Animal welfare is an increasingly important issue in current agricultural research, due to the growing consumer demand and intensification of livestock production, which simultaneously contributes to concerns about the health and well-being of animals bred for their products (Mellor, 2015; Nawroth et al., 2019, for review). Animal welfare is characterised as the result of physical and psychological well-being (for review, see Coria-Avila et al., 2022; Duncan, 1987) and is expected to occur when animals are free from hunger, thirst, malnutrition, discomfort, stress, fear, and pain or if animals express their natural behavioural repertoire (Coria-Avila et al., 2022, for review). The concept of animal welfare is based on research findings that animals have subjective emotions and can feel pleasure and suffering (Nawroth et al., 2019, for review). Importantly, welfare is not a constant state but rather the result of specific brain dynamics that render motivated behaviours. Impaired welfare can be expected when the allostatic load from repeated, chronic, unpredictable and/or uncontrollable conditions compromises the physiological and psychological basis for health due to an impaired allostatic adjustment (for a review, see Korte et al., 2007; Mcewen, 1998; McEwen and Stellar, 1993). Many physiological and behavioural adjustments labelled as stress responses may rather be seen as adaptive changes, through which the animal responds to predictable/unpredictable events (Koolhaas et al., 2011; Romero et al., 2009, for review), and these adjustments are also dependent on individual coping styles (de Boer et al., 2017, for review). Individual stress-coping abilities and variations in stress responsiveness are common phenomena, seen not only in laboratory animals (Bosch et al., 2006; Castanheira et al., 2017; Neumann et al., 2005; Silva et al., 2010 and see Koolhaas et al., 1999; Veenema and Neumann, 2007, for review), but also in livestock. More specifically, different stress-coping styles in pigs, which can be defined as sets of behavioural and physiological responses to stress, have been characterised as reactive vs. proactive. Reactive pigs show a high post-stress HPA-axis activation and a robust increase in cortisol (CORT) levels, low sympathetic adrenal-medullary system activity, low levels of aggression, low risk-taking, and high behavioural flexibility as measured by the Open-Field/Novel-Object test. In contrast, proactive pigs are characterised by low CORT responses, high sympathetic activity, aggressive-, as well as rigid and high-risk behaviour (Kanitz et al., 2019). Inter-individual variations in responsiveness can be of particular importance in animal welfare since certain personality traits may enhance or inhibit stress-coping abilities (Griffin et al., 2015, for review). Thus, a changing or challenging environment may not affect some animals while negatively impacting others. It has generally been accepted that one of the main characteristics of contrasting coping styles is behavioural flexibility (Coppens et al., 2010, for review), which may be measured by tests of cognitive functions and influenced by chronic stress. Indeed, cognition is inextricably linked with stress coping, emotions and thus mental health, as seen in human studies (Barnett et al., 2012; Karestan et al., 2009 and see Burger, 2010 and Martin et al., 2007 for review). Evidence from human literature to farm animals shows that individuals lacking environmental complexity or enrichment may be more likely to show (1) negative affective states such as aggression, (2) increased susceptibility to negative emotions, (3) reduced ability to experience positive states, (4) lower emotional intelligence, (5) reduced stress-coping abilities and (6) reduced behavioural flexibility (Campbell and Lee, 2021, for review). Despite the knowledge of these consequences, commercial indoor animal husbandry systems often offer minimal environmental stimulation and a lack of opportunity for farm animals to express their inherent behaviours. Barren slatted concrete floors with no rooting material preclude the expression of natural behaviours like those exhibited under normal environmental conditions (Van de Weerd and Day, 2009), such as rooting, wallowing and exploring, particularly in pigs (Studnitz and Pederson, 2007), thus leading to harmful social and aggressive behaviour towards conspecifics (Fraser et al., 1991). Indeed, there is growing interest in implementing EE for farm animals in intensive production systems to improve animal welfare while maintaining high productivity (Beattie et al., 1996; Fraser et al., 1991; Jansen et al., 2009; Zebunke et al., 2013). Although there is no standardised form of EE, it usually comprises a physical, sensory, social, cognitive or nutritional component (see Figure 1) (Newberry and Wood-Gush, 1988; Van de Weerd et al., 2003) to increase the number and range of normal behaviours and the ability to deal with behavioural and/or physiological challenges, while preventing abnormal behaviours. In choosing the right form of EE, it must be borne in mind that EE preferences are not only species-specific, but also sex-specific (Van de Weerd et al., 2003; Wurbel, 2007). Unfortunately, there is a lack of studies evaluating the effect on each sex (Simpson and Kelly, 2011, for review).

There seems to be a notable time-dependent factor, with EE being particularly effective during sensitive developmental windows of brain development, i.e., during the early prenatal and postnatal period as seen in rodents (Brummelte et al., 2007; Cymerblit-Sabba et al., 2013; MacRae et al., 2015). Consequently, EE has been shown to buffer the adverse effects of parental neglect and early-life stress at the time of weaning, whereas no effect was observed when EE was applied in later life (Bredy et al., 2003). Specifically, EE offspring showed better behavioural adaptability in response to novel contexts and elevated levels of GLUR1 and GABA in the PFC, indicative of a better stress regulatory system (Connors et al., 2015). Interestingly, the benefits of EE seem to be mediated by changes in maternal care (Connors et al., 2015), which is known to affect not only the physiology but also behaviour of the offspring (Francis et al., 1999a,b; Francis and Meaney, 1999).

The results from rodent studies could be particularly interesting in modern animal husbandry, where early weaning represents a stressor for the offspring. Putting the results from rodent studies in the context of livestock farming suggests that applying EE is particularly important during the prenatal phase to minimise potential behavioural disturbances in the offspring during their animal husbandry lives.

Although studies do reveal EE’s beneficial effects on outcome such as behaviour, stress response, performance and productivity (Mkwanazi et al., 2019, for review), the measurement of welfare and stress resilience in livestock animals is challenging. There are only a few valid scientific measures for reliably assessing their effects (Rault et al., 2018). Nevertheless, the positive effects of EE have been shown on physiological parameters like CORT levels through the minimisation of stress levels and promotion of inherent pig behaviour in intensive production systems. As a result, pigs reared in barren environments without any substrates show higher CORT concentrations than pigs in an enriched environment (de Jong et al., 2000; Pearce and Paterson, 1993; Peeters et al., 2006; Van de Weerd and Day, 2009; Yonezawa et al., 2012). One important factor to consider when assessing CORT levels in the context of stress is its diurnal regulation. Consequently, assessing the effect of EE on the circadian rhythm of CORT plasma levels may be a more reliable measure of outcomes than single CORT measurements (Geverink et al., 2002; Ruis et al., 1997). Bypassing this problem, measuring hair CORT may be a reasonable alternative, for one thing it is non-invasive and, for another, rather independent of the fluctuations seen in plasma levels (see Heimbürge et al., 2019, for review). Aside from changes in CORT levels, the body temperature, growth performance and meat quality have been demonstrably altered in pigs without EE (Mkwanazi et al., 2019, for review).

EE's positive effects on pig behaviour are well studied and numerous (Mkwanazi et al., 2019, for review). Aggression among pigs is a common and normal behaviour primarily driven by forming a social/dominance hierarchy (Turner and Tilbrook, 2006, for review). However, the high levels of aggression often seen in uncomfortable environments in barren housing systems can further exacerbate existing high levels of aggression, constituting a challenge to welfare in pig livestock farming (Bolhuis et al., 2005; Turner and Tilbrook, 2006). Another sign that the environment of intense farming fails to meet pigs’ behavioural needs is the occurrence of harmful social behaviour (e.g., tail-biting) (Beattie et al., 2000; van de Weerd et al., 2005). Importantly, EE has been shown to successfully reduce both the level of aggressive interaction among pigs and harmful social behaviours by increasing the time spent on the substrates provided (Beattie et al., 2000). Although play behaviour is not typical in adult pigs, research suggests that pigs in EE exhibit more play behaviour, which may increase behavioural flexibility and, thus, welfare in the long term (Bolhuis et al., 2005; Kelly et al., 2000; Newberry and Wood-Gush, 1988; Spinka et al., 2001).

Although revealing ultimate and proximate mechanisms behind these analogous behaviours and investigating causes and consequences of personality-cognitive covariation is a growing field of research in domesticated animals (Goursot et al., 2019; Kanitz et al., 2019 and see Finkemeier et al., 2018; Griffin et al., 2015; Sih, 2012, for review), the underlying neuronal mechanisms that drive these variations are largely unknown and underestimated. Neural plasticity is one parameter that could be a particularly interesting biomarker in the context of stress coping styles, behavioural flexibility, emotional intelligence and thus individual welfare, given the fact that there seems to be a correlation between measures of neural plasticity, in other words, adult neurogenesis, EE/complexity, cognition, stress and vice versa (Overli and Sorensen, 2016, for review). Therefore, examining adult neurogenesis in different relevant brain regions, such as the subgranular zone (SGZ) of the hippocampus, the subventricular zone (SVZ), the olfactory bulb (OB) and the hypothalamus may be a promising avenue for understanding differences in individual stress coping and how EE may lead to improved livestock welfare at an individual level.

The following sections summarise the literature available on neuroplasticity in specific brain regions (i.e., SGZ, SVZ, OB and hypothalamus) of farm animals (i.e., pigs, cattle, goats and sheep) and aim to draw correlations with behavioural measures like cognition and how these measures may be linked to welfare in animal husbandry.

Whereas there are significant anatomical differences between rodent and human brains, the brains of farm animals such as pigs, cattle, goats and sheep share some critical structural and functional characteristics (Abbott, 2014). Specifically, they have a similar pattern of brain development with rapid growth during the perinatal period, a highly evolved gyrencephalic neocortex, a comparable cortical organisation and anatomical structure, as well as a similar grey/white matter ratio and regional distribution of neurotransmitter systems (Conrad et al., 2012; Henry et al., 1996; Ishibashi et al., 2012; Pond et al., 2000; Quinones-Hinojosa et al., 2006; Rakic, 2009; Ryan et al., 2018; Zhang and Sejnowski, 2000 and see Dobbing and Sands, 1979; Quinones-Hinojosa et al., 2007, for review). Thus, although somewhat less documented and understood, it is not surprising that adult neurogenesis as seen in tree shrews, marmosets, macaques and humans (Cameron et al., 1993b; Eriksson et al., 1998; Gould et al., 1997; Gould et al., 1999b; Kempermann et al., 1997a), has also been observed in farm animals (Brus et al., 2010; Corona et al., 2017; Dhari et al., 2021; Guidi et al., 2011; Huguet et al., 2023; Levy et al., 2019; Rodriguez-Perez et al., 2003).

In the adult mammalian brain, two neurogenic regions are generally accepted: the SVZ of the lateral ventricle and the SGZ of the hippocampus (Altman and Das, 1965; Brazel et al., 2003; Palmer et al., 1995; Reynolds and Weiss, 1992; Sanai et al., 2004). Multipotent neural stem/progenitor cells (NSCs), also known as precursor cells, which reside in the anterior portion of the SVZ in the walls of the lateral ventricles, migrate along the rostral migratory stream (RMS) into the OB, where they differentiate into neurons (Okano, 2002, for review). Although reports are conflicting, the production of new neurons also seems to occur in the neocortex, the amygdala and the hypothalamus (Dayer et al., 2005; Fowler et al., 2002). At least the latter seems to play an important functional role in sheep, where it regulates seasonal plasticity in the context of reproductive function (Batailler et al., 2016; Kokoeva et al., 2005; Xu et al., 2005, for review see Lee and Blackshaw, 2012 and Migaud et al., 2010).

Adult hippocampal neurogenesis, which occurs in the neurogenic region of the DG, the SGZ, comprises at least four distinct processes: cell proliferation, differentiation, migration and cell survival (see also Figure 2). The SGZ harbours a resident population of cells believed to be multipotent NSCs (Kempermann et al., 2004). NSCs have some characteristics of astroglial cells and, as such, express the astrocytic marker glial fibrillary acid protein (GFAP)—however, they are not themselves glial cells. They are either in a quiescent stage, in which they are in G₀ of the cell cycle (when they express SOX2, a transcription factor essential for the maintaining self-renewal of stem cells), or self-renewing via mitosis (when they express proliferating cell nuclear antigen (PCNA)), or the proliferative marker Ki67. NSCs that are actively engaged in cycling, give rise to daughter cells (progenitor cells), which act as transiently amplifying cells (Mckay, 1997; Seaberg and van der Kooy, 2002) and express the microtubule-associated protein doublecortin (DCX), a marker for commitment to a neuronal phenotype (Brown et al., 2003; Seri et al., 2004; Seri et al., 2001). These progenitor cells re-enter the cell cycle to divide again, producing daughter cells that become post-mitotic and differentiate. In rodents, the majority of these cells (~70%) (Cameron et al., 1993a) differentiate into neurons (Cameron and McKay, 2001), as they express the neural nuclei protein NeuN (Kempermann, 2006). In contrast, only a few (~10%) (Cameron et al., 1993a) become astroglial cells that express GFAP (Kempermann et al., 1997b). After differentiation, cells migrate to the granule cell layer (GCL), where, after approximately 7 weeks following division, the new cells are functionally indistinguishable from older cells (Zhao and Overstreet-Wadiche, 2008) and are electro-physiologically integrated into the circuitry (Hastings and Gould, 1999; Stanfield and Trice, 1988; van Praag et al., 2002). However, differences in lifespan and brain expansion accompanied by gyrification and topographical complexity lead to dramatic differences in the dynamics, cellular composition and morphological organisation of adult neurogenesis in livestock animals (Levy et al., 2017, for review). Thus, generation, migration and differentiation are much longer in primates and farm animals than in rodents (for more details, see Section 3.2). SVZ-derived neural stem and progenitor cells proliferate as multipotent neurospheres, expressing the neural progenitor and neurodevelopmental markers SOX2, GFAP, DCX and Ki67 (Ara and De Montpellier, 2013). Indeed, NSCs from the SVZ have been shown to produce neurons and glia in vivo and in vitro (Guidi et al., 2011; Liard et al., 2009; Yin et al., 2011) and to migrate to the hippocampus, where they become incorporated in the hippocampal circuitry (Nakatomi et al., 2002). The fact that DCX+ cells with morphologies of migrating neuroblasts are additionally found in the striatum, the prefrontal cortex and the anterior cingulate cortex (Aoyagi et al., 2018; Lois et al., 1996; Morton et al., 2017; Sanai et al., 2011 and see Lazarini and Lledo, 2011, for review) suggests that SVZ NSCs are the main source of new neurons in the neocortex, contributing to adult gyrencephalic growth by grey matter maturation, at least in humans and pigs (Ara and De Montpellier, 2013; Morton et al., 2017; Paredes et al., 2016; Sanai et al., 2005). In species with strong olfaction, such as rodents, sheep and pigs, NSCs located on the wall of the lateral ventricles function as primary precursors in the SVZ to continuously generate large numbers of interneurons, which radially migrate through the RMS to the OB, where they mature and mostly replenish granular GABAergic interneurons throughout life (Costine et al., 2015 and Gage and Temple, 2013; Ming and Song, 2011, for review).

As mentioned above, there is also evidence for a neurogenic niche in the hypothalamus. There, resident stem cells with the capacity to self-renew have been observed across numerous mammalian species, including mice (Kokoeva et al., 2007), rats (Pencea et al., 2001a; Xu et al., 2005), voles (Fowler et al., 2005), hamsters (Huang et al., 1998; Mohr and Sisk, 2013) and sheep (Batailler et al., 2014; Migaud et al., 2011; Migaud et al., 2010). However, the level of constitutive neurogenesis is lower than those seen in the SGZ and SVZ (Lee and Blackshaw, 2012, for review). Although the identity and location of hypothalamic NSCs are still unclear or at least debatable, as there has been no proper characterisation to date, specialised hypothalamic glial cells-tanycytes-may be the key cell type in this context (Haan et al., 2013; Robins et al., 2013). Indeed, they show characteristics of neural stem/progenitor cells by expressing the specific markers nestin, vimentin and SOX2 in all mammalian species studies conducted so far (Baroncini et al., 2007; Batailler et al., 2014; Bennett et al., 2009; Bolborea and Dale, 2013; Li et al., 2012; Rodriguez et al., 2005; Wei et al., 2002). Located along the border of the third ventricle, their morphology is similar to radial glia, and they exhibit long processes that project to the hypothalamic parenchyma to reach the fenestrated capillary network. As such, they are at the focus of central physiological regulation and may be particularly important in seasonal species, where the hypothalamus plays a critical role in the availability of thyroid hormones, which are mandatory for an appropriate seasonal alteration between breeding and non-breeding seasons (Nicholls et al., 1988).

To date, adult hippocampal neurogenesis has been shown in the neonatal, new-born and developing piglet brain, in prepubertal and adult pigs, as well as in various breeds of lambs and adult sheep and in different physiological, pathophysiological and behavioural contexts (Brus et al., 2010; Brus et al., 2013; Brus et al., 2014; Conrad et al., 2015; Hawken et al., 2009; Huguet et al., 2023; Swanson et al., 2017; Val-Laillet et al., 2017; Val-Laillet et al., 2018; Val-Laillet et al., 2020) (see also Table 1), whereas there are no reports about neural plasticity in the bovine or caprine hippocampus. So far, research in pigs has mainly investigated the outcome of diet (Huguet et al., 2023; Val-Laillet et al., 2017; Val-Laillet et al., 2020) or pathological interventions and conditions such as traumatic brain injury (Kinder et al., 2019), or infection (Conrad et al., 2015) on the neurogenic niche in the hippocampus. In sheep, most studies have focused on the interplay between the social-hormonal environment and neurogenesis in the DG. Given that stress, EE, cognition and welfare are closely linked, it is striking that no studies have addressed this issue up to now.

Although the DG is widely accepted as the neurogenic niche of the hippocampus in the majority of mammalian species investigated, thus sharing regional similarities across livestock animals (Amrein, 2015; Vadodaria and Gage, 2014), there are profound inter-species differences in the organisation and dynamics of hippocampal neurogenesis, affecting timing of generation, migration and differentiation of new neurons. More specifically, whereas in rodents, mature neurons in the GCL are already seen 7–10 days post BrdU injection, with a peak 2 weeks after labelling and a decline 30 days later (Brown et al., 2003; Imayoshi et al., 2008; Kempermann et al., 2003; Petreanu and Alvarez-Buylla, 2002; Steiner et al., 2004; Suh et al., 2007; Winner et al., 2002), hippocampal neurogenesis in sheep is characterised by a slower neuronal maturation compared to non-human primates. Thus, while few (~13%) granule cells express DCX or NeuN 1–3 months following BrdU injection, the proportion of mature neurons reaches a peak level of about 37–50% 4 months after labelling, with a stable amount of surviving neurons observed up to 6–8 months later (Brus et al., 2013; Kohler et al., 2011; Kornack and Rakic, 2001). The fact that there tends to be an increase in BrdU+ cells 4 months after labelling in the GCL, whereas such an increase is seen 8 months post-BrdU in the SGZ, might suggest a migration process of cells from the GCL to the SGZ over time (Brus et al., 2013). Interestingly, around 69% of BrdU+/SOX2+ cells are located in the external layer adjacent to the SGZ rather than in the GCL 1 month post BrdU, suggesting that the SGZ of the DG mostly harbours progenitor and neural stem cells in sheep (Brus et al., 2013).

However, the above results must be interpreted cautiously, given that those studies were performed on post-parturient female sheep. Thus, hormonal changes and/or fluctuations associated with pregnancy/parturition may have modified the observed neuronal maturation.

Unfortunately, no data is available so far on the dynamics of hippocampal neurogenesis in pigs. Although one study on Göttingen minipigs revealed numerous BrdU+ cells along the rostrocaudal and dorsoventral extent of the DG with the hilus as the major site of new granule neuron production in juvenile (6–7 weeks) and adult (32-week-old) pigs (Guidi et al., 2011), it does not provide any information needed to create a picture about dynamics as reported in sheep. This is mainly because most studies that have been performed on pigs (Huguet et al., 2023; Val-Laillet et al., 2017; Val-Laillet et al., 2018; Val-Laillet et al., 2020) use Ki67 as a proliferation marker (see also Table 1). Although Ki67, PCNA and phosphohistone 3 allow conclusions to be drawn about cells that were actively proliferating, they neither provide any information on the fate of specific cells nor on the rate of survival, which can be analysed when using BrdU injections as a birth-dating method. Comparable studies performed on sheep would therefore be desirable to increase our knowledge of the dynamics of hippocampal neurogenesis in other farm animal species. Despite this absence, a study by Franjic et al. did draw a very detailed picture of anatomically defined cell populations in five hippocampal-entorhinal subregions of the adult mouse, pig, macaque and human brain (Franjic et al., 2022). Using single-nucleus transcriptome profiling via snRNA-sequencing, they not only identified highly diverse cell populations with clear regional distinctions across the DG and CA1-CA4 regions of the hippocampus; they also pointed out fundamental species differences in hippocampal DCX expression per cell type and thus delineated the molecular diversity of cytoarchitectural transition from the allocortex to the neocortex. More specifically, they found differing cell-type proportions of DCX+ cells within species, with pigs showing a dominant (55.6%) DCX+ signal within the neural intermediate progenitor-cell niche, suggesting that this cell population may play a specific role in the pig hippocampal system that will need more detailed investigation in future.

Up to now, there generally seems to be a focus on using the immature DCX neuron marker in pig studies. As a result, and because different dietary components and diet enrichments are known modulators of neuroplasticity and cognition, DCX immunolabelling has been used to assess the effect of a high-fat diet +/− omega-3 and Bifidobacterium breve enrichment in prepubertal female piglets, revealing an increase in immature neurons after 10 weeks of probiotics and omega-3 fatty acid supplementation (Huguet et al., 2023). Interestingly, the effect was limited to the dorsal hippocampus, which is known for its role in cognition and memory, whereas no effect was observed in the ventral hippocampus, which is more involved in regulating emotions (Jinno, 2016, for review). Similarly, studies by Val-Laillet investigated the effects of different dietary compositions and supplements on hippocampal granule cell layer size, proliferation (Ki67), immature neurons (DCX), working and reference memory in Landrace and Yucatan minipigs (Val-Laillet et al., 2017; Val-Laillet et al., 2018). Meanwhile, the maternal western diet (WD) during gestation led to a shrinkage of the GCL in WD piglets. However, despite an increase in proliferation at 3 months of age, DCX+ cells were demonstrably diminished. Moreover, in spite of the negative effect of WD on immature neuron production, WD piglets performed better on a working and reference memory task, which may suggest that improvements in cognitive abilities are not necessarily linked to alterations in neuroplasticity. It may, instead, reflect a higher motivation in WD piglets to be given the food rewards used in the respective tests (Val-Laillet et al., 2017). In contrast, oral administration of butyrate, known for its positive effect on memory, possibly via upregulation of neurotrophic factors in the context of stress-induced anxiety (Moretti et al., 2012; Valvassori et al., 2014), led to a larger hippocampal GCL volume, accompanied by increased proliferation and DCX+ cell numbers (Val-Laillet et al., 2018). However, these results are hardly comparable because of the difference in breeds and assessment time points used in the experimental outline.

Despite the analysis of immature neuron production in most studies that assess hippocampal neurogenesis in pigs, there has been no use of other neurogenesis markers such as PCNA, SOX2, NeuN and GFAP. Viewing hippocampal neurogenesis as a complex mechanism, where proliferation, differentiation, migration and integration are intricately linked and finely balanced processes, illustrates the need for further studies that provide images of hippocampal neurogenesis in the pig brain. Furthermore, another important factor underestimated in studies assessing hippocampal neurogenesis in farm animals is sex. Indeed, sexual dimorphism in neuroplasticity is a well-known phenomenon in rodents and other mammalian species (Bowers et al., 2010; Hillerer et al., 2013; Zhang et al., 2008 and see Hillerer et al., 2019; Yagi and Galea, 2019; Yagi et al., 2020, for review). Although a study by Conrad et al. using BrdU injections to determine the survival of newly generated cells in the DG tends to suggest comparable sex differences in neonatal pigs with increased levels of cell survival in males compared to females (Conrad et al., 2015), a detailed picture, as in rodents, (Hillerer et al., 2013) is largely missing during different pig developmental stages.

Our understanding of hippocampal neurogenesis in pigs is still at an early stage due to a lack of studies. However, our knowledge of neuroplasticity in the sheep brain, particularly regarding social interaction, gestation, parturition and lactation, is relatively broad. This is mainly due to the group centred around Levy and Brus, who significantly contributed to the research in this field. They revealed a clear effect of the peripartum period on hippocampal neuroplasticity in Île-de-France sheep, comparable to studies in rats and mice, and depicting a distinct clear modulating effect of pregnancy and/or lactation on hippocampal cell proliferation, survival and differentiation levels (Darnaudery et al., 2007; Furuta and Bridges, 2005; Hillerer et al., 2014b; Leuner et al., 2007; Shingo et al., 2003). Using BrdU as a marker of cell division and specific markers of neuroblasts (DCX), primary neuronal progenitors (GFAP) and mature neurons (NeuN), they investigated the effect of parturition and interaction with the young on hippocampal neurogenesis in four experimental groups (Brus et al., 2010; Brus et al., 2014). Moreover, as an additional readout parameter for neuroplasticity, they conducted a morphological analysis of dendritic length and number of nodes of neuroblasts (BrdU+/DCX+) in the DG (Brus et al., 2014), as motherhood is seen to be accompanied by changes in the morphology of new neurons in mice (Kopel et al., 2012). As many as 40–50% of BrdU-labelled cells contained GFAP, suggesting the presence of neural stem cells (Brus et al., 2010). As seen earlier in rats, cell proliferation 24 h post BrdU injection (Brus et al., 2010), cell survival 4 months after BrdU injection, as well as the proportion of new neurons (BrdU+/NeuN+) (Brus et al., 2014) was significantly lower in both parturient groups compared to virgins (Brus et al., 2010; Brus et al., 2014) or a group of ewes interacting with either a male or another female (Brus et al., 2010). In contrast, separating the ewe from the young for 48 h did not affect cell survival or neuronal differentiation (Brus et al., 2014), suggesting that the suppression is more associated with parturition than dependent on any interaction with the young. Given that neither parturition nor separation nor their combination affects dendritic plasticity of neuroblasts in the DG, suggests that other neurogenic zones, such as the OB (see Chapter 3.4), might be predominantly involved in olfactory recognition learning of the lamb.

Exercise is another factor well known for its positive effect on hippocampal neurogenesis in rodent studies and it could be particularly interesting in the context of EE and, thus, livestock welfare. However, only one study so far has addressed the hypothesis that an exercise-induced increase in hippocampal neurogenesis also occurs in larger mammalian brains like those of farm animals. More specifically, Swanson et al. performed a four-week running regimen in 6-week-old male Romney sheep injected with BrdU on 5 successive days, beginning 17 days pre-killing to label proliferating cells in the SGZ (Swanson et al., 2017). However, contrary to the positive effects of exercise on cell proliferation found in rodents (Farmer et al., 2004; Kempermann et al., 1997a; van Praag et al., 1999a; van Praag et al., 1999b; van Praag et al., 2005), this study failed to observe any changes in the number or distribution of BrdU+ cells in sheep SGZ. Furthermore, there was no co-labelling of BrdU with mature neuronal or glial markers (PCNA, NeuN, GFAP) used in this study, suggesting that the neurogenic niche of lissencephalic brains may have a greater proliferative potential than that of gyrencephalic brains, at least in the context of enrichment via exercise. Nevertheless, the positive correlation between the distance covered and the level of cell proliferation in the SGZ (van Praag et al., 1999a; van Praag et al., 1999b) and the fact that the distance covered by sheep in this study was comparably shorter to that in rodent studies, may have affected this outcome. Moreover, the results suggest that other EEs such as odour or auditory components may be better used to positively affect hippocampal neurogenesis in sheep. Indeed, effects on stress response (i.e., reduced CORT levels; Li et al., 2020), neuroplasticity, structure and brain neurochemistry, along with neurotrophic factors related to neuroplasticity, seem to be a common phenomenon of musical interventions in humans, pigs, cattle and chickens (Alladi et al., 2002; Cloutier et al., 2020; de Jonge et al., 2008; Kim et al., 2013 and see Alworth and Buerkle, 2013, for review), when carefully selected within a frequency range of 4,000–16,000 Hz. In this context, BDNF seems to be an important outcome measure, particularly given the fact that increased BDNF levels in the dorsal hippocampus are not only indicative of greater resilience to stress (Mosaferi et al., 2015), but moreover, given its known association with increased levels of neuroplasticity and cognitive function (Novkovic et al., 2015 and see Kraemer and Kraemer, 2023; Liu and Nusslock, 2018, for review). Adding important knowledge about the potential of EE for promoting neuroplasticity in farm animals, a recent study in pigs suggests that auditory EE has a positive effect on BDNF serum levels in piglets of sows exposed to music therapy during the last trimester and/or during lactation (Lippi et al., 2022). In particular, sows have been exposed to classical music (a 2 h playlist chosen at random; sound intensity of 60–75 dB; three times throughout the day) either (1) during the last trimester, but not lactation, (2) through lactation, but not pregnancy, (3) through pregnancy and lactation or, (4) not exposed to music at all. Although music therapy did not change BDNF levels of piglets at birth in any group, exposure to music in utero and after birth prevented a decrease in BDNF, as observed in the other groups (Lippi et al., 2022). Although BDNF levels did not affect behaviours associated with stress resilience (i.e., novel arena test, novel-object test, vocalisation, voluntary approach test), it has to be borne in mind that factors like the period of behavioural assessment and the behavioural tests used may play important roles. Moreover, BDNF levels in serum may not necessarily mirror BDNF levels in the hippocampus, which might have affected the study’s outcome.

Thus, there is a need for future studies that (1) assess the effect of EE on other parameters of neuroplasticity, e.g., adult hippocampal neurogenesis, and (2) correlate it with neurotrophin levels (e.g., BDNF) in saliva, serum and locally in the brain, and (3) correlate it with well-validated behavioural outcome measures of anxiety, depression and cognition in pigs.

Aside from the DG of the hippocampus, the SVZ of the lateral ventricle is a major site of neuroplasticity throughout mammalian species (Altman and Das, 1965; Brazel et al., 2003), including farm animals like pigs (Dhari et al., 2021; Guidi et al., 2011; Kinder et al., 2019; Maeda et al., 2020) sheep (Brus et al., 2010; Low et al., 2013; Migaud et al., 2015) and cattle (Perez-Martin et al., 2003; Rodriguez-Perez et al., 2003). The SVZ is typically divided into four rostrocaudal regions, the anterior, central, lateral and dorsolateral regions (Brazel et al., 2003; Quinones-Hinojosa et al., 2006), with the anterior-dorsolateral (rostral) part being the most active neurogenic niche, displaying a high level of proliferating NSPCs, at least in human and porcine SVZ (Morton et al., 2017; Quinones-Hinojosa et al., 2006). Using BrdU labelling in female juvenile and adult Göttingen minipigs, Guidi et al. revealed numerous BrdU+ cells organised in clusters along the lateral wall and roof of the rostral SVZ. In contrast, only a few were visible in the caudal region of the SVZ (Guidi et al., 2011). Immunohistochemistry studies in seven-week-old pigs revealed a laminar organisation of the anterior-lateral SVZ composed of four distinct layers, sharing structural features of the human SVZ, i.e., an ependymal layer (I) acting as boundary with the lateral ventricle, a layer rich in DCX+ neuroblasts (II), a GFAP+ rich astrocyte ribbon (III), the thickness of which may vary, as well as a transitional zone (IV), rich in myelinated fibres (Morton et al., 2017). SVZ-derived neural stem/progenitor cells proliferate as multipotent neurospheres that have been shown to express the SOX2, GFAP, DCX and Ki67 neural progenitor and neurodevelopmental markers, whereas PCNA labelling as observed in humans is hardly visible in sheep and pig SVZ (Curtis et al., 2005; Morton et al., 2017). DCX+ neuroblasts increase over the neurodevelopmental period in porcine SVZ, with an increased number of clusters seen in prenatal day (PD)42 Yorkshire/Landrace piglets compared to PD16 and PD0 in the anterior end of the SVZ, as well as an increase in labelling on PD16 and PD42 in the anterior portion of the SVZ (Porter et al., 2022). Interestingly, DCX+ neuroblast clusters have been shown to be co-labelled with PSA-NCAM at 2 weeks of age, suggesting a migratory process in the piglet brain (Porter et al., 2022). Indeed, DCX+ cells migrate short distances to the GCL, incorporate and functionally integrate into the hippocampal circuitry (Nakatomi et al., 2002), where they play an important role in hippocampal-dependent memory functions, as well as in long-term potentiation (LTP) that underlies those memory functions (Snyder et al., 2001; Winocur et al., 2006). Besides these migratory processes to the GCL, DCX+ cells with morphologies of migrating neuroblasts have also been found in the striatum and frontal cortices, where they primarily differentiate into calretinin (CalR+) interneurons, suggesting that SVZ is the main source of new neurons in the neocortex (Brazel et al., 2003; Morton et al., 2017). However, cortical development is a dynamic and complex process, and very little is known about the contribution of SVZ to cortical growth and associated cognitive functions in gyrencephalic species, including farm animals (Lewitus et al., 2014; Poluch and Juliano, 2015).

In cattle, neurogenesis in SVZ has been shown and characterised in vivo in the bovine lateral ventricle via IHC (Rodriguez-Perez et al., 2003), as well as in vitro in explants from the walls of the lateral ventricle (Perez-Martin et al., 2003). Using specific markers such as ßIII-tubulin, S-100ß, vimentin, GFAP and PCNA, Rodriguez-Perez et al. revealed a similar laminar organisation of the bovine ventricular wall, consisting of specific cell types comparable to those seen in pigs. In particular, four main cell types have been identified within the bovine lateral ventricle: ependymal cells, subependymal neuroblasts, astrocytes and PCNA+ proliferating cells (Rodriguez-Perez et al., 2003). Although the latter are found throughout the lateral ventricle, the rostral horn showed highest densities of PCNA+ nuclei within the subependyma that penetrate to the subjacent glial network and migrate in a rostro-ventral direction, thus mirroring the results in pigs and other mammals. Confirming the results of the in vivo study, an in vitro model using ependymal-subependymal explants from the lateral ventricles of two-year-old calves found a comparable immunoreactivity of GFAP and PCNA in nuclei and basal processes of those cells, further supporting the idea that subependymal cells are the main source of proliferative activity in cattle SVZ (Perez-Martin et al., 2003). Many of the newly generated cells developed neurites and expressed the neuronal marker ßIII-tubulin, which suggests differentiation of cells into neurons (Perez-Martin et al., 2003).

Summarising the results from ependymal and subependymal cells of the SVZ in farm animals, it seems that the organisation of SVZ in distinct layers is similar in pigs, cattle and sheep (Low et al., 2013) and thus comparable to that seen in humans. However, although there is only one study in sheep investigating proliferation dynamics of the SVZ, it is postulated that the rate of maturation of new neurons is much slower in rodents. Generally, it seems that there is an enormous variation in proliferation between species (Curtis et al., 2005; Gould et al., 1999a; Lazic et al., 2006 and Kempermann and Gage, 1998, for review). Whereas PCNA labelling revealed a rate of about 50 cells per mm² in the human SVZ (Curtis et al., 2005; Low et al., 2011; Lucassen et al., 2010, for review), the proliferation rate in mice is around 800 times higher (Gil et al., 2005). Adding information about neurogenesis in the SVZ in rodents and humans, a study in young (3-year-old) and old (5–6-year-old) Romney/Suffolk ewes successfully shed light on proliferation, short (11 days) and long-term (122 days) survival, as well as differentiation at least in one livestock species. Using a combination of BrdU (40 mg/kg; injected over 5 days), PCNA as a proliferation marker, and NeuN as a marker of neuronal fate, Low et al. revealed proliferation counts of approx. 65 BrdU+ cells per mm², thus comparable to proliferation levels in other sheep breeds and experimental setups (Brus et al., 2010). PCNA labelling was much less than in human studies (Curtis et al., 2005) and independent of age (Low et al., 2013). As suggested by previous studies in larger animals, stem cell to neuronal maturation time was significantly slower than in rodents (Eriksson et al., 1998; Kohler et al., 2011; Malberg and Duman, 2003), with only 20% of BrdU+ cells showing a neuronal phenotype (BrdU+/NeuN+) after a survival time of 4 months (Low et al., 2013).

As discussed above for the hippocampus, there is a similar lack of research in pigs assessing the effects of acute/chronic stress and/or EE on neuroplasticity in the SVZ, but moreover, a general lack of studies focussing on porcine SVZ as a neurogenic region in other contexts. Due to its structural similarity to the multi-layered human SVZ (Morton et al., 2017), most studies so far have used the gyrencephalic pig brain (Landrace and Yorkshire pigs) to model the effects of pathological insults such as hypoxic ischaemia (Ara et al., 2010), TBI (Costine et al., 2015; Kinder et al., 2019) or cardiopulmonary pathologies/insults (Dhari et al., 2021; Maeda et al., 2020; Morton et al., 2017) on neuroplasticity and cortical maturation in humans.

Although most of the studies on sheep focus on the OB as a neurogenic region (see below), a few assess the effects of the hormonal environment, for example, seasonal changes and peripartum-associated changes on the ovine SVZ. Thus, administration of BrdU in adult female sheep in July and December, during a long vs. a short photoperiod, revealed no change in SVZ proliferation rate (Migaud et al., 2011), suggesting that changes might be linked and/or limited to structures involved in seasonal neuroendocrine regulation such as the hypothalamus (see also Chapter 3.5). Indeed, endocrine (changes in oestrogen/progesterone levels) and/or social factors (e.g., pregnancy, parturition and interaction with the young) are important regulators of the production of new neurons in the SVZ, not only in rodents but also in sheep (Levy et al., 2017, for review). Injecting 20 mg/kg BrdU 24 h before sacrifice to determine cell proliferation in the ovine SVZ, Brus et al. found parturition as a down-regulatory factor (Brus et al., 2010), as seen in rodents (Darnaudery et al., 2007; Hillerer et al., 2014b; Leuner et al., 2007). Using Ki67 and GFAP as further markers, they also revealed that 40–50% of the 1-day-old BrdU+ cells in the SVZ were GFAP+, suggesting presence of neural stem cells (Brus et al., 2010).

For a summary of the results discussed regarding the SVZ, see also Table 1.

Although the ability of the OB to produce new cells is about 10 times lower than that of the SGZ and the SVZ (Gritti et al., 2002, for review), 1-day-old BrdU+ cells co-labelled with Ki67 and GFAP have been observed in the periventricular layer (Chojnacki et al., 2009), suggesting a neurogenic niche within the OB (Brus et al., 2010; Doetsch et al., 1999). Thus, even if the OB may not necessarily be seen as a neurogenic region on its own, there is no doubt that it plays a vital role in neuroplasticity and its associated behaviours. In the olfactory system, neural stem cells function as primary neurogenic and gliogenic precursors in the SVZ located on the wall of the lateral ventricles. They produce transient amplifying cells, which rapidly create chains of neuroblasts that migrate along the RMS to the OB (Brus et al., 2010; Brus et al., 2013; Doetsch et al., 1999; Sawamoto et al., 2011), where the majority of them mature into granular interneurons. Whereas reports about olfactory neuroplasticity are fairly limited or even absent in pigs and cattle, detailed knowledge exists about the dynamics and maturation of new-born cells in the SVZ-OB network in sheep, at least in the context of parturition and lactation. Using BrdU (20 mg/kg) as a marker of cell division and a combination of neuroplasticity markers (SOX2, DCX, NeuN and S100β), Brus et al. have significantly contributed to our understanding of the maturation timeframe of new-born cells in post-parturient sheep OB (Brus et al., 2013). When analysing different BrdU post-administration survival times over eight months (1 m, 2 m, 3 m, 4 m and 8 m), they found no variation in BrdU intensity, except for a decrease at 8 month post-injection, suggesting a slow process of apoptosis over this period. However, the survival time in sheep seems to be much longer than in rodents, where half of the neurons die within the first month after birth (Lemasson et al., 2005). Similarly, while very few neuroblasts (BrdU+/DCX+) were found 1 month after BrdU injection in the granule layer of the OB, a gradual increase was observed in the periventricular layer 1–3 months later, subsequently reaching a plateau and suggesting a far longer maturation time in sheep than in mice, where interneurons become morphologically mature within 4 weeks of birth (Petreanu and Alvarez-Buylla, 2002). Thus, although there is a lack of knowledge about proliferation and/or survival dynamics in pigs, the results in macaques, which reveal a neuronal maturation time of 97 days (Kohler et al., 2011), suggest a comparable neuronal maturation time across gyrencephalic species. Analysis of the other markers used revealed a variation in specific labelling across survival times, with BrdU+/SOX2+ cells (quiescent cells that may start proliferating, depending on physiological changes) decreasing within 1–4 months and a BrdU+/NeuN+ labelling remaining low 1–2 months after BrdU injection and significantly increasing thereafter up to 8 months. Analysis of the glial marker S100 confirmed a significant increase of double-labelled BrdU+/S100+ cells 2–3 months post-injection, with a decline 4–8 months later (Brus et al., 2013).

Aside from providing insights into dynamical neuroplasticity processes within the SVZ-OB network, the group around Brus also elegantly provided evidence for an intricate migratory pathway along the ventricle from the SVZ to the OB (Brus et al., 2013). By using injections of an adeno-associated virus coupled with an enhanced green fluorescent protein (AAV5-eGFP) that infects neural cells in the SVZ, they revealed fusiform cells with and without processes [neuroblasts (20%) and progenitor cells (30%)] in the periventricular and granular layer of the OB one-month post-injection of the virus. Interestingly, eGFP+ cells were only found in the OB when the injection site was located above the SVZ. In contrast, no labelling was present when located too posteriorly on the level of the corpus callosum, further supporting the idea of a neural stem cell lineage within the SVZ.

Despite the knowledge of the mechanisms of neuroplasticity within the SVZ-OB network, the nature and specific contribution of adult-generated neurons to olfactory behaviour is still being debated (Holmes, 2016; Lazarini and Lledo, 2011, for review). However, given that adult-born neurons have unique properties of synaptic plasticity compared to early-born neurons (Nissant et al., 2009), they could indeed provide an important substrate in olfactory learning, which will be discussed in the following section.

It is now well accepted that the peripartum period and the physiological/hormonal changes associated with pregnancy, parturition and lactation modulate neuroplasticity in all the neurogenic sites and species investigated so far (Galea et al., 2014 and Hillerer et al., 2014a, for review). In sheep, where social behaviours, including maternal behaviour, largely rely on an appropriate sense of smell (Corona et al., 2017; Corona et al., 2018; Gelez and Fabre-Nys, 2006; Keller and Levy, 2012), neuroplasticity may provide an additional mechanism that enhances olfactory learning, as well as the establishment of maternal behaviour and a selective bond with the young following parturition (Poindron et al., 2007, for review). Indeed, a single BrdU treatment (20 mg/kg) 24 h before sacrifice and additional Ki67 labelling revealed a parturition-dependent down-regulation of cell proliferation in the SGZ, SVZ and OB in ewes (Brus et al., 2010), which may be a mandatory mechanism for facilitating the olfactory perceptual memory demands associated with maternal behaviour by favouring the survival and integration of neurons born earlier. Furthermore, co-labelling of the one-day-old BrdU+ cells with GFAP showed that 40–50% of those cells were positive for the respective marker, suggesting the presence of neural stem cells in the SVZ and OB of the ewes (Brus et al., 2010). The density of these BrdU+/GFAP+ cells was around three times higher around the ventricle than the granular and periglomerular cell layer, which further supports the hypothesis of a neurogenic zone in the ventricular margin of the OB.

Interestingly, as previously observed in rodents (Leuner et al., 2007), one major important regulator of peripartum-associated changes in neuroplasticity in sheep seems to be the presence of and particularly interaction with the young. More specifically, Brus et al. used three different groups of sheep in their experimental outline. A virgin control group, a group of ewes allowed to interact with their young 2 days after parturition and a group of ewes that had their lambs removed 2 days before sacrifice (Brus et al., 2014). Immunohistochemical analysis uncovered a suppression of cell proliferation and survival in the OB, which was dependent on lamb presence. Interestingly and contrastingly, in the SGZ, suppression was more likely to be associated with parturition. It is hypothesised that fewer neurons reduce cell competition in the OB, thus enhancing the maturation of new neurons selected to participate in odour learning by young animals. However, olfactory learning seems to involve other mechanisms of neuroplasticity apart from changes in cell proliferation and survival. Thus, neurochemical and electrophysiological changes in mitral cells of the OB supporting the coding of the lamb were shown more than 30 years ago (Kendrick et al., 1992; Keverne et al., 1993; Levy et al., 1993; Levy et al., 1995a). Furthermore, morphological changes at the neuroblast level (DCX+) (for example, dendritic length and number of nodes) are known to be involved in enhancing olfactory neuroblast maturation by lamb interaction (Brus et al., 2014). Indeed, learning has been associated with an accelerated maturation of dendritic trees of new-born neurons in other contexts than lactation (Lemaire et al., 2012; Tronel et al., 2010) and motherhood is a prominent modulator of new-born neuron morphology in rodents (Kopel et al., 2012). Hence, it appears that the ability to process sensory information depends on the functional architecture, synaptic connectivity and electrophysiological changes within the OB. However, adult neurogenesis may be an important regulator in fine-tuning these processes.

Performing two other studies involving Île-de-France ewes, Corona et al. (2017) added further information on the importance of neuroplasticity in the SVZ-OB network in the context of motherhood, olfactory processing, odour learning and maternal selectivity in sheep. Their main questions were: (1) Do adult-generated neurons in the OB contribute to the processing of odours involved in attracting lambs and memorising their individual olfactory signatures? (2) Do olfactory stimuli coming from the neonates specifically activate new interneurons in the OB (Corona et al., 2017)? (3) Does the chemical disruption of olfactory neurogenesis impair the establishment of maternal behaviour in sheep and alter their ability to learn their own lambs’ olfactory signature (Corona et al., 2018) ?. In an attempt to answer the first set of questions, sheep were given four BrdU injections (20 mg/kg) 3 months before parturition. After birth, ewes from Group 1 spent 2 days with their lambs, were then separated for 3 h and re-exposed to either their own, to an unfamiliar lamb or an adult conspecific. In contrast, ewes from Group 2 were not exposed to any lamb for 5 h (control group). Using a combination of BrdU, DCX, NeuN (neuroplasticity markers), as well as c-Fos and Zif268 (neuronal activation markers), they showed preferential activation of three-month old OB neuroblasts (BrdU+/DCX+/Zif268+), but not mature neurons (BrdU+/NeuN+/Zif268+), mirroring the results in rodents where immature olfactory neurons are very susceptible to sensory experience compared to mature neurons (Belnoue et al., 2011; Mouret et al., 2008; Yamaguchi and Mori, 2005). Furthermore, this activation was specific to the OB, as no activation was observed in the SGZ (Corona et al., 2017). Interestingly, neither neuroblasts nor mature neurons of the OB differentiate between exposure to familiar or unfamiliar lambs, suggesting either that older neurons (>3 months) may be involved in recognising their own lambs or that adult olfactory neurogenesis is rather important for the long-term olfactory memory as seen in father mice who exhibited preferential activation of olfactory adult-born neurons when exposed to the odour of 6-week-old offspring (Mak and Weiss, 2010). Nevertheless, the responsiveness of adult-born neurons is lamb-odour specific and not an activation by odours per se, as no activation was observed when exposed to an unfamiliar ewe (Corona et al., 2017). To further prove that olfactory neurogenesis is indeed involved in the establishment of maternal behaviour and olfactory signature learning, they administered the mitotic blocker cytosine arabinoside (Ara-C) into the SVZ. Then they assessed maternal bleats and olfactory neurogenesis in the context of lamb discrimination behaviour (Corona et al., 2018). Adding information from previous studies, where they showed a reduction in maternal behaviour rendering ewes anosmic after a zinc sulphate infusion (Levy et al., 1995b), or removing the amniotic fluid from neonates (Levy and Poindro, 1987), Ara-C treatment led to (1) a 70% reduction in olfactory neurogenesis (DCX+ cells) and (2) fewer maternal bleats indicative of maternal responsiveness (Dwyer et al., 1998) during the first hours after parturition and (3) a lack of familiar/unfamiliar lamb discrimination (Corona et al., 2018).

Given the known interplay between the OXT system and the sensory system (Grinevich and Stoop, 2018, for review), the fact that OXT plays an important role in the orchestration of socio-emotional behaviours and the physiological and neuronal maternal adaptations during the peripartum period, another study in non-pregnant sheep investigated the potential involvement of central OXT in regulating adult neurogenesis in the OB (Levy et al., 2019). Ewes received steroid treatment followed by intracerebroventricular infusions of OXT to mimic endocrine events during parturition. OXT significantly decreased the density of the neuroblasts (DCX+) in the OB. In contrast, there was no effect in any other neurogenic niche, such as the SGZ or ventromedial nucleus of the hypothalamus (Levy et al., 2019), although the latter might be due to a lack of OXT-R usually expressed in neuroblasts or more mature neurons, as seen before in rodents and sheep (Broad et al., 1999; Lin et al., 2017). Aside from the region-specific effect of OXT, the effect was also cell-type specific, as the density of progenitor cells was not influenced by OXT treatment (Levy et al., 2019). Consequently, other neuro-modulatory factors such as glucocorticoids could be involved (Sandi, 1998, for review); it seems that OXT participates in the formation of olfactory memory in sheep after parturition by decreasing the density of adult-generated neuroblasts in the SVZ-OB network.

In summary, neural plasticity within the SVZ-OB network during the peripartum period appears to be an adaptive response to the high olfactory perceptual demand at that point, with adult-born neurons helping to process infantile odours.

For a summary of the results discussed regarding the OB, see also Table 1.

The hypothalamus, known as a key regulator of homeostatic activity and numerous physiological functions such as reproduction, food intake and energy balance ((Schneeberger et al., 2014) for review), has not classically been considered a neurogenic niche. However, more than 20 years ago, proliferative activity (BrdU labelling) within the hypothalamus was reported in rats (Pencea et al., 2001b). To date, constitutive hypothalamic neurogenesis has been proven to occur in many species of mammals including mice (Kokoeva et al., 2007; Lee and Blackshaw, 2012; Li et al., 2012; Werner et al., 2012), rats (Pencea et al., 2001b; Perez-Martin et al., 2010; Xu et al., 2005), voles (Fowler et al., 2005), hamsters (Huang et al., 1998) and sheep (Batailler et al., 2016; Batailler et al., 2014; Migaud et al., 2011; Migaud et al., 2010) (for review see Lee and Blackshaw, 2012 and Migaud et al., 2010). Several reports indicate that hypothalamic stem cells are located in the hypothalamic parenchyma (Li et al., 2012; McNay et al., 2012). However, lineage-tracing studies revealed that a subpopulation of tanycytes lining the ependymal layer of the third ventricle (Xu et al., 2005) and expressing the NSC markers SOX2, GFAP, vimentin and nestin (Batailler et al., 2014; Lee and Blackshaw, 2012; Li et al., 2012) are instead neural stem cells with a proliferative capacity within the hypothalamus. From the subependymal niche, they spread towards the neighbouring nuclei [i.e., arcuate (Arc) and ventromedial nucleus (VMH), median eminence], forming a migratory path of progenitors. Those cells, expressing DCX, show a round morphology close to the third ventricle, whereas they are rather fusiform and bipolar deeper in the parenchyma, further suggesting maturation in, or migration through the parenchyma (Batailler et al., 2014; Lee and Blackshaw, 2012). Most of these DCX+ cells will adopt a neuronal phenotype (BrdU+ NeuN+) 1 month after labelling (Batailler et al., 2014; Haan et al., 2013; Migaud et al., 2015). However, the hypothalamus also shows gliogenic and microglial-like properties, at least in sheep, where approx. 70% of new cells differentiate into S100ß and GFAP+ astrocytes (Steiner et al., 2004) and 10% CD45+ cells, respectively, (Hazlerigg et al., 2013), with no evidence for oligodendrocytic labelling (Steiner et al., 2004).

Naturally enough, the obvious proliferative capacity of the hypothalamus discussed above raises the question of its functional relevance. The discovery that endogenous and exogenous factors, including the photoperiod and season, may modulate hypothalamic cell proliferation and neuron production (Migaud et al., 2011; Migaud et al., 2010; Migaud et al., 2015) soon suggested the involvement of hypothalamic neuroplastic changes in seasonal cycles of physiology and behaviour, which are essential adaptations to cope with environmental variations and to fine-tune the timing of reproductive/metabolic functions. Indeed, photoperiod effects on proliferative capacity and marked structural changes within the hypothalamus have first been observed in songbirds (Alvarez-Buylla et al., 1990; Goldman and Nottebohm, 1983). Using ³H-thymidine to identify dividing cells revealed that new neurons are generated and integrated into functional avian song circuits in a photoperiodic manner (Paton and Nottebohm, 1984). To date, photoperiod-dependent changes in hypothalamic neuroplasticity have also been extensively studied in sheep, a species where sexual activity/inactivity largely relies on photoperiodic length. More specifically, measuring the expression of DCX+ cells at various times of the year, Batailler et al. showed that the density of these cells varied across the photoperiod, with higher counts during the short photoperiod in the Arc and higher counts during the long photoperiod in the VMH (Batailler et al., 2016). Thus, it appears that higher cell proliferation levels are associated with a short stimulatory photoperiod (sexual activity, for example). In contrast, lower proliferation is more likely to be observed during phases of sexual inactivity (i.e., inhibitory long photoperiods). Although peripheral oestradiol levels have been shown to be independent of seasonal changes in cell proliferation in sheep (Migaud et al., 2011), the fact that hypothalamic neurons express ERα (Batailler et al., 2014) and the well-known potential of oestrogens to regulate neuroplastic changes in rats (Cameron and Gould, 1994; Ormerod and Galea, 2001; Tanapat et al., 1999), further suggest an involvement of hypothalamic neuroplasticity in reproductive behaviour and seasonal variation in neuroendocrine secretion. Furthermore, other hormones could be involved in the regulation of hypothalamic neuroplasticity. A study by Levy et al. revealed that intracerebroventricular injections of OXT in non-pregnant female sheep significantly increased the density of neuroblasts in the VMH, with no change in any other neurogenic niche being investigated (i.e., OB and SGZ) (Levy et al., 2019).

In summary, there is some initial evidence for hypothalamic neuroplasticity’s involvement in regulating seasonal changes in neuroendocrine function. Nevertheless, more studies are needed to investigate whether these alterations are also sensitive to EE and how this may relate to animal welfare in livestock farming.

For a summary of the results discussed regarding the hypothalamus, see also Table 1.