Different DNA repair pathways support intact or truncated insertions by R2 retrotransposon protein.

Non–long terminal repeat (non-LTR) retrotransposon proteins copy their RNA template into a genome through coordinated nicking and reverse transcriptase activities of target-primed reverse transcription. Mechanisms by which the first-strand complementary DNA (cDNA) becomes a stably inserted duplex, including requirements for junction formation at the cDNA 3′ end and second-strand synthesis, are unknown. We screened for cellular factors that influence site-specific transgene synthesis into the human genome by an R2 retrotransposon protein. We discovered that insertion lengths and junction signatures differ based on alternative repair processes involving ATR-dependent polymerase θ end joining, 53BP1-directed shieldin and CST–polymerase α–primase fill-in synthesis, or limited strand annealing dependent on CtIP-MRN. These insights shed light on how genome-primed cDNA synthesis by a non-LTR retrotransposon protein can support stable new gene insertion, with major implications for native retrotransposon mobility and genome engineering. Editor's summary: Retrotransposons are selfish mobile DNA elements that propagate through an RNA intermediate. Some encode enzymes with DNA-nicking and reverse transcriptase activities that can be harnessed for therapeutic gene insertion. Using their recently developed system based on avian R2-family retrotransposons to generate new insertions at genomic "safe harbor" sites in human cells, McIntyre et al. identified DNA repair factors essential for completing the insertion process. Manipulating these repair pathways could enhance both the efficiency and safety of this approach for therapeutic genome supplementation. —Di Jiang INTRODUCTION: Transposons are selfish genetic elements that copy themselves within a host genome. Transposons that replicate through an RNA intermediate are designated retrotransposons, which include the ancestral non–long terminal repeat (non-LTR) retrotransposons. Non-LTR retrotransposons encode one or two proteins that bind the RNA transcript and copy it to add new DNA to the genome using a mechanism called target-primed reverse transcription (TPRT). In TPRT, the nicking endonuclease activity of the retrotransposon protein generates a 3′-hydroxyl primer that is transferred to its reverse transcriptase active site for synthesis of cDNA. Subsequently, the long RNA-cDNA duplex generated from TPRT must be converted to stable duplex DNA fully ligated to the target site on each end. How a retrotransposon protein and/or host-cell DNA repair factors attach the cDNA 3′ end to the target site and perform cDNA–complementary second-strand synthesis are not yet defined. RATIONALE: To study the processes by which RNA-cDNA duplexes are converted into stable duplex DNA genes in the human genome, we leveraged our newly developed technology for therapeutic safe-harbor transgene supplementation called PRINT (precise RNA-mediated insertion of transgenes). PRINT uses two RNAs: an mRNA encoding an R2 non-LTR retrotransposon protein that initiates cDNA synthesis at its specific target site in the multicopy ribosomal DNA loci and a template RNA, recognized by the R2 protein, that encodes an autonomous transgene payload. We designed high-content imaging screens and used small interfering RNA (siRNA)–mediated depletion to query the contributions of host DNA repair factors to new gene insertion. Additional molecular and cellular assays informed long-standing questions about host factor roles in new gene insertion by a native non-LTR retrotransposon protein. RESULTS: We describe three DNA repair processes that contribute separately to converting cDNA into a stably genome-embedded transgene. One repair process generates severely truncated gene insertions, a hallmark of non-LTR retrotransposon biology, with minimal or no microhomology at the site of cDNA joining to the upstream target site. This pathway involves 53BP1-directed shieldin and CST–polymerase α (Polα)–primase second-strand fill-in synthesis. The second repair process is a microhomology-mediated, ATR-dependent polymerase theta (Polθ) end joining pathway that often fuses the cDNA to the upstream target site with minimal sequence loss. In addition to these two "join" pathways, we discover a third pathway that produces what we call "anneal" junctions. It relies on CtIP and MRN for short-range end-resection of the upstream target site to promote its annealing to ~20 nucleotides of complementary cDNA sequence. Notably, by inhibiting the two join pathways, we force new gene insertions to be exclusively full-length with precise junctions on both sides of the insertion. CONCLUSION: We demonstrate that PRINT gene insertion is differentially influenced by several repair pathways to resolve the post-TPRT non-LTR retrotransposon insertion intermediate. By suppressing two pathways that support lower precision of cDNA 3′ end attachment to the target site, our entire RNA-based site-specific, safe-harbor delivery system is highly selective for installing only full-length transgenes. Our findings have major implications for understanding the repair of RNA-DNA duplexes in general and for engineering retrotransposon proteins for genomic medicine applications. Multiple DNA repair steps complete transgene insertion after TPRT.: PRINT uses a native R2 non-LTR retrotransposon protein to insert autonomous expression cassettes for a gene of interest (G.O.I.) into nucleolar ribosomal DNA. After TPRT by the R2 protein, resolution of the RNA-cDNA duplex into a stable transgene requires host DNA repair factors. The repair pathway chosen influences whether insertions are intact or truncated. ORF, open reading frame. [The modified nucleus icon comes from Servier, https://smart.servier.com/, and is licensed under CC-BY 3.0.] [ABSTRACT FROM AUTHOR]