🧬 Aging Wiki

fsh

Properties1
aliasesfollicle-stimulating hormone, follitropin, FSHB, FSH, FSH-beta

17 min read

FSH (FSHB / follicle-stimulating hormone)

Follicle-stimulating hormone is a pituitary glycoprotein gonadotropin secreted by anterior-pituitary gonadotropes under the control of hypothalamic GnRH. Its central aging relevance is dual: (1) as the earliest rising endocrine marker of the menopausal transition — reflecting declining ovarian reserve before estradiol falls — and (2) as the subject of a contested but rapidly developing hypothesis that elevated FSH exerts direct, estrogen-independent effects on bone, fat, and brain in post-reproductive life.

Identity

  • UniProt: P01225 (FSHB_HUMAN) — beta subunit; this is the specificity-conferring chain
  • Alpha subunit: P01215 (GLHA_HUMAN / CGA) — shared with LH, TSH, and hCG; not listed as primary because the beta subunit confers receptor selectivity
  • NCBI Gene (FSHB): 2488
  • HGNC: 3964
  • Ensembl: ENSG00000131808
  • Mouse ortholog: Fshb (NCBI Gene 14308)
  • Protein length: 129 amino acids (beta subunit, canonical isoform); signal peptide residues 1-18; active chain 19-129
  • Molecular weight: ~14.7 kDa (unglycosylated beta chain); intact glycosylated heterodimer ~30 kDa

Structure and biochemistry

FSH is an obligate non-covalent heterodimer. The alpha subunit (CGA, 92 aa) is identical across the four glycoprotein hormones; the beta subunit (FSHB, 111 aa active chain) determines receptor specificity via its unique “seat-belt” loop that clamps around the alpha subunit 1.

Key structural features of the FSHB subunit:

  • Cystine-knot motif — three disulfide bonds forming the growth factor superfamily fold; shared with the alpha subunit (which contributes an additional three)
  • N-linked glycosylation at Asn-25 and Asn-42 — glycan composition is heterogeneous and modulates bioactivity; pituitary FSH isoforms show acidic (more sialylated, longer half-life) and basic (less sialylated, higher bioactivity) forms
  • Seat-belt region — the C-terminal tail of FSHB wraps around the alpha subunit and is critical for dimerization; mutations here cause hypogonadotropic hypogonadism 24 (HH24)

Regulation and secretion

Synthesis and release from anterior-pituitary gonadotrope cells is governed by a regulatory hierarchy:

  1. GnRH (hypothalamus) — pulsatile GnRH stimulates FSH and LH release; frequency of GnRH pulses differentially regulates LH vs. FSH (slower pulses favor FSH)
  2. Inhibin B (ovary / testis) — the most potent specific suppressor of FSH; produced by granulosa cells (follicular phase) in females and by Sertoli cells in males. Inhibin B selectively suppresses FSH without meaningfully affecting LH. Inhibin B decline is the primary driver of FSH rise in the menopausal transition
  3. Estradiol — exerts negative feedback at both hypothalamus and pituitary; a brief positive-feedback surge triggers the preovulatory LH/FSH surge; in the menopausal transition, the loss of estrogen feedback amplifies FSH rise but is temporally secondary to inhibin B loss
  4. Activin — FSH-stimulating paracrine/autocrine signal at the pituitary; acts via SMAD2/3

Role in classic gonadal physiology

Female: folliculogenesis and estradiol synthesis

FSH binds FSHR (a G-protein-coupled receptor) on granulosa cells, stimulating:

  • Follicle recruitment and growth (antral follicle development)
  • Aromatase (CYP19A1) expression → conversion of androgen to estradiol
  • Inhibin B and AMH (amh) production
  • LH receptor upregulation (priming granulosa cells for ovulation)

Male: spermatogenesis support

In males, FSH binds FSHR on Sertoli cells, supporting:

  • Sertoli cell proliferation and survival
  • Spermatogonial proliferation and differentiation
  • Inhibin B secretion (male FSH negative feedback)

FSH levels rise modestly through male reproductive aging as Sertoli-cell number and inhibin B decline, though the magnitude is smaller and more gradual than the female menopausal transition 2.

FSH as a biomarker of reproductive aging

The menopausal transition (females)

The menopausal transition is the central aging event for this protein. The STRAW+10 (Stages of Reproductive Aging Workshop) staging system formalizes reproductive aging across menstrual, endocrine, and ovarian markers 3.

Key endocrine sequence:

  1. Earliest change: Inhibin B falls (before cycles become irregular), releasing FSH from feedback inhibition → FSH rises first — detectable years before the final menstrual period and before estradiol falls significantly
  2. Early transition: FSH levels become variable and elevated (>25 IU/L on day 2-3), menstrual cycle length becomes variable
  3. Late transition: estradiol declines, FSH rises steeply and persistently (typically >40 IU/L)
  4. Postmenopause: FSH peaks in the first 2-3 years (~50-100 IU/L range), then modestly declines; LH rises more slowly and less dramatically 4

Clinical caveats: FSH fluctuates considerably in the early transition; a single elevated FSH does not confirm perimenopause. AMH (amh) and antral follicle count may predict transition timing more reliably in early stages. Assays measure the intact heterodimer; the alpha subunit can circulate independently (biologically inactive) and interfere with some assays. dose-response-unclear for the exact FSH threshold that marks pathological transition in diverse populations.

Male reproductive aging

FSH rises gradually across the male lifespan. In the European Male Ageing Study (EMAS) cohort, higher gonadotropin levels (FSH and LH) correlated with higher frailty index scores independent of testosterone level, consistent with Sertoli-cell dysfunction as an aging phenotype 2. FSH rise in men is neither as steep nor as clearly staged as in women.

THE CONTESTED FRONTIER: FSH direct actions on non-gonadal tissue

This is the most dynamically developing area of FSH biology and the one most directly relevant to the aging hallmark of altered-intercellular-communication. The FSH-direct-action hypothesis proposes that the high FSH of post-reproductive life has estrogen-independent effects on bone, adipose tissue, and brain — raising the possibility that FSH itself, not simply estrogen deficiency, drives part of the phenotypic aging cluster of menopause.

Important framing: The majority of experimental evidence comes from mouse models, frequently from the Zaidi laboratory at Mount Sinai. Some findings have limited or zero human replication. The hypothesis is actively contested and should be read with this in mind.

Bone: FSH stimulates osteoclastogenesis

The seminal claim: Sun et al. 2006 (Cell) showed that FSHR is expressed on osteoclast precursors in mouse bone marrow, and that FSH directly stimulates osteoclast formation and activity via G_i2α-coupled FSHR → MEK/ERK, NF-κB, and Akt signaling 5. Two lines of genetic evidence support estrogen-independent FSH action: (1) FSHR−/− mice are severely hypogonadal yet do not lose bone (areal and volumetric BMD indistinguishable from WT controls), indicating hypogonadism alone is insufficient to drive bone loss without FSH signaling; and (2) FSHβ+/− mice are eugonadal (normal ovaries and estrogen levels) yet have increased bone mass and reduced serum TRAP compared with WT, demonstrating that partial lowering of FSH in a normal estrogen environment is sufficient to improve bone. Together these data suggest FSH exerts direct osteoclast-stimulating effects independent of gonadal status. A subsequent study showed FSH also promotes TNF-α production from bone marrow immune cells, expanding the osteoclast precursor pool 6.

The Zaidi group developed MS-Hu6, a humanized monoclonal anti-FSH antibody, demonstrating in mouse models that it prevents ovariectomy-induced bone loss and stimulates new bone formation 7. Gera et al. 2022 also characterize MS-Hu6’s pharmacokinetics and safety in monkeys; prevention of adipose gain (from Liu 2017 8) and neurodegeneration (from Xiong 2022 9) were established in earlier papers using polyclonal precursors, not MS-Hu6 itself.

The controversy: Several replication and mechanistic concerns have been raised:

  • The magnitude of FSH-driven bone loss vs. estrogen-deficiency bone loss is debated; in humans, the early perimenopausal bone loss that begins when inhibin B falls (and FSH rises, while estradiol is still largely maintained) is modest, and it is difficult to dissect FSH effects from subtle estradiol changes
  • FSHR expression on osteoclast precursors has been confirmed in some studies but its level and functional significance are disputed; some groups report much lower or absent osteoclast FSHR expression
  • Human GWAS data does not clearly identify FSHR loci as major determinants of bone mineral density in a direction consistent with the direct-FSH-action hypothesis (though power for this specific signal is limited)
  • needs-human-replication — no published human interventional study directly tests FSH blockade effects on bone density independent of estrogen

Adipose tissue and metabolism: FSH blockade reduces adiposity

Liu et al. 2017 (Nature) reported that a polyclonal FSH-blocking antibody in mice markedly reduced adipose tissue mass, induced beiging of white adipose tissue (WAT), activated brown adipose tissue (BAT), increased thermogenesis and mitochondrial density, and reduced fat mass 8. The mechanism: adipocyte FSHR couples to G_i (not G_s as in granulosa cells), suppressing cAMP and thereby inhibiting Ucp1/thermogenic gene expression; the antibody relieves this suppression. The effects were observed in both male and female C57BL/6J mice on a high-fat diet, as well as in sham-operated (eugonadal) and ovariectomized females — the reduction in fat mass in gonadally intact animals strengthens the claim that FSH acts directly on adipocytes rather than solely via estrogen deficiency. The paper also introduced a monoclonal antibody (Hf2) targeting the human FSHβ epitope that phenocopied the polyclonal antibody’s anti-adiposity effects in mice.

Human relevance: The fat redistribution and weight gain of the menopausal transition (increased abdominal adiposity, decreased lean mass) is a major driver of cardiometabolic risk. If FSH directly contributes, FSH blockade — rather than or in addition to hormone therapy — could address this. However, this remains entirely preclinical. needs-human-replication

Xiong et al. 2022 (Nature) demonstrated that FSH acts on hippocampal and cortical neurons via FSHR in those cells, activating the C/EBPβ → AEP (asparagine endopeptidase, a δ-secretase) pathway, which promotes amyloid-β and tau cleavage and deposition in mouse models of Alzheimer’s disease 9. The paper used two distinct AD mouse models (3xTg-AD mice ovariectomized to elevate FSH, and APP/PS1 male mice injected with exogenous FSH); a polyclonal FSH-blocking antibody (the same epitope-specific anti-FSHβ Ab used in prior bone/fat studies, not yet MS-Hu6) reversed cognitive deficits in both models. Hippocampal Fshr knockdown by stereotactic AAV2-siFshr injection reproduced the antibody’s protective effects, directly establishing the neuronal FSHR as the relevant target. The paper proposed this as a mechanistic explanation for the female sex bias in Alzheimer’s disease (women are approximately twice as likely as men to develop AD, beyond longevity differences), and explicitly showed FSH effects were independent of estrogen by clamping estrogen in ovariectomized mice.

A 2024 commentary in Nature Reviews Endocrinology by Sauvé, Kacimi, and Prévot (an independent French group, not the Zaidi laboratory) synthesized the Xiong 2022 findings and the subsequent APOE4 interaction work, framing elevated FSH as a catalyst for Alzheimer’s neuropathology in the context of the menopausal HPG-axis shift 10. The review is supportive of the FSH hypothesis rather than cautionary; it highlights both FSH-blockade and pulsatile GnRH replacement as therapeutic avenues, and notes that the APOE4 interaction was observed selectively in female mice expressing human APOE4 (not APOE3). The primary caveat acknowledged is that all mechanistic evidence remains preclinical.

DimensionStatusNotes
FSHR expressed in neurons?partialXiong 2022 confirmed Fshr mRNA (qPCR, RNAscope) in mouse and human cortex and SH-SY5Y neuroblastoma cells, and ~85 kDa FSHR protein in mouse brain by western blot; expression absent in Fshr−/− mice (specificity control); expression level and functional significance in adult human neurons disputed by independent groups contradictory-evidence
FSH→AD pathway conserved in humans?not testedNo human interventional study; epidemiological associations only
Replicated in humans?noAll mechanistic evidence is mouse-only; needs-human-replication
Replicated by independent labs?limitedMost direct evidence from Zaidi/Ye cluster; Sauvé 2024 (Lille group) synthesized and endorsed the framework; Cheliadinova 2026 (Mol Psychiatry, PMID 41402481, also Zaidi group) replicated with Fshr-KO in AD mice

contradictory-evidence — The expression level and functional significance of FSHR in adult human neurons is contested. Xiong 2022 shows mRNA and protein evidence; however, independently produced human single-cell atlas data (not cited in that paper) shows low FSHR in neurons, and the C/EBPβ→AEP pathway has not been independently confirmed in human neurons.

needs-human-replication — The female-bias-in-AD claim is plausible and well-framed, but the causal FSH→brain pathway remains mouse-only.

Integrated review and therapeutic horizon

Kim et al. 2024 (Journal of Endocrinology, Zaidi group) reviewed the evidence across all three non-gonadal axes (“FSH, bone, belly and brain”), concluding that preclinical genetic and pharmacologic data converge to support FSH-independent contributions to postmenopausal bone loss, adiposity, and cognitive decline 11. The review characterized MS-Hu6 (humanized anti-FSH antibody) as positioned for human testing following favorable pharmacokinetic and safety profiling 7. No phase-1 human trial of FSH-blocking monoclonal antibody therapy has been registered as of 2026-06-03. long-term-unknown

A 2026 review in Yale Journal of Biology and Medicine extended the framework across the lifespan, noting that gonadotropin dysregulation — including FSH elevation — spans not just menopause but a neurodegeneration-neurodevelopment continuum 12.

Pathways and signaling

  • Canonical gonadal axis: gnrh-signaling → gonadotrope FSH/LH release → FSHR (Gs/Gi-coupled GPCR on granulosa cells / Sertoli cells / osteoclast precursors / adipocytes / neurons per hypothesis) → cAMP/PKA, PI3K-Akt, MEK-ERK, NF-κB downstream
  • Inhibin B feedback loop: granulosa cell inhibin B → pituitary SMAD2/3 → suppresses FSH synthesis (FSH-specific; LH unaffected)
  • GnRH pulse-frequency encoding: slow GnRH pulses (>90-min intervals) favor FSH over LH synthesis by altering AP-1 and SF-1 transcription

Disease and loss-of-function genetics

  • Hypogonadotropic hypogonadism 24 (HH24, OMIM #229070): homozygous or compound-heterozygous FSHB mutations cause primary amenorrhea in females (no folliculogenesis, infertility) and severe oligoazoospermia with infertility in males; LH and estradiol/testosterone may be normal, consistent with FSH specificity for Sertoli/granulosa cells
  • Clinical FSH deficiency: low FSH (with normal or elevated LH) suggests selective FSH/FSHB deficiency; very rare
  • Pharmacological FSH (Gonal-F, Puregon, Bravelle) — recombinant or urinary FSH used for ovarian stimulation in assisted reproduction; well-established clinical record

Pharmacology and aging-context druggability

Druggability tier: 2 — No FSH-modulating drug is approved for an aging indication. GnRH agonists (leuprolide) and antagonists (degarelix, relugolix) suppress FSH as part of broad gonadotropin suppression, used in oncology (prostate cancer, hormone-receptor-positive breast cancer) and endometriosis. These have established clinical experience but cause menopausal symptoms and are not an aging-rejuvenation tool.

The direct anti-FSH monoclonal antibody approach (MS-Hu6, Zaidi group) targets FSH specifically, leaving LH intact — a pharmacologically important distinction. Preclinical mouse data suggest this could simultaneously address bone loss, adiposity, and perhaps cognition. The tier-2 designation reflects that a high-quality pharmacological probe (MS-Hu6) exists with published characterization 7, but no human trial data.

Aging-context note: The prospect of blocking FSH while leaving LH (and potentially testosterone/estrogen) relatively unaffected distinguishes this approach from conventional hormone therapy. Whether MS-Hu6 or a similar agent will replicate mouse findings in humans is entirely open.

  • lh — co-secreted gonadotropin sharing the CGA alpha subunit; LH drives theca cell androgen production (females) and Leydig cell testosterone (males); rises in parallel with FSH at menopause but less steeply
  • amh — anti-Müllerian hormone; produced by granulosa cells of pre-antral and small antral follicles; inversely correlates with FSH; better predictor of ovarian reserve in early transition
  • estradiol — FSH feedback target and effector; declines later in the transition than FSH rises
  • inhibin — primary FSH negative feedback; inhibin B produced by follicles and Sertoli cells; inhibin A by corpus luteum; both selectively suppress FSH
  • granulosa-cells — primary cellular target of FSH in the ovary
  • alzheimers-disease — proposed downstream effect of elevated FSH via neuronal FSHR
  • osteoporosis — postmenopausal bone loss with FSH-direct-action hypothesis as contributing mechanism
  • menopause — the physiological state in which FSH elevation is most pronounced and clinically relevant

Knowledge gaps

  • needs-human-replication — All non-gonadal FSH-direct-action evidence (bone, fat, brain) is currently mouse-only; no human RCT of FSH blockade for any aging indication is registered
  • contradictory-evidence — FSHR expression on human osteoclasts, adipocytes, and neurons is contested; expression levels may be insufficient for meaningful direct signaling
  • needs-replication — The Xiong 2022 FSH→AD findings (neuronal FSHR → C/EBPβ → amyloid/tau) await independent replication in a different lab
  • long-term-unknown — Long-term effects of sustained FSH blockade on gonadal function, bone, cardiovascular risk, and cognition in humans are unknown
  • dose-response-unclear — The FSH concentration at which non-gonadal direct effects reach physiological significance is not established in humans; mouse studies used antibody doses that suppress FSH substantially more than the postmenopausal rise

Footnotes

Footnotes

  1. doi:10.1093/humrep/9.7.1335 · Flack MR et al. · Hum Reprod 1994 · 9(7):1335-44 · review · FSH heterodimer structure, seat-belt model, cystine-knot superfamily; glycan heterogeneity and bioactivity isoforms ↩

  2. doi:10.1111/j.1532-5415.2011.03398.x · Tajar A et al. (EMAS) · J Am Geriatr Soc 2011 · 59(5):814-21 · observational · n=3369 men aged 40-79 · higher gonadotropins (FSH, LH) associate with higher frailty index independent of testosterone · not full-text verified — PDF download failed; n and key finding from abstract no-fulltext-access ↩ ↩2

  3. doi:10.1210/jc.2011-3362 · Harlow SD et al. (STRAW+10 Collaborative Group) · J Clin Endocrinol Metab 2012 · observational / consensus staging · STRAW+10 staging system; FSH >25 IU/L (variable) marks early transition; inhibin B decline precedes estradiol decline; 1196 citations ↩

  4. doi:10.1016/j.amjmed.2005.09.067 · Santoro N · Am J Med 2005 · 118(12B):16S-20S · review · progressive FSH rise; most estradiol decline occurs in late transition, not early; inhibin B fall is primary FSH driver · PMID 16414322 ↩

  5. doi:10.1016/j.cell.2006.01.051 · Sun L et al. (Zaidi lab) · Cell 2006 · 125(2):247–260 · in-vivo (mouse: FSHR−/−, FSHβ+/−, FSHβ−/−, and WT; also ex-vivo osteoclast cultures from human and mouse) · FSH directly regulates bone mass via Gi2α-coupled FSHR on osteoclast precursors (CD11b+) → MEK/ERK, NF-κB, Akt → enhanced osteoclastogenesis; FSHR−/− hypogonadal mice do not lose bone (BMD indistinguishable from WT); FSHβ+/− eugonadal mice have increased BMD and reduced serum TRAP vs WT; FSH stimulates osteoclastogenesis at 3 ng/mL in human and mouse cultures · n = 4–14 mice/group per subexperiment (stated in figure legends); 8 cross-sections from 4 animals per histomorphometry group ↩

  6. doi:10.1073/pnas.0605717103 · Iqbal J et al. (Zaidi lab) · PNAS 2006 · 103(40):14925-30 · in-vivo (mouse) · FSH stimulates TNF-α production from bone marrow immune cells, expanding osteoclast precursor pool · PMID 17003115 ↩

  7. doi:10.7554/eLife.78022 · Gera S et al. (Zaidi lab) · eLife 2022 · 11:e78022 · in-vivo (mouse + Cynomolgus monkey) + biophysical characterization · MS-Hu6 humanized anti-FSH antibody; K_D 7.52 nM (surface plasmon resonance); β-phase t½ 7.5 days (180 hr) in humanized Tg32 mice (express human FcRn); prevents ovariectomy-induced bone loss and stimulates new bone formation in mice; confirmed bone marrow and adipose tissue biodistribution; acute safety established in two Cynomolgus monkeys; fat/neurodegeneration prevention from earlier polyclonal Ab studies (Liu 2017; Xiong 2022) — not from MS-Hu6 itself in this paper · positioned for first-in-human osteoporosis trial ↩ ↩2 ↩3

  8. doi:10.1038/nature22342 · Liu P et al. (Zaidi lab) · Nature 2017 · 546(7656):107–112 · in-vivo (mouse: 3-month-old male and female C57BL/6J on high-fat diet; also ovariectomized and sham-operated females on normal chow) · polyclonal anti-FSHβ antibody (200 µg/mouse/day i.p.) reduces total fat volume, subcutaneous and visceral adiposity, induces WAT beiging (Ucp1 upregulation), activates BAT thermogenesis, increases mitochondrial density; adipocyte FSHR couples to G_i (not G_s), suppressing cAMP and Ucp1; effects seen in both sexes and in gonadally intact (eugonadal) animals; also introduced monoclonal Hf2 targeting human FSHβ epitope with similar anti-adiposity effects · n = 3–8 mice/group per subexperiment ↩ ↩2

  9. doi:10.1038/s41586-022-04463-0 · Xiong J et al. (Zaidi lab + Keqiang Ye lab) · Nature 2022 · 603(7901):470–476 · in-vivo (mouse: 3xTg-AD ovariectomized females; APP/PS1 male mice injected with exogenous FSH; also primary neurons and SH-SY5Y cells) · FSH (30 ng/mL) acts on hippocampal/cortical neurons via FSHR → G_i → AKT/ERK → C/EBPβ phosphorylation → AEP (arginine endopeptidase, a δ-secretase) activation → cleaves APP and Tau → amyloid-β and tau deposition; polyclonal anti-FSHβ Ab (200 µg i.p. every 2 days, 8 weeks) reverses cognitive deficits (Morris water maze) in ovariectomized 3xTg mice; hippocampal AAV2-siFshr knockdown phenocopies antibody; FSHR confirmed in human cortex and neuroblastoma SH-SY5Y; effects independent of estrogen (clamped with E2 pellet in separate experiment) · no human evidence ↩ ↩2

  10. doi:10.1038/s41574-024-00981-1 · Sauvé F, Kacimi L, Prévot V (University of Lille; independent of Zaidi group) · Nat Rev Endocrinol 2024 · 20(6):317–318 · commentary (2 pp.) · synthesizes Xiong 2022 and 2023 APOE4 follow-up findings; supportive of FSH-blockade and pulsatile GnRH therapy as therapeutic avenues; notes APOE4 interaction is selective for female mice; confirms all evidence preclinical; primary caveat = no human interventional studies ↩

  11. doi:10.1530/JOE-23-0377 · Kim SM et al. (Zaidi lab) · J Endocrinol 2024 · 262(1) · review · synthesizes preclinical evidence for FSH direct action on bone, fat, and brain; frames MS-Hu6 as ready for human testing · not full-text verified — closed-access, PDF unavailable; claims from this source are abstract/title-level only no-fulltext-access ↩

  12. doi:10.59249/EVST3084 · Galbraith JA et al. · Yale J Biol Med 2026 · review · gonadotropins across the lifespan; FSH/LH receptors in hippocampus and cortex; menopause-associated gonadotropin elevation linked to brain connectivity changes, amyloid/tau, and cognition ↩

Graph View

Backlinks