Chromatin Architecture and Transcription Factor Occupancy of Erythrocyte Membrane Genes Studied by Chromatin Immunoprecipitation on Microarrays (ChIP-chip)

Erythrocyte membrane protein genes serve as excellent models of complex gene loci structure and function, as most encode multiple tissue-, cell-, developmental-, and stagespecific isoforms. Dynamic chromatin modifications participate in the regulatory control of many gene loci. We hypothesize that specific DNA sequences, transcription factors, and chromatin architecture (epigenetic modifications) regulate the tissue-specific expression of erythrocyte membrane protein genes. Advances in genomics technology have permitted rapid identification of DNA sequences bound by transcription factors and other DNAassociated proteins on a genome-wide scale. One technique available for mapping protein-DNA interactions in vivo couples chromatin immunoprecipitation to microarrays that contain regions of genomic DNA (ChIP-chip). We are using DNA obtained from chromatin immunoprecipitations performed with histone and erythroid transcription factor antibodies hybridized to genomic DNA microarray chips (ChIP-chip) to study the regulation of membrane protein genes in erythroid and nonerythroid cells. Chromatin immunoprecipitations (ChIP) were done in erythroid (K562) and non-erythroid (HeLa) cell lines using antibodies against H3 tri-methyl lysine 4 (H3K4me3, a marker of active chromatin) and the erythroid transcription factors GATA-1 and NF-E2. The chromatin resulting from these ChIPs was hybridized to a custom made NimbleGen high density human genomic DNA microarray (chip) focused on 15 genes critical to the erythrocyte membrane: ankyrin (ANK1), α-spectrin (SPTA1), β-spectrin (SPTB), band 3 (SLC4A1), β-adducin (ADD2), α-adducin (ADD1), γ-adducin (ADD3), ICAM-4, Erythroid Associated Membrane Protein (ERMAP), Protein 4.1 (EPB41), Protein 4.2 (EPB42), Dematin (ERPB49), β-Actin (ACTB), tropomodulin (TMOD1), and tropomyosin (TPM3). Probes for the chip were ~50bp in length with Tm ≥ 76°C, tiled every 65bp. From 50–100kb of flanking DNA was included on the chip for each locus. The Tamalpais peak calling algorithm using L1–L3 level of stringency (Genom Res16:595, 2006) was used to analyze the resulting data and identify regions of epigenetic modifications and transcription factor binding. Fourteen of 15 genes were enriched for H3K4me3 at promoter and transcriptional start sites (TSS) in K562 cells, with one gene, TMOD1, demonstrating a large peak of enrichment 5′ of the currently identified TSS, but not at the promoter. Two compact genes, β-actin and ICAM4, had H3K4me3 enrichment at the promoter and throughout gene. A total of 19 GATA-1 sites and 18 NF-E2 sites were identified. GATA-1 sites were found in 8 of 15 genes or in their flanking DNA. Three sites were in the 5′ flanking DNA, 1 site was at the promoter (~500bp from transcription start site, TSS), 12 sites were in introns, and 3 sites were in the 3′ flanking DNA. NF-E2 sites were found in 10 of 15 genes or their flanking DNA. 6 sites were in the 5′ flanking DNA, 1 site was at the promoter (~200bp from TSS), 8 sites were in introns, and 3 sites were in the 3′ flanking DNA. 18 of 19 GATA-1 sites (95%) and 13 of 18 NF-E2 sites (72%) were validated using qPCR-based quantitative ChIP. In K562 cells, 15 of 19 (79%) validated GATA-1 sites were associated with regions of chromatin enriched for H3K4me3, suggesting that ~a fifth of GATA-1 sites were in regions of inactive chromatin, consistent with a repressor function for GATA-1 at these sites. Eleven of 13 validated NF-E2 sites (85%) were associated with regions of K562 chromatin enriched for H3K4me3. In HeLa cells, the sites of GATA-1 and NF-E2 occupancy identified in K562 cells were almost never associated with H3K4me3 enrichment. GATA-1 and NF-E2 sites identified by Tamalpais and validated in K562 cells were analyzed in CD71-bright, glycophorin A-bright cultured primary erythroid cells using conventional quantitative ChIP analyses. Of the 13 NF-E2 sites identified in K562 cells, all 13 were also occupied in primary erythroid cells. ChIP-chip is a powerful tool for studying chromatin architecture and identifying transcription factor binding sites in complex genetic loci such as the erythrocyte membrane protein genes. It will be useful in constructing a comprehensive catalogue of chromatin architecture and transcription factor binding of genes expressed in erythroid cells.