Delayed aging has a cost: reduced fertility, redirection of maternal activities and earlier onset of puberty in p66 mice

Targeted mutation of the p66 gene, in mice, results in reduced oxidative stress, increased metabolic rate and in an increase in lifespan associated to a delayed aging without any apparent “side effect”. The antagonistic pleiotropy theory of aging predicts that for each immediate benefit there is a price to pay later in life, for example a negative correlation between longevity and reproductive success has been reported. Main aim of this study was to investigate whether the increased lifespan in the p66 mice has a cost in terms of their overall fitness. To this purpose we looked at repeated breeding cycles, breeding under stressful conditions, mean number of pups per nest and time of the onset of puberty. Since p66 mice are characterized by a decreased emotionality we also investigated whether this characteristic may be mediated by changes in early environmental factors, such as maternal care. To this purpose, we tested the effects of the p66 deletion on maternal behavior from postnatal day 1 to 8. P66 mice showed an overall earlier onset of puberty in addition to a reduced fertility both over repeated breeding cycles and under challenging stressful situations; mean number of pups did not differ between the two genotypes. A redirection of maternal behavior was observed in mutant mice as a function of their feeding pattern, which might account for the anticipated onset puberty observed in the offspring. Taken together our results show that p66 is important for optimal reproductive success. The lack of the gene is able to affect sexual maturation, an effect possibly mediated by changes in maternal care.


Introduction
Aging may be defined as a late-onset genetic disease resulting from mutations that strike very late in life at ages beyond the action of natural selection, leading ultimately to death (Partridge and Gems, 2002).The trade-off theory of aging predicts that mutations producing beneficial effects early in life, but later increasing the rate of aging, will be incorporated into a population because natural selection will act more strongly on the early beneficial effects than on the later deleterious ones (Charlesworth, 1994;Partridge and Barton, 1993;Williams, 1957).
P66 Shc is the first mammalian gene whose mutation was shown to increase resistance to oxidative stress (OS) and to prolong lifespan in mammals.Targeted mutation of this gene in a 129Sv/Ev mouse strain (knock out -KO -mice: p66 Shc-/-) results in normal mice from a developmental point of view, with a remarkably high resistance to oxidative stress, and a 30% increase in longevity when compared to the same age wild-type control group (p66 Shc+/+ -WT -) (Migliaccio et al., 1999).
In former studies we characterized the behavioral phenotype of KO and a profile of reduced emotionality emerged, differences with WT subjects becoming more pronounced with age (Berry et al., 2007).This piece of data is in agreement with a number of evidence indicating increased anxiety as a behavioral trait associated to the progression of aging (Bessa et al., 2005;Francia et al., 2006).Indeed p66 Shc-/-subjects revealed a reduced emotionality already at adulthood and this was associated to lower central levels of oxidative stress markers and increased basal levels of BDNF (Berry et al., 2008) (a neurotrophin playing a role in several aspects of neuronal plasticity, mood tuning and metabolism (Cirulli et al., 2004;Gomez-Pinilla, 2008;Thoenen, 1995) thus suggesting that in these mutants reduced OS throughout lifespan might prevent some of the behavioral effects of aging, particularly in response to emotionally arousing stimuli.
Beyond the emotional trait, and consistently with the observed delay in the aging process, we observed a general better health status in mutant subjects at senescence (Berry et al., 2008), while at adulthood KO mice showed improved brain functions in addition to a more efficient homeostatic control of the immune-endocrine functions which allows to readily cope with changes in the internal milieu (Berry et al., 2008;Berry et al., 2009 submitted) In mammals and other vertebrates, aging is associated with a natural decline in the functionality of physiological systems involving reproduction, metabolism, neuroendocrine and immune responses (and many others) of the organism.In humans some traits of aging like menopause or the decline in renal function intrinsically pertain to the physiology of this process, others, like coronary artery disease, represent pathophysiological features since they are age-related diseases not always present in aging individuals (Troen, 2003).
A number of age-related pathologies have been associated to a disruption of metabolic signals.In particular overweight, increased fat deposition and fat-related disorders are considered prominent risk factors for many diseases of old age and are though to exert a life-shortening action.Diet interventions involving a caloric restriction (CR) without malnutrition have been shown to result in a delay in the onset of both physiological and pathological changes associated with aging in mice and rat models (Masoro, 1993) and in non-human primates (Mattison et al., 2007) and to robustly extend lifespan in every organism in which it has been tested.
A work by Berniakovich and colleagues (Berniakovich et al., 2008) has recently shed light on the function and complex molecular mechanism of action of p66 Shc and on its possible role in the aging process through changes in metabolism.P66 Shc-/-mice are characterized by increased basal metabolism, reduced fat development and increased insulin sensitivity of peripheral tissues (Berniakovich et al., 2008).Thus it has been proposed that the p66 Shc gene in addition to regulating the effect of insulin on energetic metabolism might also accelerate aging favoring fat deposition and fat-related disorders by increasing intracellular oxidative stress.
In this context given the role of p66 Shc on oxidative stress and metabolism, and taking into account the overall positive behavioral features observed in mutant subjects, an obvious question is why this gene has not been negatively selected by natural selection.A large body of evidence reports a negative correlation between lifespan and reproductive success (Kuningas et al., 2008;Partridge et al., 2005) thus it is intriguing to speculate that one of the reasons for p66 Shc to be still a part of the mammalian genome is related to a pleiotropic effect on reproductive success.
In order to test this hypothesis we have assessed fertility, ability to raise the off-spring and time of sexual maturation in p66 Shc mutant subjects.Lack of the p66 Shc gene has been shown to play a role in shaping both the emotional and the metabolic phenotype of mutant subjects.Changes in the response to stress/emotionality can be affected early during development by environmental factors.In rodents, the two weeks following birth represent a critical window in which maternal behaviors can impact the physical and emotional development of the offspring, including future adrenal/stress sensitivity (Cirulli et al., 2003;Schmidt et al., 2006).Maternal influence on adrenal responses of the infant is mediated primarily through changes in feeding, possibly through availability of leptin in the milk (Suchecki et al., 1993).This adipose-derived hormone regulates energy homeostasis by stimulating coordinated changes in energy intake and expenditure in response to changes in energy stores (Campfield et al., 1995;Halaas et al., 1995;Pelleymounter et al., 1995a).A role for this hormone in development has been suggested by the presence of neuroendocrine and structural neuronal abnormalities in ob/ob mice with genetic leptin deficiency (Campfield et al., 1995;Halaas et al., 1995;Pelleymounter et al., 1995a).Additionally it is also involved in signaling the neuroendocrine response to starvation (Ahima et al., 1996;Carro et al., 1997;Legradi et al., 1997), the timing of puberty (Ahima et al., 1997;Chehab et al., 1997;Cheung et al., 1997) and regulation of the HPA axis (Bornstein et al., 1997;Costa et al., 1997;Heiman et al., 1997;Raber et al., 1997).Thus a second aim of this study was to assess whether changes in maternal behavior of the mutants might also contribute to the ontogeny of their emotional and metabolic traits.To this purpose, we tested the effects of p66 Shc deletion on maternal behavior as well as on leptin levels in the offspring in response to a metabolic/stressful challenge such as a 8 hrs maternal separation (Schmidt et al., 2006).
Based on data showing reduced emotionality in p66 Shc-/-adult mice, we expected the mutants to experience higher levels of maternal care buffering the response of the offspring to a stressful challenge (Cameron et al., 2005).

Animals
Experimental subjects were p66 Shc+/+ (WT) and p66 Shc-/-(KO) mice of the 129Sv/Ev strain, generated as previously described (Migliaccio et al., 1999).Founders were obtained from Charles River and a colony was successively established in the animal facility of the Section of Behavioral Neuroscience at the Istituto Superiore di Sanità (Rome, Italy).Changes in fertility were noted on subjects bred in the EIO p66 Shc-/- colony (Milan, Italy).Animals were kept under standard conditions: housed in an airconditioned room (temperature 21 ± 1 °C, relative humidity 60 ± 10%) with a whitered light cycle (lights on from 08.30 to 20.30).Home cages were Plexiglas boxes (42 x 27 x 14 cm) with sawdust as bedding.Pellet food (Enriched Standard Diet purchased from Mucedola, Settimo Milanese, I-20019, Italy) and tap-water were continuously available.
All experimental procedures were carried out in accordance with the EC guidelines Decreased fertility and earlier onset of puberty in p66 Shc-/-mice 57 Chapter 2 (EC Council Directive 86/609 1987) and with the Italian legislation on animal experimentation (Decreto L.vo 116/92).

Reproductive behavior
Reproductive success was evaluated by taking into account, over 8 breeding cycles of verging females, the number of pregnant subjects, number of delivered pups, number of litters with dead pups and number of litters in which cannibalistic episodes took place.Cannibalism was defined as a defect in the normal care of the newborn pups, reflected in visible physical damage to the pups caused by repeated bites and leading to death.A pup was considered cannibalized on day N if it was alive on day N-1 and not present on day N (Day and Galef, 1977).Body weight of 5 WT and 10 KO virgin females was also taken into account as a possible factor affecting future reproductive success.In addition mean number of pups for each litter was also evaluated.
Two breeding cycles that took place during unexpected environmental stressful conditions (breeding pairs were moved from the Charles River animal facility to that of Istituto Superiore di Sanità -ISS -; breeding pairs were subjected to the inversion of the dark-light cycle with a shift from dark to light of 2 hrs per day) were considered for statistic analysis to study the effect of stress on reproductive success.In addition, changes in fertility over subsequent breeding cycles in WT and KO mice were taken into account by considering the number of full term pregnancies over 9 breeding cycles performed on 40 WT and 40 KO female mice from the colony at EIO, Milan, Italy.

Maternal Behavior
Four-month-old virgin females, (18 WT and 18 KO) were housed 2 per cage and a male was introduced, of the same age and genotype, and left there for mating for 10 consecutive days.At the end of the mating time the male was removed and females individually housed until the day of delivery.Before mating all the females were weighed in order to correlate body weight with reproductive success (carrying out a full-term pregnancy).Of all subjects bred, 10 WT and 10 KO dams gave birth.Out of these, 1 WT and 4 KO subjects were eliminated from the behavioral observations because all pups either died or were cannibalized.
At birth litters were not culled due to the overall meager number of pups in the litters and not to interfere with the onset of maternal behavior.Maternal behavior was observed each day from PND 1 to PND 8, considering the day of delivery as day 0 and a number of behavioral items were scored involving both maternal and non-maternal behaviors as previously described (see Capone et al., 2005).Observations of maternal behavior were performed by an observer blind to the genotype at 4 different times during each day (08.00-09.00a.m., 11.00-12.00a.m., 02.00-03.00p.m. and 05.00-06.00p.m.).Most maternal behavior in rodents is expressed during the light phase of the day, thus our observations were aimed at maximizing the likelihood of observing maternal behavior at its peak (around noon) and its decline over the day (Capone et al., 2005).Each litter was observed by instantaneous sampling 20 times/h with an inter-observational interval of 3 minutes.Maternal behaviors were scored as "1" or "0" depending on whether each behavioral component occurred or not.

Response to a metabolic stressful challenge in the offspring
In order to assess the response to a metabolic stressful challenge, leptin levels were measured in 8-day-old pups following 8 hrs of maternal separation.For this experiment 7 KO and 6 WT litters were purpose-bred and culled at PND 1 to 4 pups, 2 male and 2 females (whenever possible) in order to avoid any bias due to litter size on metabolism and body weight (Agnish and Keller, 1997).At PND 8 pups were weighed and, 1 male and 1 female pup were removed from their home cage and put in a novel cage, with clean sawdust as bedding, on a heated pad for 8 hrs while the remaining 2 pups were immediately sacrificed (controls) and trunk blood collected.Blood samples were collected also from the dam by tail nick procedure.At the end of the 8 hrs of separation pups were sacrificed and trunk blood was collected.Blood samples were collected individually in potassium-EDTA coated 10 ml tubes (1.6 mg EDTA/ml blood; Sarstedt, Germany).All samples were kept on ice and later centrifuged at 3000 rpm for 15 min at 4°C.Blood plasma was transferred to Eppendorf tubes and stored at −20°C until further analysis.Leptin was measured using a commercially available ELISA kit (mouse ELISA kit Crystal Chem.Inc. Downers Grove, IL 60515, USA).

Onset of puberty in the offspring
The onset of puberty was assessed in 19 WT and 32 KO male and 10 WT and 19 KO female pups.Vaginal opening (VO) and balano-preputial separation (BPS) were chosen as pubertal markers (Korenbrot et al., 1977;Rodriguez et al., 1997); beginning on the day of weaning (PND 21), mice were examined daily between 08.00 and 1100 a.m. and the dates of VO for females and BPS for males were recorded.In female mice, opening of the vaginal cavity to the skin is an event occurring around the fifth week of life that can be induced in sexually immature mice by beta-estradiol injections (Rodriguez et al., 1997).The separation of the prepuce from the gland penis (balanus) has been shown to be androgen dependent and to occur around the time of puberty (Korenbrot et al., 1977).

Statistical analysis
Data were analyzed using parametric analysis of variance (ANOVA) with genotype as between-subject factor (maternal behavior, leptin assessment, onset of puberty, mean number of delivered pups per litter) and time blocks as within-subject repeated measures (maternal behaviors).As for leptin measurement, values of nonseparated or 8 hrs-separated pair of littermates were averaged before being used for ANOVA analysis, which included a 8 hrs-separation challenge as within-litter factor.Post hoc comparisons were performed using the Tukey's test.

Chapter 2
Due to unbalancing of sex ratio in litters, data on the onset of puberty were also analyzed by means of a multiple linear regression with the litter as a clustering factor to take into account litter effects.
A linear regression model was used to assess the effect of drinking and eating behaviors (in dams) on pups' care as well as the effect of dams' leptin levels on pups' leptin levels.Fisher's exact probability test was used to compare genotypes for reproductive success of the colony bred at ISS (i.e.number of pregnant females, number of litters with dead pups and number of litters in which cannibalistic episodes took place: two-by-two contingency table) as well as those bred at EIO (Milan, Italy) (i.e. decline of reproductive success over 9 repeated breeding cycles occurring on stable pairings: two-by-two contingency table).As for data from the EIO colony, Bonferroni's correction was applied to take into account the multiple comparisons performed.Statistical analysis was performed using Statview II (Abacus Concepts, CA, USA).

Reproductive success
No difference was found between WT and KO when considering the number of females carrying out full terms pregnancies during 8 breeding cycles, neither for the number of litters found with dead pups (two tailed Fisher's exact probability test: p=0.4863 and p=0.4109, respectively for delivering females and for number of litters found with dead pups) or for the mean number of pups per litter (F(1,123)=0.714,p=0.3997).By contrast cannibalistic behavior episodes occurred more frequently in the KO litters (Fisher's exact probability test: p=0.0014).When reproductive success was assayed on two breeding cycles occurred under stressful conditions, the number of KO female subjects giving birth was significantly lower than in the control group (Fisher's exact probability test p=0.0246see Table 1).As for body weight, KO females were found to be overall leaner than WTs (genotype main effect: F(1,32)=5.770;p=0.0223), nevertheless this feature did not affect their ability to become pregnant (main effect of pregnancy condition F(1,32)=0.049,p = 0.0826; interaction between genotype and repeated measures F(1,32)=0.300;p=0.5874, see Figure 1B).
As for the colony bred at the animal facility of EIO (Milan, Italy), data showed that KO breeding pairs underwent a faster decline in reproductive success over repeated breeding cycles.In particular the decline was apparent from the 4 th to the 7 th breeding cycle (respectively: p=0.0339; p<0.0001; p=0.0064; p=0.0476, see Figure 1A) resulting statistically significant only at the 5 th and 6 th (Bonferroni's corrections).

Chapter 2
Table 1 KO subjects appear less successful under stressful breeding conditions; under standard conditions mutant subjects exhibited a higher amount of cannibalistic behavior, however the mean number of pups did not differ between the two genotypes.

Maternal Behavior
A significant interaction was found for the nest building behavior between genotype and day showing that at PND 7 KO mice performed a significantly lower amount of this behavior (F(7,98)=0.245,p=0.0434, post hoc comparisons, p<0.05, data not shown).Given the overall low occurrence of some of the maternal behaviors scored, the frequencies of nursing and licking items were pooled together and defined as "pup care" (= sum of the overall nursing and licking).A significant interaction between day and genotype (F(7,98)=2.156,p=0.0446) showed that KO dams appeared to be overall less prone to parental care especially on PNDs 5 and 8 (post hoc comparisons p<0.05, see Figure 2A).Analysis of "pup care" as a whole revealed a main effect of the day (F(7,98)=7.082,p<0.0001) and of the time of observation (F(3,42)=6.699,p=0.0008).In particular, maternal care tended to decrease from PND 1 to 8 regardless of genotype and was overall characterized by a trough early in the morning (08.00a.m.) followed by a significant increase (peak) around noon (11.00 a.m.) (post hoc comparisons for 08.00 vs. 11.00a.m.p<0.01; 08.00 a.m. vs. 02.00 and 05.00 p.m. p<0.05).
As for non-maternal behaviors, interestingly the KO subjects showed a higher frequency of foraging behavior (eating and drinking) compared to controls, especially on PND 5 and PND 8 (main effect of genotype: F(1,14)=6.636,p=0.0220, and interaction between day and genotype: F(7,98)=3.938,p=0.0008, post hoc comparisons for WT vs. KO both on days 5 and 8, p<0.01,see Figure 2B) i.e. exactly in those days during which maternal behavior was found to be reduced.The time course of the foraging behavior showed overall to be complementary to that of pup care.In fact, regardless of genotype, the frequency of these behaviors was characterized by a peak early in the morning (08.00a.m.) followed by a trough around noon (11.00 a.m.) (main effect of time of observation: F(3,42)=6.075,p=0.0016, post hoc comparisons for 08.00 vs. 11.00a.m., p<0.01 and for 08.00 a.m. vs. 02.00p.m., p<0.05) and showed a progressive increase over days (main effect of day: F(7,98)=8.929,p<0.0001).Although not significant, ANOVA analysis revealed a tendency for KO dams to eat and drink more both at 08.00 and 11.00 a.m.(interaction between genotype and time of observation: F(3,42)=2.306,p=0.0905; post hoc comparisons for KO vs. WT p<0.01 and p<0.05 respectively for 08.00 and 11.00 a.m.).In addition, mutant subjects moved and explored and perform more rearing episodes than controls (main effect of genotype: F(1,14)=9.999,p=0.0069;F(1,14)=8.299,p=0.0121 respectively for moving-exploring and rearing behaviors).Other non-maternal behaviors taken into account did not differ between the two genotypes (genotype main effect, digging F(1,14)=0.728,p=0.4078; self-grooming F(1,14)=2.808,p=0.1160; out of nest F(1,14)=2.697,p=0.1228).
To assess whether variation in maternal care were in some way dependent upon foraging (eating-drinking) behavior, these behavioral items were analyzed using a linear regression.Results showed a significant inverse relationship both for WT and KO subjects (F(1,8)=29.157,p<0.0010, r = -0.898;F(1,6)=183.959, p<0.0001, r= -0.987, respectively for WT and KO mice see Figure 2 C and D).

Response to a metabolic stressful challenge in the offspring
Leptin levels measured in dams and pups under basal conditions did not differ as a function of genotype (F(1,11)=0.827,p=0.3827 and F(1,11)=0.266,p=0.6161 respectively for pups and dams, see Figure 3A).To assess whether leptin levels in the pups were dependent upon those of dams, a linear regression model was used.Interestingly while a significant direct correlation was found for WT subjects (F(1,5)=7.428,p=0.0527,R 2 =0.650) no apparent correlation was found for KOs (F(1,6)= 2.608, p=0.1672,R 2 =0.343, Figure 3B).Following a 8 hrs of maternal separation leptin levels of the pups were overall equally reduced in WT and KO (F(1,11) = 0.001, p = 0.9709, Figure 3C).In addition, as a result of this metabolic challenge, leptin levels of both WT and KO pups appeared not to be affected by those of dams (regression ANOVA: F(1,5)=1.217, p=0.3318, r= +0.483 and F(1,6)=3.724, p=0.1115, r= -0.653 respectively for WT and KO subjects, Figure 3D).WT and KO pups did not differ in their body weight (F(1,47)=0.29:p=0.865, data not shown).
Figure 3 WT and KO dams and pups did not differ for their leptin levels under basal conditions (A).Leptin levels in the pups were found to be dependent upon those of dams only in WT but not in KO pups (B).A maternal separation of 8 hrs reduced leptin levels in pups of both genotypes (C), as a result of the separation leptin levels of both WT and KO pups appeared not to be affected by those of dams (D).Data shown in panel A are mean ± SEM (n = 6 for WT and 7 for KO); in panels B and D the correlation between leptin levels in mothers and dams respectively under basal conditions and following a 8 hrs separation (Pearson's correlation coefficient r=0.806; 0.586; 0.483; 0.653 respectively for WT; KO under basal conditions and for WT and KO following 8 hrs of separation); in panel C the percent change in leptin levels from basal values.

Discussion
Our data show that deletion of the lifespan determinant p66 Shc in adult mice results in a reduced fitness, over repeated breeding cycles and under stressful conditions, in addition to an earlier onset of puberty.Maternal care during the early postnatal phases is constrained and redirected by changes in feeding behavior resulting from lack of p66 Shc .
Considering the overall reproductive performance, over repeated breeding cycles and under challenging environmental conditions, KO subjects appeared to be less successful than WTs.In fact KO breeding couples from EIO showed a faster decline in reproductive success over time, in addition mutant mice that were moved during pregnancy (a stressful condition) totally failed to carry out full-term pregnancies.Interestingly, KO mice showed a higher frequency of cannibalistic behavior which did not affect their overall fitness since the two genotypes did not differ for the mean number of pups nurtured by each dam.Moreover, under standard breeding conditions the number of nests found with dead pups and the number of females able to carry out full-term pregnancies did not differ between WTs and KOs.
Negative energy balance, low body weight, reduced food intake and low fat mass are all factors that have been negatively associated to reproductive success (Kennedy and Mitra, 1963).KO virgin females are leaner than their WT counterpart (lower body weight) although this feature did not affect their ability to carry out full-term pregnancies under standard conditions.Under stress, it is possible that glucocorticoids, which have a powerful catabolic effect, might have exerted a negative influence on the onset and progression of pregnancy, especially in lean subjects.Likewise, since pregnancy and lactation are conditions involving a huge energetic demand (Thompson, 1992), it is conceivable to hypothesize that the higher frequency of cannibalistic behavior observed in KO dams might be related to their lower body weight.In several rodents, cannibalism is an organized part of the normal maternal behavior which allows an individual female to adjust for her litter size in agreement with her capacity to rear young in the environmental conditions (food shortage) prevailing at the time of her parturition (Day and Galef, 1977;Krackow, 1989).Recent findings by Berniakovich and co-workers showed that p66 Shc mice are characterized by reduced fat development and increased basal metabolism in addition to increased insulin sensitivity of peripheral tissues (Berniakovich et al., 2008).Thus, although cannibalism has been often reported to occur in transgenic animals, in this case, because of the peculiar metabolic demand of the p66 Shc mutants, this behavior could well reflect different metabolic needs of this genotype rather than a non-specific effect due to transgenosis.To strengthen this hypothesis, we observed a specific increase in the general locomotor and exploratory activity (moving/exploring and rearing behaviors) in mutant dams.This piece of data is in line with previous results showing a higher and more prompt exploratory activity (in tests of spontaneous behavior) (Berry et al., 2007;Berry et al., 2008), in addition, and more interestingly, from an evolutionary perspective metabolic needs (hunger) may increase locomotor activity resulting in an adaptive foraging behavior (Cabanac, 1985;Overton and Williams, 2004).
To further investigate the role of metabolic signals in shaping the behavioral phenotype of the p66 Shc-/-mutants, we looked at leptin levels and found no difference between WT and KO dams and pups both under basal conditions and following 8 hrs of maternal separation.Worth noticing in KO pups leptin levels were not related to those from their dams while WT subjects showed a direct relationship under basal conditions that was specifically lost as a result of stress (maternal separation).Leptin provided by the dam in the milk during the first 15 days of life is the main source of gastric leptin for the pup rodent (Oliver et al., 2002).The role of maternal milkborne leptin is largely unknown, however it has been suggested to be involved in the control of energy balance during the early post-natal period (Pico et al., 2007).During lactation, KO dams showed a higher frequency of eating and drinking than their WT counterpart.As a consequence, especially in the mutants, parental care appeared to be strictly dependent upon foraging behavior, a strong inverse correlation being present between feeding and drinking vs. active pup care.In this context it can be hypothesized that while in the WTs the source of leptin is primarily the mother in the KOs the mother might not provide a sufficient amount of this hormone triggering an autonomous production by pup tissues to supply to this lack.
Wild rodents having pups are often forced to leave the nest for variable periods Chapter 2 (hours) of time to provide their self food (Calhoun, 1963).In altricial rodents (e.g.mice and rats), maternal care, and more in general maternal environment, is crucial for an adequate development of the pups, representing the most relevant source of stimulation.Alterations in maternal care, during early developmental phases, can lead to complex and long-term influences on behavioral and neuroendocrine responses of the offspring (Cirulli et al., 2003;Levine, 2000;Pryce et al., 2001).This pattern of maternal attendance to the nest has been modeled in the laboratory by early handling, this manipulation resulting in important changes in the functionality of the hypothalamic-pituitary-adrenal (HPA) axis (Levine et al., 1957) in a way such that the ability of the adult organism to respond, cope, and adapt to stressful stimuli is increased (Fernandez-Teruel et al., 2002;Meaney et al., 1991).We have previously shown that p66 Shc-/-mice are characterized by overall lower emotionality and increased basal levels of BDNF (Berry et al., 2007;Berry et al., 2008), a neurotrophin involved in neuronal survival and differentiation, cognition, emotionality and pain sensitivity (Berry et al., 2008;Cirulli et al., 2004;Thoenen, 1995), which plays also a role in body weight control and energy homeostasis (Pelleymounter et al., 1995b;Xu et al., 2003).In addition we have recently found mutant subjects to be also characterized by a more efficient homeostatic control and better abilities to cope with changes in the internal milieu possibly via a fine regulation of the HPA axis (Berry et al., 2009 submitted).Thus, considering our initial hypothesis overall it is conceivable to hypothesize that the foraging pattern of KO dams might lead to behavioral and neuroendocrine features characterizing mutant in a combined nature (the lack of p66 Shc )-nurture (maternal cares) effect, this latter driven in turn by the effects of the lack of the gene on dam's metabolism.
A huge body of evidence suggests that an interplay between metabolic and reproductive signals exists, aimed at maximizing fitness when food sources are available, fat mass accumulated by the organism representing a peripheral "pro-reproductive" signal.In addition, fertility and longevity are two traits that often show an inverse relationship with each other in a wide range of organisms, as predicted by the tradeoff theory of aging (Westendorp and Kirkwood, 1998;Williams, 1957).In this context, since p66 Shc-/-mice are characterized by increased longevity and reduced body weight we indeed expected KO mice to show a delayed onset of puberty.By contrast, when balano-preputial separation and vaginal opening were investigated, respectively in male and female WT and KO subjects, we found mutant mice to reach the onset of puberty significantly earlier than the same age control counterparts.
The environment may affect the development of individual differences through subtle variations in parental care (nutrient supply, behavioral interactions) (Cameron et al., 2005) such that the offspring is shaped as a forecast of the environmental conditions to face soon after weaning.Cameron and colleagues found evidence for a maternal programming of timing of the onset of puberty in the female rat.In particular, the female offspring of Low nurturing (LG-ABN) mothers showed vaginal opening significantly earlier in life than the offspring of High LG-ABN dams.The authors hypothesize this anticipation to be an adaptive mechanism to maximize fitness in ad-verse environmental conditions reflected by the lower amount of parental care (Low LG-ABN) (Cameron et al., 2005).Indeed we observed a lower frequency of pup care in the p66 Shc dams and hypothesize the anticipated puberty of mutants to be an indirect effect of the lack of the p66 Shc due the high foraging behavior shown by these subjects.This behavioral pattern may be assimilated to the experimental handling condition or to a naturalistic condition in which food is not easily available.
Interestingly, although KO subjects were less successful over repeated breading cycles, possibly leading to an overall lower fitness, they appeared to buffer this physiological trait both by anticipating the onset of puberty and possibly by increasing the number of delivered pups.In fact the mean number of pups per nest (under standard breeding conditions) did not differ between WT and KO subjects despite the higher cannibalistic behavior observed in mutant dams.
Taken together these data suggest that the increased longevity of mutant subjects has a cost given by the reduced fertility over subsequent breeding cycles.Beyond this clear result and more interestingly, the lack of the p66 Shc gene results in an overall remodeling of the reproductive pattern in a way which may be reminiscent of the wellknown r (KO) and K pattern (WT): the first, characterizing p66 Shc-/-subjects, maximizing the quantity of offspring through earlier mating, the second which better reflects the WT situation, pointing to a greater parental investment in the offspring (McCarthy, 1965).
Overall our results point to a main direct genetic effect of the lack of the p66 Shc gene on metabolism which in turn, in a fine interplay of nurturing effects, also mediated by maternal behavior, is able to affect reproductive pattern.
Finally it is important to note that all the studies on the p66 Shc mutants were undertaken under standard laboratory settings which are far away to be representative of a wild-life condition in which gene selection may realistically take place.Thus further studies under naturalistic or semi-naturalistic challenging condition may help to elucidate the potential to cope and survive of the p66 Shc-/-phenotype.

Figure 1
Figure 1 Reproductive success is affected by the lack of the p66 Shc gene in reproductive pairs on the 5 th and 6 th subsequent breeding cycles (A); KO virgin females are characterized by a reduced body weight, however this feature did not affect their ability to become pregnant (B).Data shown are expressed in percent in panel A, (n= 40 for both WT and KO breeding pairs) and mean + SEM (n = 17 for WT and KO) in panel B. Post hoc comparisons: * p < 0.05.

Figure 2
Figure 2 KO dams were less prone to parental care, especially on PNDs 5 and 8 (A), on the same days a higher frequency of foraging behavior (eating and drinking) appeared, compared to controls (B).Variations in maternal care were dependent upon foraging (eating-drinking) behavior in WT (Pearson's correlation coefficient r = 0.898) (C) and particularly in KO dams (Pearson's correlation coefficient r = 0.987) (D).Data shown in panel A and B are mean ± SEM (n = 9 for WT and 6 for KO); in panels C and D the correlation between maternal care and foraging behavior.Post hoc comparisons: * p<0.05 and **p<0.01 for WT vs. KO both on days 5 and 8.

Figure 4
Figure 4 KO mice overall reached the onset of puberty earlier than WTs.Post hoc comparisons: *p<0.05 for WT vs. KO.Data shown are mean + SEM (n = 19 WT and 32 KO males and 10 WT and 19 KO females).