomes (supplementary figs.S1 and S3, Supplementary Material online).Chromosome A is the longest one at the lampbrush stage.In the subtelomeric regions of both arms, we identified noticeable marker loops (supplementary figs.S1 and S3, Supplementary Material online).The long arm of B had a sphere in its subtelomeric region and a marker loop in the pericentromeric region of the short arm (supplementary figs.S1 and S3, Supplementary Material online).In the long arm of C, we detected long marker loops and lumpy loops (supplementary figs.S1 and S3, Supplementary Material online).Chromosomes D and E have long marker loops in the short arms (supplementary figs.S1 and S3, Supplementary Material online), F does not have prominent marker structures (supplementary figs.S1 and S3, Supplementary Material online), G has a marker loop in the pericentromeric region of the short arm (supplementary figs.S1 and S3, Supplementary Material online), and H has a lumpy loop, marker loop, and active NOR locus with attached nucleoli (supplementary figs.S1 and S3, Supplementary Material online).Furthermore, FISH with an oligonucleotide probe specific to (TTAGGG) n repeat revealed three small interstitial blocks of this repeat on chromosome H (supplementary fig.S2, Supplementary Material online).On the long arm of J, we found two marker loops (supplementary figs.S1 and S3, Supplementary Material online) and a sphere in the short arm, as well as a marker loop and a lumpy loop in the long arm (supplementary figs.S1 and S3, Supplementary Material online).On the long arms of K and L, we detected marker loops in the long arms (supplementary figs.S1 and S3, Supplementary Material online).Chromosome M had a sphere in the short arm and two pairs of marker loops in the long arm (supplementary figs.S1 and S3, Supplementary Material online).

Pelophylax grafi karyotype
Somatic tissue.The diploid karyotype (2n = 26 chromosomes) of P. grafi was assessed in eight individuals.All chromosomes showed strong hybridization, which indicated that the P. grafi karyotype consists of 13 P. ridibundus and 13 P. perezi chromosomes, clearly seen both from GISH (fig.2A and B) and CGH (fig.2C and D) analyses.The pairs of chromosomes were metacentric, submetacentric, and acrocentric with pair no. 10 bearing the NORs, as in the case of the P. perezi karyotype.Using the GISH technique with the P. perezi wholegenomic probe and blocking cot10 DNA from P. ridibundus, we distinguished species-specific P. perezi heterochromatin on the chromosomes from pairs nos. 1, 3, 5, 12, 13 (fig.2A and B-white arrowheads), while chromosomes nos. 1 and 3 showed stronger signals in subtelomeric regions (on p or q arms, respectively).Pelophylax perezi chromosomes from pairs nos.5, 12, and 13 showed stronger signals in the pericentromeric regions; no. 5 showed stronger signals on the p arms, and nos.12 and 13 showed stronger signals on the q arms.The signal of the pericentromeric region on chromosome pair no. 12 always appeared as the most intensive and remained highly visible even in CGH-treated metaphase plates (fig.2C and D-white arrowhead).In P. grafi karyotypes, P. ridibundus chromosomes had a similar size as compared to P. perezi.Neither GISH nor CGH-treated metaphase chromosomes showed genome introgressions within a chromosomal structure.None of the analyzed metaphase plates indicated the existence of polyploid individuals.
Pericentromeric Heterochromatin and Telomeric (TTAGGG) n Repeats in P. perezi and P. ridibundus Chromosomes FISH with a pericentromeric probe to the RrS1 repeat, specific to the P. ridibundus genome, was performed on P. ridibundus, P. perezi, and the hybrid P. grafi individuals and showed that the probe efficiently hybridized with both P. ridibundus and P. perezi chromosomes (fig.4A-C).The fluorescence level of the signal intensity differed between the genomes of the two species and was higher in the ridibundus chromosomes, visualized by performing FISH on slides previously used for the CGH (sequential staining).Artificially lowered level of brightness for green color signals of the probe to RrS1 repeat on P. grafi karyotypes revealed that on several figures, the signals on five large and one small ridibundus chromosomes were the most prominent, while on the other P. ridibundus and all P. perezi chromosomes, the signal was weak or invisible (fig.3A and B).Using solely the FISH technique, it was rather difficult to differentiate parental sets in the hybrids' tissues, however, some chromosomes had higher proficiency in combining with the probe to RrS1 repeat (fig.4A, white arrows).The comparison with FISH-treated metaphase plates obtained from P. perezi somatic tissues shows that pericentromeric fluorescence signals were lower and rather uniform than those of P. grafi and P. ridibundus (fig.4A-C).Adjusting the threshold for the green color allowed us to distinguish eight chromosomes belonging to P. ridibundus (fig.3C and D).Additionally, we observed that the species-specific sequence on chromosome 12 of P. perezi genome is possibly AT-rich, due to the strong DAPI signal in this area (fig.3A-D, white arrowheads).
FISH with an oligonucleotide probe specific to telomeric repeats (TTAGGG) n performed on P. grafi revealed interstitial telomeric sequences on both chromosomes no. 10 (fig.4Darrowheads, E), which have secondary constrictions (NOR) on the long arm visible as dark areas on the proximal part of the q-arm.Hybrid metaphase plates showed that interstitial sequences flanked both sides of the NOR region of the ridibundus chromosome no.10 (fig.4D-white arrowheads).Although P. perezi genome showed ambiguous results, it was visible that NORs were flanked from both sides with interstitial (TTAGGG) n sequences yet less abundant (fig.4D-gray arrowheads), as reflected by lower staining intensity than those of P. ridibundus.This subtle difference in NOR staining was sufficient enough to distinguish ridibundus and perezi chromosomes in the P. grafi genome (fig.4D and E).
AMD-DAPI performed on P. ridibundus chromosomes showed quite distinctive bright spots in the centromeric regions (supplementary fig.S4A, Supplementary Material online) indicating AT-rich heterochromatin, while P. perezi chromosomes were more homogenous (supplementary fig.S4B, Supplementary Material online), and the centromeric regions stained weakly and gave faint signals.The difference in the signal intensity is well visible in metaphase plates of the P. grafi hybrid (supplementary fig.S4C, Supplementary Material online).

Discussion
Pelophylax perezi and P. grafi Somatic Karyotypes Karyotypes of P. perezi and its hybrid P. grafi consisted of 13 pairs of chromosomes (2n = 26).The studied P. grafi had integral parental chromosomal sets and maintained F1 hybrid genome constitution (fig.2, fig.5A and B showed vast similarities in chromosome morphology (table 1).Diploid karyotypes of P. perezi and P. grafi were composed of five pairs of large and eight pairs of small chromosomes, as in other Pelophylax taxa studied so far.We characterized P. perezi karyotype as with eight metacentric, ten submetacentric, and eight acrocentric chromosomes (fig. 1 and table 1).The choosing of categories to which the chromosomes were assigned was conducted based on visual comparisons with existing literature and researchers' experience concerning Pelophylax karyotypes.Measure-based categorization was deemed insufficient due to the impact of the methodology on the morphology of chromosomes (Heppich 1978;Hnátková et al. 2023;Krysanov et al. 2023).The occurrence of eight acrocentric chromosomes is not a common phenomenon in water frogs and was reported only in P. perezi (this study).In earlier reports, the numbers of acrocentrics ranged from four to zero (Ebendal 1977;Koref-Santibáñez and Günther 1980;Martirosian and Stepanyan 2009;Zales ńa et al. 2011), suggesting diversity at the population level.However, it is important to mention that those results were obtained largely during chromosomal measurements.Overall, the karyotype of P. perezi and other Pelophylax species is generally stable, rarely it shows even extended variation.It is supported by an evolutionary context of a single ancestor of the Western Palearctic water frogs Pelophylax revealed by the mitochondrial phylogeography studies (Lymberakis et al. 2007; Dufresnes and Mazepa 2020), as well as karyotype studies in which the results were superimposable (table 1).Despite the above-mentioned data in Bucci et al. (1990), it is reported that in samples from Central Poland, chromosome pair no.11 was metacentric in P. ridibundus and P. lessonae, and chromosome pair no. 12 was submetacentric in P. ridibundus and metacentric in P. lessonae (Bucci et al. 1990).We do not know yet whether this variation has an evolutionary context or it simply comes from a different methodology.As an example, P. lessonae from France had submetacentric and metacentric chromosome pair no.11 caused by the centric inversion (discussed in Tunner and Heppich 1983).All 15 P. grafi individuals studied here were diploid, whereas Schmeller et al. (2001) reported the occurrence of 20 triploids in a sample of 234 individuals of this taxon.Further data on the ploidy of grafi individuals are needed to understand if triploids occur at low frequency in most populations of the P-G system or if they reach high proportions in some populations, as is observed in the L-E system.

GISH and CGH in the Hybrid P. grafi
Applying the GISH technique with P. perezi whole-genomic probe on P. grafi, we demonstrated that P. perezi chromosomes exhibited several species-specific sequences schematically presented in figure 5A and B. Both methods clearly visualized two distinct groups of chromosomes and differentiated chromosomes belonging to P. ridibundus from those belonging to P. perezi.We confirmed the genomic integrity of the two parental chromosome sets in somatic cells of P. grafi, which indirectly supports functional hybridogenetic mechanisms of clonal reproduction in these taxon, efficiently restricting gene flow between parental genomes as known from P. esculentus (Zales ńa et al. 2011) or clonal fishes (Majtánová et al. 2016(Majtánová et al. , 2021)).
The ridibundus and perezi chromosomes were similar in size when we compared 13 pairs of homologs on the karyograms of P. grafi; yet, perezi chromosomes seem to be slightly smaller.Combining the current data with those presented by Zales ńa et al. ( 2011), lessonae chromosomes studied in hybrid P. esculentus were much smaller when compared with P. ridibundus and perezi.
GISH performed with a whole-genomic probe from P. perezi and P. ridibundus cot10 DNA highlighted heterochromatin blocks that could be tandem repeats on five perezi chromosomes, three (nos.5, 12, and 13) in pericentromeric regions and two (nos. 1 and 3) in telomeric regions.Although the CGH procedure does not predispose to greater specificity of binding by possible occurrence of repeats contained in the whole-genomic probes, the pericentromeric region of chromosome no.12 remained intensively highlighted throughout nearly all of the metaphase plates studied (fig.2A-D) and is a potentially useful speciesspecific pericentromeric marker for future studies in the P-G system.
GISH and CGH remain highly useful cytogenetic methods used in studies on hybrid and allopolyploid animals.GISH well-distinguished the ridibundus and lessonae parental chromosomes in the hybrid P. esculentus within L-E, R-E, R-E-L, and E-E systems (Zales ńa et al. 2011;Chmielewska et al. 2022).GISH was also successfully applied to hybrid and allopolyploid gynogenetic unisexual salamanders of the Ambystoma laterale-jeffersonianum complex (Bogart et al. 2007;Bogart and Bi 2013).The CGH technique was effectively implemented in the R-E system (Doležálková et al. 2016), also allowing us to clearly distinguish perezi and ridibundus genomes in P. grafi karyotypes.The difference between GISH and CGH is methodological.In the first case, a genome of one of the parental species is blocked with the whole-genomic DNA or cot10 DNA (whole-genomic DNA enriched with repetitive sequences), which blocks dispersed repeats, and the other is aligned with a species-specific probe, which results in more pronounced signals of species-specific sequences including species-specific heterochromatin blocks (Symonova et al. 2015).In the case of CGH, the repetitive sequences are not drastically removed.Most evenly distributed retrotransposons of both genomes are aligned with a speciesspecific probe, visualizing the integrity of genomes or genome introgressions in metaphase plates.

Pericentromeric Heterochromatin and Telomeric (TTAGGG) n Repeats in P. perezi and P. ridibundus Chromosomes
We detected pericentromeric RrS1 repeats in P. grafi (with perezi and ridibundus chromosomes) and in P. perezi, with a more intense fluorescent signal of the probe on ridibundus than perezi chromosomes.Our observation might have been connected to likely a higher number of RrS1-like repeats in pericentromeric regions of P. ridibundus in comparison with P. perezi chromosomes.
Interestingly, we did not observe morphologically distinct centromeric regions on P. perezi lampbrush chromosomes that would suggest a smaller size of pericentromeric repeats in this species (supplementary fig.S1, Supplementary Material online).Similarly, no centromeres at the lampbrush chromosomal stage were observed in P. lessonae from Poland (Bucci et al. 1990).Nevertheless, we visualized centromeric regions on lampbrush chromosomes using FISH with a probe against RrS1 repeat and its signal on P. perezi lampbrush chromosomes seems to be weaker than those obtained from P. ridibundus (Dedukh et al. 2013).Ragghianti et al. (1995) reported that pericentromeric RrS1 repeats are abundant in ridibundus and relatively infrequent in lessonae chromosomes (around 2% of the genome).FISH with the RrS1 probe allowed the identification of either all ridibundus chromosomes (Dedukh et al. 2019(Dedukh et al. , 2020, this study) , this study) or at least six chromosomes (five large and one small chromosomes no. 8 [Ragghianti et al. 1995[Ragghianti et al. , 2007;;Marracci et al. 2011;Chmielewska et al. 2022]).The latter pattern seems to be typical for a ridibundus/ bedriagae lineage, considering that Marracci et al. (2011) observed such a signal also from P. kurtmuelleri and P. bedriagae.We observed this pattern on ridibundus chromosomes under methodological conditions when FISH was made after CGH, otherwise, the result was ambiguous.
The interstitial telomeric repeats flanked the NOR regions on the chromosome pair no. 10 and the signal was stronger in ridibundus than in perezi, suggesting differences in the copy number of this repeat.Previously mapping (TTAGGG) n probe on the lampbrush chromosome of P. ridibundus, we observed a big cluster of ITS with approximately 8% from the length of the whole lampbrush chromosome H located in the region closer to the centromere (Dedukh et al. 2013).However, on lampbrush chromosomes of P. perezi, we found three small ITSs in the regions flanking NOR region (supplementary fig.S2, Supplementary Material online).Both RrS1 and telomeric repeats also distinguished P. perezi from P. lessonae chromosomes which did not bind the respective probes (Dedukh et al. 2015).In addition, we did not find ITS in the subtelomeric region of the long arm of lampbrush chromosome B of P. perezi (supplementary fig.S2, Supplementary Material online), contrary to P. ridibundus chromosomes (Dedukh et al. 2013).ITS are known to form due to chromosomal rearrangements, telomeres fusion of ancestral chromosomes, or after reparation of double-stranded breaks (Lin and Yan 2008).Given the similar location of one small (TTAGGG) n site on P. perezi chromosome to P. ridibundus and P. lessonae chromosomes, we propose that this site originated in their common ancestor.Additionally, P. perezi and P. ridibundus, but not P. lessonae, exhibit an additional (TTAGGG) n site, differing in size, likely resulting from its amplification on the P. ridibundus chromosome.The expansion of microsatellites and (TTAGGG) n sequences may be caused by replication errors due to DNA polymerase slippage (Lin and Yan 2008) and was suggested to explain ITS size difference in such loci on human chromosomes (Mondello et al. 2000) However, the origin of this ITS loci on water frogs chromosomes requires further investigations.
Actinomycin D treatment, which is a nonfluorescent DNA intercalator to GC pairs, combined with DAPI staining (AMD-DAPI) allowed us and others to show a visible difference concerning centromeric AT repeats in chromosomes belonging to different species of the genus Pelophylax (Heppich et al. 1982;Ogielska et al. 2004).Zales ńa et al. ( 2011) proved that P. ridibundus chromosomes contain higher quantities of such repeats in comparison to those of P. lessonae.We have repeated this experiment for the P-G complex and obtained parallel result in which P. ridibundus chromosomes contained a higher copy number of these repeats in comparison to those of P. perezi.This differentiation seems to be a useful tool to discriminate between the two sets of parental chromosomes in hybrid P. grafi.

Pelophylax perezi Lampbrush Chromosomes
Identification of genome composition in oocytes is possible using the analysis of lampbrush chromosomes isolated from diplotene nuclei (Zlotina et al. 2017).The large size and distinct morphology allowed the identification of individual chromosomes which was successfully used for species-specific discrimination of genomes transmitted by females of P. esculentus complex (Bucci et al. 1990;Dedukh et al. 2013Dedukh et al. , 2015Dedukh et al. , 2017)).
Analysis of lampbrush chromosomes isolated from growing oocytes clearly revealed 13 fully paired bivalents (five large and eight of smaller size) corresponded well with the mitotic (somatic) karyotype.These results are similar to those of P. ridibundus and P. lessonae from various localities (Bucci et al. 1990;Dedukh et al. 2013).Comparison of the chromosome morphology of P. perezi (this study) with those of P. ridibundus and P. lessonae showed four distinct groups of chromosome homology.1) Lampbrush chromosomes A, C, and J of P. perezi have a conservative distribution of marker structures which are similar among all three species.Nevertheless, 2) chromosomes B, E, I, and M of P. perezi resemble the corresponding chromosomes of P. ridibundus, while 3) chromosomes G, H, and L of P. perezi resemble more chromosomes of P. lessonae.4) Only three chromosomes (D, F, and G) of P. perezi have a different and generally poorer distribution of marker structures from corresponding P. ridibundus and P. lessonae chromosomes, which is unique for P. perezi.Similar to P. ridibundus and in contrast to P. lessonae, lampbrush chromosome H of P. perezi bears associated nucleoli, suggesting the presence of active NOR during the lampbrush stage.We believe that the observed marker structures on lampbrush chromosomes of P. perezi can be used for further recognition of the genome transmitted by P. grafi oocytes.

Conclusions
Somatic karyotypes of P. perezi and P. grafi provided a new evidence for conserved chromosome morphology in the Pelophylax taxa.Genome integrity was shown in both perezi and ridibundus chromosomes in hybrid P. grafi supporting a hemiclonal genome inheritance in this hybrid.FISH to RrS1 genomic sequence so far successfully used to distinguish ridibundus from lessonae chromosomes gave ambiguous results for the discrimination between ridibundus and perezi chromosomes.Future research on the genome elimination mechanism in germ-line cells in P. grafi may be performed by applying GISH and CGH together with AMD-DAPI, which were proven to efficiently discriminate both parental genomes.Species-specific sequences revealed in P. perezi chromosomes with the whole-genomic probe are a promising lead for the future creation of FISH probes designed for perezi genome.Finally, this study revealed another promising marker with a slightly different expression pattern in FISH for telomeric repeats, where perezi chromosome no.10 showed smaller ITS sequences flanking the NOR than in ridibundus chromosome.This sequence and variation in marker structures can be used in studies of lampbrush chromosomes in P. grafi oocytes to investigate patterns of clonal gametogenesis in P. grafi.Comparison of P. ridibundus, P. perezi, and P. grafi genomes in our and other available data clearly show that the ridibundus chromosomes do not differ significantly between the hybrids while lessonae and perezi do.This strongly suggests that the ridibundus genome promotes and enforces the elimination of lessonae or perezi chromosomes from the germ line and therefore is a clue to hybridogenesis.

Animals
For the genomic probes and cytogenetic study of P. grafi and P. perezi, we selected 20 animals (4 adults, 11 juveniles and metamorphs, and 5 tadpoles, of which 14 were females, three males, and three non-sexed) (supplementary table S1, Supplementary Material online) collected in 2021 from three mixed P. perezi-P.grafi (P-G) populations in Southern France.
Two P. ridibundus males used for AMD-DAPI staining and DNA extractions for genomic probes were collected in Ruda Milicka (51.533153°N, 17.335117°E), Poland, in 2016, following the permission of the Regional Directorate for Environmental Protection in Wrocław no.WPN.6401.177.2016
Afterward, PCR products were incubated with restriction enzymes (Thermo Fisher Scientific, USA) for RAG1: EcoO1091 (Dra II) (RGzGNCCY) and for Tyr1: Bal I (Mls I) (TGGzCCA).Digested PCR products were run in agarose gel electrophoresis in 1.5% agarose, and results were documented using the Molecular Imager GelDoc XR+ system with ImageLab software (Bio-Rad).DNA size marker Perfect 100 bp DNA Ladder (EURx, Poland) was used as a size standard.Results were interpreted according to Cuevas et al. (2022), where EcoO1091/Dra II digested 936 bp RAG1 PCR product in P. ridibundus, which resulted in two fragments, 565 and 371 bp long, and Bal I/Mls I digested 601 bp Tyr1 PCR product in P. perezi resulting in two fragments, 318 and 283 bp long.

Preparation of Chromosomes
Adult and juvenile animals were injected intraperitoneally with 1 ml or 0.5 ml (for animals with a body mass below 30 g) of 0.3% colchicine solution (Sigma-Aldrich) 24 h before dissection.Metamorphs and tadpoles were placed in a beaker with 0.1% colchicine solution overnight.
Shortly before sacrificing, adult and juvenile animals were anesthetized with a 0.5% water solution of MS 222 (Sigma-Aldrich), while metamorphs and tadpoles were treated with a 0.25% water solution of MS 222 (Sigma-Aldrich).A piece of intestine was dissected from each individual, hypnotized in 0.075 M KCl for 10 min, and fixed in Carnoy fixative (ethanol:glacial acetic acid in 3:1 proportion).After three changes of Carnoy fixative, samples were stored at −20 °C until use.
To obtain chromosomal spreads, small pieces of intestine tissue were placed in 70% acetic acid in the glass preparation chamber and gently homogenized using forceps under a stereomicroscope.The cell suspension was then spread over sloped slides placed on a heater set at +60 °C.Afterward, the chromosomes on the slides were stained with the Giemsa solution.
Metaphase plates on slides were initially scanned by the Zeiss Axioplan epifluorescence microscope equipped with a Charge Coupled Device (CCD) camera and ZEISS Axio Imager.Z2 epifluorescence microscope (Zeiss, Oberkochen, Germany) with Metafer platform (MetaSystems, Altlussheim, Germany).Selected slides were afterward de-stained in ethanol series (50%, 70%, 96%) and used for further staining.Classification of chromosome types (metacentric, submetacentric, acrocentric) was performed based on the morphological features and comparisons to existing literature.
For GISH, we used 1.2-2 µg unlabeled blocking DNA (cot10 DNA) prepared according to S1 nuclease protocol (Trifonov et al. 2009) from total genomic P. ridibundus DNA.Whole-genomic P. perezi probe was labeled by nick translation similarly to the CGH method (1 µg gDNA per reaction, 0.8 µg per slide).Further steps were identical for GISH and CGH.
Microscopic slides with chromosomal spreads were incubated with 200 ng/ml RNase solution in 2× SSC for 1 h at RT in a humid chamber, then washed three times in 2× SSC, 5 min each at RT. Solution of pepsin (0.1 mg/ml) in 0.1 M HCl was dropped on slides for 7 min RT.Slides were washed in 1× PBS for 5-10 min, dehydrated in ethanol (50%, 70%, 96%), and air dried.Chromosomes were denatured in 75% formamide at 74 °C in the water bath (GFL, Germany), followed by incubation in an ice-cold series of ethanol (70%, 80%, 96%) for 5 min each.Simultaneously, probes were denatured at 86 °C in the heating block (Grant Bio PCH1, UK) for 10 min and then put on ice.Probes were incubated on the slides covered with two 24 × 24 coverslips and glued with rubber cement (Marabu GmbH, Germany) in the humid chamber at 37 °C in the incubator (Panasonic MIR-162-PA, Japan) for 2 days.Afterward, the slides were washed in 0.2× SSC three times for 5 min each on the shaker.A blocking solution was administered (either 0.5% bovine serum albumin (BSA) in 1× PBS or 1% blocking reagent [Roche]) for 20 min.The biotin-dUTP and digoxigenin-dUTP were detected using streptavidin-AlexaFluor 488 (Invitrogen) and anti-digoxigenin-rhodamine (Invitrogen), respectively.Streptavidin-Alexa 488 and antidigoxigenin-rhodamine were diluted in blocking solution, administered on slides, and incubated in the dark humid chamber for at least 3 h at RT. Washing was carried out in 4× SSC with 0.1% Tween pre-heated to 44 °C with shaking, three times for 5 min each.Afterward, the slides were dehydrated in ethanol series (70%, 80% 96%), air dried, and mounted in Vectashield medium containing DAPI (1.5 mg/ml) (Vector).
FISH with RrS1 probe was performed on metaphase chromosomal preparations previously used for CGH and on slides with diplotene chromosomes.CGH-treated slides were washed three times in 4× SSC with 0.1% Tween preheated to 44 °C in an incubator with shaking (150 rpm), leading to careful removal of the coverslips.Afterward, slides were treated according to the FISH protocol described below.
In the case of FISH with RrS1 probe, the hybridization mixture contained 50% formamide, 10% dextran sulfate, 2× SSC, ssDNA (Sigma-Aldrich) (500-1,000 ng per slide), and PCR-labeled probe (10-50 ng per slide).Simultaneous and separate denaturations of the probe and chromosomal DNA were performed.For simultaneous denaturation, the hybridization mixture was placed directly on the chromosomal spreads, followed by heating to 75 °C for 3-5 min on the heating block (FALC, Italy).For separate denaturation, slides with chromosomal spreads were heated in 75% formamide (72 °C) for 4 min in the water bath (GFL, Germany), and placed in ice-cold ethanol (50%, 70%, 96%).The hybridization mixture was denatured in the heating block (Grant Bio PCH1, UK) at 86 °C for 7 min, then immediately put on ice for 10 min.The mixture was placed on the slides, covered with two 24 × 24 coverslips and glued with rubber cement (Marabu GmbH, Germany), and incubated in RT overnight.After hybridization, slides were washed in 4× SSC at RT to remove coverslips and then three times in 0.2× SSC heated to 47 °C (150 rpm).A blocking solution (1% blocking reagent [Roche] or 0.5% BSA in 1× PBS) was administered for 20 min.Streptavidin conjugated with AlexaFluor 488 (Invitrogen) was diluted in a blocking solution and was added for at least 3 h RT or overnight (4 °C).The slides were then washed three times in 4× SSC heated previously to 44 °C, 5 min each, then dipped in distilled water and dehydrated in ethanol (50%, 70%, 96%).Slides were mounted with Vectashield medium containing DAPI (1.5 mg/ml) (Vector).Telomeric (TTAGGG) n repeat was detected on metaphase and diplotene chromosomes using a fluorescein 5-isothiocyanate-labeled peptide nucleic acid probe (DAKO, Denmark) in accordance with the providers' instructions.

AMD-DAPI Staining
AMD-DAPI staining was performed in accordance with Heppich et al. (1982) and Chmielewska et al. (2022), with slight changes presented below.Slides were incubated in McIlvaine's buffer pH 7.0, for 10 min, carefully drained on a paper towel, and placed in a humid chamber, where they were incubated in the darkness with actinomycin D (100 µl/slide, 0.25 mg/ml, Sigma-Aldrich) for a minimum of 20 min.Afterward, the slides were washed three times in McIlvaine's buffer for 2-3 min each, drained on paper towels, and mounted with Vectashield medium containing DAPI (1.5 mg/ml) (Vector).

FIG. 1 .
FIG. 1.-Karyotype of P. perezi.Chromosomal spreads were stained with Giemsa.(A) The complete diploid genome of P. perezi consists of 13 pairs of chromosomes.The pairs are arranged from the largest to smallest, resulting in two groups: five large and eight small.(B) Chromosome pairs are divided into classes according to the position of the centromere.Arrows indicate the NOR on chromosome pair no.10. m, metacentric; sm, submetacentric; ac, acrocentric chromosomes.

FIG. 2
FIG. 2.-Karyotype of P. grafi.GISH (A, B) and CGH (C, D) on chromosomes with the whole-genomic probe from P. perezi (A, B) or the mixture of two whole-genomic probes from P. perezi and P. ridibundus (C, D) counterstained with DAPI.(A) Metaphase plate and (B) karyotype containing 26 chromosomes: 13 belonging to P. ridibundus (DAPI-chromosome on the left side of the pair) and 13 to P. perezi (AlexaFluor488-chromosome on the right side of the pair); (C) metaphase plate and karyotype containing 26 chromosomes: 13 belonging to P. ridibundus (RhodamineRed-chromosome on the left side of the pair), 13 to P. perezi (AlexaFluor488-chromosome on the right side of the pair).Arrowheads indicate possible species-specific sequences on P. perezi chromosomes.Scale bars 10 µm.
FIG. 4.-Pericentromeric RrS1 regions and interstitial telomeric regions in the chromosomes of the hybrid P. grafi and their parental species P. perezi and P. ridibundus.FISH with RrS1 probe on (A) P. grafi, (B) P. perezi, and (C) P. ridibundus chromosomes.Although P. grafi chromosomes are characterized by visually distinctive intensities of RrS1 signal on eight chromosomes (arrows), the differences are insufficient to assess the species-specific affiliation of individual chromosomes.Chromosomes no. 10 are marked with dots (A), P. perezi chromosomes no. 12 with visible heterochromatin blocks are depicted by an asterisk (B, D, E). (D, E) FISH on telomeric sequences on P. grafi chromosomes using (TTAGGG) n telomeric probe reveals interstitial sequences on chromosome pair no. 10 enlarged in the inset.Pelophylax ridibundus chromosome on the left has more prominent staining of interstitial sequences (arrowheads) than P. perezi chromosome on the right (darker arrowheads).(E) Karyogram of the chromosomes presented in (D).One of the chromosomes from pair no. 8 is lacking in the metaphase plate.Scale bars 10 µm.

FIG. 5 .
FIG.5.-Diagrams summarizing (A) differences and (B) similarities between two parental genomes, P. ridibundus (chromosome on the left side of the pair) and P. perezi (chromosome on the right side of the pair), examined in a hybrid P. grafi.(A) Centromeric ovals represent AT-rich centromeres present on P. ridibundus chromosomes, P. perezi species-specific sequences present on chromosomes nos. 1, 3, 5, 12, and 13 are marked with different color.Areas with diagonal stripes represent interstitial (TTAGGG) n repeats, more pronounced on the P. ridibundus chromosome no.10. (B) Chromosome areas with diagonal stripes depicted on chromosomal ends are telomeres, marked pericentromeric areas are RrS1 pericentromeric repeats, more abundant on eight P. ridibundus chromosomes (five large ones and three small ones, supposedly nos.6, 7, and 8).