Re: Why proteins are not branched?

The fact that newly synthesized proteins are linear is due to the linearity of the transcripts (mRNA) needed for their synthesis. Of course, the linearity of mRNA derives from the linearity of DNA. However, you are correct in thinking that protein branching could occur post-translationally. In fact it does, and many proteins have both stable and transient branched forms. In almost all known cases, the branching is a result of enzymatic cross-linking. Let me give you some examples.

Example 1. Cross-linked fibrin forms during blood clotting. This cross-linking is catalyzed by the enzyme Factor XIIIa (a transamidase, also called transglutaminase) that cross-links a glutamine residue in one molecule of fibrin to a lysine residue in another molecule of fibrin. Individuals lacking factor XIII have a pronounced tendency to bleed.

Example 2. One of the major components of our connective tissue is elastin. It is found in elastic fibers, which have tremendous elastic strength and resilience. Elastic also contains many cross-linkages that make it both insoluble and elastic. This cross-linking occurs between lysine and an oxidized lysine, which has an aldehyde instead of an amine functional group, to give lysinonorleucine and between 4 (that's right 4!) lysines to give a desmosine tetravalent cross-link. The oxidation of lysine to the aldehyde in elastin is catalyzed by the enzyme lysyl oxidase, which is a copper-requiring enzyme. In copper-deficient animals, the elastin cross-links are poorly formed and the animals have a very strange looking connective tissue with little elasticity.

Example 3. Cross-links are also found in collagen, and occur between two lysine residues (also catalyzed by lysyl oxidase) to give an aldol cross-link.

Example 4. During protein degradation in the cytoplasm, there is a pathway dependent on the addition of one protein, named ubiquitin, to the epsilon-amino groups in lysine residues of proteins to be degraded. This ubiquitination is catalyzed by an enzyme called the ubiquitinating enzyme complex, and it can add multiple ubiquitin molecules to a protein. Thus, a protein that is ubiquitinated could have several N-termini. The ubiquitinated proteins are usually degraded fairly quickly in proteosomes to small peptides. However, several "mature" proteins, such as the receptor for platelet-derived growth factors, were found to be ubiquitinated. Oh by the way, many histones are constitutively ubiquitinated, and in those cases, the function of ubiquitination is not known, since the histones are rather long-lived.

So there you have it. Many proteins are known to be branched and in every case it appears to be enzymatic and involve specific amino acids, usually lysine, as you pointed out. Cross-linking may be more common than we know, and new biochemical methods of determinating molecular weights of proteins and rapid N- and C-terminal sequencing may give us more examples of cross-linking.