IGF-1 LR3 (Long R3 IGF-I)

Chemical identity, and a CAS number that points elsewhere

The catalogue name and the registry's preferred name are not the same string. The FDA Global Substance Registration System holds one record for this molecule, filed as Long-(Arg3)insulin-like growth factor-I under UNII M9L22Y19H9, with IGF-1 LR3 among the synonyms attached to it, so the two names denote one substance. The single subunit runs 83 residues, beginning MFPAMPLSSLFVN and continuing into GPRTLCGAELVDALQFVCGDRGFYFNKPTGYGSSSRRAPQTGIVDECCFRSCDLRRLEMYCAPLKPAKSA, with three disulfide bonds recorded at Cys19-Cys61, Cys31-Cys74 and Cys60-Cys65. After the first thirteen residues comes the 70-residue human IGF-I chain, glutamate at position 3 replaced by arginine. That thirteen-residue extension is the first eleven residues of methionyl porcine growth hormone followed by a Val-Asn dipeptide, which served as the hydroxylamine-sensitive cleavage linker in the original construct (PMID 1311930).

Two CAS numbers are in circulation for it. The registry record carries 143045-27-6, which is also the number in the Wikipedia infobox. Most pages selling or describing the compound print 946870-92-4 instead. Querying PubChem for 946870-92-4 returns CID 168009904, a C80H154N4O6S2 ionisable lipid of molecular weight 1332 with no relation to any peptide. PubChem holds no standardised compound record for the protein itself; searches on the name and on 143045-27-6 return substance deposits only, which is ordinary treatment for a chain this size. The identifier field on this page is left blank rather than filled with the number that circulates.

Formula and mass do not settle either. The registry lists C400H623N111O115S9 and an estimated average mass of 9120 Da, both flagged as calculated rather than measured. Vendor pages and the Wikipedia infobox give C400H625N111O115S9 and 9117.60 g/mol. The arithmetic resolves the discrepancy. C400H625N111O115S9 is the reduced-chain formula computed from the registry's own 83-residue sequence, and 9117.60 is that formula's calculated average mass; the folded form carrying the three disulfide bonds the registry records would be C400H619N111O115S9. The registry's C623 matches neither, and its stated 9120 Da sits about 4.4 Da above the 9115.6 Da its own formula computes to. No experimentally determined mass for the untagged substance was located. Zhao and colleagues spelled the extension out residue by residue in their 1993 bovine leukocyte paper as Met-Phe-Pro-Ala-Met-Pro-Leu-Ser-Ser-Leu-Phe-Val-Asn (PMID 7508487), matching the registry sequence and not the double-leucine spelling carried in the Wikipedia article body and on several secondary pages.

What the substitution was designed to do

King and colleagues described the Escherichia coli expression system in February 1992 and Francis and colleagues characterised the resulting fusion analogues four months later, both in the Journal of Molecular Endocrinology. The engineering objective was narrow: reduce association with the IGF-binding proteins while leaving the type-1 receptor intact. It worked in that direction. Ballard and colleagues put the affinity difference towards IGFBP-3, IGFBP-4, total rat plasma binding proteins and L6 myoblast binding protein at roughly a thousandfold in 1993 (PMID 7683526). Receptor binding moved slightly the other way, and Tomas and colleagues recorded that the analogue binds the type 1 IGF receptor about three times less well than IGF-I.

Potency in culture varied with whether the line secreted binding proteins. In rat L6 myoblasts, which secrete them, Ballard's group measured a five- to tenfold advantage over IGF-I. In chicken embryo fibroblasts, a line that secretes no detectable IGFBPs into the medium, Francis and colleagues found the reverse: Long [Arg3]-IGF-I was less potent than IGF-I. Francis and colleagues reported the opposite potency orderings in those two culture systems and attributed the difference to whether the line secreted detectable IGF-binding proteins into the medium (PMID 1378742).

The variants bound plasma binding proteins to different extents by species. Lord and colleagues compared labelled IGF-I, des(1-3)IGF-I and LR3IGF-I against plasma binding proteins from rat, sheep, human, pig and chicken in 1994, and reported that the variants which bound extremely weakly in rat plasma bound significantly in plasma from every other species tested. By size-exclusion chromatography the extent of variant binding fell in the order sheep, human, pig, chicken, then rat. Nearly all of the in vivo potency data was generated in the rat, at the extreme end of that range.

Measured plasma clearance in rats

Bastian and colleagues attributed the difference to binding-protein association: peptides that do not enter the 150 kDa ternary complex were cleared and degraded faster in their rat measurements. They gave intravenous boluses of radiolabelled peptide to catheterised rats in 1993. Metabolic clearance rate for LR3IGF-I was 9.84 plus or minus 0.28 ml/min per kg in virgin animals. For IGF-I in the same animals it was 0.90 plus or minus 0.05. The analogue left the circulation roughly eleven times faster (PMID 7693845).

The pregnant animals in the same study provided a within-study comparison. On day 18 of gestation, when rat plasma IGFBP-3 is sharply reduced, IGF-I clearance rose to 2.88 plus or minus 0.12 ml/min per kg while LR3IGF-I clearance barely moved, at 9.19 plus or minus 0.15. Most labelled LR3IGF-I was recovered as free peptide; most labelled IGF-I sat in the 150 kDa complex. Ballard's group summarised the series plainly, writing that an IGF variant which associates poorly with IGFBPs is removed more rapidly from the blood. Shoubridge and Read reproduced the direction in 2003 with 2.6 microgram/kg boluses in preweaning and adult rats, five to six animals per group (PMID 12697696).

The half-life figures attached to the analogue in secondary sources come from a different experiment on a different molecule. Guler and colleagues injected radiolabelled IGF-I and IGF-II into two healthy adults in 1989 and reported three half-lives for the tracer depending on what it was bound to: 10 to 12 minutes free, 20 to 30 minutes in the 50 kDa complex, and 12 to 15 hours in the 200 kDa complex (PMID 2558477). Those three numbers, stripped of the qualifier that distinguishes them, correspond to the twelve-to-fifteen and twenty-to-thirty figures carried against LR3 by the Wikipedia article and by vendor and aggregator pages.

The rodent record, and the species that disagreed

Four rat studies reported growth and nitrogen-retention differences, summarised below. Tomas and colleagues infused LR3IGF-I into dexamethasone-treated 150 g male rats by subcutaneous osmotic pump for seven days in 1992 and found it about 2.5-fold more potent than IGF-I on body weight and nitrogen retention, noting that the result was striking because the peptide binds the type 1 receptor less well. In normal growing female rats over fourteen days the following year, 44 microgram/day of the analogue produced effects similar to 278 microgram/day of IGF-I (PMID 8371075). Gut was the tissue with the largest reported changes across the series: Read and colleagues recorded IGF-I at 695 microgram/day increasing total gut weight by up to 60% in the dexamethasone model, and described the analogues as severalfold more potent without attaching a percentage to them (PMID 1613443).

Gastrointestinal endpoints carry the most complete numbers in the retrieved reports. Steeb and colleagues treated female rats, six per group, for fourteen days and measured a 43% rise in total gut weight, a 47% rise in small intestinal weight and a 13% increase in small intestinal length at 278 microgram/day of LR3IGF-I. Carcass muscle was the endpoint where the difference did not hold. When the same laboratory compared once-daily injection with continuous infusion in 1996, the analogue held a 1.5- to 2-fold advantage on body weight, visceral organ weights and feed efficiency but was, in the authors' words, barely equipotent with IGF-I for reversal of carcass muscle loss.

The pig studies reported the opposite direction. Dunaiski and colleagues infused LR3IGF-I into finisher pigs at 180 microgram/kg per day for four days and recorded decreased average daily gain, decreased food intake, and suppression of plasma IGFBP-3, IGF-I and insulin, with the area under the growth hormone peaks down 60%. Plasma glucose was unaffected by any treatment in that study. Dunshea and colleagues found no overall effect on daily weight gain across 42 artificially reared neonatal pigs in the first of two experiments, and a 457 versus 386 g/day difference over the second nine days of the eighteen-pig second experiment (PMID 12067429).

What the acute glucose measurements recorded

Tomas and colleagues ran two arms in 1997. LR3IGF-I was given only to pigs, at bolus doses of 20 and 50 microgram/kg. The marmoset arm, dosed at 42 to 270 microgram/kg, compared IGF-I with des(1-3)IGF-I and did not include LR3IGF-I. Blood was sampled from 30 minutes before injection to four hours after. Maximum lowering of plasma glucose reached 4.8 mmol/l in the pig, and in the marmoset 3.7 mmol/l conscious and 2.5 anaesthetised, the latter two figures belonging to IGF-I and des(1-3)IGF-I rather than to this analogue. In the pig arm, the variants that bind IGFBPs poorly, LR3IGF-I among them, were two- to threefold more potent than IGF-I at the glucose nadir.

Cumulative glucose suppression over four hours was four- to eightfold greater for the variants than for IGF-I in the pig arm, and at doses equipotent at the nadir the cumulative effect was still about double. Hypoglycaemic potency ordered by binding-protein affinity: IGF-I, then long-IGF-I, then R3IGF-I and LR3IGF-I together, then des(1-3)IGF-I. Correlated changes in plasma insulin, IGF-I and IGF-binding proteins did not account for the differential in the marmoset.

Where that effect is transduced is partly visible in the 1992 culture panel. Francis and colleagues included H35 hepatoma cells specifically because the IGFs act through the insulin receptor in that line, and found the Long IGF-I analogues held a similar potency relative to IGF-I there as in L6 myoblasts. Nothing in the modification was aimed at receptor selectivity, and nothing in the retrieved data indicates it changed. No primate has received this analogue in any retrieved study.

No human record, and what has been found in circulation instead

ClinicalTrials.gov intervention searches on Long R3 IGF-1, Long R3 IGF-I, LR3-IGF-I and IGF-1 LR3 each returned zero registered studies in August 2026. The PubMed query ("Long R3 IGF-1" OR "Long R3 IGF-I" OR "LR3IGF-I" OR "LR3-IGF-I" OR "long-R3-IGF-I" OR "IGF-1 LR3") AND humans[mh], run unfielded on 18 August 2026, returned 36 records. No record in that set reports administration of the peptide to a person. The set is not confined to human material: alongside human cell-line and human-plasma binding work it carries rat studies (PMIDs 12697696, 10607940, 8138750), turkey satellite cells (11469659), bovine preantral follicles (17636166) and an intranasal mouse Alzheimer's model (39610283). Dominikowski and colleagues reached the same position in a 2026 review in Frontiers in Endocrinology (PMID 42395176), placing IGF-1 LR3 in their lowest evidence tier and recording that no route of administration has been reported in a peer-reviewed human study and that its half-life is not documented.

Regulatory status separates this molecule from its parent. Mecasermin, recombinant human IGF-I with the native 70-residue sequence, holds United States biologics licence 021839 and carries current labelling on file with DailyMed. No marketing authorisation for the LR3 analogue was located in any jurisdiction. The 2026 World Anti-Doping Agency Prohibited List reaches it by class rather than by name, at S2.3, which lists insulin-like growth factor-1 and its analogues among growth factors prohibited at all times, in and out of competition.

One analytical paper describes what was actually inside a vial. Kohler and colleagues at the Cologne doping laboratory reported in 2010 on an injection vial obtained as a black-market product, isolating the contents by immunoaffinity purification and characterising them by nano-UPLC and high-resolution mass spectrometry of both the intact and trypsinated protein. The identification was Long-R3-IGF-I carrying a hexahistidine tag on the C-terminus through a Leu-Glu linker. The authors observed that such tags are added for purification convenience during biochemical synthesis, are normally removed enzymatically when placed at the N-terminus, and that the effects of the His-tagged form in humans have not been described.

Two published uses of the molecule do not depend on any physique claim at all. Falconer and colleagues used Long-R3-IGF-I as the model recombinant product in a 1999 Biotechnology and Bioengineering paper on selective extraction from Escherichia coli inclusion bodies, recovering 81% w/w at 46% purity (PMID 9921154), and Lu and colleagues expressed it as a xylanase fusion in Pichia pastoris in 2023, reaching about 1 g/l in a fifteen-litre bioreactor (PMID 37261455). Von der Thusen and colleagues used it as a stable IGF-1 surrogate in apolipoprotein E knockout mice (PMID 21281823). In each case it is a reagent chosen for handling properties.

What is not known

The published record contains no human administration data of any kind. ClinicalTrials.gov intervention searches across four name variants returned zero registered studies in August 2026, and the PubMed query ("Long R3 IGF-1" OR "Long R3 IGF-I" OR "LR3IGF-I" OR "LR3-IGF-I" OR "long-R3-IGF-I" OR "IGF-1 LR3") AND humans[mh], run unfielded on 18 August 2026, returned 36 records, none of which reports administration of the peptide to a person. That set is not confined to human material: it carries rat, turkey, bovine and mouse studies indexed under the human heading alongside human cell-line and human-plasma binding work. There is consequently no human pharmacokinetic curve, no dose-ranging work, no safety database, and no measured half-life in a person; Dominikowski and colleagues state exactly this in their 2026 review. No primate has received this analogue in any retrieved study either: the marmoset arm of Tomas 1997 compared IGF-I with des(1-3)IGF-I only, and every LR3IGF-I dose in that paper went to pigs. Beyond that, the animal literature has its own gaps. Almost all of the potency data comes from the rat, and Lord and colleagues showed the rat is the species in which this analogue binds plasma binding proteins least, which is the mechanism the potency advantage depends on. Sample sizes are absent from most retrieved reports, and dosing routes are heavily weighted towards continuous subcutaneous infusion by osmotic pump rather than injection. Nothing was located on chronic administration beyond a few weeks, reproductive or developmental effects, immunogenicity against the non-native N-terminal extension, or carcinogenicity, despite the receptor being one whose signalling is implicated in tumour biology. Product identity in circulation is also unestablished: the single analytical characterisation of an injection vial obtained as a black-market product found a His-tagged research by-product whose behaviour in humans the authors described as unelucidated.

Questions