Coevolution and Adaptation of Transition Nuclear Proteins and Protamines in Naturally Ascrotal Mammals Support the Black Queen Hypothesis

Abstract

Protamines (PRMs) and transition nuclear proteins (TNPs) are two key classes of sperm nuclear basic proteins that regulate chromatin reorganization and condensation in the spermatozoon head, playing crucial roles in mammalian spermatogenesis. In scrotal mammals, such as humans, cryptorchidism, the failure of the testes to descend into the scrotal sac is generally associated with higher rates of defective spermatozoon quality and function. However, ascrotal mammals, such as cetaceans, with naturally undescended testes, produce normal spermatozoa similar to their scrotal counterparts. This study investigates the evolutionary pattern and functional changes in PRMs and TNPs to explore the potential molecular mechanisms underlying spermatogenesis in naturally ascrotal mammals. Although we found a conserved genomic arrangement for PRM and TNP genes across mammals, the coevolutionary loss of intact PRM2 and TNP2 was observed in several species, correlating significantly with diverse testicular positions. Notably, in cetaceans, which lack intact PRM2 and TNP2, we detected enhanced thermostability and DNA binding in PRM1, along with superior DNA repair capability in TNP1. These findings suggest that gene loss of PRM2 and TNP2, combined with functional enhancements in PRM1 and TNP1 proteins, evolved in response to physiological challenges posed by natural cryptorchidism in most ascrotal lineages. This evolutionary strategy enhances chromatin condensation efficiency and promotes DNA repair during spermatogenesis in natural cryptorchid mammals, supporting the Black Queen Hypothesis.

Introduction

Nuclear chromatin condensation during the terminal stages of spermatogenesis is essential for developing functionally active spermatozoa in mammals (Nishimura and L’Hernault 2017). This condensation ensures sperm motility and normal sperm head morphology, predominantly consisting of a densely compact nucleus (Jakubik-Uljasz et al. 2020). The extensive nuclear chromatin condensation leads to transcriptional silencing and decreased proliferation in spermatozoa, preparing for successful fertilization (Namous et al. 2022).

Transition nuclear proteins (TNPs) and protamines (PRMs) are the primary sperm nuclear basic proteins (SNBPs) responsible for DNA compaction, achieved through sequentially replacing histones (Oliva 2006; Chocu et al. 2012). In mammals, PRM1 and PRM2 encode two types of PRMs, and TNP1 and TNP2 encode two TNPs. In humans, PRM1, PRM2, and TNP2 are tightly clustered in the same chromosomal region, along with PRM3 (also known as “gene 4”), which is not typically associated with PRM function (Balhorn 2007; Kasinsky et al. 2011; Martin-Coello et al. 2011).

The nucleoprotamine structure in spermatozoa is known for protecting genetic material in the sperm head from physical and chemical damage (Braun 2001). For example, TNP1 is involved in repairing DNA single-strand breaks during spermiogenesis (Caron et al. 2001). Genetic mutations or the loss of any of these four genes have been linked to DNA damage, abnormal spermatozoon morphology, and male infertility in humans and mice (Bianchi et al. 2018). For instance, Prm2-deficient mouse spermatozoa exhibit impaired DNA condensation, acrosome formation, and motility (Schneider et al. 2016), and targeted Tnp2 disruption affects chromatin structure and reduces fertility in mice (Zhao et al. 2001). In humans, pathogenic variants in PRM1 have been linked to male infertility across populations (Aoki et al. 2005; Ravel et al. 2007; Nasirshalal et al. 2020). Evolutionary studies have found positive selection on PRM1, PRM2, and TNP2 in domesticated pigs, shaping specific reproductive traits (Wang et al. 2015). Additionally, these genes' evolution appears to be influenced by sperm competition, as suggested by correlations with relative testis mass in mammals (Lüke et al. 2011).

Chromatin condensation abnormalities, often evident as abnormal sperm head morphology, are common in both congenital and artificially induced cryptorchidism in humans and livestock (Mieusset et al. 1987; Chung and Brock 2011; Gatimel et al. 2017). Cryptorchidism, characterized by undescended testes, is a prevalent congenital condition in pediatrics and is often associated with teratospermia and azoospermia. PRM and TNP genes are predicted to be associated with the phenotype of cryptorchidism in diseases such as spermatogenic failure, according to The International Mouse Phenotyping Consortium (www.mousephenotype.org; Groza et al. 2022). Additionally, PRM1, PRM2, and TNP1 gene expression is significantly downregulated in cryptorchidism patients due to reduced spermatogenesis (Sun et al. 2023). Cryptorchidism-induced hyperthermia disrupts sperm chromatin structure and stability, reducing fertilization capacity and increasing infertility risk (Foresta et al. 1992; Ahmad et al. 2012). Interestingly, this contrasts with naturally ascrotal mammals, such as elephants and cetaceans, which despite having undescended testes, produce fertile sperm comparable to their scrotal counterparts (Hutson et al. 1992; Chai et al. 2022). This highlights the complexity of reproductive adaptations and the potential for diverse evolutionary strategies in mammalian fertility, warranting a deeper investigation into the molecular mechanisms underlying normal spermatogenesis in naturally healthy cryptorchid mammals, focusing on the core SNBP genes.

In the present study, the evolutionary pattern of PRM and TNP genes across mammalian phylogeny were explored to elucidate the evolutionary relationship between sperm morphology and diverse testicular positions. We found that in natural cryptorchid mammals, TNP2 and PRM2 inactivation, along with enhanced TNP1 and PRM1 functionality, appears to coevolve. These findings support the Black Queen Hypothesis, a theory of reductive evolution positing that gene loss can provide selective advantages by conserving limiting resources (Morris et al. 2012), suggesting that a strategic reduction in genetic complexity helps maintain spermatogenesis in naturally ascrotal mammals.

Materials and Methods

Sequences, Data Collection, Genomic Locations, and Phylogenetic Trees

The CDSs of PRM1, PRM2, PRM3, TNP1, and TNP2 from 72 representative species across 15 major mammalian orders were retrieved using GeneBank (https://www.ncbi.nlm.nih.gov/genbank/) and BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) applied to the latest genome versions, employing the megablast program with its default parameters (supplementary table S1, Supplementary Material online). CDSs and the corresponding amino acid sequences from cattle (Bos taurus), mice (Mus musculus), and humans (Homo sapiens) were used as queries. The matching nucleotide and protein BLAST results were retained as target sequences.

The codon-based alignments of CDSs were conducted using MACSE with its default parameters (Ranwez et al. 2011). Unreliable and gapped regions in the alignments were identified and trimmed using GBlocks employing the parameters “-t=c -b4=5 -b5=h” (Talavera and Castresana 2007). Phylogenetic trees based on codons were reconstructed using ML methods via the IQ-TREE web server (http://iqtree.cibiv.univie.ac.at/) with its default parameters.

DNA Extraction and PCR Amplification

Muscle tissue samples from baiji (Lipotes vexillifer), Yangtze finless porpoise (Neophocaena asiaeorientalis), bottlenose dolphin (T. truncatus), minke whale (B. acutorostrata), humpback whale (M. novaeangliae), and cattle (B. taurus), previously preserved in our laboratory, were used as templates for genomic DNA extraction, following a standard phenol–chloroform extraction method (Laird et al. 1991). Key reagents, including histiocyte lysate buffer (10 mmol/L Tris-Cl, 500 mmol/L EDTA, and 100 mmol/L NaCl; pH 7.5; Urakawa et al. 2010), proteinase K, Tris-saturated phenol, chloroform/isoamyl alcohol (24:1), and sodium acetate, were purchased from the Shanghai Sangon Biotech Co., Ltd (Shanghai, China), whereas Taq DNA Polymerase was purchased from Takara Biomedical Technology Co., Ltd. (Beijing, China). The anhydrous ethanol used in this study was of analytical reagent grade.

Primers for PRM2 and TNP2 were designed based on genomic data alignments for mice, humans, and cattle (supplementary table S2, Supplementary Material online). PCR was performed in a 25-μL reaction mixture comprising 12.5 μL of 2 × Taq Master Mix (Vazyme), 5 µM of each primer, 9.5 μL of double-distilled water, and 5 to 50 ng of DNA template. The amplification protocol consisted of an initial denaturation at 95 °C for 3 min, followed by 35 cycles at 94 °C for 15 s, 55 °C for 15 s, and 72 °C for 30 s, with a final extension at 72 °C for 5 min. The amplified PCR products were purified and sequenced by Shanghai Sangon Biotech Co., Ltd.

Ancestral State and Sequence Reconstruction

To reconstruct the ancestral presence of TNP2 and PRM2 in mammals, we used the ace function in the R package phytools (Revell 2012). Divergence times and phylogenies for 72 representative species were retrieved from the TimeTree (http://www.timetree.org). Ancestral sequences were reconstructed using PRANK (Löytynoja 2014) with the “-showanc” parameter.

Selective Pressure Test

To evaluate the selective pressure acting on PRM and TNP genes, the ratio of nonsynonymous to synonymous substitution rates (ω = dN/dS) was calculated using codeml in the PAML package (Yang 2007). Phylogenetic topologies from TimeTree (http://www.timetree.org) were employed as working trees.

The ω value provides insights into selective pressures, with ω = 1, ω < 1, and ω > 1 corresponding to neutral evolution, purifying selection, and positive selection, respectively (Nei and Kumar 2000). For intact CDSs, genes exhibiting rapid evolution with significantly higher ω values in foreground branches compared with background branches were identified using a nested branch model (null model: model = 0; alternative model: model = 2). Additionally, the branch-site model was used to identify positively selected genes by comparing a null model where ω is fixed at 1 with an alternative model where ω is free to vary, allowing for positive selection (null model: model = 2, NSsites = 2, fix_omega = 0; alternative model: model = 2, NSsites = 2, fix_omega = 1, omega = 1). Positively selected codons were identified using Bayes empirical Bayes with a P-value > 0.9.

Correlation Analyses

PGLS analysis, which fits a linear model accounting for phylogenetic nonindependence among data points (Freckleton et al. 2002), was used to investigate the association between evolutionary rates and morphological traits/gene numbers. Specifically, root-to-tip and terminal ω values were calculated to account for the full evolutionary history of each terminal branch (Montgomery et al. 2011). First, the free-ratio model in PAML (Yang 2007) was applied to estimate the independent ω value for each leaf branch and ancestral node. Root-to-tip ω, dN, and dS values were then averaged across all internal nodes for each terminal branch. dN, dS, or ω values below 0.00002 were excluded to prevent outliers from skewing the correlation analysis. The caper package in R (Orme et al. 2012) was used to apply the PGLS model, and the phylogenetic signal (λ) was tested via ML. The ultrametric time-calibrated phylogeny used in the PGLS analysis was obtained from TimeTree. P < 0.05 was considered statistically significant.

To assess differences in spermatozoon morphology and gene evolutionary rates between mammals with varying numbers of intact genes, ANOVA accounting for phylogenetic relationships was performed using the phylANOVA function in the R package phytools (Revell 2012), with a P-value threshold of 0.05. The working phylogenetic topology for these analyses was consistent with that used in the selective pressure analyses.

Specific Amino Acid Substitutions and Functional Predictions

We used FasParser (Sun 2017) to identify specific amino acid substitutions in foreground clades. The functional impacts of these substitutions, along with the positively selected sites, were predicted using HOPE (https://www3.cmbi.umcn.nl/hope/; Venselaar et al. 2010).

Wild-Type and Mutant Plasmid Generation, Transient Transfection, and Cell Line Growth

The coding sequences of PRM1 and TNP1 from the bottlenose dolphin and mouse were optimized and synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China) and then cloned into flag-tagged pcDNA3.1 V5-His C and flag-tagged pEGFP-n1 plasmids, respectively. Additionally, mutant plasmids for bottlenose dolphin PRM1 were constructed by substituting Asn (N) with Tyr (Y) at the 4th position and Pro (P) with Arg (R) at the 18th position using a Q5 Site-Directed Mutagenesis Kit. For bottlenose dolphin TNP1, mutants were created by altering Lys (K) to Arg (R) at the 14th position, Gln (Q) to Arg (R) at the 34th position, Asn (N) to Asp (D) at the 45th position, and Ser (S) to Tyr (Y) at the 51st position using the QuickMutation Kit (Beyotime Biotechnology, D0206M) according to the manufacturer's protocol. The primers for both wild-type and mutant plasmid construction are listed in supplementary tables S10 and S11, Supplementary Material online. All genes were verified via sequencing. The plasmids were transfected into HEK293T cells using Lipofectamine 3000 transfection reagent (Invitrogen). All cell lines were sourced from our laboratory and cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum (WISENT) and 1% penicillin/streptomycin at 37 °C with 5% CO₂.

Immunofluorescence Staining

Immunofluorescence staining was performed using cultured cells following standard procedures, as previously described (Li et al. 2016). As the pEGFP-n1 plasmid expresses green fluorescent protein, no antibody incubation was required. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Images were captured using laser scanning confocal microscopy (Zeiss LSM 510 META, Carl Zeiss, Oberkochen, Germany).

Protein Extraction and Western Blots for Thermal Stability

At 48 h post-transfection of pcDNA, pcDNA-PRM1-MUS, pcDNA-PRM1-TUR, and all pcDNA-PRM1 mutants, HEK293T cells were homogenized in RIPA buffer (Beyotime) containing a protease inhibitor mixture (BOSTER) and phosphatase inhibitor mixture (BOSTER). Lysates were sonicated for 30 s and centrifuged at 14,000 rpm and 4 °C for 10 min. The supernatants were divided into four parts and incubated at 4 °C, 35 °C, 37 °C, and 60 °C for 1 h, followed by centrifugation at 14,000 rpm and 4 °C for 10 min. Finally, the supernatants were heated at 95 °C for 5 min in preparation for western blot analysis. Protein samples were separated using sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes. They were blocked with 5% skimmed milk powder and probed with the following antibodies: mouse anti-Flag (Affinity), diluted 1:5000; rabbit anti-β-tubulin (ABclonal), diluted 1:5000; and goat anti-rabbit/mouse IgG (H + L) HRP (ABclonal), diluted 1:5000. Protein levels were semi-quantified from blots using NIH Image J software (National Institutes of Health, Bethesda, MD, USA).

Cell Proliferation Analysis

At 48 h post-transfection, the cells were transferred to 96-well plates. A Cell Counting Kit-8 (CCK8; Beyotime) was used to assess cell growth by measuring absorbance at 450 nm using a Synergy H1 microplate reader (Biotek). Absorbance was measured at 0, 2, 6, 12, and 24 h post-incubation. Results represent three independent experiments, each performed in triplicate.

UV-induced DNA Lesion and DNA Repair Assessments

The degree of DNA damage induced by UV irradiation was evaluated by measuring luciferase gene expression. This was achieved following the transfection of the pGL3-promoter plasmid, which had been exposed to UV light at 960 J/m². HEK293T cells were seeded in 12-well plates, and upon reaching 60% to 70% confluency, 0.9 μg of DNA per well was transfected over 48 h using Lipofectamine 3000 (cat L3000015; Invitrogen). The transfection mixture contained 0.4 μg of the pGL3-promoter construct (Beijing Tsingke Biotech Co., Ltd.), 0.4 μg of the TNP1 plasmid, and 0.1 μg of the pRL-SV40 vector (expressing Renilla luciferase as a control for transfection efficiency; supplementary table S6, Supplementary Material online). Luciferase activity was measured using a dual-luciferase reporter assay system (Vazyme, Nanjing, China), with the Renilla luciferase vector pRL-SV40 serving as an internal reference, adhering to the manufacturer's instructions. Relative luciferase activity was calculated as the ratio of firefly luciferase activity to Renilla luciferase activity, with the control samples (pGL3-Pro-UV-0s-pRL) normalized to 100%. Each transfection experiment was performed independently at least three times, each time in triplicate.

Statistical Analysis

For data from western blotting, CCK8, and dual-luciferase reporter assays, statistical analyses were conducted using an unpaired two-tailed Student's t-test with a 95% confidence interval, assuming normal data distribution. Analyses were performed using Prism 9 (GraphPad). Data are presented as means ± SDs, with group sizes indicated in the main text. Student's t-test was used to compare two groups, and significant differences between groups are denoted in figures by asterisks (*P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant).

Results

Genomic Landscape of PRM and TNP Genes

We retrieved the coding sequences (CDSs) of 318 PRM and TNP genes from the genomes of 72 mammals and examined their genomic distribution (supplementary table S3, Supplementary Material online). Overall, the genomic landscape of the PRM gene cluster (TNP2 > PRM1 > PRM2 > PRM3) and TNP1 showed conservation across representative genomes. Notably, intact PRM1, PRM3, and TNP1 were present in most mammals (Fig. 1a; supplementary table S1; Supplementary Material online). Conversely, PRM2 and TNP2 were absent in certain species. Specifically, PRM2 was missing from the monotreme, marsupial, pangolin, and eulipotyphlan genomes, whereas partial PRM2 sequences were found in cetacean genomes. Similarly, TNP2 sequences were absent in the eulipotyphlan, cetacean, monotreme, and marsupial genomes, whereas a partial TNP2 sequence was found in the pangolin genome.

We also identified incomplete CDSs of cetacean PRM2 via polymerase chain reaction (PCR) amplification. CDSs of PRM2 from four cetaceans, including the humpback whale, bottlenose dolphin, Yangtze finless porpoise, and Yangtze River dolphin, as well as from cattle, were successfully amplified and sequenced (Fig. 1b and c; supplementary table S2, Supplementary Material online). The failure to fully amplify sequences from the minke whale (Balaenoptera acutorostrata) and humpback whale (Megaptera novaeangliae) may have been due to primer incompatibility for certain baleen whales or the relatively poor quality of DNA templates from the samples. Besides segment deletions of various lengths, we found that one-base-pair (bp) and two-bp frame-shift mutations in PRM2 were shared among the cetaceans studied. We then reconstructed ancestral PRM2 sequences showing that these mutations were present in all cetacean ancestral branches but not in other ancestral and terminal branches of their sister group, terrestrial artiodactyls (supplementary fig. S1, Supplementary material online). This suggests that gene loss likely occurred in an early ancestor of cetaceans. Ancestral state reconstruction further supported this notion, with the optimal model (all rates different) predicting the loss of PRM2 and TNP2 in the cetacean ancestor (node 26 in Fig. 2). These results imply that PRM2 and TNP2 have different evolutionary histories in mammals. PRM2 likely originated in an ancestor of mammals, whereas TNP2 may have evolved separately in different eutherian lineages. Furthermore, the maximum likelihood (ML) phylogeny of the PRM gene cluster inferred from codon sequences placed PRM3 outside the TNP2, PRM1, and PRM2 cluster (supplementary figs. S2 and S3, Supplementary Material online), suggesting the possible functional divergence of PRM3 from the PRMs.

Selective Pressure, Specific Amino Acid Substitutions, and Putative Functional Effects on PRM and TNP Genes in Natural Cryptorchid Mammals

To explore the relationship between PRM and TNP genes and testicular positions in mammals, we first examined the selective pressure on PRM2 and TNP2, which have undergone molecular decay in several natural cryptorchid mammals. Pairwise comparisons of branch models [one-ratio versus two-ratio, two-ratio versus two-ratio ([fix ω = 1]) revealed significantly relaxed selective pressure on PRM2 in the ascrotal rodents, rhinoceros, and pinnipeds studied, whereas TNP2 showed relaxed selective pressure in the ascrotal star-nosed mole and bats examined (Table 1). This suggests that, although molecular decay of PRM2 and TNP2 did not occur in these highly diverged ascrotal lineages, the accumulation of inactivating mutations might have lagged behind relaxed selective pressure.

To assess evolutionary patterns influenced by the absence of intact PRM2 and TNP2, we tested selective pressure on PRM1 or TNP1 in lineages lacking intact PRM2 and TNP2 (i.e. the cetaceans, eulipotyphlans, pangolin, monotremes, and marsupial studied) as well as their close relatives (i.e. the terrestrial artiodactyls, bats, horse, carnivores, and elephant examined; supplementary table S3, Supplementary Material online). PRM1 in eulipotyphlans and TNP1 in the pangolin showed rapid evolutionary rates not observed in their close relatives. Furthermore, in the Artiodactyla dataset, positive selection on both PRM1 and TNP1 in cetaceans, but not in closely related lineages, was detected (supplementary table S4, Supplementary Material online), indicating lineage-specific evolutionary histories and highlighting the role of background branches in evaluating selective pressure. We then tested selective pressure on branches and codon sites using the branch-site model, identifying different positively selected codons in lineages lacking intact PRM2 or TNP2 compared with those with intact genes (supplementary table S3, Supplementary Material online).

We also identified several amino acid substitutions specific to natural cryptorchid lineages (Fig. 3). Among them, TNP1 R14K, R34Q/W, Y51C/S, and PRM1 Y4F/R/N/T were shared across more than two lineages. These substitutions likely affect protein function due to changes in physicochemical properties, such as volume and hydrophobicity (Fig. 3). Additionally, a specific substitution was found in the DNA-binding region of PRM1, which could have functional significance (Ren et al. 2021; Arévalo et al. 2022).

Coevolution of PRM2 and TNP2 Is Associated With Mammalian Testicular Position

To explore whether the evolution of PRM genes is linked to TNP genes, we conducted an analysis of variance (ANOVA) based on the presence of both gene classes. Results showed that the absence of one gene significantly impacted the absence of the other (P = 0.001; supplementary fig. S4, Supplementary Material online), suggesting that a coevolutionary relationship exists between PRM2 and TNP2. This relationship was not influenced by phylogeny, as the phylANOVA results ruled out phylogenetic signals as a factor in the variance.

We subsequently investigated whether the presence of PRM2 and TNP2 correlated with divergent mammalian testicular positions. After accounting for phylogenetic relationships, results from phylANOVA indicated that testicular position (considering both scrotum development and testicular descent) was a significant predictor for the presence or absence of both PRM2 and TNP2 loci (supplementary fig. S5, Supplementary Material online). This finding suggests that the evolution of PRM and TNP genes may be associated with variation in testicular positions across mammals.

Cetacean PRM1 Exhibits Superior Thermal Stability

To determine whether TNPs and PRMs in naturally ascrotal mammals evolved to withstand higher testicular temperature, we employed western blot assays to compare PRM1 thermostability in a cetacean (the bottlenose dolphin, Tursiops truncatus) and mouse (Mus musculus), representing ascrotal and scrotal mammals, respectively. Results revealed no significant differences in PRM1 thermostability between the two species at lower temperatures (4 °C and 35 °C). However, at a higher temperature (37 °C), cetacean PRM1 exhibited significantly greater stability compared with mouse PRM1 (Fig. 4a and b; supplementary fig. S6, Supplementary Material online). Enhanced thermostability in cetacean PRM1 persisted even at an extreme temperature of 60 °C. In contrast, the stability of dolphin TNP1 was no greater than that of mouse TNP1 (supplementary fig. S7, Supplementary Material online).

We also conducted reverse mutation experiments to assess whether unique amino acid substitutions in cetaceans affected PRM1's thermal stability. Although individual reverse mutations exerted minimal effects, dual-site reverse mutations (N4Y and P18R) markedly reduced the thermal stability of cetacean PRM1 at 37 °C, suggesting that Y4 and R18 play a key epistatic role in maintaining PRM1 stability (Fig. 4c and d; supplementary figs. S8 and S9, Supplementary Material online).

Mammals Without Intact PRM2 or TNP2 Maintain General Nuclear Chromatin Condensation

PRMs and TNPs are essential for chromatin condensation in spermatozoa, and their absence often leads to fertility issues and abnormalities in sperm head morphology, particularly affecting head size and elongation (Lüke et al. 2014). The mammalian sperm head, primarily comprising the nucleus, and nuclear size correlate strongly with genome size (Fawcett 1958; Gregory 2001). In spermatozoa, DNA content is reflected in the organism's C-value, which measures the total DNA in its haploid chromosome set and is often used in genome size measurements (Greilhuber et al. 2005). To determine whether differences in PRM and TNP gene integrity impact nuclear chromatin condensation efficiency in mammals, we used sperm head length as a proxy of nuclear chromatin condensation and employed phylANOVA analysis. Additionally, to control for genome size effects on sperm head dimensions, we analyzed the ratio of sperm head length to C-value (Ausió et al. 2007; supplementary table S5, Supplementary Material online). In the Mammalia dataset, species with only one intact PRM1 or TNP1 gene did not show significant differences in sperm head length or chromatin condensation efficiency compared to species with two PRM or TNP genes (Fig. 4e). However, in the Artiodactyla dataset, cetaceans displayed a lower sperm head length/C-value ratio and shorter sperm head lengths compared with their counterparts, indicating more efficient DNA condensation driven by cetacean PRM1 and TNP1 in the absence of PRM2 and TNP2 (Fig. 4f).

To further investigate the superior nuclear chromatin condensation by cetacean PRM1, we expressed different PRM1 variants in somatic cell systems and measured their PRM–DNA-binding efficiency. Immunofluorescence staining confirmed the predominant nuclear expression of both cetacean and mouse PRM1 in transfected cells (Fig. 4g). Thus, we monitored cell proliferation rates as an indicator of PRM1's DNA-binding effects, finding that cells expressing cetacean PRM1 (PRM1-TUR) exhibited a slower growth rate compared with those expressing mouse PRM1 (PRM1-MUS) after 12 h, indicating that cetacean PRM1 more effectively inhibits cell proliferation by binding and condensing DNA (Fig. 4h), at least at 37 °C. Further mutational analysis, including both single- and multipoint mutations, revealed that specific PRM1 N4 and P18 mutations in cetaceans markedly enhance DNA-binding capacity (Fig. 4h), whereas single-point mutations do not significantly impact this capacity (supplementary fig. S10, Supplementary Material online).

DNA Repair Capacity of Cetacean TNP1

To investigate potential functional differences in the DNA repair abilities of TNP1 in natural cryptorchid mammals, particularly in the absence of TNP2, we designed an ultraviolet (UV)-induced DNA damage assay based on previously established methods (Caron et al. 2001). This assay aimed to comparatively evaluate the DNA damage tolerance of TNP1 across representative species. Initially, we assessed the effects of UV irradiation on DNA damage over various exposure times. Results revealed that compared with control vector reporter plasmids unexposed to UV and those exposed for 15, 20, and 30 s, a 60-s UV exposure caused significant damage to the plasmid DNA (Fig. 5a).

We subsequently examined the dual-luciferase reporter gene activity in HEK293T cells transfected with plasmids carrying different versions of TNP1 to assess their DNA repair capabilities (supplementary table S6, Supplementary Material online). We observed that, although the reporter gene activity in cells expressing mouse TNP1 (MUS-TNP1) did not show a notable increase, cells expressing cetacean TNP1 (TUR-TNP1) restored the luciferase activity of the UV-damaged reporter, highlighting the remarkable DNA repair capacity of cetacean TNP1 (Fig. 5b). Notably, cells expressing single-mutation TNP1 exhibited similar expression levels compared with cells expressing wild-type TUR-TNP1. However, the luciferase activity in UV-damaged reporter genes was notably reduced in cells expressing the four-point mutation version of TNP1, indicating a potential epistatic functional interaction among these mutation sites (Fig. 5b). Additionally, TNP1 with three-point mutations demonstrated varying levels of luciferase activity (supplementary fig. S11, Supplementary Material online), indicating that specific mutations may have diverse effects on DNA repair functions.

Correlation Between PRM and TNP Gene Evolution and Sperm Competition

To explore the influence of sperm competition on PRM and TNP gene evolution, we correlated gene numbers and lineage-specific gene evolutionary rates (including ω, dN, and dS) with relative testicular mass values (testicular mass to body mass [TM_r]). Phylogenetic generalized least squares (PGLS) analysis, accounting for phylogenetic effects, revealed that the evolution of TNP1 and PRM1 did not appear to be influenced by sperm competition (Fig. 6). Specifically, neither the number of PRM and TNP genes nor their evolutionary rates showed a significant association with TM_r, both in the broader mammalian dataset and artiodactylan-specific analysis (supplementary tables S7 and S8, Supplementary Material online).

Discussion

Spermatogenesis is conserved in mammals and includes a hallmark step: the replacement of nuclear somatic-like histones by TNPs and PRMs (Champroux et al. 2016). This replacement facilitates chromatin remodeling and condensation in the sperm head, a process crucial for spermatozoon function and fertility (Braun 2001). In the current study, the conservation of PRM and TNP gene sequences across a broad spectrum of mammals highlights the critical role these SNBP genes play in maintaining spermatogenesis and ensuring proper spermatozoon chromatin condensation at the molecular level (Fig. 1). Our gene retrieval and ancestral state reconstruction analyses indicated that PRM2 and TNP2 likely evolved after mammals' common ancestor (Fig. 2), aligning with the notion that TNP genes are unique to mammals and that PRM2 predates TNP2 within the PRM gene cluster (Ausió et al. 2007; Kasinsky et al. 2011).

Previous studies have highlighted the essential role of PRM and TNP genes in reproductive function, with gene knockouts in vivo resulting in reduced fertility (Champroux et al. 2016). Our research revealed that PRM2 and TNP2 have undergone gene loss, inactivating mutations, or relaxed selection pressure in various mammalian lineages, including cetaceans, monotremes, and afrotherians (Fig. 1; Table 1). Notably, the loss of these two intact genes appears to be a coordinated event rather than independent occurrences, as evidenced by significant coevolutionary patterns (supplementary fig. S3, Supplementary Material online). Additionally, this joint gene inactivation correlates with differing testicular positions, a key mammalian reproductive trait (supplementary fig. S4, Supplementary Material online), suggesting that natural cryptorchid mammals may be more susceptible to molecular decay or relaxed selection of PRM2 and TNP2. However, we found no significant association between sperm head dimensions, a proxy selected for sperm nuclear DNA condensation in the present study, and variation in PRM and TNP gene numbers (Fig. 4e and f). Further studies involving more direct measures of nuclear DNA condensation, such as nuclear size, will be essential.

Evidence indicates that naturally ascrotal mammals may have evolved compensatory mechanisms to maintain normal spermatogenesis despite the loss of intact PRM2 or TNP2. First, we identified signs of positive selection and accelerated evolutionary rates in PRM1 and TNP1 within lineages lacking these intact genes (Fig. 3; supplementary table S3, Supplementary Material online). Positively selected codons or specific amino acid substitutions located in functionally critical domains were predicted to affect the functionality of PRM1 and TNP1 proteins. Based on these findings, we propose that PRM1 and TNP1 in natural cryptorchid mammals have undergone adaptive evolution, enhancing their functional capacities.

In mammals, spermatogenesis is highly sensitive to temperature (Bedford 1991). Indeed, an increase in testicular temperature of only 2 °C to 4 °C above normal can halt spermatogenesis or result in spermatozoon deformities (Abdelhamid et al. 2019). Our in vitro functional assays demonstrated that, in cryptorchid cetaceans, PRM1 exhibited superior thermostability and chromatin condensation capacity compared with scrotal mouse PRM1 (Fig. 4). Moreover, we confirmed that the functional differences observed between cetacean and mouse PRM1 are primarily due to amino acid substitutions at positions N4Y and P18R. Similarly, variations in UV-induced damage repair capacity between cetacean and mouse TNP1 are attributed to substitutions at multiple sites (Fig. 5), suggesting an epistatic effect consistent with evolutionary patterns observed in other genes (Gros et al. 2009; Gai et al. 2023).

In conclusion, although the absence of intact PRM and TNP genes in natural cryptorchid mammals resembles the significantly reduced expression of these genes in patients with cryptorchidism (Nguyen et al. 2009), we propose that ascrotal mammals have developed alternative evolutionary strategies to maintain normal spermatogenesis. Although chromatin condensation is vital for spermatogenesis, the subsequent decondensation of paternal DNA also plays a key role in nuclear reprogramming and fertilization. The more complex and condensed the chromatin, the more time and energy are required for decondensation (Champroux et al. 2016; Ribas-Maynou et al. 2021). Strengthening the function of one PRM and TNP, rather than relying on two, may offer a selective advantage by conserving limiting resources, as postulated in the Black Queen Hypothesis (Morris et al. 2012). Furthermore, TNP1 has been suggested to stimulate single-strand DNA break repair (Akama et al. 1999; Caron et al. 2001), potentially enhancing the integrity of genetic material in spermatozoa. This is supported by the high quality of dolphin sperm cells, which typically exhibits few abnormal spermatozoa (Van der Horst et al. 2018). Thus, the efficient, thermostable chromatin condensation–decondensation system established by the singular but functionally strengthened PRM and TNP in natural cryptorchid mammals, at least regarding the ascrotal cetaceans studied, may represent an evolutionary strategy for ensuring normal spermatogenesis and reproductive success. Future studies should include a broader range of natural cryptorchid and scrotal mammals for further investigation.

Although preliminary, our results suggest a link between the molecular decay of PRM2 and TNP2 and the diverse testicular positions observed in mammals. Moreover, we provide evidence supporting the hypothesis that PRM1 and TNP1 have adapted in response to these changes. Further research is warranted to explore how other factors influence the evolution of different SNBP genes across mammals.

Supplementary Material

Supplementary material is available at Genome Biology and Evolution online.

Author Contributions

G.Y. designed the study. S.C., T.W., and X.Z. were responsible for the data collection and analysis. J.K. and Y.Z. carried out the experiments. S.C. and J.K. drafted the manuscript. S.C. and G.Y. revised the manuscript. W.R. and S.X. helped revised the manuscript. All authors read and approved the final manuscript

Funding

The Youth Fund of the National Natural Science Foundation of China (grant no. 32200345 to S.C.). China Postdoctoral Science Foundation (grant no. 2022M710878 to S.C.). Guangzhou Basic and Applied Basic Research Foundation (grant no. 2023A04J0770 to S.C.). The Key Project of the National Natural Science Foundation of China (grant no. 32030011 to G.Y.). PI Project of Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou) (grant no. GML2021GD0805 to G.Y.). The National Natural Science Foundation of China (grant no. 31872219 to W.R.).

Data Availability

The data underlying the article are available in the article and in its online Supplementary material.

Literature Cited

Author notes