Metal binding shows up in electrospray MS as a defined mass shift between the metal-free (apo) and metal-bound (holo) peptide. A divalent metal ion that binds in place of two protons adds its own mass minus two hydrogen masses: +61.914 Da for 64Zn, +60.914 Da for 63Cu, +55.920 Da for 58Ni. Divided by the charge state, that shift, the metal's isotope pattern and the stoichiometry across titration points tell you which metal binds, how many and whether the binding is specific or just adduct formation.
What mass shift does a bound metal ion produce?
In positive-mode electrospray, a peptide is charged by protons. When a metal ion with charge n+ binds, the overall charge of the ion is kept by releasing n protons. The shift of the neutral mass is:
Δm = m(metal) − n × m(H)
| Ion | Monoisotopic mass (Da) | Shift Δm (Da) |
|---|---|---|
| 64Zn2+ | 63.9291 | +61.9135 |
| 63Cu2+ | 62.9296 | +60.9139 |
| 58Ni2+ | 57.9353 | +55.9197 |
| 40Ca2+ | 39.9626 | +37.9469 |
| 56Fe3+ | 55.9349 | +52.9115 |
| 23Na+ (adduct) | 22.9898 | +21.9819 |
| 39K+ (adduct) | 38.9637 | +37.9559 |
The table already shows one trap: a calcium complex (+37.947) and a potassium adduct (+37.956) differ by only 0.009 Da. At charge 2+ that is 0.0045 m/z, which needs high resolution and careful calibration to separate.
A worked example at charge 2+
Take a peptide with a neutral monoisotopic mass of 1500.700 Da. The apo form at 2+ appears at m/z 751.357.
| Species | Neutral mass (Da) | m/z at 2+ | Shift from apo (m/z) |
|---|---|---|---|
| apo | 1500.700 | 751.357 | |
| + Na adduct | 1522.682 | 762.348 | +10.991 |
| + Cu | 1561.614 | 781.814 | +30.457 |
| + Zn | 1562.614 | 782.314 | +30.957 |
Zinc and copper sit 0.5 m/z apart at this charge state, which any modern instrument resolves. The more robust discriminator is the isotope pattern: zinc has five isotopes with 64Zn at 49 %, 66Zn at 28 % and 68Zn at 19 %, so the holo peak shows a characteristic envelope two mass units wide. Copper shows 63Cu and 65Cu at roughly 69 to 31. Nickel shows 58Ni and 60Ni at roughly 68 to 26. Matching the observed envelope against the simulated one for the complex is the identity check for the metal.
Specific binding or adduct?
A metal peak in the spectrum is not yet a binding site. Three observations separate specific complexes from nonspecific adducts:
1. Stoichiometry follows the titration. Adding metal in steps of 0.5, 1, 2 and 5 equivalents, a specific site saturates: the 1:1 complex rises and then stays, while 1:2 and 1:3 species appear only at excess. Nonspecific adducts grow roughly linearly with metal concentration and spread over several stoichiometries at once. 2. Charge-state behaviour. Specific complexes keep their relative abundance across charge states; adducts formed during desolvation often vary strongly between them. 3. Mutant or competitor control. Replacing a putative ligand residue, such as a histidine or cysteine, or adding a competing chelator should remove a specific complex and leave adducts largely unchanged.
Sodium and potassium adducts are the most frequent false positives. Ammonium acetate buffer, desalting and gentle source conditions keep them low.
Where is the binding site?
The intact mass tells you that a metal binds and how many. The site needs fragmentation of the holo ion. In MS/MS, fragments that still carry the metal shift by Δm, those that lost it do not. If the b ions up to residue 6 are unshifted and from residue 7 shifted, the metal is held by residues from 7 on. Collision-induced fragmentation can move or strip the metal, so ETD or ECD, which tend to preserve noncovalent contacts, are preferred for site mapping.
Limits of the statement
- Gas-phase abundance is not solution affinity. Ionisation efficiency of apo and holo forms can differ, and the ratio of peak heights is only an approximation of the fraction bound. Dissociation constants need a titration with a correction for response, or an orthogonal method such as isothermal titration calorimetry.
- Redox-active metals change. Copper(II) can be reduced to copper(I) in the source, which changes the shift by one hydrogen mass. Check for both forms.
- The oxidation state of iron and the release of three protons for Fe3+ are assumptions; verify against the isotope pattern and the exact shift.
- A buffer with phosphate or high salt suppresses the signal and promotes adducts. The experiment depends on volatile buffers.
Frequently asked questions
How do I calculate the mass of a metal-peptide complex?
Add the metal's monoisotopic mass to the neutral peptide mass and subtract one hydrogen mass per positive charge of the metal. For zinc(II): M + 63.9291 − 2 × 1.00783 = M + 61.914 Da. Then add the protons for the observed charge state and divide by the charge.
How can I tell zinc from copper binding by MS?
At charge 2+ the two complexes are 0.5 m/z apart, and the isotope patterns differ: zinc shows a wider envelope with 64Zn, 66Zn and 68Zn, copper a pair of 63Cu and 65Cu at about 69 to 31. Comparing the observed pattern with the simulated pattern for each complex is the reliable test.
How do I distinguish a sodium adduct from a metal complex?
By mass (sodium adds 21.982 Da, a divalent transition metal much more), by titration behaviour (adducts grow with salt, specific complexes saturate) and by a control without the metal or with a chelator.
Can native MS give a binding constant?
It can give an estimate from a titration if the response of apo and holo forms is comparable or corrected. For a defensible constant, confirm with an orthogonal solution method.
Which fragmentation method localises a metal site?
ETD and ECD preserve metal contacts better than collision-induced dissociation. Fragments that carry the metal shift by the metal's Δm, so the boundary between shifted and unshifted fragments marks the site.