Discovery of new genetic loci for male sexual orientation in Han population

Abstract

Epidemiological studies have demonstrated that the genetic factors partly influence the development of same-sex sexual behavior, but most genetic studies have focused on people of primarily European ancestry, potentially missing important biological insights. Here, we performed a two-stage genome-wide association study (GWAS) with a total sample of 1478 homosexual males and 3313 heterosexual males in Han Chinese populations and identified two genetic loci (rs17320865, Xq27.3, FMR1NB, P_meta = 8.36 × 10^-8, OR = 1.29; rs7259428, 19q12, ZNF536, P_meta = 7.58 × 10^-8, OR = 0.75) showing consistent association with male sexual orientation. A fixed-effect meta-analysis including individuals of Han Chinese (n = 4791) and European ancestries (n = 408,995) revealed 3 genome-wide significant loci of same-sex sexual behavior (rs9677294, 2p22.1, SLC8A1, P_meta = 1.95 × 10^-8; rs2414487, 15q21.3, LOC145783, P_meta = 4.53 × 10^-9; rs2106525, 7q31.1, MDFIC, P_meta = 6.24 × 10^-9). These findings may provide new insights into the genetic basis of male sexual orientation from a wider population scope. Furthermore, we defined the average ZNF536-immunoreactivity (ZNF536-ir) concentration in the suprachiasmatic nucleus (SCN) as lower in homosexual individuals than in heterosexual individuals (0.011 ± 0.001 vs 0.021 ± 0.004, P = 0.013) in a postmortem study. In addition, compared with heterosexuals, the percentage of ZNF536 stained area in the SCN was also smaller in the homosexuals (0.075 ± 0.040 vs 0.137 ± 0.103, P = 0.043). More homosexual preference was observed in FMR1NB-knockout mice and we also found significant differences in the expression of serotonin, dopamine, and inflammation pathways that were reported to be related to sexual orientation when comparing CRISPR-mediated FMR1NB knockout mice to matched wild-type target C57 male mice.

Fig. 1. Genome-wide association analysis of the Han Chinese population.

a Principal component analysis of homosexual men versus heterosexual men. b Quantile-quantile plots of the association results of 521 homosexual men and 1270 heterosexual controls (inflation factor λ_GC = 1.02) from the discovery phase. Results. Overall association results from the discovery phase. c–e Manhattan plots of P values derived from Cochran-Armitage trend tests in the discovery phase, with 521 homosexual men and 1270 heterosexual controls (c); regional plots of the two loci (rs17320865 and rs7259428) associated with male sexual orientation (d, e). f The expression profile of FMR1NB using the HBT database. g The expression profile of ZNF536 using the HBT database. h Differences in the expression levels of FMR1NB and ZNF536 transcripts across various 10 brain regions by using the BRAINEAC database (ID 3994162; ID 3994168; ID 3828304). i Difference of expression level of FMR1NB and ZNF536 transcripts across 10 various brain regions by using the BRAINEAC database (ID 3994162; ID 3994168; ID 3828304).

Fig. 2. Ploygenic risk score profiling using the European results as the discovery set and the Chinese data as the testing set.

The x axis shows six P value thresholds (P = 1e-6, 1e-5, 1e-4, 1e-3, and 1e-2) and the best-threshold (P = 7e-3), which showed the maximal Nagelkerke R2. The y axis shows the Nagelkerke R2, i.e., the proportion of variance in case-control status explained by the risk score profile. The number above each bar is the P value for the capacity of the risk score profile to predict case-control status for that PT. a The GWAS results of males as the discovery set. b The GWAS conducted in males and females combined as the discovery set.

Fig. 3. Postmortem study of ZNF536.

a, b AVP-immunoreactivity (AVP-ir) is identified in the SCN of a homosexual man (a) and a heterosexual man (b). c, d ZNF536-immunoreactivity (ZNF536-ir) of a heterosexual man (c) and a homosexual man (d) in the area of the SCN. Data are presented as the means ± SEM. e, f The cOD of ZNF536 was significantly lower level in homosexual subjects (n = 13) than in heterosexual subjects (n = 13) (e), and the percentage of stained area was relatively smaller in homosexual subjects (f).

Fig. 4. Animal sexual behavior and transcriptomics associated with the FMR1NB gene.

a The mounting behavior of FMR1NB^−/− (n = 10) and FMR1NB^+/+ (n = 14) male mice in resident-intruder tests: The percentage of males who mounted target males was higher in FMR1NB^−/− males than FMR1NB^+/+ males. The difference was very nearly significant (P = 0.051). FMR1NB^−/− male mice mounted with a tendency for a higher frequency (P = 0.092) and longer duration (P = 0.071); Sexual preference of FMR1NB^+/+ and FMR1NB^−/− male mice in the mating choice assay. b The mounting latency, number and duration toward females in FMR1NB^+/+ and FMR1NB^−/− male mice are shown. c Relative expression level of DEGs. d Altered GO and KEGG pathways of serotonin, dopamine (left) and representative inflammatory (right) DEGs. e Protein-Protein Interaction (PPI) Network of DEGs.