Sebastian M. Siegner

Pancreatic beta cells can compensate for increased metabolic demand by increasing their function. While many studies have investigated mechanisms of beta cell decompensation during persistent metabolic stress, less is known how beta cells increase insulin synthesis and secretion. Here we identify the non-canonical cytoplasmic terminal nucleotidyltransferase Tent5a in a functional RNAi screen of RNA binding proteins as a positive regulator of insulin production. Tent5a expression positively correlates with obese and normoglycemic mouse models exhibiting compensated beta cell function, while its levels negatively correlate with beta cell failure. Increased Tent5a expression can be triggered by ribosome collisions and is repressed by ER stress and lipotoxic inflammation. Tent5a overexpression extended the poly(A) tails of insulin transcripts, enhanced mRNA stability and increased insulin content in INS-1E cells and human islets microtissues, while Tent5a knockout cells showed shortened Insulin mRNA half-life and reduced insulin content. Cytosolic polyadenylation activity is dependent on Tent5a tethering to the endoplasmic reticulum (ER) via Fndc3 proteins, leading to increased ER function. Our data identify a role of Tent5a in enhancing insulin production, which may be involved in boosting insulin synthesis and maintaining euglycemia in obesity.

DNA double-strand breaks (DSBs) both pose threats to genome integrity and are commonly used for genome editing applications. Structural features of DSB ends play key roles in determining DNA repair pathway usage and outcomes during genome editing, but the cellular factors involved in these processes are only partially known. Through genome-wide CRISPRi screening, we identify ERCC6L2 as critical for repairing Cas12a-induced staggered DSBs but irrelevant for Cas9-induced blunt DSBs. We show that ERCC6L2 acts as a protection factor for staggered DSBs with either 5′ or 3′ polarity, preventing large deletions and translocations stemming from DNA damage induced by Cas12a, TALENs, or dual Cas9 nicks. Furthermore, ERCC6L2 loss hyper-sensitizes cells to multiple staggered DSBs induced by promiscuous Cas12a activity or etoposide-induced TOP2 trapping. By combining genetics and biochemical reconstitution, we find that ERCC6L2 counteracts MRE11-RAD50-NBS1 (MRN)-mediated resection by binding and melting staggered DNA ends, thereby promoting accurate end joining. Our data reveal a protective role of ERCC6L2 in staggered-end DSB repair, which suggests the molecular underpinnings of pathology in patients with ERCC6L2 mutations and cautions against using overhang-inducing genome editing tools for their treatment.

Despite their well-defined genetic background, repeat expansion diseases (REDs) still represent an unmet medical need, with no causative therapy offered to patients. The strategy of repeat shortening using genome editing tools is very attractive because a single intervention can result in permanent repair of the disease-causing mutation. However, a limited understanding of DNA repair mechanisms in repetitive sequences complicates the prediction and control of the editing effects. Using a CRISPR interference (CRISPRi) screen, we identified pathways and factors responsible for the repair of staggered cuts generated by Cas12a within CAG repeat tracts. This analysis revealed a central role for interstrand crosslink (ICL) repair factors in mediating CAG repeat contraction, with DCLRE1A emerging as a key effector. We demonstrated that DCLRE1A recognizes and binds structures generated by Cas12a. Moreover, DCLRE1A interacts with SLX4, promoting the generation of pure contractions, and with POLI, leading to the formation of inverted repeats at the break site during template switching mechanisms. We then used this knowledge to increase the contribution of pure contractions to the pool of editing outcomes using fusions of Cas12a with DNA repair proteins. Our study indicates that Cas12a can be used as an effective tool for generating repeat contractions. Although the mechanisms leading to repeat shortening are complex, understanding them can help researchers develop more precise therapeutic strategies with greater control of the editing process.

The DNA damage response (DDR) is a multi-faceted network of pathways that preserves genome stability. Unraveling the complementary interplay between these pathways remains a challenge. Here, we comprehensively mapped genetic interactions for all core DDR genes using combinatorial CRISPRi screening. We discovered myriad new connections, including interactions between cancer genes and small molecule targets. We focused on two of the strongest interactions: FEN1 / LIG1 : WDR48 and FANCM : SMARCAL1 . First, we found that WDR48 works with USP1 to restrain overactive translesion synthesis in FEN1/LIG1-deficient cells, and that a preclinical inhibitor of USP1 specifically kills FEN1-deficient cells. Second, we found that SMARCAL1 and FANCM suppress DNA double-strand break (DSB) formation at TA-rich repeats in late replicating regions that otherwise escape into mitosis and cause nuclear fragmentation. We present fundamental insights into genome maintenance processes and our dataset provides a springboard for mechanistic investigations into connections between DDR factors and suggests multiple interactions that could be exploited in cancer therapy.

Fanconi Anemia (FA) is a debilitating genetic disorder with a wide range of severe symptoms including bone marrow failure and predisposition to cancer. CRISPR-Cas genome editing manipulates genotypes by harnessing DNA repair and has been proposed as a potential cure for FA. But FA is caused deficiencies in DNA repair itself, preventing the use of editing strategies such as homology directed repair. Recently developed base editing (BE) systems do not rely on double stranded DNA breaks and might be used to target mutations in FA genes, but this remains to be tested. Here we develop a proof of concept therapeutic base editing strategy to address two of the most prevalent FANCA mutations in patient cells. We find that optimizing adenine base editor construct, vector type, guide RNA format, and delivery conditions lead to very effective genetic modification in multiple FA patient backgrounds. Optimized base editing restored FANCA expression, molecular function of the FA pathway, and phenotypic resistance to crosslinking agents. ABE8e mediated editing in primary hematopoietic stem and progenitor cells from an FA patient was both genotypically effective and restored FA pathway function, indicating the potential of base editing strategies for future clinical application in FA.