Human promoter CpG islands contain primate-specific repeats, cluster and resist reprogramming

Abstract

Imperfect tandem repeats (TRs) of greater than or equal to 400nt are associated with 365 human 5-prime genic CpG islands (TR-CGIs). Most are clustered at chromosome ends, with a high density across chromosome 19. These genes are enriched in neurodevelopmental/behavioral disorders and show interindividual variation in methylation levels. A subset of TR-CGIs is highly methylated and remains so during reprogramming to primed iPSCs, but become unmethylated in naive PSCs, as do imprinting control regions. Transcript levels correlate with methylation for some TR-CGI genes. TR-CGIs occur as orthologs in primates, but the corresponding mouse promoter-CGIs are without TRs and unmethylated. Thus, non-imprinted TR-CGIs accompanied primate evolution, with unique ability to acquire methylation during embryonic development and resist reprogramming to a pluripotent stem cell state.