VDR and ovarian function: from an “underestimated polymorphism” to a causal system of receptor, ligand, identity, and environment

Project: L6-5
Stage: REPORT_EN
Cutoff date: 2026-07-24
Classification: Hormonal nutrigenomics
Question: What mechanisms connect VDR and ovarian function with women’s health and longevity, and what new hypotheses can be falsified?
Overall maturity: H1 for preclinical ovarian actions and for the mineral/HPO contribution in global animal models; H0–H1 for a specific local action in the adult human ovary; H0 for a causal common polymorphism, longitudinal translation, and mediation to healthspan.

This report is scientific research, not medical advice. It does not prescribe vitamin D, calcitriol, polyamines, sterols, or VDR modulators.

1. Executive scientific abstract

VDR belongs to ovarian biology, but the thesis that an “underestimated VDR polymorphism” has demonstrated effects on women’s hormonal health and longevity is not supported. rs2228570/FokI changes translation initiation and generates VDR forms of 424 or 427 amino acids, but its transcriptional advantage was positive in one overexpression reporter and null in later assays of affinity, DNA binding, three reporters, and endogenous induction. BsmI, ApaI, and TaqI are 3′ markers without demonstrated ovarian function; Cdx2 has regulatory activity dependent on intestinal context. Candidate studies of PCOS, AMH, and assisted reproduction are small or discordant; some signals have Hardy–Weinberg and multiplicity problems; two larger cohorts were null for susceptibility or genotype×25(OH)D interaction; and the PCOS GWAS did not prioritize VDR. No reviewed common polymorphism completes the chain allele → ovarian VDR function → ovarian trajectory → healthspan.

The evidence does support a context-dependent ovarian VDR system. Calcitriol modifies programs in human granulosa, cumulus, and theca cells ex vivo, but the direction changes with follicle size, luteinization, FSH, insulin, dose, and time. In SUNDRO, a 2.8-fold increase in follicular 25(OH)D changed transcription but not follicular hormones; in the main trial of 630 women, cumulative clinical pregnancy was 37% versus 40% with placebo (RR 0.91; CI95% 0.75–1.11). Exposure is not equivalent to function.

Two recent mechanisms expand preclinical plausibility: VDR–ODC1–polyamines–DNMT–p53/p21 in granulosa cells and VDR/RXR–PI3K–AKT–FOXO3 in primordial follicles. Both remain provisional. The first did not measure dcSAM/SAM directly, despite an available LC–MS/MS method, and contains a directional contradiction in the AMD1 arm; the second used pharmacological calcitriol, inferred PI3K activity from complex association, and did not demonstrate genetic VDR loss/rescue in human tissue. Delaying follicular activation is also not equivalent to improving function: the mouse model shifted the end of reproductive activity but did not increase total offspring.

The donor-aware reanalysis of GSE202601 also failed to reproduce VDR↓–ODC1↓–CDKN1A↑ in aged human granulosa cells. VDR was too sparse and pointed upward; ODC1↓ and CDKN1A↑ were imprecise. In contrast, SMS↓ and especially FSHR↓ indicated polyamine remodeling together with loss/change of granulosa identity. Granulosa snATAC failed the donor gate. This result weakens VDR RNA as the human upstream factor and elevates two rivals: cellular-identity remodeling and the systemic mineral–HPO environment.

The final synthesis replaces the linear model “vitamin D → VDR → better ovary” with four gates:

1. identity and stage: follicular origin, luteinization, and granulosa state;
2. active ligand: 25(OH)D entry, CYP27B1 conversion, and CYP24A1 catabolism;
3. receptor: VDR protein/localization, RXR/cofactors, chromatin, and the rapid PDIA3 alternative;
4. systemic environment: Ca/P/PTH/FGF23, the HPO axis, and extraovarian VDR pathways.

The highest-information experiment is an acute VDR × PDIA3 × calcitriol factorial in non-tumoral human granulosa cells, using the donor as the unit. It must require loss/rescue, measured intracellular ligand, a proximal output (CYP24A1), and an independent function (aromatization/E2 under FSH). If VDR changes target engagement but not function, the functional extension is falsified. Only a positive result opens CYP27B1 conversion, polyamines versus HMGCR–7-DHC, local–systemic comparison, and, much later, trajectory and healthspan.

Scientific delta of REPORT_EN

The new contribution of this synthesis is to establish that the polymorphism is not the defensible causal unit: experimental priority must move from genotyping notorious SNPs to decomposing an ovarian response into pretreatment identity, active ligand, nuclear VDR, PDIA3, and mineral/HPO homeostasis. In addition, the path to longevity cannot be modeled as a linear benefit because a longer reproductive trajectory may favor bone and type 2 diabetes while increasing hormone-sensitive cancers, and VDR acts directly in all those organs. Refuting a local action, demonstrating dominance of identity or the mineral environment, or establishing FokI equivalence are positive scientific results.

2. Scientific question and relevance

The causal question is not whether “vitamin D” correlates with AMH, PCOS, or fertility. It is:

In adult women with ovaries in situ, does a functional and reproducible difference in the VDR system—through acquired state or, secondarily, a variant—modify an ovarian function in a ligand- and stage-dependent manner, and then mediate an endocrine trajectory with organ-specific health effects?

The immediate estimand is an interaction within tissue:

cell identity × intracellular calcitriol × VDR/PDIA3 → proximal response → independent ovarian function.

The life-course estimand, conditional on validating the first, would be:

ovarian VDR function → follicular/endocrine trajectory → bone, cardiometabolic, or cancer outcome,

compared against direct VDR pathways in intestine, kidney, bone, muscle, adipose tissue, endothelium, and immunity. Without that comparison, a systemic VDR association cannot be attributed to the ovary.

The relevance to women’s longevity is therefore potential but unproven. Healthspan means years free of morbidity and requires incident outcomes and competing risks; it is not replaced by AMH, AFC, follicle number, fertility, age at menopause, or survival alone.

3. Scope, population, and life stage

The human core concerns adult women with ovaries in situ. Stages must not be mixed:

• 18–34 years, early/mid reproductive stage, as the functional reference;
• 35–45 years, late reproductive stage/early transition, separating chronological age from STRAW+10 stage;
• 40–60 years, only as a longitudinal bridge to the menopausal transition, systemic function, and early healthspan outcomes;
• PCOS, diminished ovarian reserve, and primary ovarian insufficiency as different biological strata, not degrees of one disorder.

The main compartments are cortex/primordial follicle, mural granulosa and cumulus, theca, stroma, oocyte, and corpus luteum. IVF evidence is interpreted separately because stimulation, trigger, luteinization, infertility etiology, and clinical selection alter identity and response.

Outside the core are pregnancy/lactation, endometrium without ovarian mediation, rare mutations causing hereditary vitamin D resistance, individual recommendations, and any use of current Lua data, PHI, PII, or private genomes.

4. Background knowledge and mechanism map

4.1 Four-gate architecture

EXPOSURE AND LIGAND
D3 → CYP2R1 → 25(OH)D–DBP → follicular entry
                           ├─ CYP27B1 → calcitriol
                           └─ CYP24A1 → catabolism
                                      │
IDENTITY/STAGE                        ▼                 RECEPTOR
follicular origin ───────────→ intracellular calcitriol → VDR + RXR/cofactors
FSHR/CYP19A1/FOXL2                                      ├─ VDRE occupancy
luteinization/metabolism                                └─ transcription
          │                                                    │
          └────────────────────────────────────────────────────┤
                                                               ▼
                         ODC1/SMS–polyamines–dcSAM/SAM–DNMT–p53/p21
                         mitochondria / FSH response / steroidogenesis
                                                               │
                                recruitment, growth, atresia, ovulation
                                                               │
                              E2/P4/androgen/inhibin trajectory
                                                               │
                             bone, metabolism, vasculature, and cancer

COMPETING RAPID PATHWAY:
calcitriol → PDIA3 and/or membrane VDR → Ca²⁺/AKT/ERK [human ovary: unproven]

COMPETING SYSTEMIC PATHWAY:
intestinal/renal VDR → Ca/P/PTH/FGF23 + HPO → same reproductive phenotype

EXTRAOVARIAN PATHWAY:
VDR in bone/muscle/adipose/endothelium/immunity → same healthspan outcomes

4.2 From variant to receptor

rs2228570/FokI alters the start codon. The C/F allele produces a 424-aa VDR and T/f a 427-aa VDR. This has been demonstrated in non-ovarian molecular systems. It has not been demonstrated that either isoform has greater activity in granulosa cells: Arai observed approximately 1.7-fold greater activity of the short receptor in one reporter under overexpression and 50 nM calcitriol; Gross found indistinguishable affinity, DNA binding, ED50, three reporters, and CYP24A1 induction. The correct hypothesis is equivalence until a prespecified and replicated interaction demonstrates otherwise.

BsmI/rs1544410, ApaI/rs7975232, and TaqI/rs731236 do not change the protein; they are markers of the 3′ region and ancestry-dependent LD. Cdx2/rs11568820 changes activity of a promoter dependent on the intestinal factor CDX2; its effect may operate through mineral absorption and does not demonstrate ovarian regulation.

4.3 From precursor to active ligand

Serum 25(OH)D is a substrate and systemic exposure, not intracellular calcitriol. CYP27B1 mRNA in human ovarian stroma demonstrates potential capacity, not flux. The chain is validated only if labeled 25(OH)D generates labeled calcitriol, CYP27B1 loss blocks conversion and response, calcitriol bypasses CYP27B1, and VDR loss blocks that bypass.

In murine preantral follicles, VDR was detected without CYP2R1/CYP27B1, and D3 did not alter survival, antral development, or oocyte maturation. This suggests—without proving it in humans—that receptor presence with an inability to activate precursor locally can appear as “vitamin D resistance.”

4.4 Genomic VDR, PDIA3, and stage

The canonical calcitriol–VDR/RXR–VDRE mode is molecularly established. In the human ovary, calcitriol changes AMHR2, FSHR, HSD3B, progesterone, and other programs, but not in one direction. A rapid calcium/AKT response does not identify nuclear VDR: PDIA3 participates in calcitriol signaling in non-ovarian human cells and must be an experimental competitor. Its function in the human ovary remains inferred.

Stage redefines the outcome. Lower primordial activation may increase the remaining pool and reduce throughput; greater antral steroidogenesis may coexist with different effects in theca or luteinized granulosa. Therefore:

primordial pool ≠ AMH ≠ AFC ≠ ovulation ≠ competent oocytes ≠ cumulative output ≠ menopause ≠ healthspan.

4.5 Two provisional metabolic pathways

The published granulosa pathway is:

VDR → ODC1 → putrescine/spermidine → dcSAM/SAM → DNMT → TP53 methylation → p53/p21 → senescence.

The decisive link is dcSAM/SAM flux, not SAM or spermidine alone. Chen et al. did not measure dcSAM directly, and AMD1 knockdown produced a contradiction between DNMT/methylation and p53/p21. The alternative VDR → HMGCR → 7-DHC/redox pathway also produced rescues in related models. This generates a contrary explanation: common stress/cell-cycle state with pleiotropic rescues rather than two exclusive mediators.

The proposed primordial pathway is:

calcitriol → VDR/RXR–p85α → PI3K/AKT/pFOXO3a ↓ → nuclear FOXO3 ↑ → activation ↓.

Co-immunoprecipitation and phosphorylation support association but do not demonstrate PI3K catalytic activity or genetic necessity of VDR in human tissue. The human ex vivo exposure was pharmacological and the donor unit was unclear.

4.6 Mineral/HPO system

In Vdr−/− female mice, a diet that normalized calcium increased fertility from approximately 14% to 86–100%. In Cyp27b1−/−, normalization of Ca/P/lactose reversed HPO, follicular, luteal, and angiogenic abnormalities. These rescues demonstrate that a large fraction of the global-knockout phenotype is systemic; they do not quantify the adult local fraction or its transport to women. Safari et al. 2022 does not resolve the fraction because its arms simultaneously change VDR agonism, calcium, phosphorus, lactose, and drugs.

5. Evidence method

The scoping, evidence-map, verification, mechanism, compute-decision, computation, hypothesis, adversarial-review, and experiment-design artifacts were integrated in full. Decisive sources were traced to primary studies. For each claim, the following were separated:

• human interventional;
• human observational/genetic;
• human ex vivo;
• animal in vivo;
• in vitro;
• computational;
• inference or absence of evidence.

Candidate associations were not considered causal without allelic function, harmonization, HWE, multiplicity control, and replication. Follicles, nuclei, cells, images, and fragments did not substitute for donors or animals as the biological unit. Null results and contradictions were preserved. The 2025 Gholami meta-analysis was excluded as decisive quantitative support because of internal inconsistencies in p values, models, nomenclature, and HWE.

The computational layer reused GSE202601, with four young donors aged 23–29 years and four older donors aged 49–54 years. snRNA and snATAC counts were aggregated by donor×cell. Nuclei were not treated as replicates, and genotypes were not inferred from sparse reads.

6. Evidence map

The strongest evidence differs by level: human evidence for contextual response and null association/intervention; animal evidence for tissue loss and the mineral component; in vitro evidence for candidate mechanisms; computational evidence for identity remodeling; inferred evidence for healthspan.

7. Contradictory evidence and null findings

1. Positive FokI versus a functional null. Promoter/cofactor dependence or true equivalence may explain the tension. The smallest discriminator is reciprocal editing under one prespecified limiting regime, not searching multiple conditions until an interaction appears.
2. Calcitriol increases, does not change, or reduces steroids. Systems differ in cell, stage, luteinization, FSH, insulin, and dose. A universal direction is refuted.
3. The human transcriptome changes without clinical function. SUNDRO demonstrates exposure and transcriptional response but no hormonal or assisted-reproduction improvement.
4. Global knockout versus mineral rescue. Local and systemic VDR can coexist; a global knockout does not estimate ovarian autonomy.
5. Preclinical VDR–ODC1 versus the human atlas. The atlas does not reproduce low VDR; granulosa identity and coverage are more economical explanations.
6. Remaining pool versus output. Lower activation or a later reproductive endpoint did not increase total output in the reviewed model.
7. High AMH versus reserve. In PCOS, high AMH may reflect accumulation of small follicles, not slower depletion.
8. Healthspan mediation versus pleiotropy. Later menopause has divergent effects, and VDR acts directly in systemic organs.

The ranked rival explanations are: granulosa identity/substate; mineral/HPO component; acute local VDR action; common stress/cell-cycle state with pleiotropic rescues; CYP27B1/CYP24A1 gate; PDIA3; contextual FokI; and, last, 3′ markers or longevity mediation.

8. Multiscale mechanistic synthesis

Molecular

Functional availability is the product, not the sum, of calcitriol AUC, VDR protein/localization, RXR/cofactors, chromatin, and CYP24A1 catabolism. A SNP matters only if it changes that surface. VDR RNA is an insufficient proxy; CYP27B1 mRNA does not identify flux.

Cellular

Granulosa may respond through nuclear transcription, polyamine metabolism, mitochondrial function, and steroidogenesis. PDIA3 may explain rapid responses. Cell identity is a potential cause or mediator: conditioning only on FSHR/CYP19A1 may conceal an effect or open a collider, so follicular origin and pretreatment state must be defined independently.

Follicular

In the primordial follicle, decreasing AKT/FOXO3 may slow activation; in the growing follicle, VDR may modulate survival, metabolic support, and endocrine output. This is a biphasic and possibly non-monotonic system. Preservation is defensible only if viability, re-entry, and competence are maintained after ligand withdrawal.

Ovarian/endocrine

Integrated function depends on recruitment, atresia, oocyte competence, ovulation, luteinization, and hormone production. Any mechanism may improve one component and worsen another. The relevant trajectory must be measured within women, not with a single marker.

Systemic

Ca/P/PTH/FGF23, HPO, insulin, inflammation, and angiogenesis may reproduce ovarian phenotypes. At the same time, the E2/P4/androgen/inhibin trajectory may alter bone, metabolism, and cancer. Attribution requires comparing the ovarian-mediated route with direct extraovarian VDR routes.

Life course

There is no single valence of “more ovarian function = more longevity.” The net effect depends on organ, age, competing risks, and hormonal exposure. The scientific endpoint is multi-outcome healthspan, not prolongation of a reproductive proxy.

9. Computational layer

Granulosa snRNA passed the eight-donor gate but contained 1,538 young nuclei versus 208 older nuclei. VDR, ODC1, CYP27B1, and CYP24A1 had 21, 160, 3, and 2 total UMIs in 1,746 nuclei and did not pass filterByExpr; CDKN1A did.

In the low-abundance sensitivity analysis:

• VDR: older/young log2FC +3.40, dependent on ovary52;
• ODC1: −1.17, CI95% −2.74 to +0.40;
• CDKN1A: +0.90, inconclusive;
• SMS: −1.93, panel FDR 0.046;
• FSHR: −4.59, panel FDR 0.000976.

The result does not elevate VDR–ODC1 to H2. It indicates identity and polyamine remodeling without identifying VDR upstream. Granulosa snATAC failed Gate 0—two young donors and one older donor with ≥20 nuclei—and had no promoter peaks for CYP27B1/CYP24A1 or usable motif matrix. RNA–ATAC concordance and the ligand gate are non-evaluable, not negative.

This layer does not measure protein, ligand, flux, occupancy, phosphorylation, minerals, HPO, trajectory, or healthspan. Its value was to change experimental priority, not prove causality.

10. Primary hypothesis

L6-5-AR-H1 v2 — acute and specific local action of VDR

Falsifiable statement: holding pretreatment identity, intracellular exposure, RXR, and PDIA3 constant, an acute and moderate reduction of VDR in human antral granulosa cells will first alter a receptor-specific proximal output and then, if VDR has a local ovarian function, an independent functional output.

Mechanism: intracellular calcitriol → VDR/RXR → occupancy/CYP24A1 → modulation of FSH response and aromatization. ODC1/polyamines is a candidate branch, not part of the definition.

Evidence for: human ex vivo responses; transcriptional target engagement in vivo; Vdr loss in animal granulosa and KGN.

Evidence against: human atlas without low VDR; SUNDRO null for function; mineral rescues; absence of specific perturbation in primary granulosa; possible PDIA3/UPR.

Status: weakened and narrowed.
Maturity: H0–H1.
Calibrated confidence: 0.28 for a measurable local cellular effect, not for benefit or healthspan.

11. Competing hypothesis

L6-5-AR-H2 v2 — identity/substate remodeling dominates

Falsifiable statement: most age-associated differences in ODC1/SMS/CDKN1A will be explained by follicular origin, stimulation, luteinization, cycle, and granulosa substate, not by VDR as the primary driver.

Mechanism: composition/state change → FSHR/SMS/steroidogenic program↓ and altered stress/cycle → secondary correlation with VDR/ODC1/CDKN1A.

Evidence for: donor-aware human reanalysis; marked FSHR loss; state-dependent calcitriol responses.

Evidence against: only eight donors, unequal nuclei, and no replication/informative ATAC; preclinical VDR loss can produce a phenotype.

Status: strengthened as a rival, still not causal.
Maturity: H0–H1.
Calibrated confidence: 0.50.

Preserved contrary attribution hypothesis: L6-5-AR-H3 v2 states that Ca/P/PTH/FGF23 and HPO mediate a large fraction of the phenotype in global knockouts. Its maturity is H1 in global animals but H0 for human dominance. It must be compared in magnitude with adult ovarian loss; it cannot be resolved by adjusting calcium as a covariate.

12. Translational hypothesis

L6-5-AR-HT1 v2 — functional ex vivo response as a trajectory predictor

Falsifiable statement: only after demonstrating local causality and reliability will a frozen ex vivo VDR-response slope add external prediction of the rate of ovarian change beyond age, stage, baseline AMH/AFC, 25(OH)D, and genotype.

Predictions: ICC/CCC and CV will pass blinded gates; gain will persist out of sample and outside IVF; FokI and 25(OH)D alone will not match performance.

Evidence against: there is no standardized assay, transportability, longitudinal cohort, or external validation; the IVF protocol may dominate the slope.

Status: parked; HUMAN_QA_REQUIRED before any individual or partnering assessment.
Maturity: H0.
Calibrated confidence: 0.03.

13. Falsifiable predictions and kill criteria

Global kill criteria: the local program closes if VDR loss/rescue does not modify a proximal output in two backgrounds, if it modifies only CYP24A1 without function, if identity/viability explains the effect, or if adult local loss is equivalent while systemic perturbation reproduces the phenotype. The healthspan route does not open without a local mechanism, trajectory, and organ-specific mediation.

14. Discriminating experiment

ED-0 — qualification

Non-tumoral human mural granulosa from a narrow IVF stratum, with a separate benign-surgery replication. The donor is the unit. Identity is recorded before treatment through follicular origin/diameter, stimulation/luteinization, morphology, and the panel FSHR, CYP19A1, LHCGR, HSD3B1, STAR, AMH, FOXL2, INHA.

Twelve blinded donors qualify the system. FSH response and quantifiable CYP24A1 induction must be observed in ≥10/12, technical CV ≤15%, and viability ≥85%. If it fails, responders are not selected post hoc.

ED-1 — decisive minimum

Within-donor factorial:

• control;
• acute VDR loss;
• acute PDIA3 loss;
• double loss;
• each loss + resistant rescue;
• vehicle or calcitriol 0, 0.1, and 1 nM; 10 nM only as a pharmacological ceiling.

Two orthogonal perturbations must converge. Protein, not only RNA, is measured. Timeline:

• 0–30 min: Ca²⁺, pAKT, pERK, pCAMKII;
• 2–6 h: nuclear VDR, prespecified CUT&RUN/CUT&Tag, and CYP24A1;
• 6–24 h: CYP24A1 AUC, ODC1/SMS, identity, UPR/redox;
• 24–48 h: E2/P4/A4 by LC–MS/MS under FSH+androstenedione, metabolic function, and viability.

Confirmatory hierarchy:

1. P1: CYP24A1 AUC, target engagement;
2. P2: E2 AUC per viable cell, independent function.

P2 is interpreted only if P1 passes. The design uses 20 evaluable donors, with up to 24 recruited, and a blinded variance pilot in eight. The mixed model includes condition, ligand, time, and interactions; donor and plate are random effects. Direction must survive leave-one-donor-out.

Margins: protein loss ≥70%, rescue 80–120%, viability ≥85%, apoptosis <10%; functional equivalence if the complete CI90% of the ratio lies within 0.80–1.25. A CI spanning equivalence and materiality is non-informative.

Conditional sequence

1. ED-1B: labeled precursor, CYP27B1/VDR loss-rescue, and calcitriol bypass; n=12 donors.
2. ED-2: only if ED-1 changes function. Ornithine/methionine and sterol tracers, direct dcSAM/SAM, ODC1/HMGCR loss-rescue, and cross-rescues; n=12–16.
3. ED-3: adult inducible Vdr-granulosa 2×2 model with normal/perturbed mineral homeostasis, Cre/tamoxifen/litter controls, and the same endpoints; pilot 6/cell and planned confirmatory 16/cell if dispersion permits.
4. ED-4: human cortex, exposure–washout, VDR/PDIA3, and balance of pool+progression+atresia+competence; n=10 for a large effect.
5. ED-G: FokI only after a validated functional window; two backgrounds and three nested clones per isoform.
6. T0/D2: metrology and human trajectory; healthspan last.

15. Candidate readouts and stratification

There is no validated ovarian VDR biomarker. The following are candidate experimental readouts, not clinical tools:

• exposure: intracellular/extracellular calcitriol and 25(OH)D by LC–MS/MS;
• receptor: VDR protein/localization, RXR, occupancy at prespecified loci;
• proximal response: CYP24A1 AUC;
• identity: follicular origin, luteinization, and FOXL2/FSHR/CYP19A1/LHCGR/HSD3B1/STAR/AMH/INHA panel;
• function: E2/P4/A4 under a standardized challenge, viability, and metabolic support;
• mechanisms: ODC1/SMS, polyamine flux, SAM/dcSAM/MTA, DNMT/p53/p21, HMGCR/7-DHC/oxysterols, redox;
• system: Ca/P/PTH/FGF23 and LH/FSH;
• future trajectory: repeated ovulation measures, AFC/AMH interpreted jointly, and age at natural menopause.

Primary stratification is by stage, cellular origin, and stimulation—not genotype. PCOS, diminished reserve, POI, and benign surgery are analyzed separately. FokI enters only after function is demonstrated and a limiting regime is defined a priori. Ancestry and LD are mandatory for any genetic replication.

16. Individual variability

Response may vary by:

• age and reproductive stage;
• follicular origin and degree of luteinization;
• FSH/LH, insulin, inflammation, and redox stress;
• abundance/localization of VDR, RXR, PDIA3, and chromatin;
• CYP27B1 capacity and CYP24A1 catabolism;
• 25(OH)D/DBP, season, and exposure, without confusing them with local calcitriol;
• Ca/P/PTH/FGF23, renal/hepatic function, and HPO axis;
• ancestry, haplotypes, and population structure;
• IVF protocol, ischemia-to-culture time, batch, density, and identity drift.

This heterogeneity does not justify post hoc mining. Interactions must be frozen before the experiment and replicated across donors/backgrounds. Variation may reside in a different gate at each stage; it does not imply a universal “good allele.”

17. Pharma relevance and maturity

The scientific opportunity is not “vitamin D for fertility.” It is to discover whether an ovarian window exists that is separable from mineral homeostasis and PDIA3:

Pharma no-go: only CYP24A1 changes; PDIA3/stress explains function; the local component is small relative to systemic; the effect requires a pharmacological dose inseparable from minerals; rescues do not mediate; preservation reduces throughput; or there is no human replication.

No modality is ready for partnering. Even with positive ED-1–ED-4, independent replication, a safety window, and HUMAN_QA_REQUIRED would be necessary; this project does not self-assign H5.

18. Limitations

1. There is no causal perturbation in primary adult human granulosa connecting VDR with an independent function.
2. Human granulosa comes mainly from IVF; stimulation and luteinization limit transportability.
3. KGN is tumoral, and several preclinical panels have n=3.
4. The 2025–2026 mechanistic studies lack independent replication.
5. The dcSAM/SAM pathway was not measured and contains an AMD1 contradiction.
6. There is no human ovarian isotopic conversion of 25(OH)D to calcitriol.
7. The ovarian function of PDIA3 in humans is unproven.
8. The atlas has eight donors, extreme inequality of nuclei, and non-informative granulosa snATAC.
9. Animal mineral rescues mix development, diet, HPO, and multiple organs; Safari 2022 does not isolate calcium.
10. Calcitriol doses in some systems are pharmacological, and physiological tissue exposure is poorly defined.
11. AMH, AFC, activation, oocytes, output, menopause, and healthspan are not equivalent.
12. No cohort simultaneously measures VDR function, ovarian trajectory, and incident systemic outcomes.
13. Later menopause has divergent organ-specific effects; there is no universal net direction.
14. The absence of a large GWAS signal does not exclude small, rare, ancestry-specific, or contextual effects.
15. This report does not estimate therapeutic benefit, dose, or individual risk.

19. Conclusions

1. VDR is biologically relevant in the ovary; a causal common polymorphism has not been identified.
2. FokI is a demonstrable protein difference but a low-priority ovarian functional hypothesis; BsmI/ApaI/TaqI are not mechanisms without fine-mapping.
3. The best causal unit is a response conditioned on identity, active ligand, VDR/RXR/PDIA3, and the mineral/HPO environment.
4. Current human evidence favors granulosa-identity remodeling over low VDR RNA; it does not exclude a local protein/ligand action.
5. The ODC1–dcSAM/DNMT and PI3K–FOXO3 pathways are valuable precisely because they can be falsified, not because they are validated.
6. The systemic mineral contribution is a first-order causal rival in animals; its dominance in women is unknown.
7. ED-1 can kill or narrow the program with a manageable number of donors: target engagement without function is insufficient.
8. Preserving the pool is not equivalent to improving throughput, competence, or healthspan.
9. There is no human evidence of mediation ovarian VDR → women’s longevity.
10. The most likely and useful scientific result may be negative: FokI equivalence, VDR as a passenger of identity, or systemic predominance.

20. References

1. Arai H, Miyamoto K, Taketani Y, et al. A vitamin D receptor gene polymorphism in the translation initiation codon: effect on protein activity and relation to bone mineral density in Japanese women. J Bone Miner Res. 1997;12:915–921. DOI: 10.1359/jbmr.1997.12.6.915. PMID: 9169350.
2. Gross C, Krishnan AV, Malloy PJ, et al. The presence of a polymorphism at the translation initiation site of the vitamin D receptor gene is associated with low bone mineral density in postmenopausal Mexican-American women. J Bone Miner Res. 1998;13:1691–1699. DOI: 10.1359/jbmr.1998.13.11.1691. PMID: 9797477.
3. Yamamoto H, Miyamoto K, Li B, et al. The caudal-related homeodomain protein Cdx-2 regulates vitamin D receptor gene expression in the small intestine. J Bone Miner Res. 2001;16:1256–1264. DOI: 10.1359/jbmr.2001.16.7.1256. PMID: 11450701.
4. Parikh G, Varadinova M, Suwandhi P, et al. Vitamin D regulates steroidogenesis and insulin-like growth factor binding protein-1 in human ovarian cells. Horm Metab Res. 2010. DOI: 10.1055/s-0030-1262837. PMID: 20711952.
5. Merhi Z, Doswell A, Krebs K, Cipolla M. Vitamin D alters genes involved in follicular development and steroidogenesis in human cumulus granulosa cells. J Clin Endocrinol Metab. 2014;99:E1137–E1145. DOI: 10.1210/jc.2013-4161. PMID: 24628555.
6. Xu J, Hennebold JD, Seifer DB. Direct vitamin D3 actions on rhesus macaque follicles in three-dimensional culture. Front Physiol. 2018;9:1600. DOI: 10.3389/fphys.2018.01600. PMID: 30487754.
7. Reginatto MW, Pizarro BM, Antunes RA, et al. Vitamin D receptor TaqI polymorphism is associated with reduced follicle number in women utilizing assisted reproductive technologies. Front Endocrinol. 2018;9:252. DOI: 10.3389/fendo.2018.00252. PMID: 29892263.
8. Szafarowska M, Dziech E, Kaleta B, et al. Anti-Müllerian hormone level is associated with vitamin D receptor polymorphisms in women with polycystic ovary syndrome. J Assist Reprod Genet. 2019. DOI: 10.1007/s10815-019-01472-3. PMID: 31089932.
9. Lin MW, Tsai SJ, Chou PY, et al. Vitamin D receptor polymorphisms and polycystic ovary syndrome. BMC Endocr Disord. 2019;19. DOI: 10.1186/s12902-019-0477-x. PMID: 31870342.
10. Lone NM, Riaz S, Eusaph AZ, et al. Genotype-independent association between vitamin D deficiency and polycystic ovarian syndrome in Lahore, Pakistan. Sci Rep. 2020;10. DOI: 10.1038/s41598-020-59228-4. PMID: 32042037.
11. Day F, Karaderi T, Jones MR, et al. Large-scale genome-wide meta-analysis of polycystic ovary syndrome suggests shared genetic architecture for different diagnosis criteria. PLoS Genet. 2018;14:e1007813. DOI: 10.1371/journal.pgen.1007813. PMID: 30566500.
12. Drakopoulos P, van de Vijver A, Schutyser V, et al. The effect of serum vitamin D levels on ovarian reserve markers: a prospective cross-sectional study. Hum Reprod. 2017;32:208–214. DOI: 10.1093/humrep/dew304. PMID: 27927849.
13. Makieva S, Reschini M, Ferrari S, et al. Oral vitamin D supplementation impacts gene expression in granulosa cells in women undergoing IVF. Hum Reprod. 2021. DOI: 10.1093/humrep/deaa262. PMID: 33305818.
14. Somigliana E, et al. Vitamin D supplementation in women with vitamin D deficiency undergoing IVF: the SUNDRO randomized controlled trial. Am J Obstet Gynecol. 2021. DOI: 10.1016/j.ajog.2021.04.234. PMID: 33894153.
15. Shim YJ, et al. Effect of vitamin D3 on the development of preantral follicles in a murine model. Clin Exp Reprod Med. 2021. PMID: 34875741. PMCID: PMC8651758.
16. Brain CE, et al. Vitamin D receptor and 1α-hydroxylase expression and calcitriol responses in human ovarian tissue and cells. Reproduction. 2025. DOI: 10.1530/REP-25-0002. PMID: 40100123.
17. Chen H, et al. Vitamin D receptor delays ovarian aging by regulating ODC1-mediated spermidine synthesis. Int J Biol Sci. 2026. DOI: 10.7150/ijbs.128631. PMID: 42212335. PMCID: PMC13215359.
18. Liu S, et al. Calcitriol is involved in maintaining primordial follicle reserve through the VDR/RXR–PI3K–AKT–FOXO3 pathway. J Ovarian Res. 2026;19:244. DOI: 10.1186/s13048-026-02135-5. PMID: 42169137.
19. Chen et al. Loss of vitamin D receptor induces premature ovarian insufficiency through compromising the 7-dehydrocholesterol-dependent anti-aging effects. Front Cell Dev Biol. 2025. DOI: 10.3389/fcell.2025.1545167. PMCID: PMC12018433.
20. Albaugh BN, et al. Difluoromethylornithine rebalances aberrant polyamine ratios in Snyder–Robinson syndrome. EMBO Mol Med. 2023. DOI: 10.15252/emmm.202317833. PMCID: PMC10630878.
21. Nowak JI, et al. Deletion of VDR and PDIA3 disrupts membrane-initiated vitamin D signaling in human A431 cells. Cells. 2024;13:11. DOI: 10.3390/cells13010011. PMID: 38201216.
22. Johnson LE, DeLuca HF. Vitamin D receptor null mutant mice fed high levels of calcium are fertile. J Nutr. 2001;131:1787–1791. DOI: 10.1093/jn/131.6.1787. PMID: 11385068.
23. Sun W, Xie H, Ji J, et al. Defective female reproductive function in 1,25(OH)2D-deficient mice results from indirect effect mediated by extracellular calcium and/or phosphorus. Am J Physiol Endocrinol Metab. 2010;299:E928–E935. DOI: 10.1152/ajpendo.00378.2010. PMID: 20807842.
24. Safari H, et al. The role of calcium and vitamin D in female reproductive function in mice. Sci Rep. 2022. DOI: 10.1038/s41598-022-14708-7. PMID: 35729248.
25. Jin et al. Molecular and genetic insights into human ovarian aging from single-nuclei multi-omics analyses. Nat Aging. 2025. DOI: 10.1038/s43587-024-00762-5. PMID: 39578560. GEO: GSE202601.
26. Ruth KS, Day FR, Hussain J, et al. Genetic insights into biological mechanisms governing human ovarian ageing. Nature. 2021;596:393–397. DOI: 10.1038/s41586-021-03779-7. PMID: 34349265.
27. Dvornyk V, Long JR, Liu PY, et al. Polymorphisms of the vitamin D receptor gene predict the age of natural menopause and surgical menopause. Gynecol Endocrinol. 2006. DOI: 10.1080/09513590600988258. PMID: 17135034.
28. Gholami M, et al. Vitamin D receptor polymorphisms and polycystic ovary syndrome: a meta-analysis. Metabolism Open. 2025;25:100343. DOI: 10.1016/j.metop.2024.100343. PMID: 39866289. Used only as an audit object, not as a reliable quantitative estimator.

Integrative claims of REPORT_EN

1. L6-5-REPORT-EN-C1: No reviewed common VDR polymorphism has a complete human causal chain from ovarian function to healthspan; FokI is an experimental tool, not a validated determinant.
2. L6-5-REPORT-EN-C2: The defensible causal unit is a four-gate surface—identity/stage, active ligand, VDR/RXR/PDIA3, and mineral/HPO environment—and no isolated measurement represents the system.
3. L6-5-REPORT-EN-C3: The human atlas weakens low VDR RNA→ODC1 and elevates granulosa-identity remodeling; it does not refute protein/ligand action or demonstrate systemic dominance.
4. L6-5-REPORT-EN-C4: A local VDR action is accepted only if loss/rescue changes target engagement and then an independent ovarian function; CYP24A1 without function is insufficient.
5. L6-5-REPORT-EN-C5: Preserving the pool or delaying menopause has no uniform longevity valence; mediation must be longitudinal, organ-specific, and compete with extraovarian VDR pathways.