HMG

What is HMG?#

HMG is a peptide that has been studied in preclinical and clinical research models for its potential therapeutic properties.

Mechanism of Action#

We interpret HMG here as High Mobility Group Box 1 (HMGB1), a dual-function nuclear protein that acts extracellularly as an alarmin/DAMP. Its mechanism of action is determined by redox state, binding partners, and receptor context, yielding distinct signaling programs and cellular outcomes.

Ligand forms and complexes

• Redox isoforms: (a) All-thiol (fully reduced) HMGB1 is chemotactic; (b) Disulfide HMGB1 (Cys23–Cys45 disulfide with reduced Cys106) is cytokine-inducing; (c) Fully oxidized/sulfonyl HMGB1 is functionally inert/tolerogenic. Post-translational modifications (e.g., hyperacetylation, phosphorylation) regulate nuclear export and secretion, while redox state dictates receptor preference and function (chemokine versus cytokine).
• Functional complexes: HMGB1 forms heterocomplexes with CXCL12 (SDF-1), LPS and other PAMPs, DNA/RNA/CpG, nucleosomes, and cytokines such as IL-1. These complexes alter receptor usage and amplify signaling; HMGB1 often serves as a chaperone to deliver ligands to endosomal or cytosolic sensors.

Receptor interactions

• TLR4/MD-2: Disulfide HMGB1 binds MD-2 within the TLR4 complex. Engagement activates TIRAP→MyD88→NF-κB for proinflammatory cytokines, and TRAM→TRIF→IRF3 for type I interferon responses. Binding depends on Cys106 in the B box, and the cytokine activity requires the Cys23–Cys45 disulfide.
• TLR2 and TLR9 (endosomal): HMGB1 can signal via TLR2 to induce NF-κB-dependent cytokines and via TLR9 when complexed with DNA/CpG, typically after RAGE-mediated uptake, using the MyD88 axis.
• RAGE: A promiscuous immunoglobulin-like receptor recognizing HMGB1 regions around aa23–50 and aa150–183. RAGE engagement activates small GTPases (Rac1/Cdc42), Ras–ERK1/2, PI3K–Akt, JNK, p38, and NF-κB; critically, RAGE mediates endocytosis of HMGB1 and HMGB1–PAMP/DAMP complexes into endolysosomes.
• CXCR4: All-thiol HMGB1 forms a heterocomplex with CXCL12 that potently activates CXCR4, driving chemotaxis independent of TLR4/RAGE during early recruitment.
• Regulatory receptors and co-receptors: CD24/Siglec-10 and TIM-3 can restrain HMGB1-driven TLR/RAGE signaling; CD14 can facilitate HMGB1–LPS/TLR4 responses; integrins (e.g., Mac-1) and proteoglycans contribute to binding/trafficking.

Pathways and signaling programs

• Canonical TLR signaling: TLR4–MD-2 engagement by disulfide HMGB1 activates MyD88→IRAK→TRAF6→NF-κB and MAPKs (p38, JNK, ERK) to induce TNF, IL-1β, IL-6; TRIF-dependent signaling induces IRF3 and IFN-β.
• RAGE signal transduction: RAGE ligation activates Rac1/Cdc42, Ras→ERK1/2, PI3K–Akt, JNK and p38, culminating in NF-κB activation; endothelial responses include upregulation of ICAM-1/VCAM-1/E-selectin and barrier dysfunction via p38.
• RAGE-mediated endocytosis and lysosomal permeabilization: HMGB1 (often as complexes with LPS or nucleic acids) is internalized via RAGE to acidic lysosomes, where HMGB1 destabilizes lysosomal membranes, allowing partner ligands to escape. Cytosolic LPS activates noncanonical inflammatory caspases (human caspase-4/5; murine caspase-11), while cathepsin release and cytosolic PAMPs/DAMPs activate inflammasomes and caspase-1, culminating in gasdermin D–dependent pyroptosis and robust cytokine release.
• Autophagy: HMGB1 binds beclin-1, promoting autophagosome formation and supporting cell survival; extracellular HMGB1–RAGE can induce autophagy in neighboring cells.

Cellular outcomes and molecular targets

• Cytokine/chemokine production: TNF, IL-1β, IL-6, type I IFN (IFN-β) via NF-κB and IRF3 pathways downstream of TLR4/MD-2 and TLR9.
• Chemotaxis and recruitment: All-thiol HMGB1–CXCL12→CXCR4 axis drives leukocyte recruitment and tissue repair programs.
• Endothelial activation and permeability: HMGB1–RAGE increases ICAM-1, VCAM-1, E-selectin and promotes p38-dependent barrier dysfunction.
• Autophagy: HMGB1–beclin-1 interaction and RAGE signaling promote autophagy; intracellular HMGB1 supports autophagy and limits apoptosis.
• Pyroptosis: RAGE-mediated uptake of HMGB1 complexes enables cytosolic sensing of PAMPs and activation of caspase-4/5/11 and caspase-1, leading to gasdermin D pore formation and lytic inflammatory death with HMGB1 release, creating feed-forward amplification.

Key nodes and domains

• Receptors: TLR4–MD-2, TLR2, TLR9; RAGE; CXCR4; regulatory CD24/Siglec-10, TIM-3; integrins (Mac-1).
• Adaptors/signaling: MD-2; TIRAP/MyD88 and TRAM/TRIF; IRAK/TRAF6; Rac1/Cdc42, Ras–ERK1/2, PI3K–Akt; beclin-1 complex.
• Effector enzymes: Caspase-1, -4/5/11; gasdermin D; cathepsins; kinases p38/JNK/ERK; transcription factors NF-κB and IRF3.
• HMGB1 structural determinants: Cys23, Cys45, Cys106 redox states; RAGE-binding regions around aa23–50 and aa150–183; TLR4 interaction requires C106 thiol in B box and disulfide C23–C45.

Integrated mechanism Upon stress or damage, HMGB1 exits the nucleus via PTM-driven export and is released passively (necrosis) or actively (secretory lysosomes, pyroptosis). In the extracellular space, its redox state and partners determine receptor engagement: all-thiol HMGB1 forms a heterocomplex with CXCL12 to activate CXCR4 and drive chemotaxis; disulfide HMGB1 binds TLR4/MD-2 to trigger MyD88- and TRIF-dependent NF-κB, MAPK and IRF3 signaling, inducing cytokines and type I interferon; HMGB1–PAMP/DAMP complexes are internalized via RAGE, where HMGB1 permeabilizes lysosomes, enabling cytosolic access of ligands to activate noncanonical caspases and inflammasomes, leading to gasdermin D–mediated pyroptosis. HMGB1 also binds beclin-1 to promote autophagy and can modulate endothelial activation via RAGE and p38. These interconnected mechanisms explain how HMGB1 functions as an amplifier of sterile and infectious inflammation, a chemotactic cue, and an autophagy regulator.

Embedded summary table of HMGB1 forms, receptors, pathways, and outcomes:

Preclinical Evidence#

Preclinical evidence

• In a mouse superovulation model, hMG plus hCG elicited a higher ovarian response and oocyte recovery versus pure FSH plus hCG: response 46% (32/70) vs 22.7% (15/66); oocytes recovered 454/32 (~14.2/mouse) vs 77/15 (~5.1/mouse). Corrected in vitro fertilization rates after hMG superovulation were ~58% at 12 h post‑hCG, with oocyte aging reducing fertilization over time.

• Across IVF/ICSI, hMG achieves live birth rates comparable to rFSH in RCTs, with similar OHSS risk in standard responders; rFSH generally yields more oocytes with lower total dose, while hMG sometimes shows modest advantages in implantation/pregnancy in specific analyses.

• For anovulatory PCOS outside IVF, urinary hMG and rFSH provide similar live birth; adding letrozole to HMG can increase monofollicular development and reduce gonadotropin dose without compromising pregnancy.

• In male HH, hMG (with hCG) supports testicular growth and spermatogenesis in a majority of patients over prolonged therapy; small controlled data show similar spermatogenesis rates across regimens but higher androgen levels with combination therapy.

Embedded summary table of representative studies and outcomes:

Research Evidence Quality#

Extent and quality of the evidence base • Randomized controlled trials (RCTs): Numerous small‑to‑moderate RCTs from the 2000s–2020s compare HP‑hMG to recombinant FSH (rFSH), usually under GnRH‑agonist or antagonist protocols. Trial endpoints vary; many prioritize oocyte yield or clinical pregnancy rather than live birth. A recent assessor‑blinded RCT in high responders reported similar cumulative live birth between HP‑hMG and rFSH but lower OHSS with HP‑hMG in that population (and similar or lower pregnancy loss), underscoring population‑ and protocol‑specific effects (high responders, antagonist). • Meta‑analyses: Older and mid‑period meta‑analyses consistently find rFSH yields more oocytes at a lower total gonadotropin dose than HP‑hMG, but with no clear live‑birth advantage once baseline factors are considered; pregnancy differences are small and often not robust after adjustment. Live‑birth reporting is sparse and heterogeneous, leading to low certainty for that endpoint. Cochrane‑style syntheses limited to a few RCTs similarly conclude no statistically significant live‑birth difference between HP‑hMG and rFSH and uncertain effects on OHSS, with overall low‑quality evidence due to small sample sizes and inconsistent endpoints. • Real‑world observational studies: Very large noninterventional cohorts provide complementary evidence. A nationwide German study (>28,000 women; 71 centers) found adjusted higher cumulative live birth, ongoing pregnancy, and clinical pregnancy with rFSH‑alfa versus HP‑hMG, with fewer cancellations and less drug per oocyte; however, residual confounding and product‑specific factors limit causal inference. Other national databases report broadly concordant signals favoring rFSH on cumulative outcomes but face similar limitations. • Trials registry: Multiple completed interventional and observational studies involve menotropins/HP‑hMG across indications and protocols, supporting a substantial—though heterogenous—trial base; however, many registered/older trials lack live‑birth primary endpoints or contemporary antagonist/freeze‑all designs.

Comparative effectiveness versus rFSH and rFSH+rLH • hMG vs rFSH: rFSH generally retrieves more oocytes with lower total dose; live birth per fresh transfer and cumulative live birth are typically similar in RCTs/meta‑analyses, though effect estimates are imprecise and of low certainty. Large real‑world cohorts suggest rFSH may achieve modestly higher cumulative live birth, but confounding is likely. In high‑responder RCT settings, cumulative LBR appears similar, with lower OHSS on HP‑hMG. • hMG vs rFSH+rLH: Evidence is limited and mixed. Some syntheses that include nonrandomized trials suggest potential benefits to LH activity (via hMG or rLH addition) for embryo/implantation outcomes in selected subgroups, but few trials report live birth and study quality is variable. High‑quality, product‑specific RCTs with live‑birth endpoints directly comparing HP‑hMG to rFSH+rLH under modern antagonist protocols remain scarce.

Safety • OHSS: Estimates vary by population and protocol. In high responders under antagonist protocols, an assessor‑blinded RCT found substantially lower OHSS with HP‑hMG than with rFSH, despite similar cumulative LBR; meta‑analytic certainty on OHSS more broadly is low given inconsistent definitions and small numbers. • Miscarriage/pregnancy loss: Some RCTs and summaries report lower pregnancy loss with HP‑hMG versus rFSH, but results are inconsistent and often underpowered; overall differences remain uncertain.

Key limitations, evidence gaps, and criticisms • Endpoints and certainty: Many trials use surrogate endpoints (oocyte count, biochemical/clinical pregnancy) rather than live birth or cumulative live birth, limiting clinical interpretability. Live‑birth and cumulative LBR data are sparse, underpowered, and heterogeneous, yielding low‑certainty estimates. • Heterogeneity and bias: Trials differ by stimulation protocol (agonist vs antagonist), dosing, embryo strategy (fresh vs freeze‑all), and patient phenotype (high responders, advanced age, PCOS), complicating pooling. Many RCTs are small, single‑center, with unclear allocation concealment/blinding and potential industry sponsorship; nonrandomized syntheses inflate risk of bias. • Product variability: Urinary‑derived HP‑hMG exhibits batch‑to‑batch variability and contains LH bioactivity predominantly from hCG rather than LH, raising questions about biological equivalence to rLH or to rFSH alone. Earlier meta‑analyses often pooled across distinct urinary products and did not compare specific modern products head‑to‑head, limiting applicability. • RCT vs real‑world discordance: Large observational analyses suggest rFSH may yield higher cumulative LBR, in tension with small RCTs indicating non‑inferiority and similar LBR; unmeasured confounding and prescribing channeling in routine care constrain causal conclusions from real‑world data. • Reporting of safety: OHSS definitions and grading differ; event counts are low in many studies, producing imprecision. Miscarriage differences are inconsistently reported. Standardized, adjudicated safety endpoints are needed. • Modern practice gaps: Few adequately powered, product‑specific RCTs use contemporary antagonist/freeze‑all strategies and cumulative live birth as the primary endpoint, especially in key subgroups (advanced age, poor responders, PCOS). The relative impact of LH source (hCG‑driven LH activity in hMG vs rLH) on euploidy and embryo competence remains insufficiently defined.

• Extent: The evidence base for hMG is substantial, comprising dozens of RCTs (mostly small/moderate), several meta‑analyses, and very large real‑world cohorts; multiple registered trials exist. However, live‑birth–focused, product‑specific RCT evidence under current protocols is limited. • Quality: Moderate for surrogate outcomes (oocytes), low‑to‑moderate for live birth and cumulative live birth owing to heterogeneity, small sample sizes, and inconsistent endpoints. Safety evidence suggests protocol‑ and population‑dependent OHSS patterns, with possible lower OHSS for HP‑hMG in high responders; certainty remains low outside that context. • Comparative effectiveness: rFSH generally yields more oocytes; live birth and cumulative live birth are similar in RCTs/meta‑analyses, while some large real‑world studies favor rFSH on cumulative outcomes. Evidence directly comparing hMG to rFSH+rLH is limited and mixed, with few live‑birth endpoints. • Key criticisms: reliance on surrogate endpoints, heterogeneity and risk of bias, product variability and LH‑source differences, industry sponsorship, and limited modern, adequately powered, live‑birth–based RCTs, particularly for cumulative LBR and prioritized subgroups.

Evidence Gaps and Limitations#

The current evidence base for HMG consists primarily of preclinical studies. Key limitations include:

• No completed randomized controlled trials in humans
• Most data derived from animal models, limiting direct translatability
• Publication bias may favor positive results
• Long-term safety data in humans is not available
• Optimal dosing for human applications has not been established

Key Research Findings#

Recombinant human FSH produces more oocytes with a lower total dose per cycle compared with HP-hMG: a meta-analysis, published in Reproductive Biology and Endocrinology (Lehert P et al., 2010; PMID: 20932351):

• The study showed meta analysis of 16 RCTs
• The study showed meta analysis of 16 RCTs ; rFSH yielded more oocytes (MD -1.54 to 2.10 fewer with hMG) at lower total dose; no significant difference in clinical pregnancy or live birth

Gonadotropins for pubertal induction in males with hypogonadotropic hypogonadism: systematic review and meta-analysis, published in European Journal of Endocrinology (Alexander EC et al., 2024; PMID: 38128110):

• The study showed systematic review of 103 studies
• The study showed systematic review of 103 studies ; spermatogenesis rates with hCG+FSH 86% vs hCG alone 40% ; gonadotropins increased testicular volume, penile size, and testosterone in >98% of analyses

Use of letrozole and clomiphene citrate combined with gonadotropins in clomiphene-resistant infertile women with PCOS: a prospective study, published in Drug Design Development and Therapy (Xi W et al., 2015; PMID: 26648691):

• The study achieved monofollicular development vs 65.3% and 54.7% ; reduced total gonadotropin dose without compromising pregnancy rates of 80.2%

Related Reading#

• HMG research studies and evidence
• HMG dosing protocols
• HMG side effects profile
• Gonadorelin research guide
• HCG research guide