A summary of published research, with citations. It states no amount, schedule or route, and makes no claim about what this compound does in a person. Supplied for laboratory research use only.
On this page · 7 sections
A peptide and a protein are both chains of amino acids joined by peptide bonds. The difference is length and, more importantly, structure. By the usual convention, chains of up to about 50 amino acids are peptides and longer ones are proteins, though regulators and textbooks draw the line in different places. In practice the feature that separates them is folding: most proteins fold into a stable three-dimensional shape that their function depends on, while most short peptides do not.

That definition has consequences well beyond vocabulary. It decides how a molecule is made, how it breaks down, how it is tested, and in some countries which regulatory pathway it follows. If you want the basics of what peptides are and what they do in the body, start with what are peptides or peptides explained. This page is about the boundary.
How many amino acids make a peptide?
There is no single official number, and anyone who quotes one is quoting a convention.
| Term | Rough size | How it is used |
|---|---|---|
| Dipeptide, tripeptide | 2 or 3 residues | Precise counts; GHK-Cu is a tripeptide bound to copper |
| Oligopeptide | Roughly 2 to 20 residues | Short chains, usually without stable structure |
| Peptide | Up to about 50 residues | The most common biochemistry convention |
| Polypeptide | Any single unbranched chain | A chemical description, used for long peptides and for single protein chains |
| Protein | Above about 50 residues, or any folded functional unit | Often one or more chains with a defined fold |
Some sources put the line at 40 residues, some at 50, some closer to 100. The United States FDA uses 40: under its definition, a chain of more than 40 amino acids is regulated as a protein, which is why insulin products moved to the biologics pathway there in 2020. Other bodies use other cut-offs or none at all.
The number is fuzzy because it is a proxy. What people actually want to know is whether the chain is long enough to fold into a stable, self-contained structure. Around 40 to 50 residues is roughly where that starts to become possible, but some shorter chains fold and some longer ones never do.
A useful rule of thumb for mass: an average amino acid residue weighs about 110 daltons, so 50 residues comes to roughly 5,500 daltons. That is why "about 5 kDa" also appears as an informal boundary.
Why does folding matter more than length?
Because folding is what turns a chain into a machine.
Christian Anfinsen's work, summarised in his 1973 Nobel lecture paper, established that a protein's amino acid sequence contains the information needed to reach its native three-dimensional structure (PMID 4124164). A typical protein folds so that water-avoiding side chains are buried in a core and charged or polar ones face outward. The result is a defined shape with pockets, surfaces and moving parts. An enzyme's active site, an antibody's binding loops and a receptor's signalling domain all depend on that shape. How chains find their fold so quickly became a research field in its own right, which Dill and MacCallum reviewed fifty years after the question was first posed (PMID 23180855).
Most short peptides have no such core. In water they exist as a flexible, shifting mix of shapes. Many adopt a defined conformation only when they bind their target, often a helix lying in a receptor's groove. That is a large part of why a short peptide can be a precise signal without being a stable object, and it is covered from the receptor's side in how peptide receptors work.
Folding also defines what "damage" means. A protein can be denatured, which means unfolded, without a single covalent bond breaking, and an unfolded protein usually loses its function and may clump together. A short peptide with no fold to lose is mainly at risk from chemical changes to the chain itself.
How are peptides and proteins made?
The two are usually made in completely different ways.
Peptides: solid-phase synthesis. Bruce Merrifield's method, recognised with the 1984 Nobel Prize in Chemistry, builds a chain one amino acid at a time while it is anchored to an insoluble resin bead (PMID 3961484). Each cycle removes a protecting group, couples the next residue and washes away excess reagent. Because the chain stays on the bead, purification between steps is a simple wash.
The method's limit is arithmetic. If each coupling succeeds 99% of the time, a 10-residue chain finishes with about 90% of chains full-length, but a 50-residue chain finishes with only about 60%, and the failures are near-identical sequences missing one residue, which are hard to separate. That is why routine synthesis works best for shorter chains, and why longer targets are often made in pieces. Native chemical ligation, introduced by Dawson and colleagues in 1994, joins two unprotected synthetic segments with a natural peptide bond at the junction, and was used to make a small folded protein (PMID 7973629).
Proteins: recombinant expression. Most proteins are made by putting the gene into living cells, such as bacteria, yeast or mammalian cell lines, and letting their ribosomes build the chain. Insulin was the landmark. Goeddel and colleagues cloned synthetic genes for the two insulin chains into E. coli, purified the chains separately, then combined and reoxidised them to form insulin (PMID 85300). Living cells can also add modifications such as sugar chains that synthesis cannot easily reproduce, which is why many therapeutic proteins are made in mammalian cells.
The boundary is not absolute in either direction. Some peptides are made recombinantly, and some small proteins can be synthesised chemically. As a practical matter, though, if a research compound is sold as a lyophilised synthetic powder, it is almost certainly a peptide made by solid-phase synthesis.
Is insulin a peptide or a protein?
Both, depending on who is counting, which makes it the best example of why the boundary is fuzzy.
Human insulin has 51 amino acids in two chains: an A chain of 21 residues and a B chain of 30, held together by two disulfide bonds, with a third disulfide inside the A chain. Neither chain alone exceeds 40 residues, but the molecule as a whole does. It folds into a defined structure, and it associates into larger assemblies in solution. Biochemistry textbooks often call it a peptide hormone; the FDA's 40-residue rule classes it as a protein. Health Canada also regulates insulin as a biologic.
Other molecules sit close to the line. A few examples from this catalogue and from approved medicine:
| Molecule | Length | Usually called |
|---|---|---|
| GHK-Cu | 3 residues plus copper | Peptide |
| BPC-157 | 15 residues | Peptide |
| Tirzepatide | 39 residues, with a fatty-acid side chain | Peptide |
| Tesamorelin | 44 residues | Peptide, despite exceeding 40 |
| Insulin | 51 residues in two chains | Either, depending on the source |
| A typical antibody | About 1,300 residues in four chains | Protein |
Tesamorelin is a useful counterexample: a 44-residue chemically modified analogue of growth hormone-releasing hormone that nearly everyone still calls a peptide. Names follow use and history as much as any cut-off. Catalogue pages for GHK-Cu, BPC-157, tirzepatide and tesamorelin list each sequence's formula and molecular weight.
How do stability and handling differ?
The practical differences follow directly from folding.
Proteins are vulnerable to physical stress. Heat, freezing and thawing, vigorous agitation and contact with air-water interfaces can unfold them, and unfolded protein tends to aggregate. Aggregates are both inactive and a concern in their own right. This is why protein medicines come with strict handling instructions.
Short peptides are more robust to physical stress because there is little fold to disrupt. Their main enemies are chemical: oxidation of methionine, cysteine and tryptophan side chains, deamidation of asparagine (especially next to glycine), and cleavage at certain sequences such as aspartate-proline. Water speeds most of these reactions, which is why research peptides are supplied freeze-dried. The stability of lyophilised peptides and whether shaking damages peptides each have their own article; the short version is that the answer depends on length and sequence.
Testing differs too. A short synthetic peptide has one exact intended mass, so mass spectrometry can confirm identity directly, and HPLC can resolve most closely related impurities. A protein may carry several sugar variants, can exist as a mixture of folded, misfolded and aggregated forms, and usually needs additional methods, including size-based separations and activity assays, to show it is the right molecule in the right shape. See HPLC vs LC-MS peptide testing for what each peptide method can and cannot confirm.
How do research peptides differ from approved peptide medicines?
Peptide drugs are a large and established class. Muttenthaler and colleagues counted more than 80 peptide drugs on the market since insulin's introduction, across diabetes, cancer, osteoporosis, multiple sclerosis, HIV infection and chronic pain (PMID 33536635), and Wang and colleagues reviewed the chemical and biological production and modification methods behind them (PMID 35165272).
An approved peptide medicine is more than a sequence. It is a particular manufacturer's product, made by a validated process under good manufacturing practice, with a defined impurity profile, sterility assurance where relevant, and a dossier of clinical trials that a regulator such as Health Canada has reviewed for a specific indication. A research-grade synthetic peptide with the same sequence has none of that review. It may be made by a different route, carry different impurities or counter-ions, and has not been assessed for use in people. Matching a sequence does not make two products the same product.
That distinction holds at both ends of the length scale. Tirzepatide is an authorised medicine in Canada in its branded form; tirzepatide sold as a research chemical is not that medicine. Many other research peptides have never been approved anywhere, and their evidence stops at cell or animal studies.
Nothing here is medical, dosing, or therapeutic guidance.
Frequently asked questions
What is the main difference between a peptide and a protein? Length and structure. Peptides are shorter chains, conventionally up to about 50 amino acids, and most have no stable fold on their own. Proteins are longer and usually fold into a defined three-dimensional shape that their function depends on. The cut-off varies between sources; the FDA, for example, uses 40 amino acids.
Is insulin a peptide or a protein? It sits on the boundary. Insulin has 51 amino acids in two chains joined by disulfide bonds, and it folds into a defined structure. Many biology textbooks call it a peptide hormone, but under the FDA's 40-residue rule it is a protein, and both the FDA and Health Canada regulate insulin products as biologics.
Are all peptides made synthetically? No. Many peptides occur naturally, and some peptide medicines are made recombinantly in cells. Most research peptides, though, are made by solid-phase synthesis, which builds the chain one amino acid at a time on a resin. Proteins are usually made recombinantly, because synthesis becomes inefficient as chains get longer.
Why are peptides more stable than proteins? Mostly because they have no fold to lose. Proteins can be unfolded by heat, freezing or agitation, then aggregate. Short peptides are more resistant to that physical stress but still degrade chemically, through oxidation, deamidation and cleavage at certain sequences. That chemistry is why research peptides are usually supplied freeze-dried.
Is a research peptide the same as an approved peptide drug? Not even when the sequence matches. An approved medicine is a specific manufacturer's product, made by a validated process and reviewed by a regulator on the basis of clinical trials. A research-grade peptide has not been through that review and is not for human use.
Selected Research References
- Anfinsen CB. Principles that govern the folding of protein chains. Science. 1973;181(4096):223-230. pubmed.ncbi.nlm.nih.gov/4124164
- Dill KA, MacCallum JL. The protein-folding problem, 50 years on. Science. 2012;338(6110):1042-1046. pubmed.ncbi.nlm.nih.gov/23180855
- Merrifield B. Solid phase synthesis. Science. 1986;232(4748):341-347. pubmed.ncbi.nlm.nih.gov/3961484
- Dawson PE, Muir TW, Clark-Lewis I, Kent SB. Synthesis of proteins by native chemical ligation. Science. 1994;266(5186):776-779. pubmed.ncbi.nlm.nih.gov/7973629
- Goeddel DV, Kleid DG, Bolivar F, et al. Expression in Escherichia coli of chemically synthesized genes for human insulin. Proceedings of the National Academy of Sciences of the United States of America. 1979;76(1):106-110. pubmed.ncbi.nlm.nih.gov/85300
- Muttenthaler M, King GF, Adams DJ, Alewood PF. Trends in peptide drug discovery. Nature Reviews Drug Discovery. 2021;20(4):309-325. pubmed.ncbi.nlm.nih.gov/33536635
- Wang L, Wang N, Zhang W, et al. Therapeutic peptides: current applications and future directions. Signal Transduction and Targeted Therapy. 2022;7(1):48. pubmed.ncbi.nlm.nih.gov/35165272
All products sold by NovaEvum are for research and laboratory use only. They are not intended for human or animal consumption, diagnosis, treatment, cure, or prevention of any disease.
