CRISPR/Cas9-mediated genome editing directed by a 5S rRNA–tRNAGly hybrid promoter in the thermophilic filamentous fungus Humicola insolens

Background Humicola insolens is a filamentous fungus with high potential of producing neutral and heat- and alkali-resistant cellulase. However, the genetic engineering tools, particularly the genome-editing tool, are scarce, hindering the study of cellulase expression regulation in this organism. Results Herein, a CRISPR/Cas9 genome-editing system was established in H. insolens based on a hybrid 5S rRNA–tRNAGly promoter. This system is superior to the HDV (hepatitis delta virus) system in genome editing, allowing highly efficient single gene destruction in H. insolens with rates of deletion up to 84.1% (37/44). With this system, a putative pigment synthesis gene pks and the transcription factor xyr1 gene were disrupted with high efficiency. Moreover, the extracellular protein concentration and cellulase activity largely decreased when xyr1 was deleted, demonstrating for the first time that Xyr1 plays an important role in cellulase expression regulation. Conclusions The established CRISPR/Cas9 system is a powerful genetic operation tool for H. insolens, which will accelerate studies on the regulation mechanism of cellulase expression and engineering of H. insolens for higher cellulase production. Supplementary Information The online version contains supplementary material available at 10.1186/s13068-021-02057-y.

has stable activity at high temperature [2,3]. The high temperature-resistant β-glucosidase and xylanase expressed by H. insolens have been used in the wine industry for quality improvement, while its neutral cellulase has been used in the textile and washing industry [4]. By having excellent heat-and alkali-resistance, high cellulose degradation ability, and an optimal pH close to neutral, the cellulase expressed by H. insolens is a good complement to that from T. reesei [4][5][6][7]. However, the production level of H. insolens is low and cannot meet the need of biofuel industries. Previously, a T-DNA random insertional mutant library created by Agrobacterium tumefaciens-mediated transformation was established [8]. It was also discovered that mutation of the transcriptional regulator CreA did not greatly improve the ability of H. insolens to produce cellulase [9]. One main reason for the inefficiency of strain engineering is that there is no well-established genome-editing system, limiting the study of the regulating mechanisms of cellulase expression in H. insolens. There is an urgent need for a new technology to solve this problem.
With the advantages of high efficiency, versatility and ease of operation, the CRISPR/Cas9 technology is now widely used in functional genomics studies of filamentous fungi [10][11][12]. The essence of this technology is that, a small guide RNA (sgRNA) is designed to target and direct the Cas9 nuclease to bind and cleave a specific site in the chromosome [13]. sgRNA recognizes and complexes with the DNA in the targeting site. The complex is inserted into the gap between the nuclease recognition and cutting sites of Cas9, which will activate the cleavage activity of Cas9 and lead to cutting of the target site and forming a double-stranded break (DSB) [14,15]. DSB is repaired by either the nonhomologous end joining (NHEJ) or homology-directed repair (HDR) mechanisms [16].
Herein, the CRISPR/Cas9-based genome-editing technique was established for the first time in H. insolens. The technique contains a codon-optimized Cas9 nuclease-expressing cassette directed by a tef1 promoter, a sgRNA-expressing cassette with a prevailing tRNA Gly element directed by a 5S rRNA promoter, and a donor DNA fragment with 600-bp homologous arms. Using this system, we greatly improved the genome editing efficiency in H. insolens and successfully disrupted the pks pigment synthesis gene and xyr1 transcription factor gene, with a maximum efficiency of 84.1% (37/44) and 78.3% (18/23), respectively. This method displayed great potential for genome editing in H. insolens and laid a foundation for functional genomics as well as construction of engineered strains with improved ability to produce cellulase.

Construction of a CRISPR/Cas9 system for genome editing in H. insolens
The CRISPR/Cas9 system used in H. insolens included a Cas9-expressing cassette, a sgRNA expression cassette, and a donor DNA fragment (Fig. 1). Abundant Cas9 protein and sgRNA are well-known to be critical to successful genome editing. Since expression of Cas9 depends heavily on the promoter, the strong and constitutive promoter Ptef1 has been successfully applied in Aspergilli (the Aspergillus nidulans Ptef1) and M. thermophile (the Myceliophthora thermophila Ptef1) [11,17]. Therefore, the Cas9-expressing cassette containing the tef1 promoter, cas9 gene from Streptococcus pyogenes with two nuclear localization signals (NLS), and the trpC terminator was synthesized according to the sequence of Cas9-expressing cassette used in M. thermophila [17]. It was used herein to provide stable expression of Cas9 and proved to have a good effect on gene editing ( Fig. 1; Table 1).
For sgRNA expression in other fungi such as T. reesei, Neurospora crassa, Aspergillus fumigatus, Penicillium chrysogenum, and Pyricularia oryzae, typically an RNA polymerase III type U6 promoter was used to drive its transcription [12,[18][19][20][21]. However, it was sometimes difficult to identify a U6 promoter in many species and the use of a heterologous U6 promoter may otherwise reduce the efficiency of gene editing [22,23]. In addition to the U6 promoter, there are also other types of promoters successfully used to initiate sgRNA transcription. In Aspergillus niger and Fusarium fujikuroi, a highly conserved 5S rRNA promoter was found to drive sgRNA transcription with higher efficiency than that using the U6 promoter [24,25]. Furthermore, high gene editing efficiency can be obtained by using 5S rRNA along with its upstream 338bp sequence as the promoter in A. niger [25]. Four tRNA promoters were more efficient than the U6 promoter in Ustilaginoidea virens [26]. A small RNA, i.e., tRNA Gly , was regarded to be able to enhance transcription of the sgRNA as a potential enhancer of Pol III and also ensure precise release of sgRNA spacer-scaffold structure from the sgRNA expression cassette [27,28]. In addition to change of the promoter, in recent years, ribozymes (an RNA-based nuclease) were included in the sgRNA expression cassette, expanding the use of other types of promoters. In Aspergilli, an HDV ribozyme was fused in the sgRNA expression cassette to liberate more sgRNA [11,25]. Taken together, both the HDV ribozyme and the tRNA Gly element were used in sgRNA expression cassette to improve the genome editing efficiency in different filamentous fungi. Since the U6 promoter of H. insolens was not identified at present, two strategies employing either the HDV ribozyme or tRNA Gly element were used to improve sgRNA expression directed by the 5S rRNA (-338) promoter in this study (Fig. 1).
For the donor fragments, the length of homologous arms has a high impact on the efficiency of genome editing. It was reported that, when the homologous arms for lea1 were 600-bp (or more) in T. reesei, the genome editing efficiency reached 100% [10]. Therefore, we designed homologous arms with 600-bp and fused them with the marker gene-expressing cassette Pgpd-hyg-Tgpd (Fig. 1).
In the wild-type Y1 strain, the numbers of transformation did not change significantly due to the addition of Cas9 expression cassette (617 transformants for donor DNA/Cas9 cotransformation versus 625 transformants for donor DNA only). Additionally, the controls without donor DNA using Cas9 expression cassette and 5SsgRNA-HDV (abbreviation for the 5SrRNA-sgRNA-HDV construct), 5SsgRNA (abbreviation for the 5SrRNA-sgRNA construct), tRNA Gly -sgRNA systems all displayed low disruption efficiency for 7.0% (26/371), 4.1% (14/345) and 9.3% (35/375), respectively (Table 1). Apparently, the efficiency of genome editing using the 5StRNA Gly -sgRNA system almost doubled that of 5SHDV-sgRNA (Table 1). This is worth noting since in A. niger, addition of the HDV ribozyme improves the genome editing efficiency [25]. Two reasons may account for the high efficiency of 5StRNA Gly -sgRNA system in H. insolens. For one thing, tRNA Gly can recruit the PolIII complex due to the existence of internal promoter elements box A and B and drive transcription of sgRNA by serving as a promoter [40]. For another, the structure of pre-tRNA Gly can be recognized by RNase P and RNase Z and the site between tRNA Gly and sgRNA spacer is efficiently cleaved [41]. These lead to releasing of active guide RNA molecules and explain the higher cleavage activity than observed for HDV.
In cells, the broken DNA double strands are repaired through either the nonhomologous end joining (NHEJ) pathway or homology-directed repair (HDR) pathway [42]. The NHEJ pathway is rapid and does not require a template [43]. In fungi, the Ku proteins play a key role in the NHEJ pathway for DNA repair. When the DNA Fig. 2 Mutations into the pks gene in H. insolens as introduced by CRISPR/Cas9-mediated genome editing. A Disruption of pks was correlated to an albino conidia phenotype in H. insolens. The pks gene disrupted strains displayed pigment-less conidia and resultant white colonies as indicated by red arrows. B Sequence alignment of the pks gene in WT and the pks-edited mutants. The 20-nt protospacer sequence of pks in wild-type strain is represented by blue letters and PAM sequence is labeled by red letters with underline strands are broken, the Ku70/Ku80 heterodimer recognizes and binds to the broken ends and recruits kinase, ligase, and other associated proteins for DNA repairing [16,44]. In filamentous fungi, it has been reported that disrupting the gene encoding Ku70 can significantly improve the efficiency of homologous recombination [21,[45][46][47]. Therefore, we used the Y1/Δku70 strain to further test CRISPR/Cas9-mediated genome editing. In this strain, the gene-editing efficiency for pks using the 5StRNA Gly -sgRNA system increased from 56.2% (204/363, in the WT strain) to 84.1% (37/44), tripling that of 5SHDV-sgRNA (25.6%, 10/39, Table 1). Disruption of xyr1 was all carried out by supplementing with the donor DNA (Fig. 3A) and this gene was disrupted at an efficiency of 78.1% (18/23) with 5StRNA Gly -sgRNA system but at 26.1% (6/23) with the 5SHDV-sgRNA system (Table 1). This again demonstrated that the using tRNA Gly was more efficient in genome editing than HDV. Disruption of xyr1 by homologous recombination was verified by diagnostic PCR: a 4824-bp specific fragment was amplified by PCR from the xyr1 disruption mutant, while a 1920-bp fragment was amplified from the wildtype Y1 and Y1/Δku70 strains (Fig. 3B).
When the Y1/Δku70 strain was used as the host strain, the genome editing efficiency of the 5SHDV-sgRNA system was comparable for pks. In Y1/Δku70 the knockout efficiency was 25.6% (10/39), while in Y1 the efficiency was 26.6% (91/342) ( Table 1). With this strain, the increase of genome editing efficiency of the 5StRNA Gly -sgRNA system was obvious: 56.2% (204/363) in Y1 versus 84.1% (37/44) in Y1/Δku70. However, the efficiency still did not reach to nearly 100% as observed in Magnaporthe grisea and A. nidulans [48,49]. Note that the genome editing efficiency of 11.0% (8/73) in Y1/Δku70 did not change significantly as compared with 6.3% (25/400) in Y1 when only the donor DNA for pks1 was used. There might be several reasons explaining the limited increase of genome editing efficiency in the ku70 disrupted strain. First, the expression of ku70 is so low that knockout of this gene does not have significant impact on the repair rate of NHEJ [50]. Second, the expression of HDR-related component genes is too low that the homology recombination efficiency cannot be improved even after the NHEJ pathway is repressed [51].
We also used the 5StRNA Gly -sgRNA system in the Y1/Δku70 strain for multiplexing genome editing. In the trial to simultaneously destruct two genes, the successful rate of deleting pks and xyr1 dropped to 17.4% (Table 1). These results revealed that this system could be used for editing the cellulase expression-related genes in H. insolens.

Deletion of xyr1 decreased cellulase production in H. insolens
Xyr1 is the main transcription activator of cellulase expression in mainly lignocellulose-degrading filamentous fungi [52,53]. However, the regulatory role of Xyr1 on fungal growth and cellulase expression is not consistent in all filamentous fungi due to its phosphorylation degree [54,55]. In H. insolens, the role of xyr1 in cellulase expression regulation is not clear. Therefore, in this study, the xyr1 mutant of H. insolens obtained by CRISPR/ Cas9-mediated gene disruption was further studied for growth phenotype and cellulase production.
On culturing at 42 °C in PDA, MMN, and YPD solid plates, the nascent hyphae were dense and clustered in the center of the colonies. At 24 h, the colony of Y1/Δku70Δxyr1 strain was significantly smaller than that of Y1 and Y1/Δku70 strains. This was also true for colonies grown at 72 h on these plates (Fig. 4A-C). When cultured in PDA, the sporulation of Y1/Δku70Δxyr1 was basically the same as those of wild-type and the Y1/Δku70 strains (Fig. 4A). Although the sporulation of Y1/Δku70Δxyr1 strain was slower, the length of its aerial hyphae appeared to increase (Fig. 4B) in the MMN medium. In YPD, Y1/Δku70Δxyr1 showed a layered state of mycelial extension in addition to aerial hyphae (Fig. 4C). Deletion of xyr1 and its homologous genes in other filamentous fungi such as T. reesei, M. thermophila and Aspergilli similarly altered the growth phenotypes depending on the carbon sources [55][56][57]. In flask fermentation, the wild-type Y1 and Y1/Δku70 strains produced almost the same amount of extracellular enzymes. However, both the cellulase and hemicellulase activities of the Y1/Δku70Δxyr1 were lower than those of the wild type (Fig. 5). The FPase activity of Y1/Δku70Δxyr1 reached the maximum on day 6 postinduction, 63.95% lower than that of WT (Fig. 5A). The enzymatic activities of endoglucanase, cellobiohydrolase, β-glucosidase, and xylanase of Y1/Δku70Δxyr1 were also significantly decreased to as low as 32.10%, 70.55%, 68.16%, and 59.45%, respectively, of the original strain ( Fig. 5A-E). The biomass of Y1/Δku70Δxyr1 strain was slightly lower than that of Y1 throughout the culture. Y1/Δku70 was similar to Y1/Δku70Δxyr1 till day 4 postinduction, but then it grew to a level similar to that of the wild type (Fig. 5F). For Y1/Δku70Δxyr1, the highest FPase/biomass ratio was 0.34 U/mg, which was much lower than those of Y1 (0.80 U/mg) and Y1/Δku70 (0.74 U/mg) (Fig. 5G).
An SDS-PAGE analysis indicated that, compared with Y1 and Y1/Δku70 strains, the major extracellular proteins with molecular masses lower than 70 kDa in Y1/Δku70Δxyr1 almost completely disappeared (Fig. 6). Taken together, the decrease of enzyme activity in H. insolens Y1/Δku70Δxyr1 was due to lowered ability to express the enzymes as well as the impaired biomass accumulation. These results were also consistent with the results of other filamentous fungi [56,58], and for the first time, demonstrated that Xyr1 played a positive pleiotropic regulatory role in cellulase expression of H. insolens.
The widespread Xyr1 and its homologs can activate transcription of cellulase and/or hemicellulase genes in many lignocellulose-degrading filamentous fungi [59]. It is well-known that, after xyr1 deletion, almost all cellulase and hemicellulase genes are unable to be induced in T. reesei [60]. However, deletion of xlnR (encoding a Xyr1 homolog) in Fusarium graminearum led to elevation of cellulase gene transcription [61]. The discrepancy in transcriptional regulation by Xyr1 is that the function of this transcription factor is mainly determined by its phosphorylation status, but also impacted by the interaction between Xyr1 and other transcription regulators [62][63][64].

Conclusion
In this study, an efficient CRISPR/Cas9 genome editing platform for H. insolens has been successfully developed for the first time, employing the 5S rRNA promoter and tRNA Gly in sgRNA synthesis. This system proved to be highly effective when pks and xyr1 were used as two model target genes. The CRISPR-Cas9 system provides a technical platform for further study of the regulation mechanism of cellulase expression in H. insolens, which enables us to study the function of other transcriptional regulators and cellulase genes. It is expected to aid in promoting studies on regulation mechanisms of cellulase expression and engineering industrial strains with improved cellulase-producing ability.

Strains and culture media
The H. insolens Y1 (CGMCC 4573) and its engineered strains were cultured on potato dextrose agar (PDA) plates at 42 °C for 5 days for conidiation. The yeast extract-peptone-dextrose medium (YPD) was used for mycelia growth at 42 °C. For cellulase production in flask fermentation, H. insolens were cultured at 42 °C for 6 days in a modified Melin-Norkrans medium (MMN) (containing 1 g/L tryptone, 20 g/L yeast extract, 0.6 g/L MgSO 4 ·7H 2 O and 20 g/L Avicel). For observation of colony phenotypes, 2 × 10 5 spores each of the H. insolens wild-type strain and its mutants were spotted and cultured on MMN, PDA, or YPD plates for 3 days at 42 °C. The Escherichia coli Top 10 (GenStar, Beijing, China) was grown at 37 °C for plasmid propagation in a Luria-Bertani (LB) broth supplemented with 100 μg/mL of ampicillin when necessary.

Plasmid construction
The expressing cassette Ptef1-NLS-cas9-NLS-TtrpC containing the tef1 promoter, a codon-optimized cas9 gene with two nuclear localization signals (NLS), and the trpC terminator was synthesized according to the sequence of Cas9-expressing cassette used in M. thermophila as described earlier [17] for expression in H. insolens.

Protoplast transformation of H. insolens
Both the H. insolens wild-type Y1 and the Y1/Δku70 strain were used as hosts in this study. Protoplast transformation of H. insolens was carried out as described earlier [66] with slight modifications. H. insolens strains were cultured on PDA medium at 42 °C for 3 days, and 10 7 spores were harvested and transferred to the YPD medium for a continued culture of 10 h. Lysing enzymes (5 mg/mL) from Trichoderma harzianum (Sigma, L-1412) was used for releasing protoplasts from mycelia. The PDA medium supplemented with 50 µg/ml of Hygromycin B and 0.44 M of sucrose was used to screen for successful transformants.

Assay of enzyme activity and SDS-PAGE
The mycelia of H. insolens Y1 and mutant strains were individually cultured in YPD and then transferred to MMN for cellulase induction [9]. The cellulase activities including the overall cellulase activity (FPase, using filter paper as the substrate), the endoglucanase (or carboxymethyl cellulose activity, using carboxylmethyl cellulose as the substrate), cellobiohydralase (using p-nitrophenylβ-cellobioside as the substrate), β-glucosidase (using p-nitrophenyl-β-glucopyranoside as the substrate), and xylanase activity (using birchwood xylan as the substrate) in the culture supernatant were determined according to the methods described before [67]. The extracellular proteins were resolved by SDS-PAGE on 12% (w/v) polyacrylamide gels. Proteins were visualized by staining with Coomassie Brilliant Blue G-250.