Histone propionylation pre-digestion

The samples containing isolated histones were dissolved in 40 µL of 50 mM NH4HCO3 (Sigma-Aldrich), pH 8.0. Then, a propionylation reagent was freshly prepared by mixing propionic anhydride (Sigma-Aldrich) with acetonitrile (Honeywell Burdick & Jackson, Morristown, NJ, USA) in the ratio 1:3 (v/v). Caution should be taken when working with propionic anhydride as it is a combustible and corrosive substance. The propionylation reagent was added to samples in 1:4 (v/v) ratio, after which NH4OH (Sigma-Aldrich) was quickly added to the samples with a ratio of 1:5 (v/v) to re-establish pH 8.0 to the solution. The samples were mixed immediately by vortexing, the pH was checked, and they were incubated at room temperature for 15 min. Next, the samples were dried, resuspended in 40 µL of 50 mM NH4HCO3, and the propionylation step was repeated to ensure complete derivatization, as it was done by Sidoli et al. [23].

Proteolytic digestion with trypsin

The histone samples were resuspended in 50 mM NH4HCO3 to achieve an optimal concentration of 1 μg/μL or trypsin (Trypsin Gold MS Grade, Promega, Madison, WI, USA) was added to the samples at a weight/weight (wt/wt) 1:10 ratio. They were incubated at 37 ℃ for 6–8 h. Digestion was stopped by freezing in − 80 ℃, and the samples were dried in a SpeedVac.

Histone propionylation post-digestion

The histone tryptic digests were resuspended in 30 µL of 100 mM NH4HCO3. Then, histone propionylation was performed as described in the Histone propionylation pre-digestion section, and the propionylated samples were dried.

Sample clean-up using mixed cation exchange (MCX)

The dried samples were acidified by resuspending it in 0.4% formic acid (LC/MS Optima, Thermo Fisher Scientific) in water (LC–MS grade, Honeywell Burdick & Jackson). The Oasis MCX cartridges (Waters, Milford, MA, USA) were equilibrated with 1 mL of 1:1 methanol: water (both LC–MS grade, Honeywell Burdick & Jackson). Then the samples were applied to the MCX cartridges. After that the cartridges were washed with 1 mL of 5% methanol/0.1% formic acid in water and then with 1 mL of 100% methanol. The bound peptides were eluted with 1 mL of 50 µL of 28% NH4OH solution (≥ 99,99% trace metal basis, Sigma-Aldrich) in 950 µL of methanol. The eluted peptides were dried in SpeedVac.

MS sample preparation

The samples were resuspended in 0.1% formic acid in water, and the peptide concentrations were determined using NanoDrop analysis at 230 nm.

For development and validation of the absolute quantification (AQUA) LC–MS/MS method, the prepared histone CIC, CIM and MIM donors’ samples, as well as the pooled histone matrix sample, were resuspended to 0.5 µg/µL with 0.1% formic acid in water. An aliquot of each of the investigated donor samples from all analyzed conditions (CIC, CIM, MIM) were pooled together to create a quality control (QC) sample that was run in each batch with the particular donor samples. The AQUA approach involves the use of SIL standards of the identical sequence to the investigated peptides to determine the amounts of the endogenous analytes. Following the recommendations in the literature to determine the histone H3 K14 acetylation stoichiometry using AQUA approach [15, 16] there were two types of internal standard peptides synthetized—an acetylated (Ac) and nonacetylated K14 that also included the 13C/15N isotopically labelled C-terminal arginine (R) residue (heavy). This resulted in 10 Da shift in molecular weight (MW) compared to the native peptide present in the investigated samples. There were also Ac and nonacetylated unlabeled peptides synthetized (light) that matched the sequence with those endogenous native peptides occurring in the donors’ samples, which were used during the method validation as internal standards for creating a reverse calibration curve. Both, heavy and light peptides (AQUA QuantPro standards, 5 pmol/µL in 50/50 ACN/water, custom synthetized by Thermo Fisher Scientific), were of > 97% purity and > 99% isotopic enrichment as recommended by Hoofnagle et al. [24]. Moreover, the nonacetylated K residues of both heavy and light peptides, i.e. K9 in acetylated or K9 and K14 in nonacetylated peptides, were propionylated (Poy). Because the target samples were propionylated during sample preparation, the propionyl groups at the K residues that were not post-translationally modified were introduced. Therefore, there were four histone H3 peptide standards used in the present study: K9[Poy]K14[Ac], K9[Poy]K14[Ac]-heavy, K9[Poy]K14[Poy], and K9[Poy]K14[Poy]-heavy. The sequences of each of the peptides as well as their calculated molecular weight is presented in Table 1. Prior to the LC–MS/MS analyses and to achieve final concentration of 50 fmol/µL SIL standards in the sample as recommended in Clinical Proteomic Tumor Analysis Consortium (CPTAC) guidelines, the donors’ CIC, CIM, and MIM samples were spiked with SIL standards of the histone H3 K9[Poy]K14[Poy]-heavy and K9[Poy]K14[Ac]-heavy in the volumes that comprised 10% of the sample volume [25]. The histone matrix samples used during the LC–MS/MS method development and validation process were also spiked with K9[Poy]K14[Poy]-heavy and K9[Poy]K14[Ac]-heavy peptides in the volumes that comprised 10% of the sample volume, to obtain seven dilutions of the final concentrations of 0.5, 1.25, 5, 50, 500, 1250, and 2500 fmol/µL and a QC of 12.5 fmol/µL of heavy peptides. The matrix blank samples (0 fmol/µL of the heavy peptides concentration) were spiked with the same volume of the 50/50 ACN/water instead of the heavy peptides’ dilution. All the histone samples for method development and validation were additionally spiked with 100 fmol of non-labelled H3 K9[Poy]K14[Poy] and K9[Poy]K14[Ac] peptides that served as internal standards during creation of the reversed calibration curves, similarly as in [26].

Table 1 Histone H3 peptides analyzed in the present study (the amino acid, on which the SIL was introduced, is depicted bold and underlined)LC–MS/MS analyses in the MRM mode

Histone tryptic digests were analyzed using reverse phase ultra-high-performance LC–MS/MS technique on an Acquity I-Class UHPLC (Waters) coupled with a QTRAP 6500+ mass spectrometer (SCIEX, Framingham, AM, USA), and controlled by Analyst software (SCIEX). The samples (4 μL) were injected in three technical replicates via an autosampler kept at 8 ℃ into the Omega (1.6 μm, PS C18, 100 × 2.1 mm, 100 Å) reversed-phase high performance liquid chromatography (RP-HPLC) column with SecurityGuard ULTRA Cartridge (Phenomenex, Torrance, CA, USA). 0.1% (v/v) formic acid in water (solvent A) and 0.1% (v/v) formic acid in acetonitrile (ACN) (both LC–MS grade from Honeywell Inc.) at a flow rate of 0.2 mL/min were used for separation of the samples’ components using a linear gradient of 5% to 50% solvent B in 25 min, then ramped to 95% in 3 min, after which the column was re-equilibrated with 5% solvent B for 10 min. Electrospray ionization of the samples was achieved in TurboV Ion Source (SCIEX) operated at a positive 5,5 kV spray voltage, with curtain gas at 35 psi, ion source gas 1 and 2 at 45 psi, temperature 450 ℃, CAD collision gas (CAD) flow medium and then analyzed in MRM mode. We used an absolute quantification (AQUA) approach [15] to determine the histone H3 K14Ac absolute abundance and stoichiometry in our samples. The AQUA strategy employs liquid chromatography (LC) separation with MRM detection, and the levels of endogenous peptide are determined by comparison to the signal of stable isotope labelled internal standard of this peptide that was added in a known amount to the sample [15, 16]. To achieve this goal, we developed the MRM assays for four peptides: K9[Poy]K14[Ac], K9[Poy]K14[Ac]-heavy, K9[Poy]K14[Poy] and K9[Poy]K14[Poy]-heavy. The MRM assays for quantification of the studied histone H3 peptides and their SIL standards were optimized manually by directly infusing to the QTRAP 6500+ mass spectrometer the AQUA QuantPro (> 97% purity and > 99% isotope enrichment from Thermo Fisher Scientific) standard solutions of the synthetic peptides of the above-mentioned sequences. The m/z of +2-charged precursor ions were selected from the enhanced MS (EMS) scan, while three of the +1 charged product ions m/z were selected based on their intensity observed during the enhanced product ion (EPI) MS scan performed for each of the analyzed peptides. After checking for specificity of the MRM transitions for K9[Poy]K14[Poy]APR and K9[Poy]K14[1Ac]APR peptides, we chose three of the best performing peptides for the final LC–MS/MS method, as indicated in CPTC Assay Development Guidelines ver.1.0 [27]. The details of the peptides analyzed in this study and their MRM transitions are presented in Table 1. The MRM dataset from the presented study was submitted to PASSEL repository (PeptideAtlas SRM Experiment Library, accession number PASS03789) [28] and is available under the accession number PASS03789 (https://db.systemsbiology.net/sbeams/cgi/PeptideAtlas/PASS_View?identifier=PASS03789) The MS parameters were individually optimized for each MRM transition, e.g., de-clustering potential (DP), collision energy (CE), entrance potential (EP), and cell exit potential (CXP); these are listed in Additional file 1: Table S1. Manually optimized MRM transitions list was imported to Skyline-daily ver. 21.1.1.327 along with the MS files from LC–MS/MS runs of the samples to perform peak extraction and data processing [29, 30]. Human histone H3 (UniProt accession number P68431) sequence in canonical FASTA format was downloaded from UniProt [31] along with the manually optimized MRM transitions to create targets list in Skyline. The following Skyline Peptide Settings were applied: digestion with trypsin [KR|P] with a maximum of 5 missed cleavages; peptides with 8 to 10 amino acids (AA) and 5 N-terminal AA excluded; structural modifications: cysteine carbamidomethylation, propionylation of lysine and N-term, lysine acetylation with up to 4 variable modifications, 1 maximum loses, heavy isotope label 13C615N4 at C-terminal arginine and light internal standard (for reverse calibration curve creation) or heavy (for separate donor’s samples processing). Skyline Transition Settings included: monoisotopic precursor and product masses; SCIEX collision energy; for filtering the peptides: +2 charged precursor ions, +1 product ions, from ion 1 to last ion were chosen with N-terminal proline as special ions. A range of 300 m/z to 1500 m/z with a method match tolerance of 0.25 m/z was chosen for the instrument. Peptides’ peak areas were calculated in Skyline software as a sum of peak areas of MRM transitions of their corresponding ions. Light to heavy peak area ratios for each of the modified and unmodified peptide and the known concentration of the added heavy peptide standards (K9[Poy]K14[Ac]-heavy and K9[Poy]K14[Poy]-heavy) were used by the software for calculation of the absolute quantities (fmol/µL) of the endogenous K9[Poy]K14[Ac] and K9[Poy]K14[Poy] peptides in the donors’ CIC, CIM, and MIM samples (Additional file 1: Table S4).

H3 K14Ac stoichiometry calculation

The absolute abundances of native modified (K9[Poy]K14[Ac]) and unmodified (K9[Poy]K14[Poy]) peptides determined in Skyline (Additional file 1: Table S4) were then applied for calculation of the H3 K14Ac stoichiometry in CIC, CIM, and MIM donors’ samples (Additional file 1: Table S6) using the following equation proposed by Prus et al. [17]:

$$Stoichiometry \left(\%\right)=\frac^}^+Abs}^} \times 100$$

where, Absnative_mod is absolute abundance of native modified peptide, Absnative_unmod is absolute abundance of native unmodified peptide. The calculated H3K14Ac stoichiometries for three technical replicates of each of analyzed conditions (CIC, CIM, MIM) for each donor were averaged to obtain average H3K14Ac stoichiometries in CIC, CIM, and MIM samples for donors. Additionally, the CIC, CIM, and MIM H3K14Ac stoichiometries for all six donors were also averaged, to obtain average stoichiometries for each of the analyzed conditions in all donors’ samples together. The SDs and % coefficient of variation (CVs) were calculated as well (Additional file 1: Table S6). These average H3K14Ac stoichiometries, for both, all donors together and each one separately, for CIC, CIM, and MIM conditions, are presented in a bar chart created using features available in Excel software (Microsoft Inc.) (Fig. 3).

The H3 K14Ac stoichiometry between CIC and CIM, CIC, and MIM, and CIM and MIM conditions were compared using a two-sample unequal variance T-test. If the p value was < 0.05, the differences were considered statistically significant.

Validation of the LC–MS/MS method in the MRM mode

Design of the validation procedure of the quantitative proteomic LC–MS/MS method in the MRM mode was based on the Assay Development Guidelines CPTAC Assay Development Working Group Version 1.0 [27] of the CPTAC of National Institutes of Health, National Cancer Institute. We used Skyline software for data processing and analysis as proposed there. Additionally, we created for our assay for research use; so to make it fit-for-purpose, we also applied recommendations of Carr et al. [32]. According to [27], we prepared the dilutions of the standard peptides in the matrix of interest for creating the multipoint calibration curve of seven different concentrations spanning from 0.5 fmol/uL to 2500 fmol/uL, QC at 12.5 fmol/uL and a matrix blank. The matrix blank was injected in nine replicates; the other concentrations were injected in five replicates each. The matrix for calibration curve was part of the experiment and the matrix of the target samples was the same: it constituted of the digested human histone extract propionylated pre- and post-digestion (from different donors, other than the target samples for analysis). We used the same protocol for sample preparation for both, the matrix for calibration curve creation and the target CIC, CIM, and MIM samples, with the exception that the MDM for the matrix samples were not HIV-infected and Meth exposed. The analytes quantified in this study—human histone H3 peptides K9[Poy]K14[Ac] and K9[Poy]K14[Poy] were likely to be occurring in the matrix in the detectable range. For that reason, we used a reverse calibration curve approach, where the matrix is spiked with SIL standards at variable concentrations and the native analyte—as an internal standard—at the same concentration.

Features available in Skyline software were used to prepare the reversed calibration curves for H3 K9[Poy]K14[Poy]-heavy and K9[Poy]K14[Ac]-heavy peptides and calculate figures of merit: lower limit of detection (LLOD) and lower limit of quantification (LLOQ) to determine sensitivity of the assay. The reversed calibration curves were created by plotting nominal analyte concentration (fmol/µL) versus heavy:light peak area ratio for each of the heavy peptides studied (Additional file 2: Figure S1 for K9[Poy]K14[Ac]-heavy and 2S for K9[Poy]K14[Poy]-heavy). The Skyline Peptide Settings and Transitions Settings applied for were the same as described in the “LC–MS/MS analyses in the MRM mode” paragraph with the exceptions listed below. We applied linear in Log space regression fit (none weighing) to create those reversed calibration curves. The max LLOQ bias (relative error) and LLOQ CV were set to 15% and qualitative ion ratio threshold set to 30%. The LOD was calculated from the blanks injected before the calibration curve as the average plus 3 times the standard deviation (SD) of the signal of the blank (3 × SD of the blank). The LLOQ was determined as “the lowest concentration of peptide at which the imprecision of the assay (expressed as the CV) is < 20% “ [27]. The linearity of the calibration curves was determined using linear regression and calculation of the Pearson’s determination coefficient R2 [26]. The accuracy and precision were also calculated, as a ratio of back-calculated concentration to the actual analyte concentration, and %CV of the replicate measurements of the same standard concentration, respectively. The calibration points had to be within 85–115% of the theoretical concentration or 80–120% at the LLOQ. The upper limit of quantification (ULOQ) was determined as the highest concentration measured whereas, a linear response between LLOQ and ULOQ was defined as a linear range [27]. The selectivity of the method was evaluated by visual examining the potential presence of the interferences in the matrix blank samples in the retention time region of the analytes (heavy peptides). To ensure that there is no interference on the analytes detection by other substances present in the samples, the specificity of the assay was assessed. At first, already on the MRM transitions development stage of the study, we used the MS Product feature of ProteinProspector software [33] to inspect whether the MRM transitions are unique for the peptides of interest and do not overlap with MRM transitions of the isobaric peptide K9[Ac]K14[Poy] that might be also present in the investigated samples. Then, the MRM transitions’ ratios for each of the analyzed peptides were determined to evaluate the specificity of the assay. The peak areas for each transition were normalized to total peak area of all transitions for analyte (Percentage peak areas) and compared between each other to obtain MRM transitions ratios. The mean, SD and %CV for each MRM transition ratio were also calculated for both the standard dilutions above the LLOQ and donors’ samples. The MRM transition ratios for all samples of concentrations above LLOQ should be within 30% from the mean.

The document grid option in Skyline was used to present the results of the LC–MS/MS method validation on peptide and MRM transitions levels, which are presented in the Additional file 1: Table S2 and S3.

Heavy to light peak area ratios for each of the modified and unmodified peptide and the concentration of the added light peptide standards (K9[Poy]K14[Ac] and K9[Poy]K14[Poy]) were used by the software for calculation of the absolute quantities (fmol/µL) of the K9[Poy]K14[Ac]-heavy and K9[Poy]K14[Poy]-heavy peptides in the series of the SIL standards dilutions used for reversed calibration curve preparation.