Integrative analysis and prediction of human R-loop binding proteins

Abstract

In the past decade there has been a growing appreciation for R-loop structures as important regulators of the epigenome, telomere maintenance, DNA repair and replication. Given these numerous functions, dozens, or potentially hundreds, of proteins could serve as direct or indirect regulators of R-loop writing, reading, and erasing. In order to understand common properties shared amongst potential R-loop binding proteins (RLBPs) we mined published proteomic studies and distilled 10 features that were enriched in RLBPs compared to the rest of the proteome. Applying an easy-ensemble machine learning approach, we used these RLBP-specific features along with their amino acid composition to create random forest classifiers that predict the likelihood of a protein to bind to R-loops. Known R-loop regulating pathways such as splicing, DNA damage repair and chromatin remodeling are highly enriched in our datasets, and we validate two new R-loop binding proteins LIG1 and FXR1 in human cells. Together these datasets provide a reference to pursue analyses of novel R-loop regulatory proteins.

Immunoprecipitation of RNA–DNA hybrid interacting proteins in Trypanosoma brucei reveals conserved and novel activities, including in the control of surface antigen expression needed for immune evasion by antigenic variation

“…Here, we describe a large cohort of putative RNA–DNA hybrid interactors in the protozoan parasite, Trypanosoma brucei , where R-loops have been mapped genome-wide and implicated in both conserved and lineage-specific activities ( 28 ). Consistent with predictions of conserved activities, we find overlap between our data and mammalian studies ( 22 , 27 ), including ribosome-associated proteins and predicted RNA helicases. In fact, our data provide a potentially novel link between these activities, in that we describe a predominantly cytoplasmic DDX (DDX60) that has been found to associate with the ribosome in T. brucei ’s relative, T. cruzi ( 143 ) and influences RNA–DNA hybrid levels in the parasite and, moreover, appears to have moonlighting activities in nuclear DNA repair.…”

“…Though cytoplasmic proteins were not enriched (Figure 1A ), they represented the majority of proteins recovered from each life cycle stage, perhaps suggesting recovery of proteins that are not RNA–DNA hybrid interactors; for instance, ribosomal proteins were prominent in the DRIP-MS dataset ( Supplementary Table S1 ), possibly due to S9.6 recognition of double-stranded RNA. However, the enrichment of ribosomal proteins may also reflect growing evidence for roles of R-loops in rRNA transcription and processing, leading to ribosome biogenesis ( 87 , 88 ), and is consistent with machine-learning predictions of human RNA–DNA hybrid interacting protein types ( 27 ). Moreover, recent work has shown R-loops formed in the nucleus can be found in the cytoplasm of human cells ( 89 ), where it is possible that they interact with cytoplasmic proteins.…”

“…Finally, based on all these datasets, Kumer et al. ( 27 ) searched for common features of the recovered proteins and used machine learning to predict RNA–DNA hybrid interacting proteins across the human proteome. Together, these studies have revealed a wealth of potential RNA–DNA hybrid and R-loop associated activities.…”

R-Loops and R-Loop-Binding Proteins in Cancer Progression and Drug Resistance

“…The recent discovery of other R-loop-interacting proteins, many of which drive R-loop resolution [ 122 , 123 ], will increase the list of proteins that can be targeted to curb cancer progression through the induction of DNA damage.…”

Helicases in R-loop Formation and Resolution

“…Kumar et al. ( 51 ) analyzed the five datasets and found only 12 common R-loop binding proteins: DDX5, NAT10, NPM1, NOP2, DDX18, NOP58, ALYREF, U2AF1, ILF3, RBM14, PDCD11, and MYBBP1A. More recently, Marchena-Cruz et al.…”