Optimization of ChIP-seq for the characterization of expressed genes within bulk chromatin derived from apple (Malus x domestica)

The goal of this study was to apply histone-modification-specific ChIP-seq for the identification of weakly expressed and tissue-specific expressed genes from total genomic DNA in domestic apple (Malus x domestica). We first determined if selected genes predicted to show strong tissue-specificity of expression could be recovered through ChIP using modification-specific antibodies. We targeted known sequences representing apple genes most closely related to Arabidopsis SVP, SOC1, TFL1, AP1, and AG. Using seedling tissues, we carried out extensive optimizations of crosslinking duration, sonication times and intensities, and antibody concentrations. Four of these sequences (SVP-like, SOC1-like, TFL1-like, and AP1-like) amplified selectively from the IP. An apple ACTIN sequence, which we know from EST source analysis to be expressed strongly in numerous tissues, was amplified only after extensive cycling, and did not show enrichment in the IPs. We then scaled up the ChIP using 10 g of seedling tissues and undertook a proof-of-concept experiment with ChIP-seq runs from apple flower bud and leaf. Of ~109,000 sequences from flower bud, ~21,000 matched 7,700 known apple contig/singleton sequences (BLASTN <1E-05), and ~8,800 matched ~4,000 Arabidopsis genes (TBLASTN <1E-05). Of ~226,000 sequences from leaf, ~72,100 matched ~14,300 known apple contig/singletons, and ~31,000 matched ~7,000 Arabidopsis genes. Given the abundance of unmatched, expressed-gene-like sequences in this dataset, we concluded that these limited preliminary experiments were highly successful in identifying previously unknown apple genes.