Sexual Differentiation and Substance Use: A Mini-Review

Abstract

The organizational/activational hypothesis suggests that gonadal steroid hormones like testosterone (T) and estradiol (E2) are important at 2 different times during the lifespan when they perform 2 different functions. First steroids “organize” brain structures early in life and during puberty, and in adults these same hormones “activate” sexually dimorphic behaviors. This hypothesis has been tested and proven valid for a large number of behaviors (learning, memory, social, and sexual behaviors). Sex differences in drug addiction are well established both for humans and animal models. Previous research in this field has focused primarily on cocaine self-administration by rats. Traditionally, observed sex differences have been explained by the sex-specific concentrations of gonadal hormones present at the time of the drug-related behavior. Studies with gonadectomized rodents establishes an activational role for E2 that facilitates vulnerability in females, and when E2 is combined with progesterone, addiction is attenuated. Literature on organizational actions of steroids is sparse but predicts that T, after it is aromatized to E2, changes aspects of the neural reward system. Here we summarize these data and propose that sex chromosome complement also plays a role in determining sex-specific drug-taking behavior. Future research is needed to disentangle the effects of hormones and sex chromosome complement, and we propose the four core genotype mouse model as an effective tool for answering these questions.

Our purpose in writing this review is 3-fold. First, we will present an overview of the data on sex differences in drug use in humans and animal models, with a focus on cocaine, because it is the best-studied illicit drug. These sex differences in adults are largely attributed to activational effects of gonadal hormones though surprisingly few studies have considered the possibility that organizational effects of gonadal differences may also contribute to sex differences in drug use and addiction. Thus, as our second purpose, we turn to how and when steroid hormones may affect the developmental organization of adult responses to drugs. We were surprised by the paucity of data on this topic, especially because many key aspects of adult sex differences are programmed in early development. Finally, we introduce and discuss a nonhormonal factor that we know influences sex differences in motivational behaviors both during brain development and adulthood: sex chromosome complement. We speculate that sex chromosome genes have direct effects on addiction-related behavior, and may also affect these behaviors through interactions with genes related to hormones (eg, receptors, steroid synthesis enzymes). One caveat to assert is that most of the behavioral work on addiction has been conducted in rats, but in this review, we also include a fair amount of mouse data. We acknowledge here that rats and mice are different in many respects, but we hope the data will generalize across species, not only from rats to mice and vice versa, but also will inform us about humans. Our goal is to stimulate interest in this topic because we feel a neuroendocrine perspective on these questions will lead to additional insights.

Sex Differences in Humans: Patterns of Cocaine Use

According to the 2017 National Survey on Drug Use and Health, more than 40 million Americans age 18 years and older had taken cocaine at some point in their lives, and nearly 25% of these had used crack cocaine. About 100 000 individuals between ages 12 and 17 years began taking cocaine in 2017, which represents a surge in new/young users. Proportionally, slightly more men than women have used stimulants like cocaine both in the past year and over their lifetime (1). At all ages, women are at higher risk than men because they develop substance use disorders more rapidly after initial use and report shorter periods of abstinence during recovery than men (2, 3). Women also report a shorter period between initial drug use and entry into treatment for a substance use disorder (4). Another concern, given the opioid epidemic, is that among patients treated for long-term opioid use disorder, more women than men concurrently used crack cocaine (5). The coadministration of cocaine with opiates may result in unique subjective effects (6), which could lead to this combination of drugs being more addictive than either drug alone.

In women, the subjective effects of stimulants vary across the menstrual cycle (7), with the greatest effects observed when estradiol (E2) levels are high and relatively unopposed by progesterone (8). Additionally, exogenously administered E2 enhances the subjective effects of psychostimulants like cocaine in women (9), whereas exogenously administered progesterone, and its metabolite allopregnanolone, attenuates the subjective effects of cocaine and other drugs of abuse both in men and women (10). These data and others have been examined experimentally in animal models.

Sex Differences and Hormonal Contributions in Animal Models

Many sex differences in behavior are regulated by circulating levels of gonadal steroids (11, 12). Differences between the sexes in levels of testosterone (T), E2, and/or progesterone lead to the “activation” of some behaviors in a sex-specific manner or to differences in intensity of behavioral display between males and females. Like humans, female rats acquire cocaine self-administration (SA) faster, at lower doses, and are more motivated to use cocaine than males (13-15). Findings in humans and animals suggest that ovarian hormones, and in particular E2, modulate the enhanced sensitivity in females (16, 17).

Animal studies have also revealed a link between the ratio of E2 to progesterone and initial vulnerability to cocaine, with results generally showing that when levels of E2 are high and progesterone is low (ie, during adolescence and during proestrus; [18,19]) cocaine’s reinforcing effects are heightened (20, 21). Acquisition of cocaine self-administration is markedly reduced by ovariectomy (OVX) and restored by E2 replacement. OVX rats given E2 SA cocaine at more than double the rate of the control OVX rats (22, 23). Although the effects of progesterone alone in OVX rats are controversial (24), when coadministered with E2, progesterone attenuated acquisition of cocaine SA (22) as compared with E2 only. Progesterone treatment in intact females has also been shown to attenuate cocaine SA.

Based on these data and others, the major hypothesis for sex differences in drug-related behaviors is that adult differences in circulating E2 and progesterone are critically responsible. Data on differences in SA or conditioned place preference (CPP) behavior in female rats over the 4- to 5-day estrous cycle (25, 26) support this contention. E2 has well-documented actions on brain functioning, including effects on dopamine (DA) production and signaling in the reward pathway (27). Electrophysiology studies with female mice have shown increased firing of DA neurons in the ventral tegmental area (VTA) for females in estrus/proestrus, when E2 levels are highest. This suggests an estrogenic “priming” of the reward circuits in females that leads to an increased response to rewarding stimuli (28). Work with female rats has shown that microinjections of E2 into the medial preoptic area leads to increased DA release in the nucleus accumbens (Acb) after cocaine administration (29). Most studies find that even after gonadectomy, adult female rats acquire cocaine SA more quickly and SA more cocaine than their castrated counterparts (21, 30). In one study, gonadectomy reduced motivation for cocaine (measured in a progressive ratio [PR] test) in females more so than in males (31). When E2 is given to gonadectomized adults, acquisition is enhanced in females (32). E2 administration to castrated male rats heightened preference for cocaine over food in a 2-choice test, even under conditions of food restriction (33). Similar sex and hormonal effects have been reported in mice (34, 35). Gonad-intact female CD1 mice acquire SA sooner and demonstrate more motivation to obtain cocaine in a PR test than males (36). In C57BL/6 mice, cocaine administration leads to CPPs in both sexes, but preferences are greater in females and vary with estrous cycle stage, with females conditioned during estrus or proestrus showing greater CPPs than males and females tested during nonestrus phases (28). Rhesus monkeys tested for acquisition and cocaine intake show higher intake in follicular stage females as compared with luteal-phase females and males (37). Female cynomolgus and rhesus monkeys are more motivated than males to obtain cocaine in a PR paradigm; yet, menstrual cycle stage does not correlate with PR responding (38, 39). Taken together, these findings support the E2 hypothesis. However, because sex differences persist after gonadectomy (but are reduced in both sexes) and E2 fails to elevate cocaine-related behaviors in males, adult differences in responses to E2 are not the sole reason for the sex differences.

Sexual Differentiation and Critical Periods

The “organization/activation” hypothesis first demonstrated experimentally by Phoenix and colleagues (40) (and tested hundreds of times since) described a bimodal role for gonadal steroid action on sexually dimorphic behaviors (41, 42). Not only are adult levels of steroids important, but sex differences in behavior are organized by exposure to hormones during critical periods of neural development (43, 44). The majority of the work on this topic examined the neonatal period just after gonadal differentiation. At this time male (XY) individuals differentiate testes, which produce T. T can act as an androgen on androgen receptors and on estrogen receptors after aromatization to E2 (45, 46). Decades of data show that for most sexually differentiated behaviors developmental neonatal exposure to estrogens, which are produced in situ by aromatization of androgens, are critical (47).

Surprisingly, the neonatal critical period has received little attention from the drug-addiction field. In rats prior to and after puberty, a sex difference in excitatory postsynaptic current recorded is present in the striatal region of the nucleus accumbens core (AcbC) (48). The AcbC is critical for motivation and performance of rewarded behaviors (49). Several forebrain areas, including the AcbC, receive dopaminergic input from the VTA (50). Rewards such as sex, cocaine, and other drugs of abuse stimulate DA release from the Acb (51, 52). An important neuronal population within the AcbC is the medium spiny neurons, which receive dopaminergic and excitatory glutamatergic inputs and project out of the region (53). To test the hypothesis that neonatal E2 or T would masculinize females, pups were injected for the first 3 days after birth with vehicle, T, or E2. Rats were harvested for slice recording prior to puberty, and the sex differences in frequency of miniature excitatory synaptic current (mEPSCs; female > male) were eliminated by E2 or T treatments to female pups. This shows that a key brain region for addiction is sexually differentiated by neonatal hormone exposure. These same neurons show dramatic differences across the estrous cycle in adult females, and their changing properties correlate with circulating levels of E2 and progesterone (48, 54). In a related study, female rat pups were given a single injection of E2 valerate the day after birth and adult AcbC medium spiny neurons were examined (55). Adult females that were treated as neonates with E2 had elevated heightened excitatory neurotransmission measured by frequencies of mEPSCs and greater DA release from the AcbC than control females. It is possible that these females did not experience a normal puberty and were not cycling (56, 57). In cycling rodents, progesterone in combination with E2, and given alone, influences mEPSC responses (58). Thus, the failure to account for cycles or cycle stage (55) is likely responsible for opposite effects in the 2 studies. Both levels of tyrosine hydroxylase messenger RNA (mRNA) and intensity of immunocytochemical stain were enhanced in neonatally E2-treated females as compared with controls both in the VTA and the substantia nigra (55).

The only behavioral study we know of related to neonatal hormones and drug-related responses examined reactions to morphine in both sexes of rats (59). A single injection of E2 valerate, T, or vehicle was given to pups the day after birth. Rats were tested in adulthood for DA release in the Acb, CPP, or locomotor activity in response to morphine. Controls did not exhibit a sex difference in DA release, but for both sexes the early E2 treatment enhanced DA release, as noted previously (55), whereas T had an intermediate effect that was not significantly different from controls or E2 treatment. Morphine produced conditioned preferences in both sexes and all treatment groups. In the activity assay, morphine produced a sex difference in controls (females > males) and enhanced spontaneous activity in both sexes treated with neonatal E2 or T. To our knowledge the impact of neonatal hormones on sex differences on adult drug SA has not been tested, but would be extremely interesting and likely of high utility for explaining sex differences in drug addiction.

A second period of brain organization is puberty, during which both sexes experience gradually elevating levels of circulating gonadal steroids (60). Tests of the organizing effects of T and E2 in both sexes show that pubertal steroids affect a number of social behaviors (61, 62). The expression of sex differences in sensitivity to drugs of abuse may depend on ovarian hormone exposure during adolescence, because such exposure appears to be critical for feminizing the DA reward pathway (63). For example, prepubertal female rats acquire drug SA faster and are more motivated to obtain infusions of drug as compared to their male counterparts. Levels of motivation to obtain infusions of cocaine during adolescence, as measured by responses to an escalating PR schedule, are positively associated with serum E2 levels (21). In cocaine CPP tests, prepubertal males acquired a cocaine CPP at lower cocaine doses than postpubertal males (64), suggesting that pubertal hormones decrease sensitivity to cocaine in adult males. Conversely, E2 treatment during adulthood does not affect cocaine SA in males regardless of their gonadal status during adolescence (31). The ability of E2 to enhance motivation for cocaine in adult females also depends on adolescent E2 exposure because treatment during adulthood enhances motivation for cocaine in females, but only if they were gonadally intact during puberty. Interestingly, E2 treatment during adulthood increases rates of acquisition of cocaine SA in females regardless of gonadal status during puberty, indicating that some of the effects of E2 are mediated independently of developmental effects. Motivation to obtain cocaine in PR tests was reduced in females given E2 for 10 days during puberty even when they received additional E2 as adults. Conclusions are hard to assess because the hormone treatment period was only 10 days (half the duration of puberty). During normal puberty, hormone levels increase gradually, whereas in this study gonads were removed at the time of weaning, and E2 exposure started via daily injection 6 days later. It is difficult to mimic puberty using controlled hormone injections, and although we note that this is an important study, it should be replicated and extended using other hormone exposure paradigms.

Sex Chromosomes and Sex Differences: Data From the Four Core Genotype Mice

Another factor that may contribute to sex differences in addiction is sex chromosome complement (17, 65). The four core genotype (FCG) mouse model separates the actions of gonadal hormones from the potential effects of sex chromosome genes (65). These mice, along with other mouse models, have been successfully used in many studies examining sex differences in factors including stroke, immunity, obesity, cardiovascular function, and social behaviors, among others (66-71). The FCG model was developed after the serendipitous finding (in the Lovell-Badge Laboratory at the Medical Research Council) of an XY female with a spontaneous mutation in the testes determining gene (Sry) located on the Y chromosome. The mutant XY animals developed ovaries, not testes, and reproductive tracts were phenotypically female (72, 73). A transgene for Sry inserted randomly into chromosome 3 rescued Sry function, restoring the full fertile male phenotype (74). The XY^–Sry (“ ^–” designates the mutation of Sry on Y and “Sry” designates the transgene) males are the sires for the cross; the dams are normal XX females. The offspring are equally distributed among the FCGs: XX females, XY females, XX males, and XY males. Behavioral comparisons between XX females and XY females would reveal the effects of the sex chromosomes, whereas comparisons between XX females and XX males would show the effects of the gonadal hormones.

Several studies have employed FCG mice to examine the roles of sex chromosome complement on rewarding behaviors. Quinn and colleagues (75) restricted food for FCG mice and trained them on an instrumental task to obtain food. The XX mice (with and without gonads) acquired this task faster than XY animals. Next, the task was devalued by pairing food with lithium chloride (making the mice feel sick). The XY mice stopped working for food after the devaluation, but the XX mice did not do so, showing that they had acquired this habit. More recently, gonadectomized FCG mice were tested for responses to obtain a food reward (sweetened condensed milk) (76). Under ad lib feeding conditions, males consumed more sweetened condensed milk than females, and XY mice consumed more than XX mice. XY mice were also more motivated than XX when a PR schedule was employed. A similar outcome was found using alcohol as the reward both in gonad-intact and gonadectomized mice (77). Clearly sex chromosome complement can influence behaviors, but the direction of the effects vary with the task.

We recently asked whether sex chromosome complement might affect SA for cocaine in the FCG (78). Briefly, SA was examined by fitting mice with indwelling jugular catheters that allow them to obtain cocaine under a fixed ratio 1 schedule using an escalating-dose acquisition procedure. The protocol began with a low dose followed by moderate and moderately high doses. Each dose was available for 4 consecutive daily sessions. Males with XY chromosomes had the lowest acquisition (42%) rate, and the XYF group (93%) had the highest acquisition rate. This difference is likely facilitated by differences in circulating E2 because the mice were gonad-intact. An effect of sex chromosome was also seen in the onset of acquisition. When all mice were considered, the XX mice acquired SA faster than the XY mice. Data on rewarded and unrewarded nose pokes indicated that XX mice learn to distinguish rewarded nose-poke holes sooner than XY mice (Fig. 1). Our interpretation of these data is that E2 and sex chromosome complement interact whereby possession of either 2 X chromosomes, or female-levels of E2, promotes vulnerability to cocaine. Moving forward we are now testing FCG mice that have been gonadectomized after puberty. Although using gonad-intact animals might be of more translational value, use of gonadectomized FCG mice is essential to separate the effects of activational hormones and sex chromosome complement. We expect to find more effects of sex chromosomes on cocaine-related behavior in our upcoming experiments.

Figure 1.

Use of four core genotype mice illustrates interactions between estradiol (ovaries) and sex chromosome complement. A shows the percentage of mice that attain the acquisition criteria. *Significant effect of gonadal sex. Ns per group XXF = 15, XYF = 14, XXM = 15, XYM = 19. B shows the mean day (± SEM) in the 12-day acquisition test when mice consistently discriminated between the rewarded and unrewarded nose-pose holes. Here we find an interaction between gonadal and chromosomal sex. **Significantly different from all other groups. These data suggest that either the presence of ovaries (and thus estradiol) or an XX genotype increases vulnerability to cocaine. Data redrawn from Martini et al (78).

Several studies using FCG brains have reported sex chromosome effects on gene expression. In pubertal prefrontal cortex and cerebellum, a sex difference (XX > XY) in calbindin-containing neurons is caused by sex chromosome complement and the α estrogen receptor (79). In adult striatum, expression of 2 drug-addiction related genes was higher in XX than XY mice: DA receptor D2 (Drd2), prodynorphin (Pdyn), and preprotachykinin (Tac1) (80). In chronically stressed adult FCG mice, expression of γ-aminobutyric acid (GABA)-DA– and serotonin-related genes in the prefrontal cortex were influenced by sex chromosome complement, in all cases XX > XY (81, 82). Notable among those genes were Ntrk2 (brain-derived neurotrophic factor receptor), Slc32a1 (GABA transporter), Htr1a and Htr2c (serotonin receptors), and Drd1 (DA receptor D1). It is well documented that stress accelerates drug addiction, but a role for factors other than adrenal and other stress hormones has not been investigated (83, 84).

Sex chromosome complement also affects prenatal neural development. Neurons in the embryonic midbrain, a region that in adults becomes the VTA and substantia nigra, contain dopaminergic neurons (85). Male embryos have more dopaminergic neurons than females, and female neurons project to the forebrain earlier in development than do males. Notably, these differences appear before gonadal differentiation, and they persisted in cultures after treatment with steroid hormones (86, 87). Embryonic FCG midbrains were probed and a sex chromosome effect was noted whereby XY individuals had more DA neurons than present in XY (85). Aromatase enzyme produces locally synthetized estrogens and is present in several brain areas, and a sex chromosome complement effect is present in the stria terminalis and anterior amygdaloid area. The same direction of sex chromosome difference (XY > XX) has been observed for aromatase enzyme mRNA, activity, and protein in midgestational (E16) FCG embryos (88). Primary cultures from E15 amygdala treated with either E2 or dihydrotestosterone elevated aromatase mRNA, but only in XX brains. This reveals a direct interaction between hormones and sex chromosome complement. This early increase in E2 synthesis may affect neural development in males selectively. In sum, these data suggest that sex chromosome complement affects prenatal exposure to DA and estrogen prior to gonadal hormones produced by the developing fetal testes.

Conclusions and Future Directions

The organizational/activation hypothesis has been tested for a large number of behaviors (learning, memory, social, and sexual behaviors, etc [89-91]). Understanding the roles that sex hormones and sex chromosome complement play in the development of addiction is essential for explaining the sex differences that exist in vulnerabilities and patterns of drug abuse. For drug-related behaviors, there is a solid amount of information on the activational role of E2 on cocaine SA in female rats and a few studies in mice (23, 30, 34). Puberty has been considered as a second organizational period, and there are studies showing organization of dopaminergic systems during this time (92, 93). Moreover, some of the essential neural circuits (DA reward system) are sculpted by neonatal hormones and genes. However, we note a large gap in the literature, with few behavioral studies testing the actions of neonatal hormones on sex differences in vulnerability to illegal drugs, and only one behavioral study that manipulated a single hormone during puberty (31). In addition to gonadal hormones, sex chromosomes may organize behavior during the neonatal and/or pubertal periods. Our recent study (78) shows some sex chromosome effects even in gonad-intact mice. We propose that both of these factors play a role. Prior to gonadal differentiation and secretion of androgens by the testes, individuals with an XY genotype produce more DA in the midbrain and more aromatase enzyme that presumably is enhancing estrogenic action in the anterior amygdala and stria terminalis compared with XX fetuses (94, 95). After gonadal differentiation, androgens (after aromatization to estrogens) augment this process in males. We have outlined this hypothesis in Fig. 2. These 2 neonatal factors produce neural differences in such a way that female drug use is activated by adult E2, whereas males are unaffected by this hormone. This hypothesis predicts that neonatal treatment of females with E2 or testosterone will produce less vulnerability to drugs of abuse in adulthood. Moreover, in these females, E2 will not facilitate acquisition. On the other hand, removal of the gonads prior to puberty in the FCG mice should reveal enhanced vulnerability in XX individuals of both gonadal sexes. Future studies are required to prove or disprove these predictions.

Figure 2.

A hypothetical model in which both hormones and sex chromosomes have roles in sexual differentiation of the motivation/reward circuits that dictate drug vulnerability over the lifespan. We hypothesize that prior to gonadal differentiation, sex chromosomes shape dopaminergic and estrogenic circuits in the male brain. After gonads are formed, both factors work to modify neural development. The influence of hormones in males declines over the lifespan. The reverse is true for females. Directly prior to puberty neither factor is dominant but as estradiol levels increase in females over puberty hormones become important for neural responses to drugs. The area shaded in blue is the time we predict only sex chromosomes are relevant; the period represented in yellow shading is when interactions prevail.

Search Strategies

We analyzed on topics related to this and searched the relevant cited literature. We also searched PubMed and Google Scholar databases using combinations of the terms “addiction,” “cocaine,” “sex differences,” “human,” “rat,” “primate,” and “mice.”

Glossary

Abbreviations

Acb

nucleus accumbens

AcbC

nucleus accumbens core

CPP

conditioned place preference

DA

dopamine

E2

estradiol

FCG

four core genotype

mEPSC

miniature excitatory synaptic current

mRNA

messenger RNA

OVX

ovariectomy

PR

progressive ratio

SA

self-administration

T

testosterone

VTA

ventral tegmental area

Additional Information

Disclosure Summary:  The authors have nothing to disclose.

Data Availability:  Data sharing is not applicable to this article because no data sets were generated or analyzed during the present study.