Chromatin Conformation Signatures Associated with Epigenetic Deregulation of the FIP1L1 and PDGFRA Genes

The human genome is organized into topological domains with discrete structural and regulatory units that influence gene expression epigenetically (Crutchley et al., 2010). In glioma patients with IDH1 (R132H) mutations, researchers detected aberrant methylation of transcriptional CTCF-enhancer regions between the FIPL1L1 and PDGFRA genes. Conformational changes to the chromosomal arrangement of these regions have been shown to upregulate PDGFRA expression (Flavahan et al; 2016). Flavahan et al also demonstrated the down regulation of cell growth and proliferation in the glioma cells with the addition of tyrosine kinase inhibitors that target the interaction.We explored this epigenetic interaction and its utility in the stratification of patients for treatment with tyrosine kinase inhibitors. Using a high-resolution chromosome-conformation capture or 3C analysis platform known as EpiSwitch^TMand quantitative PCR, we mapped, evaluated, and quantified the conformational juxtaposition between FIP1L1 and PDGFRA in glioma-cell lines with and without IDH mutations. Deregulation of PDGFRA by interstitial deletion at 4q12 and fusion to FIP1L1 associated with chronic esosinophic leukemias prompted our group to also investigate whether the same chromosome-conformation interactions are present in EOL-1 and other leukemic cell lines.

EpiSwitch^TM templates were produced using a modified process to purify the 3C DNA. Unlike previous protocols, only one set of primers is required. Chromosome conformation capture analysis with a probe, single step PCR and gel purification was used to identify and sequence 3C interactions in the AML cell lines EOL-1 and HL-60. qPCR templates, adjusted to 20 ng of 3C library DNA, were used with concentration-matched negative controls (ie, 3C libraries derived from adipose biopies and normal blood). A 3C interaction with MMP-1 was used as a internal control for the EpiSwitch^TM library. A dual label hydrolysis probe was used to detect the sequenced interaction. Five novel interactions were identified between FIP1L1 and PDGFRA by single step PCR and confirmed by sequencing. We quantified one interaction with a dual-labeled hydrolysis probe based qPCR assay to determine copy number in eight cell lines: BT-412 (anaplastic oligoastrocytoma), DBTRG-05MG (glioblastoma multiforme), U-3T3 (glioblastoma astrocytoma), U-87 (glioblastoma astrocytoma), GDM-1 (acute myelomonocytic leukemia), EOL-1 (acute myeloid leukemia, AML), HL-60 (AML) and KG-1 (AML).

Using EpiSwitch^TMand ISO standard MIQE-compliant qPCR design, five specific epigenetic interactions were detected in three types of cancerous cell lines including EOL-1 and HL-60. qPCR was used to detect one specific difference between normal blood (n = 4) and leukemic cell lines (n = 4). By qPCR the interaction was detected in two AML cell lines known to be highly sensitive to the tyrosine kinase inhibitor Glivec (EOL-1 and GDM-1), as well as HL-60 cells, but not in KG-1. The FIP1L1 PDGFRA 3C interaction was also found in DBTRG-05MG and U-373 at a similar copy number compared to the AML cell lines. A glioma patient biopsy tissue also tested positive for the conformational-chromosome interaction by qPCR compared with an adipose biopsy control. An interaction in the MMP-1 gene was used as an internal positive control in all samples tested.

Emerging epigenetic approaches, such as chromosome conformation signatures as detected by EpiSwitch^TM, are providing sensitive and accurate methods for improving patient treatment stratification. Detection of interactions between PDGFRA and FIP1L1 based on chromosomal conformational changes have the potential to improve upon the available methods for identifying patients who may benefit from treatment with PDGFRA-targeted tyrosine kinase inhibitors. Further research is needed to validate the method described using blood and biopsy samples in patients with glioma, AML and other myeloproliferative neoplasms associated with the FIP1L1-PDGFRA fusion gene.