Discovery of facultative parthenogenesis in a new world crocodile

1. Introduction

Once considered rare, the ability of sexually reproducing species to generate offspring without genetic contributions of males, termed facultative parthenogenesis (hereafter, FP), has been documented across multiple vertebrate lineages, including both avian [1] and non-avian reptiles (specifically snakes and lizards) [2], and elasmobranch fishes, with notable examples in sharks, rays and sawfish [3–5]. The widespread phylogenetic distribution and increasing frequency with which FP is reported suggests that the occurrence of this unusual reproductive strategy is less sporadic than previously suggested [6]. In vertebrates, common FP characteristics include (i) strict occurrence in lineages lacking genomic imprinting, a developmental mechanism believed to be restricted to the main mammalian lineages (Eutheria, Marsupialia) [7], (ii) the sex of parthenogens is constrained by the species' sex chromosome system, where XX females produce only female (XX) parthenogens [1,3,7] and ZW females produce only male (ZZ) parthenogens [1,8–13], (iii) the capacity to switch between sexual and parthenogenetic modes of reproduction [14], (iv) the ability to produce both sexual and parthenogenetic offspring within single-reproductive events ([15]; W. Booth 2016, personal observation), (v) the capacity for consecutive parthenogenetic events [5,13,16,17], and (vi) the reproductive competence of adult parthenogens, with both sexual ([10], W. Booth 2016, personal observation) and secondary parthenogenetic events documented ([18] W. Booth 2015, personal observation). Furthermore, viable parthenogens have been reported in wild populations [4,19]. Accordingly, these commonalities among taxa suggest that the phyletic distribution and evolutionary significance of FP have been underestimated.

Documentation of FP in vertebrates is increasing for two primary reasons. First, the contribution of specimens with detailed records from zoological institutions and private keepers predominates in studies, including samples otherwise difficult to obtain (e.g. California condor [20]; Komodo dragon [12]; king cobra [21]). Second, the advent of modern molecular analysis, particularly of genome-scale data, has enabled FP to be robustly disentangled from the phenomenon of long-term female sperm storage (LTSS) [22,23]. With records of LTSS exceeding 6 years [23], it is the main competing hypothesis for the production of offspring after prolonged periods of isolation from potential mates [22,23].

Nonetheless, questions remain as to the occurrence of FP outside of squamates, birds and elasmobranchs. For example, FP has not been documented in Chelonia (turtles) or Crocodylia (crocodiles, alligators and gharials). The latter group is particularly interesting as, unlike all previously documented cases of FP in vertebrates, all crocodilians lack sex chromosomes, and sex determination is controlled by temperature, a mechanism termed temperature-dependent sex determination (TSD) [24]. Furthermore, crocodilians are members of a larger monophyletic lineage (Archosauria), a crown group which at its base splits into two groups, one (Pseudosuchia) comprising crocodilians and their extinct relatives and the other (Avemetatarsalia) that contains birds and their extinct relatives, including pterosaurs and non-avian dinosaurs. Here, using whole-genome sequencing and bioinformatic analysis, we provide, to our knowledge, the first evidence of FP in a crocodilian, the American crocodile, Crocodylus acutus.

2. Methods

The American crocodile (figure 1a) is a large and wide-ranging apex predator. Owing to its saltwater tolerance, the species occurs from the southern USA (Florida) to Northern South America (Colombian, Venezuela and Peru). It is International Union for Conservation of Nature listed as vulnerable and is at risk of extinction in the wild (http://www.iucnredlist.org/species/5659/212805700). On 17 January 2018, a clutch of 14 eggs was discovered in the enclosure of an 18-year-old female American crocodile housed on public exhibit at Parque Reptilandia, Costa Rica. This female was obtained when it was 2 years old (2002) and maintained in isolation from other crocodilians for its entire life. Of the 14 eggs laid, seven appeared to be fertile and were artificially incubated. After three months of incubation, the eggs had failed to hatch and were opened to assess their contents. The contents of six eggs were not discernable, but one egg was found to contain a fully formed non-viable fetus that upon dissection was determined via gross gonadal morphology as female (figure 1b) (electronic supplementary material). DNA was extracted from a scute shed by the mother and cardiac muscle obtained from the fetus using a Qiagen DNeasy Blood & Tissue kit (Qiagen. Hilden, Germany). This was sent to Novogene (Sacramento, CA) for whole-genome sequencing on an Illumina platform (NovaSeq 6000 PE150). Raw sequences were mapped by Novogene to the saltwater crocodile, Crocodylus porosus, reference genome with single-nucleotide polymorphisms (SNPs) identified by Novogene using the following command in SAMtools: mpileup -m 2 -F 0.002 -d 10.

• Download figure
• Open in new tab
• Download PowerPoint

Three thinned SNP datasets comprising 191 255 SNPs (thinned to one variant per 10 kb), 78 603 SNPs (thinned to one variant per 25 kb) and 39 746 SNPs (thinned to one variant per 50 kb) were analysed using ParthenoGenius [25] to test for evidence and mode of parthenogenesis. ParthenoGenius first determines whether paternal alleles are present in the offspring. If the number of maternally homozygous loci for which the offspring has non-identical genotypes to the mother is less than that expected owing to genotyping error alone (i.e. the offspring is homozygous at all or nearly all of the mother's homozygous loci), the offspring is identified as a likely parthenogen. ParthenoGenius then scans maternally heterozygous loci to identify those at which the offspring has retained heterozygosity. Based on the proportion of retained heterozygosity and taking into consideration the estimated per-base error rate, the parthenogenetic mode is identified as either gametic duplication or automixis (electronic supplementary material). Automictic parthenogenesis can be subdivided into terminal fusion, where the terminal meiotic products fuse resulting in near genome-wide homozygosity [21], and central fusion, where parthenogens exhibit extensive genome-wide heterozygosity at a comparable level to that of the mother [26]. The latter is not known to occur in vertebrates [27].

3. Results

The offspring was found to have identical genotypes to the mother at greater than 99.9% of her homozygous loci, demonstrating a lack of paternal alleles. Genome-wide maternal heterozygosity was 0.349, compared to 0.045 for the offspring, demonstrating a significant reduction in heterozygosity in the offspring when compared to the mother. Regardless of the dataset evaluated, the offspring retained approximately 3% of the maternal heterozygosity (table 1), supporting FP through terminal fusion automixis as the mode of reproduction. Note that in instances of central fusion, retained heterozygosity is expected to exceed 33% of that of the mother, whereas terminal fusion will be less than 33% [28]. Owing to the lack of a chromosome-scale C. acutus genome, it is not possible at this time to identify the actual genomic regions that retain heterozygosity with any accuracy.

4. Discussion

Our results provide both novel and substantive evidence of FP in an extant archosaur, the American crocodile. Crocodilians diverged from other archosaurs approximately 240 Ma [29]. Following this split, Pterosauria and Dinosauria underwent substantive diversification, eventually giving rise to modern birds circa 65 Ma [29,30], the only extant theropod dinosaur lineage (figure 2). As such, we now have evidence of FP in the two main branches of extant archosaurians: crocodilians (Pseudosuchia) and birds (Avemetatarsalia). Furthermore, the most-recent common ancestor of squamates (snakes, lizards and amphisbaenids), lineages for which FP has been widely documented in both snakes and lizards [1], and crocodilians, diverged approximately 267.0–312.3 Ma [31], suggesting that FP may be a primitive condition. Comparable to FP studies of birds, snakes and lizards, the evidence presented here supports terminal fusion automixis as the parthenogenetic developmental mechanism [1,21]. Accordingly, at this time, these commonalities do not suggest independent evolutionary origins of this trait across these lineages, but instead a trait probably possessed by a distant common ancestor of these lineages. While beyond the scope of the current study, it is clear that more work is required to fully test the evolutionary distribution and dynamics of FP across deeper evolutionary time, allowing for a more comprehensive understanding of the ancestral states and rates of FP across vertebrate lineages.

• Download figure
• Open in new tab
• Download PowerPoint

Building on previous studies, the data presented here advance our understanding of the distribution of FP in vertebrates, particularly in that all previous studies relate to species whose sex is genetically determined (i.e. XX|XY or ZZ|ZW). In crocodilians, sex chromosomes are absent and the sex of the offspring is thus temperature dependent [24]. Crocodylus acutus exhibits a female–male–female TSD pattern, with temperatures below 30°C and above 33°C producing 100% females, and temperatures around 31.5°C producing a majority of males [24,32]. Here, eggs were incubated at 29–30°C and the fully formed yet stillborn fetus was female. TSD is an ancestral trait in crocodilians, being present also in turtle, lizard and tuatara lineages. Within turtles, recent genomic analyses have revealed that many genes associated with calcium signalling pathways and neuroactive ligand–receptor interactions are associated with TSD [31]. If these genes are also involved in the sex determination process in crocodiles, it appears that the genomic reduction inherent with terminal fusion automixis has not affected the expected offspring sex; larger sample sizes are required for confirmation of this finding.

While it is disappointing that the crocodile parthenogen produced here failed to hatch, it is not uncommon to see non-viable fetuses and developmental abnormalities within litters or clutches of parthenogens [1,21,33], and long-term failure to thrive even for individuals born outwardly healthy [34]. For example, following a 9-year study of FP in turkeys, it was found that on average 3.3% of parthenogenetic eggs hatch [33]. That study followed the fate of 2084 parthenogenetic eggs, with 68 hatching. Furthermore, parthenogens have attained adulthood in a variety of species studied, including turkeys, Meleagris gallopavo [10], California condors, Gymnogyps californianus [20], boa constrictors, Boa imperator [16] and the whitespotted bambooshark Chiloscyllium plagiosum [18]. Studies following the fitness of parthenogens are lacking; however, substantial negative costs associated with the trait have been demonstrated in sharks [34]. While fitness costs are evident, it has been shown that turkey parthenogens and both boa constrictors, B. imperator and royal pythons, Python regius, that survive to adulthood are reproductively competent ([33], W. Booth 2015/2016, personal observation). As such, the failure of this crocodile parthenogen to hatch should not be viewed as an indicator that all crocodile parthenogens will be non-viable.

The evidence presented here provides, to our knowledge, the first documentation of parthenogenesis in the reptilian lineage Crocodylia. Its discovery was dependent on the attention provided by those that maintained the animal in captivity. For instance, while it is not uncommon for captive reptiles to lay clutches of eggs, given the period of isolation from mates, these would normally be considered non-viable and discarded. These findings therefore suggest that eggs should be assessed for potential viability when males are absent. Furthermore, given that FP can occur in the presence of potential mates [4,9,19], instances of FP in crocodilians may be missed when reproduction occurs in females co-habited with males. The discovery of FP in several snake species was reliant on the presence of colour or pattern variants, where offspring exhibited variants inconsistent with sexual reproduction by the putative parents [9,16]. Owing to the rarity of widespread colour/pattern variants in crocodilians, such anomalies indicative of FP would be missed, and hence parthenogens produced by females cohabiting with males would be overlooked without genomic testing. Moreover, it has been hypothesized that FP may be more common in low-density populations, such as those on the verge of extinction [4]. Accordingly, genomic studies investigating wild populations should assess for the potential presence of cryptic parthenogens [4,19].

With terminal fusion automictic FP documented in squamates, birds and now crocodilians, this new evidence offers tantalizing insights into the possible reproductive capabilities of extinct archosaurian relatives of crocodilians, notably the Pterosauria and Dinosauria [35,36].

Ethics

This study used skin shed from the mother and tissue extracted from the stillborn fetus. As such, as no animals were harmed no institutional IACUC approval was required.

Data accessibility

Raw data used in this manuscript are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.7sqv9s4x1 [37]. Whole-genome sequencing was performed by Novogene (Sacramento, CA) on an Illumina platform (NovaSeq 6000 PE150). Raw sequences were mapped by Novogene to the saltwater crocodile, Crocodylus porosus, reference genome with SNPs identified by Novogene using the following the command in SAMtools: mpileup -m 2 -F 0.002 -d 10. Raw data files are uploaded in .vcf, .txt, .csv and STRUCTURE file formats. A full description of how the raw data was analysed can be found in the Genomic Parthenogenesis Assessment section of the electronic supplementary material.

The data are provided in the electronic supplementary material [38].

Authors' contributions

W.B.: conceptualization, data curation, formal analysis, funding acquisition, investigation, methodology, project administration, resources, software, supervision, validation, visualization, writing—original draft, writing—review and editing; B.A.L.: data curation, formal analysis, investigation, methodology, software, validation, visualization, writing—review and editing; J.B.C.: data curation, formal analysis, methodology, writing—review and editing; M.A.D.: data curation, formal analysis, resources, supervision, visualization, writing—review and editing; Q.D.: data curation, methodology, visualization, writing—review and editing; R.D.P.: data curation, investigation, methodology, writing—review and editing; G.W.S.: conceptualization, investigation, supervision, visualization, writing—original draft, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This work was supported by start-up funds provided by the College of Agriculture and Life Sciences at Virginia Polytechnic Institute and State University and the College of Engineering and Natural Sciences at The University of Tulsa to W.B.