Fig. 4: FCA and FY promote R-loop resolution at a subset of targets, so protecting the genome from DNA damage during replication.

a DRIP–qPCR analyzing R-loops at UA2, UA10B, and XRCC4 (UA R-loops) in Col-0 and fca-9, with and without RNase H treatment. p1 to p4 are amplicons in the qPCR, which are indicated in the schematic of each locus. Data are mean ± s.d. from three biological replicates. b DRIP–qPCR analyzing UA R-loops in Col-0 and fy-2. Data are mean of two biological replicates. c seedlings were treated with and without bleomycin. After treatment and recovery (see “Methods”), plants with or without true leaves were scored. Data are presented as the number of seedlings in two categories. d “Detangle model”. FCA associates with co-transcriptionally formed R-loops, potentially with Pol II stalling. FCA facilitates the deposition of m⁶A onto the nascent transcript by the m⁶A writer complex. This enhances the multivalent interactions promoting the formation of dynamic nuclear condensates. Consequently, this promotes polyadenylation at noncanonical polyadenylation (p(A)) sites, concomitantly resolving the R-loop. At FLC, this process is linked to chromatin silencing via interaction of the 3′-end processing factors with a histone K4 demethylase FLD. Without FCA or FY or the m⁶A writer complex, R-loops accumulate, and transcripts polyadenylate at efficient p(A) sites. Source data for Fig. 4a–c are provided as a Source Data file.