Exposure to female olfactory cues hastens reproductive ageing and increases mortality when mating in male mice

Abstract

Theories of ageing predict that investment in reproduction will trade-off against survival and later-life reproduction. Recent evidence from invertebrates suggests that just perceiving cues of a potential mate's presence can reduce lifespan, particularly in males, and that activation of neuroendocrine reward pathways associated with mating can alleviate these effects. Whether similar effects occur in vertebrates remains untested. We tested whether exposure to olfactory cues from the opposite sex would influence mortality and reproductive senescence in male mice. We observed that males exposed to female olfactory cues from middle- to old age (from 10 to 24 months of age) showed reduced late-life fertility, irrespective of whether they had also been allowed to mate with females earlier in life. Males that were exposed to female odours in conjunction with mating also showed an increased mortality rate across the exposure period, indicating that olfactory cues from females can increase male mortality in some environments. Our results show that exposure to female odours can influence reproductive ageing and mortality in male mice, highlighting that sensory perception of mates may be an important driver of life-history trade-offs in mammals.

1. Introduction

Senescence in fertility and late-life function, which causes increased mortality with age, is linked to early-life investment in growth and reproduction. This linkage is at least partly a consequence of trade-offs between life-history traits, such as between current and future reproduction, and between current reproduction and lifespan [1]. The exact cause for these trade-offs is poorly understood, although several prominent hypotheses have been proposed. A dominant theory is that these trade-offs occur because of limitations in allocation of a limited resource, for example allocation of finite amounts of energy. This would require organisms to optimize reproductive allocation to different life-history traits in relation to their environment, with early investment reducing the resources available for body maintenance and late-life reproduction [2,3]. It has also been hypothesized that the presence of senescence and apparent trade-offs between early- and late-life functions may be linked to stronger evolutionary selection on functions that drive early-life growth and fertility [4]. The continuing influence of functions that were under strong selection early in life could lead to ageing and senescence because the force of natural selection at later ages is insufficient to reoptimize gene expression and cellular function to support fertility and survival at later ages (for example, selection for early-life growth may lead to elevated cancer risk later in life) [5]. Both hypotheses can explain why high early-life reproduction is associated with faster ageing, with the former hypothesis assuming that there are obligate limitations in allocation of resources [6], although what these are and the mechanisms by which allocation trade-offs are mediated remain poorly understood.

Traditional theories of resource allocation [7,8] assume that costs can occur from the variety of activities associated with reproduction, including attracting or competing for mates, mating, producing gametes, and caring for offspring, which all require energy and nutrient allocation. Each of these reproductive processes can impose a resource allocation cost, and these costs must be balanced by diverting resources from somatic maintenance or future reproduction. It has recently been highlighted that resource allocation to reproduction versus protection from ageing may be modulated by the nervous system in response to detection of different signals by sensory systems. This could allow for phenotypic plasticity in life-history investment in relation to perception of environmental conditions [9] but could lead to trade-offs with late-life function even if animals do not actually engage in social interactions or reproduce [10]. These systems may also be harnessed for sexual conflict, with opposite-sex individuals using secreted substances to alter the balance of a mating partner's investment in reproduction versus ageing to favour increased investment in their offspring [11]. In several invertebrates, associations between faster ageing and (normal) reproduction are independent of resource transfer to offspring [12], do not require fertilization [13], and can even occur in the absence of conspecific interactions [14,–16]. In Caenorhabditis elegans and Drosophila melanogaster, exposure to opposite-sex pheromones alone reduces lifespan [14–16]. This suggests that the detection of sensory signals from mates may modulate ageing in relation to predictive cues of reproduction, rather than costs occurring from the process of reproduction per se. Mating, in male Drosophila, can alleviate both the metabolic and mortality costs of pheromone exposure [14] effects that are dependent on expression of the neuropeptide corazonin [16], which is involved in metabolism, reproduction and reward pathways in this species [17]. Thus, neuroendocrine reward responses that occur with mating may be beneficial for survival, and help to reverse metabolic and life-history costs that occur in response to expectation of the mating opportunities that are driven by pheromone exposure.

In mammal species, reproductive activity is negatively associated with lifespan [18] and there is emerging evidence that sensory cues may modify this trade-off. In female house mice (Mus musculus) the trade-off between age at puberty and lifespan can be modified by olfactory cues (in soiled bedding) of adult females early in life leading to delayed puberty and an extension of lifespan [19]. Adult female house mice experience a reduction in lifespan due to mating, and this is independent of fertilization since females mating with sterile males also show reduced lifespan [20]. In male mice, it has been shown that alterations in metabolic rate and body weight with mating are linked to olfactory perception of female odorants via specific neurons in the accessory olfactory system [21]. Long-term effects of female odours on male lifespan or ageing have not been reported, although housing male rodents with females throughout life can improve survival [22] and reduce reproductive senescence [23]. This suggests that mating and/or a cue associated with female presence influences aspects of ageing in male mice.

The aim of this study was to test whether exposure to female odours would influence reproductive senescence in male mice. Based on previous research, we envisaged two competing hypotheses for how female odour exposure could influence reproductive ageing and mortality in male mice. (1) Prior invertebrate research suggests exposure to opposite-sex olfactory cues could hasten reproductive ageing and increase male mortality rates. If male mice show similar responses to Drosophila, exposure to female odours (via transplanted bedding) would hasten ageing, but mating would alleviate these effects (i.e. the effects of odours would be lessened when males have the opportunity to mate). Alternatively, (2) exposure to female odours could replicate the life-history changes that occur when male mice mate and could enhance survival and slow reproductive ageing. If exposure to female olfactory cues is the cause of the previously observed ageing benefits from mating, then these effects would be replicated in males exposed to female odours.

2. Results

This study entails a two (mating or not mating) by two (odour exposure or no odour exposure) factorial design. From 10 to 24 months of age, male mice were housed in one of the following four environments (housed alone; housed with two females; housed alone plus exposure to odours from two absent females; housed with two females plus exposure to the odours of two additional absent females). At 24 months of age, all males, regardless of prior social environment, were paired with two young females and assessed for fertility over a 64 day period. By 24 months of age male reproductive senescence is usually observable in male mice, in relation to a male's ability to sire litters [23,24], the number and quality of sperm produced [25], and investment in olfactory sexual signalling [26].

(a) . The effect of odour exposure and mating on mortality

There was a significant interaction between housing with females and odour exposure on survival to the end of the experiment at 26 months of age (Wald statistic: 4.78, p = 0.029). Mating without additional odour exposure tended to reduce male mortality compared to unmated males without odour exposure (Chi squared test for proportion alive: Chi-square statistic = 3.35, d.f. = 1, p = 0.067), as previously reported [22]; however, mating with additional odour exposure reversed this effect and led to males showing increased mortality compared to other groups (figure 1a). At the end of the study, 15 out of 30 males in the mating plus odour environment had died, compared to only 4 out of 30 in the mating only group (figure 1b; Chi-squared test: Chi-square statistic = 9.32, d.f. = 1, p = 0.002).

Figure 1.

Mortality rates in males housed in different social environments from 10 to 26 months of age. (a) The survival of males within each treatment group according to their age. (b) The proportion of males within each group that were dead by the end of the study, that were alive and successfully sired at least one litter, or that were alive but not fertile (did not produce any litters over the 64 days housed with two females). N = 30 animals per group at the start of the study.

(b) . Odour exposure influences fertility

At 24 months of age all males that remained alive were switched to being housed with young females for 64 days to assess late-life fertility. Of those animals remaining alive, a similar proportion of males in each group were fertile and produced at least one litter when housed with two females over this period (17 of 20 in the alone group; 18 of 21 in the alone plus odour group; 19 of 26 in the mating group; 12 of 15 in the mating plus odour group; figure 1b). Assessing the number of litters sired by males over the fertility testing period, males that were previously exposed to female odours produced fewer litters when allowed to mate later in life (figure 2a; main effect of odour exposure: F_1,64 = 5.71, p = 0.02). This effect occurred irrespective of whether males had prior mating experience (interaction between mating and odour exposure factors: F_1,64 = 0.52, p = 0.48). Prior mating experience as a main factor did not influence fertility at 24 months of age (F_1,64 = 0.24, p = 0.63).

Figure 2.

Reproductive ageing in male mice exposed to female mating and/or female odour exposure earlier in life. (a) The number of litters sired by fertile males in each treatment group from 24 to 26 months of age when housed with two young females. (b) The number of litters sired by males in mating environments across three periods of life, when housed with similar aged young females for similar periods of time (approx. two months). (c) Daily sperm production for males in each treatment group at 26 months of age, and a cohort of younger males (3 months old), calculated from testicular sperm head counts. (d) Testes mass of males in each treatment group. (e) Percentage of a testes cross-section that is comprised of interstitial space and (f) percentage of abnormal seminiferous tubules observed in male testes at the end of the study. Each dot represents a value for an individual mouse. Error bars show mean ± s.e.m. In (b), * represents p < 0.05 and n.s. represents p > 0.1 calculated using a Student's t-test.

Using data from individuals of the two earlier-life mating groups (one of which had additional odour exposure), we were also able to compare the fertility of males at 24 months of age to their rate of litter production at earlier ages, when they were paired with two equivalent young virgin females at each of 10 and 16 months of age. We observed that the number of litters sired by males over equivalent time periods was reduced between the 16- and 24-month age periods (figure 2b), indicating males had undergone reproductive senescence. There was no evidence of an effect of odour exposure on male fertility of these males at younger age periods of the experiment when comparing these two treatment groups (figure 2b), indicating that long-term exposure to female odours influences male fertility specifically during later ageing.

At the end of the fertility period males remaining alive were dissected to assess testicular sperm number and histology, to determine whether age-related changes at the level of the testis may contribute to fertility differences. In addition to the 26-month-old males from the four main treatment groups, we also included samples from a group of young adult, three-month-old control males (housed in control ‘alone’ environments), to establish whether reproductive senescence was observable in males at 26 months of age compared to younger males within our population of mice. Comparing young males to old males (irrespective of treatment groups), old animals showed fewer testicular sperm (Mann–Whitney test: p > 0.005; figure 2c), reduced testes masses (Mann–Whitney test: p < 0.0001; figure 2d), increased interstitial space within the testis (Mann–Whitney test: p > 0.005; figure 2e) and an elevated presence of abnormal seminiferous tubules (Mann–Whitney test: p < 0.01; figure 2f). However, there was no effect of prior mating experience or odour exposure on these parameters when assessed at 26 months of age (figure 2c–f; p > 0.1 for effect of mating experience, effect of odour exposure and the interaction between these variables in each analysis for testes mass, testicular sperm counts and interstitial space; we did not conduct a formal analysis of presence of abnormal seminiferous tubules due to pathology being observed in only a few individuals per treatment group).

(c) . Scent marking and body weight

Investment in olfactory sexual signalling is another aspect of reproductive investment that declines with age in male mice and can influence male attractiveness to females [26]. Given the non-invasive nature of this assessment, this was measured in all males at regular intervals through the study. There was an interaction between age and mating treatment on the rate of scent mark deposition (figure 3a; table 1). Males in the mating treatments reduced scent-marking behaviour from 10 to 13 months and then continued to scent mark at low rates from 13 to 24 months of age (figure 3a). By contrast, scent marking by males that were not mating increased up to 19–22 months and then declined, but remained higher than the mating treatment males between 22 and 24 months of age. There was also an interaction between mating and odour treatments, with the effects of mating on scent marking being less pronounced within the odour-exposed treatment groups (figure 3a; table 1).

Figure 3.

Changes in scent marking and body weight across the experiment. Males were repeatedly sampled for their scent marking rate (number of scent marks over 1 h) (a), urinary protein concentration (b) and body weight (c) at several different time points across the experiment. Data shown as mean ± s.e.m.

Table 1.

Results from mixed effects models assessing the effects of time (age at sampling), mating environment and odour environment on scent marking, urinary protein concentration and body weight. The random effect variance estimates for each timepoint and covariance estimates between timepoints for each model are shown on the right-hand side of the table.

factor	numerator d.f.	fixed effects			random effects
		denominator d.f.	F	p value	parameter	estimate	std error
scent marking
mating	1	115.171	14.322	<0.001	Var(time 1)	57.507	7.551
odour	1	115.171	1.081	0.301	Var(time 2)	60.823	8.059
time	5	104.084	5.547	<0.001	Var(time 3)	69.858	9.392
mating × odour	1	115.171	4.769	0.031	Var(time 4)	70.455	9.566
mate × time	5	104.084	5.194	<0.001	Var(time 5)	58.198	8.335
odour × time	5	104.084	0.646	0.665	Var(time 6)	46.233	6.824
mating × odour × time	5	104.084	1.13	0.349	Corr(2,1)	0.382	0.08
					Corr(3,1)	0.369	0.082
					Corr(3,2)	0.584	0.063
					Corr(4,1)	0.335	0.087
					Corr(4,2)	0.648	0.056
					Corr(4,3)	0.675	0.052
					Corr(5,1)	0.278	0.092
					Corr(5,2)	0.572	0.067
					Corr(5,3)	0.666	0.055
					Corr(5,4)	0.671	0.055
					Corr(6,1)	0.367	0.089
					Corr(6,2)	0.524	0.074
					Corr(6,3)	0.587	0.066
					Corr(6,4)	0.617	0.064
					Corr(6,5)	0.726	0.05
urinary protein
mating	1	110.435	0.641	0.425	Var(time 1)	21.249	2.853
odour	1	110.435	0.169	0.682	Var(time 2)	51.633	6.9
time	3	96.097	29.523	<0.001	Var(time 3)	44.2	6.572
mating × odour	1	110.435	0.008	0.927	Var(time 4)	40.046	6.047
mating × time	3	96.097	1.273	0.288	Corr(2,1)	0.489	0.075
odour × time	3	96.097	0.615	0.607	Corr(3,1)	0.238	0.093
mating × odour × time	3	96.097	0.593	0.621	Corr(3,2)	0.379	0.097
					Corr(4,1)	0.335	0.097
					Corr(4,2)	0.331	0.11
					Corr(4,3)	0.315	0.096
body weight
mating	1	114.371	5.575	0.02	Var(time 1)	4.736	0.622
odour	1	114.371	0.304	0.583	Var(time 2)	5.345	0.705
time	5	89.557	19.831	<0.001	Var(time 3)	7.436	0.99
mating × odour	1	114.371	0.508	0.477	Var(time 4)	8.546	1.153
mating × time	5	89.557	9.355	<0.001	Var(time 5)	8.496	1.207
odour × time	5	89.557	0.616	0.688	Var((time 6)	7.326	1.148
mating × odour × time	5	89.557	1.093	0.37	Corr(2,1)	0.892	0.019
					Corr(3,1)	0.791	0.036
					Corr(3,2)	0.871	0.023
					Corr(4,1)	0.724	0.046
					Corr(4,2)	0.825	0.031
					Corr(4,3)	0.928	0.013
					Corr(5,1)	0.62	0.061
					Corr(5,2)	0.729	0.046
					Corr(5,3)	0.841	0.03
					Corr(5,4)	0.872	0.024
					Corr(6,1)	0.635	0.064
					Corr(6,2)	0.742	0.052
					Corr(6,3)	0.838	0.036
					Corr(6,4)	0.867	0.03
					Corr(6,5)	0.923	0.019

We also assessed male urinary protein concentration at various time points. Major urinary proteins comprise the vast majority of protein in urine [27] and play key roles in olfactory communication between same- and opposite-sex individuals [28]. The concentration of urinary signalling proteins declines from middle age and has been associated with lower attractiveness of male mice to females [26]. We observed a strong effect of age on urinary protein concentration, with concentration declining from 13 months of age. There was no significant effect of mating or odour exposure on urinary protein concentration, or interactive effects of these treatments with age (figure 3b; table 1).

Body weight of males was assessed across the experiment. There was a significant interaction between male age and the effect of mating, with the effect of mating on body weight becoming stronger over time (table 1; figure 3c). There was no significant effect of odour exposure, or interaction between age, mating or odour environment, on body weight across the study.

3. Discussion

These results show that exposure to female odours and direct experience of mating with females have different effects on male mortality and reproductive senescence. The primary goal of this study was to test whether earlier experience of mating and female odours would influence subsequent fertility later in life. We observed that males previously exposed to female odours several times a week via transplanted bedding have reduced fertility later in life, in terms of the number of litters they are able to sire with young females over a two-month period. These effects on fertility were not modified if males were also allowed to mate with females earlier in life, and exposing males to females that they could mate with did not by itself influence reproductive senescence. Thus, the perceived presence of females via olfactory cues within the environment, when not matched to the physical presence of those females in the environment, impairs later-life fertility. These results bear some similarities to findings in male Drosophila, where the perception of females within the environment reduces lifespan [14,16]. In that context, providing male flies with the ability to mate with additional females reduces the effects of pheromones on male ageing. We do not find evidence for this latter phenomenon in mice, although it is notable that the female flies needed to be provided at a high density to reduce the lifespan response to pheromones in Drosophila [14], and it is possible that a higher density of additional females could ameliorate the effects on reproductive ageing.

We did not observe that prior experience of mating with females influenced male fertility later in life. This result contrasts to that of Schmidt et al. [23], who demonstrated in B6129SF1/J hybrid mice that males allowed to mate with females across life were able to sire more litters during advanced ages than males that had previously been housed alone. Our experiment was designed to try and replicate this effect and test whether the response was modulated or replicated by exposure to female odours. There are several possible reasons for the lack of concordance between results. Our experiments were conducted using wild-derived mice that were the descendants of 50–60 mice caught in the wild. They have a different genetic background to the F1 hybrid mice used by Schmidt et al. [23], which were a cross of two inbred mouse strains that have undergone intense laboratory selection. The effects reported by Schmidt et al. [23] were also stronger in the males they studied at later ages (e.g. 30 months of age). Our analysis of reproductive senescence was assessed when males were aged between 24 and 26 months of age. At this age-point reproductive senescence was still observable in the mice used in this study, as documented by reduced fertility and scent marking compared to earlier ages, and testicular degeneration compared to young mice. It is possible, however, that an effect of mating might have been observable if we had studied our mice at even later ages when there might be more pronounced deficits in reproductive ability.

In addition to the pro-ageing effect of female odour exposure on male fertility, we observed that exposure to female odours modified the effects of mating on male mortality. Prior work has demonstrated that allowing male rodents to mate can improve survival [22,29,30]. Our results are consistent with this, but only when males are housed in the absence of additional exposure to female odours. When males are exposed to extra odours from another two unfamiliar females this effect is reversed and mortality is increased. Thus, mortality costs associated with social (including competitive) interactions and communication may be driven by olfactory cues in mice, as in some invertebrates [15,16]. However, in our experiment these additional costs for mortality were observed in males when they were also allowed to mate, which is the opposite response to that observed in male Drosophila, where allowing males direct interactions with females reduces the costs of pheromone exposure [14,16].

The mechanisms responsible for reduced longevity in males exposed to female odour remain to be established. One possibility is that earlier exposure to female odours increased male reproductive investment early in life, and this generated the associated fertility and mortality trade-off. We did not observe a change in early-life scent marking, or early-life fertility for those housed with females, but it is possible that another aspect of reproductive investment was altered by exposure to female olfactory cues. Sensing of females within the environment may stimulate changes in males that increase investment in male–male competition, an important component of reproductive investment in male mice. While males in this study did not encounter males, they could conceivably show changes in their underlying physiology that would be beneficial if they did encounter males in their environment. It would be interesting to understand whether males previously exposed to female odours show increased competitive ability in aggressive interactions, or whether exposure to male odours modifies ageing responses. In this study mating and odour responses were compared to isolated males that were housed singly and not exposed to female odours, which is unlikely to represent typical social conditions for wild mice, except during dispersal. Mice are found in extremely wide-ranging densities (from up to ten per m² to one per 100 m² [31]) and varying social conditions, and future research understanding whether mice show life-history responses to graded levels of pheromone exposure from different numbers of conspecifics may help to understand how animals respond to cues that predict the different levels of social structure that are commonly found in nature.

It has recently been shown that exposure to female odours induces immediate increases in energy expenditure, effects that are sufficient to reduce adiposity when mice are provided with high calorie diets that promote energy storage [21]. These effects were linked to detection of female odours via specific Gαo-expressing neurons of the accessory olfactory system, indicating that detection of opposite-sex pheromones by the olfactory system can lead to central regulation of metabolism. Male mice exposed to female odours through this study would have thus shown repeated increases in energy expenditure in response to female bedding and these responses may lead to the associated trade-offs later in life. It would be of interest to determine whether the life history responses we report in this study are also mediated by the accessory olfactory system. Such determination would help to firmly establish that life-history changes are driven by cues perceived from the olfactory system, rather than other chemical/biological cues found in soiled female bedding.

It would also be of interest to understand the physiological causes for reduced late-life fertility in males exposed to female odours. These males had equivalent sperm counts to animals not exposed to female odours and similar testicular pathology scores. Our late-life mating trials broadly followed the protocol of Schmidt et al. [23] in that we measured the ability of males to sire litters when presented with two females. A decline in the fertilization ability of male mice with age has also been noted in other studies, which have implicated both impaired neuroendocrine control of sexual behaviour and deficits in the production of sperm [32]. It would be of interest to determine whether neuroendocrine control of reproduction is altered in ageing mice previously exposed to female odours, and whether aspects of reproductive behaviour are impaired in these ageing male mice.

Our results add to the growing evidence that indirect cues from mates can influence the ageing process. Responses induced by opposite-sex odours may be important in priming physiology in anticipation of reproduction so that animals can effectively engage in social interactions that are important for fitness. However, the responses that are induced could also be a cause of life-history trade-offs and hastened ageing, leading to negative effects on later fitness that are detectable even in the absence of direct, earlier interaction with mates or competitors.

4. Methods

(a) . Mice

Mice were captive-bred house mice (Mus musculus) that were the 7th to 9th generation progeny of a population of 50–60 wild mice captured from a chicken farm in the northwest of Sydney, Australia [33]. The use of a wild-derived population of mice means that genes underlying growth, reproduction and longevity are more likely to represent those found in natural conditions, in contrast to inbred laboratory strains of mice that have undergone substantial selection for propagation in laboratory conditions [34]. Animals were weaned at 28 days of age at which point males were housed singly and females were housed with their female siblings until the beginning of the experiment. Mice were maintained on a 12 : 12 h reversed light cycle with lights on at 18.00. A dim red light was used for experimental procedures, which were conducted in the dark phase. Mice were housed in 315 × 180 × 125 mm³ cages lined with corncob bedding and provided with tissues and shredded newspaper for nesting material. Animals were fed Mouse and Rat Maintenance Pellets purchased from Gordon's Speciality Stock Feeds (Yanderra, New South Wales, Australia) and water was provided ad libitum.

(b) . Experimental design

Experimental male mice were kept singly housed from weaning until the start of the experiment at 40 weeks of age without exposure to other males or females. At this time point 120 male mice were randomly assigned to four equally sized (n = 30) treatment groups: housed alone (a; control), housed with two females (b), housed alone and exposed to female odour from two females (c), and housed with two females and exposed to the odours of two additional females (d). A sample size of 30 per group was used as we estimated that we would need 20–24 mice per treatment to detect a change in male litter production at 24 months of age after prior male mating experience. This was calculated with power analysis using the effect size of the difference in the number of litters produced between mated and non-mated male mice at 24–26 months of age reported by Schmidt et al. [23], with power of 0.8 and a significance value of p < 0.05. We expected that some mice would die by 24 months of age, but with an initial sample size of 30 per group we would have sufficient power to determine effects of social environments on reproductive senescence that have previously been reported. Animals were exposed to differing social environments starting at 40 weeks of age, because in our previous study we had noted slightly increased mortality in mating environments earlier in life [22], and we wished to mitigate these effects by pairing males later in life.

(i) . Mating environment

In the mating environments, each male mouse was housed within a cage with two sexually naive females that were 8–12 weeks of age at the time of pairing. To ensure that females remained sexually receptive, these females were switched with new young females after eight months of pairing when males were 16 months of age. Males remained with these females until the final tests of fertility at 24 months of age. All offspring produced within a cage were counted and removed from the cage when they were 28 days of age. When comparing rates of litter production across different age-points, we used the number of litters that females housed with males birthed over the first 90 days of pairing. This would approximately capture litters that had been sired by males over the first 60–65 days of pairing (equivalent to the late-life fertility testing period) since the length of postpartum pregnancies is approximately 25–30 days in this population of mice [35].

(ii) . Odour exposure

Males that were allocated to an odour exposure treatment were exposed to the odours of two females over the same experimental period as those animals that were mating. Twelve cages of female odour donors (two females per cage) were initially used, that were 8–12 weeks of age at the start of the experiment and were housed in pairs. Three to four times per week approximately five grams of soiled corncob substrate and nesting material was taken from the cage of females and added to the males cage, with an equivalent amount of bedding removed from the males cage. These females were replaced with 12 young females when males were 16 months of age, providing scent exposure from females that were of an equivalent age to those that were mating in the mating environment. Odour exposure continued throughout the experiment, including through the fertility testing period, until males were euthanized by cervical dislocation at 26 months of age.

(c) . Fertility assessments

(i) . Late-life fertility testing period

At 24 months of age, each male that remained alive, regardless of their earlier environment, was housed in a cage with two young (8–12 weeks of age), sexually naive females. These animals were allowed to breed continuously over following 64 days, with the number of offspring and litters weaned (at 28 days) by each female recorded over this period. After 64 days males were euthanized, and the testes dissected for sperm and pathology analysis. The females that had been paired with males were allowed to finish pregnancy and birth any final litters that males sired before they were removed from the cages. The total number of litters that males sired was used as the assessment of late-life fertility, following Schmidt et al. [23].

(ii) . Testicular sperm counts

At euthanasia, testes were dissected, with one testis frozen at −80°C for testicular sperm counts. The tunica albuginea was removed from the frozen testis and then the testis was placed into a test tube containing 4 ml DMSO/saline solution. Using a Homogeniser Turrax at 7500 rpm, the testis was homogenized for 1 min. Fifty microlitres of trypan blue was added to the tube, covered with parafilm wax and vortexed for five seconds. Twenty-five microlitres of the homogenized solution was added to each well of a haematocytometer and spermatid heads counted. Sperm counts were converted to daily sperm production by calculating the total number of sperm in the mouse testis suspension, correcting this for the weight of the testis used, and dividing this value by 4.84, the time divisor that is recommended for mouse samples [36].

(iii) . Pathology analysis

One dissected testis was immediately fixed in 10% neutral buffered formaldehyde solution and then subsequently embedded in wax. Seven-micrometre transverse sections were taken from the middle plane and stained with hematoxylin and eosin (H&E). The sections were photographed under a 10× objective and quantified using ImageJ. The area of interstitial space between seminiferous tubules was calculated by converting the picture into an 8-bit image and masking the seminiferous tubules by hand using the spraying can function, carefully following the tubule contours. The threshold and measure commands were then used to obtain the interstitial space area as a percentage of the total image. The number of seminal tubules with a disruption in normal layering was counted and presented as a percentage relative to the total number of tubules in a field of view under the 10x objective.

(iv) . Scent marking

Scent marking was assessed for each male at the start (10 months of age) and at various time points throughout the study period. Males were placed into a clean cage lined with Whatman Benchkote absorbent paper and allowed to scent mark for one hour, after which they were returned to their home cage. Under UV light an experimenter who was blind to each individual's treatment group counted the number of scent marks produced by each male.

(v) . Urinary protein concentration

Urine samples were collected from males prior to the experiment, and again at 13, 19 and 24 months of age. Males were placed on a wire mesh above a clean cage and allowed to urinate into the cage bottom below over a 2 h period. Urine was collected and frozen at −20°C before analysis. Total protein concentration was assessed using the Commassie Protein Assay (Sigma Aldrich, Darmstadt, Germany). The total protein concentration of urine reflects total major urinary protein concentration [27,37]. Urinary dilution was corrected for by measuring creatinine concentration in urine using the alkaline picrate assay [27]. Protein concentration was then divided by creatinine concentration and this value used in subsequent analysis.

(d) . Survival assessment

Across the experiment males were assessed for survival using defined criteria. Animals were scored as dead or when they were considered to be severely moribund and unlikely to survival an additional 48 h [38]. Severe moribundity was indicated by one or more of the following clinical signs: rapid weight loss; lack of responsiveness to manual stimulation; trembling/hunched or immobile posture; laboured and irregular breathing; severe abdominal enlargement; a severely ulcerated tumour. The age at which a moribund mouse was euthanized was used as an estimate of its lifespan.

(e) . Statistics

Statistical analyses were carried out in SPSS version 28. Prior to analysis, data were visually assessed to determine conformation to parametric assumptions of specific tests and where necessary data log transformed or square root transformed to fit with the associated assumptions, or a non-parametric test used.

Data for body weight, scent marking, and urinary protein concentration were analysed using a general linear mixed model (GLMM). GLMMs included animal ID as a random effect (subject) that was repeated within measurement time using the unstructured covariance structure (termed ‘Unstructured: Correlation Metric’ within SPSS). We tested for an effect of mating experience, an effect of odour exposure, and whether there was an interaction between these two factors. We also tested whether these variables showed an interaction with time. Where an interaction between mating and time was present, we conducted individual tests within each mating environment to determine how changes in the trait of interest occurred over time.

Analysis of fertility, sperm number, and testicular pathology was conducted with a general linear model, testing for an effect of mating, an effect of odour exposure and an interaction between these two factors. Where an interaction was present we conducted individual tests within mating or non-mating groups to determine whether there was an effect of odour exposure. Comparisons between young and old animals were conducted with a Mann–Whitney test due to differences in variance between young and old animals.

Overall differences in survival between groups were tested using Cox Regression Survival Analysis (SPSS, version 25), including mating and odour exposure as separate fixed factors. Follow-up analysis of differences in survival rates between two specific groups was conducted using a Chi-squared test (Prism, version 9).

Ethics

This research was approved by the UNSW animal ethics committee, application ID: 4408614.

Data accessibility

All data are provided as electronic supplementary material [39].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors' contributions

M.G.: conceptualization, data curation, formal analysis, validation, writing—original draft; H.T.: investigation, methodology, writing—review and editing; C.N.: investigation, writing—review and editing; R.C.B.: conceptualization, formal analysis, methodology, project administration, writing—review and editing.

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

Conflict of interest declaration

We declare we have no competing interests.

Funding

This research was funded by an Australian Research Council Discovery Grant DP190100899 to R.C.B. and M.G., and a Marsden Grant UOO1819 to M.G.