Large-scale genome-wide analyses of proteomic data identifies that sex hormones affect plasma glycodelin levels

This large-scale genome-wide study utilizing UK Biobank data reveals that sex hormones causally influence plasma glycodelin levels, suggesting that previously observed associations between glycodelin and reproductive diseases are likely driven by hormonal effects rather than a direct causal role of glycodelin itself.

The Gist

Imagine your body is a massive, bustling city. Inside this city, there are millions of tiny workers (proteins) doing specific jobs to keep everything running smoothly. One of these workers is called Glycodelin.

Think of Glycodelin as a specialized security guard who works in the reproductive districts of the city (the ovaries, uterus, and testes). Its job changes depending on the "shift" (whether you are male or female, and whether you are pre- or post-menopausal). Sometimes it helps sperm meet an egg; other times it stops them from meeting; sometimes it helps a pregnancy stick, and other times it prevents it.

For a long time, scientists noticed that when this security guard's numbers went up or down, it seemed to happen alongside various city problems like infertility, endometriosis, or certain cancers. They wondered: "Is the security guard causing these problems, or is the guard just reacting to them?"

This new study acts like a detective trying to solve that mystery using a special tool called "Mendelian Randomization." Think of this tool as a time machine that looks at your DNA (your genetic blueprint) to see what your body was destined to do, rather than just what is happening right now. This helps avoid confusion caused by other factors like diet or lifestyle.

Here is what the detective found, broken down into simple stories:

1. The "Boss" is the Hormone, Not the Guard

The study discovered that Glycodelin isn't usually the one pulling the strings. Instead, it's more like a thermometer that reacts to the weather.

• The Weather: The "weather" in this city is your sex hormones (like testosterone).
• The Finding: The study found that your hormone levels cause the number of Glycodelin guards to change.
  • In Men: Higher testosterone acts like a "green light," telling the body to produce more Glycodelin.
  • In Women (Before Menopause): It's the opposite! Higher testosterone acts like a "red light," telling the body to produce less Glycodelin.
  • In Women (After Menopause): The rules flip back. Higher testosterone acts like a "green light" again, increasing Glycodelin.

The Analogy: Imagine Glycodelin is a thermostat. In the winter (pre-menopause), turning up the heat (testosterone) makes the thermostat turn down. In the summer (post-menopause), turning up the heat makes the thermostat turn up. The thermostat isn't controlling the temperature; the temperature is controlling the thermostat!

2. The Guard Doesn't Cause the City Disasters

Scientists had suspected that if Glycodelin levels were weird, it might cause diseases like ovarian cancer or endometriosis.

• The Verdict: The detective found no strong evidence for this.
• The Explanation: It turns out that when people see low or high Glycodelin levels in sick patients, it's usually because the hormones are messed up, and the Glycodelin is just a bystander. The Glycodelin isn't the criminal; it's just a witness that got caught in the crossfire.

3. The "Shift Change" at Menopause

The study highlighted that the city operates differently before and after menopause.

• Before menopause, the ovaries are the main factory producing hormones.
• After menopause, the body switches to a different factory (the adrenal glands).
• Because the source of the "weather" changes, the way Glycodelin reacts to it also changes. This explains why some studies in the past got conflicting results—they were looking at two different cities with different rules!

4. Why This Matters

For years, doctors have tried to use Glycodelin levels as a crystal ball to predict fertility issues or cancer.

• The Takeaway: This study suggests that using Glycodelin as a crystal ball might be misleading. If you see a weird Glycodelin level, don't blame the guard. Instead, look at the hormones (the weather).
• The Future: To really understand reproductive health, we need to focus on fixing the hormonal balance, not just watching the security guard's numbers.

Summary in One Sentence

This study proves that sex hormones are the bosses that control how much Glycodelin is in your blood, and Glycodelin is mostly just a passive messenger reacting to those hormones, rather than the direct cause of reproductive diseases.

Technical Summary

1. Problem Statement

Glycodelin (progesterone-associated endometrial protein) is a glycoprotein secreted by reproductive tissues with four distinct glycoforms (-A, -C, -F, -S) that determine its function. Observational studies have consistently linked glycodelin levels to various reproductive traits and diseases, including fertility, endometriosis, preeclampsia, and reproductive cancers. However, these observational associations suffer from confounding factors and reverse causality, making it unclear whether glycodelin plays a causal role in these conditions or if its levels are merely a biomarker driven by underlying physiological factors (such as sex hormones).

The study aims to:

1. Identify genetic determinants of plasma glycodelin levels using large-scale genomic data.
2. Determine the causal direction of relationships between glycodelin and reproductive traits (i.e., does glycodelin cause the trait, or does the trait/hormone cause glycodelin changes?).
3. Investigate whether these relationships differ by sex and menopause status.

2. Methodology

The researchers utilized data from the UK Biobank (v3 release), focusing on individuals of European genetic ancestry aged 40–69.

• Data Source: Proteomic data for glycodelin was obtained from the UK Biobank Pharma Proteomics Project (Olink proximity extension assays) in 54,219 participants.
• Study Design:
  • Stratified GWAS: Genome-wide association studies (GWAS) were performed on glycodelin levels in five distinct groups:
    1. Sex-combined (All)
    2. Men only
    3. Women only
    4. Pre-menopausal women
    5. Post-menopausal women
       Total sample size for GWAS: 46,468 individuals.
  • Mendelian Randomization (MR): Two-sample and one-sample MR analyses were conducted to test bidirectional causal associations.
    • Exposure 1: Genetically determined glycodelin levels (using a cis-acting variant near the PAEP gene, rs9409964, as the instrument).
    • Exposure 2: Genetically determined reproductive traits (19 traits including sex hormones, cancers, and fertility markers).
    • Outcomes: The reciprocal traits.
• Statistical Analysis:
  • GWAS was performed using REGENIE with inverse rank normalization, adjusting for age, genotyping batch, recruitment center, and principal components.
  • MR analyses used the Inverse-Variance Weighted (IVW) method as the primary estimator, supported by sensitivity analyses (Weighted Median, MR-Egger, Mode-based) to detect pleiotropy.
  • Significance thresholds were Bonferroni-corrected ($P < 1.4 \times 10^{-3}$ for glycodelin $\to$ traits; $P < 1.8 \times 10^{-3}$ for traits $\to$ glycodelin).

3. Key Contributions

• Stratified Genetic Discovery: The study is the first to perform GWAS of glycodelin stratified by sex and menopause status, revealing that genetic architecture and effect sizes differ significantly across these groups.
• Causal Inference: It provides robust causal evidence distinguishing between direct effects of glycodelin on disease versus the influence of sex hormones on glycodelin levels.
• Hormonal Regulation Mechanism: It elucidates that sex hormones (specifically testosterone and SHBG) are the primary drivers of glycodelin levels, with opposing effects observed before and after menopause.

4. Key Results

A. Genetic Architecture of Glycodelin

• Nine Independent Loci: Nine genetic signals reached genome-wide significance ($P < 5 \times 10^{-8}$).
• Primary Signal: The strongest signal was rs9409964, located near the PAEP gene (which encodes glycodelin).
  • Sex Differences: The effect size was significantly larger in men (1.31 SD per A allele) than in women (0.60 SD).
  • Menopause Differences: In women, the effect was approximately double in post-menopausal women (0.91 SD) compared to pre-menopausal women (0.40 SD).
• Additional Signals: Six novel signals were identified that did not reach significance in previous unstratified studies, highlighting the importance of stratification.

B. Causal Effects: Glycodelin $\to$ Reproductive Traits

• Limited Causal Impact: There was little evidence that genetically elevated glycodelin levels causally increase the risk of reproductive conditions (e.g., endometriosis, preeclampsia, infertility, or most hormone-sensitive cancers).
• Minor Associations:
  • Elevated glycodelin showed a weak association with increased ovarian cancer risk (OR=1.06).
  • Weak associations were found with male BMI (lowering) and female waist-to-hip ratio (raising), but these did not survive strict multiple testing correction.

C. Causal Effects: Sex Hormones $\to$ Glycodelin

• Strong Bidirectional Causality: The study found robust evidence that sex hormones causally determine glycodelin levels, with effects varying by sex and menopausal status:
  • Bioavailable Testosterone:
    • Men: Increases glycodelin ($\beta = 0.14$ SD).
    • Post-menopausal Women: Increases glycodelin ($\beta = 0.10$ SD).
    • Pre-menopausal Women: Decreases glycodelin ($\beta = -0.16$ SD).
  • SHBG (Sex Hormone Binding Globulin):
    • Pre-menopausal Women: Increases glycodelin.
    • Post-menopausal Women: Decreases glycodelin.
• Other Traits: Later age at natural menopause was causally associated with higher glycodelin levels.

5. Significance and Implications

• Reconciling Conflicting Literature: The study suggests that previously observed associations between glycodelin and reproductive diseases (like endometriosis or infertility) are likely confounded by sex hormones rather than being direct causal relationships. For example, the link between high glycodelin and endometriosis may be explained by the fact that high testosterone lowers endometriosis risk but has opposite effects on glycodelin depending on menopausal status.
• Biomarker Interpretation: While glycodelin levels correlate with reproductive health, they should be interpreted as a downstream consequence of hormonal regulation (specifically the interplay between testosterone, SHBG, and menopause status) rather than a direct therapeutic target for these conditions.
• Biological Insight: The findings highlight a shift in biological pathways between pre- and post-menopausal women, likely reflecting the transition from ovarian to adrenal steroidogenesis.
• Limitations: The study notes that plasma measurements may not reflect local tissue levels, and the inability to distinguish between the four glycodelin glycoforms limits mechanistic understanding. Additionally, the pre-menopausal sample size was modest.

Conclusion: The authors conclude that sex hormones are the primary causal drivers of plasma glycodelin levels, and the observed correlations between glycodelin and reproductive diseases are likely indirect consequences of these hormonal pathways rather than direct causality.