Current dogma holds that differences in brain development and behavior between males and females depend primarily on gonadal steroid hormones, especially testosterone and its metabolites that induce the masculine pattern and inhibit the female pattern of brain development. However, there is also evidence that genetic factors may act directly on the developing brain contributing to these differences. Until recently, this alternative view has been difficult to document, but Dewing et al provide new and convincing evidence for non-hormonal genetic effects.

Their work was done in a mouse embryo 10.5 days after conception. This is just before the first sign of sexual differentiation of the genital ridges occurs, thus the influence of gonadal hormones could be excluded. Their strategy was to harvest whole heads from the embryos, isolate RNA into separate pools for males and females and then analyze for differential gene expression in the male and female brains. For screening analysis, they used gene (microarray) chip (Affymetrix) technology which allowed the relative expression of nearly 10,000 characterized mouse genes and over 3,000 less well defined expressed sequences (Expressed Sequence Tags – ESTs) to be determined. The normalized gene chip results reported as fold change or difference between male and female brain RNA revealed 36 genes or ESTs with enhanced expression in females and 18 genes or ESTs with enhanced expression in males. These genes exhibited a significant fold difference of greater than 1.1 and 7 genes or ESTs for each sex displayed a fold difference of 2.0 or more. The gene showing highest differential expression in females was Xist, which was 18.5 fold higher in females, while genes showing the highest differential expression in males included DEAD box peptide (Dby) and eukaryotic translation initiation factor 2,Y (Eif2s3Y) with fold differences of 10.0 and 8.8, respectively. Xist maps to the X chromosome, while the latter two genes reside on the Y chromosome.

Real-time quantitative analysis (RT-PCR) of littermate-matched male and female embryonic brain RNA confirmed and validated the results of the gene chip screening for a small number of genes based on their potential roles in brain development. The authors concluded that developmental differences in male and female brains in mice are due in part to the differential expression of genes before gonadal secretion starts.

Dewing P, et al. Sexually dimorphic gene expression in mouse brain precedes gonadal differentiation. Mol Brain Res 2003;118:82-90.

First Editor's Comment: This is an important paper that documents the differential expression of genes in the male and female brain prior to any influence from gonadal hormones. If confirmed, it will have a substantial impact on understanding how genetic factors influence brain development. The design of the study allows for the identification of non-hormonal factors that act before the gonads are formed. However, there is no reason to think that genes act through mechanisms that do not involve gonadal hormones after gonadal hormone secretion begins, although other investigational approaches will be needed to demonstrate this. Dewing and colleagues provide no insight into the nature of the non-hormonal mechanisms through which genes may act before the appearance of gonadal hormones, although they could presumably be multiple and diverse.

One should note that the most dramatic differences were found for genes whose expression is expected to be limited to one sex or the other. For example, one would expect genes located on the Y chromosome to be expressed only in the male brain and Xist mRNA, which is expressed only by the inactive X chromosome in XX females, to be detected only in the female brain. That they were detected at all, seemingly reflects how the assays distinguish negative results from background signals. When these results are excluded the differences were diminished. Microarray gene chip and related approaches for studying gene expression are relatively new and evolving rapidly as is bioinformatics, the discipline that deals with analysis of the vast amounts of data this technology generates. Its novelty combined with the complexity of its data has led to a certain amount of caution in the biomedical field with regard to the biological significance of microarray results. Initially, a 2-fold difference in expression was considered an informal threshold for biological significance. Many of the results in this study fall below this level and therefore would not be considered significant by this criteria even though they are statistically significant. However, as the analytical methods advance, the threshold is being progressively lowered such that a cut-off, such as the 1.1-fold difference used in this paper, is becoming acceptable. It is still probably wise, however, to view small differences in gene expression with caution until they are confirmed by others and placed in a biological context.

William A. Horton, MD

Second Editor's Comment: The findings of this study are important and exciting, and will likely contribute to a transformation of the dominant conceptual model regarding sexual differentiation of somatic phenotype, brain, and behavior. There is a risk that the findings may be misinterpreted in a manner potentially harmful to the clinical decision-making process in cases involving intersexuality. The findings force us to rethink the classic view of brain sexual differentiation and behavior which posits that the role of genes in the development of sex differences is restricted to the process of sex determination, i.e., the development of a bipotential and undifferentiated gonad into either an ovary or a testis. Evidence of a direct role of genes (not mediated by sex hormones) may lead clinicians to question the flexibility in decision-making they may currently exercise when sex assignment is in question. But should they?

The basic finding of the study is that over 50 candidate genes are differentially expressed in the brains of male and female mice, ostensibly prior to gonadal production of sex hormones. Although a remarkable observation, these findings are not necessarily relevant for one psychological outcome variable of great importance in intersex cases, that is the stability of gender identity across the lifespan. (Gender identity refers to the individual's self identification as either girl/woman or boy/man.) Readers of media reports of this article will likely draw different conclusions. The headline of one well-publicized report of this study states "Sexual Identity Hard-Wired by Genetics." 1 Quotes within the article imply that gender identity springs directly from our genome. If so, then how do we account for the consistent finding in the literature that 46,XY individuals with complete androgen insensitivity syndrome develop an unambiguous gender identity as girls, and later women?2

The conflict between research findings and their interpretation is likely more apparent than real and is promoted by an oversimplification of the process of psychosexual differentiation in humans. An individual's gender identity need not be congruent with their gender-role (which refers to behaviors that differ in frequency or level between males and females in this culture and time such as toy play or maternal interest), and sexual orientation (the pattern of sexual arousal). At the present time, the clinical research literature suggests that gender identity generally conforms with the gender of rearing, even when gender assignment is discordant with genetic sex. The picture is quite different, however, with respect to the variables of gender-role behavior and sexual orientation. It is clear that many new findings will stem from the line of research described in this report. However, it would be unfortunate if these data were to be interpreted as suggesting that gender assignment must conform with genotype to foster a stable gender identity.

David E. Sandberg, PhD

References - (linked to )

1. Reuters Health. Sexual identity hard-wired by genetics. http://www.nlm.nih.gov/medlineplus/news/fullstory_14351.html
2. Hines M, et al. Arch Sex Behav 2003;32:93-101.