Figure 1.
Direct chromatin PCR for CDKN2A gene.
CDKN2A primer locations for DC-PCR are shown on a map of its 5′ region. An arrow indicates the transcription start site for transcript NM_000077.4 (A). Analysis of CDKN2A transcription by RT-PCR (Right upper panel) and chromatin structure by DC-PCR (Left panel) in control (vehicle treated) KMS-12-PE cells, after treatment with 1 µM decitabine (DAC) or 100 nM trichostatin (TSA) and in normal fibroblasts (B). KMS-12-PE multiple myeloma cells do not express CDKN2A (Right upper panel) and when untreated cells are subjected to DC-PCR, most CDKN2A primers do not yield products (Left panel). After treatment with DAC, CDKN2A mRNA expression is re-activated (Right upper panel) and DC-PCR yields multiple amplicons (Left panel). Short-term treatment with the deacetylase inhibitor TSA, which did not affect viability of KMS-12-PE cells (Right lower panel) also made chromatin accessible for DC-PCR amplification (Left panel) but did not induce CDKN2A mRNA expression (Right upper panel). Exposure of KMS-12-PE cells to high salt concentrations (600 mM NaCl for 20 min) allowed DC-PCR amplification of almost all CDKN2A sites. DC-PCR also detects the open chromatin structure of CDKN2A expressing normal human dermal fibroblasts (NHDF) (Left and Upper right panel, respectively). Correlation of DC-PCR results with Chromatin Immunoprecipitation (C). Chromatin immunoprecipitation assay of the CDKN2A regulatory region with antibodies against acetylated lysine 9 histone 3 (H3K9Ac) in control, vehicle treated cells and cells treated with DAC or TSA as under (B). Purified DNA was analyzed by real-time PCR using primers for the CDKN2A regulatory region shown on the map in Figure 1A. Error bars represent standard deviations of the mean of duplicates. All results are representative of at least three independent experiments.
Figure 2.
1 gene. Primer locations for DC-PCR are shown on a map of the PU.1 5′ region (A). An arrow indicates the transcription start site for transcript NM_001080547.1. Analysis of PU.1 transcription by RT-PCR (Right panel) and chromatin structure by DC-PCR (Left panel) in control (vehicle treated) KMS-12-PE cells, after treatment with DAC (as described under Figure 1) and in PU.1 positive cells (B). The RPMI-8226 cell line expresses PU.1 while KMS-12-PE cells have PU.1 silenced but it can be re-activated by DAC treatment (Right panel). DC-PCR correlates with RT-PCR results: In expressing cells, whether at baseline or after treatment with DAC, most 5′ PU.1 sequences can be amplified while untreated KMS-12-PE cells yielded no products (Left panel). Brief TSA treatment of KMS-12-PE cells as under Figure 1, also opened PU.1 chromatin (Left panel). Displayed results are representative of three independent experiments.
Figure 3.
Direct chromatin PCR for CD34 gene: detection of epigenetic changes during cellular differentiation in vitro.
Primer locations for DC-PCR are shown on a map of the CD34 5′ region (A). An arrow indicates the transcription start site for transcript NM_001025109.1. Flow cytometry (Left panel) and DC-PCR analysis (Right panel) of cord blood CD34+ cells as they undergo cytokine mediated differentiation in vitro (B). One day after CD34 magnetic bead purification about 65% of cells expressed CD34 as determined by flow cytometry (Left panel). At day 15 this fraction had decreased to 37.1%, while cells remained healthy with less than 10% showing evidence for loss of membrane integrity as assessed by 7-AAD co-staining (Left panel). CD34 DC-PCR (Right panel) identified decreased accessibility of sequences within about 500 bp 5′ of the CD34 transcription start site on day 15 compared to day 1 and of 11 tested sequences in this region (primers 9–19) only 3 yielded (faint) DC-PCR products in CD34 negative (Left panel) KMS-12-PE cells (Right panel). Results are representative of three independent experiments.
Figure 4.
Direct chromatin PCR for IRF4 gene.
IRF4 Expression and Inhibition in Myeloma Cells. Myeloma cells express high levels of IRF4, while leukemia cells or normal fibroblasts (NHDF) have IRF4 silenced. Treatment with 10-E-09 suppresses IRF4 mRNA and protein expression (upper right panel) (A). Primer locations for DC-PCR are shown on a map of IRF4 5′ region (B). An arrow indicates the transcription start site for transcript NM_001195286.1. After pre-heating of cells to 80°C for 60 seconds in isotonic media, sites 5′ to the IRF4 transcription start site became more accessible to DC-PCR amplification in IRF4 expressing KMS-12-PE cells than in IRF4 negative K562 cells (Left panel). The sites closest to the transcription start site only opened after treatment with the IRF4 inhibitor 10-E-09 at 10 µM for 6 h (Left panel). After pre-heating of cells in isotonic media to 95°C for 60 seconds multiple IRF4 chromatin sites could be amplified in all cells (Right panel). Displacement of RNA polymerase II and NFkB from the IRF4 promoter and changes in the level of H3K9Ac after treatment of KMS-12-PE myeloma cells with the IRF4 inhibitor 10-E-09 investigated by ChIP (C). The chromatin immunoprecipitation assay used antibodies against RNA polymerase II, NFkB and acetylated lysine 9 histone 3 (H3K9Ac) in vehicle (DMSO) treated cells and in cells treated with 10 M of 10-E-09 for 6 hours. Purified DNA was analyzed by real-time PCR using primers #5, 6, 7 and 8 of the IRF4 promoter. Error bars represent standard error of the means of duplicates. All results are representative of at least three independent experiments.
Figure 5.
Direct chromatin PCR for FOSB gene.
FOSB mRNA Expression after 10-E-09 In KMS-12-PE detected and analyzed on HT-12 Array™ (A). Cells were treated with 10-E-09 at 10 µM for indicated times and arrays were run in duplicates. Displayed are mean fluorescence units (Left panel) and standard deviations of duplicates (Right panel). Primer locations for DC-PCR are shown on a map of the FOSB 5′ region (B). An arrow indicates transcription start site for transcript NM_001114171.1. DC-PCR performed for FOSB on KMS-12-PE cells (C): untreated and treated with 10-E-09. Active chromatin of FOSB can be detected by DC-PCR 3 h after treatment with 10-E-09. Most of the 11 tested sequences allowed primer access whereas chromatin of untreated cells remained almost completely closed. Genomic DNA used as template demonstrated amplification of all FOSB sites. Displayed DC-PCR results are representative of three independent experiments.
Figure 6.
Optimization of DC-PCR for drug screening.
Effect of cytoplasmic enzymes on DC-PCR (A). KMS-12-PE cells were diluted at 1∶25 in hypotonic PCR buffer with or without added protease inhibitors and incubated for 20 min before PCR was run to simulate the time required for pipetting during a small molecule screen. Inclusion of protease inhibitors in the PCR mastermix decreased the number of amplifiable sites in DAC treated and untreated KMS-12-PE cells suggesting that digestion of DNA binding proteins can occur in PCR buffer although epigenetic differences between samples are maintained in the studied example. Detection of DC-PCR product via UV transillumination of PCR plates and sensitivity compared to gel electrophoresis (B). Ethidium bromide was added to the PCR mastermix before serially diluted cells were submitted to DC-PCR: untreated, DAC treated or fibroblasts. After completion of thermal cycling samples were first analyzed under UV light (Left panel), then via agarose-gel electrophoresis (Right panel). Using gel-based detection, the level of detection for epigenetic differences of CDKN2A was lower, down to about 1 cell for site #6 compared to about 300 cells with direct UV transillumination. The primer numbers correspond to sequences on the map described in Figure 1A. Three independent experiments yielded the same detection levels. Correlation between direct UV transillumination and gel electrophoresis was seen in all experiments (hundreds).
Figure 7.
Principles of DC-PCR drug screens.
Cartoon outlining the general procedure of DC-PCR and possible results that depend on the interrogated region and utilized cells (A). General outline of a CS-PCR screen for molecules that make the interrogated site more soluble (B). Only a fraction of cells (1–10 µl) is required to run the screen. Remaining cells can be used for confirmatory tests or secondary screens. If heat is used and confirmatory tests on hits are planned, an additional pipetting step before heating is required.
Figure 8.
Direct Amplification of open chromatin with phi29 polymerase: Translation of DC-PCR to genome-wide chromatin analysis.
CDKN2A PCR on KMS-12-PE supernatant after direct genome-wide phi29 amplification (A). Untreated KMS-12-PE cells or cells treated with DAC (1 µM for 3 d) were placed directly in phi29 reaction mix for 4.5 hrs at 30°C. After that, the reaction was centrifuged to pellet cell carcasses. Only the upper 50% of the supernatant was used for subsequent amplification of CDKN2A sites by conventional PCR. Supernatant from DAC treated cells yielded more CDKN2A products than vehicle treated cells. When DAC treated cells were incubated in reaction buffer without Phi29 polymerase no CDKN2A products could be amplified from supernatants by taq polymerase during conventional PCR. FOSB PCR on KMS-12-PE supernatants obtained from untreated cells or cells treated with 10-E-09 (10 µM for 6 h) after direct genome-wide phi29 amplification (B). The same procedure as above yielded FOSB PCR products from supernatants of 10-E-09 treated KMS-12-PE cells amplified by phi29 while vehicle treated cells or supernatants from 10-E-09 treated cells incubated in reaction buffer without phi29 yielded no FOSB regulatory region amplicons. Results are representative of three independent experiments.
Figure 9.
Inhibitors of IRF4 expression identified by DC-PCR.
Reduction of IRF4 protein expression by hits in myeloma cells compared to 10-E-09. Immuno-blot for IRF4 was performed after treatment with vehicle (DMSO) or with drug-candidate inhibitors at 3 µM for 24 hours (A). β-Actin was used as loading control (lower panel). Similar results were obtained in three independent experiments. Confirmatory gel electrophoresis of DC-PCR after treatment of KMS-12-PE cells with drug candidates (IRF4 inhibitors) at 3 µM for 6 hours (B). The solubilization of sites 7 and 8 indicates that IRF4 transcriptional complexes are displaced by hits suggesting at least a common final path of all three IRF4 expression inhibitors. Results were confirmed in at least three independent experiments per drug candidate. Treatment with hits suppresses growth and survival of IRF4 expressing multiple myeloma cells (Left panels) more than IRF4 negative AML cell lines (Right panels) (C). Cells were treated with IRF4 inhibitors at indicated concentrations for five days before coulter counter with trypan blue exclusion was used to assess total and viable cell counts. Error bars represent standard deviations of least two independently treated wells. Similar results were obtained in at least one independent repeat per cell line.