In vivodiversification of target genomic sites using processive T7 RNA polymerase-base deaminase fusions blocked by RNA-guided dCas9

Abstract

32Diversification of specific DNA segments typically involve in vitro generation of large sequence 33 libraries and their introduction in cells for selection. Alternative in vivo mutagenesis systems on 34 cells often show deleterious offsite mutations and restricted capabilities. To overcome these 35 limitations, we have developed an in vivo platform to diversify specific DNA segments based on 36 protein fusions between various base deaminases (BD) and the T7 RNA polymerase (T7RNAP) 37 that recognizes a cognate promoter oriented towards the target sequence. The transcriptional 38 elongation of these fusions generates transitions C to T or A to G on both DNA strands and in 39 long DNA segments. To delimit the boundaries of the diversified DNA, the catalytically dead 40 Cas9 (dCas9) is tethered with custom-designed crRNAs as a "roadblock" for BD-T7RNAP 41 elongation. While the efficiency of this platform is demonstrated in E. coli, the system can be 42 adapted to a variety of bacterial and eukaryotic hosts. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58Directed evolution enables the selection of protein variants with improved properties as 59 therapeutics and biocatalysts 1, 2, 3 . The generation of genetic variability followed by a screening 60 process are the essential steps of directed evolution 4, 5 . In vitro mutagenesis techniques (e.g., 61error-prone PCR) can quickly produce large number of variants of the target gene but their 62 selection requires, in most cases, cloning and transformation into a host cell for expression (e.g., 63E. coli). These steps are time-consuming and labor-intensive, especially when iterative cycles 64 of mutagenesis and selection are needed. Cell-free selection methods are also feasible 6 , but 65 they are technically demanding and functional expression of complex proteins (e.g. membrane 66 proteins, multimeric enzymes) is often difficult to achieve. Hence, in vivo mutagenesis systems 67 are preferred because the generation of genetic variants, their expression and selection can 68 done in a continuous process, which accellerates directed evolution 5, 7 . 70Long-established in vivo mutagenesis methods (e.g., chemical mutagens, radiations) do not 71 target specific genes and are deleterious for the host cell due to accumulation of random 72 mutations in the genome 8 . Similarly, "mutator" host strains do not allow the concentration of the 73 mutagenic activity on the target gene, inducing the accumulation of unwanted mutations in the 74 host genome 9, 10, 11, 12 . A few in vivo mutagenesis systems with targeted specificity have been 75 reported for E. coli and yeast cells, which are the preferred hosts for cloning and expression of 76 gene libraries 5 . For example, an error-prone variant of E. coli DNA polymersase I enables the 77 mutagenesis of cloned genes in ColE1 plasmids, although it concentrates the mutations close 78 to the origin of replication 13 . A different approach involves the transformation of E. coli with 79 mutant oligonucleotide libraries, targeting one or multip...

A synthetic RNA-mediated evolution system in yeast

“…Here, compared to other in vivo directed evolution systems in yeast (14, 15, 17), limitations of the current version of CRAIDE exist and need consideration and further improvement for the system to be applicable for efficient in vivo directed evolution across multiple species. Indeed, with a mutation rate in the order of 3.26 × 10 −6 per base, RNA-mediated repair of genomic contexts using variant RNA donors as demonstrated in this study is still 2-3 orders of magnitude less efficient compared to state-of-the-art in vivo directed evolution methods for bacteria, yeast, and mammalian cells, like OrthoRep, ICE and TRACE (11, 14, 16, 17, 49). Furthermore, even though RNA-mediated DNA repair has previously been reported in wild-type yeast (29, 40), in its current version, CRAIDE requires disruption of host RNases for successful RNA-mediated repair of genomic DSBs.…”

“…Here, prime editor demonstrated RNA-mediated genome engineering using in vitro -edited donor-amended gRNAs (prime editing gRNAs) (7), while TRACE and T7-DIVA demonstrated that T7RNAP fused to base editors could be applied for continuous in vivo mutagenesis of target genes controlled by genomically integrated T7 promoters (11, 16). Individually, these new technologies enable >10 −4 mutations per base in engineered T7pro-driven open reading frames sized up to 2 kb, and nuclease-deficient integration of mutant bases in a prime editor window of approximately 30 bases (7, 11, 16). In the future, we envision that the in vivo variant donor delivery and editing window size of CRAIDE together with the high editing efficiencies of these technologies could present appealing mergers for development of efficient in vivo continuous evolution in broad genomic contexts, as well as providing a tool for more foundational basic research on RNA-mediated evolution.…”

“…Importantly, when developing systems for directed evolution in vivo , orthogonal mutagenesis and subsequent targeted delivery of mutant donors is of primary importance, in order to efficiently dereplicate sequence to function under selective conditions (19, 20). To address these considerations, creative bioprospecting and mixing of biological parts from diverse hosts have proven successful, including delivery of DNA mutant donors by heterologous faulty DNA polymerases and targeted base-editing using protein-fusions strategies (9–11, 13, 14, 16). Interestingly, various viral phylogenies store genetic information with low replicative fidelity (up to 10 −4 per base per infection) in the form of RNA-encoded genomes (21), and viral-derived components have been a rich source for prospecting parts for synthetic directed evolution systems (8, 10, 22, 23).…”

Potential for Applying Continuous Directed Evolution to Plant Enzymes: An Exploratory Study

“…In continuous directed evolution, the function of the target gene is typically coupled to the growth rate of the host microbe by forcing the microbe to rely on the target gene, e.g., by complementing the loss of an essential host activity with a plasmid-borne target gene. Using this method, improved target genes are immediately identifiable with no need for additional screening [ 27 , 28 , 29 ]. In addition, tracking enzyme function in vivo avoids possible artifacts of in vitro screening assays, which may be poor facsimiles of physiological conditions.…”

Potential For Applying Continuous Directed Evolution To Plant Enzymes

“…by complementing the loss of an essential host activity with a plasmid-borne target gene. Using this method, improved target genes are immediately identifiable with no need for additional screening [26][27][28]. Also, tracking enzyme function in vivo avoids possible artifacts of in vitro screening assays, which may be poor facsimiles of physiological conditions.…”