Rodents communicate extensively through odors released in urine, feces, and scent glands. Urine of the house mouse (Mus musculus) contains a complex mixture of volatile odorants and nonvolatile proteins which conveys information about sex, reproductive and social status, individual identity (Hurst, 1990; Hurst et al., 2001). In male mice, many urinary compounds are regulated by androgens (Schwende et al., 1986) which profoundly alter reproductive physiology in female mice (Whitten, 1956; Bruce, 1959). A well-known endocrine effect of male odor is the acceleration of puberty onset in female mice (Vandenbergh, 1967, 1969). Peripubertal exposure to male urine or soiled bedding accelerated vaginal opening, uterine growth, first estrus and ovulation (Vandenbergh, 1969; Colby and Vandenbergh, 1974; Bronson, 1975; Vandenbergh et al., 1975); but urine or soiled bedding from castrated male had no effect (Lombardi et al., 1976; Jouhanneau et al., 2014). To date, the following androgen-dependent volatile compounds are reported to increase uterine weight in peripubertal mice: (1R, 5S, 7R)-3,4-dehydro-exo-brevicomin (DHB), 6-hydroxy-6-methyl-3-heptanone (HMH), (S)-2-sec-butyl-4,5-dihydrothiazole (SBT) and β-farnesene (Novotny et al., 1999a,b). Additionally, the contribution of nonvolatile major urinary proteins (MUPs), which bind DHB, HMH, SBT and β-farnesene, in the Vandenbergh effect cannot be excluded (Bacchini et al., 1992; Robertson et al., 1993; Mucignat-Caretta et al., 1995; Novotny et al., 1999b; Sharrow et al., 2002).

In addition to accelerating vaginal opening, we reported recently that peripubertal exposure to male-soiled bedding induced expression of male-directed odor preference in female mice when they were adults (Jouhanneau et al., 2014). At postnatal day (PD) 45, female mice exposed to male-soiled bedding for 18 days from weaning (i.e., from PD21 to PD38) investigated longer the odor of male-soiled bedding than that of female-soiled bedding, whereas females exposed to clean bedding did so only at PD60. Exposure to castrated male-soiled bedding also triggered an early attraction to intact male odor but without effect on vaginal opening, suggesting that the physiological and behavioral effects of male chemosignals in female mice are supported by different mechanisms. Additionally, we reported that female mice exposed to male-soiled bedding during the peripubertal period also preferred intact male odors over castrated male odors at PD45. Taken together, this suggested that female mice were attracted to androgen-dependent volatile compounds of male-soiled bedding previously learned during the peripubertal period. Our characterization of volatile compounds in intact male-soiled and castrated male-soiled beddings by gas chromatography coupled to mass spectrometry revealed that the major androgen-dependent volatile compounds present in intact male-soiled bedding are DHB, HMH and SBT (Jouhanneau, 2014). These three compounds have also been detected in urine of male mice (Liebich et al., 1977; Novotny et al., 1984, 1999a; Schwende et al., 1986; Flanagan et al., 2011; Supplementary Figure 1 and Table 1).

Taken together, these results suggest that the urinary compounds DHB, HMH and SBT present in male-soiled bedding, in addition to accelerating vaginal opening, could also be responsible for the early expression of male-directed odor preference in female mice. A study by Jemiolo et al. (1985) further supports this hypothesis by showing that a mixture of DHB+SBT is attractive to female mice in adulthood. Therefore, the present study explored whether peripubertal exposure to DHB, HMH and SBT accelerated vaginal opening and induced adult expression of male-directed odor preference in female mice.

Exposure to DHB, HMH, SBT, a mixture of DHB+HMH+SBT or male urine from PD21 to PD38 advanced vaginal opening in comparison to water exposure. Half of the animals exposed to a chemosignal (i.e., DHB or HMH or SBT), a mixture of the three chemosignals or male urine showed vaginal opening between PD24 and PD27 whereas half of the animals exposed to water showed vaginal opening on PD30 (Figure 2A). One-Way ANOVA revealed significant differences in the mean age at vaginal opening [F_{(5, 94)} = 3.81, p = 0.003] (Figure 2B). Indeed, females exposed to a chemosignal, a mixture of the three chemosignals or male urine had vaginal opening earlier than females exposed to water (DHB vs. water groups: p < 0.001; HMH vs. water groups: p = 0.045; SBT vs. water groups: p = 0.016; DHB+HMH+SBT vs. water groups: p = 0.043; male urine vs. water groups: p < 0.001). No differences in the mean age at vaginal opening between females treated with chemosignal(s) or urine were observed.

Exposure to a mixture of DHB+HMH+SBT or male urine from PD21 to PD38 triggered male-directed odor preference in female mice at PD45. Indeed, females exposed to a mixture of DHB+HMH+SBT or male urine investigated longer the odor of male urine than the odor of female urine [t₍₁₆₎ = 2.18, p = 0.044; t₍₁₆₎ = 3.15, p = 0.006 respectively] (Figure 3). By contrast, no differences in the time investigating male vs. female odor were observed for female mice exposed to DHB [t₍₁₆₎ = 0.30, p = 0.77], HMH [t₍₁₄₎ = 0.62, p = 0.54], SBT [t₍₁₄₎ = 1.68, p = 0.12] or water [t₍₁₈₎ = 1.51, p = 0.15] (Figure 3). Individual data showed that estrous cycle stage was distributed randomly in each group and therefore does probably not influence expression of male-directed odor preference (Figure 4).

The present study demonstrates that the combined peripubertal exposure to DHB, HMH and SBT, in addition to accelerating vaginal opening, also induced expression of male-directed odor preference in female mice in adulthood. Interestingly, each of these three compounds accelerated vaginal opening but was inefficient in stimulating male-directed odor preference. In accordance with our previous study (Jouhanneau et al., 2014), these present results confirmed that female attraction to male odor is not the consequence of the acceleration of vaginal opening by male chemosignals.

Early studies suggested that the male chemosignal responsible for puberty acceleration was a peptide bound to a protein (Vandenbergh et al., 1975, 1976). Then, volatile nature of chemosignal was considered when Novotny et al. (1984) isolated the MUP ligand DHB in the fraction active for puberty acceleration. DHB and also HMH and SBT accelerated uterine growth in prepubertal female ICR(CD-1) mice to a level comparable to the effect of male urine (Novotny et al., 1999b). HMH also increased uterine weight in (SJL/J × SWR/J) F1 hybrid mice (Novotny et al., 1999a). Here, we showed that DHB, HMH and SBT accelerated vaginal opening in another mouse strain derived from the Swiss lineage (Lynch, 1969; Beck et al., 2000; Yalcin et al., 2010). In Swiss mice, each compound was as efficient as male urine to accelerate vaginal opening and the mixture of the three compounds had no synergistic effect. However, no effect of MUP ligands on uterine weight has been reported in Swiss mice (Mucignat-Caretta et al., 1995). Future research will need to determine the reasons for this discrepancy. Moreover, the effect of MUP ligands on puberty acceleration appears to be lineage-specific. Indeed, in contrast to male urine, DHB and SBT did not accelerate uterine growth in female BALB/cJ mice (Flanagan et al., 2011), which are derived from the Castle lineage (Beck et al., 2000). According to Flanagan et al. (2011), an unknown nonvolatile compound present in male urine would participate in the acceleration of puberty onset in female mice.

Additionally, the present study reported that peripubertal exposure to male urine also induced a preference for male over female odors in female mice when they were adults. In contrast, female mice exposed to water did not show attraction to male odor. These results confirmed those of our previous study showing that female mice expressed a male-directed odor preference at PD45 after peripubertal exposure to male-soiled bedding whereas females exposed to clean bedding did not (Jouhanneau et al., 2014). Furthermore, the present findings revealed that DHB, HMH and SBT are sufficiently responsible for the expression of male-directed odor preference when used in combination. Our chemical analyses showed that SBT is exclusively produced in urine of male Swiss mice and DHB and HMH levels are higher in male urine than in female urine. Hence, female mice exposed to a mixture DHB+HMH+SBT were attracted to male odor probably because male urine is the only stimulus which releases these three compounds. Similarly, female mice exposed to male urine were attracted to male odor because they probably perceived the attractive mixture DHB+HMH+SBT in male urine. As female mice exposed to water did not display male-directed odor preference, we hypothesized that the attraction to the mixture DHB+HMH+SBT was learned during the peripubertal exposure to male odor. An alternative hypothesis would be that groups exposed to the mixture or to male urine expressed a preference for male odor because their hormonal status was different than those of control group. However, analysis of individual data revealed that estrous cycle stage had no effect on expression of male-directed odor preference. In agreement with this finding, Moncho-Bogani et al. (2004) demonstrated the hormonal independence of male-directed odor preference. Indeed, female mice still preferred investigating the odor of male mouse rather than the odor of female mouse following ovariectomy.

Moreover, our findings showed that exposure to each individual compound of the mixture (i.e., DHB or HMH or SBT) were not enough to induce female attraction to male odor. It suggests that at least two compounds in the mixture DHB+HMH+SBT act together to release female attraction to male odor. Interestingly, Jemiolo et al. (1985) reported that DHB and SBT released no behavioral response when they were presented separately to adult female mice but were attractive when they mixed together; therefore the combined exposure to DHB+SBT in the mixture could be responsible for the induction of male-directed odor preference observed in the present study.

The induction of male-directed odor preference after prior exposure to some volatile compounds of male urine reported in the present study contradicts previous findings (Ramm et al., 2008; Roberts et al., 2010). Indeed, these authors showed a lack of female attraction to male odor after prior exposure to the volatile fraction of male urine. The differences in methods of exposure to the odorant stimulus could explain the discrepancies in results. Indeed, in the present study, female mice had physical contact with the odorant stimulus since it was applied on their oronasal groove. In contrast, in previous studies, volatile components of male urine freely diffused into the cage. Allowing female mice to have direct physical contact with male-soiled bedding activated the vomeronasal organ (VNO), whereas preventing direct contact to male-soiled bedding did not (Kimoto et al., 2005). This suggests that airborne molecules from male-soiled bedding are not detected by the VNO in female mice. In the light of these findings, we hypothesized that the activation of the accessory olfactory system through physical contact with the liquid mixture DHB+HMH+SBT during peripubertal exposure was probably responsible for the acquisition of an attraction response to these three compounds in female mice. To support this hypothesis, it has been previously reported in female mice that 1/DHB, HMH and SBT individually activated VNO neurons in vitro (Moss et al., 1997; Zhou and Moss, 1997; Leinders-Zufall et al., 2000; Del Punta et al., 2002); 2/the mixture DHB+SBT applied on the oronasal groove induced Egr-1 expression in the accessory olfactory bulb (AOB) (Brennan et al., 1999) and 3/bilateral lesion of the AOB before exposure to male-soiled bedding in adulthood disrupted subsequent expression of male-directed odor preference (Martínez-Ricós et al., 2008).

However, the contribution of the main olfactory system in induction of male-directed odor preference cannot be excluded. Indeed, destruction of the main olfactory epithelium (MOE) within 1 week before the behavioral test abolished expression of male-directed odor preference in female mice (Keller et al., 2006a) whereas ablation of the VNO did not (Keller et al., 2006b). Interestingly, airborne urinary volatiles from male mice activated the AOB in female mice through the MOE-main olfactory bulb-medial amygdala-AOB pathway (Martel and Baum, 2007, 2009; Kang et al., 2009). Identity of the volatiles that activate this pathway has not yet been determined but it was previously reported that DHB, HMH and SBT, in addition to activating VNO neurons, also activated MOE neurons in vitro (Ziesmann et al., 2002). Hence, after acquisition of attraction response to the odor mixture by VNO activation during odor exposure, DHB, HMH and SBT emanating from male urine probably induced appetitive memory recall through MOE activation during the behavioral test; which resulted in induction of male-directed odor preference. Destruction of the MOE or the VNO in female mice before peripubertal exposure to a mixture of DHB+HMH+SBT or before the behavioral test will provide insight into the role of each olfactory sense organ in mediating the effect of MUP ligands on male-directed odor preference.