Central intracrine DHEA synthesis in ageing-related neuroinflammation and neurodegeneration: therapeutic potential?

Inflammation is an integral part of the innate immune system: it not only acts as the first line of defence against pathogenic attack and injury, but also serves to execute final humoral immune effects (e.g. antibody-mediated pathogen destruction). It is thus a normal physiological process that is essential for the maintenance of homeostasis. An equally normal consequence of the inflammation-facilitated repair of tissues after any insult is the transient disruption of the local cellular homeostasis of non-infected or uninjured cells. This injury-repair cycle is efficient during youthful years, but is affected by ageing, resulting in a relatively impaired ability of the body to regenerate damaged tissues [13]. Systemically, immunosenescent changes in the innate immune system are characterised by a loss of phagocytic capacity, reduced efficiency in leukocyte (neutrophil and macrophage) chemotaxis, reduced intracellular, but increased extracellular levels of free radicals and an increase in production of “early” or acutely responsive pro-inflammatory cytokines, such as interleukin (IL)-1β, IL-6 and tumour necrosis factor (TNF)-α [17‐20], all of which contribute to a more pro-inflammatory phenotype and which increase the burden of tissue damage-recovery cycles. The loss in efficiency of this process, coupled with an increased exposure to antigenic load and modern lifestyle associated chronic stressors, results in the reduced ability of the body to maintain normal immune function. This manifests as an increased risk for malignancy, auto-immune responses, an impaired ability to mount proper immune responses to stimuli and prolonged inflammatory activity during acute infections [21‐23].

Since the turn of the century, several papers suggesting beneficial effects after DHEA supplementation, in the context of neuroprotection, have been published—we have summarised findings illustrating the variety of models used in Table 2. Reported benefits include modulation of neurogenesis, neuronal function, metabolism and longevity. However, a significant limitation to available information is that the majority of these effects have only been illustrated in in vitro and animal studies.

It is clear that DHEA is neuroprotective within these various disease model contexts, but how is DHEA relevant to inflammation—and in particular to neuroinflammation-related neurodegeneration?

To the best of the authors’ knowledge, very little evidence for direct anti-inflammatory action of DHEA or DHEAS in the brain or central nervous system (CNS) has been elucidated per se. We were however able to find at least three studies that have demonstrated the anti-inflammatory effects of DHEA in vitro in models relevant to the brain or CNS. The first demonstrated that DHEA inhibits the synthesis of TNF-α and IL-6 by cultured foetal rat astroglia in response to exposure to Mycoplasma fermetans [62]. The second and third studies, which are also the most recent, were performed both in vitro and in vivo [63, 64]. In the in vitro study, DHEA treatment of cultured mouse microglia exposed to LPS showed a significant reduction in TNF-α, IL-6, IL-12 and monocyte chemoattractant protein 1 (MCP-1) mRNA expression when compared to LPS-exposed microglia receiving no treatment [64]. The in vivo study was an experimental model of autoimmune encephalomyelitis (EAE) in mice, which showed a significant reduction in IL-1β and interferon gamma (IFN-γ) mRNA expression in spinal cord tissue in animals treated with DHEAS when compared to untreated EAE-affected animals [63].

Numerous other studies have illustrated the anti-inflammatory action of DHEA and DHEAS in systems and disease models other than those related to the brain or CNS (Table 3). Although this does not necessarily prove that these same effects may be seen in the brain under the same or similar circumstances, the possibility cannot be discounted.

From these studies, it seems that DHEA exerts its anti-inflammatory effects primarily by modulating either the effects or production of pro-inflammatory cytokines. However, it also appears that the same pro-inflammatory cytokines have the ability to regulate DHEA production in turn. For example, in macrophage-depleted primary murine leydig cell cultures, both IL-1 and TNF-α were shown to inhibit androgen steroidogenesis by downregulating the expression of cytochrome P450c17 mRNA—which encodes for an enzyme that is critical to androgen formation [65, 66]. TNF-α has shown similar inhibitory effects in cultured porcine leydig cells by reducing the binding of luteinising hormone (LH) or human chorionic gonadotropin (hCG) to the respective receptor—which would stimulate steroidogenesis [67]. These authors further suggested that the primary inhibitory effect of TNF-α was mainly by decreasing availability of cholesterol, the initial substrate required for steroidogenesis [67].

Regardless of how cytokines may inhibit the synthesis of androgens, these studies provide a plausible explanation for the age-related decline of DHEA not only systemically, but centrally as well—given the relatively pro-inflammatory status associated with ageing. Furthermore, since it has been shown that the inflammatory maladaptation seen with ageing affects the brain to a larger extent (specifically the hippocampus), it is plausible that the age-related decline of DHEA occurs even more rapidly in the brain than it does the rest of the body.

In contrast to the idea that DHEA is largely beneficial, a minority of studies have suggested that DHEA may have neurotoxicity under specific conditions. In a model of global cerebral ischaemic insult in rats [68], intraperitoneal administration of a supraphysiological dose of DHEA 3 to 48 h after ischaemic injury was neuroprotective (e.g. decreased neuronal death in the CA1 hippocampal region). However, DHEA administration 1 h before or after the insult exacerbated the effects of the injury. Given the known anti-inflammatory role of DHEA, this may suggest that (at least at this high dose) DHEA inhibited early inflammatory or oxidative responses that were required for cytoprotection. Additionally, decreased cell viability was reported in primary murine neuronal cultures and human neuroblastoma (SK-N-SH) cells after (24–72 h) exposure to a large dose-range of DHEA—a result not seen in mixed neural or glial cultures. Furthermore, the neurotoxic effects of DHEA were abrogated by simultaneous treatment with DHEAS [69]. From this, it is difficult to formulate an interpretation on potentially toxic effects of DHEA. On the one hand, it is certainly feasible that a high dose of an antioxidant such as DHEA may have damaging effects, as this has been illustrated in the nutraceutical literature [70]. On the other hand, the second study clearly highlights the limitations of interpretations possible from single cell culture models, although it does provide insight on the complexity of the mechanisms at play. In our opinion, given the relatively larger body of evidence in support of a beneficial effect of DHEA, significant concern is probably not warranted in our context of intracrine DHEA production.

In order to further our understanding of how DHEA may impact on the ageing brain, it is necessary to delve deeper into the potential molecular actions and interactions of DHEA and DHEAS.

DHEA supplementation has historically been used in the treatment of reproductive-related diseases, particularly in women, where it has been shown to alleviate menopause-related pathologies such as vaginal atrophy [91]. In support of our interpretation of benefit from maintaining DHEA availability, in diseases not related to reproduction, DHEA supplementation has been shown to be beneficial to both sexes in the treatment of numerous other pathologies or conditions (see Table 4).

Despite these many clinical studies involving DHEA supplementation, to the best of the authors’ knowledge, the effects of exogenous DHEA supplementation has not been directly assessed in clinical manifestations of neurodegeneration or any related pathologies. Similarly, there are also currently no clinical studies assessing the effects of DHEA or DHEAS supplementation as an adjuvant therapy for neurodegenerative diseases in either the European or American clinical trial database.

In addition, it should be noted that other studies have reported oral DHEA supplementation to show little to no benefit in the treatment of other ageing-related pathologies such as cognitive performance, lipid metabolism, glucose metabolism, bone health and muscle function [92‐94].

However, the available clinical studies collectively do provide some insight. The conditions included in our summary (Table 4), which all suggests benefit from DHEA administration, are linked by the fact that inflammation is a characteristic feature in all. Similarly, neuroinflammation is a critical hallmark of the generalised pathophysiology of neurodegeneration and its occurrence appears to have a causative role in this degenerative condition. Thus, given the already discussed anti-inflammatory effect of DHEA (refer to Table 3), as well as the anti-inflammatory therapeutic effect suggested by the results just presented, theoretically, if one can increase DHEA production at cellular level in the central compartment, one should be able to alleviate or prevent neuroinflammation and thus neurodegeneration.

Since exogenous DHEA supplementation has yielded varying benefit, an obvious alternative therapeutic—or preventative—strategy to consider is whether it might be naturally or artificially possible to increase endogenous DHEA production.

From the sport and exercise literature, it is known that exercise in older males can significantly increase serum DHEA and DHEAS concentrations [95‐101]. These results do not provide proof of extragonadal androgen synthesis though, as moderate exercise is known to increase gonadal testosterone production. However, interestingly, similar effects of exercise have been reported for DHEA in older, post-menopausal females [102‐104]. In this population, where gonadal androgen production is less possible, it is possible that extragonadal DHEA production may occur in response to exercise.

Interestingly, a study conducted in Sprague-Dawley rats has illustrated skeletal muscle to have endogenous steroidogenic capacity and that acute bouts aerobic exercise can significantly enhance the localised production of DHEA in both male and female animals [105]. This adds substantial support for our theory that extragonadal DHEA synthesis may be elicited through intervention.

Furthermore, higher DHEAS levels have been reported in long-term practitioners of alternative therapies such as transcendental meditation and the ancient martial art of tai chi, when compared to age-matched controls [106, 107]. Although these studies do not prove that DHEA production is stimulated by these practices, the relatively higher levels suggest at least an effect to sustain long-term production rate. In line with these results obtained in psychology-based practices, environmental enrichment in caged rats was linked to increases in steroidogenic enzyme expression associated with DHEA production, relative to unstimulated rats of the same age [108]. Indeed, enrichment therapy is already being employed to treat patients with dementia and Alzheimer’s disease [109‐113], although the authors could not find data on the effect of this strategy on DHEA synthesis.

Considering this evidence, perhaps it is necessary to first understand which tissues may possess the ability for DHEA synthesis, before considering the feasibility of this approach in the context of neurodegeneration.

DHEA is produced during steroidogenesis, in a manner dependent on mainly two classes of steroidogenic enzymes, namely the P450 enzymes and the hydroxysteroid dehydrogenases [114]. Also, DHEA can be sulphated into DHEAS via enzymes called steroid sulphotransferases (SULT), while conversion of DHEAS back into DHEA is mediated via steroid sulphatases (STS). DHEAS levels continue to increase after puberty, peaking in the mid-20s, after which levels begin to progressively decline with age in both men and women [3, 12]. Serum DHEA and DHEAS levels are thought to be almost exclusively maintained by their synthesis in the adrenal zona reticularis [114]. However, in theory any tissue that expresses the steroid enzymes can be classified as steroidogenic. In fact, in the early 80s Fernand Labrie, a well-known endocrinologist, was the first to note that prostate glands of patients with prostate cancer still had high levels of dihydrotestosterone (DHT) even after castration [115]. He illustrated that some peripheral tissues, such as the prostate, possesses the steroidogenic enzymes needed to transform DHEA and DHEAS into its androgenic and estrogenic metabolites, thereby capable of exerting a relatively more localised effect. The study of this unique ability of peripheral tissues was termed “intracrinology” and has subsequently gained some momentum [116]. Given the importance of intracrine derived hormone production in terms of female reproductive health, it is clear that investigations into intracrinological mechanisms in other contexts are warranted, such as in neurodegeneration and ageing.

Although it is known that intracrinology takes place in peripheral tissues, all tissue types do not express the same isoforms of the steroidogenic enzymes, nor are they expressed in the same quantity or constitutively expressed as we age. For example, a recent study in rats demonstrated that the expression of several steroidogenic enzymes decline significantly in the brain with age [108]. It is important to have an understanding of where the capacity for steroidogenesis exist, as well as to consider sex and species differences, as these aspects significantly impact on choices for experimental models, as well as on the relevance and applicability of research data gained in any particular model. It is not possible to address steroidogenesis in its totality within the scope of this review. Rather, in the next few sections we describe some of these intricacies in the context of DHEA synthesis specifically.

In terms of substrate delivery for steroidogenesis, the steroidogenic acute regulatory protein (StAR) and cholesterol side-chain cleavage enzyme (P450scc) are required for the transport of cholesterol to the mitochondria and conversion into pregnenolone, respectively. Interestingly, these enzymes are expressed at significantly lower levels in the Leydig cells of aged rat testes relative to those of younger rats [117], which—since Leydig cells are the primary cells for testosterone production in the male body—provides a plausible explanation for the age-related decline of the hormone in both rodents and humans. Interestingly, in contrast, P450scc protein and mRNA expression in the human adrenal glands appears to be relatively stable with ageing [118, 119].

In the context of central intracrine production of DHEA, P450scc mRNA has been detected in cultured human astrocytes and oligodendrocytes (but not neurons) in at least one study [120]. Many other studies have detected steroidogenic enzyme expression in both neurons and neuroglial cells in vertebrates other than humans (refer to Fig. 1). We tabulated these studies for brevity, but it is important to note that although steroidogenic enzyme mRNA was detectable in these different animal brain cell types, it is not clearly indicated to what extent the level of gene expression may differ from one cell type to another. Our interpretation of the data presented in most of these cited studies, is that neuroglial cells appear to express higher levels of steroidogenic enzyme mRNA relative to neurons, where levels were barely detectable in some cases. This suggests that mainly neuroglial cells are capable of significant intracrine synthesis, as already reported for peripheral tissue other than gonads and adrenals [57, 58]. In contrast to the P450scc findings, StAR mRNA expression in rat brain brains has been shown to indeed decline significantly with age [121], suggesting a relative decline in substrate delivery similar to that reported for the testis. Whether this occurs in human brain tissue could not conclusively be answered from the literature, but it certainly cannot rule out. Nevertheless, the presence of both StAR and P450scc in astrocytes and oligodendrocytes argues in favour of sufficient substrate availability to indeed allow for localised production of DHEA in these cells.

Another protein of potential relevance in the context of cholesterol transport to the mitochondria has been described. Translocator protein (TSPO), also known as peripheral-type benzodiazepine receptor (PBR), a protein mainly bound to the outer mitochondrial membrane, has also been suggested to play a role in the facilitation of cholesterol transport via StAR [138]. Whether the presence of TSPO is critical in the transport of cholesterol is a matter of controversy within literature. It was shown in mice, with a specific TSPO gene knockout in their Leydig cells, that testosterone could still be produced despite the absence of what thought to be a critical protein for the process [139]. Despite the uncertainty surrounding its role in steroidogenesis, TSPO has recently been identified as a potential biomarker of neurodegeneration, as its expression increases with inflammation and neurodegeneration associated with Alzheimer’s disease, HIV encephalitis and MS [140, 141].

More specifically in terms of its function centrally, in vitro experiments in primary human astrocytes have demonstrated that various synthetic TSPO ligands of the N,N-dialkyl-2-phenylindol-3-ylglyoxylamide class (PIGAs) are capable of activating TSPO activity and stimulating neurosteroidogenesis [142]. In fact, PIGAs have shown to be neuroprotective in l-buthionine-(S,R)-sulfoximine (BOS) induced cytotoxicity of C6 rat glioma cells [143]. The mechanism of neuroprotection was associated with a reduction in lipid peroxidation and inflammation, which was abrogated in presence of the P450scc and StAR inhibitor aminoglutethimide and was therefore thought to occur due to steroidogenesis. Given the known anti-inflammatory and antioxidant properties of DHEA, TSPO may be a role player in the maintenance of DHEA levels.

Another fact in favour of local DHEA production in the brain is the relatively increased demand for DHEA in brain relative to the periphery. Firstly, steroid sulphotransferase isoform 2A1 (SULT2A1)—which increases circulating DHEAS by sulphating DHEA, rendering it bio-inactive [118]—seems to not be present in brain tissue at all, while SULT2B1 expression has only been reported for the prefrontal cortex, hippocampus and cerebellum, but not for the cerebral subcortex or neocortex [144, 145]. Secondly, sulphatase (STS) expression—which is responsible for conversion of DHEAS into the bioactive DHEA—is generally very high in the CNS [6]. Together, this suggests a relatively higher demand for DHEA in the brain when compared to the periphery. This scenario would also increase the likelihood of mechanisms for local production to be in place.

The idea of de novo synthesis of DHEA and DHEAS in the brain itself remains a highly debated topic in the field. As mentioned, the human adrenal glands were initially thought to be the only site of endogenous steroidogenesis, apart from the gonads (i.e. testes, ovaries) which produce DHEA and DHEAS to a lesser extent. However, the brain has recently and controversially been highlighted as a possible alternative site for production. Given the neuroprotective and central anti-inflammatory effects of DHEA already discussed, this possibility for intracrine DHEA production has far-reaching implications for both neurodegenerative disease and ageing—not only to increase our understanding of the disease processes, but also in terms of development of preventative and/or therapeutic strategies.

The first suggestion of DHEA synthesis in the brain was already published in the 1980s—when DHEA and DHEAS concentrations were reported to be higher in rat brains than in circulation [146]. A few years later, similar circumstantial evidence for DHEA synthesis in the human brain emerged [147]. DHEA was subsequently commonly referred to as a “neurosteroid”, based on its relatively increased central localisation. Although DHEA can be detected in both animal and human brain tissue, it remains unclear whether de novo synthesis indeed occurs. For evident reasons, it is practically difficult to accurately assess actual DHEA production in the human brain, so that most data is generated from rodent models. To add complexity to this topic, subsequent evidence may suggest that a species difference in terms of central DHEA levels may exist. In terms of criticism of analytical techniques, it has for example been suggested that DHEA levels in rat brain tissue reported in earlier studies may have been overestimated due a lack of sensitivity of the radio-immunoassays (RIA) technique used at the time, which relied on antibody specificity to detect the steroid hormones [148]. A newer method encompassing solid phase extraction (SPE), high-performance liquid chromatography (HPLC) and gas chromatography-mass spectrometry (GC-MS) was developed to generate more accurate data [148]. Ironically, nearly a decade later it was reported by the same authors that this technique may indeed also be flawed when it came to the detection of sulphated steroids [149]. They showed in rodent samples that the levels of DHEA and PREG were higher in the lipoidal fraction of the SPE. It was suggested that cholesterol contamination may have contributed to these findings through cholesterol autoxidation, generating DHEA and PREG, and that actual levels of DHEA and DHEAS are likely extremely low or essentially non-existent in rat brain tissue [149]. Interestingly, in the same paper, the authors were able to confirm the presence of significant amounts of both DHEA and PREG in human brain tissue samples using this improved methodology. The exact mechanisms of how cholesterol autoxidation leads to DHEA and PREG generation remains to be elucidated. Nevertheless, these results highlight the necessity of considering species in the design of experimental models. The fact that DHEA cannot be detected in rodent brains may be a function of generally lower levels when compared to humans. Thus, although rodents can still be feasible models, the parameters to be assessed, might have to differ depending on analytical sensitivity.

Given these methodological limitations in the measurement of hormone levels directly, many indirect methods have been employed to investigate the possibility of DHEA production centrally. However, results have again posed more questions than answers. For example, on one hand, P450c17 mRNA—which transcribes an enzyme that is crucial in the formation of DHEA—has been detected in cultured human glial cells [120]. Subsequently, P450c17 mRNA expression has also been demonstrated in the amygdala, caudate nucleus, cerebellum, corpus callosum, hippocampus and thalamus of the human brain [119]. In addition and similarly to circulating DHEA levels, the enzyme expression also declines with age in the rat brain, which argue in favour of the possibility for local DHEA synthesis [150]. Interestingly, the age-related decline in P450c17 mRNA was reported to be more pronounced in the cerebral cortex and cerebellum, but not the hypothalamus [150]—this fact should be further explored in order to determine its relevance.

However, the expression of P450c17 in the human brain has been a topic of debate – at least two other studies could not detect any P450c17 mRNA in the human cerebellum, hippocampus or temporal lobe and reported no P450c17 activity in the temporal lobe [145, 151]. There are a variety of possible explanations for these different results. Firstly, it is known that age has a significant effect on steroidogenic enzyme expression, so it is possible that this may have affected the results. Indeed, in the study where the P450c17 mRNA was detected, samples were pooled from patients between the ages of 10 to 78 years. On the other hand, in at least one of the studies failing to detect enzyme activity, samples were collected from a much older patient cohort (80 years and older) [119, 151]. Secondly, in terms of the failure of any group to report activity of the enzymes which catalyse the reaction in the brain [6], there is currently no way to determine at which rate potential local DHEA synthesis would occur. An important consideration here is that P450c17 expression reported for the CNS is substantially lower than for the testes, which exceeds it by nearly 200-fold [6]. It is therefore likely that the conventional methods previously employed in studies with negative outcome, may not have been sufficiently sensitive to detect these possibly very low levels of activity. Thus, the issue remains to be elucidated across the ageing spectrum, to eliminate the potential for a false negative outcome by using aged individuals.

Additionally, it is possible that localised DHEA biosynthetic mechanism in the brain may differ from that at other sites of local DHEA production. One group has indeed suggested that DHEA and DHEAS can be synthesised in manner independent of P450c17. Their evidence suggests that cultured human astrocytes and oligodendrocytes, but not neurons, could synthesise DHEA the presence of a P450c17 and 3β-HSD inhibitor. It is suggested that this alternative pathway may be dependent on reactive oxygen species and produce DHEA from an unknown “precursor” molecule other than pregnenolone [120]. Although this study provided interesting data, there is a limitation to the scientific rationale, as the authors admitted that the P450c17 inhibitor used could not block endogenous DHEA production in the glial cells. At least one other lab has also speculated on the possibility for an alternative pathway of DHEA production linked to free radicals [152]. This is not impossible, since a precedent in this regard exists: an alternative steroid production mechanisms in the brain was clearly demonstrated in the case of 21-hydroxylation, which appears to be mediated through CYP2D instead of CYPC21 [153].

Apart from the debate on the capacity for intracrine DHEA synthesis in the CNS, several theories have been formulated in attempts to explain the relatively higher DHEA and DHEAS concentrations in the central compartment. For example, it has been suggested that DHEA in the brain may be produced from DHEAS delivered from circulation, through conversion by steroid sulphatase (STS). Indeed, it has been shown that the human temporal lobe expresses high amounts of STS mRNA as well as activity in the cerebral cortex relative to subcortical white matter [145]. However, there are two arguments against this hypothesis. Firstly, sulphated hydroxysteroids such as DHEAS are hydrophilic and do not readily cross the blood–brain barrier [154]. Although sulphated steroids such as DHEAS and pregnenolone sulphate may enter the brain through circulation via organic anion transporting peptides (OATP) transporting them both ways, it would appear that OATP transport of sulphated steroids favours movement out of the CNS, not in [155]. However, a recent study investigating the transport of DHEAS and PREGS through the blood brain barrier (BBB) of rats found that although PREGS entered the blood brain barrier more readily, DHEA was rapidly metabolised into androstenedione and androstenediol [156]. Further investigation found that the enzymes responsible for the rapid desulphatasion were mainly localised the BBB capillaries as opposed the brain parenchyma [156]. This argues that the DHEA in brain is likely from peripheral sources as opposed to de novo synthesis. However, it is important to note that the overall efflux of DHEAS through the BBB may still not be rapid enough to account for the differences in concentration in the brain versus the periphery.

Secondly, as already mentioned, DHEA and DHEAS concentrations have been reported to be markedly higher in the human brain than in circulation [147]. In this study, DHEA concentrations were determined in the brains of 10 subjects (9 females and 1 male, ages ranging from 80 to 93). Nevertheless, in rat brains DHEA concentrations were higher relative to circulation and remained so even 15 days after an adrenalectomy and/or orchiectomy [146]. This argues against the idea of systemic delivery. However, as mentioned earlier the techniques that were used to determine the concentration of DHEA in these studies may reassessed due to possible cholesterol contamination.

Taken together there still exists a strong body of evidence to support that de novo occurs in the brain. Although it is near impossible to prove this specifically in humans, as one will be faced with clear practical and ethical hindrances, it is the opinion of the authors that the human brain does indeed produce DHEA centrally. In order to understand how DHEA may be produced in the brain, the use of animal models may be the only way to elucidate key pathways. This, however, does not come without its challenges.

Several practical challenges have already been alluded to in this review, such as methodological issues pertaining to mRNA and protein determination, as well as gender-specificity. A remaining consideration which will be touched on in this section, is the importance of species selection for investigations of this nature.

For example, it is important to consider that systemic steroidogenesis in humans, higher primates and bovines differ from rodents in two significant ways. Firstly, humans primarily convert PREG to androstenedione through the Δ⁵ pathway in the following manner: PREG » 17OH-PREG » DHEA » androstenedione. Rodent steroidogenesis on the other hand favours the Δ⁴ pathway; PREG » progesterone » 17OH-progesterone (17OH-PROG) » androstenedione (Fig. 2). This is speculated to be the reason why humans have high levels of circulating DHEAS, while in rodents levels are significantly lower when compared to humans [157].

Secondly, in humans the adrenal gland is the main source of DHEA production, but in rodents the gonads are the main source, as the adrenal glands of most species mice and rats appear to lack P450c17 enzyme activity [158, 159]. Although human gonads do produce androgens such as DHEA, the reason for adrenal preference of androgen production is not clear. The use of rodent models in studying the systemic role DHEA has thus garnered concern due the significant differences between humans and rodents. However, the fact that DHEAS has been found in both human and rodent brain tissue suggests that potential for de novo DHEA production exists in the brain in both species. Therefore, rodents could still serve as reliable model systems in studying DHEA and its role in the brain.

Of specific interest in this context, it was recently discovered that the precocial species of mouse, Acomys cahirinus or Spiny mouse, possess adrenal glands that express P450c17 and are in fact capable of producing DHEA [160, 161]. This unique mouse has a longer gestational period (± 39 days) when compared to other mice and rats and unlike other rodents, has a brain growth pattern that is comparable with humans at the time of birth [162, 163]. It has also been shown in spiny mice that cortisol is the main glucocorticoid in the blood, unlike in rats where corticosterone is the main circulating glucocorticoid [160, 161]. Most importantly, explanted foetal, neonate and adult spiny mouse brain tissue in culture have been shown capable of producing DHEA in the presence of pregnenolone [164]. In addition, spiny mouse also exhibit an age-related decline in DHEA synthesis [163]. These mice may prove to be a particularly reliable and relevant model for studying the mechanisms of DHEA in many neurodegenerative disease states.