Global histone modification fingerprinting in human cells using epigenetic reverse phase protein array

Abstract

The balance between acetylation and deacetylation of histone proteins plays a critical role in the regulation of genomic functions. Aberrations in global levels of histone modifications are linked to carcinogenesis and are currently the focus of intense scrutiny and translational research investments to develop new therapies, which can modify complex disease pathophysiology through epigenetic control. However, despite significant progress in our understanding of the molecular mechanisms of epigenetic machinery in various genomic contexts and cell types, the links between epigenetic modifications and cellular phenotypes are far from being clear. For example, enzymes controlling histone modifications utilize key cellular metabolites associated with intra- and extracellular feedback loops, adding a further layer of complexity to this process. Meanwhile, it has become increasingly evident that new assay technologies which provide robust and precise measurement of global histone modifications are required, for at least two pressing reasons: firstly, many approved drugs are known to influence histone modifications and new cancer therapies are increasingly being developed towards targeting histone deacetylases (HDACs) and other epigenetic readers and writers. Therefore, robust assays for fingerprinting the global effects of such drugs on preclinical cell, organoid and in vivo models is required; and secondly, robust histone-fingerprinting assays applicable to patient samples may afford the development of next-generation diagnostic and prognostic tools. In our study, we have used a panel of monoclonal antibodies to determine the relative changes in the global abundance of post-translational modifications on histones purified from cancer cell lines treated with HDAC inhibitors using a novel technique, called epigenetic reverse phase protein array. We observed a robust increase in acetylation levels within 2-24 h after inhibition of HDACs in different cancer cell lines. Moreover, when these cells were treated with N-acetylated amino acids in addition to HDACs, we detected a further increase in histone acetylation, demonstrating that these molecules could be utilized as donors of the acetyl moiety for protein acetylation. Consequently, this study not only offers a novel assay for diagnostics and drug screening but also warrants further research of the novel class of inexpensive, non-toxic natural compounds that could potentiate the effects of HDAC inhibitors and is therefore of interest for cancer therapeutics.

Figure 1

The effect of HDAC inhibitors on de-repression of GFP transgenes (a and b) HeLa-GFP cells were treated with 10 μM of a panel of different HDAC inhibitors for 24 h (a). The percentage of GFP-positive cells (green) was then calculated as a percentage of total cells (DAPI) using fluorescent microscopy. Cells treated with vehicle alone (DMSO) served as a negative control, whereas HeLa cells transfected with an empty vector acted as null controls. TSA (10 μM), which was a different batch to the one in the panel, was used as a positive control. The ‘top hits’ from the screen in (a) (CAY10603, HC toxin and TSA) were then used to treat HeLa-GFP cells for 24 h at a concentration of 500 nM and the effects on GFP silencing were examined by fluorescence microscopy: five biological replicates with three technical replicates each time. (b) The same three chemicals were then used to treat colorectal cancer cells HCT116 for 2 h at a concentration of 500 nM and the effects on acetylation of isolated histones were examined by RPPA: three biological replicates with three technical replicates each time.

Figure 2

Induction of histone acetylation (a and b): Each sample, normalized for protein levels, was loaded onto the slide using twofold serial dilutions in triplicate. Panel a shows a typical image obtained by RPPA using a pan-acetylated lysine antibody. The amount of antibody staining was normalized with Fast Green and the staining intensity (Relative Fluorescent Units) of each sample was calculated as a percentage of cells treated with DMSO alone±S.E.M. (b, chart). Each value represents the mean of three individual experiments, conducted and spotted in triplicate. Ten micrograms of the isolated histone protein from each treatment was used to confirm the RPPA findings by western blot analysis using a pan-acetylated lysine antibody (b, below the chart); three biological replicates with three technical replicates each time.

Figure 3

Metabolites potentiate the effects of HC toxin on histone acetylation and methylation. HCT116 cells were exposed to 100 nM HC toxin for 24 h in the presence or absence of combined metabolites (500 nM ALCAR, 500 nM NAL and 500 nM NAM) and then RPPA was conducted on isolated histones. A typical image from RPPA for pan-acetylated lysine is shown (a). Each sample, normalized for protein levels, was loaded onto the slide using twofold serial dilutions in triplicate. The amount of antibody staining was normalized with Fast Green and the staining intensity (Relative Fluorescent Units) of each sample was calculated as a percentage of cells treated with DMSO alone±S.E.M.; three biological replicates with three technical replicates each time. (b) HC toxin and HC toxin+metabolites caused a significant increase in histone acetylation, compared with DMSO or metabolites alone (P<0.001). The RPPA findings were confirmed using western blot analysis of the isolated histones (b, lower panel); three biological replicates with three technical replicates each time. Acetylation (c–f) and methylation (g and h) at specific histone residues were also examined by RPPA. Relative staining intensity was calculated and expressed as a percentage of DMSO (open bars), metabolites alone (spotted bars), HC toxin (solid bars) and HC toxin+metabolites (hatched bars). Statistical significance was determined using two-tailed Student’s T-test: *P<0.05; **P<0.01; ***P<0.001. Three biological replicates with three technical replicates each time for each panel (c–g).