Peptide solubility: predicting it from sequence

Assign a charge to each ionisable residue at neutral pH: aspartate and glutamate negative; lysine, arginine and histidine positive, with histidine partial; plus the free N-terminus positive and the C-terminus negative unless amidated. Sum them.

A net charge with magnitude of roughly one or more per five residues generally predicts water solubility. Charged groups are well solvated, and electrostatic repulsion between molecules discourages the intermolecular association that precedes precipitation.

A net charge near zero is the difficult case. Without electrostatic repulsion, hydrophobic association dominates and the peptide is prone to aggregation regardless of how many polar residues it contains.

Where net charge is unhelpfully close to zero at neutral pH, shifting pH away from the isoelectric point creates it. For a basic peptide, dilute acetic acid protonates basic side chains and confers net positive charge. For an acidic peptide, dilute ammonium bicarbonate or a mildly basic buffer deprotonates acidic side chains and confers net negative charge.

Both reagents are chosen partly because they are volatile and can be removed by lyophilisation if the peptide needs to be recovered as a solid, which non-volatile buffers make awkward.

The direction matters and getting it backwards makes things worse: acidifying an already acidic peptide drives it towards neutrality and out of solution. Determine whether the sequence is net basic or net acidic before reaching for a bottle.

Beyond charge, the fraction of hydrophobic residues — leucine, isoleucine, valine, phenylalanine, tryptophan, methionine, alanine — sets the baseline difficulty. Above roughly half hydrophobic content, direct aqueous dissolution becomes unlikely regardless of net charge.

Secondary structure propensity compounds it. Sequences that readily form beta-sheet, particularly alternating hydrophobic patterns, associate into ordered aggregates that resist redissolution once formed. This is the mechanism behind peptides that dissolve on first reconstitution and refuse to redissolve after a freeze-thaw cycle.

For these, a strong solvent first is the reliable route: a minimum volume of DMSO, or acetonitrile, or in stubborn cases a chaotrope such as guanidinium chloride, to produce a clear concentrated solution, then dilution into the working buffer with mixing. Dilute slowly — adding the organic stock rapidly into a large aqueous volume creates a transient high-concentration boundary where precipitation happens before mixing completes.

The solvent that dissolves the peptide has to be compatible with what happens next. DMSO is effective and widely tolerated in cell-based work at low final percentages, but it is not free — it has biological effects of its own, and the tolerable ceiling depends on the cell type and the readout.

This constrains the stock concentration. A stock that requires a final DMSO percentage above what the assay tolerates is not usable, however clear it looks. Work backwards from the acceptable final solvent fraction to the required stock concentration before deciding how much solvent to add.

Always include a vehicle control at the same final solvent concentration. Without it, any effect observed is confounded with the solvent, and the experiment cannot distinguish the two.