Reactive Oxygen Species Distribution Involved in Stipe Gradient Elongation in the Mushroom Flammulina filiformis

The mushroom stipe raises the pileus above the substrate into a suitable position for dispersing spores. The stipe elongates at different speeds along its length, with the rate of elongation decreasing in a gradient from the top to the base. However, the molecular mechanisms underlying stipe gradient elongation are largely unknown. Here, we used the model basidiomycete mushroom Flammulina filiformis to investigate the mechanism of mushroom stipe elongation and the role of reactive oxygen species (ROS) signaling in this process. Our results show that O2− and H2O2 exhibit opposite gradient distributions in the stipe, with higher O2− levels in the elongation region (ER), and higher H2O2 levels in the stable region (SR). Moreover, NADPH-oxidase-encoding genes are up-regulated in the ER, have a function in producing O2−, and positively regulate stipe elongation. Genes encoding manganese superoxide dismutase (MnSOD) are up-regulated in the SR, have a function in producing H2O2, and negatively regulate stipe elongation. Altogether, our data demonstrate that ROS (O2−/H2O2) redistribution mediated by NADPH oxidase and MnSODs is linked to the gradient elongation of the F. filiformis stipe.


Introduction
As a remarkable feature in the development of basidiomycete fruiting bodies, the stipe elongation raises the pileus into a position suitable for spore dispersal and thus ensures the propagation of the species [1,2]. For cultivated edible mushrooms, the stipe elongation rate is important for efficient growth; in addition, stipe length determines product quality by producing the classic cap-and-stem mushroom shape [3]. Moreover, stipe elongation is primarily based on cell elongation, providing an excellent model for studying hyphal cell elongation [4]. Therefore, the mechanisms governing stipe elongation are of great interest in the field of mushroom biology [5][6][7][8][9].
Stipe elongation is non-uniform: the elongation rate decreases in a gradient from the apex to the base of the stipe, as demonstrated in basidiomycete mushrooms [10]. Niu et al. [11] suggested that the gradient stipe elongation rate in Coprinopsis cinerea may result from gradients of cell wall composition, architecture, and thickness from the apical to the basal regions. Our previous results also showed that expression of the genes encoding the enzymes S-adenosylmethionine-dependent methyltransferase and cytochrome c peroxidase was higher in the apical region than in the basal region of Flammulina velutipes A commercial dikaryotic strain of Flammulina filiformis, Fv01 (mated to monokaryotic Fv01-10 and Fv01-N strains, with a white fruiting body), was obtained from the Fujian Edible Fungi Germplasm Resource Collection Center of China. The mycelium was maintained on potato dextrose agar medium (200 g/L potato; 20 g/L glucose; 20 g/L agar) at 25 • C. The fruiting bodies were grown on substrate medium (58% cottonseed hulls, 20% sawdust, 20% wheat bran, 1% calcium sulphate dihydrate, and 1% sucrose (all percentages are w/w), then adjusted to a total moisture content of 60%) and cultivated in bottles according to the method in Park et al. [35]. The temperature was maintained at 10 • C during stipe elongation.

Stipe Elongation Measurement and Cell Length Detection
The stipe was marked in 1.5 cm intervals starting at the apex. Then, the marked fruiting bodies were grown continuously on the substrate medium, and the change in length of each section was measured over time.
Approximately 3 mm long fresh segments of the stipe region were cut using a scalpel and collected; the surface mycelium was examined under an Olympus microscope BX51 (Olympus Corporation, Tokyo, Japan), the micrographs were taken, and the lengths of stipe mycelium cell were measured by Image J 1.51j8 software.

ROS Detection
Nitro blue tetrazolium chloride (NBT, Amresco, Fountain Parkway, Solon, OH, USA) and 3, 3 -diaminobenzidine (DAB) were used for histochemical detection of O 2 − and H 2 O 2 , respectively, according to the method in Kumar et al. [36]. The NBT working solution was 0.05% (w/v), and staining was performed at room temperature with light for 20 min. The DAB working solution was 1 mg/mL, and staining was performed at room temperature without light for 8 h.
The fluorescent probes dihydroethidium (DHE) and 2 , 7 -dichlorodihydrofluorescein diacetate (DCF) were used to detect O 2 − and H 2 O 2 , respectively [37]. The DHE working solution was 10 µM, and staining was performed at room temperature without light for 30 min. The DCF working solution was 10 µM and staining was performed at room temperature without light for 20 min. A Leica SP8 confocal microscope (Leica, Heidelberg, Germany) was used to detect the fluorescence. For DHE staining, dye excitation was set at 300 nm and emitted light was detected at 610 nm. For DCF staining, dye excitation was set at 488 nm and emitted light was detected at 550 nm. The images were processed with Leica Confocal Software.
A commercial oxygen free radical assay kit (Suzhou Comin Biotechnology, Suzhou, China) was used to determine the O 2 − content. O 2 − reacts with hydroxylamine hydrochloride to generate NO 2 − , and NO 2 − forms a red azo compound in the presence of p-aminobenzene sulfonic acid and α-naphthylamine. This compound was measured at 530 nm and the absorbance was used to calculate the O 2 − content (oxidation of hydroxylamine method). A commercial hydrogen peroxide assay kit (Suzhou Comin Biotechnology, Suzhou, China) was used to determine the H 2 O 2 content. H 2 O 2 bound to titanium sulfate forms a yellow titanium peroxide complex, which was measured at 415 nm and the absorbance was used to calculate the H 2 O 2 content (titanium sulfate colorimetric method).

Gene Identification and Sequence Analysis
The amino acid sequences of the genes encoding NADPH oxidase subunits and SOD family proteins of filamentous fungi were downloaded from GenBank and used for BLAST queries with the 14,270 amino acid sequences of the F. filiformis strain Fv01-10 genome (Accession: PRJNA769814) using BLASTP (e-value set as 1 × 10 −2 ). The gene sequences were verified by Sanger sequencing after PCR amplification; the PCR primers were designed using Primer Premier 6.0 and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) (Table S1). Gene structure analysis was performed according to the method of Yan et al. [32] by using strand-specific RNA-seq data of mixed samples of Fv01 mycelium and fruiting bodies [14].

Gene Expression Analysis
A commercial E.Z.N.A. Plant RNA kit (Omega Bio-Tek, Norcross, GA, USA) was used to isolate the total RNA of samples. TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech, Beijing, China) and TransScript All-in-One First-Strand cDNA Synthesis SuperMix for qPCR (One Step gDNA Removal) kits (TransGen Biotech, Beijing, China) were used for cDNA synthesis. Gene expression was detected by quantitative real-time PCR using TransStart Top Green qPCR SuperMix (TransGen Biotech, Beijing, China). The genes encoding glyceraldehyde-3-dehydrogenase (FfGAPDH) and Ras family Small GTPase (FfRas) were used as the reference genes [38]. The qPCR primers were  (Table S2). The relative expression levels were calculated by the 2 −∆∆Ct method [39].

SOD Enzyme Activity Measurement
SOD activity was detected according to the method of Li et al. [40]. Inhibition of the photochemical reduction of NBT was monitored, and one unit of SOD activity was defined as the amount of enzyme required for 50% inhibition of the reduction of NBT monitored at 560 nm.

DPI Treatment
The fruiting body elongation stage was selected for DPI treatment. The stipes were marked at 1.5 cm intervals starting at the junction with the pileus; the marked fruiting bodies were either untreated or soaked in water (two controls), or subjected to 50 µM DPI treatment for 30 min to inhibit NADPH. After that, the fruiting bodies were placed vertically, supported by damp paper, and grown at 10 • C in the dark for 72 h before the level of O 2 − was determined by staining.

Overexpression and RNA Interference Constructs and Fungal Transformation
The binary transformation vector pBHg-eGFP (NCBI accession no. MZ420392), stored in the Mycological Research Center of Fujian Agriculture and Forestry University, Fuzhou, China, was used to construct the RNA interference (RNAi) and overexpression (OE) plasmids. The pBHg-eGFP plasmid included the hygromycin B phosphotransferase (Hyg) gene driven by the Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter from Agaricus bisporus, and the gene encoding enhanced green fluorescent protein (eGFP) driven by the GAPDH promoter from F. filiformis. For construction of the OE plasmid, the eGFP sequence of the pBHg-eGFP plasmid was replaced by the target gene sequence. For the RNAi plasmid, the eGFP sequence was replaced by the hairpin RNA (hpRNA) construct. In detail, the 4th intron (58 bp) of FfNoxA in strain Fv01-10 was selected as the loop of the hpRNA construct. The 4th exon (295 bp) of FfNoxA was selected as the arm of the hpRNA for pBHg-FfNoxA-RNAi plasmid, the coding sequences (without the ATG region) of FfMnSOD1 were selected as the arms of the hpRNA for pBHg-FfMnSOD1-RNAi plasmid, and coding sequences (without the ATG region) of FfMnSOD2 were selected as the arm of the hpRNA for pBHg-FfMnSOD2-RNAi plasmid. All plasmids were verified by sequencing, the schematic representation of OE and RNAi constructs were drawn by SnapGene 1.1.3 ( Figure S1), and the sequences were submitted to the NCBI database (MZ374061-MZ374064).
The plasmids were transformed into F. filiformis dikaryotic strain Fv01 by the Agrobacterium tumefaciens-mediated transformation method according to Wu et al. [41], with minor modifications. First, each recombinant plasmid was introduced into A. tumefaciens strain GV3101, and transformation of modified mycelia pellets was performed. Transformed colonies were selected on CYM medium (0.2% yeast extract, 0.2% tryptone, 1% maltose, 2% glucose, and 2% agar; all percentages are w/v) containing 50 µg/mL hygromycin B four times. The T-DNA insertion site(s) of transformants were verified by whole-genome resequencing according to Chou et al. [42]. Gene expression was normalized to the transcript level of the reference gene FfGAPDH and calculated relative to the transcript level in the wild type (WT) according to the 2 −∆∆CT method [39].

Phenotype Analysis of Transformants
To investigate the phenotype changes of the transformants in the mycelium stage, 6 mm diameter modified mycelia pellets of the WT (strain Fv01) and transformants were inoculated onto a new potato dextrose agar plate and incubated in the dark at 25 • C for 6 days. The O 2 − and H 2 O 2 concentrations were measured by NBT and TMB histochemical detection [43], respectively. The microscopic images were recorded with an Olympus BX51 microscope. To assess the relation between gene function and stipe elongation, the fruiting bodies of the WT and transformants were cultivated under the same conditions as in Park et al. [35]. Fruiting bodies with 5 cm long stipes were selected, the area 1.5 cm away from the stipe apex was marked, the region from the apex to the mark was measured over 6 days, and the elongation rates were calculated. To assess the relation between gene function and stipe elongation ability, the fruiting bodies were cultivated until stipe elongation ceased, and the longest 7 fruiting bodies of each bottle were selected for total stipe length measurement.

Statistical Analysis
The statistical analysis was performed by Graphpad Prism 6.0 (Graphpad software, San Diego, CA, USA). The mean ± S.E.M. was determined for each treatment group in the individual experiments. Statistical significance between samples was investigated by Student's t-test or Tukey's multiple comparisons test.

Gradient Elongation of the F. filiformis Stipe
F. filiformis fruiting body development is divided into primordium, differentiation, elongation, and maturation stages, each accompanied by distinct morphological changes ( Figure S2). The mushroom stipe starts elongating at the differentiation stage; elongation increases at the elongation stage, and then slows down substantially at the maturation stage.
To investigate the elongation pattern of F. filiformis stipes and determine which regions elongate the fastest, we used mushrooms in the elongation stage, when the stipe reached approximately 8.5 cm. Then, we subdivided the stipes from the apex to the base into 1.5 cm regions, labeled them with a marker (Figure 1a), and allowed the fruiting bodies to grow for 72 h to reach the near-maturation stage. Examination of individual regions revealed that the apical 0-1.5 cm region was the most extended, elongating by an additional~3.3 cm, and that the 1.5-3.0 cm region exhibited only~0.2 cm of elongation. By contrast, the 3.0-4.5 cm, 4.6-6.0 cm, and 6.0-7.5 cm regions did not elongate ( Figure 1b). To more precisely determine the main elongating region within the apical 0-1.5 cm and 1.5-3.0 cm regions, we subdivided these regions into 0.3 cm intervals, labeled them, and measured their length at 0 and 8 h of growth. We found that all 0.3 cm intervals in the top 0-1.5 cm region elongated rapidly, with the 0.6-0.9 cm segment extending the most, while the 0.3-0.6 cm and 0.9-1.2 cm segments elongated more slowly (Figure 1c). The  To more precisely determine the main elongating region within the apical 0-1.5 cm and 1.5-3.0 cm regions, we subdivided these regions into 0.3 cm intervals, labeled them, and measured their length at 0 and 8 h of growth. We found that all 0.3 cm intervals in the top 0-1.5 cm region elongated rapidly, with the 0.6-0.9 cm segment extending the most, while the 0.3-0.6 cm and 0.9-1.2 cm segments elongated more slowly (Figure 1c). The elongation speed slowed dramatically in the region located below 1.5 cm, and the stipe stopped extending in the region below 2.4 cm.
These measurements show that the 0.6-1.5 cm region is the fastest elongating segment of the stipe; therefore, we designated this as the elongation region (ER) and designated the slowly elongating region (1.5-2.4 cm) as the transition region (TR) (Figure 1d). The segment with no elongation (5-6 cm) was designated the stable region (SR) (Figure 1d). These regions were chosen for further studies to investigate the molecular mechanism of gradient elongation of the stipe. To confirm our observations, we allowed mushrooms of the same stage to grow for 12 h and determined the elongation speed of the ER, TR, and SR. Our results confirm that the elongation rate of the ER was 2.2 times faster than that of the TR, while the SR had no further elongation ( Figure 1e).  (Figure 2c).

O 2
These results reveal a differential gradient distribution of O 2 − and H 2 O 2 in different regions of the stipe and suggest a potential role of ROS signaling in controlling F. filiformis stipe elongation.

Genes Encoding NADPH Oxidase and MnSODs Are Differentially Expressed in Different Regions of F. filiformis Stipes
− is generated in eukaryotic cells via various mechanisms and NADPH oxidase is among the major O 2 − sources [44,45]. In fungi, NADPH oxidase is a multi-subunit membrane-bound enzyme complex. SOD is the main enzyme that catalyzes the dismuta- [19]. To investigate the O 2 − and H 2 O 2 gradient distribution in the F. filiformis stipe, we examined the expression of the genes encoding NADPH oxidase subunits and SODs. To this end, we sequenced and assembled the F. filiformis genome and used it to bioinformatically identify F. filiformis homologs of NADPH oxidase subunits and SOD family genes. We identified six F. filiformis NADPH oxidase subunit homologs: FfNoxA, FfNoxB, FfNoxR, FfRac1, FfBem1, and FfCdc24. The gene structures are shown in Figure 3a and the GenBank IDs are MN661156-MN661161, respectively. Bioinformatics analysis of the NADPH oxidase subunits showed that all proteins contain complete conserved domains, suggesting that the F. filiformis NADPH oxidase is a functional complex (Table S3). In addition, FfNoxA and FfNoxB subunits were predicted to be plasma membrane proteins with 7 transmembrane helices, consistent with NADPH oxidase localization in the cell membrane (Table S3). We performed qRT-PCR to investigate the expression patterns of the NADPH subunit genes and found that all six genes mimicked the pattern of O 2 − localization, which was highest in the ER and lowest in the SR (Figure 3c). Pearson's correlation coefficient analysis showed that the expression patterns of all six genes were positively correlated with the stipe elongation rate and O 2 − production (Table S4). We also examined O2 − and H2O2 distributions at the cellular resolution by laser scanning confocal microscopy using the fluorescent dyes dihydroethidium (DHE) and 2′, 7′dichlorodihydrofluorescein diacetate (DCF), respectively. Consistent with the histochemical staining, we found that the ER contains a high concentration of O2 − and a low concentration of H2O2, while the SR contains a high concentration of H2O2 and a low concentration of O2 − (Figure 2b). To quantitatively confirm the distributions of O2 − and H2O2 in the stipe, we measured their concentrations by spectrophotometry. The O2 − and H2O2 concentrations differed significantly between the ER and SR, with O2 − gradually decreasing from the ER to the SR, and H2O2 increasing from the ER to the SR (Figure 2c).
These results reveal a differential gradient distribution of O2 − and H2O2 in different regions of the stipe and suggest a potential role of ROS signaling in controlling F. filiformis stipe elongation.

Genes Encoding NADPH Oxidase and MnSODs Are Differentially Expressed in Different Regions of F. filiformis Stipes
O2 − is generated in eukaryotic cells via various mechanisms and NADPH oxidase is among the major O2 − sources [44,45]. In fungi, NADPH oxidase is a multi-subunit membrane-bound enzyme complex. SOD is the main enzyme that catalyzes the dismutation of O2 − into H2O2 [19]. To investigate the O2 − and H2O2 gradient distribution in the F. filiformis stipe, we examined the expression of the genes encoding NADPH oxidase subunits and SODs. To this end, we sequenced and assembled the F. filiformis genome and used it to bioinformatically identify F. filiformis homologs of NADPH oxidase subunits and SOD We then investigated the F. filiformis SOD homologs. We found five genes encoding two SODs of the manganese (Mn) type (FfMnSOD1 and FfMnSOD2, GenBank IDs AYG96714 and AYG96715, respectively), and three SODs of the copper (Cu)/zinc (Zn) type (FfCuZn-SOD1, FfCuZnSOD2, and FfCuZnSOD3, GenBank IDs AYG96711-AYG96713). The gene structures are shown in Figure 3b. Bioinformatics analysis of these genes shows that all the encoded proteins contain complete conserved domains, suggesting that the F. filiformis genome encodes five functional SOD enzymes (Table S3). Subcellular localization analysis predicted that the two FfMnSODs are located in mitochondria with a high prediction score, and the three FfCuZnSODs are most likely located in the cytoplasm. Moreover, we found a secretory Sec/SPI signal peptide in the FfCuZnSOD2 sequence, which indicates that FfCuZnSOD2 might be secreted, unlike the MnSODs. Furthermore, the gene expression patterns showed that both FfMnSOD genes had the highest expression in the SR and the lowest expression in the ER, which show a positive correlation with SOD enzyme activity (Figure 3d,e). Of note, FfMnSOD1 expression was 124 times and 99 times higher in the SR than in the ER and TR, respectively (Figure 3d). The expression patterns of FfCuZnSOD1 and FfCuZnSOD2 are negatively correlated with SOD enzyme activity, while FfCuZnSOD3 shows no significant differences among the different stipe regions (Figure 3d,e). The gene In addition, FfNoxA and FfNoxB subunits were predicted to be plasma membrane proteins with 7 transmembrane helices, consistent with NADPH oxidase localization in the cell membrane (Table S3). We performed qRT-PCR to investigate the expression patterns of the NADPH subunit genes and found that all six genes mimicked the pattern of O2 − localization, which was highest in the ER and lowest in the SR (Figure 3c). Pearson's correlation coefficient analysis showed that the expression patterns of all six genes were positively correlated with the stipe elongation rate and O2 − production (Table S4). We then investigated the F. filiformis SOD homologs. We found five genes encoding two SODs of the manganese (Mn) type (FfMnSOD1 and FfMnSOD2, GenBank IDs AYG96714 and AYG96715, respectively), and three SODs of the copper (Cu)/zinc (Zn) type (FfCuZnSOD1, FfCuZnSOD2, and FfCuZnSOD3, GenBank IDs AYG96711-AYG96713). The gene structures are shown in Figure 3b. Bioinformatics analysis of these genes shows that all the encoded proteins contain complete conserved domains, suggesting that the F. filiformis genome encodes five functional SOD enzymes (Table S3). Subcellular localization analysis predicted that the two FfMnSODs are located in mitochondria with a high prediction score, and the three FfCuZnSODs are most likely located in the cytoplasm. Moreover, we found a secretory Sec/SPI signal peptide in the FfCuZnSOD2 sequence, which indicates that FfCuZnSOD2 might be secreted, unlike the MnSODs. Furthermore, the gene expression patterns showed that both FfMnSOD genes had the highest Altogether, these results suggest that the specific expression of NADPH oxidase genes in the ER and MnSOD genes in the SR could result in the differential distribution of O 2 − and H 2 O 2 in different regions of F. filiformis stipes.

NADPH-Oxidase-Derived O 2 − Positively Regulates Stipe Elongation
To determine if NADPH-oxidase-derived O 2 − regulates stipe elongation, we treated F. filiformis stipes with the NADPH oxidase inhibitor diphenyleneiodonium chloride (DPI) and measured the O 2 − distribution and stipe elongation. We found that O 2 − concentration and stipe elongation were significantly inhibited after pre-incubation in DPI for about 30 min (Figure 4), which indicates that the NADPH oxidase and the O 2 − it generates are important factors in stipe elongation.
To determine if NADPH-oxidase-derived O2 -regulates stipe elongation, we treated F. filiformis stipes with the NADPH oxidase inhibitor diphenyleneiodonium chloride (DPI) and measured the O2 − distribution and stipe elongation. We found that O2 − concentration and stipe elongation were significantly inhibited after pre-incubation in DPI for about 30 min (Figure 4), which indicates that the NADPH oxidase and the O2 − it generates are important factors in stipe elongation. FfNoxA encodes an NADPH oxidase subunit with core transmembrane and catalytic domains [30]. Therefore, to confirm the function of NADPH oxidase in controlling the O2 − concentration and stipe elongation, we created FfNoxA overexpression and knock-down RNAi lines. We obtained five independent overexpression transformants and four independent RNAi transformants. Two of the overexpression transformants (FfNoxA OE#1 and FfNoxA OE#2 ) and two of the RNAi transformants (FfNoxA RNAi#1 and FfNoxA RNAi#2 ) with a verified single T-DNA insertion site for each transformant were selected for further study (Table S5). The qRT-PCR results confirm that the transcript abundance of FfNoxA in the overexpression lines was significantly increased, and that in RNAi lines it was significantly reduced, compared with the WT mushrooms (Figure 5a). FfNoxA encodes an NADPH oxidase subunit with core transmembrane and catalytic domains [30]. Therefore, to confirm the function of NADPH oxidase in controlling the O 2 − concentration and stipe elongation, we created FfNoxA overexpression and knockdown RNAi lines. We obtained five independent overexpression transformants and four independent RNAi transformants. Two of the overexpression transformants (FfNoxA OE#1 and FfNoxA OE#2 ) and two of the RNAi transformants (FfNoxA RNAi#1 and FfNoxA RNAi#2 ) with a verified single T-DNA insertion site for each transformant were selected for further study (Table S5). The qRT-PCR results confirm that the transcript abundance of FfNoxA in the overexpression lines was significantly increased, and that in RNAi lines it was significantly reduced, compared with the WT mushrooms (Figure 5a).
To confirm that the transgenic lines differed in O 2 − production relative to the WT strain, we detected the endogenous O 2 − level via NBT staining in transgenic and WT mycelium. In agreement with the role of NoxA in O 2 − production, as a catalytic subunit of NADPH oxidase [31], FfNoxA overexpression increased O 2 − levels, while RNAi of FfNoxA reduced O 2 − levels compared with WT mycelium (Figure 5b), supporting the idea that NADPH oxidase mediates O 2 − production. To examine whether FfNoxA is involved in stipe elongation, we cultivated FfNoxA overexpression and knock-down lines, grew their fruiting bodies until stipe elongation ceased, and compared them to the WT mushrooms. We found that the transgenic lines displayed dramatic differences in their appearance and the rates of growth compared to the WT. The measurements of the total length of mature fruiting bodies showed that the FfNoxA OE#1 and FfNoxA OE#2 overexpression lines were 28.4% and 11.5% longer than the WT, respectively (Figure 5c,d). The total length of the RNAi lines FfNoxA RNAi#1 and FfNoxA RNAi#2 were 11.7% and 21.4% shorter than the WT (Figure 5c,d).
Additionally, to confirm the function of FfNoxA in stipe elongation, we compared the growth rate of the main elongating stipe region (0-1.5 cm) in the overexpression and knockdown lines relative to the WT mushrooms. The elongation rates of the two overexpression lines were significantly higher than in the WT (0.80 ± 0.01 cm/day and 0.80 ± 0.02 cm/day for FfNoxA OE#1 and FfNoxA OE#2 , respectively) (Figure 5e). The elongation rates of the RNAi lines were 0.62 ± 0.02 cm/day and 0.39 ± 0.03 cm/day for FfNoxA RNAi#1 and FfNoxA RNAi#2 , respectively, and were significantly lower than in the WT (0.75 ± 0.01 cm/day). These data and the observations that inhibition of FfNoxA decreased O 2 − production strongly indicate that FfNoxA and its associated O 2 − production are required for stipe elongation in F. filiformis. To confirm that the transgenic lines differed in O2 − production relative to the WT strain, we detected the endogenous O2 − level via NBT staining in transgenic and WT mycelium. In agreement with the role of NoxA in O2 − production, as a catalytic subunit of NADPH oxidase [31], FfNoxA overexpression increased O2 − levels, while RNAi of FfNoxA reduced O2 − levels compared with WT mycelium (Figure 5b), supporting the idea that

MnSODs Convert O 2 − to H 2 O 2 to Negatively Regulate Stipe Elongation
Since the Mn-type SOD genes FfMnSOD1 and FfMnSOD2 were highly expressed in the SR and reduced in the ER, particularly FfMnSOD1 (Figure 3c), we examined the function of FfMnSOD1 and FfMnSOD2 in controlling the levels of O 2 − and H 2 O 2 as well as mushroom growth. To this end, we created FfMnSOD1 and FfMnSOD2 overexpression and knock-down RNAi lines. We obtained three independent FfMnSOD1 overexpression lines and four independent FfMnSOD1 RNAi lines as well as three independent FfMnSOD2 overexpression lines and two independent FfMnSOD2 RNAi lines.
To confirm that the transformants each contained a single T-DNA insertion site, we conducted genome resequencing and T-DNA insertion analysis (Tables S6 and S7). We selected one overexpression line and two RNAi lines of FfMnSOD1 (FfMnSOD1 OE#1 , FfMnSOD1 RNAi#1 , and FfMnSOD1 RNAi#2 ) for further analysis. For FfMnSOD2, we selected two overexpression lines and one RNAi line (FfMnSOD2 OE#1 , FfMnSOD2 OE#2 , and FfMnSOD2 RNAi#1 ). The qRT-PCR results confirm that, in comparison with the WT, the transcript abundance of the target genes in all overexpression lines was significantly increased, and that in the RNAi lines it was significantly reduced (Figure 6a,c).
A RNAi#2 , respectively, and were significantly lower than in the WT (0.75 ± 0.01 cm/day). These data and the observations that inhibition of FfNoxA decreased O2 − production strongly indicate that FfNoxA and its associated O2 -production are required for stipe elongation in F. filiformis.

MnSODs Convert O2 − to H2O2 to Negatively Regulate Stipe Elongation
Since the Mn-type SOD genes FfMnSOD1 and FfMnSOD2 were highly expressed in the SR and reduced in the ER, particularly FfMnSOD1 (Figure 3c), we examined the function of FfMnSOD1 and FfMnSOD2 in controlling the levels of O2 − and H2O2 as well as mushroom growth. To this end, we created FfMnSOD1 and FfMnSOD2 overexpression and knock-down RNAi lines. We obtained three independent FfMnSOD1 overexpression lines and four independent FfMnSOD1 RNAi lines as well as three independent FfMnSOD2 overexpression lines and two independent FfMnSOD2 RNAi lines.
To confirm that the transformants each contained a single T-DNA insertion site, we conducted genome resequencing and T-DNA insertion analysis (Table S6 and S7). We selected one overexpression line and two RNAi lines of FfMnSOD1 (FfMnSOD1 OE#1 , FfMnSOD1 RNAi#1 , and FfMnSOD1 RNAi#2 ) for further analysis. For FfMnSOD2, we selected two overexpression lines and one RNAi line (FfMnSOD2 OE#1 , FfMnSOD2 OE#2 , and FfMnSOD2 RNAi#1 ). The qRT-PCR results confirm that, in comparison with the WT, the transcript abundance of the target genes in all overexpression lines was significantly increased, and that in the RNAi lines it was significantly reduced (Figure 6a,c).   (Figure 6b,d).
To examine the growth of the FfMnSOD1 and FfMnSOD2 overexpression and knockdown RNAi lines, we allowed them to grow until stipe elongation ceased and compared the transgenic lines to the WT mushrooms (Figure 6e-h). The transgenic lines displayed significant differences in their appearance and growth rates. The stipe length of mature fruiting bodies was shorter in the FfMnSOD1 and FfMnSOD2 overexpression lines than in the WT. Considering that these lines had lower O 2 − levels in the stipes, these results strongly indicate the importance of O 2 − in stipe elongation. As expected, the stipes were longer in the FfMnSOD1 and FfMnSOD2 RNAi lines than in the WT (Figure 6e-h).
These results suggest a critical role of FfMnSOD1

Discussion
Several studies have shown that the stipe elongates faster in the apical region in mushrooms such as F. velutipes [33,46], C. cinerea [47], and A. bisporus [2]. However, the ERs differ among species and the mechanisms of gradient stipe elongation are unknown. For Flammulina, the area of elongation is uncertain. For example, Fang et al. [33] showed that the cell wall extension activity is located exclusively in the 1.0 cm apical region for mushrooms with 3.0 cm long stipes, while Kern et al. [46] found that elongation is restricted to a 0.2 to 0.3 cm apical zone of the stipe. Although these studies show conflicting results, they indicate that the area of elongation could be specific to the growth stage. In our research, we used mushrooms in the elongation stage, when the fruiting bodies reach about 8.5 cm, thus making it easier to study the elongation mechanism. We found that the top 1.5 cm of the stipe elongates quickly, with the 0.6-0.9 cm segment being the most extended, while the regions below 3.0 cm do not elongate. These results also indicate that the stipe ER could be determined by multiple factors, including species, growth stage, and environmental conditions.
Although high levels of ROS (O 2 − and H 2 O 2 ) are harmful to cells, ROS are important signaling molecules necessary for many biological processes [16]. In animal cells, O 2 − signaling regulates cell proliferation and H 2 O 2 signaling supports quiescence [48]. In plants, O 2 − and H 2 O 2 have antagonistic roles in stem cell regulation and the ROS balance defines stem cell fate [44]. It has also been suggested that the distribution of O 2 − and H 2 O 2 is involved in gradient root elongation in A. thaliana [26]. In eukaryotic cells, mitochondria and NADPH oxidase are two major sources of O 2 − [44,45]. It has also been suggested that O 2 − generated from membrane-localized NADPH oxidase plays a key role in fungal cellular differentiation and development [28]. Gradient stipe elongation is an important process in macro fungi; therefore, we investigated the mechanism of this process and examined the role of ROS. In this study, we found that the differential expression of genes encoding NADPH oxidase, which produces O A recent study reported that lateral cell extension is the main mechanism of stipe elongation in mushrooms, and that the extent of cell wall loosening controls the ability of cells to elongate [9]. In plants, the wall-loosening reaction controlling cell elongation is affected by the production of reactive oxygen intermediates [20,21,49]. For example, in maize (Zea mays), O 2 − serves as a precursor for hydroxyl radical (•OH), which loosens cell walls and thus facilitates cell elongation [50]. Here, we demonstrated that the O 2 − generated by NADPH oxidase plays an important role in stipe elongation, and the membrane-bound catalytic subunit of NADPH, NoxA, positively controls stipe elongation. At the same time, H 2 O 2 mediates cell-wall stiffening and decreased cell-wall extensibility in maize coleoptiles [51]. In accord with this, we found that two genes encoding mitochondria-localized FfMnSODs were significantly up-regulated in the SR of the stipe, which also showed very high SOD enzyme activity, suggesting that H 2 O 2 enrichment in the SR and inhibited stipe elongation could be due to mitochondria-produced O 2 − converted to H 2 O 2 by FfMnSODs. However, we also found that genes encoding a different kind of SOD, FfCuZnSOD1 and FfCuZnSOD2, had significantly higher expression in the ER, which is contrary to the ROS distribution between the elongation region (ER) and the stable region (SR). Therefore, what could be the potential role of the FfCuZnSOD genes in the elongating region? We hypothesize that FfCuZnSODs and FfMnSODs play different roles and may be important in cell-wall loosening for elongation due to their distinct cellular localization and properties. Such precedents have been demonstrated in plants. For example, in cotton (Gossypium hirsutum), the extracellular Cu/Zn superoxide dismutase (GhCSD3) specifically translocates to cell walls, despite the absence of a signal peptide. There, it produces high levels of H 2 O 2 and reacts with metal ions, producing •OH in elongating fiber cell walls via the Fenton reaction and loosening cell walls for elongation [52]. Our analysis also showed that FfCuZnSOD2 contains a secretory Sec/SPI signal peptide, making it potentially able to translocate to cell walls, which suggests that the high expression of FfCuZnSOD2 in the elongating region may play a positive role in accelerating cell-wall loosening in the ER. However, further investigation is needed to test this hypothesis.
Moreover, stipe elongation requires not only the extension of cell walls, but also the expansion of cell membranes, the increase of cell cytoplasm, and the polar growth of mycelia, as well as probably more processes. Several studies have also shown that ROS regulates root and pollen tube development (elongation) by activating Ca 2+ and other ion channels in plants, which indicated that ROS-mediated cell elongation is a complex biological process [53][54][55]. Further studies will be needed in the future to reveal the function of different ROS molecular signals in regulating mushroom stipe elongation in more depth.

Conclusions
In summary, our results demonstrate that the interplay between NADPH oxidase and Mn-type SODs plays a critical role in controlling the distribution of two main ROS (O 2 − and H 2 O 2 ) which regulate the gradient elongation of the mushroom stipe. Our findings not only provide insight into the mechanism of mushroom stipe elongation, but also highlight a potential target gene, manipulation of which may benefit mushroom breeding.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/cells11121896/s1, Figure S1: The vector maps of the overexpression and RNAi plasmids constructed in this study. (a,b). The overexpression and RNAi plasmids of FfNoxA. (c,d) The overexpression and RNAi plasmids of FfMnSOD1. (e,f) The overexpression and RNAi plasmids of FfMnSOD2. "Hyg" represents the hygromycin B phosphotransferase gene, "Ab GPD promoter" represents the GAPDH promoter of A. bisporus, "Ff GPD promoter" represents the FfGAPDH promoter, "Ff GPD Intron" represents the first intron of FfGAPDH. The restriction enzyme cutting sites are shown in the maps. Figure S2: Different growth stages of a F. filiformis fruiting body. The fruiting body development beginning from (a) the primordia stage (PR) followed by (b) the differentiation stage (DI) and (c) the elongation stage (EL) and maturation stage (MA). Table S1: Primers used in PCR. Table S2: Primers used in real-time quantitative RT-PCR. Table S3: Bioinformatics analysis of amino acid sequences.