ABSTRACT
The production of new neurons in the brains of adult animals was first identified by Altman and Das in 1965, but it was not until the late 20th century when methods for visualizing new neuron production improved that there was a dramatic increase in research on neurogenesis in the adult brain. We now know that adult neurogenesis is a ubiquitous process that occurs across a wide range of taxonomic groups. This process has largely been studied in mammals; however, there are notable differences between mammals and other taxonomic groups in how, why and where new neuron production occurs. This Review will begin by describing the processes of adult neurogenesis in reptiles and identifying the similarities and differences in these processes between reptiles and model rodent species. Further, this Review underscores the importance of appreciating how wild-caught animals vary in neurogenic properties compared with laboratory-reared animals and how this can be used to broaden the functional and evolutionary understanding of why and how new neurons are produced in the adult brain. Studying variation in neural processes across taxonomic groups provides an evolutionary context to adult neurogenesis while also advancing our overall understanding of neurogenesis and brain plasticity.
Introduction
In adult centers, the nerve paths are something fixed and immutable: everything may die, nothing may be regenerated. It is for the science of the future to change, if possible, this harsh decree.
Santiago Ramón y Cajal
While early neuroscience research progressed from rejecting to accepting that neurogenesis occurs in adult mammals, other researchers were demonstrating this phenomenon in other taxonomic groups in parallel. Reptiles were of interest because, traditionally, many species in this group have been thought to exhibit indeterminate growth, which is reflected by increased brain size and cell numbers over their lifespan (e.g. López-García et al., 1984; Ngwenya et al., 2013). Although an increase in these neural factors could be attributable to increases in other neural attributes, such as dendritic branching and gliogenesis, researchers did not discount the idea that the production and addition of new neurons may also contribute to brain growth. As early as the 1950s, researchers had identified undifferentiated cells in the brains of adult reptiles (e.g. Fleischhauer, 1957) and, several decades later, proliferating cells were identified as new neurons that could integrate and function within the existing neural network (e.g. García-Verdugo et al., 1986, 1989; López-García et al., 1988, 1990; Ngwenya et al., 2018; reviewed in Font et al., 2001).
Since those early studies, research on adult neurogenesis has exponentially increased, and research in evolutionary neuroscience and neuroethology has started to clarify the functional and evolutionary significance of adult neurogenesis across taxa. This Review will begin by discussing the similarities and differences between reptilian and mammalian postnatal neurogenesis, highlighting some advantages of a reptilian model of neurogenesis. As a caveat, reptiles are a very large group including more than 11,000 non-avian species (reptile-database.org), but studies examining reptilian postnatal neurogenesis have been limited to a small number of species; thus, it is unclear how well these studies generalize to other reptilian species. This Review will also address how studying animals in the field can broaden our understanding of why and how new neurons are produced in the adult brain, and will go on to consider the utility of reptiles as study animals in this context. Because adult neurogenesis in mammals appears to be highly restricted in the brain, occurring primarily in the olfactory bulb and the hippocampus, and to a lesser extent in other areas (reviewed in Migaud et al., 2010), I will primarily focus on adult neurogenesis in the putative reptilian homologue of the mammalian hippocampus, the medial cortex.
The reptilian medial cortex
The evolutionary trajectory of brain structures is currently unclear and is the subject of debate (e.g. Medina et al., 2013). Thus, identification of putative homologues of the mammalian hippocampus in other vertebrates tends to be approached with caution. Although this Review will not provide an exhaustive treatise of the structural similarities and differences between reptiles and mammals, many recent reviews have highlighted that the medial cortex of reptiles does exhibit some similarities with the mammalian hippocampus, based on cytoarchitecture, neuronal types, neuronal morphologies, connectivity, electrophysiological characteristics, genoarchitecture and neurotransmitter types (Striedter, 2016; Tosches et al., 2018; Medina et al., 2017; Martínez-Cerdeño et al., 2018). The medial cortex of the reptilian brain exhibits trilaminar organization, with a soma-dense cell layer flanked by the relatively soma-sparse outer and inner plexiform layers (Fig. 1) (e.g. Luis de la Iglesia et al., 1994; Luis de la Iglesia and Lopez-Garcia, 1997). Although parts of the mammalian hippocampus also demonstrate laminar organization, it remains unclear whether all mammalian hippocampal subfields exist in reptiles (Streidter, 2016). The classes of neurons found within the medial cortex vary across studies, with from one up to six different types of cortical neurons reported (e.g. Ulinski, 1979; Srivastava et al., 2009; Luis de la Iglesia et al., 1994; Luis de la Iglesia and Lopez-Garcia, 1997). This variation is likely to be attributable to a number of factors, including the criteria used to classify neurons, the study species and the methodology; thus, it remains unclear exactly how many neuronal types exist in the reptilian medial cortex. A recent study used transcriptomics to examine the evolution of cell types between mammals and reptiles, and found that although mammals have evolved new glutamatergic neuronal types, the two groups share GABAergic neurons (Tosches et al., 2018; reviewed in Martínez-Cerdeño et al., 2018), suggesting that at least some hippocampal cell types have been conserved across the vertebrate lineage. Neuron connectivity also appears to be conserved in some cases; some neuronal types within the reptilian medial cortex share connectivity characteristics with the bipyramidal neurons of mammals, with dendrites extending into the outer and inner plexiform layers, whereas other neurons exhibit zinc-positive axonal projections similar to those of hippocampal mossy cells (Lopez-Garcia and Martinez-Guijarro, 1988; Luis de la Iglesia et al., 1994; Luis de la Iglesia and Lopez-Garcia, 1997; Martinez-Guijarro et al., 1991). However, mammals and reptiles differ in some aspects of connectivity; in mammals, information flow is generally unidirectional towards neocortical substructures, whereas in reptiles there is greater reciprocity and connectivity to subcortical structures (reviewed in Striedter, 2016). Currently, evidence suggests that there are structural similarities and differences between the reptilian medial cortex and the mammalian hippocampus, indicating some evolutionary relationship between the two, but no studies have demonstrated or suggested a direct homology.
Interestingly, the reptilian medial cortex appears to be more functionally than structurally homologous to the mammalian hippocampus (e.g. Rodríguez et al., 2002; Streidter, 2016). The hippocampus is one of the most highly studied regions of the mammalian brain, and research has demonstrated that the hippocampus is important in memory processing (particularly that of spatial memory) and pattern separation (e.g. reviewed in Treves et al., 2008). Although the reptilian literature is not as robust, studies to date have also found a relationship between spatial processing and the medial cortex (e.g. Rodríguez et al., 2002). Reptiles have been shown to use spatial maps to orient (Holtzman et al., 1999; Zuri and Bull, 2000; Lopez et al., 2001; LaDage et al., 2012) and increased demands on spatial processing upregulate cortical volume (e.g. Holding et al., 2012; LaDage et al., 2009). Further, lesions to the medial cortex impair performance on spatial tasks (Rodríguez et al., 2002; Lopez et al., 2003). Thus, although structural homology has been difficult to demonstrate, both the mammalian hippocampus and the reptilian medial cortex appear to subserve spatial cognitive demands.
Characteristics of postnatal neurogenesis in the reptilian brain
Neurogenesis is a generalized term for a process that consists of distinct developmental phases including proliferation, migration, differentiation and survival (reviewed in Kempermann et al., 2004). Each of these phases can be differentially regulated by many factors, both genetic and environmental (see below) (e.g. Zhao et al., 2008), and a cell's neuronal destiny cannot be predicted by its proliferative capacity. For example, a proliferating cell may be destined to become a glial cell, may become a non-functional neuron or may go through apoptosis without incorporating into the existing neural architecture (e.g. Kempermann et al., 2004; Pérez-Domper et al., 2013). Mammalian studies are beginning to reveal factors that differentially affect each phase of postnatal neurogenesis, providing a more holistic picture of the generalized process of neurogenesis. Although reptilian postnatal neurogenesis appears to proceed through phases that mirror the mammalian phases, the paucity of studies currently precludes meaningful extrapolation.
As discussed above, in mammals, there are two primary and well-studied areas in which new neurons are produced (although other areas of the brain such as the hypothalamus have demonstrated the presence of newly divided cells, e.g. Migaud et al., 2010). First, new neurons are produced at the subventricular zone, then enter the rostral migratory stream and incorporate into the olfactory bulb. The other common site of postnatal neurogenesis, the subgranular zone, is located within the hippocampus and produces granule cells that are incorporated into the dentate gyrus (e.g. Barker et al., 2011).
In the studies to date, the production of new neurons is more diffuse within the reptilian brain than in the mammalian brain, with new neurons produced along the proliferative zone of the lateral ventricles. New neurons radially migrate away from the proliferative zone and, between one to three weeks later, cells can be found some distance away, with many soma located within the inner plexiform layer of the medial cortex. After one month of age, most soma of newly generated cells are found within the middle cell layer and have incorporated into the existing neural network (e.g. Lopez-Garcia et al., 1990).
Just as reptiles have more diffuse regions of proliferation compared with mammals, the areas to which new neurons migrate and are incorporated are also less constrained throughout the brain. In mammals, new neurons are primarily incorporated in the dentate gyrus of the hippocampus and the olfactory bulb, whereas the incorporation of new neurons occurs throughout the reptilian telencephalon, including the cortices, dorsal ventricular ridge, nucleus sphericus and septum, as well as the olfactory bulb (Fig. 2) (Lopez et al., 1988, 1990; García-Verdugo et al., 1989; Pérez-Sanchez et al., 1989; Pérez-Canellas et al., 1996, 1997; Font et al., 1997). Thus, the regions of proliferation and the location of incorporation are much more restricted in mammals compared with reptiles, with some researchers postulating that neuronal migration is constrained in bigger, structurally more complex mammalian brains (e.g. Paredes et al., 2016).
Interestingly, although the location of postnatal neurogenesis is relatively constrained yet consistent across mammalian species, the patterns in production rate of new neurons can vary dramatically from species to species. Early studies documented ∼300 proliferating cells mm–3 within the primate hippocampus (e.g. Gould et al., 1998; Eriksson et al., 1998), but studies on other mammals have demonstrated that this number can vary, even among closely related species (e.g. Amrein et al., 2004), with one study finding few to no new hippocampal neurons in nine species of bats (Amrein et al., 2007). Of the handful of studies in reptiles, the number and/or density of new neurons also appear quite variable and can differ based on location of production along the ventricle, brain region, age, season and species (Pérez-Cañellas and Garcia-Verdugo, 1996; Pérez-Cañellas et al., 1997; Font et al., 2001; Marchioro et al., 2005). For instance, although some lizard species can generate thousands of new neurons per day (e.g. López-García et al., 1984; Font et al., 2001), the number of neurons generated is highly heterogeneous among brain structures. In one study, Delgado-González et al. (2008) documented almost 2500 cells mm–2 in the anterior olfactory nucleus, yet the number of new neurons in the medial cortex was closer to 300 cells mm–2. In a different species, proliferation was much higher in the medial cortex, with almost 40% of the medial cortical neuron population labeled as less than 30 days old (Pérez-Cañellas and Garcia-Verdugo, 1996). This represents a large proportion of the total number of neurons present within the medial cortex and it is unclear whether the proliferation of large numbers of cells correlates with the actual incorporation of new neurons into the existing neural architecture; if reptiles are similar to mammals, there is likely a high degree of apoptosis of newly proliferated cells. Currently, it is unclear why mammalian and reptilian species vary in the number of new neurons produced and why structures of the brain vary in the number of new neurons. We also do not know whether species-typical differences can differentially modulate the stages of neurogenesis, or what function these new neurons serve.
As mentioned above, most reptiles demonstrate indeterminate growth, and the overall number of neurons within the brain of some species positively correlates with snout–vent length (Lopez-Garcia et al., 1984). It is possible, therefore, that the production of new neurons may not be specific to the brain and may reflect an overall higher rate of mitotic activity throughout the body (e.g. Dunlap, 2016). Recently, some studies have demonstrated the absence of growth plate cartilage in some adult lizard species, which suggests that growth can be determinate in some species (Frýdlová et al., 2016; 2019). As such, examining closely related reptilian species that differ in growth patterns could reveal the link between generalized mitotic activity due to growth across the lifespan and postnatal neurogenesis. If species with determinate growth do not exhibit increased neurogenesis and number of neurons across the lifespan, then increased mitosis due to overall body growth could partially explain higher rates of neurogenesis and number of neurons in indeterminate growers. Similarly, some lizards demonstrate caudal autotomy with a concurrent acute increase in mitotic activity at the site of injury (e.g. Lozito and Tuan, 2017); this could also potentially stimulate a whole-body increase in mitosis and neurogenesis. Currently, the few studies that have examined this have found that leopard geckos (Eublepharis macularius) with autotomized tails have high rates of mitosis at the site of amputation without a concurrent increase in cellular proliferation in the cortices or in myocardial cells (McDonald and Vickaryous, 2018; Jacyniak and Vickaryous, 2018). Therefore, in leopard geckos, there are likely to be other factors that contribute to the modulation of neurogenesis beyond a transient increase in mitotic activity due to tail autotomy; it remains unclear whether this process is similar in other species.
Functional significance of reptilian postnatal neurogenesis
In mammals, the functional significance of postnatal neurogenesis is presently equivocal, although several factors have been shown to modulate rates of neurogenesis in the laboratory. As mentioned above, increased demands on spatial learning and memory cause an upregulation of mammalian adult neurogenesis (e.g. Gould et al., 1999; Dupret et al., 2007). In addition, motor stimulation increases cellular proliferation rates, and an enriched environment increases the survival of new neurons (reviewed in van Praag et al., 2000), whereas intensive stressors downregulate neurogenesis in adult mammals (e.g. Gould et al., 1997, 1998). However, there is a paucity of studies examining whether these same factors regulate rates of postnatal neurogenesis in reptiles. Powers (2016) found that two groups of painted turtles (Chrysemys picta) – those housed with other individuals and those given the opportunity for increased motor activity – have more new neurons than turtles housed alone, suggesting that social environmental enrichment and motor stimulation are sufficient to upregulate adult neurogenesis in this species. In contrast, a different study found that translocated northern Pacific rattlesnakes (Crotalus oreganus) have larger activity ranges than non-translocated snakes, with an assumed increase in motor activity, yet translocated snakes do not demonstrate a concomitant increase in new neuron production in the medial cortex (Holding et al., 2012). In side-blotched lizards (Uta stansburiana), larger enclosures are sufficient to upregulate medial cortical neurogenesis. However, this is only found in territorial males; non-territorial males housed in larger enclosures have a similar number of new neurons as non-territorial males housed in small enclosures (LaDage et al., 2013). The results of this study suggest that although environmental enrichment can upregulate neurogenesis, differential life-history characteristics may modulate the effects of environmental enrichment on rates of neurogenesis. Overall, it appears that some of the same factors that modulate postnatal neurogenesis rates in mammals (i.e. environmental enrichment and motor stimulation) can do so in some reptilian species as well; however, the functional significance of reptilian postnatal neurogenesis remains elusive.
Neurogenesis and remodeling of the reptilian brain after injury
Mammalian models have demonstrated some capacity for structural and functional repair via postnatal neurogenesis after a traumatic brain injury; indeed, neurogenesis may be involved with spontaneous recovery from such an injury (e.g. Rice et al., 2003; Blaiss et al., 2011). However, reptilian models of brain damage and remodeling are unparalleled; the reptilian medial cortex can be nearly destroyed and then completely remodeled, to the point of being indistinguishable from the previously undamaged cortex. For example, chemical lesions of the medial cortex can destroy up to 95% of all neurons within the structure, leading to impairment of spatial cognitive function (Font et al., 1989; 1997; Molowny et al., 1995; Austin et al., 2020), but within a week of injury, there is a burst of proliferative activity at the ventricular zone, with cells migrating out into the medial cortical structure and forming new connections (e.g. Font et al., 1997, 1997; Molowny et al., 1995). The structural repair of the medial cortex is highly contingent on the generation of new neurons and glial cells, with full recovery of the medial cortex achieved within three months (e.g. Romero-Aleman et al., 2004; López-García et al., 2002; Ramirez-Castillejo et al., 2002; Austin et al., 2020). Thus, damage-induced adult neurogenesis is very prevalent in reptilian models of brain injury and neural remodeling.
Broadening our understanding of postnatal neurogenesis
Currently, much of what we know about postnatal neurogenesis has been discovered in rodent models within the laboratory (e.g. Oppenheim, 2019). However, it remains unclear whether the factors that modulate neurogenesis in model rodent species are translatable to other species as well. Further, the environment in the laboratory is tightly controlled and not representative of the range of factors that could modulate neurogenesis in the field. Reptiles are an evolutionarily important link in the vertebrate lineage, and species in this taxon demonstrate variability in many important factors (e.g. ecological niche, life history characteristics) that could modulate and ultimately select for differential rates of postnatal neurogenesis. Understanding how these factors correlate with postnatal neurogenesis will broaden our understanding of neurogenesis in general. Next, I consider the effects of captivity, seasonality and sex on variability in reptilian postnatal neurogenesis, although this is far from an exhaustive review of all the factors that could modulate neurogenesis.
Effects of captivity
Selective breeding in a laboratory setting may cause significant and heterogeneous changes in the neural architecture when compared with that of wild conspecifics (e.g. Welniak-Kaminska et al., 2019), suggesting that results attained from laboratory animals may not mimic patterns of neurogenesis found in the field. Further, laboratory conditions restrict physical activity, social interactions and foraging, and remove the effects of predation and environmental fluctuations, all of which could influence the process of neurogenesis. Several recent mammalian studies have demonstrated that the process of adult neurogenesis is differentially modulated in the field compared with the laboratory environment (e.g. Cope et al., 2019). Interestingly, although increased physical activity robustly correlates with increased cellular proliferation in the laboratory, this phenomenon has not been replicated in wild rodents (e.g. Schaefers, 2013; Klaus and Amrein, 2012; Hauser et al., 2009). Of the studies examining neurogenesis in wild rodents, most suggest that habitat- and species-specific factors have selected for neurogenesis rates; thus, neurogenesis in wild rodents is less susceptible to short-term perturbations compared with what has been found in captive animals (e.g. Cavegn et al., 2013; Amrein, 2015; Oppenheim, 2019). As such, researchers have advocated for examining other species in their natural habitats to understand the true range of variability in the process of adult neurogenesis and how selection has molded the process based on differential ecological variables (e.g. Calisi and Bentley, 2009; Bonfanti and Amrein, 2018; Faykoo-Martinez et al., 2017).
Although there is a paucity of studies examining this phenomenon in reptiles, some studies have demonstrated that captivity can modulate neural variation in lizards. For example, side-blotched lizards are polymorphic and, in some populations, males can be found in one of three different morphotypes (Sinervo and Lively, 1996; Sinervo et al., 2000; 2006). The orange morph is highly territorial, defending relatively large territories, the blue morph is less aggressive than the orange morph and defends smaller territories, whereas the yellow male does not hold a territory and sneaks copulations on the territories of the other male morphotypes (Sinervo and Lively, 1996; Zamudio and Sinervo, 2000; Sinervo et al., 2000; 2006; Calsbeek and Sinvero, 2002). Our previous work on field-caught individuals found that cortical volume (specifically, the dorsal cortex with a trend in the medial cortex) correlates with territorial behaviors – orange and blue males have larger cortices than yellow males – likely because of the increased spatial processing demands of holding and defending a defined boundary (LaDage et al., 2009). Interestingly, adult laboratory-reared individuals whose wild-caught parents were from these same populations have a significantly lower medial cortical volume compared with field-caught individuals; furthermore, the differences among morphotypes that we found in the field disappear in laboratory-reared individuals (LaDage et al., 2016). Interestingly, although medial cortical volume was suppressed in captivity, dampening differences between morphotypes, neurogenesis could be modulated within a captive environment. Territorial and non-territorial males held in typical laboratory enclosures do not differ in the number of new neurons produced; however, when males reside in larger enclosures, territorial males upregulate neurogenesis whereas non-territorial males do not (LaDage et al., 2013). Thus, although captivity suppresses cortical volume, neurogenesis can still be modulated by environmental experiences within the laboratory. Delgado-González et al. (2008) also found that captivity can modulate neural attributes in Tenerife lizards (Gallotia galloti): laboratory-housed lizards have lower levels of neuronal proliferation than wild conspecifics of the same age and from the same population. Further, variation across the year in the number of proliferating cells is less variable in the laboratory than that seen in the field. These results strongly suggest that the variability in adult neurogenesis observed in captive lizards may not reflect the true range of variability found in more naturalistic settings.
There are also few studies examining captive versus wild individuals in all taxonomic groups; however, there is tentative agreement that captivity likely downregulates neural attributes. Many reptilian species can be leveraged to examine this phenomenon and questions therein, as many species are easily captured in the field and quickly adapt to a captive environment. Exploiting these reptilian model species will allow researchers to compare neural attributes between natural and captive individuals, understand the timeline across which downregulation of attributes occurs in captivity, elucidate whether differential variation in neural attributes occurs in response to captivity, and understand how factors that affect neurogenesis in the laboratory translate to those found in the field. Also, because reptiles are a large taxonomic group, they exhibit a plethora of variation in ecology and life-history traits in the field; these factors could potentially modulate neurogenesis beyond what we see in model rodent species in the laboratory. Although not exhaustively examined in other taxonomic groups, some studies have demonstrated the relationship between variation in spatial processing demands and the hippocampus in the field. For instance, in black-capped chickadees (Poecile atricapillus), variation in hippocampal neurogenesis across populations is associated with naturally selected variation in spatial processing demands (Chancellor et al., 2010). In vole species, a larger home range is associated with a larger hippocampus, suggesting that differential spatial demands in the context of home range can modulate neural attributes in the field (Jacobs et al., 1990). Reptilian species also exhibit variation in spatially based behaviors (e.g. differences in territoriality, foraging and predation avoidance) that could be leveraged to test hypotheses concerning the relationship between adaptive spatial behaviors and the brain. Conducting more studies in natural and semi-natural environments, and including more reptilian species, will help to broaden our understanding of the process of adult neurogenesis, allowing for greater nuance in our knowledge, while clarifying the adaptive significance of postnatal neurogenesis within an ecological context (e.g. Barnea, 2010).
Seasonal effects
Because reptiles are poikilothermic and many species are seasonal breeders, variation in temperature and photoperiod across the year is mirrored by a myriad of changes in physiology and behavior, including seasonal variation in rates of neurogenesis. In the laboratory, it has been reported that warmer temperatures and a longer photoperiod increase cell proliferation and the distribution of cells throughout the cortex (Ramirez et al., 1997; Marchioro et al., 2012; Penafiel et al., 2001). Interestingly, studies in the field have found varying patterns of cell proliferation rates across the year. In the Italian wall lizard (Podarcis sicula), proliferation rates are higher in the summer compared with the spring (Margotta, 2014), whereas in the Tenerife lizard neurogenesis rates are greatest in the spring and are reduced in the autumn and winter (Delgado-González et al., 2008). In Tsingling dwarf skinks (Scincella tsinlingensis), the proliferation, migration and survival of new neurons is greatest in the autumn, followed by summer, spring and winter (Yang et al., 2017); red-sided garter snakes (Thamnophis sirtalis parietalis) also demonstrate a higher number of proliferating cells in the autumn compared with the spring (Maine et al., 2014). Finally, in the Iberian wall lizard (Podarcis hispanica), no difference in cellular proliferation was found when comparing individuals in April and November (Sampedro et al., 2008), yet both of those time points have higher proliferation rates than other times of the year (Font et al., 2012).
Interestingly, better studied taxonomic groups also provide no clear picture of seasonal neurogenesis. For instance, in one of the first studies to demonstrate adult hippocampal neurogenesis in the field, Barnea and Nottebohm (1994) found a seasonal peak in neuronal production that coincided with a peak in food caching behavior in black-capped chickadees. However, since then, there has not been clear cut evidence demonstrating a correlation between seasonality and food caching behavior in other species (reviewed in Pravosudov, 2006; Sherry and Hoshooley, 2009). This also reflects what has been discovered in field-caught mammals; some species demonstrate variation in neuronal proliferation across the year, whereas others do not (Galea and McEwen, 1999; Lavenex et al., 2008; reviewed in Migaud et al., 2015). Taken together, we can conclude that, although neurogenesis rates vary across the season in some species, there is no clear predictability in what that pattern may look like. It remains unclear whether the different seasonal patterns of postnatal neurogenesis found across studies are due to differences among species, ecology or methodology.
Sex-based differences in the field
Previous studies indicate that some brain regions in reptiles are sexually dimorphic (e.g. Crews et al., 1990), yet relatively few reptilian studies have examined sex-based differences in rates of adult neurogenesis. In the Iberian wall lizard, Sampedro et al. (2008) found that males have bigger olfactory bulbs with more new cells compared with females, even after correcting for overall brain size. The authors suggest this may be related to sex-based differences in chemosensory abilities, in that males demonstrate the ability to chemically distinguish between species, whereas females do not (Barbosa et al., 2006). Lutterschmidt et al. (2018) also found sex-based differences in cell proliferation rates in red-sided garter snakes (T. sirtalis). In red-sided garter snakes, males emerge from hibernacula about 2 weeks earlier than females. Upon female emergence, the snakes breed, then migrate to foraging grounds. Lutterschmidt et al. (2018) found that males have more proliferating cells than females at the time of female emergence; females subsequently upregulate proliferation during migration. The authors suggest that the differential timing of the life-history transitions and behaviors between males and females may explain the differences in cell proliferation in the brain from the time of emergence through to migration. Although these studies indicate the existence of sex-based differences in neurogenesis rates, little evidence has accumulated for a correlation between testosterone level and neurogenesis (e.g. Powers, 2016), suggesting that other factors can also modulate the relationship between postnatal neurogenesis and sex. Again, these results reflect those found in field-caught mammals and birds: no consistent pattern has been found associating sex or gonadal hormones with new neuron production in the hippocampus (reviewed in Duarte-Guterman et al., 2015; Spritzer and Roy, 2020).
Conclusions and future directions
Although reptiles may not be an obvious choice in which to advance our understanding of adult neurogenesis, we can harness the benefits of this taxonomic group to gain a different perspective, which can then broaden our understanding of the process of neurogenesis across vertebrates. Because reptiles maintain high rates of postnatal neurogenesis, this may allow for greater sensitivity when examining the factors that modulate neurogenesis, as the resultant effects of a manipulation may be better visualized in a species with many new neurons. Tangentially, the high degree of plasticity and neurogenesis exhibited by reptiles could serve as a starting point in basic research attempts to understand the mechanisms underlying neural repair. Because reptiles demonstrate a unique ability for dramatic neural remodeling, studying the mechanisms involved may serve to inform clinical studies concerning neural repair in mammals. Overall, the unique neural characteristics of reptiles can be leveraged to gain further insight into the basic process of postnatal neurogenesis.
The number of reptilian species in which postnatal neurogenesis has been studied to date is relatively low, yet reptiles are a highly diverse group that vary in ecology and life-history characteristics. This variation can be leveraged to test different hypotheses regarding the influence of these characteristics on neurogenesis in reptiles, particularly in wild-caught animals. Because animals brought into the laboratory or raised in the laboratory do not typically reflect the neural architecture of their wild counterparts, studying reptiles in the field will expand the boundaries of our understanding of the factors that modulate neurogenesis in a natural environment. Further, more naturalistic studies will allow for a greater knowledge regarding how differences in the environment, season, population, morphotype, species, predation risk, spatial use and other ecological factors may differentially modulate neurogenesis. Studying differences in neurogenesis across these different contexts will allow us to better understand the functional significance of adult neurogenesis, how selection has acted on this process and how neurogenesis operates in the natural environment.
From a comparative viewpoint, reptiles are an evolutionarily important link in the vertebrate lineage, and being aware of the similarities and differences between reptiles and other taxonomic groups in the factors that modulate adult neurogenesis has ramifications for our understanding of the evolution of the neural substrate. Compared with mammals, and (to a lesser degree) birds and fish, there is a paucity of literature on the neural anatomy and physiology of reptiles. Although some studies have found that factors that modulate neurogenesis rates in mammals also do so in reptiles, other studies have not (e.g. Holding et al., 2012). Further, in the studies that have found similarities between mammals and reptiles, it is still unclear whether these factors differentially affect the different stages of neurogenesis (i.e. proliferation, migration and survival). Similarly, it remains unclear whether variable numbers of neurons produced in different contexts and in different areas of the brain have specific functional relevance. Only by studying postnatal neurogenesis in a range of taxonomic groups can we begin to thoroughly understand the similarities and differences in the process, its evolutionary trajectory and the possible behavioral outcomes across taxa; doing so will provide a more holistic picture of the mechanisms and the functional significance of adult neurogenesis.
Acknowledgements
I thank Sarah Alderman and Matthew Vickaryous for organizing the session ‘Brain Building: Plasticity in Form and Function of the Central Nervous System’ for the Society for Experimental Biology's (SEB) Annual Meeting. The session was timely and the conversations with the participants were stimulating. I would also like to thank three anonymous reviewers for feedback that greatly improved the content of the manuscript.
Footnotes
Funding
The American Association of Anatomists Outreach Grant Program generously supported travel to the SEB Annual Meeting.
References
Competing interests
The author declares no competing or financial interests.