Peptide Backbone

Strip away every side chain and what remains is the backbone: a repeating unit of amide nitrogen, alpha carbon, and carbonyl carbon, run end to end from the N-terminus to the C-terminus. It is identical in every peptide regardless of sequence. What varies is the geometry it is allowed to take, and that is tightly constrained. The amide bond itself is planar and effectively cannot rotate, so conformational freedom is confined to two dihedral angles at each alpha carbon, conventionally called phi and psi. Plot the sterically permitted combinations and you get the Ramachandran map, whose allowed regions correspond to the familiar secondary structures.

The backbone is also the hydrogen bonding network. Each amide contributes a donor in its NH and an acceptor in its carbonyl oxygen, and it is these groups, not the side chains, that hold an alpha helix or a beta sheet together. That same abundance of polar donors is why peptides are poor at crossing membranes: burying those hydrogen bonds costs energy that a small lipophilic drug never has to pay.

Modifying the backbone is therefore a distinct design lever from modifying side chains. N-methylation deletes a donor and adds bulk, beta-amino acids insert an extra carbon and change the repeat spacing, azapeptides swap an alpha carbon for a nitrogen, and peptoids move the side chain onto the nitrogen entirely. Each buys protease resistance, because proteases read backbone geometry, and several improve permeability.

The misconception is that a peptide is defined by its sequence and the backbone is inert scaffolding. Two molecules with the same letters can behave completely differently if one is N-methylated or contains beta-residues, and a certificate of analysis that reports only a sequence and a mass will not necessarily distinguish them.