More than 1,000 small DNA changes identified that influence age of first period

In a recent study published in Genetics of natureResearchers examined the deoxyribonucleic acid (DNA) of nearly 800,000 women to study the complexities of puberty timing. They identified signals related to the timing of menarche and studied how they influenced the onset of puberty.

Background

Age at menarche (AAM) is a key predictor of pubertal timing in females, influencing reproductive maturity in association with health problems such as cardiovascular disease, diabetes, hormone-related malignancies, and obesity. AAM is polygenic, with 400 genetic loci identified in European populations. It has a strong genetic association with pubertal timing and obesity in males, with the melanocortin-3 receptor (MC3R) being the major hypothalamic sensor linking nutritional status to pubertal timing.

About the study

In the current study, researchers investigated genetic variables influencing women's age at menarche and potential links between reproductive timing and health outcomes later in life.

To discover independent markers of AAM, the researchers analyzed a large genome-wide association study (GWAS) of 799,845 women, 166,890 of whom were of East Asian descent. They also conducted an in-depth study of unusual variations in pubertal timing in 222,283 women using exome sequencing data.

The researchers conducted a meta-analytic GWAS search on age at menarche among 799,845 women using data from five strata: the ReproGen consortium groups (n = 38), the United Kingdom Biobank, the Ovarian Cancer Association Consortium (OCAC), the Breast Cancer Association Consortium (BCAC), 23andMe, and three East Asian biobanks.

The biobanks used were the Korean Genome and Epidemiology Study (KoGES), China Kadoorie Biobank (CKB), and Biobank Japan (BBJ). They also indirectly validated the AAM signals by assessing the age of voice change (AVB) in men from UK Biobank and 23andMe research.

The researchers conducted an exome-wide association study on 222,283 women of European descent from the UK Biobank. They investigated unusual genetic variations, including high-confidence protein truncating variants (HC PTVs) and deleterious variants with combined annotation-dependent depletion (CADD) scores ≥25. They also examined the relationships of rare variants with AAM or BV for ANOS1, CHD7, FGF8, and WDR11, all clinically assessed in hypogonadotropic hypogonadism. They used lassosum and data from a meta-analysis of European-descent cohorts to calculate the AAM polygenic score (PGS).

The researchers compared AAM signals to phenotypic predictions in 3,140 girls from the Avon Longitudinal Study of Parents and Children (ALSPAC), creating a framework known as “GWAS to genes” (G2G). They grouped 1,080 AAM signals from the Norwegian Mother, Father, and Child (MoBa) cohort study based on their relationships with body weight from birth to age eight. They also investigated biological pathways based on early weight trajectories and expression dynamics of AAM-associated genes in GnRH neurons.

Results

The study identified 1,080 distinct genome-wide significant adrenal amino acid (AA) signals, accounting for 11.0% of trait variation in the independent sample dataset. Women in the lowest and highest 1.0% polygenic risk had 14 and 11 times the odds of early and delayed puberty, respectively. The rarest alleles had an effect size of 3.50 months, while more common variations had an effect size of five days.

The researchers observed a 1.2-fold (median) increase in χ2 values ​​for their association with age at menarche in the combined ancestry study compared with those limited to Europeans. This result is proportional to the increasing number of East Asian samples (21%). Among the 1,080 age at menarche signals, 84% (n = 909) revealed concordant directional relationships with age at voice change in the UK Biobank, while 79% (n = 852) were present in 23andMe. Analysis of the combined dataset, comprising 205,354 individuals, showed that 83% (n = 893) of signals exhibited concordant directional effects.

Several genes in 200,000 women had unusual loss-of-function variations, including mutations in zinc finger protein 483 (ZNF483), which counteracted the effects of polygenic risk. Variant-gene maps and neuronal ribonucleic acid (RNA) sequencing of murine gonadotropin-releasing hormone identified 665 genes such as G protein-coupled receptor 83 (GPR83), an unidentified receptor that increased MC3R signaling, a critical nutritional sensor. Shared signals and timing of menopause in genes related to the DNA damage response indicate that ovarian reserves may communicate centrally to initiate puberty.

The Danish Blood Donor Study (DBDS) revealed that AAM variation increased fourfold from 5.60% to 11.0% for 969 accessible signals. Six genes were significantly associated with age at menarche at the exome level, including two genes previously implicated in rare monogenic pubertal disorders: tachykinin receptor 3 (TACR3) in normosmic idiopathic hypogonadotropic hypogonadism (IHH) and Makorin Ring Finger Protein 3 (MKRN3) in familial central precocious puberty.

AAM signals with ALSPAC data explained more variance in AAM than child body mass index (BMI), parental BMI, or maternal AAM. They were as effective in predicting extremes of AAM as a multi-phenotype predictor, and a combined genotype and phenotype model performed well for both early and late AAM.

Conclusion

The study identified age-related signals at menarche, which accounted for 11% of trait variation. Polygenic risk affects timing of puberty, with the top and bottom 1% of risk indicating higher rates of late and early puberty. Rare loss-of-function genetic mutations in ZNF483 affect polygenic risk and timing of menarche.

The study shows possible genetic links between reproductive timing and health outcomes at later ages, highlighting the importance of understanding genetic impacts on pubertal development. The extensive multi-ancestry GWAS signal doubles the variation explained by AAM.