Stroke: Understanding the Differences between Males and Females

Abstract

Stroke is a significant cause of death and long-term disability in the United States. The incidence, mortality and outcomes of stroke are significantly different between men and women. As with many diseases that affect men and women differently, an understanding of the reasons underlying those differences is critical to effective diagnosis and treatment. This review will examine the sex differences in stroke in both humans and animal models of stroke and review what is known about potential mechanisms underlying these differences. It is clear that there is a complex interaction between hormonal, genetic and unknown factors at play in generating the sex differences in stroke.

INTRODUCTION

Stroke is currently the fourth leading cause of death in the United States following heart disease and cancers [25]. Each year approximately seven hundred thousand people suffer a stroke and it causes approximately one out of nineteen deaths in the United States [25]. Overall, women have a lower risk of stroke compared to men, but this disparity diminishes following menopause when they have more strokes and worse outcomes as compared to men [21, 45, 68]. Retrospective studies have shown, however, that post-menopausal women on hormone replacement therapy appear to have better functional outcomes following a stroke [27]. Data from the Women’s Heath Initiative in which either placebo, estrogens or progestins plus progesterone were administered to post-menopausal women in a double-blind trial, however, failed to show a protective effect of current hormone replacement therapies against stroke [76]. This may be due to the timing of hormone replacement as the average age of women in the study was 63, although this has not been conclusively determined [28, 32]. In addition to sex differences observed in humans, there are numerous studies in animal models of ischemic stroke suggesting sex differences in brain damage and recovery [29, 42, 35]. In this manuscript we will discuss some of the most current literature of what may be the source of this sex difference in stroke. A complete understanding of these complex events are not entirely known.

SEX DIFFERENCES IN STROKE IN HUMANS

Someone dies of a stroke in the United States every four minutes. The vast majority of these strokes are ischemic strokes where blood flow to part of the brain is blocked resulting in damage to the brain [25]. The single largest risk factor of stroke is age, however, males are more likely to suffer a stroke as compared to females [59, 74]. Males are more likely to suffer a pediatric stroke and this trend continues through adolescence [26]. Interestingly, in young adults females are more likely than males to experience stroke [52, 56]. This is likely due to the pro-thrombotic actions of oral contraception [8]. After age thirty, however, male predominance returns [56, 67]. Following menopause, however, the protection in females is lost [30]. This is likely due to the increased number of women who live longer life spans than men, but also due to loss of female hormones after menopause.

When women suffer a stroke they are significantly more likely to have what is classified as a “severe” stroke more likely to result in death [20]. The women who survive are also more likely to experience more severe outcomes with less of a chance of complete recovery [21]. Women are more likely to experience long-term handicaps, a reduced quality of life and are more likely to experience depression [4, 57]. The contributing reasons for this likely have to do with age in that they may not have family caregivers and are older at the time of the stroke, in addition to an underlying sex difference in recovery. Overall there are significant sex differences in the incidence, severity and recovery from stroke.

SEX DIFFERENCES IN ANIMAL MODELS OF STROKE

Animal models have been extensively used to begin to understand the mechanisms, outcomes and sex differences underlying stroke. Methods to transiently or permanent block one or more of the cerebral arteries have been used by many investigators to produce a reproducible infarct in the brain. Females consistently experience less brain injury than males in numerous rodent models. Female gerbils suffered less neuronal pathology compared to males after ischemia was induced by unilateral carotid artery occlusion [31]. Similar results were obtained in rats in studies utilizing several models of transient cerebral ischemia [3, 40, 84] where it was demonstrated that both young and middle-aged females with similar levels of circulating estrogen had significantly less infarct volume when compared to gonadally intact males or to ovariectomized female rats whose estrogen levels are virtually undetectable.

Permanent occlusion of the middle cerebral artery (MCAO) is a well-established model of stroke. In this model, there is a 50% reduction in cerebral blood flow to the striatum and overlaying cortex [11]. This decrease in blood supply causes necrotic cell death in the striatum followed by programmed cell death, or apoptosis, in the overlaying cortex. In this model of permanent occlusion, ovariectomized females and intact males had a much larger MCAO-induced injury than animals with higher circulating estradiol concentrations [11, 60, 63, 73]. Pretreatment of these animals with low doses of 17β-estradiol is sufficient to exert dramatic protection in the brains of both female [80, 81] and male rats [51, 73], suggesting that indeed the female sex steroid hormone, estradiol, at least in part contributes to the sex difference in stroke vulnerability.

ROLE OF STEROID HORMONES

Estrogen

Many sex differences are mediated by the presence of the sex-specific steroid hormones; estrogens and progesterone, in females and androgens, including testosterone in males. Studies have shown that even low physiological doses of estradiol replacement are sufficient to exert dramatic protection in the brains of young and middle-aged female [13] and male [73] rats. Treatment with lower physiological doses of estradiol must occur prior to the onset of injury [11] while higher doses of estradiol can still be protective when given immediately before injury or even after injury [44, 73]. There are likely numerous mechanisms by which estrogen can prevent cell death following stroke, some requiring the action of the classical estrogen receptor and others that do not.

Steroid hormone receptors classically mediate the actions of steroid hormones by acting as transcription factors regulating target genes. This estrogen receptor function is necessary for the prevention of secondary apoptosis observed in cortex following focal ischemia [12]. Protection by estradiol is dependent on the presence of estrogen receptor alpha (ERα) specifically in the cortex. In ERα knockout mice, estradiol is not neuroprotective [14]. In the uninjured adult cerebral cortex, ERα is virtually absent as it is only transiently expressed during neonatal development [47, 55]. Twenty-four hours after MCAO, both ERα mRNA and protein are increased in the cortex of rats and mice [12, 15]. In ovariectomized females, this increase in ERα expression occurs in both oil and estradiol-treated groups [15]. One potential mechanism of this re-expression is thought to be due to the release of repressive DNA methylation [77]. DNA methylation of the promoter regions of the ERα gene is inversely correlated with expression in a time course that matches the re-expression following injury.

Estradiol also appears to enhance cellular viability by regulating the expression of anti-apoptotic gene, bcl-2 [71]. In the hypothalamus, Bcl-2 mRNA was elevated in ovariectomized, estradiol-treated rats [22]. Additionally, estradiol prevented the injury-induced decrease in Bcl-2 gene expression in the cerebral cortex after ischemic injury [12]. Similarly, in hippocampal cells, estradiol increased expression of Bcl-xL, a Bcl-2 family member that interacts with Bcl-2 and Bcl-xL to prevent the pro-apoptotic actions of Bax, inhibit free radical production and suppress the activation of cysteine proteases [53]. Therefore, the regulation of this family of genes by estradiol may have multiple downstream effects that together suppress apoptosis and favor cell survival. In an in vitro model of ischemia it was demonstrated that pharmacologically blocking estradiol action via its receptors did indeed, reduce cell death and specifically, apoptosis [78, 79].

While the inflammatory response following ischemia is very complex, in general estradiol is believed to have an anti-inflammatory effect in the brain [58]. Estradiol can prevent the inflammatory response in cerebral blood vessels in young rats, however, this effect appears to be lost in older animals [69]. Additionally, estradiol has also been shown to regulate plasma levels of IL-6, TNF-α, granulocyte-macrophage colony-stimulating factor, IL-4 and IL-5 leading to a reduction in cell death in females following MCAO [70]. Together, these studies suggest that estradiol has the potential to regulate multiple aspects of inflammation that could result in less neuronal cell death following a stroke.

Estradiol may also exert trophic and protective effects by acting via classical receptor-mediated mechanisms on a variety of genes including the neurotrophins and their receptors [66]. Ovariectomy reduces brain derived neurotrophic factor (BDNF) mRNA levels in the cortex and hippocampus of female rats. In addition, estradiol may influence neurotrophin receptors, (trk A, trkB, trkC) and/or the pan-neurotrophin receptor, p75NTR [48]. During development estradiol receptor mRNA colocalizes with NGF, BDNF and NT3 in subsets of cells in the cortex and hippocampus and with the receptors for these neurotropins, p75NTR, trkA and trkB in the basal forebrain [48, 66]. The colocalization of estradiol receptors, neurotrophins, and their cognate receptors suggests potential complex autocrine and paracrine interactions between estradiol and the neurotrophins [65]. Estradiol may exert trophic and protective effects by influencing the expression of genes that encode for survival factors in the brain.

In addition to the receptor-mediated genomic actions of estradiol, several studies have suggested receptor-independent actions of estradiol as well. In female rats, pharmacological doses of estradiol can protect against the injury induced by transient cerebral ischemia [63]. Pretreatment or acute treatment with either estrogen isomer, 17β-estradiol or the receptor inactive 17α-estradiol, achieved equivalent neuroprotection, suggesting a mechanism of protection independent of transcriptional activation [44]. This may be due to the antioxidative properties of the estrogen molecule or effects at the mitochondria that prevent cytochrome C release following injury [41, 64]. Additionally, acute treatment of male rats with pharmacological levels that are thousands of times higher than physiological levels and pretreatment with physiological levels of estrogen protect against injury induced by transient cerebral ischemia [73]. Together, these findings suggest that the sex-related differences in the extent of brain injury and the protective effects of estrogen replacement may be mediated via multiple cellular and molecular mechanisms.

Progesterone

Much less is known about how the other female sex steroid hormone, progesterone, acts as a potential neuroprotective agent in the brain. Several studies have suggested it can have a neuroprotective role in stroke [43]. Administration of progesterone can reduce cortical infarct size in middle-aged, reproductively senescent females and this protection is not related to the ability of progesterone to regulated cerebral blood flow [6]. Also administration of progesterone during the reperfusion in a model of transient ischemia in ovariectomized young females is neuroprotective [50]. Progesterone administration after a stroke has also been shown to reduce infarct size and enhance functional recovery in males [24]. While the pontential protective mechanisms are not as well known as those of estradiol, progesterone may act as an antioxidant and has been shown to modulate gamma-aminobutryic acid (GABA) receptor channel activity expression in neurons [16]. It can also on stimulate PI3 kinase activity leading to an increase in vascular endothelial growth factor and BDNF [36]. The overall effect of progesterone in stroke, however, is not entirely clear as some studies have demonstrated that progesterone administration can have either no effect or actually increase infarct size in certain models of stroke [10, 49]. Finally, the estrogen plus progesterone arm of the Women’s Health Initiative suggests that combined hormone therapy may slightly increase the incidence of stroke [46].

Testosterone

The male sex is a known risk factor for stroke in humans. There are relatively few studies using this animal model testing the effects of androgens on ischemia and of those there are contradictory findings. For example in male rats, castration and the resulting low testosterone has been shown to either have no effect or decrease infarct size following MCAO [33, 73]. Testosterone administration to intact male rats has also been shown to increase infarct size, potentially due to the ability of testosterone to exacerbate glutamate excitotoxicity [83]. On the other hand, in humans, testosterone levels are inversely associated with stroke severity, infarct size and improved functional recovery [37]. Additionally, several in vitro studies suggest that testosterone in protective against oxidative stress, beta-amyloid toxicity and serum deprivation [1, 2, 54]. Testosterone and its metabolite, dihydrotesosterone, can protect against ischemia in adult male rats when administered in the low, physiological range and this protection was mediated through androgen receptors [19], however these effects are dose-specific. Higher levels of androgen replacement resulted in increased infarct volumes [75]. It is possible that these divergent effects can also be due to the activation of membrane-associated versus intracellular androgen receptors as the membrane-associated receptor activation can enhance cell death in an in vitro model while activation of the traditional intracellular receptor leads to neuroprotection [23]. These data suggest that there is a complex effect of testosterone on neuronal survival after stroke, where dose, timing and the type of androgen all influence its neuronal effects. Clearly, more needs to be done to decipher the overall effects and potential mechanisms of action of androgens.

SEX HORMONE-INDEPENDENT MECHANISMS

In addition to the hormone-regulated events described above, several aspects of the sex differences observed following stroke appear to be independent of direct hormone action. There are clear differences in body size and vascular anatomy that are associated with an increased risk of stroke in males [72]. Increasingly the contribution of the sex chromosomal makeup of the organism is being recognized as a potential source of the sex difference. Cultured neurons from either male or female rats are differentially susceptible to an in vitro model of ischemia where male neurons are more vulnerable to cell death than female neurons cultured under identical conditions [18]. It was subsequently demonstrated that, neurons from males expressed higher levels of soluble epoxide hydrolase that contributes to neurotoxicity. Additionally it is possible that X or Y chromosome-linked genes may contribute to the inherent sex difference in response to stroke. Genes on the Y-chromosome are involved in regulating hypertension, which is a risk factor associated with the increased stroke risk in men [9]. Furthermore, polymorphisms of this gene are associated with increased hypertension. In females, an X-chromosome linked gene, XIAP, has been identified that inhibits apoptosis [34]. XIAP has recently been shown to be specifically regulated by a microRNA in females following stroke [62]. Disruption of this expression in females resulted in ischemic cell death comparable to that in males.

Another hormone-independent mechanism that appears to be in play in the brain is the epigenetic regulation of genes involved in neuroprotection or apoptosis. Epigenetic regulation involves post-replication and post-translational modification of DNA and histones that lead to lasting changes in chromatin structure and thus changes in gene transcription (for review see [82] and [39]). DNA methylation is central to epigenetic control of gene expression. Methylation typically occurs at repeats of cytosine-phosphate-guanine sites primarily in the promoters of genes. DNA methyltransferase (DNMT) transfers a methyl group to the cytosine resulting in 5-methylcytosine [61]. The methylated region of DNA is then bound by DNA methyl binding proteins which can then prevent transcription machinery from binding, resulting in gene silencing in many cases [7, 39]. Increased DNA methyltransferase activity is associated with stroke and inhibiting DNA methyltransferase activity confers neuroprotection following MCAO [17]. In the hippocampus, expression of the methyl DNA binding proteins is altered in a spatial and temporal fashion following a transient ischemia [38]. We have demonstrated that the promoter of the ERα gene is progressively methylated during development and is essentially turned off in the adult rodent cortex [55]. Following MCAO the ERα gene is demethylated and subsequently re-expressed [15, 77]. Interestingly, this demethylation occurs only in female rats and is independent of estradiol treatment in the females[77]. MCAO also regulates the expression of the DNA methyltransferase that regulates the maintenance of DNA methylation, DNMT 1, in a similar sex-specific manner. Sex differences in DNMT gene expression have also been observed in the developing hypothalamus [5]. Together this data suggest that epigenetic modification of genes may play a role in regulating genes involved in neuronal survival and cell death following brain injury in a sex specific manner.

CONCLUSIONS AND FUTURE DIRECTIONS

Many factors impact the incidence and outcomes of individuals that experience an ischemic stroke. It is clear that age, sex and hormone exposure all contribute to the responses to stroke. We have reviewed several potential mechanisms underlying these contributions. Further investigation is needed to come to a more complete understanding of this complicated relationship. It is also increasingly clear that the age and sex of the animal models needs to be taken into consideration very carefully when drawing conclusions about causes and mechanisms of stroke. A deeper understanding of these mechanisms will eventually result in the design of better prevention, treatment and therapeutic paradigms when particularly applied to aging men and women.