---

In contrast to the homologous directed repair (HDR) pathway, the oleaginous yeast Yarrowia lipolytica prefers the non-homologous end-joining (NHEJ) pathway as a double-strand break (DSB) repair system. Due to this selectivity for the DNA repair mechanism, it is difficult to use homologous recombination to remove important genes or insert foreign genes into certain places precisely. In recent years, a base editing system based on the CRISPR/Cas9 system has been developed, which can create nonsense mutations within the target gene to precisely block gene expression without the need for DSB repair. It works by a nucleotide deaminase being fused to dead Cas9 (dCas9) , which is recruited to a specific DNA site via gRNA. Among the many base editor tools, the Target-AID system, based on Petromyzon marinus Cytidine deaminase (PmCDA1) , has been efficiently used in a variety of organisms such as Saccharomyces cerevisiae, Escherichia coli and mammalian cells (Edit window range of -15 to -20bp) [2-4]. This component (BioBrick BBa_K3645011) was first registered by iGEM20_Peking Team in 2020 (Fig 1).

We are trying to improve the ability of BioBrick BBa_K3645011 to work more efficiently in Y.lipolytica by using optimised related components. We first supposed whether replacing dCas9 with nCas9 will increase the targeting and editing efficiency of CBE. We also hypothesized that the expression level of CBE could be increased through using the stronger promoters thus increasing editing efficiency. For this,the improving of BioBrick BBa_K3645011 are divided into three sections as follows: (i) the replacement of dCas9 by nick Cas9 (nCas9) , (ii) promoter optimization and (iii) Cas9 without protospacer adjacent motif (PAM) limitation (SpRY)[5] to expand the editing range.

Fig. 1 The principle diagram of cytosine base editor

---

First, to ensure that BBa_K3645011 functions in Y.lipolytica , we added the SV40 nuclear localization signal (NLS, PKKKRKV) at both terminals of the component via the primers. Further, the subsequent recombinant plasmids were all constructed on the basis of the original plasmid pUC-PTEF by the restriction-ligation method. On this basis, the plasmid pUC-PTEF-CBE (D10A, H840A) was successfully constructed. To compare the effect nCas9 on editing efficiency, we performed a point mutation using MegaWHOP PCR on the plasmid pUC-PTEF-CBE to obtain pUC-PTEF-nCBE (D10A).

Then, the above linearized recombinant plasmids were transformed into Y. lipolytica po1f . Briefly, the cells in the exponential growth phase were aspirated in 1 ml and washed twice with 100 mM phosphate buffer (pH 7.0). The above cells were suspended in a transformation solution, containing 90 µL 50% PEG4000, 5 µL 2M lithium acetate, 5 µL boiled single stand DNA and 5 µL DNA products (no less than 200 ng). And spread on selected plates after 1h at 30℃. Other recombinant plasmids or linearized fragments were subsequently transformed using this method. The sgRNA plasmid pYLXP'-PTEF-sgRNA1 targeting TRP1 (5′-gggccaactcaacggactaa-3′) was further transformed into the obtained strains (polf-CBE/polf-nCBE) and tested on screening plates.

On the basis of these results, we envisage further improvements in the editing efficiency of the component and its application range. To address the problem of editing efficiency, we hope to optimize the expression level of base editing proteins to improve base editing efficiency. For this purpose, we replaced various promoters, including PEXP, PGAP, and PTEFin, and generated pUC-PEXP-nCBE, pUC-PGAP-nCBE and pUC-PTEFin-nCBE. To expand the editing range of the gene, we replaced the original nCas9 with a PAM-restricted Cas9 variant (SpRY) and combined the best promoters to construct pUC-PTEFin-nSpRYCBE. Meanwhile, we selected three sites in TRP that do not contain 5'-NGG-3' for testing (5′-agggccaactcaacggacta-3′ (sgRNA2), 5′-ggccaactcaacggactaat-3′ (sgRNA3), 5′-gccaactcaacggactaatg-3′ (sgRNA4)) to ensure the validity of this optimization The above improvements to the BBa_K3645011 achieve both optimization for high editing efficiency and high editing range in Y.lipolytica .

---

To ensure that the component BBa_K3645011 can be applied in Y. lipolytica po1f , we added NLS at both terminals of the CBE. Meanwhile, we selected the TRP1 gene as a target for the introduction of nonsense mutations. The TRP1 gene can be inactivated by introducing a nonsense mutation at -16 bp in the PAM sequence through a C to T mutation. To calculate the editing efficiency of TRP1 , we tested the growth of transformants on synthetic complete medium without tryptophan (SC-W) (Fig 2). However, the results showed that only one of the 20 selected transformants did not grow on the plate (Fig 3). It has been shown that the introduction of nick in targeted genes using nCas9 can facilitate DNA repair to improve editing efficiency. Therefore, we replaced dCas9 in the component with nCas9 and tested it again. It was found that three of the twentieth transformants were not grown from SC-W plates (Fig 2B). Further, we incubated the transformants on YPD plates containing the screened resistance and then selected to SC-W plates after transferring them for five generations. We found that extending the plasmid curing time can effectively improve editing efficiency, up to 35% (7/20) (Fig 3).

Fig. 2 The workflow for testing the efficiency of base editing.

Fig.3The editing efficiency of TRP1 gene under different Cas9 conditions (dCas9 or nCas9).

Next, we wanted to further improve the editing efficiency of this element in Y. lipolytica . We hypothesized that optimizing the expression intensity of the components would improve the editing efficiency. We constructed several different promoter-driven plasmids such as pUC-PEXP-nCBE, pUC-PGAP-nCBE and pUC-PTEFin-nCBE, and integrated them into the genome for testing. Excitingly, the editing efficiency increased under all three promoters, reaching 50% (10/20), 65% (13/20) and 75% (15/20), respectively. Similarly, we extended the incubation time and the editing efficiency was further improved after five generations to 70% (14/20), 85% (17/20) and 85% (17/20) (Fig 4). To expand the use of base editing, we chose the PAM-less SpRY to replace the conventional nCas9 for editing. We adjusted the original targeting site to ensure no connection to PAM in sgRNA (5'-NGG-3'). The experimental results showed that nSpRY effectively extended the editing range, with editing efficiencies of 75% (15/20), 60% (12/20) and 40% (8/20) for three different sgRNAs after five generations (Fig 5). It is speculated that there are two main reasons for the change in editing efficiency: the replacement of Cas9 protein and the change in editing window. In order to facilitate the use of the tool by subsequent teams, we have registered the final optimized composite component with the number BBa_K4343130. In summary, we have enhanced the editing efficiency and the editing range of BBa_K3645011, proving the authenticity of our strategy and component improvement.

Fig.4 The editing efficiency of TRP1 gene after promoter replacement.

Fig.5 The editing efficiency of TRP1 gene under different nCas9 types (nCas9/nSpRY)

---

[1] Verbeke, J., Beopoulos, A., & Nicaud, J. M. (2013). Efficient homologous recombination with short length flanking fragments in Ku70 deficient Yarrowia lipolytica strains. Biotechnology letters, 35(4), 571-576.

[2] Liu, Y., Lin, Y., Guo, Y., Wu, F., Zhang, Y., Qi, X., ... & Wang, Q. (2021). Stress tolerance enhancement via SPT15 base editing in Saccharomyces cerevisiae. Biotechnology for Biofuels, 14(1), 1-18.

[3] Banno, S., Nishida, K., Arazoe, T., Mitsunobu, H., & Kondo, A. (2018). Deaminase-mediated multiplex genome editing in Escherichia coli . Nature microbiology, 3(4), 423-429.

[4] Fukushima, T., Tanaka, Y., Adachi, K., Masuyama, N., Tsuchiya, A., Asada, S., ... & Kitamura, T. (2021). CRISPR/Cas9-mediated base-editing enables a chain reaction through sequential repair of sgRNA scaffold mutations. Scientific reports, 11(1), 1-12.

[5] Walton, R. T., Christie, K. A., Whittaker, M. N., & Kleinstiver, B. P. (2020). Unconstrained genome targeting with near-PAMless engineered CRISPR-Cas9 variants. Science, 368(6488), 290-296.

[6] Anzalone, A. V., Koblan, L. W., & Liu, D. R. (2020). Genome editing with CRISPR–Cas nucleases, base editors, transposases and prime editors. Nature biotechnology, 38(7), 824-844.