Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 9 (9), 8206-8222
eCollection

Distinct Histone Modifications Denote Early Stress-Induced Drug Tolerance in Cancer

Affiliations

Distinct Histone Modifications Denote Early Stress-Induced Drug Tolerance in Cancer

Abdullah Al Emran et al. Oncotarget.

Abstract

Besides somatic mutations or drug efflux, epigenetic reprogramming can lead to acquired drug resistance. We recently have identified early stress-induced multi-drug tolerant cancer cells termed induced drug-tolerant cells (IDTCs). Here, IDTCs were generated using different types of cancer cell lines; melanoma, lung, breast and colon cancer. A common loss of the H3K4me3 and H3K27me3 and gain of H3K9me3 mark was observed as a significant response to drug exposure or nutrient starvation in IDTCs. These epigenetic changes were reversible upon drug holidays. Microarray, qRT-PCR and protein expression data confirmed the up-regulation of histone methyltransferases (SETDB1 and SETDB2) which contribute to the accumulation of H3K9me3 concomitantly in the different cancer types. Genome-wide studies suggest that transcriptional repression of genes is due to concordant loss of H3K4me3 and regional increment of H3K9me3. Conversely, genome-wide CpG site-specific DNA methylation showed no common changes at the IDTC state. This suggests that distinct histone methylation patterns rather than DNA methylation are driving the transition from parental to IDTCs. In addition, silencing of SETDB1/2 reversed multi drug tolerance. Alterations of histone marks in early multi-drug tolerance with an increment in H3K9me3 and loss of H3K4me3/H3K27me3 is neither exclusive for any particular stress response nor cancer type specific but rather a generic response.

Keywords: DNA methylation; acquired drug resistance; epigenetic reprogramming; histone modification; stress-induced resistance.

Conflict of interest statement

CONFLICTS OF INTEREST Gordon B. Mills serves as a consultant for AstraZeneca, Blend Therapeutics, Critical Outcome Technologies Inc., HanAl Bio Korea, Illumina, Nuevolution, Pfizer, Provista Diagnostics, Roche, Signal Chem Lifesciences, Symphogen, Tau Therapeutics; owns stock in Catena Pharmaceuticals, PTV Healthcare Capital, Spindle Top Capital; and has received research funding from Adelson Medical Research Foundation, AstraZeneca, Critical Outcome Technologies Inc., GSK, and Illumina.

Figures

Figure 1
Figure 1. A common stress-induced transition of cancer cells into induced drug-tolerant cells (IDTCs)
a. WM1366, SKBR3 or A549 cells were exposed to either dimethyl sulphoxide (DMSO), docetaxel (DC; 5nM, 10nM, 20nM; WM1366, SKBR3) or doxorubicin (Dx; 500nM, 1µM and 2.5µM; A549) for a period of 12 to15 days. Experiments were performed in biological triplicate. Surviving cells were stained with crystal violet (well images), then dissolved in 10% acetic acid and measured at 570nm to provide the relative cell number of viable cells. Crystal violet staining at day one after seeding was used as loading control compared with 12-15 days’ time point. Error bars represent the standard deviation from the mean. b. In a separate experiment, 5nM docetaxel and 500nM doxorubicin IDTCs were further exposed to higher concentrations of doxorubicin at 2.5µM, docetaxel at 30nM, cisplatin at 80µM and sorafenib at 5µM in biological triplicate. Relative survival of IDTCs compared to parental cells was measured by crystal violet staining after 72hrs. Statistical analysis was performed by unpaired t-test and P-value is represented as (*) where, ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05. c. Morphological changes during the transition of parental cells to IDTCs in WM164 and A549 cells (10x magnification). d. Expression of CD271 in WM1366 and A549 IDTCs compared to untreated control by IF. Representative CD271 (red) and Hoechst staining (blue) overlay images are shown. IgG isotype control is shown in Supplementary Figure 1e. (10x Magnification and bar is 100µm long) e. Protein lysates from IDTCs and the untreated control were subjected to immunoblotting for expression levels of the indicated antibodies. f. Indicated cell lines were maintained in low glucose (1mg/ml) containing media for 12-15 days. Cell lysates from low glucose IDTCs and normal glucose controls were probed for the indicated proteins. All western blot images were quantified by ImageJ software. Values were normalized by subtracting from loading control.
Figure 2
Figure 2. Stress and drug holidays dynamically modulate histone modifications
a. WM1366, A549 IDTCs and their parental cells were tested for H3K4me3, H3K9me3 and H3K27me3 antibodies by IF. Representative antibodies staining and Hoechst (nuclear) staining were taken separately and overlaid picture are shown. b. A549 and SKBR3 cells were cultured with 1mg/ml glucose containing media for at least 12 days. Media was replenished after every three days. IF was performed for the indicated Abs to compare to parent cells supplemented with 5mg/ml glucose media. c. WM1366, SKBR3 and A549 IDTCs were allowed a drug holiday for ten days. During this period of time, no drug was provided and media changed every 72 hrs. Histone modifications were analysed by IF with the indicated Abs compared to corresponding parental cells. (All of the IF image were taken at 10x Magnification and bar is 100µm long) d. After 10 days of drug holidays WM1366, A549 and HT29 IDTCs were exposed to toxic concentrations of doxorubicin (2.5µM), docetaxel (30nM) for WM1366, A549 and dabrafenib (500nM) and docetaxel (30nM) for HT29. Cell survival was analysed by MTT assay. All experiments were done in triplicate. Statistical analysis was performed by the one-way ANOVA test and P-value is represented as (*) where, ****P < 0.0001.
Figure 3
Figure 3. Differential expression of genes and genome wide DNA methylation in IDTCs compared to parental cells
a. Hierarchical clustering of differential gene expression. Four cancer cell lines were analysed by the Australian Genome Research Facility (AGRF) for genome-wide differential expression using Illumina expression arrays. An unbiased hierarchical clustering was developed using the Mann-Whitney unpaired test with a fold change cut off FC≥ 1.5 and P≤0.05 as shown in the figure. b. A volcano plot was generated with a similar threshold as mentioned above depicting the overall 293 up and down-regulated genes. c. Pathways enriched for each of the IDTCs compared to parent cells are shown by using cluster Profiler7, with a p-value and a q-value cut off of 0.1 and 0.05 respectively. Genome-wide DNA methylation analyses reveal cell type-specific changes at the IDTC state. d. DNA methylation levels of IDTCs compared to untreated control. Bars represent the mean DNA methylation level for each cell line, with (IDTC) and without (CTRL) treatment. Error bars = S.E.M. e. Venn’s diagrams representing commonly differentially methylated (hypo- or hyper-methylated) CpG sites in IDTCs compared to control. f. Hierarchical clustering for cell-type using differentially methylated CpG sites. Differentially methylated CpG sites between the groups were identified using the Rank Products test [54]. After 100 permutations, CpG sites showing an adjusted p-value<0.05 (q-value) were considered significant. Scale bar for beta values: Green = 0 Black = 0.5 and Red = 1.
Figure 4
Figure 4. Genome-wide re-distribution of histone modifications in IDTCs
a. Genome-wide levels of histone modifications are represented by the mean number of segments with ChIP-seq peaks 20kb around 100,000 randomly selected genome segments (upper panel). The association with transcriptomic changes is represented as the distribution of each histone modification around the transcription start sites (TSSs) of 1,423 genes down-regulated (fold change <-2) in IDTC cells (lower panels). b. Re-distribution of histone modifications in IDTCs was evaluated on 1,245 polycomb repressive domains (PRDs) associated with IDTC down-regulated genes. c. A significant overlap was observed between down-regulated genes (n = 1,423) and genes associated with H3K9me3 (n = 4,969)(n = 487; exact hypergeometric probability; P = 5.38x 10-22) as well as for regions between up-regulated genes (fold change >2; n = 1,087) and genes associated with H3K4me3 (n = 1,487)(n = 138; exact hypergeometric probability; P = 3.15x10-12) and between downregulated genes (n = 1423) and genes associated with H3K27me3 (n = 370; exact hypergeometric probability P = 0.03). d. Representative genomic view of the inversely correlated H3K4me3 and H3K9me3 modifications in IDTCs. The chromatin states were identified by calculating the Epilogos score (https://epilogos.altiusinstitute.org/) from the integration of the Core 15-state model using 111 reference human epigenomes generated by the Roadmap Epigenomics Project [23] plus 16 human epigenomes generated by the ENCODE Project. Two PRDs showing enrichment in H3K9me3 and depletion of H3K4me3 leading to a regional (2.6 Mb window) transcriptional down-regulation in IDTCs.
Figure 5
Figure 5. SETDB1 and SETDB2 regulate H3K9me3 in IDTCs and knockdown restores drug sensitivity
a. RNA was isolated from different IDTCs and their parental cells. q-RTPCR was performed for the different histone modifiers of H3K9me3. Actin RNA was used as an internal control and relative gene expression levels were calculated as delta CT. Bars represent the mean of three biological replicates. Statistical analysis was performed by two-way ANOVA test and P-value is represented as (*) where, ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05. b. Proteins were isolated from different IDTCs and their respective parental cells. Immunoblots were performed to investigate the expression of SETDB1, SETDB2, and H3K9me3. c. WM164, WM1366, and HT29 cells were exposed to 100nM dabrafenib, 10nM docetaxel and 100nM dabrafenib respectively for two weeks. Proteins were isolated from IDTCs and untreated control and immunoblotted for SETDB1, SETDB2 and H3K9me3. All western blot images were quantified by ImageJ software. Values were normalized by subtracting from loading control. d. Tumour tissue was formalin fixed and tissue slides were stained for H3K9me3. Images were taken in a Vectra III Spectral scanner. e. RNA-sequencing data of paired pre- and post-treatment tumour biopsies derived from melanoma patients from GEO [55] were normalized, background-corrected and analysed using the R package “lumi”. Statistical analyses were performed by paired t-test where, **P < 0.01, *P < 0.05. Each paired match-patient expression value for an individual gene is shown by different colours. Numeric values indicate the patient number as provided in Supplementary Table 2. BT- Before treatment, AT- After treatment. f. Lentiviral transduction was performed according to manufacturer protocol (Sigma Aldrich, MA, USA). Protein lysates of the transduced cells were subjected to immunoblotting for SETDB1, SETDB2, H3K9me3 and Actin. g. WM164 shcontrol, shSETDB1, and shSETDB2 transduced cells were exposed for 16 days to 50nM dabrafenib and stained for H3K4me3, H3K9me3, and H3K27me3 by immunofluorescence. h. Same as in (g) but in addition challenged with toxic concentrations of docetaxel (30nM) and doxorubicin (2.5µM). Cell survival was analysed by MTT assay. Statistical analysis was performed by two-way ANOVA test and P-value is represented as (*) where, ****P < 0.0001. All western blot images were quantified by ImageJ software. Values were normalized by subtracting from loading control.
Figure 6
Figure 6
Model (a) Stress induce cancer cells to undergo a dynamic histone reprogramming with the increase of repressive histone mark H3K9me3 and decrease of active mark H3K4me3 and repressive mark H3K27me3. These distinct histone modifications help cancer cells to maintain a transcriptionally repressed slow cycling state. Drug holiday reversed the histone modifications similar to that of parental cells which induce sensitivity. b. Upon drug treatment silenced SETDB1/2 cancer cells reprogram their histone modifications which are sensitive to other drugs.

Similar articles

See all similar articles

Cited by 10 PubMed Central articles

See all "Cited by" articles

References

    1. Gottesman MM, Fojo T, Bates SE. Multidrug resistance in cancer: role of ATP-dependent transporters. Nat Rev Cancer. 2002;2:48–58. doi: 10.1038/nrc706. - DOI - PubMed
    1. Fojo T, Bates S. Strategies for reversing drug resistance. Oncogene. 2003;22:7512–23. doi: 10.1038/sj.onc.1206951. - DOI - PubMed
    1. Longley DB, Johnston PG. Molecular mechanisms of drug resistance. J Pathol. 2005;205:275–92. doi: 10.1002/path.1706. - DOI - PubMed
    1. Talpaz M, Silver RT, Druker BJ, Goldman JM, Gambacorti-Passerini C, Guilhot F, Schiffer CA, Fischer T, Deininger MW, Lennard AL, Hochhaus A, Ottmann OG, Gratwohl A, et al. Imatinib induces durable hematologic and cytogenetic responses in patients with accelerated phase chronic myeloid leukemia: results of a phase 2 study. Blood. 2002;99:1928–37. - PubMed
    1. Berrieman HK, Lind MJ, Cawkwell L. Do beta-tubulin mutations have a role in resistance to chemotherapy? Lancet Oncol. 2004;5:158–64. doi: 10.1016/s1470-2045(04)01411-1. - DOI - PubMed

LinkOut - more resources

Feedback