From 2017... Sexual Differentiation of the Brain: A Fresh Look at Mode, Mechanisms, and Meaning

Margaret M. McCarthy, Geert J. De Vries, Nancy G. Forger, 5.01 - Sexual Differentiation of the Brain: A Fresh Look at Mode, Mechanisms, and Meaning, Editor(s): Donald W. Pfaff, Marian Joëls, Hormones, Brain and Behavior (Third Edition), Academic Press, 2017, Pages 3-32, https://doi.org/10.1016/B978-0-12-803592-4.00091-2

Abstract: This is an exciting time to study sex differences in the brain. Fifty-plus years of building on the foundations established by the organizational/activational hypothesis proposed by Phoenix and colleagues to explain steroid hormone action on the brain has provided an increasingly complex and nuanced view of how the brain develops differently in males and females. In this chapter we first discuss the things we know; there are sex differences in physiology and behavior, in susceptibility to diseases of the nervous system including mental health disorders, and in neuroanatomical and neurochemical measures. These sex differences depend on androgens, estrogens, and sometimes sex chromosomal complement (XX vs XY) acting during development as well as in adulthood, and yet the manifestation of these sex differences may be context dependent. There are four key cellular processes that could potentially underlie sexual differentiation of the brain: cell birth, cell death, cell migration, and cell differentiation, and we discuss the evidence for each in detail. Lastly, we review what we consider major emerging areas and unanswered questions in the field, including the function of sex differences, why they persist, and what they mean.

Keywords: Androgen; Anteroventral periventricular nucleus; Astrocyte; Bed nucleus of the stria terminalis; Bulbocavernosus; Cell death; Dendritic spine; Differentiation; Epigenetic; Estrogen; Hypothalamus; Partner preference; Preoptic area; Sex chromosome; Sex difference; Spinal cord; Vasopressin

5.01.4.5 Is Partner Preference Sexually Dimorphic?

Sexual orientation, also referred to as sexual partner preference,

is defined by the sex of the individuals that are arousing or attractive

to the reference individual, whether it be an individual of the

opposite sex (heterosexual), the same sex (homosexual), or both

sexes (bisexual). The estimated frequency of homosexuality in

humans ranges from 2% to 10%, suggesting that the large

majority of males are sexually oriented toward females and the

majority of females are sexually oriented towardmales. The overwhelming

prevalence of one sex preferring the other is a constant

across all vertebrate species, as would quite naturally be expected

from the point of view of reproductive success. Nonetheless,

what draws the majority of attention is the much less frequent

phenotype of same-sex preference. Notably, the biological basis

of sexual orientation is a matter of impassioned debate only

when it involves discussion of the etiology of homosexuality.

Few seem to question whether opposite-sex orientation is biological.

But actually we understand little about opposite-sex

attraction, and it can be argued that understanding the biological

basis of same-sex orientation would be greatly advanced by

understanding opposite-sex orientation. Thus, a fundamental

question is whether sexual orientation per se is sexually

differentiated.

The answer to this may depend on how you pose the question.

If we state that the majority of males prefer females as

sexual partners and the majority of females prefer males, then

this sounds like a profoundly sexually dimorphic and presumably

differentiated response. Antecedent to this view would be

the assumption that distinct biological processes drive the

neural substrate of partner preference to either a male bias or

a female bias. The existence of distinct processes for a male preference

versus a female preference provides a ready explanation

for why some females prefer other females, some males prefer

males, and why some individuals have no preference.

However, if we take the view that the majority of animals prefer

the opposite sex as partners, then there is no sex difference as

the same drive exists in males and females but it is manifest

differently as a function of one’s own sex. This means that

a component of the neural response is computation of one’s

own sex, which then determines the response to others’ sex.

Given the intensity and early onset of both internal and

external influences of sex on brain development, this is not

outside the realm of possibility. In humans, we are unlikely

to ever be able to definitively separate the impact of nature

from nurture, and our best alternative is the study of naturally

occurring or experimentally manipulated variation in sexual

preference in animals.

The current state of the art of partner preference research is

found on several fronts. These include studies of the programming

effects of gonadal steroids and early experience on partner

preference, the neuroanatomical loci controlling partner preference,

and the study of naturally occurring variation in partner

preference in animal models. Consistent evidence supports

the view that partner preference is organized by gonadal

steroids, such that perinatal androgens, with aromatization to

estrogens in rodents, direct the formation of preference for

a female sexual partner (Brand et al., 1991; Vega Matuszczyk

et al., 1988). In many mammals, odors are the primary signal

indicating sex. Preference can be assessed by determining the

amount of time a test subject prefers to spend with male versus

female stimulus animals or by the amount of time spent investigating

odors generated by stimulus animals. Male- versus

female-specific odors can induce a differential brain response

in the same animal, and likewise, animals of opposite sex

will respond to the same odor differently (Bakker et al.,

1996; Woodley and Baum, 2004). The latter speaks to the

sexual differentiation of partner preference and suggests that

the olfactory system may be the initiation point for subsequent

behavioral responses. In many species, if olfaction is blocked,

there is no partner preference to measure.

Olfaction is important to humans as well, but visual stimuli

are far more potent and the arousal potential of same-sex versus

opposite-sex images depends on the partner preference of the

observer (see for review Baum, 2006). Zebra finches are also

heavily dependent upon vision for expressing partner preference,

and steroids influence partner preference in this species as well

(Adkins-Regan and Leung, 2006). The effect is context dependent,

however, because early experience, i.e., being raised in an

environment with a skewed sex ratio, can also strongly influence

adult partner preference in zebra finches.

The neuroanatomical substrate of partner preference begins

with that portion of the brain detecting and decoding the

sex-specific sensory signals originating from the stimulus

animal, be they olfactory, visual, or auditory. But from there,

all signals appear to converge on the POA, and in particular

an SDN within the POA (see Baum, 2006). An SDN is present

in the POA not only in rats, but also in sheep, gerbils, ferrets,

hamsters, and humans. Lesions of the SDN and its surround

in rats and ferrets either eliminate or reverse sexual preference

(for review, see Baum, 2006). In humans, the third interstitial

nucleus of the hypothalamus (INAH3) may be homologous

to the SDN-POA of rats and is larger in men than women

(Allen et al., 1989). Levay (1991) found that INAH3 is

smaller in homosexual men than in heterosexual men, and

a second study found a mean difference in the same direction

that did not, however, reach statistical significance (Byne

et al., 2001). Thus, the size of INAH3 may be a marker of

partner preference in men, although this conclusion is not

without its detractors.

Another approach is the use of biomarkers to determine if

an individual was exposed to an endocrine environment

in utero that varies from the norm for that sex. These biomarkers

include long bone length, hand digit ratio, and the detection of

small noises made by the inner ear. In general, these studies

support the conclusion that the prenatal hormonal milieu

contributes to the propensity to show same-sex orientation

(Balthazart, 2016). But in many human and animal studies,

a major and unavoidable confound is either the use of surgical

manipulations, such as lesions, or the health status of the

affected individuals, such as the number of HIV-infected

subjects in the homosexual group in human studies. Neither

of these criticisms applies to the study of a naturally occurring

variant of homosexuality, the male-preferring domestic ram.

In at least two different study populations, approximately 8%

of rams prefer to mount other male rams. The frequency of

the phenotype is similar to that observed in humans, and there

are no clear external markers of male-preferring rams. Analyses

of the brain reveal that the SDN of male-preferring rams is

smaller than that of rams that prefer ewes, and it contains fewer

aromatase-expressing neurons. This suggests reduced neuronal

exposure to estradiol developmentally and in adulthood may

be a critical variable in the establishment of same-sex preference

in this species (Roselli et al., 2004).

Thus, on balance, we can conclude that partner preference

is sexually differentiated and that there is an important role for

gonadal steroid exposure in the organization of partner preference,

but early experience may also be important. The primary

detection of the sensory stimulus emanating from a potential

partner is a critical initiating step but the integration and

response to the stimulus appear to be encoded within the

POA. While these are important advances, there remains

much to be learned. Work on the genetics of partner preference

generated a great deal of interest in the early 1990s

(Hammer et al., 1993), but there has been little progress on

that front. There is a continuing interest in the role of birth

order and correlations with handedness, particularly for

male homosexuality, and the proposal of the maternal

immune hypothesis (Blanchard et al., 2006). But again, we

know far less than we should. Moreover, the preponderance

of information is weighted toward understanding male, as

opposed to female partner preferences, although this may be

defensible given the health implications for male versus

female homosexuality. Regardless, progress on both is likely

to remain slow given the paucity of researchers and resources

currently dedicated to this topic.

5.01.4.6 Is the Human Brain Sexual Differentiated?

To some an affirmative answer to this question is self-evident:

how could the human brain not be subject to the same process

that occurs in the majority if not all other mammals as well as

many birds, reptiles, and amphibians? Even invertebrates have

brains that differ in males and females. But others argue, ‘not so

fast, humans are exceptional in many ways.’ We are the only

species with complex computational abilities and a sophisticated

language that includes generation of an historical record.

We also have rich cultural and societal rituals and expectations

that include prescriptions of the appropriate behavior for boys

and girls, men and women. These rules and expectations are

imposed on children even before they are born with the choice

of a gender-typical name and continue with modes of dress and

even the manner in which adults interact with an infant of one

sex versus the other. Thus as we have mentioned many times in

this chapter, it is impossible to parse out the influence of environment

and experience from biology when considering the

brain and behavior of humans. Nonetheless, it is worth a try.

One powerful approach is the study of children at a young

age. While the influence of environment and experience cannot

be eliminated, it is at least lessened compared to that of a fullgrown

adult. Toy choice reflects an interest in different types of

objects, and this varies on average between boys and girls. The

data are messy, with many boys willing to play with girls’ toys

on occasion, and vice versa. One of the strongest influences is

whether a child has a sibling of the opposite sex, demonstrating

the importance both of exposure and of modeling the behavior

of other children. Nonetheless, on average, boys spend more

time with certain types of toys and girls with others. Melissa

Hines has spent most of her career studying this phenomenon

and whether prenatal exposure to androgens in girls can influence

toy choice. She and others have consistently found that

androgen-exposed girls shift their toy preference to that of

boys (reviewed in Hines, 2006). So does this mean there is

a ‘toy’ nucleus in the brain that directs boys to like trucks and

girls to like dolls? That seems unlikely. Hines and colleagues

recently added a new dimension to the sexual differentiation

of play with the observation that girls prefer to mimic the

behavior of other girls or women and that it is this aspect of

their brain development that is influenced by hormones, not

the desire to play with a particular object or in a particular

way (Hines et al., 2016). In girls exposed to androgens in utero,

the desire to mimic other females is lost, and for reasons not

well understood, their preference shifts to toys normally

preferred by boys.

The work of Hines and others speaks to the behavior of

humans, but, as noted above, brain and behavior are not

always closely aligned. Many neuroanatomical sex differences

have been reported in the human brain, but the majority of

these relied on postmortem tissue. Because of this, the majority

also involved the brains of adults although one remarkable

study looked at a nucleus in the hypothalamus from many

different individuals ranging from birth to old age and found

a sex difference that did not appear until late adolescence

and then waned again in older adults (Swaab et al., 2003).

More recently, researchers have taken advantage of noninvasive

imaging techniques that allow for longitudinal analyses of the

same individual as they mature. The magnitude and direction

of sex differences varies with the mode of data acquisition and

analyses (i.e., correcting for total brain volume or not), but

differences in the developmental trajectory of the peak of

cortical gray matter are reliably found and modulated by

androgen receptor allelic variation as well as androgen levels.

White matter increases more rapidly in male brains as development

proceeds and combined with differences in gray matter

amplify the magnitude of sex differences across the life span

(reviewed in Giedd et al., 2012).

More recently advances in imaging have allowed formeasurement

of functional connectivity. Images from almost 1000

humans revealed a profound sex difference in the ‘connectome,’

with females showing strong interhemispheric connectivity and

males the opposite, strong intrahemispheric connections

(Ingalhalikar et al., 2014). The authors interpreted their findings

as supportive of the view that females engage multiple tasks at

once and are highly social whilemales are focused systematizers.

This stereotypical view generated a firestorm of criticism. In

response the authors went on to use a computerized battery of

neurocognitive tasks in combination with imaging and largely

supported their original conclusions (Tunc et al., 2016), but the

controversy here is certainly not resolved.

On the opposite end of the spectrum and equally controversial

was a recent report that reexamined several studies

involving imaging and gender-typical activities. Here the

authors concluded that there was no clear predictability of

sex based on the mean responses (Joel et al., 2015). Instead,

they concluded, every brain is a mosaic of male-like, femalelike,

and neutral features, and therefore there is no such thing

as a ‘male brain’ or a ‘female brain.’ In some ways this conclusion

is intuitively obvious and consistent with the high degree

of regionally specific mechanisms establishing sex differences

in the brain as determined in rodent models. But the finding

was largely misinterpreted by the lay media as demonstrating

there are no sex differences in the brain, which was not the

case even in the Joel et al. study.

This serves as a fitting conclusion to our long treatise on the

modes, mechanisms, and meanings of sex differences in the

brain as it so aptly demonstrates how much we still have to

learn. In some ways, the topic of sex differences in the brain

remains as controversial today as when the first reports were

made in the late 1960s early 1970s. With the changing policies

at major granting agencies, there is likely to be more, not fewer

reports of brain and behavior sex differences. This makes even

more salient the admonishment that scientists bear the burden

of assuring their work is not used or interpreted inappropriately

(Maney, 2016). It is essential that we ‘get it right’ as the implications

of sex differences research reach far beyond the laboratory

to medical, educational, and public health policies that

impact the daily lives of all members of society.