Altered metabolic health during pregnancy in mice with lean polycystic ovary syndrome-like traits from high prenatal AMH

This study demonstrates that while prenatal anti-Mullerian hormone exposure in Ins1-null mice induces mild reproductive PCOS-like traits without causing significant metabolic dysfunction, it does alter metabolic homeostasis during pregnancy by diminishing the insulinogenic response and reducing beta-cell mass, even as these mice maintain superior blood glucose control.

The Gist

The Big Picture: A "Lean" PCOS Mystery

Imagine Polycystic Ovary Syndrome (PCOS) as a complex machine that is running a bit rough. In many women, this machine has two main problems:

1. Reproductive Gears: The ovaries aren't releasing eggs regularly, and hormone levels are off.
2. Fuel System: The body struggles to manage sugar (glucose), often leading to high insulin levels (hyperinsulinemia) and weight gain.

Scientists have long wondered: Is the fuel system problem (high insulin) causing the reproductive gears to jam, or are they just two separate issues happening at the same time?

To find out, the researchers in this paper built a special "test car" (a mouse model) to see what happens if they remove the fuel pump (insulin production) while the reproductive gears are already acting up.

---

The Experiment: Building the Test Car

1. The "PCOS" Setup (The Prenatal AMH):
The researchers gave pregnant mother mice a dose of a hormone called AMH (Anti-Müllerian Hormone). Think of AMH as a "blueprint" that tells the baby's ovaries to develop a specific way. In this case, it programmed the baby mice to have "PCOS-like" traits: slightly longer distances between their genitals (a sign of high testosterone exposure), delayed puberty, and irregular monthly cycles.

2. The "Fuel Pump" Modification (The Insulin Genes):
Mice have two insulin genes (like two fuel pumps): Ins1 and Ins2. Humans only have one.

• The researchers used mice that had no Ins1 gene (one pump removed).
• They split these mice into two groups:
  • Group A: Had one working Ins2 pump (Full dosage).
  • Group B: Had only half a working Ins2 pump (Reduced dosage).
• The Goal: They wanted to see if having less insulin production would fix the PCOS symptoms. If high insulin was the villain, reducing it should make the mice healthier.

3. The Diet Challenge:
They fed some mice a normal diet and others a "junk food" diet (High-Fat, High-Sugar) to see if the PCOS mice would get sick faster than normal mice.

---

The Results: What Happened?

1. The "Lean" Surprise

The researchers expected the PCOS mice to become very sick with high blood sugar and insulin resistance, especially on the junk food diet.

• The Twist: They didn't! Even on the junk food diet, the PCOS mice remained "lean." They didn't get the massive spikes in insulin or blood sugar that usually come with PCOS.
• The Analogy: It's like giving a car a bad fuel mix, but the engine is so efficient it just doesn't stall. The "PCOS" trait existed, but the "Metabolic Disease" trait didn't show up.

2. The Reproductive Gears Still Jammed

Even though the fuel system was fine, the reproductive gears were still slightly off.

• The mice still had delayed puberty and irregular cycles.
• The Lesson: Reducing the insulin "fuel pump" didn't fix the reproductive issues. This suggests that in this specific model, the reproductive problems aren't caused by high insulin.

3. The Pregnancy Plot Twist

This is where the story gets really interesting. The researchers waited until the mice became pregnant. Pregnancy is a time when a body needs to work extra hard to manage sugar for the baby.

• Normal Mice (Control): When they got pregnant, their bodies naturally built more insulin pumps (beta-cells) and pumped out more insulin to handle the extra sugar load. This is a healthy, normal response.
• PCOS Mice: When these mice got pregnant, their bodies failed to ramp up insulin production. They didn't build the extra pumps they needed.
• The Paradox: You would think that failing to make enough insulin during pregnancy would lead to dangerous high blood sugar (Gestational Diabetes). But it didn't!
  • The PCOS mice actually had better blood sugar control than the normal mice.
  • The Analogy: Imagine a bakery (the body) that usually bakes extra bread (insulin) when a big order comes in (pregnancy). The PCOS bakery didn't bake extra bread, yet somehow, they still managed to feed everyone perfectly without running out of flour. They were somehow more efficient at using the bread they had.

---

The Conclusion: What Does This Mean?

1. Two Different Types of PCOS:
This study highlights that PCOS isn't just one thing. There is a "Lean" type where the reproductive system is messy, but the metabolism is surprisingly healthy. In these cases, lowering insulin doesn't seem to help the reproductive issues.

2. Pregnancy Reveals Hidden Flaws:
Even though the PCOS mice looked healthy on the outside, pregnancy acted like a "stress test." It revealed that their bodies handle pregnancy differently. They don't follow the standard "build more insulin" rule, yet they manage to keep their blood sugar perfect.

3. The Takeaway for Humans:
This suggests that some women with PCOS might have a "lean" version where their metabolism is actually quite good, but their bodies react strangely to the stress of pregnancy. It also tells us that simply lowering insulin might not be the magic cure for all PCOS symptoms, because the reproductive issues might be driven by other factors (like the initial hormonal programming in the womb) rather than just high insulin.

In a Nutshell

The researchers tried to fix a "PCOS-like" mouse by turning down its insulin production. It didn't fix the reproductive issues, and the mice stayed surprisingly healthy. However, when the mice got pregnant, their bodies showed a unique, unexpected way of handling sugar that was different from normal mice—proving that pregnancy can reveal hidden metabolic quirks that we can't see when the mice are just sitting around.

Technical Summary

1. Problem Statement

Polycystic Ovary Syndrome (PCOS) is a heterogeneous reproductive disorder affecting up to 20% of reproductive-age women, characterized by ovulatory dysfunction, hyperandrogenism, and polycystic ovarian morphology. A significant subset of PCOS patients presents with "lean" phenotypes that lack overt metabolic dysfunction (e.g., obesity, severe insulin resistance) but may still harbor underlying metabolic vulnerabilities.

The precise role of hyperinsulinemia in driving PCOS pathogenesis and pregnancy complications (such as gestational diabetes and fetal growth restriction) remains unclear. While insulin is known to act as a co-gonadotropin exacerbating hyperandrogenism, it is unknown whether curtailing insulin secretion is sufficient to alleviate PCOS severity. Furthermore, how prenatal AMH exposure (a known driver of PCOS-like traits) influences maternal metabolic adaptation specifically during the high-demand state of pregnancy is not well understood.

2. Methodology

The study utilized a genetically modified mouse model to decouple insulin production from PCOS-like traits.

• Animal Model: Female mice with a complete ablation of the rodent-specific Ins1 gene (Ins1-/-) were used. Within this background, two genotypes were compared:
  • Ins1-/-:Ins2+/+ (Full Ins2 expression).
  • Ins1-/-:Ins2+/- (Partial Ins2 expression, reducing total insulin production).
• PCOS Induction: PCOS-like phenotypes were induced via Prenatal Anti-Müllerian Hormone (PAMH) exposure. Pregnant dams received intraperitoneal injections of recombinant human AMH (0.12 mg/kg/day) from embryonic day 16.5 to 18.5. Control dams received PBS (PPBS).
• Experimental Groups: Four groups were generated: PAMH and PPBS controls, each with either full or partial Ins2 dosage.
• Dietary Regimens:
  • Chow Diet: Mice were assessed longitudinally for 6 months to evaluate baseline metabolic and reproductive traits.
  • High-Fat, High-Sucrose (HFHS) Diet: A subset of mice was switched to an HFHS diet at 2.5 months of age to induce metabolic stress and obesity, followed by timed breeding to assess pregnancy outcomes.
• Key Assessments:
  • Reproductive: Anogenital distance (AGD), vaginal opening (puberty), estrous cycling (vaginal cytology), ovarian histology (follicle counts, corpora lutea), and fertility metrics.
  • Metabolic: Fasting glucose/insulin, Glucose Tolerance Tests (GTT), Insulin Tolerance Tests (ITT), and glucose-stimulated insulin secretion (GSIS).
  • Pregnancy Specifics: Maternal fasting insulin/glucose at gestational day 14.5 (G14.5), $\beta$-cell mass quantification (immunohistochemistry), and fetal/placental health metrics.

3. Key Contributions

• Genetic Dissection of Insulin in PCOS: The study is the first to utilize Ins1-null mice with modulated Ins2 dosage to test if limiting endogenous insulin production mitigates PCOS-like reproductive and metabolic phenotypes.
• Decoupling Reproductive and Metabolic Phenotypes: The authors demonstrate that in this specific genetic background, prenatal AMH exposure induces reproductive PCOS traits without the expected robust metabolic dysfunction (hyperinsulinemia/hyperglycemia) typically seen in other PCOS mouse models.
• Pregnancy-Specific Metabolic Shifts: The study reveals that while non-pregnant PAMH mice appear metabolically "lean" and healthy, prenatal AMH exposure fundamentally alters maternal metabolic adaptation during pregnancy, specifically suppressing the physiological $\beta$-cell expansion and insulin elevation required for normal gestation.

4. Key Results

A. Non-Pregnant Metabolic and Reproductive Phenotypes (Chow & HFHS)

• Reproductive Traits: PAMH mice exhibited classic PCOS-like reproductive features regardless of Ins2 dosage:
  • Increased anogenital distance (indicating fetal hyperandrogenism).
  • Delayed vaginal opening (delayed puberty).
  • Disrupted estrous cycling (prolonged metestrous/diestrous phases).
  • Note: Reducing Ins2 dosage did not significantly ameliorate these reproductive defects.
• Metabolic Traits: Contrary to previous PAMH studies, no robust metabolic dysfunction was observed:
  • No significant hyperinsulinemia or hyperglycemia was detected at 6 months on chow or HFHS diets.
  • Glucose tolerance and insulin sensitivity were largely normal, with only a transient reduction in insulin sensitivity at 8 weeks in PAMH mice.
  • HFHS diet induced mild glucose intolerance in all groups but did not exacerbate the phenotype in PAMH mice compared to controls.
  • Conclusion: In Ins1-null mice, reducing Ins2 dosage did not significantly alter metabolic outcomes because the PAMH model failed to induce the hyperinsulinemia necessary to test the "insulin mitigation" hypothesis.

B. Pregnancy Outcomes (HFHS Diet)

Despite the lack of overt metabolic dysfunction in non-pregnant states, pregnancy revealed significant metabolic shifts:

• Suppressed Insulin Response: Control (PPBS) pregnant mice showed the expected physiological rise in fasting insulin at G14.5. In contrast, PAMH mice failed to elevate insulin levels during pregnancy.
• $\beta$-Cell Mass: There was a trend toward reduced $\beta$-cell mass in pregnant PAMH mice compared to controls, suggesting impaired $\beta$-cell adaptation/hyperplasia.
• Superior Glucose Tolerance: Paradoxically, despite lower insulin levels, PAMH pregnant mice exhibited superior glucose tolerance at G14.5 compared to controls, indicating preserved insulin sensitivity.
• Fetal Health: Despite the altered maternal metabolism, fetal metrics (litter size, fetal weight, fetal-to-placental ratio) were not significantly different between groups at G14.5.

5. Significance and Implications

• Lean PCOS Subtype Validation: The findings support the existence of a "lean" PCOS subtype where reproductive pathology exists independently of severe metabolic dysfunction. This suggests that hyperinsulinemia is not strictly required for the development of PCOS-like reproductive traits in this model.
• Pregnancy as a Metabolic Stress Test: The study highlights that pregnancy acts as a "stress test" that unmasks underlying metabolic dysregulation in PCOS models that are otherwise asymptomatic. Prenatal AMH exposure appears to impair the maternal $\beta$-cell compensatory mechanism required for pregnancy, even if it does not cause gestational diabetes in the short term.
• Mechanistic Insight: The lack of hyperinsulinemia in Ins1-null PAMH mice suggests that androgen-driven insulin secretion in rodents may rely heavily on Ins1 expression (which is absent here) rather than Ins2. This provides a potential explanation for why metabolic phenotypes were muted in this specific genetic model.
• Clinical Relevance: The results suggest that women with lean PCOS may have altered $\beta$-cell dynamics during pregnancy that differ from the general population, potentially increasing the risk for late-gestation complications or requiring different monitoring strategies, even in the absence of pre-pregnancy metabolic syndrome.

Limitations Noted: The study acknowledges that the Ins1-null background may have artificially suppressed the metabolic phenotype, preventing the observation of insulin-mitigation effects. Additionally, the lack of gestational diabetes in PAMH mice suggests the metabolic shift observed might be a compensatory mechanism rather than a failure, though long-term outcomes were not assessed.