Why the Maternal Brain?

Authors


C. H. Kinsley, Department of Psychology, Center for Neuroscience, Richmond Hall, University of Richmond, 28 Westhampton Way, Richmond, VA 23173, USA (e-mail: ckinsley@richmond.edu).

Abstract

In the rat, the change from a virgin/nulliparous female to the maternal animal takes place at many levels. A subtle developmental wave washes over the female nervous system and transforms her from largely self-centred to offspring-directed, from personal care and protection to care of genetically-related offspring, from indifference to ardour. Such change is preceded by substantial and apparently permanent neural alterations, the depth of which results in the maternal brain, and is the basis of the present review. The neuroplasticity of pregnancy, inherent to the female brain and, we believe, representative of the full expression of the female nervous system’s capacity, is a result of significant hormonal and other neurochemical actions. It results in the striking brain changes that are associated with, and necessary for, successful reproduction. We discuss some of these changes and their ramifications. Collectively, they represent the culmination of mammalian evolution and have led to the development of the social brain characteristic of higher orders of mammal, including the human. We also examine different facets of the maternal brain, beginning with a review of the genes involved in maternal behaviour, and in the subsequent ‘expression’ of the maternal brain. We next discuss olfaction and the manner in which this major sense draws from the rich sensory milieu of the mother to regulate and support maternal behaviour. Last, we discuss the ‘whys’ of maternal behaviour, a theoretical foray into the reasons for such substantial maternal brain alterations. We focus on the male’s potential role as the raison d’etre for the manifest alterations in his mate’s brain. In the end, it is clear that the female brain undergoes a significant reorganisation en route to motherhood, the results of which are deep and enduring.

Gene signalling and maternal behaviour

The ways in which the female becomes a mother represents a singular experience for the animal. The transition through which the female mammal passes changes her on a number of levels. The fundamental alterations that signal the transformation to the maternal brain begin with pregnancy hormone–gene interactions. Those genes expressed during reproduction, which, interestingly, are similar among worms, wasps and mammals (and include multiple links among insulin-signalling systems (1), are linked by two factors that arguably unite all life on this planet: resource acquisition (feeding and supporting behaviours) and reproduction. There are complementary behavioural enhancements in stress/anxiety systems, upticks in learning and memory regulation and aggression-promotion, and basic neuronal and glial modifications that appear to augment behaviour and to protect the female’s brain from damage and trauma, perhaps for life. This array of neuroplasticity, naturally-wrought, is of major importance, and it appears to be both deep and enduring (2,3). That the maternal brain may change out of necessity, a set of adaptations dictated, in part, by the historical unreliability of males (given that they are more likely to leave the female, become injured, thereby compromising their usefulness – or die), is speculative. To be sure, not all males abandon their mates, although apparently a sufficient number have done so to contribute to the neural independence of the parous female. We expand on this idea later in this review.

Work by Brown et al. (4) established a regulatory role for the fosB gene in the display of maternal behaviour in the mouse. In their study, a knockout mutation of the fosB gene expression in the preoptic area (POA) led to a significant reduction in what the authors call ‘nurturing’. The lack of maternal responsiveness in the knockout mothers had a marked effect on offspring survival, despite there being no apparent defects in basic hypothalamic activity, pregnancy, cognition or olfaction. Other aspects of the mothers’ behaviour toward pups (e.g. exploration and investigation of the pups) was similarly normal. Therefore, it appears that some more intimate aspect of the maternal–offspring interaction was wanting. The defect resembled an apparent inability to completely attend to the pups’ sensory communicative nature, followed by a lack of appropriate responses. Similar to a language being spoken, which is foreign to the listener, the pup-related cues appear to fall on the knockouts’‘deaf ears’. Thus, those pup cues normally stimulatory of maternal behaviour fail to elicit adequate responses in the fosB knockouts. The possibility exists that deficiencies in fosB may lead to fewer receptors for oxytocin and, hence, a reduced sensitivity to this important maternal neuropeptide. Irrespective of the exact chain of events leading from a knockout of fosB to a diminution in maternal responsiveness, it is apparent that fosB acts in the POA to mediate the constellation of behaviours collectively known as nurturing. We discuss the potential role of the fosB gene in ineffective maternal behaviour (5, 6).

Further regulating oxytocin binding levels in the medial (m)POA (and thus maternal behaviour) is the expression of the Peg3 gene (7). Peg3 is an imprinted gene that regulates male sexual behaviour and olfaction and regulates offspring development and maternal care in females (8). On the day of birth, maternal rats with a knockout mutation of the Peg3 gene were shown to display longer latencies to build nests and an increased latency to retrieve pups (7). Observations during the first six postnatal days revealed that the mutation was associated with decreased levels of maternal licking/grooming, as well as decreased levels of nursing (7). Finally, as noted above, investigation of oxytocin receptor binding indicated significantly reduced levels in the mPOA in Peg3 mutants (7). The above-mentioned studies provide evidence for particular genes being involved in the expression critical aspects of maternal behaviour. More recent work (9), investigating the mPOA and ventromedial hypothalamus, demonstrates a suite of genes associated with the transition from nulliparity to maternity. In particular, dopamine-related and GABA-related genes are up-regulated in mothers compared to non-mother rats.

Other neurotransmitters have been examined and found to be involved in maternal responsiveness in varying ways, including serotonergic gene expression. Lerch-Haner et al. (10) examined the role of serotonergic signalling in maternal behaviour in mouse dams containing a specific disruption in the development and maturation of serotonin neurones. The Pet-1 ETS (E-twenty six) brain transcription factor appears to be restricted to expression of midbrain serotonin neurones. Thus, using Pet-1 deficient dams, Lerch-Haner et al. (10) investigated the impact of arrested serotonin (5-HT) neurone development. on offspring viability. Virgin wild-type and Pet-1−/− females were bred with wild-type or Pet-1+/− males. Although birth rates and offspring body weights were normal for primiparous Pet-1−/− dams, striking deficits in maternal care were observed. Whereas almost all wild-type pups survived when born to and nurtured by wild-type dams, not one pup survived when born to the Pet-1−/− dams, and pups cross-fostered to wild-type dams also survived. Overall, focusing on detailed pup-directed behaviours, motivation toward pup odours, nest building and related maternal-related behaviours, Lerch-Haner et al. (10) report that intact serotonergic function is required for the nurturing and survival of offspring. Maternal behaviour and survival of the offspring depended largely on the mother’s expression level of the upstream serotonergic transcriptional cascade. Indeed, the intrinsic transcriptional programming of maternal serotonergic activity appears necessary to the quality of nurturing.

In a related study, Girirajan and Elsea (11) investigated the Rai1-transgenic mouse, which demonstrates a significant decrease in its reproductive fitness, with respect to both quality and quantity of offspring. It is possible that the effects on the offspring are a result of poor maternal behaviour. The Rai1 mice were observed to have poorer maternal behaviour, which included protracted pup retrieval times. The deficits were tied to impaired serotonin metabolism. Thus, serotonin and its gene expression pathways appear to be important for the proper display of maternal behaviour.

Hippocampal neuronal gene expression supporting maternal behaviour

There are many different ways in which the female brain transforms itself into the maternal brain. We and others have observed inherent plasticity, the summation of which contributes to the successful generation and protection of the parents’ genes (mainly the mother’s but, by default, also the father’s genes). As we observed above, this expression of plasticity requires genomic involvement because this activity is the foundation for the neuronal alterations that occur and the downstream behavioural effects. Because the changes occurring in the female brain are general and robust, we have begun to examine the genetic expression of proteins in maternal behaviourally-relevant, but supportive, regions of the female brain. That is, there are behaviours directed toward the offspring (retrieving, grouping, crouching, licking, etc.) and also those that enable the female to engage in maternal behaviour by providing metabolic resources (foraging, predation, etc.). Therefore, we have examined regions such as the hippocampus and forebrain, which support the display of maternal behaviour. The hippocampus displays significant natural plasticity, both structurally, in terms of significant dendritic spine changes owing to the reproductive state (12–18) and neurogenetic restructuring (19, 20); Pawluski & Galea, 2007; and Barha et al., 2007; (21, 22, 23), and in functions and behaviours that are tied to its behavioural regulation (i.e. spatial ability) (16, 24).

Given the requisite role of the hippocampus in the maternal behaviour of a parous female, regulating the acquisition of resources and general spatial reckoning, we aimed to examine gene expression patterns associated with reproduction in this region. We performed a DNA microarray study to characterise the genes that may be comparatively expressed or relatively inhibited in the brains of female rats with different reproductive experience (25). Such techniques provided a glimpse into the fundamental brain activity of mothers versus non-mothers. In this case, we used age-matched nulliparous (NULL) and day 5-lactating female rats, whose CA1-hippocampi regions were punched-out to yield 15–20-mg bilateral tissue samples. The RNA was subsequently extracted, and the tissue was analysed in accordance with standard microarray techniques (25) (Tables 1 and 2).

Table 1.   Relative Differences (and Fold Difference) in Gene Expression Profiles from the CA1 Hippocampus of Nulliparous and Lactating Female Sprague-Dawley Rats.
Lactating > Nulliparous
  1. The differences denote gene expression levels that are greater in lactating females compared to nulliparous females. Numbers in parentheses refer to the fold difference between the two groups. Here, we focus only on those differences that are at least two-fold greater or higher, as is convention (23).

Gene
 Insulin-like growth factor (10.6)
 Sensory neurone synuclein (8.6)
 Synaptojanin (6.8)
 Proenkephalin (4.7)
 Calmodulin-dependent protein kinase (4.6)
 Insulin-like growth factor binding protein (4.5)
 Insulin growth factor-binding protein (4.0)
 Synaptosome-associated protein (3.2)
 Huntington’s disease mRNA (2.8)
 Potassium channel mRNA (2.7)
 Glutamate receptor (2.5)
 Interleukin-1b-converting enzyme-related protease (2.2)
Table 2.   Relative Differences (and Fold Difference) in Gene Expression Profiles from the CA1 Hippocampus of Nulliparous and Lactating Female Sprague-Dawley Rats.
Nulliparous > Lactating
  1. These differences denote gene expression that is greater in nulliparous females compared to lactating females. Numbers in parentheses refer to the fold difference between the two groups. Here, we focus only on those differences that are at least two-fold greater or higher, as is convention (23).

Gene
 5-hydroxytryptamine receptor (11.2)
 Olfactory inositol 1,4,5-triphosphate receptor (3.4)
 Na–Ca exchanger isoform NACA-1 (2.8)
 RET ligand-2 (2.6)
 Neural receptor protein-tyrosine kinase (2.4)
 Glycine receptor a (2.3)
 P2x (ATP) receptor (2.1)

The data obtained indicated that many differences existed in gene expression profiles between NULL and lactating females. We focus here on just those genes which displayed a five-fold difference or greater, in either direction, between the NULL and lactating females. One factor that changed dramatically (10.6-fold) was insulin-like growth factor (IGF). This gene, a member of a family of structurally homologous proteins, regulates the activity of IGF-1 and IGF-2 levels, which are the main proteins in the amniotic fluid and in foetal and maternal circulations, and the levels of these fluctuate in diabetics (i.e. serum levels are decreased in Type II but are elevated in Type I diabetes). Interestingly, Hills et al. (26) showed that the levels are high in the foetus and newborn, and decline through puberty. IGF expression in the hippocampus of a lactating female may also reflect such latter influences and regulatory events, in addition to other influences. For example, in the rat, hippocampal glucose-infusions (and consequent insulin activity) enhance memory for spatial behavioural tasks in rats (27,28). Marks et al., (29) demonstrated how intranasal infusions of insulin were able to enhance spatial memory, object recognition and olfactory memory in mice. Furthermore, Xu et al. (30) reported on sex and oestrous cycle differences in insulin receptor concentration in mouse brain, particularly in the hippocampus. Indeed, Xu et al. (30) suggest that the insulin system in the hippocampus may be worth examining as having a possible regulatory role in neurodegeneration-like disorders such as Alzheimer’s disease, the link to which is strengthened in a recent study by Lazarov et al. (31) suggesting that neurogenesis, which is itself enhanced by insulin (32), may play a role in the above conditions. We have reported that aged parous females’ brains appear to be generally healthier, especially with regard to the hippocampus. Deposits of the deleterious substance, amyloid precursor protein (APP), an Alzheimer’s disease harbinger, are significantly reduced in 24 month-old parous (both primiparous and multiparous) compared to age-matched NULL females (2). We have argued that the ‘enriched environment’ of both pregnancy (i.e. elevated and longer-duration exposure to many steroid and protein hormones) and the postpartum period (i.e. the sensory mélange of sights, smells, sounds, tactile stimulation, gustatory intake and suckling stimulation presented to the female) may increase the cognitive reserve of the mother. Together with the effects of IGF-2 and insulin, in general, such changes may produce a hedge against some ageing-related detriments.

Last, a recent study provides strong evidence for IGF-2 being involved in memory enhancement. Chen et al. (33) demonstrate that, in hippocampal slices, IGF-2 has the ability to enhance long-term potentiation, an effect not dissimilar from the findings of Tomizawa et al. (34), who showed that similar systems underlay the spatial memory improvements that we and others have reported (2,12,13,16, 35–39). Therefore, the role of IGF-2 in the various memory enhancements reported in mothers represents a fertile area for additional research.

Other types of memory also appear to be influenced by parity. For example, recent work from our laboratory (R.A. Franssen, K. Rafferty, S. Byce, E. McDaniel, S. Khan, M. Lucia, B. Marks, A. D.-P.L. Phan, D. Sedio, E. Sinclair, H. Tujuba, E.A. Meyer & C. H. Kinsley, unpublished data) has implicated a novel type of learning in which mothers are clearly superior, namely that of prospective memory (PM). Briefly, males and NULL and parous females were lightly water-deprived and placed into an open field (OF) environment in which they could forage for water. They then were put back into their home cages in which they had either ad lib. access to water (control) or no water (PM group). After a period of time, the animals were put back into the OF and their behaviours toward the water sources were observed. To date, the PM mothers drink the most water of any group, suggesting that they anticipate a future environment in which water will be unavailable, and therefore must exploit the current situation to their and their pups’ benefits. We are currently running a replicate of these data. PM is a sophisticated type of cognition, and was previously reported only in species such as humans and scrub jays (40); thus, its presence and enhancement in maternal rats is intriguing.

The gene for synuclein represents yet another gene whose expression is significantly elevated in the mother’s hippocampus compared to non-mothers (by 8.6-fold). Proteins such as synuclein, which are factors prevalent in neurodegeneration, are implicated in a group of illnesses termed synucleopathies, which include Alzheimer’s and Parkinson’s diseases, and may indicate age-related changes in the regulation of synuclein. Continuing with the age-related changes, significant alterations in synuclein activity (i.e. mRNA expression) have been reported in senescent animals. For example, Malatynska et al. (41) found that synucleins fluctuate in the central nervous system (including in the hippocampus), concluding that they may play important roles in both early developmental periods and the ageing process. Other studies have implicated synucleins in Alzheimer’s disease (42). That synuclein gene expression changes so dramatically in the hippocampus of a lactating female compared to a virgin female is testimony to the immense amount of development and reorganisation occurring in the maternal brain, and reflects the possible connections to (or protections against) neurodegeneration. In a review, also in this current volume, work by Morales and her colleagues provides provocative evidence for enhancements in neuroprotection in mothers compared to non-mothers (XX).

In our own work, as referred to above, we have reported significant decreases in neurodegeneration-related proteins (i.e. APP) that may be related to the modifications in hippocampal gene expression levels that we have also observed (2). Females with multiple reproductive experiences displayed fewer deposits of APP in neurones of the dentate gyrus and CA1, the effects of which were correlated with the animals’ performance in a spatial task at the latter stage of their lives (i.e. more APP = worse maze performance).

Regardless, the above changes suggest alterations in neuroplasticity that occur as a function of reproductive experience. Included too are gene expression patterns for the neuronal product, synaptojanin, the expression of which is also increased in lactating compared to NULL females (a 6.8-fold increase). Synatpojanin, a protein involved in recycling of synaptic vesicles, is a limiting factor in the turnover of neurotransmitter released during action potentials. This protein removes the vesicle coating of encapsulated neurotransmitters in the terminal buds of neuronal axons, as the vesicles await docking and release. The cumulative effect is the efficient managing of vesicular packaging and the maintenance of sufficient neurotransmitter stores (43,44). That its hippocampal gene displays enhanced activity during lactation (relative to virgins) may be related to the role of enhanced excitatory activity in the hippocampus, which is related to long-term potentiation and learning (34), and increased dendritic spine density (14), although it is unknown whether the latter effects persist beyond weaning. In other words, more neuronal activity equals an enhanced turnover of neurotransmitter levels and associated degradation to keep pace with the requisite needs in this active region.

As Tomizawa et al. (34) have reported, the latter activity may also relate to enhanced MAP kinase activity in response to increased stimulation by circulating oxytocin. Enhanced expression in the gene for calmodulin-dependent protein kinase (CaM kinase-GR) is also observed in lactating females compared to NULL females (by 4.6-fold). CaM kinase-GR is a neurone-specific enzyme that mediates Ca2+-signalling within various subcellular compartments, including the calcium-dependent gene transcription and plasticity associated with the effects reported by Tomizawa et al. (34). Ho et al. (45) reported that CaM kinase-Gr deficient mice exhibit deficits in two forms of synaptic plasticity: long-term potentiation and long-term depression in cerebellar Purkinje neurones, both of which would be expected to have effects on spatial learning and memory or their behavioural display.

The gene for the opioid-associated precursor, proenkephalin (PRO), similarly displays enhanced expression (4.7-fold difference compared to virgins). PRO, an endogenous opioid itself, is also the precursor of the opiate neuropeptides, metenkephalin and leuenkephalin. PRO is synthesised in the endoplasmic reticulum, is transported within the neurone to the trans-Golgi network, and is bundled into dense-core granules within the regulated secretory pathway of the cell. There it awaits release and its regulation of nociception, modulating the intensity of an experience and signalling pain. PRO is ubiquitous and is expressed throughout the nervous and endocrine systems in the glia and cells of the immune system, cardiac muscle, and even osteocytes (46). Within the hypothalamus, a wide range of stressful physiological stimuli (e.g. hypertonic saline infusions, naloxone-precipitated morphine withdrawal) induces PRO expression and release. PRO expression in the paraventricular nucleus of the hypothalamus may represent part of a cascade of responses/adaptations to stress that regulates the release of hypothalamic substances such as corticotrophin-releasing hormone (47). Given the ability of PRO to bind to and activate multiple receptors, μ (mu), δ (delta) and κ (kappa) opioid receptors, and the nociceptin receptor, for example, makes it a far-ranging substance. Its modified activity, as indicated by the up-regulation of its gene owing to reproduction, may demonstrate a role in controlling maternal interactions (48,49), as well as the mother’s physiological states of lactation and food intake (50).

Complementing the effects where lactating females had gene expression patterns greater than those of the virgins, there were some interesting effects in the other direction, in which NULL females displayed greater gene expression patterns compared to the parous females. For example, NULL females showed greater expression of the 5-HT receptor (5HTr5b; 11.2-fold difference). 5HTr5b is a G-protein-coupled membrane receptor that facilitates the binding of the neurotransmitter, serotonin. 5HTr5b binds to the extracellular ligand, which initiates a change in cellular activity, thereby mediating the efficiency of the transfer of chemical signals between 5-HT neurones (51). A change here in the modulation of the receptor, as suggested by the gene expression difference, and the specific receptor, in this case for 5-HT, indicates likely effects on mood, aggression and food intake, and these are all factors that are manifested differently between NULL mice. and mothers. The latter are more aggressive, and display hyperphagia, for example, relative to NULLs (50).

The expression differences observed in this DNA microarray analysis are good examples of the intricacies of the maternal brain, and display an interesting set of patterns. The hippocampus (and especially the CA1 region) of the rodent [for a summary of these genes’ baseline expression profiles, see the Allen Brain Atlas (52)] likely plays a supportive role in maternal behaviour. [For a demonstration and discussion of lesion effects on behaviours such as nest building, for which the data suggest a supporting role for hippocampus in maternal behaviour, together with and complementary to, spatial learning and foraging, see Terlecki and Sainsbury (53)]. That the gene expression pattern changes so dramatically between reproductive states suggests a role in the transition to/maintenance of the maternal brain. In other words, there is a large difference between the brains of mothers and non-mothers. Genetic analyses of this type represent a rich insight into the brain’s regulation of development, cognition, ageing and neuroplasticity, all of which are affected by reproductive experience. We should note, too, that discussions of gene expression differences must be interpreted with caution. The DNA microarray, as fascinating as this technique is, nevertheless yields just a series of interesting correlations. The technique is useful as a compass, directing the investigator toward some yet-to-be-defined goal.

Building on the microarray data, however, we have recently identified a complex of hippocampal CA1 genes, related by their regulation of neuroplastic factors, which further demonstrate enhanced activity (54). In this work, we examined NULL mice, pregnant (PREG) and lactating females. At their respective and specific stages of reproduction, we killed the females and isolated the CA1 region from each brain, excised bilateral tissue and determined relative levels of target mRNA expression via quantitative reverse transcriptase-polymerase chain reaction analysis. We examined the following neuroplasticity-related genes: brain derived neurotrophic factor, cAMP-response-element binding protein, neurotrophic tyrosine kinase receptor type 2, spinophilin and syntaxin. The data obtained indicate that, in general, reproductive experience [i.e. females that are PREG or parous (primi/multiparous)] modifies the expression of these CA1 genes. For each gene, there were significant differences or parallel but nonsignificant trends among NULL, PREG and parous females, with the reproductive females expressing more mRNA than NULL rats. We are currently evaluating the differences between the up-/down-regulation of these genes and their downstream products. Furthermore, we are currently evaluating preliminary data obtained from the prefrontal cortex in which gene expression levels for APP were reduced in parous compared to NULL females, which is the only gene for which a decrease relative to NULL females has been observed (55). In total, however, the data suggest that the parity-related changes in activity of these specific genes may at least be partly responsible for the observed augmentation in spatial memory (16) (and possibly other functions) in response to pregnancy and the presence of young. Such alterations identify a robust and far-ranging modification of basic neuronal activity in service to the mother and her offspring.

It should be noted, too, that recent work by Meaney and Champagne and colleagues has begun to paint a picture of a fascinating form of gene-by-environment interactions that affect physiological activity in the offspring (54–56). That the brain of the developing organism is shaped by its experiences is beyond debate, although the manner in which the translation of environment into biology occurs involves the epigenetic regulation of behavioural and neural systems. This form of gene expression represents an interface between the animal’s genetic programming and its environmental/nurturing experiences. For example, DNA methylation is important for gene transcription and silencing. In the case of oestrogen receptor α, which is critical in the regulation of sensitivity to oestrogen through membrane-bound and/or nuclear receptor activities, the degree of sensitivity to this major steroid hormone can be affected by maternal attention to the infant (56). Indeed, the behavioural control of gene expression and down-regulation by subtle maternal behaviours such as licking and infant stimulation appears to play a role in the nongenomic transfer of behavioural phenotypes across generations (57–60). Thus, the notion of nature versus nature requires emendation: it may be that nurture = nature.

This work adds to the mosaic of data indicating a most changeable, adaptable and sensitive brain in the reproductive female. The picture beginning to emerge is that of set of neural systems that reach maximal expression in the maternal brain, beginning with likely pregnancy hormonal stimulation of neuronal DNA, followed by changes and enhancements to neural activity, in turn followed by improved maternal behaviours, which are broadly defined. Ultimately, the final goal of all of this activity is the successful maintenance and nurturance of the female’s offspring, so that they too may maximise their reproductive fitness.

Olfactory regulation of parental responsiveness

It is important to understand that unless (and until) sensory stimuli make their way into the nervous system of the organism, it is as if the external world does not exist. Recent work has shown that, even at the very front end of sensory regulation for the rodent, the olfactory system, the interaction of odour and receptor can render an improper interpretation of those signals (61). In most mammalian species, the olfactory system is the rostral-most portion of the telencephalon, and its brain components are comprised of the main olfactory bulb, the accessory olfactory bulb, the anterior olfactory nucleus, and the piriform cortex (62). These areas play significant roles in the processing of olfactory information and have many reciprocal neural connections, particularly within the limbic system below (62). The relatively large size of the olfactory system in relation to the rest of the brain, together with its deep and intimate connections, attests to its survival value.

For most mammals, olfaction is the primary sense involved in reproduction (63,64). Through mate choice, reproduction, sex behaviour and maternal behaviour, females rely on olfaction to guide them through a variety of tasks. Even the Bruce effect, an unfamiliar male/pregnancy blocking mechanism, is regulated through the olfactory system (65).

So, what role does olfaction play in maternal behaviour? Surely animals that cannot smell can still rear their young successfully? Whereas the olfactory system does not regulate maternal behaviour per se, it appears to be an instigator of change in the brain circuitry associated with maternal behaviour (63). Initially, virgin animals that are exposed to pups show no maternal behaviour, displaying instead an aversion to their odours. After birth, however, these same pup odours elicit maternal behaviour in the mother, comprising the suite of responses crucial to the offspring’s survival (66). This is a plastic system in some species, but not all. As virgin rats, gerbils and hamsters are repeatedly exposed to pup odours, maternal behaviour is elicited (67), although virgin rabbits (and sheep) are unable to adjust their behaviour with prolonged exposure to young (68). Interestingly, if the accessory olfactory bulbs of these animals is removed, maternal behaviour is displayed (68,69). This surgery-induced change suggests that the olfactory system controls maternal behaviour through other brain areas, such as the medial amygdala, hypothalamus and periaqueductal grey (PAG), because these systems receives significant input from both the primary and accessory olfactory systems (Brunton & Russell, 2007 (87); (70).

The olfactory system becomes sharpened in maternal humans as well. Fleming et al. (71) reported that new human mothers are much more likely to rate their infants’ odours as pleasant compared to non-mothers (71). The changes in olfaction are linked to the emotional centres of the brain, thus possibly aiding in creating a ‘better’ mother; or at least one who finds her own baby’s odours alluring (e.g. amygdala) (71,72). Of interest, higher levels of cortisol increased sensitivity to the mother’s own infant, further suggesting that it is not just the offspring themselves prompting the change in behaviour, but also internal chemical factors that shape behaviour (63,71).

So what specific changes might be occurring within the olfactory system to elicit changes in behaviour? The well know maternal hormone, oxytocin, plays a role in the processing of olfactory information (as well as other areas) (73). Infusion of oxytocin into the main olfactory bulb of rats produces a rapid onset of maternal behaviour, which in turn may modulate noradrenaline release, a neurotransmitter involved in olfactory learning (63). Further, additional functional magnetic resonance imaging work in lactating rodents has indicated that oxytocinergic neurotransmission contributes to a reduction of anxiety and fear levels (74). Again, these data point to the significance of the interaction of the olfactory system with other downstream non-olfactory related structures in the expression of maternal behaviour. Interestingly, findings such as these are not limited to the female. One study suggests changes in the male brain and paternal behaviour in response to a primarily ‘female-oriented’ hormone, prolactin (75). They found that newly-born olfactory interneurones in males were preferentially activated by their offspring’s odours and that the interruption of prolactin stopped the production of new neurones, thus inhibiting offspring recognition. Furthermore, the recognition behaviour was restored with the return of neurogenesis (75).

Just as pup odours can change from being inhibitor to facilitator of maternal behaviour, other behavioural circuits suggest additional on/off buttons to appropriate maternal behaviour. For example, how does the female ‘know’ how to toggle between caring for young and predation? As noted above, the PAG acts like a switch for these behaviours (76). More specifically, Sukikara et al. (76) proposed that the PAG weighs the balance between eating and acting maternally: as the PAG is activated, it switches from maternal to hunting behaviour and back again. When the rostral lateral portion of the PAG is lesioned, predatory behaviour is inhibited, whereas maternal behaviour continues (76). Thus, avoiding a conflict, the necessary adaptive behaviour of hunting occurs only when the mother has already taken care of her pups, and vice versa. Hence, we observe the neural equivalent of decision-making in this most basic of behaviours. Furthermore, because it is well established that morphine injections into the mPOA inhibit maternal behaviour (48,77), a similar effect has been observed in the PAG as well (77), thus suggesting that maternal behaviour per se is a result of many different brain areas acting together to create a maternal female.

Related to the above-mentioned ‘switching’ behaviour is another neural conflict occurring within the new mother as she decides when it is appropriate to leave the nest to forage for food, thus putting her offspring at risk in her absence. As noted above, the medial amygdala, which modulates fearfulness, defensiveness and avoidance behaviour (78), exchanges impulses with the olfactory system to regulate maternal behaviour (70). Furthermore, there are extensive reciprocal connections between the amygdala and the PAG (80), thus making the amygdala another key player in the expression of maternal behaviour. Wartella et al. (2003) (88) reported that mother rats (compared to NULL rats) displayed less freezing, more exploratory behaviour, and reduced c-fos expression after exposure to an open field. Furthermore, lesions to the medial amygdala (MeA) have been shown to facilitate maternal behaviour in NULL rats after several days (Numan, Numan, & English, 1993 (70). It also appears that there may be subregion-specific changes within the MeA because the anterodorsal portion MeA exhibits more dendritic spines in the postpartum female compared to virgin counterparts (81).

Finally, let us consider alterations in hormones and associated but basic brain regions that facilitate maternal behaviour. There is copious information available in the literature about hormonal and structural modifications associated with hypothalamus and mPOA, as well as the role that these changes play in maternal behaviour, broadly defined. For example, the hormone oxytocin, which stimulates complementary activities such as uterine contractions and milk-letdown, and the protein hormone prolactin, which stimulates milk production, adapt to the ever-changing demands of pregnancy, exerting changes at many levels. Theodosis et al. (82) showed that neuronal and glial relationships in hypothalamic control centres redraw connections, adding flexibility to a vast set of levels already dramatically modified by pregnancy. Additionally, the pituitary gland exhibits chemo-architectural changes in response to reproductive experience. Increased levels of dopamine D2 receptor mRNA (83), and ERα mRNA (84) have been identified in the anterior pituitary of primaparous rats, compared to NULL female rats. These parity-induced changes produce an efficient maternal brain, thereby optimising maternal behaviour and the successful rearing of the offspring. The bottom line is that the mother’s brain expresses its inherent plasticity in response to the demands of motherhood. We conclude with a discussion about the possible reasons why.

Possible evolutionary pressures on the maternal brain

It is abundantly clear that reproduction wrings many effects from the brain of the mammalian mother. In total, the female changes in a way that is both subtle and striking, and apparently for one simple reason: reducing the cost : benefit ratio required for the successful rearing of young. The many facets affected raise one over-riding and fascinating question: why? We will examine these issues in more detail as we attempt to reconcile the ultimate goals of life and evolution, with the manifest reshaping of the female mammal’s brain in service to her young. Research has begun to illuminate the alterations to the female’s brain that accompany and facilitate the display of maternal behaviour, a modification that rivals the other developmental epochs through which the female passes, including early pre- and postnatal sexual differentiation, where the sex appropriate substrate is laid down, and puberty, where the aforementioned neurobiological landscape is saturated with the warm rain of steroid hormones, activating and guiding adult sex-specific behaviour. In short, we argue that the female brain appears to have evolved to adapt and change in ways that promote procreation.

Consider for a moment the biological requirements of reproduction to the female. (Although we focus on mammals, it is worth noting that any organism is faced with costs associated with its reproduction. Furthermore, evolution may have contributed to solutions or lessening the burden of these costs through a sliding scale of accompanying benefits mainly to the female: the greater the investment on the part of the organism, the greater the associated changes to both brain and behaviour.) For example, when still in the womb, the mammalian female develops the eggs that she will eventually mature and have fertilised (85,86). These eggs remain with the female until and if they are used to reproduce. Once a suitable mate has been identified [females, by the way, exert great choice in this matter, although fallibly (85,86)] and the egg(s) fertilised, the processes of gestation and embryonic/foetal/infant growth occur: internal to the female. These activities draw on significant energy reserves. The genetic and metabolic costs of producing, nurturing and carrying the female gametes (eggs) are huge (and any compromised/damaged genes could mean reproductive or offspring problems later on); compared to the male, however, whose gametes (the spermatozoa) are inexpensive, plentiful and easy to replenish, the female’s burden is large (note the size difference in gametes as well). After maintaining the eggs in their pre- and post-reproductive states, the female bears the cost of internal reproduction and, subsequently, the costly dangers of offspring delivery. After all this, to have a maternal behavioural system that cares only cavalierly for the young would be akin to discarding one’s investment. Thus, maternal care and the brain mechanisms that produce it are a stop-gap measure aiming to prevent the loss of a significant investment; the very raw distillation of neuroeconomics.

A stark example of such neuroeconomic principles can be found in the predation proclivities of the parous female. As discussed above, in previous laboratory investigations, we have identified enhanced cognition and reduced stress in parous rats, which is likely an adaptation in mothers needing to efficiently exploit resources to maintain, protect and provision their immature offspring. In a series of behavioural tests on rats, we examined a natural interface between cognition and resource gathering: predation. Comparisons among age-matched virgin, pregnant and postpartum lactating mothers revealed a significant enhancement of predation in mid-late pregnant females and mothers (at a time when maternity-associated bodily changes would be expected to moderate predation ability) relative to virgins (Kinsley CH, Karp NE, Hester NW, McNamara IM, Orthmeyer AL, McSweeney MC, Bardi M, Karelina K, Christon LM, Sirkin MR, Victoria LW, Skurka DJ, Fyfe CR, Hudepohl MB, Felicio LF, Franssen RA, & Lambert KG, unpublished data.). Furthermore, our experiments indicate that hunger motivation, olfaction, audition and somatosensation play little role in the behaviour, whereas vision appears to be important, and neural alterations therein are a likely candidate for further study. The data show a remarkable compensation for changes to size, unwieldiness and shifted survival priorities in the mothers.

An argument can be made that the changes widely described for the mammalian maternal brain are the result, at least in part, of the historical unreliability of the male partner. Males are more likely to be injured or killed, as they compete or fight or forage, or they may leave their mate for the promise of greener pastures elsewhere. It makes little difference to the female: absent independent and redundant offspring care mechanisms; an absent male could spell doom for the under-maternal female’s genetic legacy, whereas the male can exploit the inherent female investment–protection mechanisms. As we noted above, the female nurtures her eggs from embryonic life onward, including their fertilised version. Once the offspring are born, the investment she has made is substantial. To abandon them or to care for them poorly at this point would be foolish, although increased vulnerability to mood disorders, and often poor maternal behaviour, do plague the otherwise attentive maternal mammal (5); it is not a perfect system. Life requires a unique combination of fortuity, adaptation, determination and design. The evolution of the maternal brain, therefore, may truly represent a hedge against the unpredictability of the male brain. Either way, the construction of the maternal brain represents a significant advance in effective offspring production, protection, care and nurturing.

The maternal brain is an example of change and adaptability, all of it in service to successful reproduction. In the present review, we have explored only a portion of the information regarding the brain’s response to the buffeting of the hormones of pregnancy and its subsequent pressures. In a relative way, the post-reproduction brain is the same as the pre-reproduction brain, although the maternal behavioural repertoire that it now regulates is more efficient and shaped for survival. Why do such effects take place? Simply, to guarantee that the next generation is itself ready to reproduce and perpetuate life. In the meantime, the brain changes in ways that are just beginning to be understood and appreciated for the vast and intimate insinuations into the everyday life of the parent.

Acknowledgements

The authors would like to thank the National Institutes of Health (NIMH: 1-R15-HD37578-01) and the NSF (NSF: BCS-0619544) for their support, as well as the University of Richmond for its generous and long-time support of student and faculty scholarship.

Ancillary