Development of a CRISPR/Cpf1 system for targeted gene disruption in Aspergillus aculeatus TBRC 277

Background CRISPR-Cas genome editing technologies have revolutionized biotechnological research particularly in functional genomics and synthetic biology. As an alternative to the most studied and well-developed CRISPR/Cas9, a new class 2 (type V) CRISPR-Cas system called Cpf1 has emerged as another versatile platform for precision genome modification in a wide range of organisms including filamentous fungi. Results In this study, we developed AMA1-based single CRISPR/Cpf1 expression vector that targets pyrG gene in Aspergillus aculeatus TBRC 277, a wild type filamentous fungus and potential enzyme-producing cell factory. The results showed that the Cpf1 codon optimized from Francisella tularensis subsp. novicida U112, FnCpf1, works efficiently to facilitate RNA-guided site-specific DNA cleavage. Specifically, we set up three different guide crRNAs targeting pyrG gene and demonstrated that FnCpf1 was able to induce site-specific double-strand breaks (DSBs) followed by an endogenous non-homologous end-joining (NHEJ) DNA repair pathway which caused insertions or deletions (indels) at these site-specific loci. Conclusions The use of FnCpf1 as an alternative class II (type V) nuclease was reported for the first time in A. aculeatus TBRC 277 species. The CRISPR/Cpf1 system developed in this study highlights the feasibility of CRISPR/Cpf1 technology and could be envisioned to further increase the utility of the CRISPR/Cpf1 in facilitating strain improvements as well as functional genomics of filamentous fungi. Supplementary Information The online version contains supplementary material available at 10.1186/s12896-021-00669-8.


Background
Filamentous fungi are of great value in the commercial production of enzymes and heterologous proteins [1,2], a wide range of primary and secondary metabolites such as organic acids and pharmaceuticals [3,4], as well as bio-based chemicals, and biofuels [5,6]. One of the ways to increase the fungi potential in commercial applications is by means of strain improvement via genetic engineering [7]. However, the conventional genetic engineering tools via classical mutagenesis and homologous recombination remain challenging due to additional complexity and time, as well as low efficiency, particularly in targeting multiple genes [8]. Therefore, the development of versatile genetic tools, e.g., CRISPRbased, for filamentous fungi modification would be of great value [9].
CRISPR/Cpf1 is characterized as a novel class 2 (type V) system with distinct features compared to Cas9 [28]. It is a single RNA-guided endonuclease which recognizes a thymidine-rich protospacer-adjacent motif (PAM) and produces staggered cuts distal to the PAM site [29]. This type V CRISPR/Cpf1 system was first harnessed and shown to have robust genome editing activity in mammalian cell lines [27], as well as targeted mutations in plants and other eukaryotic cells [30]. Interestingly, Cpf1 is a dual nuclease that not only cleaves target DNA but also processes its own crRNA array [29,31]. CRISPR/Cpf1-mediated DNA cleavage is guided by a short single crRNA (42-44 nt), in contrast to Cas9 that uses both crRNA and tracrRNA. More importantly, Cpf1 makes a staggered DNA double-stranded break resulting in a five-nucleotide 5′-overhang distal to the PAM site [29], whereas Cas9 creates blunt ends proximal to the PAM site [15]. However, both Cas9 and Cpf1 need a seed sequence at the PAM-proximal side of the protospacers, which is critical for DNA recognition and cleavage. With these advantages, Cpf1 system has been used for multiplex gene editing in mammalian cells, where up to four genes were simultaneously edited using just a single crRNA array spaced by direct repeats (DR) [29].
Cpf1 nuclease and the crRNA are two essential components in the formation of an adjustable CRISPR toolbox for genetic manipulation (Fig. 1). In general, Cpf1 expression is driven by an RNA polymerase II promoter [32]. The guide RNAs are noncoding small RNAs with guide sequence at their 3′-end crRNA and are naturally expressed using an RNA polymerase III (Pol III) promoter [33]. However, due to the complexity and uncertainty of genome information, some endogenous Pol III promoters from filamentous fungi are difficult to identify or are not suitable for crRNA transcription [12]. Therefore, suitable crRNA expression cassettes are needed to construct efficient CRISPR-based tools. Until recently, several methods have been developed to enhance the expression of crRNAs by engineering different RNA processing machineries, including a self-cleavable ribozyme from a virus [34], the endogenous tRNA processing enzymes [35], and pre-crRNA array consisting of direct repeats [29]. These systems are used to generate crRNAs from a single primary polycistronic transcript driven by either Pol II or Pol III promoter. Among these methods, a single pre-crRNA array and a tRNA-based CRISPR system have been shown to boost multiplex genome-editing capability and efficiency without introducing exogenous ribonucleases or additional tracrRNA in the CRISPR/Cas9 system [21].
Over the past several years, various studies have demonstrated the great potential of CRISPR/Cpf1 as a tool for genome engineering, owing to user-designated site specificity of Cpf1 endonuclease activity and the simplicity of crRNA designs [27,30]. Since the first harnessing of Cpf1 in human and animal cell lines [27], until recenty, AsCpf1 and LbCpf1 were reported to effectively catalyze oligonucleotide-mediated genomic site-directed mutagenesis and simultaneous gene deletions/insertions in a few filamentous fungi [36][37][38]. However, there have been no reports for the use of FnCpf1 from F. tularensis subsp. novicida U112 for genome editing in Aspergillus filamentous fungus despite its successful application in yeast [39][40][41]. Here, we established a single vector for the CRISPR/Cpf1 platform derived from FnCpf1 and a crRNA with the endogenous tRNA-processing system for CRISPR-mediated gene editing in Aspergillus aculeatus TBRC 277, a wild-type filamentous fungus. As a proof of principle, pyrG gene was used as the target gene for the CRISPR/Cpf1-mediated gene editing to generate uracil auxotrophic fungi.

Construction of CRISPR/Cpf1 plasmid backbone
To establish the CRISPR/Cpf1 system in A. aculeatus TBRC 277 (Fig. 2a), an expression plasmid carrying the P TEF1 -Cpf1-TEF1 TT cassette and bleomycin selection marker was initially constructed. For Cpf1 gene, we codon-optimized the F. tularensis subsp. novicida Cpf1, so called FnCpf1, and attached a SV40 nuclear localization signal (NLS) fragment at the C-terminal to ensure successful nuclear compartmentalization in the fungal host. Specifically, construction of CRISPR/Cpf1 backbone carrying 5.5 kb autonomously replicating plasmids (AMA1) was successfully obtained by removing Cas9 from pFC333 plasmid [22], followed by MCS insertion that allows for expression of a 3.93 kb Cpf1-SV40 NLS gene fusion controlled by a constitutive A. nidulans TEF1 promoter. The successfully constructed plasmid, referred to as pCRISPR0 ( Supplementary Fig. S1), was verified by Sanger sequencing. Then, insertion of kanamycin resistant gene cassette at the BglII-site was performed. This was confirmed by positive selection on LB agar containing 50 μg/ml kanamycin, as well as 25 μg/ml ampicillin; and also by HindIII digestion of recombinant plasmids. The obtained plasmid was named pCRISPR01 ( Supplementary Fig. S1) and used throughout this study as the vector backbone for FnCpf1 and crRNA expressions.

Expression of codon-optimized eGFP and eGFP-Cpf1 in A. aculeatus TBRC 277
To test the expression and localization of the recombinant FnCpf1, the enhanced green fluorescent protein (eGFP) and FnCpf1-eGFP fusion, with attached SV40 NLS, were used as the gene reporter. These were constructed by ligation of each DNA fragment to the NcoI/ PmeI-site of pCRISPR01, yielding pCRISPR01-eGFP and pCRISPR01-FnCpf1-eGFP plasmids, respectively. The positive clones of each construct were confirmed by restriction analyses and further verified by sequencing. The obtained vectors, pCRISPR01-eGFP and pCRISPR01-FnCpf1-eGFP, were individually transformed into A. aculeatus TBRC 277 host. Positive transformants were further confirmed by colony PCR, followed by single spore isolation hereafter referred to as egfp and Cpf1-egfp accordingly. The expression of eGFP and Cpf1-eGFP was detected after culturing TBRC 277 transformants in PDB for 24 h. Western blot detected specific protein bands of approximately 28 kDa and 180 kDa as predicted for eGFP and Cpf1-eGFP, respectively (Supplementary Fig. S2), suggesting that codon-optimized FnCpf1 was expressed in A. aculeatus TBRC 277. Additionally, to investigate whether the SV40 nuclear localization signal targeted FnCpf1 to the nucleus, the transformants were further assessed using confocal microscopy. The result showed that both transformants, egfp and Cpf1-egfp, showed green fluorescence signal when compared to the WT, after overnight incubation in PDB medium with bleomycin. Furthermore, these recombinant fungi were also stained with 4,6-diamidino-2-phenylindole (DAPI) and showed that the majority of the eGFP signals were overlapped with nuclei, suggesting that FnCpf1 correctly localized to the nucleus of A. aculeatus TBRC 277 (Fig.  2b). Finally, to obtain the expression vector for further genome engineering in TBRC 277, we also cloned just FnCpf1-NLS without eGFP to generate pCRISPR01-FnCpf1 plasmid.
Construction of guide RNA (crRNA) CRISPR/Cpf1 system requires crRNA consisting of a 20bp direct repeat (DR) followed by a 20-23 bp protospacer to direct Cpf1 to cleave sequence-specific target [41]. In this study, pyrG, an essential gene for the biosynthesis of uridine, was chosen as the target gene. As pyrG mutants undergo auxotrophic selection, its inactivation can be observed by growing transformants in minimal medium containing 5-fluoroorotic acid (5-FOA) [42,43]. PyrG converts 5-FOA into fluoroorotidine monophosphate which is subsequently converted into fluorodeoxyuridine by ribonuclease reductase. Fluorodeoxyuridine is a pyrimidine analog that is rapidly converted to 5fluorouracil, a suicide inhibitor of the thymidylate synthase, and therefore inhibits nucleic acid synthesis, leading to cell death; however, 5-FOA is non-lethal in the absence of pyrG [9,44].
To perform this experiment, first, we needed to validate the sequence of A. aculateus TBRC 277 pyrG gene from the genomic draft sequence as generated by Next Generation Sequencing (NGS). The pyrG gene was successfully PCR amplified and sequenced (Fig. 3). The 915bp amplified sequence was identified to encode orotidine 5′-phosphate decarboxylase (pyrG); it showed 93.63% sequence identity to A. aculeatus ATCC 16872 and had a 81-bp putative intron at the positions 158-238. Secondly, since FnCpf1 could be programmed to edit multiple target sites [29], this aspect was tested by selecting three out of eight predicted protospacers with the following criteria: no off-targets and highest specificity score (100%). Three protospacer guide sequences were obtained targeting positions 59-bp (pyrG-1), 167-bp (pyrG-2), and 648-bp (pyrG-3) downstream of the pyrG start codon.
Next, due to the lack of well-identified RNA Pol III promoters in A. aculeatus, we instead employed a U3 Pol III promoter derived from A. fumigatus to generate a single crRNA construct, i.e. a 182-bp crRNA-flanking tRNA fusion transcript (GenBank KT031983.1) [21]. Using this approach, the endogenous tRNA processing system was expected to cleave both ends of the tRNA precursor and release a single crRNA with the aid of RNAses Z and P [35]. In this study, U3 Pol III promoter and its terminator, together with tRNA-sgRNA(Cas9)-tRNA motif originating from pFC902, were successfully amplified and inserted into pJET1.2B resulting in pJET-U3P sgRNA (Table 1). This plasmid served as a template for the subsequent PCR reactions to generate six DNA fragments, in which each of the associated fragments were assembled with BglII-cut pOK12 by Gibson Assembly (see Methods). The obtained tRNA-crRNA-tRNA Cpf1-associated cassettes in pOK12 specifically referred to as crRNA-tRNA-pyrG-1, crRNA-tRNA-pyrG-2, and crRNA-tRNA-pyrG-3 which corresponded to the sequence-specific protospacers pyrG-1, pyrG-2, and pyrG-3, respectively. To finalize the CRISPR/Cpf1 construction, each of the purified crRNA-pyrG cassettes successfully incorporated the BglII-cut pCRISPR01-FnCpf1 to obtain the plasmid constructs named pCRISPR01-FnCpf1-pyrG-1, pCRISPR01-FnCpf1-pyrG-  2, and pCRISPR01-FnCpf1-pyrG-3. All of these plasmids were verified by sequencing and by BglII digestions, which generated 16.8-kb and 0.9-kb DNA fragments representing the vector pCRISPR01-FnCpf1 and crRNA-pyrG inserts, respectively ( Supplementary Fig. S3).
CRISPR-crRNA complex induced pyrG-targeted gene disruption in A. aculeatus TBRC 277 This study relies on the ability of Cpf1/crRNA-complex targeting pyrG locus to induce DSB, followed by NHEJrepair mechanism. While some methods employ either Cas9 or Cpf1 expression in vivo but crRNA preparation in vitro which requires laborious steps for crRNA transcript preparation as well as an additional transformation step when introducing the crRNA. This leads to low transformation efficiencies due to crRNA stability issues [9,14]. We therefore opted for in vivo preparation both of crRNA and Cpf1 in a single plasmid (see Methods). To test whether heterologous expression of the CRISPR/Cpf1 system is able to induce pyrG-targeted DSB in A. aculeatus TBRC 277 chromosome, as well as to explore the general utility of RNA-guided genome editing in fungal cells, three pCRISPR01-FnCpf1-pyrGs plasmids expressing Cpf1 and three different crRNA guides were independently transformed into A. aculeatus TBRC 277 protoplasts. Next, pyrG-mutant selection was performed by a two-step selection by first randomly selecting bleomycin-resistant colonies (Fig. 4a) and reculture the colonies on new MM+Czapek-Dox plates with bleomycin and Uri/Ura supplementation, referred as the primary selection plate, but in the absence of sorbitol (Fig. 4b). Then, the colonies were grown on the secondary MM+Czapek-Dox plates supplemented with Uri/Ura as well as 5-FOA (Fig. 4c). The colonies that survived on this medium were considered mutant auxotroph [45]. For a negative control, the protoplast was transformed with empty vector or only Cpf1-bearing plasmid, and no colony was observed on the secondary selection plates (Fig. 4d). Our preliminary study showed that selective medium for protoplast-mediated transformation comprised of a mixture of sorbitol and 5-FOA gave rise to false-positives. Therefore, a two-step selection was required to remove fungal clones without the CRISPR plasmid. Using this approach, several bleomycin-resistant colonies obtained from the primary selection were not viable on secondary selective medium containing 5-FOA considered as non-mutated pyrG transformants. Additionally, we also tested the transformation efficiency of Cpf1-bearing plasmids (in pCRISPR01-FnCpf1) which was 30 colonies/μg of DNA ( Supplementary Fig.  S4). This was slightly lower than that of the empty vector without Cpf1 (37.5 colonies/μg of plasmid DNA). This may be due to size of the plasmid since the Cpf1bearing plasmid is approximately 4-5 kb larger. However, the effect of the constitutive Cpf1 expression from the AMA1-based plasmid may also contribute to the minor difference in transformation efficiency. Similar results were also reported in A. nidulans [36], and Ashbya gossypii [37] expressing LbCpf1. However, the potential toxicity effect of FnCpf1 was reported in yeast expressing multicopy plasmid [40].
To confirm whether the obtained colonies were auxotrophic mutants, PCR was performed to detect the presence of introduced exogenous CRISPR plasmid followed by sequencing of the pyrG gene. Using this approach, Fig. 4 An example of A. aculeatus TBRC 277 protoplasts transformed with CRISPR/Cpf1 plasmids. a pCRISPR01-FnCpf1-pyrGs (with crRNA specific targeting pyrG gene) on primary selective transformation medium containing 1.2 M sorbitol + 10 mM uri/ura + 50 μg/ml bleomycin, b a single colony selected from panel a, was then cultured on new minimal medium containing 10 mM uri/ura+ 50 μg/ml bleomycin, (c) secondary selection from panel b, re-cultured on minimal medium containing 10 mM uri/ura and 1.5 mg/ml 5-FOA. d pCRISPR01-FnCpf1 encoding only Cpf1 (without crRNA targeting pyrG), after growing on secondary selective medium containing 5-FOA, no colony was observed (control). Uri: uridine, Ura: uracil each randomly selected clones exhibited the presence of indels with 100% efficiency in every pyrG locus (Supplementary Table S1). Sequencing of the pyrG-specific locus revealed that three out of ten randomly selected mutant of each auxotroph contained pyrG indels (Fig. 5). These results indicated that Cpf1-crRNA complex induces DSB and endogenous error-prone repair that leads to indels of the pyrG gene in A. aculeatus TBRC 277.
Moreover, these DNA sequences were further analyzed. The results showed that mutants carrying either crRNA-pyrG-1 or crRNA-pyrG-2, typically had small insertions or deletions in the range of 2-6 bp and 7-16 bp, respectively. Interestingly, in a mutant carrying crRNA-pyrG-1, we noticed a 6-bp insertion in the exon of the pyrG-targeting cleavage site; this small in-frame insertion somehow could abolish PyrG activity. Additionally, for mutants carrying crRNA-pyrG-3, their deletions were in the range of 23-28 bp. These similar observations were also reported in other fungi [17,23], plant [46], and animal cell line [27]. However, large DNA insertions with sizes greater than 100-bp were also found in which polyketide synthase gene (pksP) was targeted in A. fumigatus [16].

Discussion
CRISPR-Cas based systems have been broadly applied for use in targeted genome modification of cells or organisms due to their programmable sequence specificity. More recently, the CRISPR-Cas based system is no longer limited to Cas9 from S. pyogenes. As an alternative to Cas9, two Cas nuclease genes originating from Acidominococcus and Lachnospiraceae that belong to the Cpf1 family have been reported [27]. The introduction of CRISPR/Cpf1 system provides another option to the CRISPR toolbox [47]. So far, applications of CRISPR-Cas in filamentous fungi have focused mostly on Cas9 [28,32,48]. Here, we presented data of another type of Cpf1, called FnCpf1, and demonstrated the NHEJ pathway involvement in the repair of Cpf1-induced DNA DSBs in a non-model filamentous fungus, A. aculeatus TBRC 277.
The success of genome editing by CRISPR/Cpf1 system requires the heterologous expression of the codonoptimized Cpf1 gene fused with a NLS (PKKKRKV) to efficiently target the fused protein into the cell nucleus.
crRNA together with FnCpf1 were expressed via a single vector in the host cell, hence, it is not necessary for in vitro crRNA preparation which is laborious and time consuming [12]. In this study, to increase the ability of A. fumigatus U3 Pol III promoter to drive the functional transcription of crRNA, we chose a tRNA-processing system to yield a mature crRNA transcript rather than introducing any additional RNases along with Cpf1/ crRNA components. This strategy was also employed for CRISPR/Cpf1 based multiplex genome editing applications in another filamentous fungi [21]. Here, we successfully produced the monocistronic tRNA-processing system and verified its ability to edit pyrG gene in the host TBRC 277 genome. Hence, this report demonstrated the successful in vivo expression and assembly of FnCpf1 and crRNA to form functional FnCpf1-crRNA ribonucleoprotein (RNP) complexes. These complexes serve as RNA-guided nucleases to specifically target and cleave genomic DNA. In addition, these results also suggested that a 20-bp protospacer sequence is sufficient in CRISPR/Cpf1 based system, in contrast to the previous reports using a 23-bp protospacer [29,36]. Our study therefore supports the previous report that the Cpf1 in complex with the crRNA and double-stranded DNA target was dictated by only the first 20-bp of heteroduplex formation of crRNA and target DNA [49].
We first developed CRISPR/Cpf1 vector specifically for FnCpf1 nuclease, as a proof of concept for gene editing in A. aculeatus species. To our knowledge, the use of FnCpf1 nuclease gene for targeted gene editing has not been reported in any Aspergillus species despite its successful applications in bacteria [50], yeast [40], plant [30], and animal cell lines [29]. In this study, we have shown that the CRISPR/Cpf1 system is able to sitespecific edit A. aculeatus chromosome at three different loci within the pyrG gene by constructing specific protospacer guides. However, for genes that are located in a region of low transcriptional activity this would probably reduce the editing efficiency and require optimization [23]. Although it is advisable to carry out off-target mutation analyses by PCR and sequencing, local BLAST analyses of each guide showed no off-targets. Interestingly, indels obtained from all pyrG mutants retain the PAM site since FnCpf1 cleaves target DNA at a distal position, and this report is in accordance with the previous studies in plants [46,51]. In contrast, Cas9-mediated NHEJ usually destroys the PAM site due to its proximity to the cleavage site thus preventing the possibility for future edits. Since we employed AMA1-based CRISPR/ Cpf1 expression vector for transient expression of both FnCpf1 and crRNAs, this would allow us to perform iterative genome editing. This iterative process would allow for targeted gene disruption of multiple targets as well as marker-free genome editing, thus greatly improving the engineering throughput. Some Cpf1 features described above make Cpf1 has several advantageous over Cas9 for gene editing applications even though in some cases both Cpf1 and Cas9 achieved comparable targeting efficiencies [41]. In addition, we envision that gene editing by CRISPR/Cpf1 could be efficiently used as a powerful tool and would enhance the range of possible target sites because of its unique PAM sequence, as well as simplicity in designing crRNA cassettes, when compared to the well-studied CRISPR/Cas9 system.

Conclusions
In summary, we have demonstrated the use of CRISPR/ Cpf1 mediated gene editing technology in the wild type strain of A. aculeatus TBRC 277. Using this system, the pyrG gene involved in uridine biosynthesis pathway was successfully edited to generate an auxotrophic strain. The use of FnCpf1 as an alternative class II (type V) nuclease was reported for the first time in A. aculeatus species. The CRISPR/Cpf1 system developed in this study highlights the feasibility of CRISPR/Cpf1 technology and could be envisioned to further increase the utility of the CRISPR/Cpf1 in facilitating strain improvements as well as functional genomics of filamentous fungi.

Plasmid constructions
All oligonucleotide primers used in this study are shown in Supplementary Table S2. All PCR reactions for cloning purposes were performed for 35 cycles with Phusion high-fidelity DNA polymerase (Thermo Fisher, USA) using touch-down PCR with annealing temperature starting from 65°C, with a decrement of 1°C per cycle for the first 10 cycles followed by a constant annealing temperature at 55°C up to 35 cycles. Standard reaction volumes were 50 μl including 1x Phusion HF or GC Buffer, 0.2 mM dNTPs, 0.5 μM primers (IDT, USA), 1 unit of Phusion DNA Polymerase, 10-50 ng of genomic DNA or plasmid as the template. All PCR products were purified using Qiaquick gel extraction and PCR clean-up kits (Qiagen, Germany) according to manufacturer's instruction. All expression vectors were assembled by standard molecular cloning or Gibson assembly methods [54].
All plasmids used in this study are listed in Table 1. A detailed schematic map of CRISPR/Cpf1 based expression vector construction is illustrated in Supplementary  Fig. S1. pFC333 was used to establish CRISPR/Cpf1based expression vectors [22]. Briefly, to create multiple cloning sites (MCS), a 421-bp of TEF1 terminator from pFC333 was amplified by using a pair of primers pFC333-TT-NcoI-F and pFC333-TT-BamHI-R. The amplified DNA fragment was digested by NcoI/BamHI and ligated to a 11.1-kb of NcoI/BamHI-cut pFC333 plasmid to generate expression vector backbone named pCRISPR0. Additionally, a 1.4-kb BamHI/BglII-cut kanamycin resistance gene cassette, amplified from pJET1.2B-Km by using KanR-Bgl2-F and KanR-R primer pairs, was ligated to BglII-cut pCRISPR0, generating pCRISPR01 vector (Supplementary Table S3). The presence of the intended MCS sequences flanked by Aspergillus nidulans TEF-1 promoter and terminator in the pCRISPR01 was verified by sequencing.
To detect the expression and localization of Cpf1, enhanced GFP (eGFP) gene was fused in-frame to Cpf1 at the C-terminal together with SV40 NLS at the 3′ end of the fused sequence to generate plasmid pCRISPR01-FnCpf1-eGFP (Supplementary Table S3, Fig. 2a). This was carried out by amplification of pUC57-FnCpf1 and pUC57-EGFP with a pair of primers FnCpf1-NcoI-F/ FnCpf1-Kpn-R and EGFP-NK-F/ EGFP-SV40-Pme-R, respectively, followed by restriction enzyme digestion (NcoI/KpnI for FnCpf1 and KpnI/PmeI for eGFP) and ligation. In addition to Cpf1 and Cpf1-EGFP fusion constructs, the expression vector bearing only eGFP gene was also employed as a control by amplification of eGFP in pUC57-EGFP using the same primers used for eGFP fusion construction, but the amplified PCR product was digested by NcoI/PmeI prior to ligation with plasmid pCRISPR01.
Protoplast-mediated transformation (PMT) was carried out as described previously [9]. Specifically, 10-μg of purified DNA in 20 μl (500 ng/μl) was mixed with 100 μl of protoplast suspension (10 7 -10 8 protoplast/ml) and incubated on ice for 50 min. Subsequently, 1.25 ml of PCT buffer (25% PEG 6000, 50 mM CaCl 2 , 10 mM Tris HCl pH 7.5) was added to the mixture followed by incubation at room temperature for additional 20 min. Next, 1.25 ml of STC buffer was added, followed by addition of 5 ml PDB in 1.2 M sorbitol. The solution was centrifuged at 3000 x g for 10 min and the supernatant was discarded. Next, 2 ml of STC buffer was added to the cell pellet, and gently mixed with 30 ml of preheat MM+Czapek-Dox soft agar medium at 45°C. Approximately 5-ml of mixture was overlayed on the high osmotic pressure MM+Czapek-Dox plates with 1.2 M sorbitol containing 50 μg/ml bleomycin (InvivoGen, USA) and Uri/Ura, followed by incubation at 30°C for 3-5 days. Fungal transformants were selected and purified to obtain single colonies onto new MM+Czapek-Dox containing 50 μg/ ml bleomycin and Uri/Ura. For pyrG mutant selections, fungal transformants were selected on MM+Czapek-Dox plates containing 5-FOA and Uri/Ura.

Western blotting
Expression of FnCpf1 in A. aculeatus TBRC 277 was also detected by western blot as described [16,59]. Specifically, total proteins from recombinant A. aculeatus mycelia was extracted by first inoculating 10 5 conidia in 50-ml PDB medium with bleomycin (100 μg/ml) and incubated for 24 h at 30°C with 250 rpm on a rotary shaker. Then, the mycelia were collected by filtration with Whatman Grade 41 filter paper (Sigma Aldrich, USA) followed by gently blotting with tissue paper to remove the remaining liquid media. The mycelia were ground into fine powders under liquid nitrogen with a mortar and pestle. Subsequently, approximately 200 mg of mycelia powders was added into an eppendorf tube containing 500 μl ice-cold extraction buffer (50 mM HEPES pH 7.4, 137 mM NaCl, 10% (v/v) glycerol 99.5%) with freshly added cOmplete mini protease inhibitor cocktail (Sigma-Aldrich, USA), followed by gently mixing for 5 min. The slurry was incubated on ice for additional 10 min, and the supernatant which contains a total protein was collected in a new tube after centrifugation at 14,000 x g for 15 min at 4°C. The protein quantity was determined by Bradford method [60] using protein assay dye reagent (Bio-Rad, USA).
Approximately 20 μg of crude protein extracts was loaded in each lane for a 10% SDS-PAGE [16,59]. After electrophoresis, proteins were transferred to a 5.5 × 8.5 cm nitrocellulose membrane with 0.2 μm pore size (Bio-Rad, USA) by using a Mini Trans-Blot Immersion (Bio-Rad) in 38.90 mM glycine, 47.88 mM Tris base, 0.0375% (w/v) SDS and 20% methanol at 90 V for 70 min. Then the membrane was soaked in phosphate buffer saline (PBS) for 5 min, then incubated in blocking buffer (PBS with 0.5% Tween 20 (PBST), 4% (w/v) non-fat dry milk) for 2 h at room temperature on shaking incubator with gentle agitation. Subsequently, the membrane was probed with anti-GFP mouse monoclonal antibody (Thermo Fisher, USA) or FnCpf1 mouse monoclonal antibody (Genscript, USA) in PBST buffer at 1:1000 dilution for 1 h. The membrane was washed with PBST buffer twice for 5 min each, then added with a 1:1000 dilution of anti-mouse IgG conjugated with alkaline phosphatase (Thermo Fisher, USA) in PBST buffer. The membrane was then washed twice with PBST buffer for 5 min each and developed using a 1-Step NBT/BCIP chromogenic substrate (Thermo Fisher, USA) by incubating the membrane for 5-15 min or until the desired color developed.  Table S2). The amplified fragment was purified and cloned into pJET1.2/B vector (Thermo Fisher, USA). Positive clones carrying DNA inserts were sequenced from both strands. Nucleotide sequence of pyrG obtained were pairwise-aligned to the sequence previously obtained by next-generation sequencing (NGS). The consensus sequence was submitted to GenBank under accession number MN364695 and used as template in selecting protospacer guide sequence ( Table 2).
Construction of pre-crRNA for pyrG gene editing in A. aculeatus TBRC 277 The protospacer guide sequences were designed based on the criteria used in S. cerevisiae [40]. Specifically, crRNA sequence targeting pyrG was first determined by searching for potential PAM sites (5′-TTTV-) together with a 20-23 bp protospacer sequence directly at the 3′end of the PAM motifs using Geneious Prime ver. 2019.2.1 software [20,61]. To ensure no off-targets, the obtained protospacer sequences were BLAST against A. aculeatus TBRC 277 genome sequence. The guide sequences of pyrG used in this study are shown in Table 2.

Verification of recombinant TBRC 277 and analysis of indels
Fungal genomic DNA of randomly selected mutants was purified by Phire plant direct PCR kit (Thermo Fisher, USA) according to manufacturer's instructions. For detection of the recombinant plasmids in TBRC 277, two pairs of primers (seq TEF-F/seq TEF-R or seq CRISPR sgRNA F-1/seq CRISPR sgRNA R-1) were employed to verify Cpf1 gene inserts or crRNA cassettes. For detection of the genotype of uridine auxotrophic mutants, a pair of 5′-169 bp up F3/3′-148 bp down R2 primers was used to amplify pyrG gene, followed by DNA sequencing to determine the presence of mutations (indels) within the targeted pyrG loci. The CRISPR efficiency was calculated by dividing the number of mutant colonies detected by sequencing and the total number of putative mutants selected for analysis [20].