Protamine sequence determines species-specific nuclear shape and histone retention

Article iScience Protamine sequence determines species-specific nuclear shape and histone retention Graphical abstract Highlights •Protamine sequence, species specific, plays a role in sperm nuclear morphology •Protamine sequence effects, notably two cysteines, emerge in the in vitro somatic model •In vivo, sequence variation has reduced but detectable effects on sperm head shape •Discovery of higher histone retention with mutant protamine opens new research paths Authors Marta Czernik, Luca Palazzese, Dawid Winiarczyk, ..., Eduardo R. S. Roldan, Maria Eugenia Teves, Pasqualino Loi Correspondence [email protected] (E.R.S.R.), [email protected] (P.L.) In brief Cell biology; Developmental biology Czernik et al., 2025, iScience 28, 113102 August 15, 2025 ©2025 The Author(s). Published by Elsevier Inc. https://doi.org/10.1016/j.isci.2025.113102 ll iScience Article Protamine sequence determines species-specific nuclear shape and histone retention Marta Czernik, 1,2,8 Luca Palazzese, 1,8 Dawid Winiarczyk, 2 Domenico Iuso, 1 Saadi Khochbin, 3 Josef Fulka, 4 Helena Fulka, 5 Rocı´o Villafranca, 6 Juan Antonio Rielo, 6 Ana Sanchez-Rodriguez, 6 Nerea Latorre, 6 Clara Agudo-Rios, 6 Eduardo R.S. Roldan, 6,8, *Maria Eugenia Teves, 7 and Pasqualino Loi 1,8,9, * 1 Laboratory of Embryology, Department of Veterinary Medicine, University of Teramo, 64100 Teramo, Italy 2 Institute of Genetics and Animal Biotechnology, Experimental Embryology Department, J05-552 Jastrzebiec, Poland 3 University of Grenoble Alpes, CNRS UMR 5309, INSERM U1209, Institute for Advanced Biosciences, La Tronche 38706, France 4 Institute of Animal Science, Prague, Czech Republic 5 Institute of Experimental Medicine of the Czech Academy of Sciences, Prague, Czech Republic 6 Department of Biodiversity and Evolutionary Biology, Museo Nacional de Ciencias Naturales (CSIC), Madrid, Spain 7 Department of Obstetrics and Gynecology, Virginia Commonwealth University, Richmond, VA, USA 8 These authors contributed equally 9 Lead contact *Correspondence: [email protected] (E.R.S.R.), [email protected] (P.L.) https://doi.org/10.1016/j.isci.2025.113102 SUMMARY The nuclear shape observed after the forced expression of mouse or human protamine 1 (PRM1) in fibroblasts led us to propose the hypothesis that the PRM1 sequence plays an important role in imposing the overall shape of the protaminized nucleus. Comparison of mouse and human PRM1 sequences pointed to cysteines 15 and 29 as potential critical residues in the mouse PRM1 sequence, inducing the characteristic mouse ‘‘hooked’’ nuclear sperm shape. To explore this idea, mice with mutations in PRM1 Cys15 and Cys29 were generated. These mice remained fertile with no significant changes in sperm count or protamine expression levels. However, modifications in sperm head shape were observed. Transmission electron microscopy revealed disrupted chromatin condensation in mutant sperm, with several morphological changes and a remarkable increase in histone retention. Overall, the findings suggest that species-specific PRM1 cysteine residue positions are crucial for nuclear shape determination and histone retention in spermatozoa. INTRODUCTION In contrast to the invariable spherical shape of the oocyte, the morphology of spermatozoa varies widely across species, particularly with regard to head shape. 1–3 Several selective forces have been proposed to account for sperm diversity, including postcopulatory sexual selection in the forms of sperm competition or cryptic female choice. 1,3 Thus, when spermatozoa from rival males compete to gain fertilizations, changes in sperm head shape may confer hydrodynamic advantages in swimming 4,5 or sperm interactions with the female tract. Selection by females may also be based on sperm morphology, and therefore, some head morphs may become better suited to negotiate barriers in the female tract. 6 Studies in the mouse have demonstrated that the cytoskeleton plays a pivotal role in shaping nuclear morphology during sperm differentiation. At the beginning of the spermatid elongation phase, the nucleus and the early acrosome polarize to one end of the cell. 7 As early as in step 5 of spermiogenesis, the F-actin filaments and associated proteins forming the acroplaxome prime nuclear shaping. Next, tensions generated by the microtubules and actin filaments in the manchette and the interaction with the nuclear envelope through the linker of nucleoskeleton and cytoskeleton complex provide a major shaping factor. 8 At this point, the nucleus compacts progressively, as the nucleosomal chromatin is transformed into compacted fibers by the sequential replacement of histones by testis-specific histone variants, transition proteins, and protamines (PRMs) (either PRM1 alone in most mammals, or PRM1 and 2 in primates and rodents and a few other species). 9 However, the possibility that chromatin compaction could have a more relevant role in nuclear shaping has been already suggested. 3 Our previous studies exploring strategies for nuclear reprogramming of somatic cell nuclei 10 have found that, when somatic cells express a heterologous mouse or human PRM1 gene (Prm1, PRM1), their nuclei acquire a sperm-like shape reminiscent of the species, that is, curved in fibroblasts expressing mouse Prm1 and straight with human PRM1. 11–13 These findings suggest that the PRM sequence might play a role in shaping sperm nuclei. The PRM1 protein sequence is not well conserved between mammals, which can lead to different 3D structures of the protein in various species. iScience 28, 113102, August 15, 2025 © 2025 The Author(s). Published by Elsevier Inc. 1 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). ll OPEN ACCESS Mammalian PRM1 is typically 50 amino acids long, and it is made of a central arginine-rich domain that binds to 11 bp located in the major groove of the DNA. 14 Both ends of PRM1 have a cysteine-rich domain. Intraand inter-PRM disulfide bonds between cysteine residues are responsible for the condensation of DNA in the final maturation phases of the male gamete. However, how the DNA-bound PRMs cross-link through S–S bonds and compact remains a matter of debate. Moreover, the surprising lack of crystallographic data of PRM/DNA complexes further complicates modeling work. In this study, we hypothesized that the position of these S–S bonds along the protein might have an impact on the overall structure and shape of the resulting compacted genome. To test this hypothesis, we first explored whether there are differences in nuclear remodeling depending on the source of PRM1 (mouse, mPrm1, or human, hPRM1) by quantifying changes in the nuclear shape induced by PRM expression in mouse or sheep fibroblasts. Because mouse and human PRM1 differ in the number and position of cysteine residues, we also examined whether the presence/absence of certain cysteines affects nuclear remodeling. To this end, we progressively substituted cysteine residues in the mouse gene by site-directed mutagenesis and expressed the mutant vectors in sheep and mouse fibroblasts. Based on these findings, PRM1 mutant mice were generated by CRISPR and used to investigate the effect of the lack of cysteine residues in the mouse PRM1 sequence on sperm head morphological traits, nuclear compaction, and fertility. Our findings suggest that, while not compromising overall fertility, mouse PRM1 mutant spermatozoa display head morphological differences, in shape as well as in thickness, probably as a consequence of a change in nuclear compaction. A reduced compaction might have been the consequence of a higher histone retention in mutated spermatozoa, in comparison to the control. Overall, our data indicate that the PRM sequence affects head morphology and the PRM-to-histone ratio in spermatozoa and raise the possibility that the PRM sequence, specifically the number and the position of cysteines, might be important for the wide morphological variations of sperm heads across species. RESULTS PRM1 drives fibroblast nuclear remodeling and the role of cysteine residues in nuclear shape We have shown that the expression of PRM1 significantly influences nuclear morphology in fibroblasts, reflecting species-specific sperm nuclear shapes (Figure 1). We categorized fibroblasts with PRM1 expression into two nuclear morphologies: ‘‘straight’’ (Figure 1R) and ‘‘curved/hocked’’ (Figures 1S and 1T). Nuclei classified as ‘‘hooked’’ display two poles that converge and do not align with a tangent passing through the nucleus center. Conversely, straight nuclei lack both convergence and divergence of the poles and align along a linear axis intersecting the nucleus center. These categories reflect the distinct nuclear condensation states observed at the fully protaminized stage (Figures 1R– 1T). Only cells that exhibited clear nuclear condensation, indicating successful PRM incorporation, were included in this analysis. The result showed that when hPRM1 was expressed in sheep (Figure 1A) or mouse (Figure 1B) fibroblasts, the nuclei largely adopted a compact, straight shape (Figures 1A–1C). As shown in Figure 1D, over 80% of those cells exhibited a straight nuclear morphology (Figures 1A–1D). In contrast, expression of mouse PRM1 predominantly led to curved/hooked nuclei both in sheep (Figure 1E) and mouse (Figure 1F) fibroblasts. This effect was particularly observed in mouse fibroblasts, where over 70% of cells displayed a hooked shape (Figure 1H). The remodeled nuclear shapes closely resembled the mature sperm nuclear morphology typical of each species (Figures 1M–1P and 2). These observations were closely tied to PRM1 expression, as expression of the tag proteins (GFP or RFP) alone had no impact on nuclear remodeling (Figure 1Q). This observation led us to hypothesize that PRM1 may underlie these morphological differences. To investigate this, we compared the amino acid sequences of mouse and human PRM1, noting a relatively low sequence homology of 62% (Figure S1A). One notable difference is the presence of three additional cysteine residues in mouse PRM1. Given that cysteines can form disulfide bonds, stabilizing protein structures, these residues may influence chromatin compaction and potentially contribute to the distinct hooked nuclear shape observed in murine sperm. To test the role of these cysteines, we performed site-directed mutagenesis to replace each additional cysteine. In the first mutant vector (Mut1-Prm1), we replaced cysteine at position 15 with tyrosine (C15Y), which aligns with the human sequence. We also substituted glycine at position 44 with alanine and cysteine at position 45 with tyrosine. In the second mutant vector (Mut2-Prm1), we introduced an additional substitution by replacing cysteine at position 29 with serine, resulting in a doublemutant form (Table S1; Figure S1B). To evaluate the effects of these cysteine modifications on nuclear shape, we transfected both sheep and mouse fibroblasts with either Mut1-Prm1 or Mut2-Prm1. Transfection of sheep fibroblasts with Mut1-Prm1 resulted in a straight nuclear morphology in approximately 70% of cells (Figure 1L), while Mut2-Prm1 increased this proportion to 77.3% (Figures 1I–1L). These findings support the hypothesis that cysteine residues in mouse PRM1 contribute to the species-specific curved nuclear morphology and that modifying these residues induces a straighter nuclear shape, more akin to human sperm. Mutation of PRM1 cysteine residues (C15 and C29) does not affect male fertility To test our hypothesis of the critical role of C15 and C29 in imposing the mature sperm nuclear shape, we genetically modified mice using the CRISPR-Cas9 system to explore the in vivo function of cysteine residues (C15 and C29) of mouse PRM1 in sperm DNA condensation and packaging, fertility, and sperm head morphology. A Mut2-Prm1 mutant was generated by intercrossing the founder individuals and the progeny of mice used for further breeding for at least three generations. Mut2-Prm1 male mice were fertile and sired offspring with an average litter size of 5.8 pups/litter, which is within the range of the reproductive performance of wild-type (WT) C57BL/6 mice of 7.2 pups/ litter. 15,16 There were no statistical differences between WT 2 iScience 28, 113102, August 15, 2025 iScience Article ll OPEN ACCESS and Mut2-Prm1 mice, with averages of 7.2 and 5.8 pups/litter, respectively (p = 0.0797 based on 17 litters). This indicates that the mutations in cysteine residues (C15 and C29) of PRM1 did not affect reproductive performance in these mice. Histological analysis of the testes showed no alterations in the seminiferous tubules, which contained the expected balance of germ cells (Figure 2A). Additionally, there were no distinguishable differences between the spermatids from Mut2-Prm1 and WT adult mice (n = 3 each), suggesting no alterations during the process of spermiogenesis (Figure 2A). Next, levels of PRM expression were investigated in WT and mutant mice. RNA samples used for gene expression had optimal purity (A260/280 > 2). There were no significant differences in the expression of PRM mRNA in WT (Prm1: 1.07 ± 0.29; Prm2: 1.07 ± 0.29) and Mut2Prm1 mice (Prm1: 0.94 ± 0.39; Prm2: 1.53 ± 0.39) (p > 0.05) (Figures 2B–2C). The ratio between Prm1 and Prm2 was also not significantly different between groups (WT: 0.97 ± 0.04 vs. Mut2-Prm1: 0.59 ± 0.16; p > 0.05; Figure 2D). These results suggest that the Prm1 mutation causes no major changes in the expression of Prm1 or Prm2 mRNA. In the final stages of spermiogenesis, PRMs are transported from the cytoplasm to the nucleus, where they replace histones to facilitate chromatin condensation. 17,18 This process is critical for sperm development and function. To investigate whether Figure 1. Remodeling of somatic cell nucleus driven by protamine 1 (A and B) Representative photograph of two-dimensional structure of sheep (A) and mouse (B) fibroblasts transfected with human protamine 1 tagged with RFP (hPrm1-RFP). Left: bright field; blue: nuclei in cells stained with Hoechst 33342; red: nuclei with hPrm1-RFP. Scale bars: 5 μm. (C) Three-dimensional structures of straight somatic nucleus in fibroblasts expressing hPrm1-RFP: front view (on the top) and side view (on the bottom). (D) Percentages of straight and curved (hooked) nuclei in sheep and mouse fibroblasts expressing human protamine vector (****p < 0.0001 for differences between straight and hooked nuclei). Data are represented as mean ± SEM. (E and F) Representative photograph of two-dimensional structure of sheep (E) and mouse (F) fibroblasts transfected with mouse protamine 1 tagged with GFP (mPrm1-GFP). Left: bright field, blue: nuclei in cells stained with Hoechst 33342; green: nuclei with mPrm1-GFP. Scale bars: 5 μm. (G) Three-dimensional structures of hooked somatic nucleus in fibroblasts expressing mPrm1-GFP: front view (on the top) and side view (on the bottom). (H) Percentages of straight and curved (hooked) nuclei in sheep and mouse fibroblasts expressing mouse protamine vector (**p < 0.05, ****p < 0.0001 for differences between straight and hooked nuclei). Data are represented as mean ± SEM. (I and J) Representative photograph of two-dimensional structure of sheep (I) and mouse (J) fibroblasts transfected with Mut2-Prm1 tagged with GFP (Mut-2Prm1). Left: bright field, blue: nuclei in cells stained with Hoechst 33342; green: Mut2-Prm1-GFP. Scale bars: 5 μm. (K) Three-dimensional structures of straight somatic nucleus in fibroblasts expressing Mut2-Prm1: front view (on the top) and side view (on the bottom. (L) Percentages of straight and curved (hooked) nuclei in sheep and mouse fibroblasts expressing mutated protamine vector (**p < 0.05, ****p < 0.0001 for differences between straight and hooked nuclei). Data are represented as mean ± SEM. (M) Two-dimensional structure of somatic cells (control, non-protaminized). (N–P) Three-dimensional structures of somatic cell nucleus (N), ram sperm nucleus (O), and mouse sperm nucleus (P) in front view (left), side view (right), and rotated 45◦(center). (Q) Negative control group: fibroblasts transfected with a GFP-expressing plasmid lacking the Prm1 gene. Scale bars: 20 μM. (R–T) Representative images illustrating the appearance of ‘‘fully protaminized’’ cells (arrowheads) exhibiting either the straight (R) or hooked (S and T) nuclear morphology, following transfection with hPrm1-GFP, Mut2-GFP, and mPrm1-GFP, respectively. Scale bars: 10 μm. iScience 28, 113102, August 15, 2025 3 iScience Article ll OPEN ACCESS Figure 2. Mutation of PRM1 in cysteine residues (C15 and C29) does not alter spermiogenesis or protamine expression (A) Testes from adult wild-type (n = 3) and Mut2-Prm1 (n = 3) mice were collected for histological studies. Sectioned slides were stained with hematoxylin and eosin and evaluated using a bright-field microscope. No differences were observed between control and Mut2-Prm1 testes. Representative images for wild-type and Mut2-Prm1 are shown in the figure. (B) Prm1 gene expression in wild-type and Mut2-Prm1 mice (n = 3). (C) Prm2 gene expression in wild-type and Mut2-Prm1 mice (n = 3). (legend continued on next page) 4 iScience 28, 113102, August 15, 2025 iScience Article ll OPEN ACCESS mutations in the cysteine residues (C15 and C29) of PRM1 impact the nuclear transport of this protein, subcellular immunolocalization of PRMs was examined (Figure S2). Quantitative analysis showed no significant differences between WT and Mut2-Prm1 mice in the percentage of either PRM1 or PRM2 localized within the nucleus (Figures 2E and 2F). These findings suggest that the mutations in the cysteine residues of PRM1 do not interfere with its transport to the nucleus during spermiogenesis. Substituting cysteine residues in the PRM1 sequence results in modifications in sperm head morphometry To further investigate the phenotype of Mut2-Prm1 mice, we examined the morphological characteristics of their spermatozoa. Spermatozoa from WT and Mut-Prm1 mice were examined using microscopy, and the proportion of normal sperm and sperm with abnormalities in each component (head, midpiece, principal plus terminal piece) was quantified (Figure S3). We found that, when compared to WT males, Mut2-Prm1 mice only exhibited a slight increase in the percentage of head abnormalities (p = 0.0586; t = 2,623, df = 4) (Table 1). The proportion of abnormalities in the flagellum (midpiece and principal piece) did not significantly differ between WT and Mut2-Prm1 mice. Proportions of abnormal sperm were in the range reported previously for adult C57BL/6 mice. 19–21 Nuclear morphology was analyzed using methods and software developed by Skinner et al. 22 (Figures 3A–3C). This software automatically identifies nuclei and locates key landmarks for orientation and measurement. The nuclear shape is then described using internal angles around the perimeter, generating an ‘‘angle profile’’ (see STAR Methods for details of methodology). Overall, results showed minor shape differences between sperm from WT and Mut2-Prm1 homozygous mice using this tool (Figure 3B). The highest variation in the angle profile was seen around position 350 of the profile (Figure 3D), which corresponds to the flagellum attachment region and the nuclear basal region, as validated previously. 22,23 The angle profiles identified five segments in the perimeter of sperm nuclei from WT and Mut2-Prm1 mice (Figure 3C). The length of three of these segments exhibited statistically significant differences, with lower values in Mut2-Prm1 nuclei in comparison to those of their WT counterparts (Table 2). The nuclear morphology software also provided measurements of various sperm nuclear parameters (Figure S4; Table S2). Significant differences in sperm head dimensions were noted between WT and Mut2-Prm1 mice (Table 3). Differences were observed between WT and Mut2-Prm1 sperm in all linear dimensions, with the exception of bounding height and width of the body. Linear dimensions were lower in Mut2-Prm1 in comparison to WT sperm. Dimension-derived parameters that infer shape followed two patterns: ellipticity and elongation were higher in Mut2-Prm1 sperm, whereas the aspect ratio and regularity were lower, in comparison with WT sperm. Circularity was not very different between the two types of sperm. These results indicate that sperm nuclei from Mut2-Prm1 mice were more elongated (streamlined) and less regular in shape than those from WT mice. Using geometric morphometrics analysis, Procrustes ANOVA revealed significant differences (p < 0.0001) in both size (centroid size) and form between WT and homozygous Mut2-Prm1 spermatozoa (Figures 4A and 4B). This indicates notable variations in shape between the two groups, suggesting an impact of the mutation on sperm head conformation. The principal component analysis clearly distinguished between the two groups (Figure 4C). To assess the significance of the separation between subpopulations based on Procrustes distances, canonical variable analysis was conducted. Significant distances (p < 0.05) were noted between any group of mutated individuals and the WT groups. Ellipses representing 90% confidence with respect to the consensus are depicted (Figure 4D). Cys15 and Cys29 mutations in mouse PRM1 affect chromatin condensation and histone retention Transmission electron microscopy (TEM) unveiled that the mutation of cysteines 15 and 29 in the PRM sequence disrupted chromatin condensation, resulting in a looser structure with multiple low-density foci scattered throughout the nuclei (Figure 5). This phenotype, absent in mouse control spermatozoa, resembles the ultrastructure of human sperm nuclei, characterized by nuclear domains retaining a histone organization. To delve deeper into this observation, western blot analysis was performed on sperm from WT and Mut2-Prm1 mice. Remarkably, a high level of histone H3, as an indicator for histone retention, was detected in Mut2-Prm1 sperm (Figures 5B and 5C), consistent with the presence of low-density foci. In WT mice, only a minimal fraction (1%) of nucleosome-organized domains is typically observed, 24 while in human spermatozoa, histone retention accounts for up to 10%–15% of the sperm DNA. 25,26 These findings underscore the influence of PRM1 cysteines on overall histone retention, potentially contributing to reduced DNA condensation. (D) Prm1/Prm2 mRNA ratio. (E) Percentage of PRM1 localized in the nucleus (n = 3) at different spermatid stages during spermiogenesis (from steps 13 to 16). (F) Percentage of PRM2 localized in the nucleus (n = 3) at different spermatid stages during spermiogenesis (from steps 13 to 16). Data are presented as means ± SEM and were analyzed using t tests for two-group comparisons and one-way ANOVA for multiple comparisons with GraphPad Prism 10.2.2 software (GraphPad, San Diego, CA, USA). Table 1. Percentage of sperm abnormalities in wild-type and Mut2-Prm1 mice Normal (%) Head (%) Midpiece (%) Principal piece (%) Other (%) Wild-type 45.7 ± 12.74 2.7 ± 1.52 46.7 ± 8.32 3 ± 2.0 2 ± 3.46 Mut2-Prm1 44.3 ± 15.88 11.0 ± 6.0 37.66 ± 10.69 5 ± 3.6 2 ± 2.64 Results are means ± SD. A total of 100 spermatozoa per male (n = 3 wild type; n = 3 Mut2-Prm1) were assessed. iScience 28, 113102, August 15, 2025 5 iScience Article ll OPEN ACCESS DISCUSSION In this study, we have shown that the PRM sequence affects sperm head shape in a simplified in vitro model of nuclear remodeling in fibroblasts and in vivo in mutated mouse spermatozoa. This study was motivated by the observation of different nuclear shapes following the forced expression of different PRM1 species (mouse and human). The idea emerged that the difference in the number and position of cysteines, specifically the mouse-specific C15 and C29, between mouse and human PRM1 sequences, could explain the observed phenomena. Our in vitro studies unequivocally demonstrated a link between the PRM sequence and nuclear morphology after compaction, a trait that we have further demonstrated by site-directed mutagenesis of cysteines in mouse PRM1 at positions 15 and 29, which brings the mouse sequence closer to the human one. The nuclear remodeling by the mutant PRM1 confirmed our prediction that the lack of cysteine 15 would abolish PRM looping resulting from the first intramolecular 7–15 S–S bond (Figure S1B), which in turn would influence nuclear shape by modifying the way it would associate with DNA and interact with its neighbors, as proposed in the electrostatic zipper model, with a looser condensation capacity. 27,28 The transfection Figure 3. Nuclear morphology analysis in spermatozoa from wild-type and mutated mice Nuclear morphology analysis (NAM) was carried out using NAM software 22,23 on wild-type sperm (WT; n = 3 males; 228 spermatozoa) and Mut2-Prm1 sperm (n = 4 males; 238 spermatozoa). (A) Sperm nuclei from WT and Mut2-Prm1 mice stained with Hoechst 33258 (above) and consensus nuclei showing segments in the nuclear perimeter (below). (B) Superimposition of consensus nuclei from WT (in blue) and Mut2-Prm1 (in red) sperm. (C) Consensus nuclei from WT sperm showing segments. Dimensions of each segment for WT and Mut2-Prm1 sperm are given in Table 2. (D) Comparison of angle profiles between WT (blue) and Mut2-Prm1 (red) sperm showing the median and interquartile range of the nuclear angle profiles. The highest variation is seen around position 350, which corresponds to the flagellum attachment region and the nuclear basal region, as validated previously. 22,23 Table 2. Angle profile segment measurements (in microns; means ± SEM) in wild-type sperm (n = 3 males, 228 spermatozoa) and Mut2-Prm1 sperm (n = 4 males; 238 spermatozoa) Segment WT Mut2-Prm1 0 2.00 ± 0.03 1.89 ± 0.04 a 1 4.76 ± 0.10 4.66 ± 0.11 a 2 2.73 ± 0.10 2.73 ± 0.09 3 2.01 ± 0.09 1.84 ± 0.08 a 4 10.76 ± 0.17 a 10.80 ± 0.20 a Indicates a significant difference between groups (p < 0.01). 6 iScience 28, 113102, August 15, 2025 iScience Article ll OPEN ACCESS experiments of fibroblasts with Mut2-Prm1 also led to an increased number of nuclei with a straight shape similar to the one obtained with the hPrm1 vector (Figures 1A, 1B, 1E, and 1J). To verify our conclusion in relevant physiological models, we used gene editing to replace C15 and C19 in the mouse PRM1-encoding gene. Careful nuclear morphology examination with an ad hoc software 22 revealed differences in the majority of sperm measurements between WT and Mut2-Prm1 sperm as well as in dimension-derived parameters inferring shape (Table 3). Further analysis using geometric morphometrics, as developed for mouse sperm, 29 revealed clear distinctions between WT and mutant mice, reinforcing the idea that changes in the PRM1 sequence impact nuclear sperm shape. The ultrastructural analysis showed differences in chromatin condensation and confirmed morphological changes in mutated spermatozoa. Changes affected the general outline of the cells and the thickness of the nuclei. It is plausible that in a flattened head, such as that of the mouse, sperm with a reduced nuclear compaction would be affected mildly, albeit significantly, in the profile and thickness of the nucleus in mutated sperm. The changes we observed in the shape and morphometry of PRM1-mutated sperm in comparison to WT controls are in agreement with the observation that sperm nuclear morphology is altered in mice in which the Prm1 is deficient or is absent. 30,31 Interestingly, a mouse with a lysine-to-alanine mutation (K49A) in PRM1 exhibited sperm structural changes, including abnormal head morphology, in the absence of effects on weight, germ cell populations, sperm counts, or PRM1 levels. 32 These results emphasize the likely role of specific PRM1 residues in nuclear shaping. It was somewhat unexpected to find altered histone retention in spermatozoa with mutated PRM1. Western blot analysis revealed a marked increase in histone retention (Figure 5). This increased histone retention may contribute to the observed looser nuclear compaction in spermatozoa from the mutated mice, as indicated by the presence of low-density areas in TEM images (Figure 5). These low-density foci are consistent with previous studies suggesting that disruptions in chromatin compaction, such as those caused by altered PRM-histone interactions, can result in regions of reduced density within the sperm nucleus. 17,33,34 This loosening of chromatin likely underlies the morphometric changes observed in the sperm heads (Figures 3and 4), further supporting the link between histone retention and chromatin decondensation. Our results provide strong evidence to suggest that, in addition to actions exerted by extranuclear structures, such as the manchette, in shaping the sperm nucleus, 35–39 intranuclear factors are also important for changes in nuclear morphology. In our in vitro studies, nuclear remodeling occurs in the absence of a manchette, indicating that the inter-/intra-PRM S–S bonds are sufficient to remodel the nucleus and that differences in PRM sequences may result in varying nuclear shapes in this somatic cell in vitro model. This hypothesis may also be supported by the relatively low conservation of the PRM gene across mammals, a surprising finding for a gene with such a highly conserved function. Apparently, the evolution of transition proteins, PRM sequences, and their regulation seems to have exerted a major influence on sperm head shape and size. 40,41 In conclusion, we have demonstrated that the PRM sequence is an important factor in intrinsic nuclear remodeling and, in particular, that PRM cysteines 15 and 29 play a major role in configuring nuclear morphology, histone retention, and chromatin condensation. Because nuclear compaction/remodeling is crucial for male fertility, our results help to characterize mechanisms underlying sperm nuclear formation and to identify the evolutionary forces that influence changes in sperm shape. Limitations of the study While this study significantly advances our understanding of how protamine sequence influences sperm nuclear shape and histone retention, particularly highlighting the role of specific cysteine residues in PRM1, future research could benefit from detailed crystallographic data of protamine-DNA complexes to fully elucidate the precise biophysical mechanisms underlying these changes. Moreover, future studies should thoroughly investigate the impact of PRM2, alone or through its synergistic effects with PRM1, on nuclear shape and histone retention in relation to the various PRM1 sequences. RESOURCE AVAILABILITY Lead contact Requests for future information and resources should be directed to and will be fulfilled by the lead contact, Pasqualino Loi ([email protected]). Materials availability This study did not generate any new reagents. Data and code availability •Data: this article does not report original code. Table 3. Morphometry of sperm nuclei of wild-type sperm (n = 3 males; 228 spermatozoa) and Mut2-Prm1 sperm (n = 4 males; 238 spermatozoa) Parameter WT Mut2-Prm1 Number of nuclei analyzed 228 238 Area 21.26 ± 0.12 20.52 ± 0.18 a Perimeter 21.89 ± 0.11 21.55 ± 0.17 a Maximum feret 8.21 ± 0.03 8.07 ± 0.04 a Minimum diameter 3.43 ± 0.02 3.32 ± 0.03 a Bounding height 7.34 ± 0.03 7.34 ± 0.05 Bounding width 5.48 ± 0.04 5.15 ± 0.05 a Width of body 3.78 ± 0.04 3.70 ± 0.03 Length of hook 1.74 ± 0.03 1.49 ± 0.04 a Circularity 0.56 ± 0.00 0.57 ± 0.01 a Ellipticity 1.36 ± 0.01 1.46 ± 0.02 a Aspect ratio 0.75 ± 0.01 0.71 ± 0.01 a Elongation 0.15 ± 0.00 0.18 ± 0.01 a Regularity 1.48 ± 0.01 1.44 ± 0.01 a Measures (in microns) are means ± SEM. a Differences between WT and Mut2-Prm1 (p < 0.01 for all differences with the exception of minimum diameter for which p < 0.05). iScience 28, 113102, August 15, 2025 7 iScience Article ll OPEN ACCESS •Code: this article does not report original code. •Additional information: any additional information required to reanalyze the data reported in this article is available from the lead contact upon request. ACKNOWLEDGMENTS This project has received funding from the European Union’s Horizon 2020 Research and Innovation Program under the Marie Skłodowska-Curie grant agreement no. GA101131087, WhyNotDry. M.C. acknowledges the support from the National Science Centre, Poland, through grant 2019/35/B/NZ3/ 02856 (OPUS). This work has been funded by the European Union - NextGenerationEU under the Italian Ministry of University and Research (MUR) National Innovation Ecosystem grant ECS00000041 - VITALITY - CUP C43C22000380007. J.F. and H.F. is supported from the Research programme ‘‘Strategy AV21 FUTURE OF ASSISTED REPRODUCTION (ART)’’, AV21VP38/2025. C.A.-R. was supported by predoctoral studentship PRE2020095265 from the Spanish Agencia Estatal de Investigacio ´n, cofounded by the European Social Fund. E.R.S.R. was supported by Spanish Ministry of Science, Innovation and Universities grants CGL2016-80577-P and PID2019108649GB-I00. M.E.T. is supported by the National Institutes of Health (grant R03HD101762). The authors warmly acknowledge Dr. L. Valbonetti, University of Teramo, for help in acquisitions of the confocal microscopy images as well as Aurora Scutieri, University of Teramo, for western blot analysis. AUTHOR CONTRIBUTIONS M.C. and L.P. conducted experiments, prepared figures, analyzed data, and wrote the manuscript. D.W. conducted animal breeding and sperm collection. D.I. performed the synthesis of mutated expression vectors. S.K. conceived the experiments, provided advice on vector design, and wrote the manuscript. J.F. and H.F. analyzed data and prepared the manuscript. R.V., J.A.R., A.S.-R., N.L., and C.A.-R. conducted experiments and analyzed data. E.R.S.R. designed the experiments, analyzed data, and wrote the manuscript. M.E.T. Figure 4. Geometric morphometrics analysis in spermatozoa from wild-type and mutated mice Geometric morphometrics analysis was carried out as described in the STAR Methods section on sperm from WT (n = 3 males; 261 spermatozoa) and Mut2-Prm1 (n = 4 males; 196 spermatozoa) animals. (A) Mouse sperm nucleus stained with Hoechst 33342 exhibiting position of landmarks (red) and semilandmarks (green) used for geometric morphometrics analyses. (B) ‘‘Relative warps’’ analysis with consensus shapes in deformation matrix; WT males are shown using blue symbols, Mut2-Prm1 males are shown using red symbols. (C) Principal component analysis with consensus shapes for each group. Each dot corresponds to an individual male in each group; blue, WT; red, Mut2-Prm. (D) Canonical variate analysis for each male; blue, WT; red, Mut2-Prm. 8 iScience 28, 113102, August 15, 2025 iScience Article ll OPEN ACCESS Sperm collection and preparation Spermatozoa were recovered from females after mating with males (either wild type or Mut-Prm1). A female was mated with an appropriate male until a vaginal plug was observed (mating success was checked every hour). Females were sacrificed and the uterus was dissected to recover spermatozoa from one ejaculation. The uterus was placed in a 35 mm Petri dish containing PBS and 0.4% BSA and the contents were flushed using a 1 mL syringe. The sperm suspension was transferred to a 1.5 mL Eppendorf tube and centrifuged at 1,500 rpm for 10 min. Supernatant was discarded and fixative was added (2% v/v glutaraldehyde in 0.165M sodium cacodylate/HCl, pH 7.4). Fixation was performed overnight at 4◦C. Samples were centrifuged as above, supernatant was discarded and suspended in 0.1 M cacodylate buffer, pH 7.3. Fixed samples were stored at 4◦C until analysis. QUANTIFICATION AND STATISTICAL ANALYSIS Sperm abnormalities Sperm morphology was evaluated by placing 10 μL of fixed sperm suspension between a slide and a 22 ×22 mm coverslip (without applying manual pressure). Spermatozoa in the preparation were allowed to settle for 5 min and then examined at 400x and 1000× magnification using phase contrast optics 44 (Kawai et al. 2006). A total of 100 spermatozoa per male (n = 3 wild type; n = 3 Mut2Prm1) were examined to quantify morphological abnormalities of the head, midpiece and principal plus terminal piece 20,44–46 (Figure S3).When a sperm cell exhibited more than one abnormality, the more severe one was recorded to avoid overestimations. The percentage of normal sperm was calculated as the proportion of spermatozoa with no morphological abnormalities out of all spermatozoa examined. Nuclear morphology analysis The Nuclear Morphology Analysis program was employed to measure and compare the shapes and dimensions of sperm nuclei. 22,23 The program uses a modified Zahn-Roskies transformation to convert the outlines of objects into linear profiles, with a set of rules to identify landmarks of interest from these profiles, detecting subtle variation in nuclear shape. 22,23 The method for morphological analysis allows to automatically identify nuclei in microscopy images, and then find key landmarks for orientation and measurement. An angle profile is generated that describes the shape of the nucleus by measuring internal angles around its perimeter. Different sperm features (the ‘‘hook,’’ site of flagellum attachment) can be identified in the profile regardless of the orientation of the sperm nucleus in the image. Profiles for different datasets (WT vs. Mut2-Prm1 sperm) were aligned against each other. The software calculates median and interquartile range of the angles which are used to illustrate differences between sets of data. Detailed information about the basis for shape analyses, as well as validation of correspondence between segments and sperm features, are available 22,23 (https://bitbucket.org/bmskinner/nuclear_morphology/wiki/Home). In addition to the shape profiles, several measurements of nuclear parameters are automatically calculated and displayed (Figure S4, Table S2). Nuclear morphology analysis using geometrics morphometrics Geometric morphometrics methods were used to analyze head shape variation based on a set of landmarks that correspond to the spatial position of particular anatomical traits. 47,48 Analyses did not include sperm with a nuclei that departed from a normal form. A total of 22 bidimensional landmark coordinates were obtained from spermatozoa of wild-type and Mut2-Prm1 mice. Landmark data were processed as described previously. 29,40 All morphometric analyses were conducted with MORPHOJ. 49 Transmission electron microscopy (TEM) and measurements of sperm head thickness Sperm were collected, as described above, and fixed in glutaraldehyde (2.5% in 0.1 M cacodylate buffer, pH 7.4) for 24 h. After washing in double-distilled water, cells were post-fixed in 2% OsO 4 in double-distilled water for 2 h. Next, cells were dehydrated through a graded series of ethanol solutions (30%, 10 min; 50%, 15 min; 70%, 24 h; 80%, 10 min; 96%, 10 min; 100%, 10 min; acetone, twice for 15 min) and were infiltrated with graded concentrations of Epoxy 812 Resin (EPON) resin in 100% acetone (1:3, 20 min; 1:1, 24 h; 3:1, 2 h), infused twice for 1 h in pure EPON resin, and polymerized at 65◦C for 24 h. Next, 60-nm sections were prepared and examined using a LEO 912AB electron microscope. Images were captured using a Slow Scan CCD camera (Proscane) and EsiVision Pro 3.2 software (Soft Imaging Systems GmbH). Western blot analysis Sperm collected from Mut2-Prm1 (n = 3) and wild type mice (n = 3) were lysed using 8 M urea buffer at 4◦C overnight, then samples were sonicated for 20 s each and centrifuged at 13,000 xg for 5 min at 4◦C. Protein concentration was assessed using the BCA Protein Assay Kit (Thermo Fisher, Milan, Italy) according to the manufacture’s protocol. For each sample, 50 μg of protein were heated at 95◦C for 5 min and then loaded into a 10% SDS polyacrylamide gel. Proteins were transferred onto a 0.45 μm nitrocellulose membrane (Bio-Rad, Milan, Italy) at 4◦C for 2 h and 200 mA. After transfer, the nitrocellulose membrane was blocked with 5% non-fat dry milk in 0.1% Tween 20 PBS (PBST) for 1 h at room temperature. Membranes were incubated overnight with primary antibodies rabbit anti-Histone H3 (PTM-1001RM, PTM BIO, Chicago, Illinois, USA), mouse anti-α-Tubulin (DM1A; 3873S, Cell Signaling Technology, Inc., USA) and mouse anti-GAPDH (D4; MA1-16757, ThermoFisher Scientific, USA) diluted 1:1000 in PBST with 0.5% non-fat dry milk. Then, membranes were washed three times for 15 min with PBST and incubated with secondary antibodies HRP-conjugated iScience 28, 113102, August 15, 2025 e4 iScience Article ll OPEN ACCESS Affinipure goat anti-mouse IgG(H + L) (SA00001-1, Proteintech, Manchester, UK) and mouse anti-rabbit IgG-HRP (sc-2357, Santa Cruz Biotechnology, USA) diluted at 1:10000 in PBST for 1 h at room temperature. Final detection was performed using enhanced chemiluminescence (ECL) Western Blotting Substrate (Amersham, Pharmacia, Piscataway, NJ, USA) and image acquisition was carried out using the ChemiDoc System (Bio-Rad, Milan, Italy). Western blot analyses were repeated 4 times. Statistical analysis Data were analyzed using GraphPad Prism for Windows (Version 6.01, GraphPad Software, Inc., CA, USA). Statistical analyses of transfection yield and the rates of different nuclear morphologies were based on three to five replicates per experiment and compared using Fisher’s exact test or ANOVAs with transformed data (arcsine for percentages and log10 for other parameters). Statistical analyses of protamine expression were performed by a two-tailed T-test. Statistical analyses of nuclear translocation of protamines was assessed using One-way ANOVA with GraphPad Prism version 10.2.2. The significance level for all statistical tests was set at p < 0.05. e5 iScience 28, 113102, August 15, 2025 iScience Article ll OPEN ACCESS