Development of Genetic Tools in Glucoamylase-Hyperproducing Industrial Aspergillus niger Strains

Simple Summary Glucoamylase is one of the most needed industrial enzymes in the food and biofuel industries. Aspergillus niger is a commonly used cell factory for the production of commercial glucoamylase. For decades, genetic manipulation has promoted significant progress in industrial fungi for strain engineering and in obtaining deep insights into their genetic features. However, genetic engineering is more laborious in the glucoamylase-producing industrial strains A. niger N1 and O1 because their fungal features of having few conidia (N1) or of being aconidial (O1) make them difficult to perform transformation on. In this study, we targeted A. niger N1 and O1 and successfully developed high-efficiency transformation tools. We also constructed a clustered regularly interspaced short palindromic repeat (CRISPR)/Cas9 editing marker-free system using an autonomously replicating plasmid to express Cas9 protein and to guide RNA and the selectable marker. By using the genetic tools developed here, we generated nine albino deletion mutants. After three rounds of sub-culturing under nonselective conditions, the albino deletions lost the autonomously replicating plasmid. Together, the tools and optimization process above provided a good reference to manipulate the tough working industrial strain, not only for the further engineering these two glucoamylase-hyperproducing strains, but also for other industrial strains. Abstract The filamentous fungus Aspergillus niger is widely exploited by the fermentation industry for the production of enzymes, particularly glucoamylase. Although a variety of genetic techniques have been successfully used in wild-type A. niger, the transformation of industrially used strains with few conidia (e.g., A. niger N1) or that are even aconidial (e.g., A. niger O1) remains laborious. Herein, we developed genetic tools, including the protoplast-mediated transformation and Agrobacterium tumefaciens-mediated transformation of the A. niger strains N1 and O1 using green fluorescent protein as a reporter marker. Following the optimization of various factors for protoplast release from mycelium, the protoplast-mediated transformation efficiency reached 89.3% (25/28) for N1 and 82.1% (32/39) for O1. The A. tumefaciens-mediated transformation efficiency was 98.2% (55/56) for N1 and 43.8% (28/64) for O1. We also developed a marker-free CRISPR/Cas9 genome editing system using an AMA1-based plasmid to express the Cas9 protein and sgRNA. Out of 22 transformants, 9 albA deletion mutants were constructed in the A. niger N1 background using the protoplast-mediated transformation method and the marker-free CRISPR/Cas9 system developed here. The genome editing methods improved here will accelerate the elucidation of the mechanism of glucoamylase hyperproduction in these industrial fungi and will contribute to the use of efficient targeted mutation in other industrial strains of A. niger.


Introduction
Glucoamylases (1,4-α-D-glucan glucohydrolases; EC 3.2.1.3) are mainly used to hydrolyze α-1,4 glycosidic bonds at the non-reducing end of starch to cut off the glucose unit [1,2]. Fungal glucoamylases are among the most widely used enzyme preparations in industry, notably in the food industry, including in the production of glucose syrup, fructose syrup, and ethanol [3][4][5][6]. The product portfolio of filamentous fungi is undoubtedly extensive. They are widely used in the production of recombinant proteins and enzymes [7,8]. Notably, Aspergillus niger, which has Generally Recognized as Safe status, is harnessed as a cell factory for the production of a diverse range of enzymes, including glucoamylase, cellulase, xylanase, amylase, protease, and lipase [9][10][11][12][13][14]. A. niger CBS 513.88 is a model strain, and its whole genome has been sequenced [15]. Strain CBS 513.88 is considered an ancestor of the enzymes that are currently used as production strains, and various studies have been performed on the strain CBS 513.88, such as for the improvement of glucoamylase production [13,16]. An efficient transformation system is imperative for the genetic manipulation of filamentous fungi [17]. Similar to other model filamentous fungi such as Neurospora crassa and Aspergillus nidulans, genetic transformation methods have been developed for A. niger, including Agrobacterium tumefaciens-mediated transformation (AMT), protoplast-mediated transformation (PMT), electroporation, and particle bombardment [18]. A. tumefaciens is a plant pathogen that is capable of causing crown gall tumors on plants by transferring a part of its DNA (T-DNA) that is located on a tumor-inducing (Ti) plasmid through a type IV secretion system to the host [19]. Meyer et al. reviewed the advantages and disadvantages of these transformation methods [20]. AMT and PMT are the most commonly used methods and have been successfully applied in the transformation of many filamentous fungi, including A. niger, with variable efficiencies [18,21,22]. Bundock et al. demonstrated that AMT was based on the ability of A. tumefaciens cells to transfer part of their T-DNA; this method is commonly used in plants and yeast [23]. Then, AMT was used in the genetic transformation of several filamentous fungi, including A. awamori; one study found that DNA could be transferred between the prokaryote and the fungus, and, in most transformants, a single copy of the T-DNA was randomly integrated into the genome of the fungus [24]. Since then, the AMT method has been used in a great variety of filamentous fungi [21,[25][26][27][28][29].
A protoplast is a naked cell in which the cell wall has usually been removed using lysing enzymes; it can be applied in cell fusion or transformation. After pioneering fungal protoplast isolation and transformation in Saccharomyces cerevisiae, protoplast isolation and transformation based on protoplasts have been used for filamentous fungi, including in A. niger strains [18,30,31]. AMT does not require the excessive treatment of fungi in the process of transformation; thus conidia, germinated conidia, and vegetative mycelia can all be used for the transformation [25,32]. AMT also has high transformation efficiency and produces stable transformants. However, compared to PMT, AMT is complicated and time consuming. PMT might also be the better method when multiple copies of the expressed gene need to be integrated into the genome [33,34].
Notwithstanding that the genetic manipulation of wild-type A. niger has been established [20,25], many industrially used strains of A. niger are still difficult to transform because the industrial strains have undergone long-term mutagenesis. Compared to the wild-type varieties, the cell structure of these industrial strains has undergone great changes, such as losing most of its conidiation capacity, and even the conidia coats are significantly different from those of the wild-type varieties [35]. A. niger N1 and O1 are glucoamylasehyperproducing industrial strains derived from the same original strain. A. niger strain O1 produces no conidia, whereas A. niger strain N1 produces some conidia. The glucoamylase components and production levels are also different between these strains.
Precision genome editing by clustered regularly interspaced short palindromic repeat (CRISPR)-associated RNA-guided DNA endonucleases (Cas9) has rapidly become a widely used technology. Functional CRISPR/Cas9 systems for gene editing have been successfully developed in filamentous fungi, including in the industrial filamentous fungi Aspergillus spp. [36,37], Trichoderma reesei [38], and Penicillium chrysogenum [39]. In previous studies, Cas9 was either integrated into the genome [37,38,40] or transiently expressed from a nonreplicating plasmid introduced into protoplasts [41][42][43]. Cas9 can also be expressed from plasmids [36] containing autonomous maintenance in the Aspergillus (AMA1) sequence or from the autonomously replicating sequences (ARSs) in Ustilago maydis. Additionally, some studies have reported the use of marker-free deletion for single or multiple genes in one transformation using repair DNA fragment(s) in combination with Cas9-expressing plasmids with self-replicating extrachromosomal AMA1 elements [44][45][46][47][48]. However, gene editing without the use of integrative selection markers has rarely been reported in glucoamylase-hyperproducing industrial A. niger strains.
In previous studies of PMT, 10 8 conidia from Aspergillus giganteus strain IfGB 15/0903 and Myceliophthora thermophila wild-type strain ATCC 42,464 were incubated in liquid culture to generate a sufficient yield of protoplasts, with 55 and 25 transformants being obtained in A. giganteus and M. thermophila, respectively [49,50]. The AMT system was constructed in M. thermophila ATCC 42,464, and the transformation efficiency was about 58% for a pool of 116 transformants analyzed within the time period of 19-21 d [21]. In Aspergillus awamori strain CBS115.52, using the AMT system, hygromycin-resistant transformants were obtained at a frequency of 200-250 transformants per 10 6 conidiospores within 10-12 d [19]. Although the AMT system is usually successful in filamentous fungi, there are also some cases in which AMT is less successful or even fails to produce transformants, e.g., in A. niger and Sclerotinia sclerotiorum (reviewed by Michielse et al. [25]). Moreover, it is very difficult to obtain abundant 10 8 conidia per mL in glucoamylase-hyperproducing industrial strains with small colonies and few conidia or even in strains that are aconidial. Therefore, it is imperative to develop a genetic transformation method suitable for industrial filamentous fungi, especially those with poor growth or in aconidial industrial strains.
In this study, we targeted the glucoamylase-producing industrial strains A. niger N1 and O1. We successfully developed high-efficiency transformation tools via the systematic optimization of the whole process. To the authors' knowledge, the PMT method has not been reported for filamentous fungi that do not produce conidia. Thus, we herein successfully established PMT for the non-conidial strain A. niger O1 by using a culture of strain O1 on agar plates to obtain young mycelia for the release of protoplasts, an approach that has not been studied previously. The final transformation efficiency reached 89.3% for A. niger N1 and 82.1% for A. niger O1 when using PMT. The final transformation efficiency obtained using AMT was 98.2% for A. niger N1 and 43.8% for A. niger O1. Using PMT and the marker-free CRISPR/Cas9 system developed in this study, an albA deletion strain was constructed in the A. niger N1 background. The optimized processes developed here provide a reference for the manipulation of industrial strains, not only for the further engineering of these two glucoamylase-producing strains of A. niger, but also for other industrial Aspergillus strains and other filamentous fungi.

Strains, Media, Culture Conditions, and Plasmids
The A. niger N1 and O1 used in this study were kindly provided by Longda Biotechnology (Shandong, China). As a wild-type control strain, A. niger CBS 513.88 was cultured on potato dextrose agar (PDA) to display the normal production of conidia compared to the industrial strains A. niger N1 and O1, which produced few conidia or no conidia. A. niger N1 was cultured on Czapek-Dox new (CPZ-new) agar in 9 cm Petri dishes or in eggplantshaped culture flasks for 7-10 days to produce conidia. The composition of the CPZ-new O, 5 mL glycerol, and 7.5 g agar. It was adjusted to pH 6.0, made up to 500 mL with double-distilled H 2 O, and autoclaved. A. niger O1 was cultured on Czapek-Dox old (CPZ-old) agar in 9 cm Petri dishes; the medium composition was the same as described previously [51]. Lysing enzymes from T. harzianum (Sigma, St. Louis, USA), Lysozyme (Solarbio, Beijing, China), Snailase (LABLEAD, Beijing, China), Driselase (Sigma), and Yatalase (Takara, Nishinomiya, Japan) were used for protoplast isolation. The MM salts (2.5×; 1 L) used for the induction medium (IM) contained 3.625 g K 2 HPO 4 , 5.125 g KH 2  The binary vector pAN52-Pahr-gfp-Tahr was constructed for the PMT of A. niger N1 and O1. The pAN52-Pahr-gfp-Tahr vector contained the Pahr promoter of the alkyl hydroperoxide reductase from M. thermophila, enhanced green fluorescent protein (GFP), the promoter (PtrpC) of the tryptophan synthetase gene from A. nidulans, and the neomycin resistance gene (neo). The gfp-neo cassette fragment was amplified from pAN52-Pahr-gfp-Tahr using the primers 1F/1R (Table S1) and was also used for the transformation of A. niger N1 and O1 into protoplasts. A Ti vector, pPK2-hph-gfp, containing the PtrpC promoter from A. nidulans and the tef1 promoter (Ptef1) from A. niger was developed as the test plasmid for the AMT of A. niger N1 and O1. The plasmid pPK2-hph-gfp contains a GFP reporter marker as well as a hygromycin resistance gene (hph). The skeleton of pPK2-hph-gfp was reported in a previous study [21].
For the conidia counting analysis, A. niger N1 and A. niger CBS513.88 were cultured on CPZ-new agar and PDA in 9 cm Petri dishes to produce conidia, respectively. Then, 10 mL of sterile normal saline was added to each plate to harvest the conidia by gently scraping the agar with a sterile stick to prepare a suspension. The resulting suspension was filtered through sterile two-layer lens paper to remove mycelial debris, and the conidia were centrifuged at 4000 rpm for 10 min. The pellet was resuspended gently in 1 mL sterile normal saline, and 10 µL of the conidia solution was quantified using a hemocytometer. For morphological analysis, conidia from the A. niger strain N1 were cultivated on several commonly used media for filamentous fungi, including CPZ-old, MM1 (Vogel's minimal medium) [52], MM2 (complete medium) [14], PDA, and EP (EP complete medium) [53] at 30 • C for 7 days. For marker screening, A. niger N1 and O1 were cultivated on bottom-agar plates at 30 and 34 • C, respectively [54].

Preparation and Transformation of A. niger N1 and O1 Mediated by Protoplasts
PMT was performed on A. niger by modifying a method previously described for A. nidulans [55]. A. niger N1 conidia were cultivated in an eggplant-shaped culture flask in CPZ new medium for 7-10 days, and the conidia were collected. The culture conditions used to produce mycelia for protoplast isolation were specific to each strain (A. niger N1 and O1), so mycelia were cultured in five different media to investigate the optimal conditions. For A. niger strain N1, a conidial suspension (10 7 /mL) was prepared with normal saline, and 100 µL of that suspension was coated on a 9 cm Petri dish containing CPZ-old, MM1, MM2, PDA, or EP medium, covered with a piece of cellophane, and incubated at 30 • C. Then, mycelium was harvested at different times to test the effect of mycelial age on protoplast yield. For A. niger O1, which does not produce conidia, using a stick, we scraped young mycelia from CPZ-old agar onto PDA, MM1, MM2, or EP in a 9 cm Petri dish covered with a piece of cellophane. Unlike in the culturing method for A. niger mycelia using liquid medium [14] for the release of protoplasts, herein, the young mycelia of A. niger O1 were scraped onto the solid plate. Ten sheets of cellophane with mycelia attached were completely immersed in a 9 cm Petri dish containing 25 mL of enzymatic hydrolysate dissolving solution A (0.1 M KH 2 PO 4 and 1.2 M sorbitol, pH adjusted to 5.6 with KOH) at 30 • C. Unlike methods that use inorganic KCl as the osmotic stabilizer [55], we used organic sorbitol. The lytic enzyme preparation Novozyme 234 is not currently available for purchase, so we applied lysing enzymes from T. harzianum, Lysozyme, Snailase, Driselase, and Yatalase. Protoplast cleavage was observed under a microscope every half hour. The effects of enzymatic incubation for 0.5, 1, 2, 3, 4, and 5 h were also determined. The resulting suspension was filtered through sterile two-layer lens paper to remove mycelial debris, and the protoplasts were centrifuged at 3000 rpm for 10 min.
For the PMT of A. niger N1 and O1, the supernatant from the above centrifugation was discarded, and the pellet was washed with 5-10 mL of solution B (50 mM CaCl 2 , 1 M sorbitol, 10 mM Tris-Cl (pH 7.5), pH adjusted to 5.6 with HCl) and then centrifuged at 3000 rpm for 10 min. The pellet was resuspended gently in 220 µL of solution B, and 10 µL of the protoplast solution was counted using a hemocytometer. Linear gfp-neo cassette DNA (10 µg) was added and mixed gently. Next, 70 µL of PEG solution (25% PEG 6000 and 50 mM CaCl 2 in 10 mM pH 7.5 Tris-Cl) was added and mixed gently, and the mixture was incubated on ice for 20 min. PEG solution (2 mL) was added to the tube dropwise, and the tube was gently swirled before being left at room temperature for 5 min. Then, 4 mL of solution B was added, and the tube was gently swirled. Medium (containing 2% sucrose, 2 mL 50 × Vogel's salts, 1 M sorbitol, 1.5% agarose, and 200-300 µg/mL G418 in 100 mL) was added to the above mixture to a volume of 50 mL. The tube was gently inverted, and the mixture was poured over seven bottom media (containing 2% sucrose, 2 mL 50 × Vogel's salts, 1 M sorbitol, 1.5% agar, and 200-300 µg/mL G418 in 100 mL) plates. Transformants were selected after culturing at 30 • C (A. niger N1) or at 34 • C (A. niger O1) for 4-7 days.

Preparation and Transformation of A. niger N1 and O1 Mediated by A. tumefaciens
Agrobacterium-mediated genetic transformation was performed by modifying a method previously described for M. thermophila [21]. First, a single colony of A. tumefaciens harboring pPK2-hph-gfp was inoculated into Luria-Bertani medium supplemented with 50 µg/mL of kanamycin and was cultivated at 220 rpm for 12-20 h at 28 • C to an optical density of 0.5-1.0 at 600 nm (OD600). Xu et al. reported that A. tumefaciens needs to be washed twice with IM containing AS [21]. However, herein, the transformants were obtained without the A. tumefaciens being washed. Then, the cells were harvested by centrifugation at 4000 rpm for 10 min and were directly diluted with IM containing 200 µM AS and 10 mM MES to OD600 = 0.1-0.2. The culture was grown at 28 • C with shaking at 220 rpm for 6-8 h; it could then be used for cocultures for the transformation of A. niger (see below).
Conidia of A. niger N1 or young mycelia of A. niger O1 were prepared as described above (see Section 2.1). A. niger N1 conidia were cultivated in eggplant-shaped culture flasks containing CPZ-new agar for 7-10 days. Sterile normal saline was added to the flasks to harvest the conidia by gently scraping the agar with a sterile stick to prepare a conidial suspension (10 6 /mL). Using a sterile stick, we scraped young mycelia of A. niger O1 grown on CPZ-old agar for 2 days onto fresh CPZ-old agar covered with a piece of cellophane in a 9 cm Petri dish and incubated the dish for 19 h at 34 • C. In contrast to a previous study [21], the young mycelia of A. niger O1, but not the conidia, were co-incubated with A. tumefaciens, which provided a reference for the AMT of non-conidial filamentous fungi.
Finally, A. tumefaciens and A. niger were cocultured. The A. tumefaciens culture (100 µL) was mixed with an equal volume of A. niger N1 conidia (10 6 /mL) or young mycelia of A. niger O1 and then spread onto an IM agar plate (9.0 cm; supplemented with 200 µM AS and 10 mM MES) that had been covered with cellophane beforehand. After cocultivation at 28 • C for 2 days, the cellophane was transferred onto an M-100 medium plate (9.0 cm) supplemented with 50 µg/mL hygromycin (Roche, Basel, Switzerland) for A. niger N1 or with 200 µg/mL hygromycin for A. niger O1 and 300 µg/mL cefotaxime (Sigma-Aldrich, St. Louis, USA) and overlaid with M-100 medium supplemented with 50 µg/mL hygromycin for A. niger N1 or 200 µg/mL hygromycin for A. niger O1 and 300 µg/mL cefotaxime before being cultivated at 30 • C (for A. niger N1) or 34 • C (for A. niger O1) for 4-7 days.

Screening of Transformants by PCR, Fluorescence Stability, RT-qPCR, and Southern Blotting
First, all of the transformants were checked by PCR analysis. The genomic DNAs of the transformants were extracted as previously described [56] and were used as templates for PCR. PCR-based detection systems were designed: primer pairs GFP-1F/1R and NEO-1F/1R were used to amplify the gfp and the neo resistance genes from the genomic DNAs of the fungi transformed with the gfp-neo cassette, respectively. Only in the case of correct transformation could the 720-bp (GFP-1F/1R) and 759-bp (NEO-1F/1R) fragments be amplified. There should be no amplification of the target genes from the untransformed genomic DNA Of A. niger N1 or O1 (Table S1). Similarly, GFP-2F/2R and HPH-1F/1R were used to amplify the gfp and hph from the genomic DNA of fungi transformed with pPK2hph-gfp, respectively. The expected amplified fragments for GFP-2F/2R and HPH-1F/1R were 1253-and 1201-bp long, respectively. Genomic DNA of untransformed A. niger N1 or O1 was used as a negative control (Table S1). Then, the PCR-confirmed transformants were observed by fluorescence microscopy. To verify the stability of the fluorescence, 10 PCR-and fluorescence-confirmed transformants were randomly selected, passaged twice on CPZ-new or CPZ-old agar containing antibiotics, and the fluorescence signals were observed again.
To determine the copy number(s) of the integrated gfp gene in transformants, fungal genomic DNAs were used as templates for RT-qPCR using a previously described method [54]. To determine the amplification efficiencies of all of the reactions, genomic DNA samples were diluted serially to construct standard curves and were then subjected to RT-qPCR three times. All of the primers used are listed in Table S1.
Although PCR reactions allow the recognition of T-DNA integration events, they cannot determine if the transformed DNA is integrated into a random place in the genome or into the copy number of the T-DNA. It is therefore imperative that the transformants be further analyzed by Southern blotting. Southern blotting was performed with 20 µg of genomic DNA that was digested by XhoI for PMT and by EcoRV for AMT. Genomic DNA was extracted as previously described [54]. The digested DNA was separated by agarose gel electrophoresis, and DNA transfer was performed as previously described [57]. The gfp fragment was used as a template, and the 310-bp PCR-amplified product used as the probe was generated with the primers Probe-F/Probe-R (Table S1). Probe preparation, membrane hybridization, and visualization were performed according to the manufacturer's instructions for the DIG High Prime DNA Labeling and Detection Starter Kit II (Roche, Mannheim, Germany).

CRISPR/Cas9 Plasmid Design
To express Cas9 and a guide RNA from the same autonomously replicating vector using hygromycin as a selection marker for fungal transformation, plasmid pFC332 was used [36]. pFC332-AnalbA contains a unique PacI site that was used to insert a single guide RNA (sgRNA) expression cassette based on the native A. niger N1 U6 promoter. To select specific sgRNA targeting in A. niger albA (AnalbA), the sgRNA target site (AnalbA Target) in the genome of A. niger N1 was identified using the sgRNACas9 tool [58]. A. niger N1 U6p-AnalbA Target-sgRNA was synthesized and was inserted into pFC332 digested with PacI to generate pFC332-AnalbA. The 5 -and 3 -flanking fragments of AnalbA were separately amplified from A. niger N1 genomic DNA via PCR with the primer pairs AnalbA-5F/R and AnalbA-3F/R, respectively (Table S1). The amplified 5 -and 3 -fragments were assembled and ligated into pUC118 (digested with EcoRI and XbaI) using an NEB Gibson Assembly Kit to generate a pUC118-donor DNA sequence. The donor DNA sequences amplified with the primers AnalbA-5F and AnalbA-3R were used for transformation (Table S1).
The plasmid pFC332-AnalbA and the donor DNA fragment were transformed into the protoplasts of A. niger N1. We used 10 µg of pFC332-AnalbA with approximately 10 µg of donor DNA fragments for each transformation. Transformants were identified using primers AnalbA-DE-F/R; amplification from the genomic DNA of A. niger N1 was expected to generate a PCR fragment of 856 bp, whereas amplification from the albA-knockout strains was expected to generate a PCR fragment of 560 bp. AlbA deletion caused the A. niger colonies to be white. White colonies were streaked on CPZ-new agar containing hygromycin (50 µg/mL) to ensure that the spores harbored pFC332-AnalbA and were thus more likely to be transformed. Next, a single colony was picked and transferred to non-selective CPZ-new medium to allow for loss of pFC332-AnalbA. Finally, a streak of a single colony on both CPZ-new medium containing hygromycin (50 µg/mL) and CPZ-new medium without hygromycin was used as a control for plasmid loss. DNA was isolated from strains that had lost the pFC332-AnalbA plasmid, and then a hygromycin gene fragment was determined by PCR using the primers HPH-DE-F/R (1006 bp) (Table S1).

Growth Phenotypes of A. niger N1 and O1
The morphology of the glucoamylase-producing strains A. niger N1, O1, and CBS 513.88 was checked on the medium-covered plates after 7 days of growth ( Figure 1A). A. niger N1 and O1 formed smaller colonies than strain CBS 513.88. The conidia concentration of A. niger CBS 513.88 was about 44.77 × 10 6 /mL for each plate, while A. niger N1 had far less conidia production (6.67 × 10 6 /mL) ( Figure 1A and Figure S1). Although the mycelia of A. niger N1 grew well on various commonly used media, conidia could only be produced on CPZ-new medium ( Figure 1B). Moreover, A. niger O1 did not produce conidia at all ( Figure 1A). The carbon source in the CPZ-old and MM1 media was sucrose; the carbon source in MM2 was glucose; there were two carbon sources (glucose and maltose) in EP medium; PDA contains a variety of carbon sources, while the major carbon source in CPZ-new medium is a high concentration of sorbitol. Previous studies have reported that the inhibition of mycelial growth was partially recovered by supplementing low or medium concentrations of sorbitol [59]. Additionally, sorbitol can stimulate the germination and growth of some xerophilic fungi [60]. Here, we found that when used as the major carbon source, sorbitol promoted the production of conidia by A. niger strain N1.

Marker Screening of A. niger N1 and O1
To improve the genetic transformation methods for A. niger N1 and O1, the first thing we identified was appropriate antibiotics for selection. A. niger N1 and O1 were cultivated on bottom-agar plates with hygromycin B, G418, or phosphinothricin at different concentrations. A. niger N1 was completely inhibited by 50 µg/mL hygromycin and 200 µg/mL G418 (Figure 2A). A. niger O1 was completely inhibited by 200 µg/mL hygromycin and 300 µg/mL G418 ( Figure 2B). Both A. niger N1 and O1 grew normally at less than or equal to 200 µg/mL and 300 µg/mL phosphinothricin, respectively (Figure 2A,B). Taken together, both A. niger N1 and O1 showed sensitivity to hygromycin and G418, and they were insensitive to phosphinothricin (Figure 2A,B). Additionally, A. niger N1 appeared to be more sensitive to hygromycin and G418 than A. niger O1 (Figure 2A,B).
We chose concentrations of antibiotics that completely inhibited the growth of A. niger N1 and O1 during the process of genetic transformation. Therefore, final concentrations of 50 µg/mL hygromycin B and 200 µg/mL G418 were used as screening concentrations during the genetic transformation of A. niger N1. Meanwhile, final concentrations of

Marker Screening of A. niger N1 and O1
To improve the genetic transformation methods for A. niger N1 and O1, the first thing we identified was appropriate antibiotics for selection. A. niger N1 and O1 were cultivated on bottom-agar plates with hygromycin B, G418, or phosphinothricin at different concentrations. A. niger N1 was completely inhibited by 50 µg/mL hygromycin and 200 µg/mL G418 (Figure 2A). A. niger O1 was completely inhibited by 200 µg/mL hygromycin and 300 µg/mL G418 ( Figure 2B). Both A. niger N1 and O1 grew normally at less than or equal to 200 µg/mL and 300 µg/mL phosphinothricin, respectively (Figure 2A,B). Taken together, both A. niger N1 and O1 showed sensitivity to hygromycin and G418, and they were insensitive to phosphinothricin (Figure 2A,B). Additionally, A. niger N1 appeared to be more sensitive to hygromycin and G418 than A. niger O1 (Figure 2A,B).
We chose concentrations of antibiotics that completely inhibited the growth of A. niger N1 and O1 during the process of genetic transformation. Therefore, final concentrations of 50 µg/mL hygromycin B and 200 µg/mL G418 were used as screening concentrations during the genetic transformation of A. niger N1. Meanwhile, final concentrations of 200 µg/mL hygromycin B and 300 µg/mL G418 were used as screening concentrations during the genetic transformation of A. niger O1.

Preparation of Protoplasts of A. niger N1 and O1
To obtain a high transformation efficiency, it is essential to generate enough protoplasts. The quantity of protoplast production can be affected by several factors: the composition of the culture medium, the age of the mycelia used, the lytic enzymes used, and the enzymolysis time. To find the best medium to culture A. niger N1 for protoplast production, mycelia were cultured in five different media at 30 • C. The highest yield of A. niger N1 protoplasts (9.2 × 10 6 /mL) was obtained from CPZ-old medium; there was also little mycelial debris in this medium. The yields of the protoplasts decreased successively in media MM1, PDA, MM2, and EP. PDA and EP media released fewer protoplasts and significant mycelial debris ( Figure 3A,B). Consequently, CPZ-old medium was the best among the tested media ( Figure 3A,B). Compared to the media (MM1 (Vogel's minimal medium) [52], MM2 (complete medium) [14], PDA, EP [53]) used in previous studies, the nutrients in CPZ-old medium were relatively simple, with fewer inorganic salts in terms of type.

Preparation of Protoplasts of A. niger N1 and O1
To obtain a high transformation efficiency, it is essential to generate enough protoplasts. The quantity of protoplast production can be affected by several factors: the composition of the culture medium, the age of the mycelia used, the lytic enzymes used, and the enzymolysis time. To find the best medium to culture A. niger N1 for protoplast production, mycelia were cultured in five different media at 30 °C. The highest yield of A. niger N1 protoplasts (9.2 × 10 6 /mL) was obtained from CPZ-old medium; there was also little mycelial debris in this medium. The yields of the protoplasts decreased successively in media MM1, PDA, MM2, and EP. PDA and EP media released fewer protoplasts and significant mycelial debris ( Figure 3A,B). Consequently, CPZ-old medium was the best among the tested media ( Figure 3A,B). Compared to the media (MM1 (Vogel's minimal medium) [52], MM2 (complete medium) [14], PDA, EP [53]) used in previous studies, the nutrients in CPZ-old medium were relatively simple, with fewer inorganic salts in terms of type.
The effect of mycelial age on protoplast yield was then investigated. With increasing culture time, the yield of protoplasts increased first and then decreased; it peaked at 9.2 × The effect of mycelial age on protoplast yield was then investigated. With increasing culture time, the yield of protoplasts increased first and then decreased; it peaked at 9.2 × 10 6 /mL when the culture time was 21 h ( Figure 3C). Additionally, the enzymatic combinations had a significant influence on the yield of protoplasts. It was found that 1% lysing enzymes from T. harzianum alone was best, producing a protoplast yield of 9.0 × 10 6 /mL. The protoplast yields were slightly decreased when other lytic enzymes were added (Table 1). Furthermore, the incubation time was an important factor for protoplast release because shortened or prolonged incubation with lytic enzymes resulted in the incomplete formation of protoplasts or the degradation of early formed protoplasts. The optimum enzymatic incubation period was determined by incubating mycelium for up to 5 h with cell wall lytic enzymes. The highest yield (11 × 10 6 /mL) was observed after 2 h of incubation ( Figure 3D). Importantly, the enzymatic hydrolysate needed to be removed as soon as possible to avoid excessive protoplast incubation. In summary, the maximum protoplast production from A. niger N1 (11 × 10 6 /mL) was obtained from old mycelia that were 21 h old that had been cultured on CPZ-old medium and digested with 1% T. harzianum lysing enzymes and incubated at 30 • C for 2 h. cell wall lytic enzymes. The highest yield (11 × 10 /mL) was observed after 2 h of incub tion ( Figure 3D). Importantly, the enzymatic hydrolysate needed to be removed as so as possible to avoid excessive protoplast incubation. In summary, the maximum pro plast production from A. niger N1 (11 × 10 6 /mL) was obtained from old mycelia that we 21 h old that had been cultured on CPZ-old medium and digested with 1% T. harzianu lysing enzymes and incubated at 30 °C for 2 h.    Three media-CPZ-old, MM1, and MM2-were tested for protoplast preparation from mycelium of aconidial A. niger O1. The highest yield of protoplasts (16.9 × 10 6 /mL) was obtained when using CPZ-old medium. Therefore, we chose CPZ-old as the culture medium for A. niger O1 ( Figure 4A). We optimized the culture time of A. niger O1 and found that mycelium cultured for 19 h produced the highest yield of protoplasts; a steep decrease in the protoplast yield was observed when the culture time was changed to 12 or 26 h ( Figure 4B). Interestingly, unlike for A. niger strain N1, few protoplasts were obtained from A. niger O1 using lysing enzymes from T. harzianum alone, and the addition of other lytic enzymes was necessary to obtain a sufficient yield of protoplasts. Up to 15 × 10 6 /mL protoplasts were obtained when using five lytic enzymes preparations together (lysing enzymes from T. harzianum, Lysozyme, Snailase, Driselase, and Yatalase); however, much of the mycelial debris could not be filtered out (Table 2). To decrease the amounts of mycelial debris, we removed each lytic enzyme preparation from the cocktail in turn. The mycelial debris disappeared after the removal of Yatalase, while the removal of any of the other lytic enzymes did not alter the amounts of mycelial debris. The yield of protoplasts was maximal (17 × 10 6 /mL) when the lytic enzyme combination used included 1.2% lysing enzymes from T. harzianum, 0.5% Lysozyme, and 0.5% Snailase ( Table 2). The final results showed that the maximum production of protoplasts (17 × 10 6 /mL) from A. niger O1 was obtained from mycelia that were 19 h old that had been cultured on CPZ-old medium and digested with a multiple-lytic-enzyme combination including 1.2% lysing enzymes from T. harzianum, 0.5% Lysozyme, and 0.5% Snailase. or 26 h ( Figure 4B). Interestingly, unlike for A. niger strain N1, few protoplasts were obtained from A. niger O1 using lysing enzymes from T. harzianum alone, and the addition of other lytic enzymes was necessary to obtain a sufficient yield of protoplasts. Up to 15 × 10 6 /mL protoplasts were obtained when using five lytic enzymes preparations together (lysing enzymes from T. harzianum, Lysozyme, Snailase, Driselase, and Yatalase); however, much of the mycelial debris could not be filtered out (Table 2). To decrease the amounts of mycelial debris, we removed each lytic enzyme preparation from the cocktail in turn. The mycelial debris disappeared after the removal of Yatalase, while the removal of any of the other lytic enzymes did not alter the amounts of mycelial debris. The yield of protoplasts was maximal (17 × 10 6 /mL) when the lytic enzyme combination used included 1.2% lysing enzymes from T. harzianum, 0.5% Lysozyme, and 0.5% Snailase ( Table  2). The final results showed that the maximum production of protoplasts (17 × 10 6 /mL) from A. niger O1 was obtained from mycelia that were 19 h old that had been cultured on CPZ-old medium and digested with a multiple-lytic-enzyme combination including 1.2% lysing enzymes from T. harzianum, 0.5% Lysozyme, and 0.5% Snailase.

PMT and AMT of A. niger N1 and O1
We constructed the binary vector pAN52-Pahr-gfp-Tahr ( Figure 5A) using the methodology described above (see Section 2.1). The GFP cassette fragment was amplified from this vector using the primers 1F/1R and was used for the transformation of A. niger N1 and O1 into protoplasts. Unlike in a previous method that used shaken liquid culture to obtain young mycelia [14,50], herein, young mycelia of A. niger N1 (which produces few conidia) and aconidial A. niger O1 generated on solid medium were used for the release of protoplasts. For A. niger N1, 10 7 conidia were needed to obtain sufficient protoplasts using solid-medium plates, whereas in previous work, 10 8 conidia were incubated in liquid culture to generate a sufficient yield of protoplasts [14,50]. A. niger strain O1 forms small colonies on agar plates, but it is difficult to produce dispersed mycelia in shaken-flask culture, while 3-6 days are needed to obtain mycelia via static culture in liquid. Thus, compared to static culture, the solid-medium method used here saved time. Therefore, the solid-medium culture method for the release of protoplasts that we applied has advantages for both A. niger N1 and O1. gfp 7-12) were randomly selected. Similarly, six PMT transformants of A. niger O1 with the gfp reporter gene ( Figure S3B: PMT-gfp 1-6) and six AMT transformants with the gfp reporter gene ( Figure S3B: AMT-gfp 7-12) were randomly selected. The gfp gene of all the AMT transformants was present in a single copy in both A. niger N1 and O1. This is consistent with a previous study in A. giganteus [49]. Meanwhile, the gfp gene of most of the PMT transformants was present in multiple copies in both A. niger N1 and O1 ( Figure S3).
To further confirm this observation, we randomly selected three and two PCR-confirmed transformants produced using PMT and AMT, respectively, and performed Southern blot analysis of them. Among the three transformants prepared using PMT, there was a single-copy gfp gene insertion, a two-copy insertion, and a three-copy insertion ( Figure  6F). The two transformants obtained by AMT were single-copy insertions. Additionally, it seems that the integration of the introduced DNA is completely at random within the chromosome ( Figure 6G). The probe did not yield bands with genomic DNA from untransformed A. niger N1 (Figure 6F,G).  Putatively transformed colonies were screened by PCR analysis. It was found that 25 of 28 transformants into A. niger N1 by PMT were correct, as identified using primer pairs GFP-1F/1R and NEO-1F/1R. There was no amplification of the target genes in the negative controls ( Figure S2A). For A. niger O1 and PMT, 32 of 39 transformants were correct ( Figure S2C). Ten PCR-and fluorescence-confirmed transformants were randomly selected to observe the stability of the fluorescence. All of the transformants had stable fluorescence signals determined by the methods described in Section 2.4 ( Figure 6A,C). In total, 69 and 16 positive transformants in each transformation plate (there are seven plates in one transformation process) of A. niger N1 and O1 were obtained by PMT, respectively ( Figure 6E). Taken together, the transformation efficiencies obtained by PMT were 89.3% (25/28) and 82.1% (32/39) for A. niger N1 and O1, respectively ( Figure S2A,C).
Separately, pPK2-hph-gfp ( Figure 5B) was transformed into conidia of A. niger N1 and young mycelia of A. niger O1 by AMT. The age of the mycelia of A. niger O1 was the vital factor for the success of the AMT method-the optimum culture time of A. niger O1 mycelia was 8 h, while the AMT method did not result in transformants when applied to A. niger O1 if the culture time was 15 h. Acetosyringone (AS) served as an inducer of virulence genes, the expression of which is a prerequisite for T-DNA transfer [24]. The AS concentration during the Agrobacterium co-cultivation period and the preculture period of A. tumefaciens influenced the frequency of transformation during the AMT of S. cerevisiae and several filamentous fungi [21]. A previous study using M. thermophila reported that the maximum transformation efficiency was obtained with 200 µM AS [21]. Therefore, the concentration of AS used in our experiments was 200 µM. Here, 55 of 56 AMT-mediated transformants of A. niger N1 and 28 of 64 transformants of A. niger O1 were found to be positive using the primers GFP-2F/2R and HPH-1F/1R ( Figure S2B,D). Additionally, 129 and 38 positive transformants in each transformation plate were obtained for A. niger N1 and O1 by AMT, respectively ( Figure 6E). The transformation efficiencies by AMT were 98.2% (55/56) and 43.8% (28/64) for A. niger N1 and O1, respectively ( Figure S2B,D).

CRISPR/Cas9 Genome Editing System and Marker-Free albA Gene Knockout Assay in A. niger N1
We constructed the vector pFC332-AnalbA, and the donor DNA fragment transformed into protoplasts of A. niger N1 ( Figure 7A) as described above (see Section 2.5). To demonstrate the functionality of the CRISPR/Cas9 system in A. niger N1, we chose to target the albA gene (ANI_1_726084). Knockout of albA results in the formation of whitecolored colonies, thus providing a direct indication on transformation plates. A series of To validate the integration of the transformed DNA into the host genome in more detail, we determined the copy number(s) of the integrated gfp gene in the positive transformants by real-time quantitative PCR and Southern blot analysis. Six PCR and fluorescenceconfirmed PMT transformants of A. niger N1 with the gfp reporter gene ( Figure S3A: PMT-gfp 1-6) and six AMT transformants with the gfp reporter gene ( Figure S3A: AMT-gfp 7-12) were randomly selected. Similarly, six PMT transformants of A. niger O1 with the gfp reporter gene ( Figure S3B: PMT-gfp 1-6) and six AMT transformants with the gfp reporter gene ( Figure S3B: AMT-gfp 7-12) were randomly selected. The gfp gene of all the AMT transformants was present in a single copy in both A. niger N1 and O1. This is consistent with a previous study in A. giganteus [49]. Meanwhile, the gfp gene of most of the PMT transformants was present in multiple copies in both A. niger N1 and O1 ( Figure S3).
To further confirm this observation, we randomly selected three and two PCR-confirmed transformants produced using PMT and AMT, respectively, and performed Southern blot analysis of them. Among the three transformants prepared using PMT, there was a singlecopy gfp gene insertion, a two-copy insertion, and a three-copy insertion ( Figure 6F). The two transformants obtained by AMT were single-copy insertions. Additionally, it seems that the integration of the introduced DNA is completely at random within the chromosome ( Figure 6G). The probe did not yield bands with genomic DNA from untransformed A. niger N1 ( Figure 6F,G).

CRISPR/Cas9 Genome Editing System and Marker-Free albA Gene Knockout Assay in A. niger N1
We constructed the vector pFC332-AnalbA, and the donor DNA fragment transformed into protoplasts of A. niger N1 ( Figure 7A) as described above (see Section 2.5). To demonstrate the functionality of the CRISPR/Cas9 system in A. niger N1, we chose to target the albA gene (ANI_1_726084). Knockout of albA results in the formation of white-colored colonies, thus providing a direct indication on transformation plates. A series of transformations of pFC332-AnalbA were performed in A. niger N1 ( Figure S4). To verify whether the albA gene had been knocked-out, 22 randomly selected single transformation colonies were analyzed the PCR using primer pair AnalbA-DE-F/R ( Figure S5). A. niger N1 (which has a Ku70 ortholog) can repair a Cas9-induced double-strand break through nonhomologous end joining (NHEJ). However, as shown in Figure S4B, the transformation of A. niger N1 with pFC332-AnalbA yielded about five colonies in each transformation plate (total seven plates), but the AnalbA gene was not knocked out in 22 randomly selected colonies without a homology template ( Figure S5B). Transformation with the control plasmid, pFC332, which lacks an U6p-AnalbA Target-sgRNA expression cassette, resulted in 40 viable colonies ( Figure S4A), and, as expected, PCR showed that there were no AnalbA-knockout transformants among them ( Figure S5A).
Co-transformation of a homology repair DNA fragment consisting of fused 5 -and 3 -flanking regions of the AnalbA gene, together with pFC332-AnalbA, yielded 9 (out of 22) white transformants on the transformation plates ( Figures S4D and S5D). Streaking of one such transformant on a CPZ-new agar plate containing hygromycin showed the white coloration more clearly ( Figure S4F). A control co-transformation of pFC332 without the sgRNA targeting plasmid and the homology repair DNA fragment did not yield knockout colonies ( Figures S4C and S5C), showing that the repair DNA fragment does not integrate autonomously at the site of homology without the assistance of the Cas9-sgRNA targeting plasmid. A streak of one such transformant is shown in Figure S4E; the difference in color compared to the AnalbA knockout shown in Figure S4F is clear.
White transformants produced by the co-transformation of the homology repair DNA fragment and pFC332-AnalbA were cultured on CPZ-new plates without hygromycin as described in "Methods 2.5" to assess whether the transformants could lose the pFC332-AnalbA plasmid in the absence of selective pressure. The nine AnalbA-knockout transformants then failed to grow in medium containing 50 µg/mL hygromycin, thus indicating that they had lost the plasmid ( Figure 7B). They were further analyzed by PCR using the primers HPH-DE-F/R to determine if the hygromycin resistance gene had been lost. There was no amplification of the hygromycin resistance gene from the nine AnalbA-knockout transformants or the A. niger N1 negative control, but a 1006-bp fragment was amplified from the positive control plasmid ( Figure S5E). Thus, we demonstrated that the AnalbAknockout transformants could lose the plasmid ( Figure 7B) that enables the construction of marker-free gene knockouts.

Discussion
The application of fungal protoplasts has contributed to the development of microbial strains and the genetic recombination of fungi. Because of the limitations of traditional mutation screening methods [61], improvement in the genetic transformation methods for industrial strains of filamentous fungi, especially aconidial A. niger O1, is urgently needed. In this study, we improved PMT and AMT techniques for A. niger N1 (which produces few conidia) and A. niger O1 (which is aconidial).
The physiological stage of the mycelium plays a significant role in the release of protoplasts [62,63]. Here, the culture medium used had a great influence on the growth of A. niger. Dispersed mycelium was observed when A. niger N1 was cultivated on CPZ-old agar plates with sucrose as the sole available extracellular carbon source (along with sev- Figure 7. Generation of the marker-free AnalbA-knockout mutants of A. niger N1 by autonomously replicating vector harboring Cas9-and sgRNA-encoding genes and a hygromycin sensitivity test of AnalbA-knockout mutants before and after plasmid removal. (A) pFC332-AnalbA and the donor DNA fragment transformed into A. niger N1 by PMT. AnU6p is the promoter of A. niger N1; Ptef1 is the promoter of translation elongation factor gene tef1 from A. nidulans; the cas9 gene encoding Streptococcus pyogenes Cas9 was codon-optimized for expression in A. niger; AMA1, self-replicating extrachromosomal AMA1 elements from A. nidulans; Hph, hygromycin resistance gene. (B) Phenotypes of nine AnalbA-knockout transformants with loss of pFC332-AnalbA plasmid with and without hygromycin pressure. Mutant with AnalbA knockout and lost pFC332-AnalbA plasmid failed to grow in medium containing 50 µg/mL hygromycin.

Discussion
The application of fungal protoplasts has contributed to the development of microbial strains and the genetic recombination of fungi. Because of the limitations of traditional mutation screening methods [61], improvement in the genetic transformation methods for industrial strains of filamentous fungi, especially aconidial A. niger O1, is urgently needed.
In this study, we improved PMT and AMT techniques for A. niger N1 (which produces few conidia) and A. niger O1 (which is aconidial).
The physiological stage of the mycelium plays a significant role in the release of protoplasts [62,63]. Here, the culture medium used had a great influence on the growth of A. niger. Dispersed mycelium was observed when A. niger N1 was cultivated on CPZ-old agar plates with sucrose as the sole available extracellular carbon source (along with several inorganic salts), while the mycelia appeared densely when A. niger N1 was cultivated on MM1, MM2, EP, or PDA plates ( Figure 1B). MM1 and MM2 contain a variety of trace elements, a single carbon source, and several inorganic ions. EP medium contains many trace elements, two carbon sources, and several inorganic ions. The nutrient composition of PDA is even more complex. These observations suggest that the dense mycelia cultured in nutrient-rich media were not conducive to enzymolysis and the release of protoplasts.
Mycelium age is an important factor for protoplast release, and different filamentous fungi behave differently [22,62,64]. With regard to the culturing time of A. niger N1, mycelia harvested at 21 h produced the highest yield of protoplasts, and a steep decrease in the protoplast yield was observed with the use of 15 or 27 h-old mycelia. Previous studies have reported that young and exponentially growing mycelia favored the release of protoplasts [22,62,64]. For instance, A. niger conidia were cultured in 100 mL complete medium (CM) for 12 h (at 100 rpm) to generate protoplasts [14]. It seems that young mycelia are more susceptible to enzymatic digestion than old mycelia and that the cell wall of old mycelia becomes thicker [22,[63][64][65].
As previously reported, the biggest influencing factor on the release of protoplasts from mycelia was the choice of the enzymatic combination and the concentrations used [61,62]. The main components of fungal cell walls are polysaccharides, proteins, and lipids. The cell wall polysaccharides of A. niger mainly include chitin, cellulose, and glucan [66]. Taking N. crassa as an example, the outermost layer is an amorphous glucan, followed by a coarse web of glycoproteins, and the innermost layer is chitin [66]. According to the information provided by the manufacturers, the main components of the lysing enzymes from T. harzianum are cellulase, protease, chitinase, and glucanase. Because this preparation contains most of the lytic enzyme activities required to lyse the main components of the cell wall of filamentous fungi, it has been the most commonly used lytic enzyme mixture. However, other lytic enzymes may also need to be added for different filamentous fungi. Lysozyme mainly lyses peptidoglycan; Snailase mainly contains cellulase, protease, and pectinase; Driselase mainly contains laminarin, xylanase, and cellulase; and Yatalase mainly demonstrates chitinase and chitobiase activities. The lysing enzymes from T. harzianum alone were enough for the release of A. niger N1 protoplasts (Table 1). However, Lysozyme and Snailase were also needed for the release of protoplasts from A. niger O1 (Table 2). We speculate that the compositions of the cell walls of A. niger N1 and O1 are different even though both strains are descended from the same ancestor.
PMT has been generally used in many fungi, and the transformation efficiencies vary greatly in different fungi for different purposes [49,67], such as in A. giganteus, where there were 55 transformants in 10 8 protoplasts/µg DNA obtained by PMT [49]. In this study, the PMT technique yielded up to 10 and 8 transformants in 10 7 protoplasts per microgram of DNA in A. niger N1 and O1, respectively. Meyer et al. reported that the deletion of the Ku70 homologue from A. niger (KusA) dramatically improved the homologous integration efficiency of AngfaA, which reached >95% compared to 19% in the wild-type background when 1000-bp homologous flanks were used [67]. For the AMT system in filamentous fungi, previous studies reported that the transformation efficiencies varied greatly in fungal species: 0.04% for Blastomyces dermatitidis [68], 29% for A. awamori [69], 58% for M. thermophila [21], 74% for Fusarium avenaceum [70], and 85% for F. graminearum [70]. However, a previous study reported that the transformation frequency obtained via AMT of A. niger was not as high as expected and was not reproducible for unknown reasons [25]. Herein, the transformation efficiencies by AMT were 98.2% (55/56) and 43.8% (28/64) for A. niger N1 and O1, respectively ( Figure S2B,D).
While this manuscript was in revision, two papers regarding the strain engineering of A. niger N1 [71] and O1 [72] using the protoplast transformation procedure described in M. thermophila were published. In A. niger O1, Guo et al. described that the four overexpression and four knockout transformants of the acid α-amylase gene were simply obtained after protoplast transformation, while four overexpression and two knockout transformants of the neutral α-amylase gene were generated [72]. In A. niger N1, 6 out of 52 transformants were triple-gene mutants, with a gene editing efficiency of 11.5% [71]. Compared to previous studies [71,72], the PMT method we developed here is relatively simple. First, in previous studies, the young mycelia were washed twice with 1 mol/L MgSO 4 before lysing with enzymatic hydrolysate [71,72], while in our study, enough protoplasts could be obtained without carrying out these washing steps; therefore, the PMT method that we developed is relatively simple and saves time. Second, previous studies washed the protoplasts with STC solution twice after the young mycelia had been lysed with enzymatic hydrolysate [71,72]; however, the protoplasts were only washed with solution B once in our study, so the method that we developed can shorten the time required for the protoplasts to remain in the osmotic stabilizer solution, thus reducing the damage to the protoplasts obtained by centrifuging twice. Third, there are some differences between the PMT method that we established in this manuscript and the method used in previous studies. For example, in the transformation procedure, the protoplast mixture was heated at 60 • C for 2 min after the DNA mixture was added into the protoplast STC solution [71,72]; however, no heat treatment is used in our method. Finally, although the flow cytometry-based plating-free system developed by Yang et al. [71] was more convenient than the PMT method used in our manuscript, the selection marker gene was still contained in the transformants generated by Yang et al., while an efficient markerfree gene deletion system was constructed in this manuscript. To increase the efficiency of PMT, we also tried using plasmids and linear DNA in the transformation; we found that the transformation efficiency was far lower for plasmid DNA than for linear DNA. Previous reports have stated that a larger plasmid size results in decreased transformation efficiency [73,74], indicating that it is harder for larger plasmid DNA to transform into cells.
The mode and frequency of individual integration events resulting from homology relate to both the transformation the host itself and the transformation technique [20]. Using AMT, single-copy integration events were detected in many filamentous fungi, including in Verticillium fungicola, A. giganteus, and A. awamori [24,49,75], whereas multiple integration events were frequently observed in the transformants obtained by PMT [49]. The assay of the copy numbers of the integrated gfp gene in Figure 6G and Figure S3 showed that here, AMT resulted in single-copy integration events. Meanwhile, both single and multiple integrations occurred in PMT (in Figure 6F and Figure S3), consistent with previous works [49,75]. Two pathways-the homologous recombination (HR) and NHEJ pathways-have been described as mediating double-strand break repair [20]. HR requires the presence of homologous fragments and thus leads to targeted integration; in contrast, the NHEJ-mediated ligation of DNA strands results in random integration. Very low frequencies of HR result in gene targeting being difficult in filamentous fungi [20]. Therefore, it is likely that the exogenous GFP cassette fragment was randomly integrated into the genomes of A. niger N1 and O1 (which has a Ku70 ortholog) via NHEJ. Meanwhile, a double-strand break generated by the CRISPR/Cas9 system can be repaired by NHEJ to directly carry out indel mutagenesis; alternatively, homology-directed repair (HDR) can be accomplished with a DNA repair template (donor DNA) [44]. Therefore, the deletion of AnalbA using a marker-free CRISPR/Cas9 system in A. niger N1 strain can be achieved by HDR if AnalbA donor DNA is provided with the sgRNA simultaneously.

Conclusions
Here, we targeted two industrial A. niger glucoamylase-producing strains, N1 (which produces few conidia) and O1 (which lacks conidia), for which the existing genetic manipulation methods are laborious. We provided high-efficiency genetic tools via the systematic optimization of the whole process. The method of culturing A. niger N1 and O1 mycelia using solid-medium plates rather than liquid medium for the release of protoplasts provides a case study for manipulating filamentous fungi with few or even no conidia. Following our systematic optimization, the transformation efficiency reached 89.3% for A. niger N1 and 82.1% for A. niger O1 by PMT; when using AMT, it was 98.2% for A. niger N1 and 43.8% for A. niger O1. These are good numbers of positive transformants, especially when considering that the transformation efficiency of A. niger by AMT is normally low [25]. We also developed a marker-free CRISPR/Cas9 genome editing system using an AMA1-based autonomously replicating plasmid to express the Cas9 protein, sgRNA, and the selectable marker hph gene for hygromycin. By using this technology, we constructed a marker-free albA deletion mutant that was visible due to its white coloration in the A. niger N1 strain.
Our results indicate that the approaches that we improved are efficient for the genetic manipulation of A. niger N1 and O1 and will contribute to the use of efficient targeted mutation in other industrial strains of A. niger.

Supplementary Materials:
The following supporting information can be downloaded at https://www. mdpi.com/article/10.3390/biology11101396/s1: Figure S1: The number of conidia for Aspergillus niger N1 and CBS513.88. Figure S2: PCR validation of transformants of A. niger strains N1 and O1. Figure S3: Assay of gfp gene copy number in mutants by RT-qPCR. Figure S4: PMT of A. niger strain N1 targeting the albA gene by CRISPR-Cas9 system. Figure S5: PCR validation of albA-knockout transformants. Table S1: Sequences of primers used in PCR.