Lua Labs Report — Parabacteroides and progesterone: direct modulation beyond estrogen

Date: 2026-05-16 Researcher: Lua Labs Classification: Microbiome Line: L1 — Gut-hormonal axis (estrobolome) Sub-topic: 1.2 — Parabacteroides and progesterone: direct modulation beyond estrogen

External sources

Wang P, Chen Q, Yuan P, et al. (2022). "Gut microbiota involved in desulfation of sulfated progesterone metabolites: A potential regulation pathway of maternal bile acid homeostasis during pregnancy." Frontiers in Microbiology, 13:1023623. DOI: 10.3389/fmicb.2022.1023623
Ly LK, Krieger-Burke T, Mahmud Y, Wong M, Devlin AS (2024). "Gut bacteria convert glucocorticoids into progestins in the presence of hydrogen gas." Cell, 187(13):3344-3358.e20. DOI: 10.1016/j.cell.2024.05.005 (Balskus lab, Harvard)
Liu Y, Liu C, Wu Y, et al. (2025). "Parabacteroides as a promising target for disease intervention: current stage and pending issues." npj Biofilms and Microbiomes, 11:135. DOI: 10.1038/s41522-025-00772-0
Wang K, Liao M, Zhou N, et al. (2019, cited in 2024 review). "Parabacteroides distasonis Alleviates Obesity and Metabolic Dysfunctions via Production of Succinate and Secondary Bile Acids." Cell Reports, 26(1):222-235.e5. DOI: 10.1016/j.celrep.2018.12.028
Duan R, Guan X, Huang K, et al. (2024). "Therapeutic potential of Parabacteroides distasonis in gastrointestinal and hepatic disease." MedComm, 5(12):e70017. DOI: 10.1002/mco2.70017

Baseline knowledge (what I know before searching)

Progesterone (P4) is a C21 steroid synthesized primarily in the corpus luteum of the ovary (after ovulation) and, during pregnancy, in the placenta. The biosynthetic pathway starts from cholesterol → pregnenolone (via StAR + CYP11A1) → progesterone (via HSD3B). In the liver, progesterone is conjugated with sulfate (via SULT2A1) or glucuronic acid (via UGTs) and excreted through bile into the intestine. Its peripheral catabolism generates key neurosteroids: 5α-reductase (SRD5A1/2) converts P4 → 5α-dihydroprogesterone (5α-DHP), and 3α-HSD converts it into allopregnanolone — a positive allosteric modulator of the GABA-A receptor with potent anxiolytic, sedative, and anticonvulsant effects.

Unlike estrogen (whose enterohepatic cycle via gmGUS is well characterized — see L1.1), the enterohepatic cycle of progesterone is less well understood. Progesterone metabolites excreted in bile are massively sulfated (not glucuronidated like E2). Intestinal reabsorption requires bacterial desulfation (microbial sulfatases). Bacteria with robust sulfatase activity are scarce: Bacteroides, Prevotella, and critically, Parabacteroides possess the anSME operon (anaerobic sulfatase-maturating enzyme), which activates type I sulfatases.

In parallel, progesterone and cortisol share a precursor (pregnenolone) and partially compete for receptors: progesterone has significant affinity for the glucocorticoid receptor (GR) and the mineralocorticoid receptor (MR), acting as a partial antagonist. This competition is central in perimenopause: when P4 falls (the first perimenopausal event, before E2), GR becomes more available for cortisol → greater stress reactivity, worse sleep, more inflammation. This is exactly the axis that L2 will explore.

Parabacteroides (family Tannerellaceae, order Bacteroidales) is a genus closely related to Bacteroides but phylogenetically distinct. The most studied human species are P. distasonis, P. goldsteinii, P. merdae, and P. johnsonii. It is a Gram-negative, aerotolerant anaerobe that produces succinate and secondary bile acids (LCA, UDCA). It is one of the few Bacteroidetes with a genome encoding ≥19 sulfatases + a functional anSME gene.

Findings from recent papers

Wang et al. 2022 (Frontiers Microbiology) — Central direct finding for L1.2: in a porcine gestation model, treatment with vancomycin (which preferentially eradicates Gram-positives but also depresses Parabacteroides) produced a significant increase in fecal excretion of allopregnanolone-sulfate (PM4S) and pregnanolone-sulfate (PM5S), while serum levels remained relatively stable. The mechanistic interpretation: the genus Parabacteroides — enriched in genes encoding steroid sulfatases — desulfates sulfated progesterone metabolites in the intestine, enabling their reabsorption. When Parabacteroides falls, sulfates are not deconjugated, they are excreted in feces, and enterohepatic recycling of progestogens is lost. This is the first paper to conceptualize a "progesterobolome" analogous to the estrobolome. Limitation: pregnant porcine model; direct human validation is lacking.

Ly et al. 2024 (Cell, Balskus lab) — Disruptive 2024 discovery: two human gut bacteria, Eggerthella lenta and Gordonibacter pamelaeae (both Coriobacteriaceae, not Parabacteroides, but from the same intestinal niche), perform 21-dehydroxylation of biliary glucocorticoids → progestins. In other words: they convert cortisol/cortisone excreted in bile directly into pregnenolone, progesterone, and allopregnanolone. The reaction requires H₂ as an electron donor, supplied by Escherichia coli and other intestinal H₂ producers. They identified the Elen_2451-2454 gene cluster (Mo-dependent oxidoreductase + Fe-S protein). In third-trimester pregnant women, fecal allopregnanolone levels are 100× higher than in non-pregnant women, correlating with the abundance of these taxa. The implications for perimenopause and postpartum depression are enormous — allopregnanolone is the active principle of brexanolone (Zulresso®), FDA-approved for postpartum depression.

Liu et al. 2025 (npj Biofilms and Microbiomes) — Comprehensive 2025 review: confirms that Parabacteroides encodes at least 19 sulfatases + the anSME gene. Lists P. distasonis, P. goldsteinii, and P. merdae as next-generation probiotics (NGP). Confirms their modulatory role in obesity, insulin resistance, IBD, fatty liver, hepatic fibrosis, inflammatory arthritis, and intestinal barrier dysfunction. Their dominant mechanism is the production of succinate + secondary bile acids (LCA, UDCA, isoLCA, 3-oxoLCA) that activate FXR/TGR5 and modulate systemic inflammation. Epidemiological data: Parabacteroides is lower in premenopausal women than in postmenopausal women (it increases with the transition), although the difference does not always reach significance. P. johnsonii does fall significantly in postmenopause.

Wang K et al. 2019 / Duan et al. 2024 — P. distasonis produces LCA and UDCA, secondary bile acids with indirect hormonal effects: LCA activates intestinal FXR (reduces gluconeogenesis) and TGR5 in macrophages (anti-inflammatory). What is critical for L2: LCA at high concentrations also modulates the glucocorticoid receptor (GR) — prior data not captured in this specific search but consolidated in hepatology literature show that LCA has low but measurable affinity for GR and can act as an allosteric modulator of its signaling.

Complete molecular/endocrine mechanism

CHOLESTEROL
   ↓ StAR + CYP11A1
PREGNENOLONE
   ↓ HSD3B (ovary, adrenal)
PROGESTERONE (P4) ──────────────────┐
   ↓ SRD5A1/2                       │
5α-DHP                              │ pathways:
   ↓ AKR1C2/3 (3α-HSD)              │   1. systemic circulation
ALLOPREGNANOLONE ───→ GABA-A        │   2. liver: sulfation (SULT2A1) /
   (anxiolytic, anti-conv)          │      glucuronidation (UGTs)
                                    │   3. biliary excretion (sulfates)
                                    ↓
                    INTESTINE (lumen)
                                    │
                                    │  PM4S / PM5S / P4-glucuronide
                                    │
   ┌────────────────────────────────┼────────────────────────────────┐
   │  PATHWAY 1 — Desulfation       │   PATHWAY 2 — Direct conversion│
   │  (Parabacteroides spp.)        │   (Eggerthella + Gordonibacter)│
   │  Sulfatases + anSME            │   + Elen_2451-2454 cluster     │
   │  PM4S → free allopregnanolone  │   + H₂ (from E. coli)          │
   │  → reabsorption                │   biliary CORTISOL/CORTISONE   │
   │  → enterohepatic recycling     │   → 21-dehydroxylation         │
   │  → systemic progesterone ↑     │   → PROGESTERONE / PREGNENOLONE│
   │                                │   → INTESTINAL allopregnanolone│
   └────────────────────────────────┴────────────────────────────────┘
                                    ↓
                    Free intestinal allopregnanolone
                                    ↓
                    GABA-A receptor (enteric nervous system
                    + systemic absorption)
                                    ↓
                    Modulation of anxiety / sleep / inflammation
                                    ↓
                    Competition with CORTISOL for GR receptor
                    (direct bridge to L2 — HPA-HPO axis)

Critical note: the two pathways are complementary and reinforcing. Parabacteroides releases pre-synthesized progestogens; Eggerthella/Gordonibacter manufacture progestogens de novo from cortisol. Both require a functional intestinal ecosystem (H₂ producers, low luminal pH, available bile acids). Perimenopausal dysbiosis (documented in L1.1: decline in Lactobacillus, Bifidobacteria, Roseburia) affects both.

Cross-synthesis with previous findings

Connection 1 — L1.1 → L1.2: the estrobolome is only one half. In L1.1 we documented the enterohepatic cycle of estrogens via gmGUS (microbial β-glucuronidase). Today we discovered that an analogous "progesterobolome" exists — and it operates with different chemistry (sulfatases, not glucuronidases) and different taxonomy (Parabacteroides + Coriobacteriaceae, not the generalist Bacteroidetes/Firmicutes). Implication: a dietary intervention that rescues Lactobacillus (favorable for estrogens) may be neutral or even negative for progesterone if it does not affect Parabacteroides. The hormonal microbiome is modular, not monolithic.

Connection 2 — L1.2 → L2 (anticipated): the cortisol-progesterone axis has a third intestinal pathway. The roadmap anticipated that L1.2 would illuminate GR competition. The data exceed expectations: Eggerthella + Gordonibacter convert cortisol directly into progesterone in the intestinal lumen. This means that intestinal dysbiosis does not only affect progesterone through desulfation — it redirects the metabolic fate of biliary cortisol. A woman with dysbiosis: (a) loses P4 recycling (pathway 1), (b) loses cortisol→P4 conversion (pathway 2), and (c) likely has reduced luminal H₂ (altered E. coli). Result: double decline in endogenous progesterone + unconverted biliary cortisol that may potentially recirculate. Three gut→progesterone pathways, one dysbiosis affects them all.

Connection 3 — L1.1 paradox resolved. In L1.1 it remained a curious observation that Parabacteroides increases in postmenopause while Lactobacillus/Bifidobacterium decline. Today's hypothesis: the increase in Parabacteroides in postmenopause could be a compensatory host mechanism — as ovarian progesterone falls, a microbiome that recycles more biliary progestogens is selected. If so, postmenopausal Parabacteroides is adaptive, not pathological, and attacking it with broad-spectrum antibiotics (common in LATAM) could precipitate vasomotor and mood symptoms. This hypothesis is longitudinally falsifiable.

Connection 4 — Loop with thyroid (pre-lab report 2026-05-14). The thyroid report identified that subclinical hypothyroidism reduces luteal progesterone. Classical pathway: high TSH → lower conversion to allopregnanolone. New pathway (from L1.2): subclinical hypothyroidism also alters intestinal motility and microbiome composition → potentially reduces functional Parabacteroides → reduces progesterone recycling. The "luteal progesterone deficit" in hypothyroidism may have an undiagnosed intestinal component.

Lua Labs Hypotheses

Hypothesis 1 — The "progesterobolome" as a modular system distinct from the estrobolome

Statement: There is a specific intestinal microbial sub-ecosystem dedicated to the recycling and production of progestogens ("progesterobolome"), composed of Parabacteroides spp. (desulfatases + anSME), Eggerthella lenta, Gordonibacter pamelaeae (21-dehydroxylases), and cooperating H₂ producers (E. coli, Blautia), whose composition is independent of the estrogenic estrobolome and whose dysfunction explains a significant fraction of progesterone-deficit symptoms in perimenopause even in the presence of stable E2.

Proposed mechanism: Perimenopause is clinically characterized by a decline in progesterone BEFORE estrogen (4-7 years earlier in many women). The classical explanation is ovarian deterioration (follicles without ovulation → no corpus luteum → no P4). L1.2 hypothesis: part of this decline is iatrogenic via the microbiome. A woman entering age 38-42 with (a) a history of broad-spectrum antibiotics (common in LATAM for UTIs and sinusitis), (b) a low-fiber diet (= low luminal H₂), (c) chronic constipation (= greater reabsorption of NON-desulfated sulfates that are otherwise eliminated in feces), loses progestogen recycling years before the ovary fails. "Ovarian failure" becomes symptomatic earlier because there is no backup intestinal recycling. This hypothesis predicts that certain countries/regions with excessive antibiotic use and low-fiber diets will have earlier symptomatic menopause, not earlier biochemical menopause (FSH not necessarily higher in these women, but symptoms are).

Confidence level: Medium — the components are confirmed (pathway 1 in pigs; pathway 2 in humans; epidemiology of Parabacteroides in menopause); the integration into a "progesterobolome" as an independent functional system is an original proposal that requires validation.

How to validate:

With formal study: observational N=200 (100 peri / 100 active-cycle control), measurement of urinary progestogens (sulfated vs free metabolites = progesterobolome ratio), fecal 16S/metagenomic sequencing, validation of Elen_2451-2454 cluster. Duration 6 months, 2 complete cycles.

Limitations:

Pathway 2 (Coriobacteriaceae) was documented in pregnant women; magnitude in non-pregnant women may be substantially lower.
Parabacteroides in postmenopause: causal direction (is it compensatory or pathological?) is unresolved.
Luminal H₂ is difficult to measure in field clinical practice (requires a specific breath test).
P4 sulfates vs glucuronides: the relative ratio of each conjugation in non-pregnant human women has limited data.

Hypothesis 2 — Cortisol→progesterone axis via 21-dehydroxylation: the "intestinal rescue of the luteal phase"

Statement: In women with a functional gut microbiome rich in E. lenta, G. pamelaeae, and H₂ producers, a quantifiable fraction of excreted biliary cortisol is converted into progesterone/allopregnanolone in the intestine, modulating systemic GABAergic tone and reducing the symptomatic severity of chronic stress — and this mechanism is deficient in women with perimenopausal dysbiosis, contributing to the typical anxiety/insomnia "storm" of that transition.

Proposed mechanism: Ly et al. 2024 demonstrated 100× more fecal allopregnanolone in pregnant women. The key question: is that allopregnanolone absorbed systemically or does it only act locally? If it is absorbed (even partially), we would have an endogenous intestinal anti-stress buffer that is not in the literature. In perimenopause, when: (a) ovarian production of P4 falls, (b) the microbiome changes, (c) chronic cortisol rises, this buffer fails exactly when it is most needed. Clinical result: the 47-year-old perimenopausal woman with a processed diet and recurrent antibiotics develops "menopausal anxiety," while one with a traditional diet + intact microbiome transitions relatively asymptomatically — a difference explainable by intestinal 21-dehydroxylation capacity.

Confidence level: Low-Medium — the pathway is biochemically confirmed; its quantitative relevance in non-pregnant humans is speculative. But falsifiability is clear.

How to validate:

With formal study: dual determination of allopregnanolone in urine + feces in perimenopausal women with vs without high probiotic+fiber intake, n=80, 2 cycles. Prediction: urinary allopregnanolone/cortisol ratio correlates positively with fermentable-fiber intake, negatively with recent antibiotic use.

Limitations:

Intestinal allopregnanolone may act mainly locally (enteric nervous system) and may not reach the CNS in pharmacological quantities.
Multiple confounders: residual ovarian, adrenal, and intestinal progesterone are indistinguishable in total serum/urine without isotopic tracers.

Candidate formulation — "Progesterobolome Stack" (concept, not product)

Compounds:

Parabacteroides distasonis (candidate NGP strain — several in preclinical pipeline) — sulfatases for deconjugation of P4-sulfates
Akkermansia muciniphila — synergy with Parabacteroides documented in barrier restoration; releases bile acids available for LCA/UDCA
Agave inulin + resistant starch (contextual LATAM food) — fermentation to H₂ + butyrate; substrate for H₂ producers
Glycine 1g — 3α-HSD substrate for conversion 5α-DHP → allopregnanolone
Magnesium bisglycinate 200mg — cofactor of 5α-reductase and 3α-HSD

Target population: Carmen (47, perimenopause with progesterone-deficit symptoms — luteal anxiety, premenstrual insomnia, short periods) and Sofía (28, active cycle with severe cyclic PMS). NOT Rosa postmenopausal (no ovulation → no ovarian P4 to recycle, pathway 1 less relevant; pathway 2 remains active but requires validation).

Complementary mechanisms:

Parabacteroides rescues pathway 1 (desulfation)
Inulin + resistant starch rescues H₂ producers → enables pathway 2
Glycine + Mg support peripheral conversion P4 → allopregnanolone
Akkermansia repairs barrier → reduces LPS absorption → reduces chronic cortisol

Regulatory status:

P. distasonis as probiotic: NOT commercial in the U.S./LATAM yet — in NGP pipeline, not GRAS.
Akkermansia: commercial (Pendulum), dietary supplement.
Inulin, resistant starch, glycine, magnesium: GRAS.

Requires validation:

RCT n=120 perimenopausal women (CGM no longer relevant, but measure: diurnal salivary cortisol, 24h urinary progestogens, validated symptom score [Greene Climacteric Scale], fecal microbiome composition), 16 weeks, vs placebo + active control (GUS+ probiotic type L. brevis KABP052 from L1.1).
Primary endpoint: reduction in Greene score ≥30%; secondary endpoints: urinary allopregnanolone/cortisol ratio, Parabacteroides abundance.

Individual variability

Why do two women of the same age respond differently to the same hormonal event? Modulating factors of the progesterobolome:

Antibiotic history. Each broad-spectrum course reduces functional Parabacteroides for months. Cumulative burden is likely decisive.
SULT2A1 / SRD5A1 / SRD5A2 polymorphisms. SULT2A1 determines how much progesterone is sulfated vs glucuronidated (affects dependence on the progesterobolome). SRD5A1/2 determines peripheral conversion to allopregnanolone — women with hypoactive variants (~20-30% of the population) depend more on microbial contribution.
Polymorphisms in AKR1C2/3 — the 3α-HSD that completes allopregnanolone. Documented variability.
Age at menarche and duration of ovarian exposure — more years of ovulatory cycles = greater maturation of the progesterobolome.
Traditional vs Western diet. Traditional Latin American diets (beans, nixtamalized corn, nopal, fermented foods) likely sustain higher Parabacteroides than a processed Western diet — a factor still poorly characterized in the literature.
Chronic stress. Persistent cortisol reduces intestinal motility and alters microbial composition → affects both progesterobolome pathways. (Bidirectional loop with L2.)
Previous pregnancies. Pregnancy massively expands populations of E. lenta and G. pamelaeae (Ly et al. 2024). Does the expansion persist postpartum? If yes, women with ≥2 pregnancies could have a more robust progesterobolome in perimenopause. Sub-original hypothesis.