PAM recognition by miniature CRISPR-Cas12f nucleases triggers programmable double-stranded DNA target cleavage.

In recent years, CRISPR-associated (Cas) nucleases have revolutionized the genome editing field. Being guided by an RNA to cleave double-stranded (ds) DNA targets near a short sequence termed a protospacer adjacent motif (PAM), Cas9 and Cas12 offer unprecedented flexibility, however, more compact versions would simplify delivery and extend application. Here, we present a collection of 10 exceptionally compact (422-603 amino acids) CRISPR-Cas12f nucleases that recognize and cleave dsDNA in a PAM dependent manner. Categorized as class 2 type V-F, they originate from the previously identified Cas14 family and distantly related type V-U3 Cas proteins found in bacteria. Using biochemical methods, we demonstrate that a 5' T- or C-rich PAM sequence triggers dsDNA target cleavage. Based on this discovery, we evaluated whether they can protect against invading dsDNA in Escherichia coli and find that some but not all can. Altogether, our findings show that miniature Cas12f nucleases can protect against invading dsDNA like much larger class 2 CRISPR effectors and have the potential to be harnessed as programmable nucleases for genome editing.

Miniature type V-F CRISPR-Cas nucleases enable targeted DNA modification in cells.

Class 2 CRISPR systems are exceptionally diverse, nevertheless, all share a single effector protein that contains a conserved RuvC-like nuclease domain. Interestingly, the size of these CRISPR-associated (Cas) nucleases ranges from >1000 amino acids (aa) for Cas9/Cas12a to as small as 400-600 aa for Cas12f. For in vivo genome editing applications, compact RNA-guided nucleases are desirable and would streamline cellular delivery approaches. Although miniature Cas12f effectors have been shown to cleave double-stranded DNA, targeted DNA modification in eukaryotic cells has yet to be demonstrated. Here, we biochemically characterize two miniature type V-F Cas nucleases, SpCas12f1 (497 aa) and AsCas12f1 (422 aa), and show that SpCas12f1 functions in both plant and human cells to produce targeted modifications with outcomes in plants being enhanced with short heat pulses. Our findings pave the way for the development of miniature Cas12f1-based genome editing tools.

ERCC1/XPF limits L1 retrotransposition.

Retrotransposons are currently active in the human and mouse genomes contributing to novel disease mutations and genomic variation via de novo insertions. However, little is known about the interactions of non-long terminal repeat (non-LTR) retrotransposons with the host DNA repair machinery. Based on the model of retrotransposition for the human and mouse LINE-1 element, one likely intermediate is an extension of cDNA that is heterologous to the genomic target, a flap intermediate. To determine whether a human flap endonuclease could recognize and process this potential intermediate, the genetic requirement for the ERCC1/XPF heterodimer during LINE-1 retrotransposition was characterized. Reduction of XPF in human cells increased retrotransposition whereas complementation of ERCC1-deficiency in hamster cells reduced retrotransposition. These results demonstrate for the first time that DNA repair enzymes act to limit non-LTR retrotransposition and may provide insight into the genetic instability phenotypes of ercc1 and xpf individuals.

The repurposing of type I-E CRISPR-Cascade for gene activation in plants.

CRISPR-Cas systems are robust and facile tools for manipulating the genome, epigenome and transcriptome of eukaryotic organisms. Most groups use class 2 effectors, such as Cas9 and Cas12a, however, other CRISPR-Cas systems may provide unique opportunities for genome engineering. Indeed, the multi-subunit composition of class 1 systems offers to expand the number of domains and functionalities that may be recruited to a genomic target. Here we report DNA targeting in Zea mays using a class 1 type I-E CRISPR-Cas system from S. thermophilus. First, we engineer its Cascade complex to modulate gene expression by tethering a plant transcriptional activation domain to 3 different subunits. Next, using an immunofluorescent assay, we confirm Cascade cellular complex formation and observe enhanced gene activation when multiple subunits tagged with the transcriptional activator are combined. Finally, we examine Cascade mediated gene activation at chromosomal DNA targets by reprogramming Zea mays cells to change color.

Characterization of pre-insertion loci of de novo L1 insertions.

The human Long Interspersed Element-1 (LINE-1) and the Short Interspersed Element (SINE) Alu comprise 28% of the human genome. They share the same L1-encoded endonuclease for insertion, which recognizes an A+T-rich sequence. Under a simple model of insertion distribution, this nucleotide preference would lead to the prediction that the populations of both elements would be biased towards A+T-rich regions. Genomic L1 elements do show an A+T-rich bias. In contrast, Alu is biased towards G+C-rich regions when compared to the genome average. Several analyses have demonstrated that relatively recent insertions of both elements show less G+C content bias relative to older elements. We have analyzed the repetitive element and G+C composition of more than 100 pre-insertion loci derived from de novo L1 insertions in cultured human cancer cells, which should represent an evolutionarily unbiased set of insertions. An A+T-rich bias is observed in the 50 bp flanking the endonuclease target site, consistent with the known target site for the L1 endonuclease. The L1, Alu, and G+C content of 20 kb of the de novo pre-insertion loci shows a different set of biases than that observed for fixed L1s in the human genome. In contrast to the insertion sites of genomic L1s, the de novo L1 pre-insertion loci are relatively L1-poor, Alu-rich and G+C neutral. Finally, a statistically significant cluster of de novo L1 insertions was localized in the vicinity of the c-myc gene. These results suggest that the initial insertion preference of L1, while A+T-rich in the initial vicinity of the break site, can be influenced by the broader content of the flanking genomic region and have implications for understanding the dynamics of L1 and Alu distributions in the human genome.

The human LINE-1 retrotransposon creates DNA double-strand breaks.

Long interspersed element-1 (L1) is an autonomous retroelement that is active in the human genome. The proposed mechanism of insertion for L1 suggests that cleavage of both strands of genomic DNA is required. We demonstrate that L1 expression leads to a high level of double-strand break (DSB) formation in DNA using immunolocalization of gamma-H2AX foci and the COMET assay. Similar to its role in mediating DSB repair in response to radiation, ATM is required for L1-induced gamma-H2AX foci and for L1 retrotransposition. This is the first characterization of a DNA repair response from expression of a non-long terminal repeat (non-LTR) retrotransposon in mammalian cells as well as the first demonstration that a host DNA repair gene is required for successful integration. Notably, the number of L1-induced DSBs is greater than the predicted numbers of successful insertions, suggesting a significant degree of inefficiency during the integration process. This result suggests that the endonuclease activity of endogenously expressed L1 elements could contribute to DSB formation in germ-line and somatic tissues.

HPV 5 and 8 E6 expression reduces ATM protein levels and attenuates LINE-1 retrotransposition.

The expression of the E6 protein from certain members of the HPV genus ß (ß HPV 5 and 8 E6) can disrupt p53 signaling by diminishing the steady state levels of two p53 modifying enzymes, ATR and p300. Here, we show that ß-HPV 5 and 8 E6 are also capable of reducing the steady state levels of another p53 modifying enzyme, ATM, and as a result restrict LINE-1 retrotransposition. Furthermore, we show that the reduction of both ATM and LINE-1 retrotransposition is dependent upon the ability of ß-HPV 8 E6 to bind and degrade p300. We use inhibitors and dominant negative mutants to confirm that ATM is needed for efficient LINE-1 retrotransposition. Furthermore, neither sensitivity to LINE-1 expression nor LINE-1 induced DSB formation is altered in an ATM deficient background. Together, these data illustrate the broad impact some ß-HPVs have on DNA damage signaling by promoting p300 degradation.

Alu-linked hairpins efficiently mediate RNA interference with less toxicity than do H1-expressed short hairpin RNAs.

RNA interference has become a powerful tool for specific inhibition of gene expression in mammalian cells. Expression constructs allow for the long-term delivery of short interfering RNAs, usually through the expression of Pol III-transcribed hairpins. In some instances, these expression systems have been shown to have side effects, including induction of the interferon response and cytotoxicity. Here we demonstrate that H1-expressed hairpins, as well as the cloning vector, reduce the plating efficiency of HeLa cells. This toxicity is abrogated by coexpression of the hairpin in the same transcript as a human Alu repetitive element. These Alu-linked hairpins retain the ability to knock down expression of target mRNAs. This modification, which we term SINE (short interspersed repetitive element)-enhanced short hairpin RNA, provides an alternative expression system for hairpins with reduced side effects.