Pathology Tissue-quantitative Mass Spectrometry Analysis to Profile Histone Post-translational Modification Patterns in Patient Samples

Abstract

Histone post-translational modifications (hPTMs) generate a complex combinatorial code that has been implicated with various pathologies, including cancer. Dissecting such a code in physiological and diseased states may be exploited for epigenetic biomarker discovery, but hPTM analysis in clinical samples has been hindered by technical limitations. Here, we developed a method (PAThology tissue analysis of Histones by Mass Spectrometry - PAT-H-MS) that allows to perform a comprehensive, unbiased and quantitative MS-analysis of hPTM patterns on formalin-fixed paraffin-embedded (FFPE) samples. In pairwise comparisons, histone extracted from formalin-fixed paraffin-embedded tissues showed patterns similar to fresh frozen samples for 24 differentially modified peptides from histone H3. In addition, when coupled with a histone-focused version of the super-SILAC approach, this method allows the accurate quantification of modification changes among breast cancer patient samples. As an initial application of the PAThology tissue analysis of Histones by Mass Spectrometry method, we analyzed breast cancer samples, revealing significant changes in histone H3 methylation patterns among Luminal A-like and Triple Negative disease subtypes. These results pave the way for retrospective epigenetic studies that combine the power of MS-based hPTM analysis with the extensive clinical information associated with formalin-fixed paraffin-embedded archives.

Fig. 1.

Proteomic analysis of hPTMs from mouse FFPE tissues. A, Schematic representation of the procedures used to isolate histones from mouse frozen spleen cells or FFPE mouse spleen. Five microgram of core histones purified from frozen cells and 20 μg of protein extract obtained from four 10 μm-thick FFPE sections were separated by SDS-PAGE and Comassie-stained. Gel bands corresponding to histone H3 and H4 were then excised, subjected to an in-gel Arg-C like digestion and analyzed by LC/MS/MS. A similar procedure was used for mouse frozen liver cells or FFPE mouse liver. FA: formalin. B, List of H3 and H4 peptides identified from frozen or FFPE samples for mouse spleen and liver stored for few weeks or mouse spleen stored for 6 years. The spleen tissues stored for few weeks or 6 years derive from different mice.

Fig. 2.

Proteomic quantification of hPTMs from mouse FFPE tissues. A, Percent relative abundances (%RA) profiles for H3 and H4 peptides from frozen or FFPE samples for mouse spleen (top panel) or liver (middle panel) stored for few weeks or mouse spleen stored for 6 years (bottom panel). Error bars represent the standard error from two independent experiments (top panel) or duplicate measurements (middle and bottom panels). B, hPTM %RA correlation between frozen cells and FFPE samples obtained from the mouse and liver spleen samples shown in panel A. Pearson correlation coefficients (r) and p values are shown. The spleen tissues stored for few weeks or 6 years derive from different mice.

Fig. 3.

Quantification of hPTMs from human breast cancer FFPE tissues using a super-SILAC spike-in approach. A, Schematic representation of the super-SILAC spike in approach and the histone isolation procedures used to quantify histones from human fresh-frozen or FFPE breast cancer tissues, whose appearance on a Comassie-stained SDS-PAGE gel is shown. A mix of histones obtained from four heavy-labeled breast cancer cell lines is spiked into breast cancer primary samples prior to gel separation and used as an internal standard for MS-based hPTM analysis. B, Heatmap display of the log₂ of ratios obtained for the indicated hPTMs for frozen and FFPE breast cancer biopsies (averages from triplicate measurements). L/H relative abundances ratios obtained with the super-SILAC strategy (light channel: breast cancer biopsy, heavy channel: spike-in super-SILAC standard) normalized over the average value across the samples are shown. C, Correlation of ratios obtained in frozen and FFPE samples. Pearson correlation coefficients (r) and p values are shown.

Fig. 4.

Epigenetic profiling of breast cancer subtypes by PAT-H-MS. A, Heatmap display of the log₂ of ratios obtained for the indicated hPTMs for FFPE breast cancer biopsies belonging to four different subtypes (LuA: Luminal A-like, LuB: Luminal B-like, TN: Triple Negative, HER: HER-positive). L/H relative abundances ratios obtained with the super-SILAC strategy (light channel: breast cancer biopsy, heavy channel: spike-in super-SILAC standard) normalized over the average value across the samples are shown. B, Ratios obtained for the indicated peptides containing K27me3 or K9me3 in Luminal A or Triple negative breast cancer samples. C, D, FFPE extracts from the Luminal A and Triple Negative samples analyzed in A–B were probed by immunoblot with anti-H3K27me3, H3K9me3 and total H3 antibodies. D, The levels of H3K27me3 were quantified from the immunoblots shown in C and normalized to the amount of total histone H3. E, F, Extracts from additional frozen Luminal A and Triple Negative samples were probed by immunoblot and quantified as in C–D. Samples in B were compared by one-way ANOVA and Bonferroni's post-hoc test. Samples in D and F were compared by T test. Error bars represent standard error from 5 patient samples. *p < 0.05, **p < 0.01.