Polycystic Ovary Syndrome (PCOS) is the most common endocrine disorder among females of reproductive age globally, affecting between 5-26% of these women, depending on the criteria applied. It is the leading cause of anovulatory infertility and a risk factor for type 2 diabetes. However, compared with other common conditions, PCOS remains poorly understood and relatively understudied. National Institute of Health (NIH) data show that between 2006 and 2015, PCOS research received about $215 million (US), compared with other common conditions; $454 million (rheumatoid arthritis), $775 million (tuberculosis), and $610 million (systemic lupus erythematosus)1. The genetic signals that underlie risk also tend to be found in non-coding DNA, which has historically received less attention than translated regions.
A new study in Nature Communications2 combines genome-wide association studies (GWAS) with functional assays and CRISPR-based perturbation. Sankaranarayanan and colleagues mapped how PCOS-associated variants affect gene regulation and hormone production. Their work provides mechanistic evidence linking variants at the DENND1A locus to increased testosterone production, a core biochemical feature of PCOS.
PCOS has a strong genetic component, with heritability estimates from twin studies of around 79%. To date, GWAS have identified 30 loci associated with altered risk, but most signals lie in non-coding regions. These regions are difficult to interpret because they do not alter proteins directly, but influence gene regulation.
To investigate, the researchers applied a high-throughput reporter assay (STARR-seq) across 14 risk loci, spanning nearly 5 megabases of the genome. They tested two steroidogenic cell models: an adrenal line (H295R), which produces testosterone, and an ovarian line (COV434), which produces estradiol. This approach identified nearly 1000 regulatory elements, roughly half with enhancer activity and half with repressor activity. Many overlapped with open chromatin regions, conserved sequences, or enhancer annotations from ENCODE, indicating biological relevance.
Fine-mapping with case-control genetic data from almost 4000 individuals then pinpointed 19 variants in functional elements that associated significantly with PCOS risk. Colocalization analyses further connected several of these variants to gene expression changes, nominating GATA4, FSHB, and DENND1A as candidate effector genes.
Among these, DENND1A was selected for deeper study. Previous work had linked rare DENND1A variants to increased androgen biosynthesis in ovarian theca cells, but a direct mechanism tying common non-coding variation to hormone production was missing.
Using targeted STARR-seq on the DENND1A regions from multiple individuals, the team identified dozens of candidate regulatory elements and 62 variants with allele-specific activity. Several overlapped with expression quantitative trait loci (eQTLs) for DENND1A and neighbouring genes, strengthening the case that they influence expression.
To test causality, they used CRISPR-based epigenome editing. Activating candidate regulatory elements with dCas9-p300 increased DENND1A expression up to 4x in adrenal cells, while repression with dCas9-KRAB reduced expression. Crucially, changes in DENND1A expression translated into altered testosterone output: activation increased production by up to 3x, and repression decreased it by up to 2x. Estradiol levels also rose with DENND1A activation, consistent with testosterone serving as a precursor.
These experiments provide a direct functional link between PCOS-associated regulatory variation, DENND1A expression, and androgen excess.
The study highlights several broader points:
Sankaranarayanan et al. demonstrate a scalable approach for connecting genetic risk variants to biological function in complex diseases. By focusing on PCOS, they address both an understudied condition and under-researched regions of the genome. The finding that common non-coding variants can modulate androgen production through DENND1A strengthens the case that regulatory variation is central to PCOS pathogenesis.
For researchers, the study offers a blueprint for fine-mapping GWAS loci in other diseases. For women’s health, it serves as a reminder of research imbalances. PCOS receives less attention than other female-specific conditions, and such conditions themselves have historically been understudied compared with male-specific or -dominated diseases. NIH analyses (2021) show that in nearly three-quarters of diseases that predominantly affect one sex, funding favours male-associated diseases. That is, either female-dominant conditions are underfunded, or male-predominant ones are overfunded relative to burden, representing a deeper bias that goes beyond any one disorder3.
Still, this gap is closing. In August 2025, the Gates Foundation pledged $2.5 billion (US) through 2030 to support innovations in women’s health, targeting areas like maternal care, contraception, menstrual and gynecological health, and STIs4. But philanthropic funding, while welcome, cannot substitute for sustained public investment. Maintaining momentum will require governments to protect and expand women’s health research funding, ensuring studies like this one translate into meaningful clinical advances. It also highlights that variant interpretation is not only a rare disease challenge, but that connecting genetic signals to mechanisms is equally important for progress in common and understudied conditions like PCOS.
Company News