Sequence-dependent attack on peptides by photoactivated platinum anticancer complexes

Octahedral anticancer platinum(iv) complexes such as trans,trans,trans-[Pt(N3)2(OH)2(pyridine)2] (1) can target peptides (and proteins) by sequence-dependent platination and radical mechanisms when activated by UVA or visible light; the specific products are highly dependent on their amino acid composition of the peptide.


Electronic Supplementary Information
. UV-vis spectra of (a) complex 1, (b) + Substance P, (c)+ [Lys] 3 -Bombesin, on irradiation with blue light for the times shown (min, decreasing absorbance at 300 nm).      Tables S8-S11. Aassignments for individual ECD MS/MS spectra of the isolated oxidised peptide species. S8: Substance P+O; S9: K 3 -Bom+O; S10: K 3 -Bom+2O; S11: K 3 -Bom+3O.  Electronic Supplementary Material (ESI) for Chemical Science. This journal is © The Royal Society of Chemistry 2018 Table 13. Relative intensities of peaks for assigned species in the FTMS of irradiated complex 1+ (A) Substance P, and (B) K 3 -Bom reaction mixtures, with and without added tryptophan.   Table S14. Peptide fragment assignments and associated mass errors for the ECD MS/MS spectra of the [K 3 -Bom+2H] 2+ species with and without the use of co-isolated calibrant ions.

Effects of platinum on MS/MS fragmentation
Scheme S1: Platinum-centred side chain losses from methionine commonly observed during ECD MS/MS studies of platinated peptides. Scheme S2. Proposed mechanism for the loss of the Pt(II) modification from the peptide during ECD; ligands shown vary with the nature of the Pt adduct. Figure S1. UV-vis spectra of (a) complex 1, (b) + Substance P, (c)+ [Lys] 3 -Bombesin, on irradiation with blue light for the times shown (min, decreasing absorbance at 300 nm).   Tables S7-9 lists the assignments for each species. Pt* indicates the platinum based modification associated with each species (see Figure 3, main text, also included in assignment Tables S7-9, above). Assignments for each species are in Tables S10-12. Pt* indicates the platinum-based modification associated with each species (see Figure 4, main text, also included in assignment Tables S10-12, above).

a) b) c)
Tables S2-S7. Assignments for individual ECD MS/MS spectra of the isolated platinated peptide species. Marked species were used for internal calibration; in tables with no marked species, coisolated calibrant ions were used instead of MS/MS fragments, as discussed in the Experimental section.   Tables S3-S6 for spectra S2a-d respectively.
Tables S8-S11. Aassignments for individual ECD MS/MS spectra of the isolated oxidised peptide species. S8: Substance P+O; S9: K 3 -Bom+O; S10: K 3 -Bom+2O; S11: K 3 -Bom+3O. Marked species were used for internal calibration; in tables with no marked species, co-isolated calibrant ions were used instead of MS/MS fragments, as discussed in the Experimental section.

Effects of platinum on MS/MS fragmentation
The presence of the Pt II centre caused a dramatic effect on both electron based and collision based MS/MS experiments (ECD and CAD). The charged Pt centre altered the ECD MS/MS fragmentation pattern by providing a fixed charge at the point of interaction (e.g. His 11 of K 3 -Bom), but also enhanced side chain losses of amino acid groups present. Electron capture at the Pt II centre/electron transfer of previously captured electrons on the peptide also caused additional ligand loss from the bound complex and additional fragmentations from amino acid residues. CAD MS/MS of platinum complex-containing species caused gas-phase dissociation of the platinum-bound ligands, even at low energies (<5-7 V), creating a reactive Pt II centre, which quickly cyclised with available peptide groups and produced uninformative fragmentation spectra.
The lack of backbone fragmentation is attributable to both the positive charge on Pt II , without additional protons along the poly(amino acid) backbone, causing the normal electron capture dissociation mechanism to be disrupted/impeded. Electron capture by Pt can dominate, causing release of bound ligands and side chain losses from nearby amino acids.

Effect of Pt on electron dissociations
The ECD MS/MS spectrum of each Pt adduct showed a series of high intensity, singly-charged peaks close to the charge-reduced species (Fig. S3a-d and S4a-d). These peaks did not match any calculated usual c or z ions, nor any side chain loss commonly observed during ECD, 1 but the highest m/z peak did correspond to the mass of the peptide+Pt (e.g. [Substance P+Pt] + for the [Substance P+Pt(py)2(N3)+H] 2+ species), indicating an ECD-induced loss of ligands from Pt. The peak spacings were 15.02374m/z and 31.97186m/z. This corresponds to two separate losses of a terminal methyl group . CH3 (-0.16 ppm) and loss of . CH3-SH (48.003920 m/z, -0.25 ppm) (SI Fig. S7). Considering the sequences of Substance P (RPKPQQFFGLM-NH2) and K 3 -Bom (pEQKLGNQWAVGHLM-NH2) the only sulphur atom present is within the thioether side chain of methionine. The observed side chain loss differs from the usual side chain losses commonly observed for methionine which are C3H6S or C2H4S, 1 suggesting that the Pt centre has caused the unusual fragmentation of the peptide under ECD. 2 For this to occur, Pt would have to be in close proximity to the methionine residue, as seen in previous work by Li et. al., 3 yet all current data points clearly to the binding of Pt to the histidine (His 11 ) and lysine (Lys 3 ) residues, depending on the amino acid sequence. Methionine sulphur is known to be a strong ligand for Pt(II), 4,5 and the side chain losses here indicate a localised interaction. As a result this localised reaction must occur after the electron capture-induced loss of the Pt-bound ligands. The methionine side chain can then coordinate to vacant sites on the Pt centre and account for the side chain losses observed. A proposed mechanism for the interaction of a methionine residue and the subsequent loss of the terminal methyl and CH3-SH species from the peptide is shown in Scheme S1. This scheme correlates with the side chain losses observed during this study, with the predicted interaction with methionine found in several reported Pt peptide MS/MS studies, 5,6,7 Pt-centred capture of an electron during ECD, 3 and commonly observed in Ptligand binding studies in solution. 8,9 Scheme S1: Platinum-centred side chain losses from methionine commonly observed during ECD MS/MS studies of platinated peptides.
Scheme S2 shows a proposed mechanism for the loss of the platinum complex from peptides during electron capture, often observed for singly bound Pt modifications in this study (single binding site on peptide).
ECD MS/MS of the diplatinated K 3 -Bom species produced a similar array of side chain losses as observed in mono-platinated species, however a more extensive number and variety of unusual sidechain losses and fragmentation channels were observed (SI Figure SF10). Unfortunately as observed in the ECD MS/MS of the [K 3 -Bom+Pt(py)2] 2+ species, without an additional proton, ECD MS/MS was unable to provide enough sequence-informative fragments to assign binding positions of the two platinum complexes, though based on the data observed for the previous modifications, tentatively these modifications would be expected to be present on the His 12 and Met 14 residues (for {Pt(py)2}) and the Lys 3 (for {Pt(py)2(N3)} + ).

Effect of Pt on slow heating dissociations
It should be noted that although more commonly used Collisionally Activated Dissociation (CAD) and Infra-Red Multi Photon Dissociation (IRMPD) MS/MS techniques were attempted on the observed platinum-peptide adducts, these ergodic/"slow heating" fragmentation methods break bonds in the lowest energy pathways in a molecule. 10,11 For peptides, these usually involve the amide bonds along the backbone. However, CAD and IRMPD on the Pt adducts, resulted in a series of peaks corresponding to the loss of each platinum-bound ligand present, loss of the platinum modification (unmodified peptide), and a series of unassignable Pt and non-Pt containing peaks due to sequence scrambling commonly observed during the study of peptides and (gas-phase) labile post translational modifications such as metallation. 12,13 In contrast, ligand loss was observed due electron capture at the metal centre during ECD MS/MS of the Pt adducts studied here. Electrons could also be captured at points along the backbone according to the accepted ECD MS/MS mechanism, 14 allowing normal ECD fragmentation and production of interpretable c/z ions.
Although the Pt complex showed the potential to bind to multiple groups present on the peptides studied; no crosslinking was observed, despite the low drug-to-peptide mol ratio (0.5:1) which might have promoted this possibility. The lack of crosslinking could be due to the short irradiation (reaction) times used (compared to the ~24 hour times used for cisplatin crosslinking studies) 5 and the low concentrations of reagents. 6 Square-planar Pt(II) has previously shown to act as an effective crosslinking agent, even for proteins such as calmodulin. 6