Simultaneous mass spectrometry analysis of cisplatin with oligonucleotide-peptide mixtures: implications for the mechanism of action

Although genomic DNA is the primary target of anticancer platinum-based drugs, interactions with proteins also play a significant role in their overall activity. In this study, competitive binding of cisplatin with an oligonucleotide and two peptides corresponding to segments of H2A and H2B histone proteins was investigated by mass spectrometry. Following the determination of the cisplatin binding sites on the oligonucleotide and peptides by tandem mass spectrometry, competitive binding was studied and transfer of platinum fragments from the platinated peptides to the oligonucleotide explored. In conjunction with previous studies on the nucleosome, the results suggest that all four of the abundant histone proteins serve as a platinum drug reservoir in the cell nucleus, providing an adduct pool that can be ultimately transferred to the DNA. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s00775-022-01924-9.


Aom 2 s tool: layout, algorithm, input parameters and data interpretation
The Aom 2 S tool as recently described elsewhere [1], [2] and a complete tutorial can be found at https://ms.epfl.ch/applications/oligonucleotides/. Aom 2 S is made up of a webpage consisting of three main modules (Home, Peak picking and recalibration, Report). The Home module is shown Figure S1 for the oligonucleotide S1, whereas the Settings window is described in Figure S2.
The first step for processing HR-MS/MS data with Aom 2 S involves the Settings window ( Figure   S1A) where calculation parameters need to be fine-tuned. For the stands S1 and S1c studied here, typical adducts allowed were the removal up to 11 protons and written (H + )-1-11 and the 3 variable groups allowed were (Pt ++ ), (Pt(NH3) ++ , (Pt(NH3)2 ++ ). Various MF filters can be added, if needed.
Then spectrum filters should be set in the Spectrum Filters section (Figure S2II) to correct peak picking and isotopic profile matching between theoretical and experimental data. These filters include: relative peak intensity threshold (typically 0.01), allowed experimental mass error in ppm (typically 5 ppm), and minimum similarity for peak matching (typically 80 %). Zone parameter refers to the comparison zone to be taken for analysis based on the monoisotopic peak position (typical values -0.5 to 3.5). In our particular case with Pt containing fragments, it should be extended to -1.5 to 5.5 to provide the coverage of the whole isotopic pattern. The expected fragment ion types should be indicated in the MS/MS fragment types section and the fragment nomenclature used is the one established by McLuckey et al [3]. and shown Figure S3. The Table   S1 summarizes the settings parameters used for the binding studies.
In a second step the experimental raw data is loaded by simply copying the data and directly pasting it as a txt file into the specially marked window (Figure 2IV). Once data are inserted, the experimental mass spectrum is reconstructed in red within the Mass spectra window. For matching, each individual isotopic peak in the theoretical spectrum is represented as an isosceles trapezoid with bottom and top widths related to the mass accuracy and resolution of the experimental spectra. As the top and bottom width values depend on the m/z and the type of mass analyzer used, these widths can be determined by automated peak picking of the experimental spectrum and generation of a regression curve of the peak width then applied to the simulated spectra ( Figure S4). Typically, the bottom width of the trapezoid is set as 3 times the width obtained from regression, while bottom width is set at the top width. For the matching process, the S3 comparison zone is defined relative to the monoisotopic peak. Subsequently, the sum of intensities of all peaks in the defined zone is normalized to 1 and each individual peak in the experimental spectra is matched against its simulated peak and the similarity score is calculated as: In a third step the tool matches experimental to theoretical peaks and generates a list of matched fragment ions with similarity scores. An additional post calibration step available in the Peak picking and recalibration module can be performed to improve mass accuracy and, therefore, similarity scores of matched ions. This post-calibration step requires a minimum number of peaks spread over the m/z range to be taken, with a minimum similarity, parameters that need to be defined by the user. After recalibration, a new corrected distribution of errors (in Da and ppm) between theoretical and experimental ions as a function of m/z is displayed. Figure S1. Main Aom 2 s Home module. Windows are labelled A to E corresponding to: (A) Settings (detailed description in Figure S2), (B) result as a list of matched fragments with variable groups, fragment sequence, fragment type, theoretical MF, experimental m/z, mass error (ppm), charge, intensity and percentage similarity (C) experimental MS spectra (red) overlaid with theoretical matched fragments (blue).
Fragment ions assigned are displayed on the top of the spectra and color coded based on fragment ion type, (D) fragment details for the selected fragments in the list B, (E) Similarity information window with zoom on overlaid theoretical (blue) and experimental (red) isotopic patterns for the selected ion, with differences in the intensity of isotopologues highlighted in yellow. Figure S2. Aom 2 s Settings window ( Figure S1, A), labelled I to IV: (I) oligonucleotide type and sequence, with any user-defined adduct and variable group, (II) bottom and top widths for of trapezoid for matching theoretical to experimental spectra, zone which specifies the mass range in m/z where theoretical spectra is overlapped with experimental spectra to calculate similarities (low and high specifies mass range in m/z before and after the monoisotopic mass of the peak of interest respectively), common zone (which specifies how similarity matching between theoretical and experimental spectra is performed), (III) MS/MS experiment parameters including fragment types, internal fragments and length of internal fragments, (IV) drag and drop spectra (accepts.txt files). Internal fragments resulting from double fragmentation usually occur at the a/w site. Figure S4. Peak picking and recalibration module, of Aom 2 S for the S1 oligonucleotide tandem mass spectra. Regression curve for peak widths vs m/z and error distribution graphs (in Da and ppm) before and after the post calibration step are shown for HR-MS/MS data analysis the ions species [S1+Pt(NH3)2-8H] 6-at m/z 701.4401.

5'
Ionization (H+)-1--11 (H+)-1--6 (H+)1-12 (H+)1-6 (P1) (H+)0-6 (P2)  Figure S5. Oligonucleotide binding study to cisplatin: Aom 2 s results from analysis of the full scan HRMS spectrum. A) overlay of experimental (ref) and assigned theoretical (blue) mass spectra after recalibration B) List of fragments assigned by the tool for each strand S1 and S1c. *: ions selected for further CID fragmentation.   Table S4. List of CID fragments assigned by Aom 2 s for the parent ion [S1c+Pt(NH3)2-8H] 6-(m/z 689.6103) including fragment type, molecular formula (MF), theoretical and experimental m/z, error (in ppm), intensity (expressed as percentage of the base peak in the spectra) and similarity scores in percentage.       Figure S9. Location of cisplatin binding site in S1c: histogram with the occurrence of nucleotide bases for all assignable platinum-free as well as platinated fragments (all fragments) following CID fragmentation of the peak at m/z 689.6103 for the strand S1c