Using CRISPR-Cas9 and Cas12a for precise gene editing and large DNA deletions in Aspergillus niger

Escherichia coli strain DH5α was used as the host for routine cloning procedures. A. niger wild-type strain ATCC 11414 from the American Type Culture Collection (Rockville, MD, 20852) was cultivated on complete medium (CM) plates at 30°C for culture maintenance and spore preparation. Cultures on CM agar plates were incubated at 30°C for 4 days, followed by harvesting of spores through washing with sterile 0.4% Tween 80. Spore counting was performed using a hemocytometer. Both CM and minimal medium (MM) were prepared according to the method described by Bennett and Lasure (Bennett and Lasure, 1991).

A total of 18 vectors were utilized in this study (Table 1). The pFC332 vector (Plasmid #87845), obtained from Addgene, served as an empty vector for cloning purposes. Initially, pFC332 was linearized using the PacI restriction enzyme. Subsequently, the pGY5 vector was constructed by assembling the linearized pFC332 with a gBlocks Gene Fragment 7_gblocks via HiFi DNA assembly. Similarly, the pGY6 vector was generated by assembling the linearized pFC332 with two gBlocks Gene Fragments, 16_gblocks and 17_gblocks, using HiFi DNA assembly. pGY18 was assembled by combining the linearized pFC332 (via BsaI) with three PCR products, namely, PCR8768, PCR6926, and PCR7086, using HiFi DNA assembly. pGY53 was constructed by combining the linearized pGY18 (via BsaI and BamHI) with two PCR products (PCR162216 and PCR165186), along with three gBlocks Gene Fragments (77_gblocks, 78_gblocks, and 79_gblocks), using HiFi DNA assembly. pGY54 was created by combining the linearized pGY53 (via BsaI) with a gBlocks Gene Fragment 80_gblocks using NEBridge Golden Gate Assembly Kit (BsaI-HF v2). Similarly, pGY55 was assembled by combining the linearized pGY53 (via BsaI) with two gBlocks Gene Fragments, 81_gblocks and 82_gblocks, also utilizing Golden Gate assembly. pGY42, pGY56, pGY76, pGY77, pGY108, pGY109, and pGY177 were all generated by ligating linearized pGY18 (via BsaI) with their respective gBlocks Gene Fragments using Golden Gate assembly. Similarly, pGY158, pGY175, pGY176, and pGY178 were created by ligating linearized pGY53 (via BsaI) with their corresponding gBlocks Gene Fragments using Golden Gate assembly. Q5 Master Mix was employed for all cloning PCR according to the manufacturer’s instructions. All cloning enzymes and restriction enzymes utilized in this study were sourced from New England Biolabs (Ipswich, MA, United States). All plasmids were verified by Sanger sequencing. Additionally, all gBlocks Gene Fragments were acquired from IDT (Supplementary Table S1). For experimental vectors, the empty vector for Cas9, designated as pGY18, and the empty vector for Cas12a, known as pGY53, will be made accessible through Addgene.

All gRNAs were designed using the online tool CHOPCHOP (Labun et al., 2019). A total of 35 gRNAs were utilized in this study (Table 2).

A standard polyethylene glycol (PEG)-mediated protoplast transformation protocol was employed in A. niger as previously described. About 1-2 µg of plasmid in a volume of 10 µL or less was added to 100 µL of protoplasts in 15 mL centrifuge tube, which was gently mixed by tapping the tube and incubated on ice for 15 min. Next, 1 mL of 40% PEG was added, gently mixed by taping the tube, and kept at room temperature for 15 min. Following this, 5 mL of minimum medium containing 1 M Sorbitol was added and the tube was placed horizontally in the incubator shaker and shaken gently (∼80 RPM) at 30°C for 1 h. Subsequently, the protoplasts were pelleted down by centrifugation at 800 g and 4°C for 5 min. The pelleted protoplasts were resuspended in 15 mL of minimal medium with 1 M sorbitol and 0.8% agar. A 300 μg/mL of hygromycin or geneticin (G418) were used for antibiotic selection in A. niger. Finally, the resuspended cells with proper antibiotics were poured onto a petri dish and culture at 30°C for 1∼2 days. Three biological replicates were performed for all vectors except pGY177 and pGY178.

In general, the transformation plate becomes fully covered with fungal hyphae and spores 7 days post-transformation, making single colony isolation impossible at that stage. To quantify the CRISPR editing efficiency, single colonies were randomly selected and isolated under a dissection microscope (Leica MZ16) 19–24 h post-transformation, when the tiny colonies are not visible to the naked eye but can be identified under the microscope (). Each colony was transferred to a slant containing 1.5 mL of MM medium supplemented with 0.8% agar and 200 μg/mL hygromycin or G418. The slants were then cultured at 30°C until spores were observed. Supplementary Figure S1

To isolate the white spores from slants containing both white and black spores, a syringe was employed under a dissection microscope to carefully collect the white spores. Subsequently, these collected white spores were inoculated onto fresh CM slants to undergo purification for further analysis and experimentation.

The spores were collected by rinsing with sterile 0.4% Tween 80. Subsequently, these harvested spores were directly utilized for Squash-PCR genotyping following previously established procedures (Yuan et al., 2023). Genotyping PCR was performed using GoTaq Green Master Mixes (Promega). For each reaction, 1 μL of squashed spore solution was used as a template in a total PCR reaction volume of 20 μL. The PCR cycling conditions were as follows: initial denaturation at 95°C for 2 min, followed by 35 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 1 min, with a final elongation step at 72°C for 5 min. All DNA oligos used for genotyping PCR are listed in Supplementary Table S2.

PCR products were purified using the QIAquick gel extraction kit from QIAGEN N.V. (Venlo, Netherlands) and subsequently sequenced by GENEWIZ (South Plainfield, NJ, United States). The DNA oligos used for genotyping PCR are additionally employed for Sanger sequencing.

CRISPR genome editing tools are capable of editing the genome at both single and multiple locations. To enhance the efficiency of CRISPR gene editing systems, we utilized a tRNA-based gRNA expression strategy to boost the expression of multiple gRNAs. In this method, individual gRNAs are released by the cellular endogenous tRNA processing machinery, a mechanism that has been demonstrated to work efficiently across multiple species (Čermák et al., 2017; Song et al., 2019; Zhang et al., 2019). Since both the Cas9 and Cas12a systems have been established in A. niger, our objective is to compare the editing efficiency of these two systems by targeting the albA gene. This gene is a classical reporter gene with white A. niger spores as marker. (Vanegas et al., 2019; Li et al., 2021). Additionally, we aim to characterize single gene editing mediated by both single gRNA and double gRNAs to compare editing efficiency and provide practical guidance for their use. The tef1 promoter and its terminator were employed to control the expression of Cas9 or Cas12a in all vectors, while a U3 promoter and its terminator derived from Aspergillus fumigatus were utilized to drive the expression of the tRNA-gRNA array (Figure 1A). Recently, it has been demonstrated that among the endogenous tRNAs in A. niger, tRNA^Ala, tRNA^Phe, tRNA^Arg, tRNA^Ile, and tRNA^Leu are particularly effective in enhancing gRNA release compared to others (Li et al., 2021). Therefore, we applied these tRNA sequences to boost the expression and release of gRNAs in both Cas9 and Cas12a systems. For the Cas9 systems, single and double gRNAs targeting the coding region of albA were integrated into vectors pGY5 and pGY6, respectively, as described in Method 2.2 (Figure 1A). However, the original empty vector of Cas9 (pFC332) lacks the U3 promoter and terminator necessary for the expression of gRNAs. Additionally, its lack of suitable restriction enzyme sites makes ligation with multiple gRNA fragments inconvenient. Therefore, we modified it by incorporating the U3 promoter and terminator. Furthermore, we adapted this vector to pGY18 for BsaI-based Golden Gate assembly, allowing efficient ligation of multiple gRNA fragments, as demonstrated by its capability to tandemly assemble 1 to 16 gRNAs in a one-pot reaction (Yuan et al., 2022). By replacing Cas9 with Cas12a in pGY18, we further engineered a novel Cas12a vector, pGY53, designed for the ligation of multiple gRNA fragments through Golden Gate assembly. Using pGY53 as the backbone, we generated vectors pGY54 and pGY55, which contain single and double gRNAs, respectively, following the procedures described in Method 2.2 (Figure 1A).

Next, we conducted protoplast transformation as described in Method 2.4 to transform vectors pGY5, pGY6, pGY54, and pGY55 into A. niger strain 11414, with pFC332 and pGY53 serving as the empty vector controls. Autoclaved Milli-Q water was used as a negative control, and no colonies were observed, indicating that the plasmid transformation efficiency was nearly 100%. Due to the high transformation efficiency, tens to hundreds of colonies were observed after transformation, resulting in the entire plate being covered with hyphae and spores (Figure 1B). Pure black spores were observed on the plates of pFC332 (Cas9-only) and pGY 53 (Cas12a-only), while large quantities of white spores were present on all plates of pGY5, pGY6, pGY54, and pGY55, indicating successful genome editing events (Figure 1B; Supplementary Figure S2). More specifically, the abundance of white spores on plates of pGY6, pGY54, and pGY55 is noticeably higher than that of pGY5, suggesting a significant increase in editing efficiency. To quantify the editing percentage, we performed single colony isolation following the procedure outlined in Method 2.5. After single colony isolation, three distinct phenotypes were observed: those containing pure black spores, pure white spores, and mixed black and white spores (Figure 1C; Supplementary Figure S3). It was evident that slants containing pure white spores and mixed spores represented genome-edited events. After quantification, we observed average editing efficiencies of 31.7% for Cas9 and 86.5% for Cas12a when a single gRNA was used to target the albA gene (Figures 1D, E). Notably, both efficiencies increased to up to 100% when double gRNAs were employed, with average editing efficiencies of 96.7% for Cas9 and 93.3% for Cas12a (Figures 1D, E). It should be noted that no events of pure white spores were observed in the Cas9 systems with a single gRNA. In contrast, white spores were observed in an average of 6.7%, 25.6%, and 48.5% of isolated colonies in the Cas9 system with double gRNAs, and in the Cas12a system with single and double gRNAs, respectively (Figure 1D; Supplementary Figure S3). This difference underscores the distinct phenotypic outcomes associated with Cas9 and Cas12a editing systems. To ensure the purity of editing events, white spores were isolated according to the procedure outlined in Method 2.6. Following this, we performed Squash-PCR genotyping and subsequent Sanger sequencing to validate the purified editing events, providing further confirmation of the desired genomic modifications. In the genotyping PCR, three selected events of pGY5 exhibited bands similar to the positive control, while 4 out of 5 tested events of pGY6 displayed lower bands compared to the wild-type band, indicating the occurrence of deletions in the targeted region (Figure 2A). Notably, in event #14 of pGY6, double bands were observed, indicating the mixture of both wild-type and mutant. The PCR genotyping was thoroughly validated by Sanger sequencing. For example, small deletions of 1 bp and 18 bp were observed in the editing events of pGY5, explaining that there were no obvious size differences between the wild type and mutants due to these small deletions (Figures 2A, B). In contrast, larger deletions ranging from 90 to 125 bp were precisely detected between the two applied gRNAs, corroborating the lower mutant bands observed in the PCR genotyping (Figures 2A, B). In the testing of Cas12a system, five editing events from pGY54 and pGY55, respectively were selected for PCR genotyping. Surprisingly, six out of the ten tested events failed to yield any PCR products. Notably, the same DNA templates successfully amplified products for another gene, ruling out low-quality DNA as the cause of amplification failure. We then designed new primer pairs to amplify a larger flanking region, but still no PCR products were obtained, suggesting that the deletion extends beyond the amplified region. Large deletions, such as 750 bp and 127 bp, were observed in event #6 of pGY54 and event #5 of pGY55, respectively (Figure 2B). This observation is completely consistent with the mutant bands detected in the genotyping PCR (Figure 2A). Interestingly, a 17 bp deletion and a 4 bp deletion were detected in event #4 of pGY55, corresponding to the sites targeted by the first and second gRNAs, respectively (Figure 2B). However, the sequence between the two gRNAs was not removed, indicating that the editing of these two sites likely did not occur simultaneously.

Overall, both Cas9 and Cas12a demonstrated exceptionally high efficiency for single gene editing, particularly when double gRNAs were applied in A. niger. Cas12a exhibited potential superiority over the Cas9 system, as it showed significantly higher editing efficiency than Cas9 when only one gRNA was used for gene editing. Furthermore, achieving pure mutants seemed to be considerably easier with Cas12a after transformation, thereby facilitating the single spore isolation required for the strain production test.

To evaluate the capacity for targeted CRISPR-mediated large-scale genomic deletions in A. niger, we selected both the upstream and downstream sequences of the albA gene as the target region (Figure 3A). Multiple regions were selected according to the anticipated size of deletions, ranging incrementally from smaller to larger: 3.5 kb, 10 kb, 20 kb, 30 kb, and 40 kb, and diversified the selection to include a broad array of deletion sizes within this range (Figure 3A). To eliminate a long DNA sequence, a double-strand break (DSB) needed to be induced both upstream and downstream of the target region. Therefore, we designed gRNA1 and gRNA2 to target the upstream and downstream regions of the target locus, respectively, with high targeting efficiency as predicted by CHOPCHOP (Labun et al., 2019) (Figure 3B). In case these gRNAs did not exhibit sufficient efficiency, we additionally designed gRNA3 and gRNA4 proximal to gRNA1 and gRNA2, respectively, to ensure comprehensive coverage. Technically, target regions exceeding 10 kb pose challenges for straightforward amplification via standard PCR methods. Upon complete removal of the target region, amplifying the edited segment would be facilitated due to shorter amplification lengths with external primer pairs, while the internal sequence of the target regions should remain undetectable by PCR when using internal primer pairs (Figure 3B). Based on above principles, multiple gRNAs were designed for both Cas9 and Cas12a systems. For instance, four gRNAs were utilized for vectors pGY42, pGY56, pGY76, pGY77, pGY175, and pGY176; three gRNAs were employed for vector pGY109, while two gRNAs were utilized for vectors pGY108 and pGY158 (Figures 3A, C).

Following protoplast transformation, all plates containing nine different CRISPR constructs exhibited white spores, contrasting with the black spores observed on the wild-type (WT) plate, indicating successful elimination of the target region containing the albA gene (Figure 4; Supplementary Figure S4). The quantity of white spores significantly varies across the plates, with notably higher abundance observed in those of pGY42, pGY175, and pGY176 compared to others, suggesting a higher editing efficiency in these plates (Figure 4; Supplementary Figure S4). Due to the dense coverage of hyphae and spores on the plates hindering direct editing percentage calculation, we conducted single colony isolation for pGY42, pGY56, pGY76, pGY77, pGY108, and pGY175 at the early post-transformation stage, following Method 2.5. We observed similar phenotypes across plates, including those with pure black spores, pure white spores, and mixed black and white spores, as illustrated in Figure 1C; Supplementary Figure S5. Following quantification, we determined that the average editing percentage of all constructs ranged from 27.8% to 69.1% (Figures 3D, E). Notably, three constructs demonstrated editing efficiency exceeding 50%, with average percentages of 69.1%, 53.7%, and 60.4% for pGY42, pGY56, and pGY175, respectively (Figures 3D, E).

After isolation of the white spores (Method 2.6), PCR genotyping and Sanger sequencing were employed to validate the large fragment deletions and determine the precise sequence (Figure 5). As depicted in Figure 3B, external primers targeting various regions were utilized to amplify the full length of the target region. Concurrently, identical internal primers were used to amplify the internal sequence of the target region (located within the albA gene) for all constructs, except for pGY42, where the target region is situated within the albA gene itself. For instance, external primers 211_F/212_R were specifically designed to amplify the entire target region of pGY42, which spans 4,924 bp in length. Three edited events, #5, #14, and #15, yielded PCR products ranging in size from approximately 1.3–1.6 kb (Figure 5A). This indicates that fragments ranging from approximately 3.3–3.6 kb in length were deleted from the target region. The external primers 224_F/225_R were used to amplify the complete target region of pGY56, spanning a length of 12,455 bp. However, events #3 and #5 produced 1 kb PCR products only, and the internal region (albA gene) was not detected using internal primers 211_F/214_R, indicating the elimination of a sequence 11 kb in length (Figure 5A). Indeed, the sequence spanning 11,506 bp between the two gRNAs was deleted in event #3, as confirmed by Sanger sequencing (Figure 5B). Using similar strategies, we investigated all the rest of the constructs. Due to variations in the target region, the full length of target regions ranged from approximately 21 kb–42 kb. However, all the tested editing events yielded PCR products of less than 1.5 kb, and no obvious internal regions were detected (Figure 5A). This finding indicates deletions occurring within a range from as small as 20 kb to as large as 40 kb in these events. Interestingly, the deleted sequences and their sizes were accurately determined by Sanger sequencing, perfectly matching with the corresponding PCR genotyping results. For example, a 21,100 bp deletion was identified in event #1 of pGY76, a 22,157 bp deletion in event #2 of pGY77, a 31,311 bp deletion in event #1 of pGY108, a 40,489 bp deletion in event #2 of pGY109, a 32,141 bp deletion in event #2 of pGY158, a 20,750 bp deletion in event #19 of pGY175, and a 21,873 bp deletion in event #11 of pGY176 (Figures 5A, B). In addition, for those constructs employing two gRNAs to target upstream or downstream sequences, we observed deletions occurring at different locations targeted by the respective gRNAs, as demonstrated in constructs pGY76 and pGY176 (Figure 5B). Taken together, our findings demonstrate that large-scale genomic sequences can be efficiently deleted using CRISPR-Cas9 or Cas12a systems, with deletions as large as 40.5 kb achieved in A. niger.

Although we demonstrated deletion of large chromosomal segments up to approximately 40.5 kb within the vicinity of albA, we were unable to isolate larger deletions from 50 to 100 kb due to the presence of genes likely to be essential for growth and sporulation. When extending the study to larger regions, the transformation plates yielded no typical white spores with these constructs. To determine the capacity of this technique, we chose an alternative genomic region containing a NRPS (nonribosomal peptide synthase) gene cluster, which involves in synthesis of siderophores such as desmalonichrome and is presumed to be transcriptionally silent under certain cultivation conditions (Mo et al., 2016; Kwon et al., 2021). The NRPS cluster spans a total size of 97,809 bp. To target this region, we designed two vectors: Cas9 vector pGY177 and Cas12a vector pGY178, intended to induce deletions of 102.3 kb and 101.8 kb, respectively (Figure 6A). Following protoplast transformation, 20 colonies were randomly chosen from the transformation plate and transferred for single colony isolation of pGY177 and pGY178, respectively. As expected, no obvious phenotype was observed in isolated single colonies compared with the control. PCR genotyping indicated that four tested events from pGY177 exhibited PCR amplification, resulting in an editing efficiency of 20%, while only one event showed PCR amplification from pGY178, with an editing efficiency of 5% (Figures 6B, C). Sanger sequencing confirmed the large fragment deletions ranging in size from 102,228 to 102,249 bp in four edited events from pGY177 (Figure 6D). Interestingly, in addition to the large deletion, a 149 bp insertion was detected in event #1, and a 4 bp insertion was found in event #16. Additionally, a large sequence spanning 101,813 bp was excised via the Cas12a system in event #19 (Figure 6D). Taken together, these findings indicate that large genomic fragment deletions exceeding 100 kb can be efficiently achieved through CRISPR systems in A. niger.

In synthetic biology and metabolic engineering, removing the engineered plasmid is essential for enabling iterative genome editing. Since AMA1-based plasmids are reportedly cured from transformants through continuous subculturing on a nonselective medium, we used AMA1 as the origin of replication in all the CRISPR vectors used in this study (Aleksenko and Clutterbuck, 1997). Next, we explored the steps needed to eliminate the plasmids from cells after gene editing using an albA mutant. We selected a mixed colony phenotype containing both white and black spores (Figure 7) and inoculated 1 µL of spores onto Plate 1 (complete medium, Hyg −) and Plate 2 (complete medium supplemented with 300 μg/mL hygromycin, Hyg +) for the initial inoculation (day 1). On day 2, an agar block with hyphae from Plate 1 was transferred to Plate 3 (Hyg −) and Plate 4 (Hyg +) for a second inoculation. This process was repeated on day 3 and day 4, yielding Plates 5 and 7 (Hyg −) and Plates 6 and 8 (Hyg +), representing the third and fourth inoculations. As shown in Figure 7, all inoculations on the Hyg − plates exhibited active hyphal growth. In contrast, no hyphal growth was observed on the Hyg + plates after the third inoculation. This suggests that three serial inoculations are required to completely remove the AMA1 plasmid in A. niger.