IGF-1 LR3: A Research Primer

The production route is the fact everything else follows from

A short synthetic peptide is built by solid-phase chemistry: an anchored residue, then repeated cycles of deprotection and coupling, then cleavage from the resin and removal of the side-chain protecting groups. Every failure mode of that process leaves a chemical signature. A coupling that does not go to completion leaves a chain missing one residue, which is a deletion sequence. A protecting group that survives cleavage leaves an adduct of known mass. Scavengers and cleavage reagents leave residues. The certificate that accompanies such a material is designed around exactly those possibilities, which is why it pairs a reversed-phase area percentage with an intact mass and stops there.

None of that applies to a recombinant protein. This molecule is made by putting an engineered gene into a host organism, letting the host's own ribosomes build the chain, then lysing the culture and purifying the product away from everything else the host made. There are no protecting groups because there was no protection step. There are no deletion sequences in the synthetic sense because translation either produces a full-length chain or terminates, and a prematurely terminated chain is a different object from a chain missing an internal residue. What there is instead is a population of contaminants that has no counterpart in synthetic work at all: residual host-cell protein, host-cell nucleic acid, endotoxin if the host is a Gram-negative bacterium, forms clipped by host proteases, forms carrying or missing an initiator methionine, oxidized and deamidated variants generated during fermentation and purification, and, most importantly, species whose covalent composition is correct but whose three-dimensional structure is not.

That last category is the one with no synthetic analog whatsoever. A 12-residue peptide does not have a fold to get wrong. This protein does. Native IGF-1 is stabilized by three intramolecular disulfide bonds, and the LR3 construct carries the same architecture. Six cysteines can pair in several ways; only one pairing is the correct one, and the incorrect pairings are chemically the same molecule by every measure of composition. A batch can therefore be correct in sequence, correct in mass, clean by chromatography, and structurally wrong.

The practical consequence is that a certificate transplanted from the synthetic side of a catalog is not incomplete in a small way. It is answering a different question set. It will tell you the covalent composition matches and that a single chromatographic peak dominates, and it will be silent on host-cell protein, on endotoxin, on aggregate content, and on whether the molecule is folded. Reading it as though it were the equivalent document is the most common error made with this material, and it is an error of category rather than of degree.

The same certificate fields mean different things on either side of this line

The rows do not simply differ in wording. Each one is a different assay run on a different instrument answering a different question. If the material is being entered into an inventory alongside synthetic peptides, the record should carry a flag saying so, because otherwise a colleague reviewing that inventory a year later will apply the wrong set of expectations to it and conclude the paperwork is fine when several of the load-bearing fields were never generated. The same caution applies to any internal specification written once for the catalog as a whole: a specification that names an area percentage and an intact mass is a complete specification for a synthetic peptide and a partial one here.

Intact mass confirms composition but not connectivity

An intact mass measurement ionizes the whole molecule and reports what it weighs. Compared against a theoretical mass calculated from the stated sequence, a match is strong evidence that the covalent composition is right: the correct residues, in the correct number, with no extra adduct and no missing fragment. For a synthetic peptide that is close to sufficient, because a synthetic peptide has almost nothing else to get wrong.

Here it is necessary and clearly not sufficient, and the reason is a piece of arithmetic worth doing explicitly. Forming a disulfide bond from two free cysteine thiols removes two hydrogen atoms, so it lowers the mass of the chain by approximately two daltons. Three disulfide bonds therefore account for about six daltons across the whole molecule. An 83-residue chain sits in the high single-digit kilodalton range, so six daltons is well under a tenth of a percent of the total. On a high-resolution instrument reporting low-parts-per-million accuracy, that difference is enormous and trivially resolved. On a benchtop system reporting a deconvoluted average mass against a tolerance window of a few daltons, which is what a large number of routine certificates describe, a fully reduced molecule and a fully oxidized one can both land inside the accepted window. The certificate then says the mass matched, and it did.

The scrambled case is worse, because there is no arithmetic at all. A molecule whose six cysteines are paired in the wrong combination has the identical molecular formula and the identical exact mass as the correctly paired one. The difference between them is purely topological. No mass measurement of any resolution, on any instrument, can distinguish them, because there is nothing to distinguish. Neither can a sequence-level check that ignores connectivity: a tryptic map run under reducing conditions gives the same peptide list for both, since reduction destroys exactly the information being asked about.

What does address the question is orthogonal, and it falls into a few families. The cheapest is a comparison: measure intact mass before and after deliberate reduction, and confirm the roughly six-dalton shift, which establishes that three disulfides were present even if it does not establish which cysteines were paired. A free thiol assay approaches the same point from the other side by confirming that essentially no cysteine is unpaired. Non-reducing versus reducing SDS-PAGE run in adjacent lanes shows whether disulfide-linked multimers exist and whether the monomer migrates as a compact species. A peptide map run without reduction, in which the disulfide-linked fragments stay linked and are identified as such, is the method that actually assigns pairings. And a cell-based functional readout is the only evidence that speaks to the fold as a whole rather than to the bonds one at a time.

What each piece of analytical evidence settles, and what it leaves open

No single row is a substitute for the others, and a certificate carrying only the first row is not a defective document so much as a partial one. The useful posture on receiving a lot is to ask which of these rows exists for this lot number, accept that most suppliers of research-grade material will have the first and perhaps the fourth, and record the absence of the rest as a known gap rather than assuming it away. A gap that is written down is a limitation on an experiment; a gap that is not written down becomes an unexplained result later.

Reading the LR3 name as a record of two separate edits

The name is a compressed description of two structural changes made to the native sequence, and separating them is what makes the literature legible. The Long refers to an N-terminal extension of thirteen residues appended ahead of the native chain. The R3 refers to a substitution at position 3 of the native numbering, where the glutamate present in native human IGF-1 is replaced by an arginine. The parent scaffold underneath both edits is the 70-residue single-chain IGF-1 fold with its three disulfide bonds, which is why the total comes to 83 residues.

Both edits sit at the same end of the molecule, and that is not a coincidence. The N-terminal region of IGF-1 is described in the structural literature as a principal contact surface for the IGF binding proteins. Exchanging an acidic side chain for a basic one in the middle of that surface changes the local electrostatics; adding a thirteen-residue tail immediately ahead of it adds steric bulk to the same region. The reported outcome of the pair is a substantial reduction in affinity for the binding proteins while affinity for the receptor itself is far less affected. That selectivity is the entire design intent: the modification separates two interactions that the native molecule cannot separate.

It is worth being precise about what "reduced binding-protein affinity" is a statement about, because it is routinely stretched into claims it does not support. Affinity is an equilibrium property of a molecular interaction, measured in a binding experiment and expressed as a dissociation constant. A lower affinity for a binding protein means that, at a given concentration of both partners, a smaller proportion of the growth factor is captured in a complex and a larger proportion is free in solution. That is a mass-action statement about a two-component equilibrium. It is not a statement about duration, it is not a statement about potency at the receptor, and it is certainly not a statement about anything happening in an organism.

The consequence that shows up in a culture experiment follows directly from the equilibrium. Serum contains binding proteins, and cultured cells secrete more of them into their own conditioned medium over the course of a run. Native IGF-1 added to such a system has a free concentration set jointly by how much was weighed out and by how much binding-protein capacity that particular medium happens to contain, which varies between serum batches. The LR3 construct engages that reservoir far more weakly, so the free concentration tracks the amount weighed out much more closely. That is a reproducibility property, and it is why the variant became the standard supplement for serum-free and low-serum media development. It is also why data generated with it and data generated with the native protein are not interchangeable: the two experiments arrived at their free ligand concentration by different routes.

Phrases attached to this molecule, and what each one is and is not a claim about

When this compound is written into a methods section, the useful discipline is to name the construct in full at first mention, state that it is a recombinant 83-residue variant rather than the native 70-residue sequence, and say explicitly whether the binding-protein evasion is the reason it was selected. A reader can then judge whether a comparison with native-sequence literature is legitimate. A methods section that simply says IGF-1 and cites native-protein papers has quietly discarded the one property that distinguishes the reagent.

Surface loss and the arithmetic of very dilute working solutions

Culture work with this material happens at concentrations far below anything the rest of a peptide catalog operates at, and the arithmetic makes the problem obvious. Take a one-milligram presentation and bring it into a stock at one milligram per milliliter. Diluting that stock a thousandfold gives one microgram per milliliter. Diluting again by a factor of ten to one hundred puts the working solution somewhere in the tens to low hundreds of nanograms per milliliter, which is the range the cell-culture literature typically describes. Between the first vial and the final well there are therefore three or four transfers, and the amount of protein present falls by four to five orders of magnitude across them.

At the bottom of that chain, the mass of protein in a tube is small compared with the amount of protein a plastic surface can adsorb. This is the dominant loss mechanism and it is not a rounding error. A dilute protein solution standing in an untreated polypropylene tube deposits an unpredictable fraction of its content on the wall, and a further fraction is left behind on the interior of every pipette tip it passes through. The loss is worst where the surface-to-volume ratio is highest, which means small volumes in large tubes and anything handled in a multiwell plate. It is also not reproducible between tubes, so it introduces scatter rather than a correctable offset.

The standard countermeasure is a carrier protein, most often serum albumin at a low percentage, added to the diluent so that it occupies the adsorptive sites and the molecule of interest stays in solution. This works, and it is why almost every supplier data sheet for a recombinant growth factor mentions it. It is also a substantial change to the matrix: it raises total protein by orders of magnitude relative to the analyte, it interferes with colorimetric protein assays and with some downstream analytical methods, and it introduces its own lot-to-lot variability. Carrier protein belongs in the written preparation record as a deliberate component, not in a footnote.

Freeze-thaw behaves differently for a folded protein than for a short peptide. The damaging event is not the low temperature but the moving ice interface, which concentrates solutes, shifts local pH as buffer components crystallize at different points, and presents a surface at which partially unfolded protein accumulates. Once a molecule exposes hydrophobic surface it nucleates further aggregation, and unlike the loose association that short peptides sometimes show, protein aggregation of this kind is effectively irreversible. Dilute solutions are more vulnerable than concentrated ones, because there is less protein to protect itself and proportionally more interface per molecule.

The failure mode that catches people is that none of this is visible. A solution that has lost most of its content to a tube wall looks exactly like one that has not. A solution carrying a large fraction of aggregate is often perfectly clear, because the aggregates are far too small to scatter visible light. Appearance is not a release criterion for this material.

Where material goes between the vial and the well

This is bench-measurement territory rather than storage guidance, and the general handling questions have their own dedicated posts. The point specific to this molecule is that the quantity being manipulated at the working end of the chain is small enough that container chemistry competes with the experiment. If a concentration-response series is noisy at the low end, surface loss is a more probable explanation than any property of the cells, and it can be tested directly by repeating the series in tubes prepared with carrier protein.

Endotoxin and host-cell protein as documentation gaps

Two impurity classes exist for this material that have no counterpart anywhere else in a synthetic peptide catalog, and both are usually absent from the certificate.

Endotoxin is lipopolysaccharide from the outer membrane of Gram-negative bacteria. If the expression host is such an organism, the material passes through a purification train in the presence of large amounts of it, and lipopolysaccharide is chemically robust, sticks to many surfaces, and is not removed by the steps that remove protein contaminants. It also has potent biological activity of its own, acting through innate immune receptors present on a great many cultured cell types. The practical hazard in an experiment is therefore not a safety question but an attribution question: a preparation carrying endotoxin will produce signaling and transcriptional changes that get recorded as effects of the protein. Because the activity is potent, the amount required to confound a sensitive readout is far below anything that would register as a purity problem on a chromatogram.

Host-cell protein is the population of the host organism's own proteins that survives purification. It is heterogeneous by definition, so it is not measured by chromatography against a single reference; the standard approach is an immunoassay raised against a lysate of the production host, reported as a quantity of host protein per quantity of product. Any of those proteins may be enzymatically active, and host proteases in particular can clip the product slowly in solution long after purification is complete, which is one mechanism behind material that degrades in storage without any obvious external cause.

The reason absence of these figures is a real gap rather than a formality is that neither impurity is visible in the assays that are reported. Endotoxin does not absorb meaningfully at the wavelengths used for peptide detection and it does not appear as a peak on a reversed-phase chromatogram. Host-cell protein is present at levels a purity chromatogram would show as baseline noise, if at all, and the immunoassay used to measure it is a completely separate method. A certificate that reports a high area percentage and an intact mass has not tested for either one and does not imply anything about either one. Absence of an assay on a document is evidence about the document, not about the material.

None of this makes research-grade material unusable. It makes the scope of the supplied documentation something to record before the experiment rather than to reconstruct afterward. Where a readout is sensitive to innate immune activation, the honest options are to obtain a lot-specific endotoxin figure, to run an endotoxin measurement in-house, or to include a control that would reveal the confound, such as a comparison against a heat-treated or protein-free preparation of the same diluent chain.

What each document establishes, and what its absence does and does not mean

The pattern across the table is that each missing row corresponds to a specific class of experimental confound rather than to a general reduction in confidence. That is the useful way to hold it. Requesting a lot-specific endotoxin figure is a reasonable thing to ask a supplier for; being told the assay is not run for research-grade lots is a legitimate answer that lets a laboratory decide whether to run one itself or to design the control in instead. What is not defensible is inferring from a clean chromatogram that neither impurity is present, since neither impurity would have appeared on it under any circumstances.

Area percent, gel and column answer three different questions

Purity is not one measurement, and for a protein of this size the available methods disagree with each other routinely. The disagreement is informative once it is clear what each method is separating on.

Reversed-phase HPLC separates on hydrophobicity, detects by absorbance, and reports the main peak's share of total integrated area. For a short synthetic peptide this is close to a complete description of purity, because the plausible impurities are chemically similar chains that the column resolves well. For a folded protein it is a narrower measurement than it appears. Species that differ only in disulfide pairing may co-elute, since their surface hydrophobicity can be similar. Aggregates may not elute at all, or may be retained on the column and never counted, which paradoxically raises the reported figure as aggregate content rises. Anything that does not absorb at the detection wavelength, including endotoxin and nucleic acid, is invisible. A high area percent from this method is a statement about the absorbing, eluting fraction only.

SDS-PAGE separates on apparent size after the protein has been denatured with detergent, and it is read by staining. Under reducing conditions it shows the polypeptide chain population: clipped forms appear as lower bands, and covalent aggregates are broken apart and disappear. Under non-reducing conditions the disulfide-linked species stay intact, so dimers and higher multimers show up as higher bands, and the monomer often migrates differently from its reduced form because the intramolecular bonds keep it compact. Running both conditions side by side is the informative comparison. Gel purity estimated from band density is a coarse figure and should not be read to a decimal place, but it detects things the chromatogram misses entirely.

Size-exclusion chromatography separates on hydrodynamic size in a non-denaturing mobile phase, which makes it the method that speaks to aggregate state as it exists in solution rather than after detergent treatment. It is the appropriate method for asking what fraction of the material is monomer. It has poor resolution between species of similar size, so it will not separate a correctly folded monomer from a misfolded one of the same mass, and it is sensitive to whether the mobile phase is a reasonable match for the solution the sample actually sits in.

The result of all this is that three figures on three methods, all honestly produced from the same vial, can differ substantially, and none of them is wrong. A high reversed-phase figure with a non-reducing gel showing multimer bands is a coherent picture, not a contradiction: it says the absorbing eluting material is homogeneous and that some fraction of the sample is disulfide-linked aggregate the column did not report. The methods are complementary, and a document set carrying only one of them has characterized one axis.

Three purity methods, three separations, three answers

The practical reading rule is to treat each figure as scoped to its method rather than as a competing estimate of one underlying truth. When two figures for a single lot disagree, the first question is not which is correct but which axis each was measuring, and the answer is almost always that they were measuring different ones. A document set that includes a reversed-phase chromatogram, a two-condition gel image and a size-exclusion trace describes the material along three independent axes, and that is a meaningfully different level of characterization from a single percentage.

Questions this primer gets asked

Where to read next

• Mass spectrometry for peptide identity confirmation how an intact mass match is produced, and the tolerance it is judged against
• Peptide purity versus peptide identity for RUO labs the two-figure framework this guide extends to a folded protein
• Freeze-thaw cycles in research materials the general handling case behind the aggregation section above
• Peptide storage and handling for laboratory research general storage practice, which this guide deliberately does not cover
• IGF-1 LR3 product page specification table, sequence and the long-form background layer
• Third-party lab testing and COAs