Clustering of sperm by chromatin integrity features reveals breed and monthly effects, as well as a link between double-strand DNA breaks and litter size in pigs Authors: Jordi Ribas-Maynou, Pablo Martínez-Díaz, Judit Drago, Ana Parra, Xiomara Lucas, Marc Yeste, Jordi Roca, Isabel Barranco Published: 2026 ABSTRACT AbstractBackgroundIdentifying sperm biomarkers that predict fertility in swine is important for improving animal selection and production. In this context, however, sperm chromatin in swine remains understudied, and whether it varies by breed and season remains unresolved. Additionally, the impact of sperm chromatin protamination, condensation, and DNA integrity on reproductive outcomes (farrowing rate and litter size) is unclear in swine. This study aimed to improve the understanding of sperm chromatin features by evaluating breed- and month-related differences and determining their influence on fertility outcomes. To this end, we evaluated sperm chromatin protamination and condensation, as well as global and double-strand DNA damage, in 91 ejaculates from 41 boars (25 Landrace and 16 Large White) using chromomycin A3, dibromobimane, and alkaline and neutral Comet assays, respectively.ResultsSingle sperm analysis revealed six clusters associated with chromatin protamination, four clusters related to chromatin condensation, and two clusters corresponding to global DNA damage. Differences between the Landrace and Large White breeds were observed in the percentage of sperm with global DNA damage (P = 0.015), and in chromatin protamination (P = 0.012) and condensation (P < 0.001). Overall, sperm chromatin integrity was better in Large White than in Landrace boars. Monthly analysis revealed that sperm chromatin quality worsened in September due to an increase in global DNA damage (P < 0.01), affecting condensation in one cluster and sperm protamination in two clusters (P < 0.05). Finally, multivariate analysis showed that the percentage of sperm with double-strand DNA breaks was associated with litter size in both breeds.ConclusionsThis study identified sperm subpopulations based on chromatin features in swine. Differences were observed between Large White and Landrace breeds, with the sperm from Large White boars having healthier chromatin quality in terms of DNA damage, protamination, and condensation. Monthly variations were also seen, primarily in relation to global DNA damage. Finally, our study showed an association between the incidence of sperm with double-strand DNA breaks and litter size in both breeds.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12917-026-05667-6.Keywords: Sperm, Pig, Chromatin, Protamination, DNA damage, Condensation, Litter size, Farrowing rate Background The swine industry plays a central role in the food bioeconomy and modern livestock production. Efficient reproductive practices are essential for maximizing profitability on pig farms. In this context, recent advances in reproductive biology have revealed the importance of paternal factors in achieving optimal farrowing rates and litter sizes [1]. Sperm quality in ejaculates intended for artificial insemination (AI) programs is traditionally assessed through motility and morphology analyses using computer-assisted sperm analysis (CASA). Despite this, new research shows that evaluating multiple parameters using flow cytometry provides additional informative biomarkers that better define sperm physiology and function and are more strongly associated with fertility outcomes [2]. Comparative studies among porcine breeds have revealed that certain genetic backgrounds may produce semen samples with superior sperm physiological parameters, such as motility and morphology [3]. Additionally, seasonal variations in sperm quality have been observed [4], and some breeds appear to be better adapted to heat stress conditions [5, 6]. Regarding nuclear integrity, recent advances in the study of sperm DNA damage using highly sensitive methods have shown that DNA breaks in sperm cells can hinder embryo development, leading to a reduction in the number of embryos that reach the blastocyst stage [7]. Sperm chromatin is particularly resilient to damage due to crucial differences compared to somatic cells, primarily the replacement of most histones by protamines, which renders the genetic material transcriptionally inert [8, 9]. In sperm, protamines bind negatively-charged DNA via strong electrostatic interactions, which are facilitated by the high content of arginine, a basic positively-charged amino acid. Sperm chromatin is also stabilized by inter- and intraprotamine disulfide bonds, which are enabled by the abundance of cysteine residues in protamines [10]. According to the most recent models, sperm chromatin is organized into toroidal structures connected by linker DNA. These structures are called toroidal linker regions [11]. Toroidal structures are resistant to nuclease activity but sensitive to oxidative modifications caused by reactive oxygen species (ROS) [12]. Although DNA damage has been extensively studied in livestock, chromatin deprotamination and chromatin decondensation have received far less attention, despite their evident role in defining sperm chromatin alterations. Chromatin deprotamination refers to the absence of protamines in regions of DNA that should be fully protaminated. Chromatin is decondensed when the thiol groups of cysteines are not forming disulfide bonds, remaining in the reduced state. Previous research in pigs found that the degree of chromatin condensation was higher in sperm from the rich fraction than in those from the post-sperm-rich fraction, yet protamination and DNA damage levels appeared similar across the fractions [13]. Additionally, a study by Lacalle et al. showed that condensation, as evaluated by disulfide bond levels, was affected by semen storage at 17 °C [14]. In other species, such as rodents, assessing free thiols with monobromobimane has demonstrated that chromatin packaging decreases with age [15]. Concerning DNA evaluation, various methods, particularly those relying on indirect measurements or enzymatic labeling, have been proven to lead to an underestimation of DNA damage, which challenges the validity of this biomarker [16, 17]. Furthermore, some studies have found that DNA breaks have a detrimental effect on fertility outcomes, whereas others have not observed such an effect [18]. Regarding protamination, while research in humans has proposed that defects could be associated with reduced fertility rates [19, 20], no study has focused on these parameters in pigs. As for chromatin condensation, there is no available data in pig sperm linking this parameter to fertility outcomes, namely, farrowing rates or litter size. The present study, therefore, aimed to determine whether distinct clusters based on chromatin protamination and condensation, and DNA integrity in sperm exist in porcine ejaculates and whether these parameters vary across breeds and months. The study also sought to establish the relationship between these variables and male fertility, as measured by farrowing rates and litter size in artificially inseminated sows. Methods Animals, ejaculates, artificial insemination, and in vivo fertility outcomes Semen samples were obtained from the ejaculates of 41 boars of the Landrace (n = 25) and Large White (n = 16) breeds. The boars were housed at the AIM Ibérica AI center in León, Spain, and served as semen donors for the commercial AI programs of the company Topics Norsvin España (Madrid, Spain). The ejaculates were collected using a semi-automated method (Collectis®, IMV Technologies; L’Aigle, France). All ejaculates met the quality criteria for producing commercial semen AI doses: more than 200 × 106 sperm/mL, more than 70% motile sperm, and more than 75% normal morphology sperm. The AI doses used contained 2.4 × 10⁹ spermatozoa in an 80 mL of volume. Each boar had inseminated at least 30 sows, ranging from 32 to 706. The sows were inseminated twice cervically during post-weaning estrus. The 41 boars were selected because their fertility records differed from the mean observed in the boar population of each breed (with more than 200 boars per breed), either positively or negatively. Fertility was recorded as the Direct Boar Effect (DBE), a metric value used by the company Topics Norsvin to measure the source of fertility variation related exclusively to boars and semen. The DBE is derived from statistical equations that separate sources of fertility variation related to farm and sow parameters (including sow parity, farm, insemination items, season, and farmer) from sources related to boar and semen parameters (including age, interval between ejaculations, AI center and laboratory, and laboratory technician) [21]. The DBE for each boar is expressed as the deviation in the percentage of sows that farrowed (the number of sows that farrowed divided by the number of sows inseminated) and litter size (the total number of piglets born per farrowing), relative to the mean for all boars within each breed (see Fig. S1). To conduct the experiment, ejaculates were collected approximately every two months throughout one year. A total of 17 ejaculates were collected in January, 19 in March, 25 in June, 15 in September, and 15 in December. Immediately after collection, 15 mL of each ejaculate was centrifuged at 3,000×g for 10 min. The pellets were stored at -80 °C until used for chromatin evaluation. Study design The 91 sperm samples were evaluated for global and double-stranded DNA damage, chromatin condensation, and chromatin protamination using the methods described below. Subsequently, sperm subpopulations were identified for each analyzed parameter using individual raw DNA-damage and fluorescence data obtained by flow cytometry. The percentage of sperm in each cluster, as well as the mean incidence of DNA damage parameters and fluorescence intensity, were used to compare Landrace and Large White breeds, and to generate multivariate models that tested for seasonal variations and evaluated the impact of each parameter on the farrowing rate (percentage of inseminated sows that farrow) and litter size (number of piglets born per litter). Global and double-strand sperm DNA damage (Comet assay) Sperm DNA fragmentation was evaluated using single-cell gel electrophoresis, adapted for pig sperm. The protocol, which ensures high chromatin decondensation, was previously described elsewhere [22]. The protocol begins with sperm fixation and lysis, continues with electrophoresis, and ends with neutralization, dehydration, and staining. Electrophoresis pH determine the two Comet variants. If pH is alkaline, then the method will evaluate global DNA damage (alkaline Comet). However, if pH is neutral, then the method will assess double-strand breaks (neutral Comet). The following subsections detail every step conducted for the Comet assays. Sperm fixation and lysis First, the sperm samples were adjusted to 5 × 105 sperm/mL in Phosphate-Buffered Saline (PBS) and mixed at a ratio of 1:2 (v/v) with 1% liquefied low-melting-point agarose while maintaining the temperature at 37 °C. Then, 6.5 µL of the mixture was placed onto two pretreated agarose slides and rapidly covered with an 8-mm diameter round coverslip. The slides were then incubated on a metal plate at 4 °C for 5 min to allow the samples to solidify. Then, the coverslips were removed, and both slides were incubated at room temperature in three sequential lysis solutions: lysis 1 (0.8 M, Tris-HCl, 0.8 M DTT, and 1% SDS, pH 7.5) for 30 min, lysis 2 (0.4 M Tris-HCl, 0.4 M DTT, 50 mM EDTA, 2 M NaCl and 1% Tween 20, pH 7.5) for 30 min, and lysis 3 (0.4 M Tris-HCl, 0.4 M DTT, 50 mM EDTA, 2 M NaCl, 1% Tween 20, and 100 µg/mL Proteinase K, pH 7.5) for 180 min. After the third lysis step, both slides were washed with distilled water at room temperature for 5 min. Electrophoresis, neutralization, dehydration, and staining The electrophoresis conditions differed between the two slides, as one was designated for the alkaline comet assay to evaluate global DNA damage, and the other was intended for the neutral comet assay to assess double-strand DNA damage. For the alkaline comet, the samples were first denatured in a solution of 0.03 M NaOH and 1 M NaCl (pH 13) at 4 °C for 5 min, and then electrophoresed at 1 V/cm in an alkaline buffer solution of 0.03 M NaOH (pH 13) for 4 min. For the neutral comet, the slide was electrophoresed directly in a Tris-borate-EDTA buffer solution (0.445 M Tris-HCl, 0.445 M boric acid, and 0.01 M EDTA; pH 8) for 4 min. After electrophoresis, both alkaline and neutral comet slides were placed in a neutralization solution of 0.4 M Tris-HCl (pH 7.5) for 5 min and then dehydrated in 70%, 90%, and 100% ethanol solutions for 2 min each. Staining, imaging and analysis Staining was performed by submerging the slides in a solution of 25 µmol/L SYTOX Orange (ThermoFisher Scientific, Waltham, MA, USA) at room temperature in the dark for 15 min. Excess stain was removed by submerging the slides in distilled water for 10 min. Finally, the slides were allowed to air-dry horizontally. Comets were visualized under an Olympus AX70 epifluorescence microscope (Tokyo, Japan). For the alkaline comet assay, one hundred sperm were imaged at 100× magnification using Pylon Viewer software (https://www.baslerweb.com) and a Basler camera (Ahrensburg, Germany) with an exposure time adjusted to prevent overexposure of comet heads and tails. The images were analyzed using the open-source CometScore v2.0 software (Rexhoover, www.rexhoover.com) with the automatic analysis function. Then, each cell was manually reviewed to eliminate images that did not correspond to comets, images showing overlapping comets, and misidentifications of the comet head or tail. A minimum of 50 correctly analyzed comets was considered necessary to obtain a statistically significant result. The data was processed using Microsoft Excel, and the percentage of DNA in the comet tail (%Tail DNA) was calculated for each cell. This variable was subjected to cluster analysis based on Euclidean distance and the between-groups linkage method, yielding two subpopulations with low and high DNA fragmentation. For the neutral comet assay, 400 sperm were classified as either with non-fragmented DNA (round halos) or with fragmented DNA (halos with a comet tail), and the percentage of sperm with fragmented DNA was recorded. Flow cytometry analysis Evaluation of chromatin condensation and protamination was performed using a CytoFlex flow cytometer (Beckman Coulter, Fullerton, CA, USA) that was equipped with three lasers: violet (405 nm), blue (488 nm), and red (637 nm). For each sample, 10,000 sperm were assessed at a controlled flow rate of 10 to 60 µL/s. The cytometer was calibrated using specific CytoFlex Daily QC fluorospheres (Beckman Coulter, Fullerton, CA, USA) before every experiment. Raw data containing fluorescence for each channel was recorded in an FCS file using CytoExpert Software (Beckman Coulter). The sperm cell population was gated based on forward and side scatter (FSC and SSC) and appeared as the typical flame shape. Chromatin condensation The evaluation of chromatin condensation was performed by assessing the oxidation-reduction status of disulfide bonds using dibromobimane (DBB). DBB is a cell-permeable reagent that alkylates two reduced thiol groups located approximately 6 Å apart [23, 24]. Hence, DBB only reacts when a disulfide bond is in its reduced state (R–SH HS–R), forming a covalent adduct (R–S–DBB–S–R). In this state, DBB becomes fluorescent, with an emission peak at 490 nm upon excitation at 394 nm. Higher DBB fluorescence values are thus inversely correlated to chromatin condensation. In brief, a 4 mmol/L stock solution of DBB was prepared in dimethyl sulfoxide. After that, the sperm samples were diluted in PBS to a concentration of 1 × 106 sperm/mL. The samples were subsequently incubated at a final concentration of 20 µmol/L at room temperature in the dark for 20 min. Then, the DBB fluorescence was quantified by flow cytometry. The samples were excited with the violet laser, and the fluorescence was collected through the KO525 filter (525/40 bandpass). Chromatin protamination Chromatin protamination was assessed by incubation with chromomycin A3 (CMA3), which is an antibiotic that binds to the minor groove of the DNA in the presence of Mg2+ ions. In sperm cells, this binding occurs only when the chromatin region lacks protamines; therefore, the higher the CMA3 signal, the lower the chromatin protamination. When bound to DNA and excited at 430 nm, CMA3 emits fluorescence at 590 nm. First, a CMA3 stock solution was prepared at 5 mg/mL in ethanol, and the sperm concentration was adjusted to 20 × 106 sperm/mL in PBS. The samples were then mixed at a ratio of 1:1 with 2× McIlvine Solution (60 mM citric acid, 280 mM Na2HPO4, and 20 mM MgCl2). CMA3 was then added to achieve a final concentration of 12.5 µg/mL, and the mixture was incubated at room temperature for 30 min. Next, the samples were diluted 1:10 in PBS and examined with the flow cytometer. For this purpose, samples were excited using the violet laser, and the fluorescence was collected through a Violet610 channel (610/20 bandpass filter). Determination of sperm clusters from flow cytometry data The raw data recorded in FCS for each channel were exported to a .csv file. Then, .csv files for all samples were merged into a single database. First, events with FSC-A values below 1 × 105 and with SSC-A values below 8 × 104 were excluded from the database. Subsequently, a cluster analysis using FSC-A and SSC-A values was performed, revealing two clusters: one corresponding to values typical of single sperm cells, and the other with higher FSC and SSC values. Events belonging to the first cluster were selected for further analysis of chromatin condensation and protamination. To do so, fluorescence data collected with either KO525 or Violet610 were included in a new cluster analysis. This approach revealed six subpopulations based on CMA3 (chromatin protamination) and four subpopulations based on DBB (chromatin condensation). The mean fluorescence and the percentage of sperm in each cluster were recorded for each sample. Statistical analysis Statistical analyses were conducted using the Statistical Package for the Social Sciences (SPSS) version 25 (IBM Corp., Armonk, NY, USA). Graphs were plotted using GraphPad Prism v. 8 software (GraphPad software, La Jolla, CA, USA). Flow cytometry data, including individual events for all biological samples, were processed using the above-described strategy, which included cluster analysis. As indicated before, cluster analyses were based on Euclidean distance and the between-groups linkage method, which allowed for the identification of different subpopulations. In all clustering procedures, no predefined number of clusters was imposed during the analysis. Instead, clusters emerged naturally from the hierarchical clustering process and were determined by the statistical algorithm. Clusters are numbered according to increasing mean values, from the lowest to the highest. Events falling into each cluster were grouped to a subpopulation, in which descriptive statistics were calculated for each biological sample. The subsequent analysis was divided into two parts. The first part evaluated the effects of the month of sample collection on sperm chromatin. The second part generated chromatin-based models to predict farrowing rate and litter size. Normal distribution and homogeneity of variances were checked with the Shapiro-Wilk and Levene tests. For the first analysis, a linear mixed model was used, including month and breed as fixed effects and each biological replicate as a random effect. As not all animals provided samples in all months, a repeated-measures structure was specified. Analyses were based on Type III sums of squares, which are appropriate for unbalanced data. This accounts for between-animal variability and enables analysis of repeated measurements. For the second analysis, separate linear regression models were generated for Landrace and Large White breeds. These models included either the average of each continuous variable or the average percentage of sperm per cluster across all parameters. For all linear regression models, Receiver Operating Characteristics (ROC) curves were used to assess their ability to predict farrowing rate and litter size. In all statistical analyses, the significance level was set at P < 0.05. Results Sperm in porcine ejaculates can be clustered based on DNA integrity, chromatin condensation, and chromatin protamination The strategy used enabled the evaluation of chromatin parameters (i.e., protamination, condensation, and global DNA integrity) in individual sperm from each semen sample. This approach grouped spermatozoa into different clusters (Fig. 1). Specifically, two clusters for DNA integrity (% Tail DNA), four clusters for chromatin condensation (DBB fluorescence intensity), and six clusters for chromatin protamination (CMA3 fluorescence intensity). The summary statistics for each cluster are shown in Table S1. Box and whiskers plot indicating values used to define sperm clusters for (a) Global DNA damage (%Tail DNA), (b) chromatin condensation (dibromobimane [DBB] fluorescence), and (c) chromatin protamination (chromomycin A3 [CMA3] fluorescence) Landrace and Large White breeds differ in the degree of chromatin protamination and condensation of their sperm, with the Landrace breed exhibiting a greater percentage of sperm with increased DNA damage After classifying sperm into clusters, we interrogated whether the percentage of sperm in each cluster differed between Landrace and Large White breeds. Figure 2 presents comparisons of the main parameters, and summary statistics are provided in Table S2. The global incidence of DNA breaks per each sample (%Tail DNA) did not differ significantly between the two breeds (mean ± SD, Landrace: 48.30 ± 9.61, Large White: 42.78 ± 11.93; P = 0.112). Regarding the distribution of sperm across clusters, most sperm (53.6% in Landrace and 70.9% in Large White) belonged to the cluster with low global DNA fragmentation (Fig. 3). Of note, the percentage of sperm with increased DNA damage was significantly (P = 0.015) higher in the Landrace (46.43 ± 25.30%) than in the Large White breed (29.03 ± 26.67%) (Fig. 2). Mean ± SD of sperm DNA integrity stratified by Landrace and Large White breeds. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively Differences in the distribution of sperm among clusters for DNA damage chromatin condensation and chromatin protamination between Landrace and Large White breeds On the other hand, differences in chromatin condensation were also observed between the two breeds. On average, DBB fluorescence intensity, which indicates chromatin decondensation, was 33.7% higher (P < 0.001) in the sperm of the Landrace (101,233 ± 16,509 arbitrary units) than in those of the Large White breed (75,739 ± 16,121 arbitrary units). In addition, significant differences in the percentages of sperm belonging to each of the four clusters identified for chromatin condensation were observed between breeds (P < 0.05) (Fig. 3), with clusters 1 and 2 being the predominant ones and comprising 82.9% and 93.5% of sperm in Landrace and Large White breeds, respectively. Chromatin protamination also differed significantly between breeds. On average, CMA3 fluorescence intensity was 13.6% higher (P = 0.012) in the sperm of the Landrace (2,319 ± 436) than in those of the Large White breed (2,040 ± 487). Moreover, the percentages of sperm belonging to clusters 2, 3, 4, and 5 differed between breeds (P < 0.05; Fig. 3), with clusters 2 and 3 together grouping 75.3% of sperm in the Landrace breed and 82.3% in the Large White breed. Variations in sperm DNA integrity, as well as in specific clusters of chromatin protamination and condensation, occur throughout the year By collecting repeated samples from the same animals, we were able to determine whether variation occurred across the months of the year despite the controlled temperature and lighting conditions of the boar stud (Fig. 4). Type III sums of squares in linear mixed models revealed month effects for the following variables: %Tail DNA (F = 5.524, P = 0.004), percentage of sperm with increased DNA damage (F = 5.055, P = 0.006), percentage of sperm with double-strand DNA breaks (F = 2.926, P = 0.043), percentages of sperm belonging to clusters 2 and 4 based on chromatin protamination (CMA3; F = 3.007, P = 0.028 and F = 4.538, P = 0.037, respectively), and percentage of sperm belonging to cluster 2 based on chromatin condensation (DBB; F = 2.543, P = 0.032). Specifically, September consistently showed statistically poorer chromatin-related parameters across all evaluated variables, with no parameter showing improvement. Summary statistics are provided in Table S3. Monthly variations in (A) global DNA damage (%Tail DNA), (B) chromatin protamination (chromomycin A3 [CMA3] fluorescence), and (C) chromatin condensation (dibromobimane [DBB] fluorescence), and in the percentage of sperm belonging to each of the clusters. Bars indicate SEM. * indicates differences between breeds at P < 0.05 Chromatin protamination and condensation, and DNA integrity of pig sperm can predict farrowing rate and litter size After determining that the month of ejaculate collection and breed influence chromatin protamination, condensation and DNA integrity in pig sperm, we examined whether these factors were associated with farrowing rate or litter size. If so, they could serve as predictors of male fertility in pigs. To this end, average DBE values across all samples collected from each boar were calculated, and separate regression models for Landrace and Large White breeds were constructed for both farrowing rate and litter size. The β estimates and P values for each parameter are shown in Table S4. Notably, the percentage of sperm with double-strand DNA breaks was associated with litter size in both breeds (P = 0.025 for Landrace and P = 0.024 for Large White), with greater percentages of sperm with double-strand DNA breaks being associated with smaller litter sizes. On the other hand, sperm chromatin protamination was found to be associated with farrowing rates in the Landrace breed (P = 0.016), and chromatin condensation was observed to be associated with litter size in the Large White breed (P = 0.028). Although these breed-specific effects were statistically significant, their magnitude was very small (β < 0.001) relative to the corresponding model intercepts (β = −8.010 and β = 4.030, respectively). Predicted farrowing rates and litter sizes were also validated using the four models based on the main parameters described in the Table S4. Predicted values were plotted against the observed fertility outcomes (Fig. S2). Figure 5 shows the ROC curves revealing that the model designed to predict litter size in the Large White breed was statistically significant (P = 0.022), with an area under the curve (AUC) of 0.841 (95% CI: 0.639–1.000). Summary statistics of ROC analysis are provided in Table S5. Receiver Operating Characteristics (ROC) curves for the multivariate models generated using sperm chromatin parameters and male reproductive outcomes (farrowing rate and litter size) for Landrace and Large White breeds Discussion The relationship between specific sperm quality variables and reproductive outcomes remains challenging in swine reproduction. In our study, we examined sperm chromatin and DNA and identified breed- and month-of-ejaculate-collection-related changes. We also demonstrated that the incidence of double-strand DNA breaks in pig sperm is associated with litter size in the Large White breed. We used flow cytometry and the Comet assay to evaluate chromatin protamination and chromatin condensation, as well as DNA integrity, at the level of individual sperm. This multimethodological approach was crucial for identifying that sperm can be classified into different clusters based on chromatin features. The six clusters identified following CMA3 staining and the four observed after DBB staining were associated with distinct levels of protamination and condensation in sperm, respectively. Previous studies have assessed chromatin packing in sperm using a bimodal classification (protamine-rich/protamine-poor or condensed/decondensed), reporting only the percentage of cells with positive staining [19, 25]. Our results, however, suggest that chromatin protamination and condensation are better represented as continuous variables than binary categories. Nevertheless, it is worth noting that despite two clusters accounted for more than 75% of serm in both protamination and condensation, it remains to be studied if the other less represented clusters may be biologically relevant. Future experiments separating sperm from these less-represented clusters could help identifying DNA region-specific differences that may help understanding the differences between them and the sperm belonging to clusters showing a completely healthy and condensed chromatin. Regarding the evaluation of global DNA damage, our analysis allowed us to identify two sperm populations based on the percentage of DNA forming the Comet tail, a parameter directly related to sperm DNA fragmentation. Most sperm were observed to have a low incidence of DNA breaks, indicating good DNA integrity. Moreover, two clusters based on mean %Tail DNA were identified, with an average of 34% (low DNA damage) and 64% (increased DNA damage). It should be noted that these figures reflect the percentage of DNA with a certain number of DNA breaks rather than the amount of degraded DNA or the percentage of sperm with fragmented DNA. Notably, we identified two clusters associated with DNA damage, whereas a similar study using the Comet assay in humans identified three distinct clusters [26]. We attribute these differences to the origin of the samples and species, as, in the current study, pig sperm were obtained from fertile boars involved in AI programs, which are expected to have high DNA integrity. The third cluster observed in humans contained sperm with highly damaged DNA, which is more typical of men experiencing fertility issues but is not observed in pigs, as confirmed in this and previous studies [27, 28]. Our study revealed differences between breeds. Specifically, the sperm of Large White boars exhibited better overall chromatin health than those of the Landrace breed, as evidenced by the lower degree of DNA damage and the higher degree of chromatin protamination and condensation. Differences in sperm physiological parameters, such as concentration, motility, and morphology, have been observed across pig breeds [29, 30], underscoring the importance of genetic background in shaping these variables. Consistent with this, previous research has revealed breed-specific differences in sperm preservation capacity at 17 °C, with certain breeds being more susceptible to storage-induced damage than others [31, 32]. Despite the lack of data on chromatin protamination and condensation, some investigations have examined how DNA damage varies across breeds, yielding inconclusive results. For example, one study reported differences between Duroc and Landrace [33] whereas another failed to find differences among Duroc, Hampshire, Landrace, and Large White breeds [34]. It is important to note that these previous studies used the Sperm Chromatin Structure Assay (SCSA) to assess sperm DNA integrity. Our previous comparison of tests for sperm DNA integrity in pigs suggests that indirect methods, such as the SCSA, may be less accurate than direct methods, such as the Comet assay [22]. In our study, the sperm DNA of Large White boars appeared less damaged than that of Landrace boars. While routine evaluation of sperm chromatin could be a beneficial tool for enhancing reproductive efficiency in pig production, further research is necessary to thoroughly assess how sperm chromatin stability changes during semen preservation at 17 °C. This will enable a comprehensive characterization of breed-specific variations in sperm resilience to storage, considering DNA damage, and chromatin condensation and protamination. We also found an effect of the month of ejaculate collection. Specifically, sperm chromatin quality decreased in September, which could be compatible with a delayed effect of summer heat exposure. Unlike other mammals, pigs have very few functional sweat glands and rely on alternative physiological mechanisms for thermoregulation. These mechanisms are inefficient at high temperatures and vary across breeds [35]. Previous reports have shown that Landrace and Large White boars exhibit the highest rates of teratozoospermia in warmer months due to tail defects and increased acrosome abnormalities [36]. Similarly, Gonzalez-Castro et al., (2022) found that warm months negatively affect sperm motility, morphology, and viability [37]. Our data point towards sperm DNA damage might not be exempt from the heat exposure, given that previous research that showed that thermal stress alters sperm chromatin structure [38]. Related to this, supplementing boar’s diet with antioxidants has been demonstrated to mitigate thermal stress-related DNA damage in sperm, which suggests an oxidative etiology [39]. Consistent with these observations, we found that the incidence of double-strand DNA breaks in sperm did not vary significantly across seasons, whereas overall DNA damage increased following warmer periods. This implies that single-strand DNA breaks, predominantly of oxidative origin [40], are the type of DNA damage most affected by seasonality. This underscores the importance of assessing different aspects of sperm chromatin integrity. Consistent with this suggestion, our study also identified monthly differences in specific sperm subpopulations based on chromatin protamination and condensation. Specifically, the percentages of sperm belonging to clusters 2 and 4 based on chromatin protamination, and those belonging to cluster 2 based on chromatin condensation, observed in September differed from the other months. Noticeably, these findings are reported here for the first time. Given that our regression models indicated that the percentages of sperm belonging to cluster 4 based on chromatin protamination and those belonging to cluster 2 are correlated with farrowing rates in the Landrace breed, future research should focus on characterizing the molecular features of these sperm subpopulations. Finally, our study examined the relationship between chromatin parameters and fertility outcomes in terms of farrowing rate and litter size. We found that increased percentages of sperm with double-strand DNA breaks were associated with smaller litters in both Landrace and Large White breeds. Yet, the incidence of global DNA damage did not show a similar association. This finding is also reported here for the first time but is consistent with our previous data on pigs and with observations in other species. In pigs, double-strand DNA breaks have been linked to impaired in vitro embryo development, particularly the ability to reach the blastocyst stage [41], suggesting that this type of damage may be detrimental to embryo development. In a previous study, we proposed that double-strand breaks in paternal DNA might induce chromosomal aberrations during early embryo development, resulting in cell death and impaired embryo progression [42, 43]. In humans, double-strand breaks in sperm DNA showed a relationship to first-trimester pregnancy losses, although not being associated with pregnancy establishment [40, 44, 45]. This could be explained by their low incidence and highly localized nature [46]. In our experimental setting, overall sperm DNA integrity appeared to be sufficient for establishing pregnancy, but localized double-strand DNA breaks might disrupt the regulation of key developmental genes during embryogenesis. This could lead to embryo death and consequently smaller litters. Further studies evaluating the genetic location of double-strand DNA breaks and their colocalization with genes important for embryo development could help confirm this hypothesis. The data presented here enabled us to build models for predicting farrowing rates and litter sizes in pigs. Of the evaluated models, only the one for litter size in the Large White breed reached statistical significance. To our knowledge, this is the first model to incorporate chromatin parameters to predict litter size in pigs. While this highlights the potential utility of these models, it also reveals some limitations. First, the β coefficients are relatively small compared to model intercepts, indicating that the effects of the chromatin parameters are small relative to other biological factors, although they remain statistically significant. Second, the model exhibited good discriminative capacity, as reflected by the AUC, and identified the percentage of sperm with double-strand DNA breaks as the main contributing parameter. Despite this, the fact that it reached statistical significance in only one of the two breeds under study suggests a breed-dependent association between sperm DNA integrity and reproductive outcomes, which is consistent with the breed-specific effects discussed above. The definitive predictive value of this model should be validated using an independent subset of boars with known fertility records. Finally, our study is not without limitations. Firstly, the results obtained are breed-dependent, and therefore cannot be directly extrapolated to other breeds commonly used in pig production, such as Duroc or Pietrain. Regarding the observed month effects, our study was able to describe monthly variations in chromatin quality that may be compatible with delayed effects of warmer periods, however, no causal relationship could be established due to the lack of environmental data supporting a heat-stress interpretation. Finally, although the sample size used was sufficient to establish predictive models, the practical utility and robustness of these models should be further validated through large-scale, multisite studies. Conclusions Through a comprehensive evaluation of sperm chromatin, we identified breed-related differences between Landrace and Large White boars, as well as month variations that might be associated with delayed effects of the warmer months. These effects were linked to decreased DNA integrity and altered chromatin structure. Our data also suggests that the percentage of sperm with double-strand DNA breaks may be associated to litter size in the Large White breed. Supplementary Information Supplementary Material 1: Fig. S1. Plots showing the direct boar effect (DBE) for farrowing rate and litter size for boars selected as semen donors in the fertility experiment. The zero on the y-axis represents the average DBE of all boars of the same breed included in the artificial insemination program (AI) of the AI company. The dots represent the DBE of each one of the selected boars. **** and *** indicate statistical significance at P < 0.0001 and P < 0.001, respectively. Supplementary Material 2: Fig. S2. Plots showing observed and predicted farrowing rates and litter sizes, calculated using four models based on sperm DNA integrity and chromatin parameters. Supplementary Material 3: Table S1. Summary statistics of sperm clusters based on DNA damage, chromatin protamination, and chromatin condensation. Supplementary Material 4: Table S2. Summary of overall and cluster statistics for sperm DNA damage, chromatin condensation, and chromatin protamination in Landrace and Large White breeds. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Supplementary Material 5: Table S3. Summary statistics for the different sperm DNA parameters in semen samples collected in each month, including their β-estimates and P-values from the mixed linear regression models. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Supplementary Material 6: Table S4. Multivariate models generated using sperm DNA parameters to explain fertility outcomes, in terms of farrowing rate (A) and litter size (B), of Landrace and Large White boars. The models differ based on whether they include the main parameters or the cluster variables. * Indicates differences between breeds at P < 0.05. Supplementary Material 7: Table S5. Summary of the statistics of Receiver Operating Characteristics (ROC) curves for the multivariate models generated using sperm chromatin parameters and male reproductive outcomes (farrowing rate and litter size) for Landrace and Large White breeds. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Acknowledgements The authors would like to thank AIM Ibérica (Topigs Norsvin Ibérica, Madrid, Spain) for providing pig ejaculates. Abbreviations Artificial Insemination Area Under the Curve Computer-Assisted Sperm Analysis Chromomycin A3 Dibromobimane Forward Scatter Phosphate-Buffered Saline Receiver Operating Characteristics Reactive Oxygen Species Sperm Chromatin Structure Assay Statistical Package for the Social Sciences Side scatter Authors’ contributions JR-M, JR, MY and IB conceived the study. JR-M, JD, AP, PM-D, and MY conducted the laboratory experiments. JR-M, XL, JR, MY and IB participated in the discussion of the results. JR-M wrote the draft manuscript. XL, MY, JR and IB critically revised and edited the manuscript. All authors contributed to the finalized manuscript, read, and approved the final version. Funding The study was funded by (1) the Ministry of Science, Innovation and Universities, Spain (MICIU/AEI/10.13039/501100011033) and the European Regional Development Fund (ERDF/EU): PID2022-137738NA-I00, PID2023-149853OB-I00 and PID2024-159343OB-I00; (2) the Seneca Foundation (Murcia, Spain) (FSRM/10.13039/100007801):21935/PI/22, and 23014/GERM/25; and (3) the Consolidación Investigadora scheme: MICIU/AEI/10.13039/501100011033: CNS2024-154471. Jordi Ribas-Maynou was a María Zambrano Fellow, supported by the MICIU (AEI/10.13039/501100011033) and the European Union Next Generation (124/MTAI/22). Isabel Barranco has a postdoctoral fellowship “Ramon y Cajal” supported by the MICIU (MCIU/AEI/10.13039/501100011033), and the European Union Next Generation/PRTR Fund (RYC2021-034546-I). Marc Yeste is supported by the Catalan Institute of Research and Advanced Studies (ICREA), under the ICREA Academia scheme. Data availability All data supporting the findings of this study are available within the paper and its Supplementary Information. Declarations Ethics approval and consent to participate The experiments were approved under research codes CBE 538/2023 and CBE 668/2025 by the Bioethics Committee of the University of Murcia at its meetings on June 16, 2023 and November 1, 2025, respectively. Commercial semen samples were obtained from AI León (AIM Ibérica, Topigs Norsvin Spain S.L.U., Madrid, Spain), an AI center that complies with the European (ES13RS04P, July 2012) and Spanish (ES300130640127, August 2006) regulations for the commercialization of AI semen doses, as well as with the European regulation concerning animal health and welfare. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1: Fig. S1. Plots showing the direct boar effect (DBE) for farrowing rate and litter size for boars selected as semen donors in the fertility experiment. The zero on the y-axis represents the average DBE of all boars of the same breed included in the artificial insemination program (AI) of the AI company. The dots represent the DBE of each one of the selected boars. **** and *** indicate statistical significance at P < 0.0001 and P < 0.001, respectively. Supplementary Material 2: Fig. S2. Plots showing observed and predicted farrowing rates and litter sizes, calculated using four models based on sperm DNA integrity and chromatin parameters. Supplementary Material 3: Table S1. Summary statistics of sperm clusters based on DNA damage, chromatin protamination, and chromatin condensation. Supplementary Material 4: Table S2. Summary of overall and cluster statistics for sperm DNA damage, chromatin condensation, and chromatin protamination in Landrace and Large White breeds. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Supplementary Material 5: Table S3. Summary statistics for the different sperm DNA parameters in semen samples collected in each month, including their β-estimates and P-values from the mixed linear regression models. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Supplementary Material 6: Table S4. Multivariate models generated using sperm DNA parameters to explain fertility outcomes, in terms of farrowing rate (A) and litter size (B), of Landrace and Large White boars. The models differ based on whether they include the main parameters or the cluster variables. * Indicates differences between breeds at P < 0.05. Supplementary Material 7: Table S5. Summary of the statistics of Receiver Operating Characteristics (ROC) curves for the multivariate models generated using sperm chromatin parameters and male reproductive outcomes (farrowing rate and litter size) for Landrace and Large White breeds. ** and * indicate differences between breeds at P < 0.01 and P < 0.05, respectively. Data Availability Statement All data supporting the findings of this study are available within the paper and its Supplementary Information.