Efficient genome editing in filamentous fungi via an improved CRISPR-Cas9 ribonucleoprotein method facilitated by chemical reagents

Abstract

DNA double-strand break (DSB) repair induced by the RNA-programmed nuclease Cas9 has become a popular method for genome editing. Direct genome editing via Cas9-CRISPR gRNA (guide RNA) ribonucleoprotein (RNP) complexes assembled in vitro has also been successful in some fungi. However, the efficiency of direct RNP transformation into fungal protoplasts is currently too low. Here, we report an optimized genome editing approach for filamentous fungi based on RNPs facilitated by adding chemical reagents. We increased the transformation efficiency of RNPs significantly by adding Triton X-100 and prolonging the incubation time, and the editing efficiency reached 100% in Trichoderma reesei and Cordyceps militaris. The optimized RNP-based method also achieved efficient (56.52%) homologous recombination integration with short homology arms (20 bp) and gene disruption (7.37%) that excludes any foreign DNA (selection marker) in T. reesei. In particular, after adding reagents related to mitosis and cell division, the further optimized protocol showed an increased ratio of edited homokaryotic transformants (from 0% to 40.0% for inositol and 71.43% for benomyl) from Aspergillus oryzae, which contains multinucleate spores and protoplasts. Furthermore, the multi-target engineering efficiency of the optimized RNP transformation method was similar to those of methods based on in vivo expression of Cas9. This newly established genome editing system based on RNPs may be widely applicable to construction of genome-edited fungi for the food and medical industries, and has good prospects for commercialization.

Fig. 1

Comparative analysis of ura5‐deficient mutants obtained through different methods. A. Phenotype of Rut‐C30 and its Trura5‐deficient mutants on PDA plates (with/without uridine). Rut‐C30: wild type. 1D4–6: Trura5‐deficient mutant (UV). AR3–5: Trura5‐deficient mutant (edited with pcbh1 promoter‐controlled Cas9). UPDC: Trura5‐deficient mutant (edited with ppdc promoter‐controlled Cas9). 3x–1 ˜ 3x–3: Trura5‐deficient mutant (edited with RNP). B. Filter paper activities of Rut‐C30 and its mutants induced with 2% (w/v) wheat bran and 3% (w/v) Avicel. C. Filter paper activities of Rut‐C30 and its mutants induced with 1% (w/v) lactose. Scale bar = 1 cm. Error bars show the standard deviation of three replicates. All P values are obtained from two‐tailed t tests using Microsoft Excel: *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 2

Trlae1 disruption based on RNP transformation without DNA fragment or selectable marker addition. A. Schematic representation of Trlae1 gene disruption through the RNP‐based CRISPR/Cas9 strategy. B. Sequence verification of the lae1 gene from white colonies. The protospacer sequence is underlined and the PAM sequence is shown in red. ‘‐’ indicates deletion and ‘>’ indicates substitution mutation.

Fig. 3

Overview of RNP‐assisted genome editing in filamentous fungi. Downscaling of the final optimized transformation procedure. Mature conidia were collected with 0.02% Tween 80. The germinated conidia are treated with 1 mg l⁻¹ benomyl on PDA plates to induce mitotic arrest (required for fungi with multinucleate protoplasts). After 15 h, mycelia were washed with sterilized water three times to remove benomyl. Protoplast preparation followed a previously reported procedure, and the protoplast concentration was diluted to 10⁶ ml^‐1. RNP(s) and donor DNA(s) (optional) were added to 200 μl protoplast suspensions. After addition of PEG 6000 solution, the surfactant Triton X‐100 is incorporated at a final concentration of 0.006% (w/v). The incubation time was prolonged to more than 25 min for full release of mitotic arrest (25 min for NHEJ, 50 min for HDR). Protoplast regeneration and transformant screening also followed a previously reported procedure. M: mitosis phase, G1: G1 phase in interphase, S: S phase in interphase, G2: G2 phase in interphase.