IGF-2

A 67-amino-acid growth factor encoded by the imprinted IGF2 gene, central to fetal development and held in check by a dedicated clearance receptor, with no approved product anywhere and no human trials of its administration.

Also referenced as: IGF-II, IGF2, Insulin-like growth factor 2, Insulin-like growth factor II, Somatomedin A

Also appears in: Performance

What is IGF-2?

IGF-2 is a 67-amino-acid single-chain polypeptide of roughly 7.5 kDa, folded by three intramolecular disulfide bonds into a shape closely related to proinsulin and to IGF-1. It is made as a precursor of about 180 residues; a 24-residue signal sequence and a C-terminal E-domain are cleaved away to release the mature 67-residue peptide. Older literature almost always writes it as IGF-II.

Its gene sits at chromosome 11p15.5 and is imprinted: in most tissues only the paternally inherited copy is transcribed, with the maternal copy silenced — the liver and parts of the brain, where both alleles are expressed, are the notable exceptions. That one feature explains most of what is distinctive about IGF-2 — its dosage is set by parent of origin, and the classic 11p15 imprinting disorders are in large part IGF-2 dosage disorders.

IGF-2 reaches the research-compound market almost entirely as a recombinant life-science reagent — a cell-culture protein sold by the microgram — rather than as an injectable research peptide in the way IGF-1 LR3, IGF-1 DES, and MGF are. Its visibility comes from family resemblance to those compounds and from the recognizability of the IGF name, not from any body of human work behind it.

How it works

• Two signaling receptors, and one of them belongs to insulin. Like IGF-1, IGF-2 activates the type 1 IGF receptor (IGF-1R). Unlike IGF-1, it is also a high-affinity agonist of the A isoform of the insulin receptor: IR-A, but not IR-B, binds IGF-2 with an affinity close to that of insulin, and activation of IR-A by IGF-2 produces primarily mitogenic effects, where activation of the same receptor by insulin produces primarily metabolic ones (Frasca et al., Molecular and Cellular Biology, 1999;19(5):3278–3288). IR-A is the fetal-pattern splice variant, and it is re-expressed in many tumors — which is why the IGF-2/IR-A loop is discussed as an autocrine cancer signal rather than a growth-hormone-style anabolic axis.
• A third receptor whose main job is to remove it. The type 2 IGF receptor (IGF2R) is the cation-independent mannose-6-phosphate receptor. It has no tyrosine kinase domain and does not transduce the canonical mitogenic signal that IGF-1R and IR-A carry; its principal role is to bind IGF-2 at the cell surface and traffic it to the lysosome for degradation, holding circulating and tissue levels down (Brown et al., Trends in Biochemical Sciences, 2009;34(12):612–619). It is not, however, inert in every setting — the rodent memory work described below is explicitly IGF2R-dependent, so “clearance receptor” describes its dominant function rather than a total absence of signaling. Mice inheriting a non-functional Igf2r from the mother do not express the receptor, carry elevated circulating IGF-2, grow 25–30% larger than their normal siblings, and mostly die around birth with major cardiac abnormalities (Lau et al., Genes & Development, 1994;8(24):2953–2963). This clearance arm has no counterpart in IGF-1 biology and is the main reason IGF-2 is treated as a separate compound rather than an IGF-1 variant.
• A fetal growth program that growth hormone does not drive. Disrupting one Igf2 allele in mice and transmitting it through the male germ line produced heterozygous progeny at about 60% of normal body weight, otherwise apparently normal and fertile — the founding demonstration that IGF-2 is an embryonic growth factor (DeChiara et al., Nature, 1990;345(6270):78–80). Deleting only the placenta-specific P0 transcript reduced placental growth first and produced fetal growth restriction several days later, identifying the placenta as a principal site of action (Constância et al., Nature, 2002;417(6892):945–948). In humans, growth hormone is a comparatively weak regulator of serum IGF-2: it shows neither the pubertal surge nor the brisk dose-response to GH that defines IGF-1. Most of the circulating pool — more than three-quarters of total IGF — is held in 150 kDa ternary complexes with IGFBP-3 and the acid-labile subunit (Livingstone & Borai, Clinical Endocrinology, 2014;80(6):773–781; van Doorn, BioFactors, 2020;46(4):563–578).
• An autocrine differentiation signal in muscle. Myoblasts switched to low-serum medium begin expressing IGF-2 themselves, and an antisense oligonucleotide against IGF-2 blocks their “spontaneous” differentiation unless exogenous IGF is supplied — establishing IGF-2 as an autocrine trigger of terminal myogenic differentiation acting through induction of myogenin, rather than as a systemic hypertrophy signal (Florini et al., Journal of Biological Chemistry, 1991;266(24):15917–15923). This cell-culture role is the entire basis of IGF-2’s association with muscle; it has never been tested as a muscle intervention in an animal or a person.

Research status

The human record for IGF-2 is unusual: there is a great deal of it, and none of it involves giving IGF-2 to anyone.

No human trials of IGF-2 administration. No interventional study registered on ClinicalTrials.gov administers IGF-2 as a therapeutic agent, in any phase, for any indication. There is no published human pharmacokinetic or dose-ranging data for exogenous IGF-2. Every clinical program that touches this molecule either measures it as a biomarker or blocks it.

Human loss-of-function genetics. A multigenerational family carrying an IGF2 nonsense variant (c.191C→A, p.Ser64Ter) produced growth restriction in four members — and only in those who inherited the variant from their father, exactly as the maternal imprinting of the locus predicts (Begemann et al., New England Journal of Medicine, 2015;373(4):349–356). The severity of the postnatal phenotype indicated that IGF-2 contributes to growth after birth, not only before it, and the dysmorphic features overlapped those of Silver-Russell syndrome. This remains the cleanest human evidence that IGF-2 dosage matters.

Imprinting disorders as dosage experiments. Beckwith-Wiedemann syndrome (overgrowth) and Silver-Russell syndrome (growth restriction) are widely described as phenotypic and molecular mirror images at 11p15, with IGF-2 dosage the common thread (Sélénou et al., Cells, 2022;11(12):1886). IGF2 loss of imprinting — reactivation of the silenced maternal allele, doubling expression — is also among the more frequently reported epigenetic lesions in human tumors.

Human IGF-2 excess produces hypoglycemia, not growth. Non-islet cell tumor hypoglycemia is a rare paraneoplastic syndrome in which a tumor secretes incompletely processed pro-IGF-2 — “big IGF-2,” roughly 10–15 kDa, retaining part of the E-domain and variably O-glycosylated. It makes up perhaps 10–15% of serum IGF-2 in healthy adults but can reach the large majority of it in NICTH; the abnormal form is poorly incorporated into the usual 150 kDa ternary complexes, circulates instead in binary complexes with IGF-binding proteins, reaches the insulin receptor, and causes refractory hypoglycemia (van Doorn, BioFactors, 2020;46(4):563–578). A retrospective series of 37 patients with big-IGF-2-producing NICTH found markedly suppressed growth hormone during hypoglycemic episodes — mean 0.85 versus 9.62 ng/mL — relative to a small comparison group of six hypoglycemic patients with normal IGF-2, with cortisol likewise lower; a hormone signature that helps separate the two (Fukuda et al., Endocrine Journal, 2017;64(7):719–726). Complete surgical resection is the only definitive treatment. This is the closest thing to an uncontrolled human experiment in IGF-2 elevation, and the result is a metabolic emergency.

Clinical development runs in the opposite direction. Xentuzumab (BI 836845) is a humanized monoclonal antibody that binds and neutralizes both IGF-1 and IGF-2. It went through a series of Phase 1 studies across solid tumors, prostate cancer, and breast cancer, and one randomized Phase 2. XENERA-1 randomized 103 patients with hormone-receptor-positive/HER2-negative metastatic breast cancer and non-visceral disease to xentuzumab or placebo on top of everolimus and exemestane; the trial was unblinded early for discordance between independent and investigator PFS assessment, and by independent review median progression-free survival was 12.7 months with xentuzumab versus 11.0 months with placebo (hazard ratio 1.19; 95% CI 0.55–2.59; p = 0.6534) — no benefit, with the hazard ratio pointing, imprecisely, toward placebo (Schmid et al., Breast Cancer Research, 2023;25(1):67; NCT03659136). An earlier IGF-1/IGF-2-neutralizing antibody, MEDI-573, ran several Phase 1 studies (NCT00816361, NCT01340040, NCT01498952) and a completed Phase 2 in hormone-sensitive metastatic breast cancer (NCT01446159), and likewise did not advance to registration. A separate program, ersodetug (RZ358), blocks the insulin receptor allosterically and is in Phase 3 for tumor hyperinsulinism including IGF-2-driven cases (NCT06881992) — again treating IGF-2’s effect rather than supplying it.

Rodent neuroscience. Injecting recombinant IGF-2 into the rat hippocampus after training or after retrieval enhanced memory retention and prevented forgetting, within a defined consolidation window and dependent on IGF-2 receptors and new protein synthesis (Chen et al., Nature, 2011;469(7331):491–497). A follow-up reported memory enhancement in mice after systemic rather than intracranial administration (Stern et al., Neuropsychopharmacology, 2014;39(9):2179–2190). This line of work is frequently cited in nootropic marketing for IGF compounds; it has not been carried into a human trial, and the compound tested was IGF-2 itself, not any IGF-1 analog.

No regulatory authority has approved an IGF-2 product for any indication.

Common dosage forms

• Recombinant human IGF-2 as lyophilized powder, typically supplied in 10–100 µg vials for cell culture, most often expressed in E. coli, and offered either carrier-free or reconstituted with a BSA carrier. Shelf life is quoted in months at −20 °C or −80 °C.
• Carrier-free liquid preparations, shipped cold, as an alternative to the lyophilized presentation.
• Effectively absent from the injectable research-peptide market. Vendor catalogs in that space carry IGF-1 LR3, IGF-1 DES, and MGF; multi-milligram injectable-style IGF-2 vials of the kind sold for those compounds are not a standard listing.
• Named as one ingredient in an unapproved homeopathic topical. A search of FDA’s drug label database for “IGF-II” returns exactly one product: an OTC homeopathic topical listing IGF-II among dozens of ingredients. Homeopathic products of this kind are not FDA-approved drugs, and no IGF-2 content is established for them.

This section describes formats only and is not dosing guidance.

Key considerations

• Nothing containing IGF-2 is approved anywhere. The approved member of this family is recombinant IGF-1 — mecasermin, marketed as Increlex (BLA 021839, first approved August 2005) for severe primary IGF-1 deficiency. Marketing that borrows the credibility of “IGF” approval is describing a different molecule with a different receptor profile.
• The human evidence is observational, and the excess phenotype is hypoglycemia. Everything known about elevated IGF-2 in people comes from imprinting disorders and paraneoplastic secretion, not from controlled administration. The dominant clinical consequence of too much IGF-2 in an adult is refractory low blood sugar driven by insulin-receptor activation, not tissue growth.
• Every clinical program aims to reduce IGF-2, not supply it. Two neutralizing antibodies and one insulin-receptor blocker have entered trials on the premise that IGF-2 signaling should be suppressed — in oncology because of the IGF-2/IR-A autocrine loop, and in tumor hyperinsulinism because of the hypoglycemia. The randomized readout that exists, XENERA-1, was negative.
• Growth-hormone reasoning does not transfer. IGF-2 sits in the somatomedin family, but its regulation is largely GH-independent: serum IGF-2 tracks neither puberty nor GH dosing the way IGF-1 does, and it has a dedicated clearance receptor that IGF-1 does not. Inferences built on GH secretagogues or on IGF-1 analogs do not carry over.
• Several names in this area point at different molecules. IGF2R is the clearance receptor, not the peptide. Preptin is a separate 34-amino-acid peptide corresponding to part of the pro-IGF-2 E-domain. “Big IGF-2” is the pathological incompletely processed form found in tumor hypoglycemia, not a long-acting analog. “Somatomedin A” is a historical name for IGF-2 whose original attribution is contested. And IGF-1 LR3, IGF-1 DES, and MGF — the IGF compounds that actually circulate in this market — are IGF-1 derivatives with a different receptor profile. Reagent-grade recombinant protein is also sold under research-use-only terms, with purity and identity established for cell culture rather than for any human application.