I read the following here.

In 1902, Emil Fischer and Frank Hofmeister independently proposed that proteins are formed from many amino acids, whereby bonds are formed between the amino group of one amino acid and the carboxyl group of another, resulting in a linear structure that Fischer termed “peptide”.

In pages 163-165 of this I read

A major part of the lecture dealt with the mode of linkage of amino acids in proteins. After considering various earlier proposals, Hofmeister presented several arguments in favor of the view that the amino acids are joined largely by amide bonds. He attached special significance to the biuret reaction - the purple color given with alkaline copper sulfate by proteins and by intermediate products of their enzymatic digestion (the so-called albumoses and peptones). … In favor of this theory, Hofmeister also offered evidence from physiological studies on the enzymatic cleavage of proteins and of hippuric acid (benzoyl-glycine).

Here, the following is stated on page 243.

… Fischer spoke about the isolation of amino acids from protein hydrolysates and suggested that proteins were made from amino acids linked together. Thus was born the Fischer-Hofmeister theory of protein structure.

What I’m getting from all this is that proteins undergo a similar reaction to biuret, some non-protein compound with peptide bonds (C(=O)N), and that you can find amino acids in the hydrolysate of proteins. Thus, proteins must be chains of amino acids linked by peptide bonds. Is that definitely true? To me there leaves some space for other stuff to be in proteins as well as amino acids and peptide bonds, but I could be missing some context.

Q: How is it known that proteins are polymers of amino acids?

Bonus question: I know that Berzelius coined protein, but did he give a definition of any kind?

  • 2
    $\begingroup$ There is a difference between how we discovered that proteins are largely amino acid polymers and how we can now observe that this is what they are. You describe the process of discovery but we can now observe many complete protein structures from high resolution X-ray analysis of their crystal structures. $\endgroup$
    – matt_black
    Commented May 1 at 10:50
  • $\begingroup$ To all: I know there is a difference between how this fact was discovered versus how it can be verified now. I don’t care how it was discovered, and I only went down this path because I don’t know X-ray or any other analysis. But if you have X-ray or NMR or other data that shows this I will be glad to look! $\endgroup$ Commented May 1 at 13:15
  • $\begingroup$ It wasn't always accepted that proteins are linear polymers of amino acids. A competing theory, the (now discredited) cyclol hypothesis due mainly to Dorothy Wrinch held (some) sway until the 1950s, Dorothy Hodgkin being a major debunker. $\endgroup$
    – tomd
    Commented May 17 at 8:18

3 Answers 3


Today, we are very certain proteins are polymers of amino acids. We also know that some proteins are bound to cofactors, are linked to sugars (glycoproteins), and have all kinds of other modifications. We know that proteins are made by ribosomes from amino acids linked to tRNA, in a particular sequence determined by mRNA.

There are many analytical methods that provide evidence that proteins are amino acids linked by peptide (amide) bonds. CD and IR spectroscopy shows signals characteristic of the amide functional group, NMR (e.g. 1H-15N HSQC) allows us to count the number of different linkages, X-ray crystallography visualizes the complete structure. You can use mass spectrometry to fragment a protein and figure out the sequence. People used to do Edman degradation, residue-by-residue hydrolysis of proteins to figure out their sequence. There are enzymes like trypsin and pepsin that hydrolyze after (or before) a given amino acid, also demonstrating that proteins contain amino acids linked by peptide bonds.

The OP already traced the very beginnings of figuring this out. It is a philosophical question when the transition was from hypothesizing the presence of peptide linkages to "knowing" that this is how proteins are made from amino acids.

  • $\begingroup$ Could you expand on the sentence about trypsin and pepsin? $\endgroup$ Commented May 1 at 3:55
  • $\begingroup$ I added links to the Wikipedia articles. If you are interested in the answer to your bonus question, ask a separate question. $\endgroup$
    – Karsten
    Commented May 1 at 10:42
  • 1
    $\begingroup$ Thanks, I know what they are, I just wasn’t sure what you meant by saying they hydrolyze after (or before) a given amino acid or how that shows that they are AA’s linked to peptide bonds $\endgroup$ Commented May 1 at 13:13
  • $\begingroup$ You can establish with model compounds that trypsin is a peptidase. Then, if it hydrolyses proteins, you can infer that proteins are also using peptide linkages. $\endgroup$
    – Karsten
    Commented May 1 at 13:40

Insulin was one of the the first proteins to be studied by x-ray crystallography and the history is well known and illustrates why we know the chemical composition of proteins

Dorothy Crowfoot Hodgkin was the first British woman to get a Nobel Prize in chemistry and that was awarded for her pioneering work on the structures of important biological molecules.

The structure of the protein insulin was one of her most important triumphs. And the history of her work illustrates how we now have a very good idea of the chemical structure of proteins.

The basic idea of her work and x-ray crystallography in general is that the atoms in crystals reflect x-rays to give particular patterns that can reveal the electron-density in crystals and that those densities can tell us the atomic structure of the crystal including which atoms are joined to which other atoms. Hence the chemical structure of the components of the crystal.

This works well for well-ordered large crystals. But proteins are uncooperatively hard to crystallise. Her first significant breakthrough was to get good crystals of insulin and a decent picture of their x-ray diffraction pattern as described by her 1935 Nature article.

But those first x-ray pictures were not good enough to reveal the complete structure.

Hodgkin herself describes some of the process of improving the results in this 1971 paper in the BMJ. As she says "But, within a little, we have recorded positions in three dimensions for the now exactly known numbers of atoms in the insulin molecules". This is essentially the breakthrough needed to verify other theories that proteins are, essentially, amino-acid polymers. Read that paper for a very good overview of how this was done.

In her day this was hard manual work. But as computers advanced and techniques got better such x-ray based approaches became almost routine (and an important part of drug design studying how small molecules interact with proteins). To take a random recent example, here is a modern paper dealing with high resolution structures of some modified insulins.

Hodgkin's pioneering work is the reason we now have certainty about the structure of proteins and that they are,essentially, amino acid polymers.

  • 1
    $\begingroup$ The work of Fred Sanger on the primary structure of insulin was surely more important in establishing that proteins are linear polymers of amino acids. The three-dimensional structure of insulin (by Dorothy Hodgkin and coworkers) came after Sanger's work. (The story goes that Sanger, invited to Oxford to view the structure, commented upon seeing the 3D model "At least we got the disulfides right") $\endgroup$
    – tomd
    Commented May 17 at 8:25
  • $\begingroup$ @tomd Good point. Sanger also got a Nobel for his contributions. $\endgroup$
    – matt_black
    Commented May 17 at 10:40

Proteins are by definition polypeptides, linear chains of aminoacids linked by amide bonds. There are various layers of information that allow one to conclude that a sample consists of protein. Atomic composition (molecular formula) is not unequivocal evidence (consider the melamine scandal in China) but is a useful constraint. Chemical reactivity is stronger evidence but spectroscopic information is now routinely used.

Any compound with an amide proton (ie a primary or secondary amide group) will result in a crosspeak in the amide region of a $\ce{^1H-^{15}N}$ HSQC NMR spectrum, so such peaks alone are insufficient evidence of connectivity between aminoacids (the chemical shifts and multiplicity do narrow the structural options however). There is however a large family of related higher-dimensional NMR experiments that allow peptide backbone connectivity through the amide bonds between neighboring aminoacids to be determined, as well documented in the literature ([1]).

Also NOE experiments that establish distance constraints between an amide proton and its neighbors or between nuclei on separate aminoacids can be useful ([2]) by providing 3-D structural information that can often be regarded as unambiguous evidence of the presence of a peptide bond.


[1] Editor(s): John Cavanagh, Wayne J. Fairbrother, Arthur G. Palmer, Mark Rance, Nicholas J. Skelton, Protein NMR Spectroscopy (Second Edition), Academic Press, 2007, ISBN 9780121644918, https://doi.org/10.1016/B978-012164491-8/50016-6.

[2] Wüthrich K, Billeter M, Braun W. Polypeptide secondary structure determination by nuclear magnetic resonance observation of short proton-proton distances. J Mol Biol. 1984 Dec 15;180(3):715-40. doi: 10.1016/0022-2836(84)90034-2. PMID: 6084719.

  • $\begingroup$ source on 2nd paragraph? $\endgroup$ Commented May 3 at 21:22
  • 1
    $\begingroup$ I added a reference. $\endgroup$
    – Buck Thorn
    Commented May 4 at 9:14
  • $\begingroup$ Is this link just the entire literature of NMR of biomolecules? $\endgroup$ Commented May 14 at 19:35
  • $\begingroup$ @powerful_bob If you want a proper answer you have to understand NMR and how it's applied to proteins. That book is an important reference that many practicing spectroscopists have used to understand how protein NMR works. $\endgroup$
    – Buck Thorn
    Commented May 15 at 6:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.