Identification of Aspergillus niger Aquaporins Involved in Hydrogen Peroxide Signaling

Aspergillus niger is a robust microbial cell factory for organic acid production. However, the regulation of many industrially important pathways is still poorly understood. The regulation of the glucose oxidase (Gox) expression system, involved in the biosynthesis of gluconic acid, has recently been uncovered. The results of that study show hydrogen peroxide, a by-product of the extracellular conversion of glucose to gluconate, has a pivotal role as a signaling molecule in the induction of this system. In this study, the facilitated diffusion of hydrogen peroxide via aquaporin water channels (AQPs) was studied. AQPs are transmembrane proteins of the major intrinsic proteins (MIPs) superfamily. In addition to water and glycerol, they may also transport small solutes such as hydrogen peroxide. The genome sequence of A. niger N402 was screened for putative AQPs. Seven AQPs were found and could be classified into three main groups. One protein (AQPA) belonged to orthodox AQP, three (AQPB, AQPD, and AQPE) were grouped in aquaglyceroporins (AQGP), two (AQPC and AQPF) were in X-intrinsic proteins (XIPs), and the other (AQPG) could not be classified. Their ability to facilitate diffusion of hydrogen peroxide was identified using yeast phenotypic growth assays and by studying AQP gene knock-outs in A. niger. The X-intrinsic protein AQPF appears to play roles in facilitating hydrogen peroxide transport across the cellular membrane in both Saccharomyces cerevisiae and A. niger experiments.


Introduction
Aspergillus niger is a potent cell factory that is successfully employed to produce organic acids, including citric-, oxalic-, and gluconic acids [1,2]. Citric-and oxalic acids are intracellularly produced and actively secreted [3][4][5][6]. In contrast, gluconic acid is the product of an extracellular conversion of glucose in a collaboration of three extracellular enzymes: glucose oxidase (GOx), lactonase, and catalase [7]. At an extracellular pH between 4.5 and 6.5 and in the presence of sufficient amounts of oxygen, glucose oxidase catalyzes the oxidation of glucose to gluconolactone and hydrogen peroxide. The co-expressed lactonase subsequently converts gluconolactone to gluconate [7,8].

Strains and Media
All strains used in this study are listed in Table 1. A. niger ATCC 64974 (N402;cspA1) [27] was used as a control strain. A. niger strain MA164.9 (kusA::DR-amdS-DR, pyrG -) is a descendant of N402 [28] and was used to construct AQP knock-outs using the pyrG gene from A. oryzae as a selectable marker gene. The pWay-pyrA plasmid was used as a control for transformation. A. niger strains were maintained on a complete medium (CM) containing 1 g/L casamino acid, 5 g/L yeast extract, 1% glucose, 20 mL/L ASPA + N, 1 mL/L, Vishniac solution, and 1 mM MgSO 4 at 30 • C [29].

Identification of A. niger Aquaporin Sequences
The A. niger N402 genome (Aniger_ATCC64974_N402) with accession number GCA_ 900248155.1 was used to search for aquaporins. The presence of the MIP domain in the resulting putative sequences was identified by Pfam [32], PROSITE [33], and SMART [34]. Transmembrane domains were determined by the HMMTOP transmembrane topology prediction server [35]. The N402 aquaporins were aligned with other fungal aquaporins, selected from Verma et al. [21] using ClustalX [36]. The alignment was applied to the phylogenetic analysis conducted using MEGA version 7.0 [37]. The phylogenetic tree was prepared by the neighbor-joining method with 1000 bootstrap replications referring to the reliability tests of an inferred tree [38].

Heterologous Expression of A. niger Aquaporins in S. cerevisiae
Spores of A. niger N402 were grown in CM and incubated at 30 • C for 40 h. After that, the mycelium was harvested and then used in the RNA extraction, following the protocol provided in the Maxwell ® 16 LEV simplyRNA Cells Kit (Promega, Madison, WI, USA). RNA concentrations were measured by NanoDrop, and purified RNA was stored at −80 • C. The purified RNA was used to synthesize cDNA following the RevertAid H Minus First Strand cDNA Synthesis Kit (ThermoFisher, Waltham, MA, USA). The reaction mixture was incubated at 42 • C for 1 h and then terminated by heating at 70 • C for 5 min. The cDNA was stored at −20 • C.
The aquaporin genes were cloned into an overexpression vector using the Yeast ToolKit (YTK) described by Lee et al. [39]. In short, the A. niger aquaporin gene sequences were first inspected to find internal restriction sites (IRS) for the enzymes used in the YTK. Specific primers (Supplementary Table S1) were designed to remove these IRS by introducing synonymous substitutions. The cDNAs of A. niger N402 and S. cerevisiae BY4741 were used as templates for the amplification of the aquaporin genes and the FPS1 gene, respectively. Q5 high-fidelity DNA polymerase (New England BioLabs, Ipswich, MA, USA) was used in the PCR reaction, following the manufacturer's protocol. Amplified fragments were purified by NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel, Düren, Germany), and DNA concentrations were measured by NanoDrop.
The aquaporin genes were cloned into the YTK entry vector, according to Lee et al. [39]. Each restriction-ligation product was transformed into E. coli DH5α and grown at 37 • C on LB agar supplemented with 25 µg/mL of chloramphenicol.
Plasmids were isolated from an overnight culture of single colonies using GeneJET Plasmid Miniprep Kit (Thermo Scientific, Waltham, MA, USA) according to the manufacturer's protocol. The integrity of the plasmids was checked by restriction digestion analysis and confirmed by Sanger sequencing (GATC Biotech, Landkreis Ebersberg, Germany).
To assemble the yeast expression plasmids, the restriction-ligation reactions contained 20 fmol of each DNA module (promoter pRPL18B, aquaporin gene, and terminator tPGK1) and 40 fmol of backbone plasmid pMV009. After the reaction, the gene-cassettes were transformed into E. coli DH5α and grown at 37 • C on LB agar supplemented with 100 µg/mL of ampicillin. A plasmid map is shown in Supplementary Figure S1. The yeast expression plasmids were transformed into S. cerevisiae BY4741 electrocompetent cells as described by Suga et al. [31,40] (Table 1).

Growth and Hydrogen Peroxide Sensitivity Assays in S. cerevisiae
Yeast cells were grown in the selective YNB medium at 30 • C and 250 rpm for 18 h. Overnight cultures were diluted in fresh medium to an OD 600 of 0.1, and subsequently, serial dilutions were made. Five-µL of the diluted cell suspensions was spotted onto the solid selective YNB medium containing different concentrations of hydrogen peroxide (0-2 mM) and incubated at 30 • C. Growth and survival were scored after 6 days of incubation. For Ag + treatments, the cell suspensions were spotted on the solid selective YNB medium supplemented with either no or 1 mM hydrogen peroxide and various concentrations of AgNO 3 (0-15 µM). The plates were incubated at 30 • C and then incubated for 6 days before growth and survival were scored. The experiments were performed in triplicate.
The transport of hydrogen peroxide was studied using a fluorescence-based assay adapted from Bienert et al. [18]. Yeast cells were pre-cultured on the solid selective YNB medium for 2 days. Three-mL liquid cultures were inoculated with a single colony and supplemented with 2 ,7 -dichlorodihydrofluorescein diacetate (DCFHDA; Sigma-Aldrich, St. Louis, MO, USA) with a final concentration of 1 µM. Cells were grown at 30 • C and 250 rpm in darkness overnight to an OD 600 of 1.6. The cells were washed five times in 20 mM HEPES buffer, pH 7.0, and finally resuspended in HEPES buffer at an OD 600 of 1.4. The yeast cell suspensions were aliquoted into 96-well plates with 200 µL per well. The suspensions were followed over time at room temperature using a microplate reader with fluorescent mode (Synergy TM Mx Monochromator-Based Multi-Mode Microplate Reader; BioTek, Broadview, IL, USA) at excitation/emission of 492/527 nm. After the initial (t = 0) measurements, hydrogen peroxide was automatically dispensed in the wells at concentrations of 0.1, 0.5, and 1.0 mM. The OD 600 and fluorescence intensity were recorded every minute within 2 h. A heatmap was generated from an average fluorescent intensity per OD 600 from the experiment with two biological and two technical replicates.

Construction of A. niger AQP Knock-Out Strains
The AQP knock-out strains were constructed using the split-marker approach [29,41]. Three selected aquaporin genes: aqpD, aqpE, and aqpF, were individually deleted from the genome of the A. niger MA169.4 strain, which is defective in the Non-Homologous End-Joining (NHEJ) pathway through a transiently silenced kusA gene [41]. The preparation of DNA fragments and protoplasts and transformation steps were done according to Laothanachareon et al. [8]. The list of the primers used to construct the knock-out strains and to confirm the correct deletion of the genes by PCR is shown in Supplementary Table S1.

Transcriptional Analysis of Aquaporins in A. niger
Shake-flasks were used to monitor the expression of A. niger aquaporin genes upon hydrogen peroxide treatment. Using an inoculation of 1 × 10 6 spores/mL, A. niger wildtype N402 and knock-out strains were pre-cultured for 18 h in minimal medium (MM) containing 4.5 g/L NaNO 3 , 1.13 g/L KH 2 PO 4 , 0.38 g/L KCl, 0.38 g/L MgSO 4 .7H 2 O, 2 g/L casamino acid, 1 g/L yeast extract, 1 mL/L Vishniac trace element solution, and supplemented with 50 mM fructose as a carbon source, at 30 • C and 200 rpm. Mycelium was harvested, rinsed with water, and then transferred to fresh MM supplemented with 50 mM fructose at an initial pH of 6.0. The experiments were performed in the presence of various concentrations of hydrogen peroxide (between 0 and 75 mM, see Section 3) and with various incubation times (between 0 and 3 h, see Section 3). After incubation, the mycelium samples were taken for RNA isolation, quickly washed, and then dried with a single-use towel, snap-frozen with liquid nitrogen, and stored at -80 • C until further processing. In all cases, two biological replicates with three technical replicates per condition were studied.
RNA was isolated from mycelium, as described by Sloothaak et al. [42]. Quantitative realtime PCR (RT-qPCR) and calculations were executed following the protocols and instrument setup as previously described by Mach-Aigner et al. [43]. Primer sequences are listed in Supplementary Table S1. Cycling conditions and control reactions were as described previously by Steiger et al. [44].
The histone H4-like transcript (ATCC64974_101030, ATCC 1015 gene ID 207921) was used to normalize the RT-qPCR expression data. The uninduced state (no addition of hydrogen peroxide) was used to compare expression levels.

The Genome of A. niger Harbors at Least Seven Putative aqp Genes
Using sequence similarity methods and by comparing the conservation of encoded structural protein features, seven genes were identified in the A. niger genome sequence and named aqpA-aqpG (Table 2 and Supplementary Dataset S1). The length of the deduced protein sequences ranged from 250 to 617 amino acids. All deduced proteins have the MIP protein family signatures with six transmembrane domains except for AQPC and AQPG, presenting only five domains. The presence of the asparagine-proline-alanine sequences (NPA motifs), highly conserved in the aquaporin water channel family, was also analyzed (Supplementary Dataset S1). Two NPA motifs were found in AQPA and AQPD, one in AQPB and AQPF, whilst AQPC, AQPE, and AQPG do not contain this motif. Unusual NPA substitutions and alternative NPA-like loops were found in some of the sequences. The previously reported substitution NPS (asparagine-proline-serine) was found twice in AQPE and once in AQPG. The rest of the present NPA-like loops consist of NP-and N-A residues only (Supplementary Dataset S1). Exceptionally, AQPF showed only one single NPA motif. Taken together, protein sequence analyses suggested that the seven selected genes most likely encoded A. niger AQPs.

Classification of A. niger AQPs
The seven AQPs of A. niger N402 were classified based on amino acid sequence similarity with other (fungal) AQPs ( Figure 1). AQPD grouped with the γ-AQGPs, which are related to the β-AQGPs and the yeast FPS1-like AQGPs. AQPB and AQPE seem to be related to the Yfl054-like AQGPs. AQPA is the only representative of the orthodox AQPs and is in the same group as AQP8 of Botrytis cinerea (XP_001547129) [19]. The two remaining AQPs, AQPC and AQPF, appeared to belong to the X-Intrinsic Proteins (XIPs) clade. Ambiguously, AQPG appears to be related to the XIP class but is separate from the well-characterized groups.

AQP D, E, and F Facilitate Hydrogen Peroxide Import in Saccharomyces cerevisiae
S. cerevisiae transformants expressing A. niger AQPs were constructed to investigate their abilities in hydrogen peroxide import by indirect and direct experiments. In the indirect experiments, colony growth was monitored in the presence of 0-2 mM hydrogen peroxide ( Figure 2). The S. cerevisiae strain used as a negative control contained the 'empty' pMV009 vector expressing a GFP control sequence, whereas the strain overexpressing the yeast aquaporin FPS1 worked as a positive control [13,19]. Due to their ability to facilitate the import of toxic amounts of hydrogen peroxide, the growth of the strains individually 3.3. AQP D, E, and F Facilitate Hydrogen Peroxide Import in Saccharomyces cerevisiae S. cerevisiae transformants expressing A. niger AQPs were constructed to investigate their abilities in hydrogen peroxide import by indirect and direct experiments. In the indirect experiments, colony growth was monitored in the presence of 0-2 mM hydrogen peroxide ( Figure 2). The S. cerevisiae strain used as a negative control contained the 'empty' pMV009 vector expressing a GFP control sequence, whereas the strain overexpressing the yeast aquaporin FPS1 worked as a positive control [13,19]. Due to their ability to facilitate the import of toxic amounts of hydrogen peroxide, the growth of the strains individually carrying the A. niger aqpD, aqpE, or aqpF genes was severely inhibited, starting at 1 mM hydrogen peroxide with a cell concentration below OD 600 of 0.1 ( Figure 2). However, the strains expressing aqpA, aqpB, aqpC, or aqpG showed a growth pattern similar to that displayed by the negative control. The positive control FPS1 showed inhibition at 1.5 mM hydrogen peroxide. At the highest concentration, 2 mM, all strains, including the negative control, were affected, although some of them still appeared to show some growth when high cell concentrations were spotted.
A fluorescence−based assay [18] was used to directly follow the transport of hydrogen peroxide across the yeast plasma membrane over time. Here, 2′−7′− dichlorodihydrofluorescein diacetate (DCFHDA) was used as a detector. The DCFHDA fluorescent dye is cell−permeable ROS−sensitive. In its normal acetylated form, the dye can diffuse into the cells. Deacetylation traps the fluorochrome inside the cells and makes it susceptible to oxidation by ROS [18]. The fluorescence intensity of each AQP expressing yeast strain under the three different hydrogen peroxide concentrations of 0.1, 0.5, and 1.0 mM were compared over a 2 h period. The data were calculated as fluorescence intensity per OD600nm using the condition without hydrogen peroxide to normalize the data (Figure 3). The strain transformed with the expression vector pMV009 was used as a negative control. At a concentration of 0.1 mM hydrogen peroxide, the expression of the aqpD gene−mediated transport of hydrogen peroxide into the cells already during the first hour, whereas higher hydrogen peroxide concentrations resulted in a slight decrease in the fluorescence signal. The aqpA and aqpF−expressing strains showed the same trend, but compared to the aqpD−expressing strain, their fluorescence signals were weaker. In the aqpE expressing strain, the fluorescence intensity barely increased during the first hour, although it started to be noticeable during the second hour. Obviously, the yeast strains carrying the aqpB, aqpC, or aqpG gene were not able to take up hydrogen peroxide.  A fluorescence-based assay [18] was used to directly follow the transport of hydrogen peroxide across the yeast plasma membrane over time. Here, 2 -7 -dichlorodihydrofluorescein diacetate (DCFHDA) was used as a detector. The DCFHDA fluorescent dye is cell-permeable ROS-sensitive. In its normal acetylated form, the dye can diffuse into the cells. Deacetylation traps the fluorochrome inside the cells and makes it susceptible to oxidation by ROS [18]. The fluorescence intensity of each AQP expressing yeast strain under the three different hydrogen peroxide concentrations of 0.1, 0.5, and 1.0 mM were compared over a 2 h period. The data were calculated as fluorescence intensity per OD 600nm using the condition without hydrogen peroxide to normalize the data (Figure 3). The strain transformed with the expression vector pMV009 was used as a negative control. At a concentration of 0.1 mM hydrogen peroxide, the expression of the aqpD gene-mediated transport of hydrogen peroxide into the cells already during the first hour, whereas higher hydrogen peroxide concentrations resulted in a slight decrease in the fluorescence signal. The aqpA and aqpF-expressing strains showed the same trend, but compared to the aqpDexpressing strain, their fluorescence signals were weaker. In the aqpE expressing strain, the fluorescence intensity barely increased during the first hour, although it started to be noticeable during the second hour. Obviously, the yeast strains carrying the aqpB, aqpC, or aqpG gene were not able to take up hydrogen peroxide.
in their sequences that are targets for Ag + inhibition (Supplementary Dataset S1). Increasing concentrations of Ag + can cause a significant decrease in growth and cell viability [18]. In this study, the toxicity of Ag + to yeast became evident at around 15 µM, which was higher than the previously reported 6 µM [45] and lethal at 30 µM. The hydrogen peroxide−induced growth phenotype of the yeast strains expressing aqpD and aqpE was considerably improved in the presence of Ag + starting at 3.7 µM. In contrast, no restoration of growth was observed for the aqpF expressing strain. In the presence of Ag + , the negative control strain showed a similar growth phenotype in the absence and presence of hydrogen peroxide. This was also the case for AQP−expressing yeast strains previously found not to be growth inhibited by the addition of hydrogen peroxide (Figure 4).  Apart from hydrogen peroxide transport analysis under various hydrogen peroxide concentrations, we tested whether the specific aquaporin was responsible for the facilitated hydrogen peroxide transport. In this experiment, silver ions (Ag + ) were used as an aquaporin inhibitor since it was reported that Ag + could significantly inhibit the transmembrane flux of hydrogen peroxide by binding to cysteine or histidine residues in protein [18]. The A. niger AQPs contain 2-8 but structurally not conserved cysteine residues in their sequences that are targets for Ag + inhibition (Supplementary Dataset S1). Increasing concentrations of Ag + can cause a significant decrease in growth and cell viability [18]. In this study, the toxicity of Ag + to yeast became evident at around 15 µM, which was higher than the previously reported 6 µM [45] and lethal at 30 µM. The hydrogen peroxide-induced growth phenotype of the yeast strains expressing aqpD and aqpE was considerably improved in the presence of Ag + starting at 3.7 µM. In contrast, no restoration of growth was observed for the aqpF expressing strain. In the presence of Ag + , the negative control strain showed a similar growth phenotype in the absence and presence of hydrogen peroxide. This was also the case for AQP-expressing yeast strains previously found not to be growth inhibited by the addition of hydrogen peroxide (Figure 4).

Transcriptional Analysis of AQPD, AQPE, and AQPF Single Knock−Out Strains
Single knock−out strains of aqpD, aqpE, and aqpF, actively facilitating hydrogen peroxide transport in yeast, were constructed. As hydrogen peroxide is both a reactive oxygen species known to cause damage to cellular components in high concentrations as well as a second messenger, we first analyzed AQP gene expression in the presence of various concentrations of hydrogen peroxide in the medium. Mycelium was incubated for 3 h with three different initial hydrogen peroxide concentrations. Fructose was used as a carbon source to avoid the generation of extracellular hydrogen peroxide by GOx. Upon the addition of hydrogen peroxide to the medium, the transcript levels of all AQP genes in-

Transcriptional Analysis of AQPD, AQPE, and AQPF Single Knock-Out Strains
Single knock-out strains of aqpD, aqpE, and aqpF, actively facilitating hydrogen peroxide transport in yeast, were constructed. As hydrogen peroxide is both a reactive oxygen species known to cause damage to cellular components in high concentrations as well as a second messenger, we first analyzed AQP gene expression in the presence of various concentrations of hydrogen peroxide in the medium. Mycelium was incubated for 3 h with three different initial hydrogen peroxide concentrations. Fructose was used as a carbon source to avoid the generation of extracellular hydrogen peroxide by GOx. Upon the addition of hydrogen peroxide to the medium, the transcript levels of all AQP genes increased ( Figure 5). As transcript levels of most AQP genes appeared to peak upon incubation with 10 mM hydrogen peroxide, a range of 0-10 mM hydrogen peroxide was chosen for transcriptional analysis of the AQP knock-out strains.
increased during the first hour after the addition of hydrogen peroxide ( Figure 6) thereafter decreased. After 3 h, the system was returned to the uninduced state. Ov we observed no or little cross−regulation. In the ΔaqpD background, transcript leve the aqpE and goxC genes showed some increase after one hour, whereas the level o aqpF gene was stable ( Figure 6B). In the ΔaqpE strain, the expression pattern of the gene was similar to that of the N402 control, while the two− and three−hour trans levels of the aqpF gene were below the uninduced values ( Figure 6C). The ΔaqpF s behaved the same as N402, although the transcript levels of the goxC gene showed a lo expression level 1 h after the induction started ( Figure 6D).  The growth phenotype of the three AQP knock-out strains was not noticeably different from the parental strain. The wild-type N402 and the three AQP knock-out strains were treated with a 10 mM hydrogen peroxide pulse and followed over time for 3 h by RT-qPCR. The goxC gene encoding glucose oxidase was used as a reporter gene to monitor hydrogen peroxide transport to the cell because, as a second messenger system, hydrogen peroxide can directly induce the expression of the glucose oxidase gene [8]. Due to the coinduction of extracellular CATR catalase activity [8], the hydrogen peroxide concentration will decrease over time. Relative transcript abundances (log10) of four genes: aqpD, aqpE, aqpF, and the goxC gene, were measured. The transcript levels of all genes increased during the first hour after the addition of hydrogen peroxide ( Figure 6) and thereafter decreased. After 3 h, the system was returned to the uninduced state. Overall, we observed no or little cross-regulation. In the ∆aqpD background, transcript levels of the aqpE and goxC genes showed some increase after one hour, whereas the level of the aqpF gene was stable ( Figure 6B). In the ∆aqpE strain, the expression pattern of the goxC gene was similar to that of the N402 control, while the two-and three-hour transcript levels of the aqpF gene were below the uninduced values ( Figure 6C). The ∆aqpF strain behaved the same as N402, although the transcript levels of the goxC gene showed a lower expression level 1 h after the induction started ( Figure 6D). Samples were taken before (0) or at various times after the addition of 10 mM hydrogen peroxide. The expression analyses were performed by RT-qPCR with primers specific for aqpD (white background filled with dots), aqpE (gray bar), aqpF (white background filled with stripes), and goxC (black). The goxC gene was used as a proxy for hydrogen peroxide uptake. The data presented is the mean of two biological replicates using a logarithmic scale (log10).
The responses of the wild−type and knock−out strains toward various initial concentrations of extracellular hydrogen peroxide were investigated further. Even though the expression levels of all genes were already decreasing 30 min after hydrogen peroxide addition, a fixed incubation time of 1 h was used for further studies because it showed significant differences in the expression of each gene. Samples were obtained from mycelium induced by initial hydrogen peroxide concentrations of 0 to 10 mM (Figure 7). In the N402 strain, all genes showed an increasing upregulation along with the increase in initial hydrogen peroxide concentration ( Figure 7A). In the aqpD deletion strain, the lowest hydrogen peroxide concentration (0.2 mM) was not able to induce the expression of any of the studied genes; however, one hour after adding an initial hydrogen peroxide concentration of 10 mM increased transcription levels of aqpE and goxC were observed ( Figure  7B). In the wild−type strain, aqpD gene transcript levels hardly respond to varying hydrogen peroxide concentrations (Figures 5 and 7), and in yeast, AQPD mediates the transport of hydrogen peroxide into the cells already at a concentration as low as 0.1 mM hydrogen peroxide (Figure 4). Taken together, it appears that upon deletion of the aqpD gene, the GOx system is less sensitive to lower hydrogen peroxide concentrations. The transcription pattern of the aqpD and aqpF genes in the ΔaqpE background was not clearly dependent on the H2O2 concentration, while the expression pattern of the goxC gene was comparable to that observed in the N402 strain. The strain carrying a deletion in the aqpF gene was unable to induce the expression of all genes in the presence of the lowest hydrogen peroxide concentration. However, while the aqpD and aqpE genes recovered a normal expression pattern in the presence of 2 and 10 H2O2 mM, the goxC expression levels remained reduced compared to the wild−type strain ( Figure 7D). Samples were taken before (0) or at various times after the addition of 10 mM hydrogen peroxide. The expression analyses were performed by RT-qPCR with primers specific for aqpD (white background filled with dots), aqpE (gray bar), aqpF (white background filled with stripes), and goxC (black). The goxC gene was used as a proxy for hydrogen peroxide uptake. The data presented is the mean of two biological replicates using a logarithmic scale (log10).
The responses of the wild-type and knock-out strains toward various initial concentrations of extracellular hydrogen peroxide were investigated further. Even though the expression levels of all genes were already decreasing 30 min after hydrogen peroxide addition, a fixed incubation time of 1 h was used for further studies because it showed significant differences in the expression of each gene. Samples were obtained from mycelium induced by initial hydrogen peroxide concentrations of 0 to 10 mM (Figure 7). In the N402 strain, all genes showed an increasing upregulation along with the increase in initial hydrogen peroxide concentration ( Figure 7A). In the aqpD deletion strain, the lowest hydrogen peroxide concentration (0.2 mM) was not able to induce the expression of any of the studied genes; however, one hour after adding an initial hydrogen peroxide concentration of 10 mM increased transcription levels of aqpE and goxC were observed ( Figure 7B). In the wild-type strain, aqpD gene transcript levels hardly respond to varying hydrogen peroxide concentrations ( Figures 5 and 7), and in yeast, AQPD mediates the transport of hydrogen peroxide into the cells already at a concentration as low as 0.1 mM hydrogen peroxide ( Figure 4). Taken together, it appears that upon deletion of the aqpD gene, the GOx system is less sensitive to lower hydrogen peroxide concentrations. The transcription pattern of the aqpD and aqpF genes in the ∆aqpE background was not clearly dependent on the H 2 O 2 concentration, while the expression pattern of the goxC gene was comparable to that observed in the N402 strain. The strain carrying a deletion in the aqpF gene was unable to induce the expression of all genes in the presence of the lowest hydrogen peroxide concentration. However, while the aqpD and aqpE genes recovered a normal expression pattern in the presence of 2 and 10 H 2 O 2 mM, the goxC expression levels remained reduced compared to the wild-type strain ( Figure 7D). . Samples were taken one hour after the addition of 0−10 mM hydrogen peroxide. The expression analyses were performed by RT−qPCR with primers specific for aqpD (white background filled with dots), aqpE (gray bar), aqpF (white background filled with stripes), and goxC (black). goxC gene expression was used as a proxy for hydrogen peroxide uptake. The data presented is the mean of two biological replicates using a logarithmic scale (log10).

Discussion
After the discovery of the first AQP over three decades ago, many members of the MIPs superfamily have been identified, cloned, and functionally studied [13]. However, information pertaining to the functional role of AQPs in ROS transport in fungal organisms is limited and currently absent for A. niger. In this study, we identified in the genome of A. niger ATCC64974 seven putative AQP genes. Comparative sequence analysis revealed that the encoded A. niger AQPs could be divided into three subclasses: one orthodox AQP (AQPA), three aquaglyceroporins (AQPB, AQPD, and AQPE), and two fungal XIPs (AQPC and AQPF). AQPG appeared to have only five transmembrane domains, and the conserved NPA−like sequences in loops B and E are absent in AQPG, and although AQPG showed sequence similarity with XIP AQPs, it could not be directly linked to a known AQP subclass (Figure 1) The genomes of Trichoderma spp. also contain a similar number of aquaporin genes. The class distribution is, however, different. There are three orthodox AQPs, three AQGPs, and one XIP in Trichoderma ssp. [20]. In A. niger, only AQPA belongs to the orthodox AQP subclass. Orthodox AQPs are considered to be specific water channels [46] and therefore play important roles in cell osmoregulation [47,48]. Yeast cells expressing the A. niger aqpA gene can grow on plates containing high concentrations of hydrogen peroxide and/or in the presence of Ag + . However, aqpA−expressing yeast cells oxidized the ROS−sensitive fluorescent dye in the liquid medium, suggesting that AQPA encodes an orthodox water channel that can but does not efficiently facilitate the transport of hydrogen peroxide into the cells.
According to their amino acid sequence, AQPC, AQPD, and AQPE are aquaglyceroporins, which are expected to transport glycerol, urea, and other small solutes across . Samples were taken one hour after the addition of 0-10 mM hydrogen peroxide. The expression analyses were performed by RT-qPCR with primers specific for aqpD (white background filled with dots), aqpE (gray bar), aqpF (white background filled with stripes), and goxC (black). goxC gene expression was used as a proxy for hydrogen peroxide uptake. The data presented is the mean of two biological replicates using a logarithmic scale (log10).

Discussion
After the discovery of the first AQP over three decades ago, many members of the MIPs superfamily have been identified, cloned, and functionally studied [13]. However, information pertaining to the functional role of AQPs in ROS transport in fungal organisms is limited and currently absent for A. niger. In this study, we identified in the genome of A. niger ATCC64974 seven putative AQP genes. Comparative sequence analysis revealed that the encoded A. niger AQPs could be divided into three subclasses: one orthodox AQP (AQPA), three aquaglyceroporins (AQPB, AQPD, and AQPE), and two fungal XIPs (AQPC and AQPF). AQPG appeared to have only five transmembrane domains, and the conserved NPA-like sequences in loops B and E are absent in AQPG, and although AQPG showed sequence similarity with XIP AQPs, it could not be directly linked to a known AQP subclass ( Figure 1).
The genomes of Trichoderma spp. also contain a similar number of aquaporin genes. The class distribution is, however, different. There are three orthodox AQPs, three AQGPs, and one XIP in Trichoderma ssp. [20]. In A. niger, only AQPA belongs to the orthodox AQP subclass. Orthodox AQPs are considered to be specific water channels [46] and therefore play important roles in cell osmoregulation [47,48]. Yeast cells expressing the A. niger aqpA gene can grow on plates containing high concentrations of hydrogen peroxide and/or in the presence of Ag + . However, aqpA-expressing yeast cells oxidized the ROS-sensitive fluorescent dye in the liquid medium, suggesting that AQPA encodes an orthodox water channel that can but does not efficiently facilitate the transport of hydrogen peroxide into the cells.
According to their amino acid sequence, AQPC, AQPD, and AQPE are aquaglyceroporins, which are expected to transport glycerol, urea, and other small solutes across cell membranes [14,49]. Based on sequence similarity, the A. niger aquaglyceroporins could be further subclassified: AQPB and AQPE are Yfl054-like aquaglyceroporins, while AQPD was classified as a γ-aquaglyceroporin. Yfl054-like aquaglyceroporins are found in both yeasts and filamentous fungi [22]. The Yfl054-like subgroup is characterized by an N-terminal extension of around 350 amino acids harboring the PVWSLXXPLPV motif and a C-terminal extension of around 50 amino acids. In filamentous fungi, this motif is partly conserved in A. nidulans and Fusarium gramineum Yfl054-like aquaglyceroporins [22]. PVWSLXXPLPV motif sequences were also found in the N-terminal extensions of the Yfl054-like aquaglyceroporins of A. niger (Supplementary Dataset S1). Although the functions of Yfl054-like aquaglyceroporins are still poorly described, they have been postulated as functional glycerol facilitators [50]. In addition, the Yfl054-like aquaglyceroporins play a specific role related to transmembrane solute fluxes, and they may be involved in regulatory processes through their long N-terminus based on their conservation of domain structure and sequence [22]. In our yeast experiments, only AQPE seemed to be able to facilitate hydrogen peroxide transport.
Only one γ-AQGP, AQPD, was found in A. niger. This subgroup can be further subdivided into γ1 and γ2 AQGPs. The γ1 AQGP subclass is found in the species of Mucoromycotina [26], while the γ2 AQGP subclass is found in filamentous Ascomycota [21]. The AQPD of A. niger N402 is a member of the γ2-AQGPs. Conserved sequence motifs have been identified in loop B and E regions of γ2-AQGPs [21]. Accordingly, these motifs were also present in AQPD (Supplementary Dataset S1). In this study, AQPD was found to be able to transport hydrogen peroxide in yeast. Since the expression level of AQPD in N402 appeared to be identical under various hydrogen peroxide concentrations, it could act as to be constitutively expressed AQP facilitating hydrogen peroxide transport (Figure 7). AQPC and AQPF belong to the XIP subfamily. XIPs are commonly found in protozoa, plants, and fungi but not in bacteria and animals [51]. Fungal XIPs are frequently found in Ascomycota, Basidiomycota, and Microsporidia [25]. The XIPs have conserved motifs in loops B and E [21]. These sequence motifs were present in AQPF, whereas a single motif was found in loop B of AQPC (Supplementary Dataset S1). The biological functions and roles of fungal XIPs are still enigmatic. In plants, the XIPs are expected to facilitate the transfer of solutes such as urea, glycerol, hydrogen peroxide, boric acid, and ammonia because of their hydrophobic selectivity property that was characterized in Populus [52] and Solanaceae [53]. No transport of water could be observed by Solanaceae XIPs [53]. In contrast, two Populus XIPs apparently facilitate water transport [51]. The yeast phenotypic growth assays showed that only AQPF is able to facilitate hydrogen peroxide transport in yeast.
Hydrogen peroxide is a by-product of many intracellular and extracellular oxidative reactions, including the GOx system of A. niger. The A. niger GOx system uses the hydrogen peroxide by-product of the extracellular enzymatic conversion of glucose to gluconate as a second messenger to further induce the expression of the GOx system [8]. The facilitated diffusion of hydrogen peroxide into the cell could be explained by the expression of specific AQPs. Expression analysis of A. niger AQPs in the wild-type and AQP knock-out strains showed upregulation of all identified AQPs upon the addition of varying concentrations of hydrogen peroxide, while yeast phenotypic growth assays suggested that at least three A. niger AQPs: AQPD, AQPE, and AQPF, can transport hydrogen peroxide. Two of them, AQPD and AQPF, appear to play a more prominent role in the amplification of the hydrogen peroxide signal of the GOx system. A knock-out of AQPD showed a reduced sensitivity in the GOx system towards the lower hydrogen peroxide concentrations, and thus the constitutively expressed AQPD may play a role in the initial amplification of the GOx signal, while the AQPF knock-out has a major negative effect on goxC expression at the higher hydrogen peroxide concentrations.