The Bicarbonate Transporter (MoAE4) Localized on Both Cytomembrane and Tonoplast Promotes Pathogenesis in Magnaporthe oryzae

Bicarbonate (HCO3−) transporter family including the anion exchanger (AE) group is involved in multiple physiological processes through regulating acid-base homeostasis. HCO3− transporters have been extensively studied in mammals, but fungal homologues of AE are poorly understood. Here, we characterized the AE group member (MoAE4) in Magnaporthe oryzae. MoAE4 exhibits more sequence and structure homologies with the reported AE4 and BOR1 proteins. In addition to the common sublocalization on cytomembrane, MoAE4 also localizes on tonoplast. Yeast complementation verified that MoAE4 rescues boron sensitivity and endows NaHCO3 tolerance in the BOR1 deleted yeast. MoAE4 gene is bicarbonate induced in M. oryzae; and loss of MoAE4 (ΔMoAE4) resulted in mycelial growth inhibited by NaHCO3. Lucigenin fluorescence quenching assay confirmed that ΔMoAE4 accumulated less HCO3− in vacuole and more HCO3− in cytosol, revealing a real role of MoAE4 in bicarbonate transport. ΔMoAE4 was defective in conidiation, appressorium formation, and pathogenicity. More H2O2 was detected to be accumulated in ΔMoAE4 mycelia and infected rice cells. Summarily, our data delineate a cytomembrane and tonoplast located HCO3− transporter, which is required for development and pathogenicity in M. oryzae, and revealing a potential drug target for blast disease control.


Fungal Strains and Culture Conditions
M. oryzae strain JJ88 was used as the wild type. It was isolated and purified from Oryza sativa cultivar Jijing88, a variety that is widely planted in Jilin Province, China. All the fungal strains were cultured on complete media (CM) agar plates and maintained on paper filters at −20 • C (CM [10 g/L glucose, 2 g/L peptone, 1 g/L yeast extract, 1 g/L casamino acids, 0.1% (V/V) trace elements, 0.1% (V/V) vitamin supplement, 0.5 g/L MgSO 4 , 6 g/L NaNO 3 , 0.5 g/L KCl, and 1.5 g/L KH 2 PO 4 , pH 6.5]). For conidiation, the strains were inoculated on oatmeal-tomato agar medium (OMA) at 24 • C for 7 days in the dark [27]. The strains were grown continually for 3 days while illuminated under fluorescent lights after the aerial hyphae of the strains had been removed by washes with sterile distilled water.
S. cerevisiae BY4741 and the ScBor1 deletion mutant strains (Invitrogen, Beijing, China) were used for functional complementation test. The yeast S. cerevisiae transformation was performed by the lithium acetate procedure. For yeast gene expression, YPB-ADHpt promoter and terminator regions of ADH1 gene in YPB1 was used [28]. All yeast strains were cultured according to Li et al. [29]. ∆ScBor1 of S. cerevisiae was transferred in MoAE4 and MoACT, respectively. The mutants of ∆ScBor1, ∆ScBor1/MoAE4, ∆ScBor1/MoACT, and wild type were inoculated on to the Solid YPD medium Plates with 100 mM H 3 BO 3 and 50 mM NaHCO 3 , respectively.

Assays for the Subcellular Localization of MoAE4
The localization of MoAE4 in the wild type strain was observed by tagging it with the Bgl II-SpeI sites of green fluorescent protein (GFP) of vector pCAMBIA1303 at its C-terminus. We generated transgenic strains expressing GFP-tagged MoAE4 fusion gene in the wild type of M. oryzae (pCAMBIA1303-MoAE4:: GFP). Fluorescent microscopic observation was carried out by using hyphae (6d) and conidia (6d). To visualize the cytoplasmic membrane and vacuolar membrane, vegetative hyphae and conidia were treated with 2 µg/mL FM4-64 (AAT Bioquest, Sunnyvale, CA, USA) solution for 30-60 min before observed [30] under laser scanning confocal microscope (Olympus fluoview FV3000, Olympus, Tokyo, Japan).

Targeted Gene Deletion and Complementation
To generate the MoAE4 replacement construct pXEH20, the upstream (1155 bp) and downstream (1195 bp) fragments of MoAE4 were amplified using primers MoAE4-L-S/MoAE4-L-A and MoAE4-R-S/MoAE4-R-A, respectively. The resulting PCR products were cloned into the SpeI-KpnI and XbaI-HindIII sites of vector pXEH2.0. The knockout vector was introduced into Agrobacterium tumefaciens strain AGL-1 and then transformed into the wild type M. oryzae using the A. tumefaciens-mediated transformation (ATMT) method as previously described [31]. Transformants were selected and cultured in 200 µg/mL hygromycin. The transformants were identified using PCR with primers HYG-S/HYG-A, MoAE4-LHYG-S/MoAE4-LHYG-A, and MoAE4-G-S/MoAE4-G-A.
The entire MoAE4 sequence was amplified using a PCR technique with MoAE4-C-S/MoAE4-C-A and inserted into the hygromycin resistant vector pCAMBIA1303 for complementation into the mutant strain. The reconstructed pCAMBIA1303-MoAE4 was transformed into the ∆MoAE4 mutant strain and designated ∆MoAE4/MoAE4. The complemented strain was also confirmed by PCR with HYG-S/HYG-A and MoAE4-G-S/MoAE4-G-A.
To further verify the gene deletion and complementation, the expression of the wild type, ∆MoAE4 mutant, and ∆MoAE4/MoAE4 strains was amplified using qRT-PCR with qRT-MoAE4-S/qRT-MoAE4-A and Actin-S/Actin-A, and the strains were identified. The primers for gene deletion and complementation are listed in Table S2.

Quantitative Real-Time PCR (qRT-PCR)
The total RNA was isolated from mycelia that had been harvested from 5-day-old CM media using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). First strand cDNA was synthesized using an oligo (dT) primer from total RNA, which had been treated with DNase I. Subsequently, qRT-PCR was performed using an ABI7500 System (Applied Biosystems, Foster City, CA, USA) and SYBR Premix Ex Taq (TaKaRa, Dalian, China). The relative mRNA levels were calculated using the 2 −∆∆Cq (C q = C qgene − C qactin ) method. The M. oryzae actin gene (MGG_03982.6) was utilized as a reference gene for normalization. Each sample was tested in three replicates in each experiment. The primer sequences used for qRT-PCR are shown in Table S2.

Assays for Conidial Production, Growth, and Development
The strains (wild type, ∆MoAE4, and ∆MoAE4/MoAE4) were cultured on PDA media to understand the effect on MoAE4 conidial production, and the conidia were cultured on OMA media as previously described. A volume of 200 µL of a 1 × 10 5 /mL conidial solution was placed on OMA medium. After 3 days of cultivation at 28 • C, sterile water was added to remove the hyphae, and a piece of the culture medium was cut with a blade and placed on a glass slide. It was then placed in a moisturizing box and incubated at 28 • C. The piece was observed under a Nikon Eclipse 80i microscope at 6, 12, 24, and 48 h after it had been cut. The strains were then stained with lactophenol cotton blue to observe the conidiophore stalks and hyphae under a light microscope [32]. Additionally, the conidia were collected with 2 mL of sterile water after 3 days of culture on OMA media and counted with a hemocytometer. Each strain was repeated three times, and the experiment was conducted in triplicate.
Conidia of the wild type, ∆MoAE4, and ∆MoAE4/MoAE4 were cultured on OMA media and collected to observe the germination of conidia and formation of appressoria. The conidial suspension was adjusted to 1 × 10 5 /mL and added drop wise to a hydrophobic cover slips under a microscope at 1, 2, 3, 4, 5, and 6 h. Each strain was repeated three times, and the experiment was conducted in triplicate.

Rice Sheath Penetration and Plant Infection Assays
To determine the pathogenicity of MoAE4, the wild type, ∆MoAE4, and ∆MoAE4/ MoAE4 strains were inoculated on OMA media to collect the conidia as previously described. The fourth leaf stage of rice seedlings (Oryza sativa cv. Lijiangxintuanheigu) was assayed for infection following the spraying of 2 mL of a conidial suspension (5 × 10 4 conidia/mL in 0.2% gelatin). The inoculated plants were placed in the dark in a dew chamber for 24 h at 28 • C and then transferred to a growth chamber with a photoperiod of 16 h for 7 days.
Conidial suspensions (100 µL, 5 × 10 4 conidia/mL) were injected into seedling leaf sheaths using a 1 mL syringe, and the inoculated plants were placed in a moist chamber as described previously. The formation of lesions and necrosis around the inoculation sites was examined when the injection-wounded leaves unfolded at different time points after the injection. The mean infectious hyphal (IH) growth rates and movement to the adjacent cells were determined from 100 germinated conidia per treatment at 12, 24, and 48 h post inoculation (hpi) and repeated in triplicate as previously described. The leaf sheaths were trimmed at the time points indicated and observed using a Nikon Eclipse 80i microscope. This experiment was performed with three independent replicates, and the representative results from one of these experiments are presented.

Assays for NaHCO 3 Treatment
To illustrate the effect of different concentrations of NaHCO 3 on the expression of MoAE4 gene, wild type strains of M. oryzae were cultivated on PDA with 0, 12.5, 25, 37.5, 50, 62.5, and 75 mM NaHCO 3 at 28 • C for 7 days.
To investigate the effects of sodium bicarbonate stress on the wild type, ∆MoAE4, and ∆MoAE4/MoAE4 strains, each strain was cultured on PDA with NaHCO 3 at final concentrations of 0, 12.5, 25, 37.5, 50, 62.5, and 75 mM at 28 • C for 7 days, and the diameters of fungal strains were photographed using a digital camera (EOS 800D, Canon, Inc., Tokyo, Japan) and measured after inoculation. Each assay was repeated three times independently for each strain, and the experiment was performed in triplicate. Further, the wild type of mycelium treated under different concentrations of NaHCO 3 was collected for expression patterns of AE4.

Assays for HCO 3 − Transport and Intracellular pH Measurements
To determine the relationship between MoAE4 and HCO 3 − transport, confocal microscope was performed with lucigenin (bis-N-methylacridinium nitrate) (MCE, Shanghai, China), a compound that is used as a chemiluminescent probe to indicate the presence of superoxide anion radicals in cells in alkaline conditions 28 [33,34]. The microscopy enabled the detection of direction of HCO 3 − transport in the conidia and hyphae. First, conidia and hyphae from the strains (wild type, ∆MoAE4, and ∆MoAE4/MoAE4) were treated at 28 • C with 0.4 M mannitol in a solution of 50 mM NaHCO 3 for 2 h, then washed and added 0.4 M mannitol to continue recovery at 28 • C for 2 h. A solution of only 0.4 M mannitol served as the control. The strains were incubated with 10 mM lucigenin and observed with a 470 nm fluorescence microscope.

H 2 O 2 Treatment and Endogenous H 2 O 2 Measurements
To investigate the effects of exogenous oxidative stress on the wild type, ∆MoAE4, and ∆MoAE4/MoAE4 strains, each strain was cultured on CM agar that contained 2.5 mM or 5 mM H 2 O 2 for 7 days at 28 • C.
The H 2 O 2 content was determined as previously described for plants [37]. Hydrogen peroxide (H 2 O 2 ) was extracted by homogenizing 3 g of mycelia from the wild type, ∆MoAE4, and ∆MoAE4/MoAE4 strains in 6 mL of cold acetone. The homogenate was then centrifuged at 3500× g for 5 min at room temperature, and the resulting supernatant was designated as the sample extract. Next, 0.1 mL of titanium reagent (5% [w/v] titanic sulfate in concentrated H 2 SO 4 ) was added to 1 mL of the sample extract, followed by the addition of 0.2 mL of strong aqueous ammonia to precipitate the peroxide-titanium complex. The precipitated sample was centrifuged at 3000× g for 10 min at room temperature; the supernatant was discarded, and the precipitate was then solubilized in 5 mL of 2 M H 2 SO 4 . The absorbance of the samples was determined at 415 nm against a blank of 2 M H 2 SO 4 . The H 2 O 2 concentration in the samples was determined by comparing the absorbance against a standard curve of a 0-5 mM titanium-H 2 O 2 complex that was prepared according to Cui et al. [38].
The production of H 2 O 2 was monitored by staining with 3,3 -diaminobenzidine (DAB) as an assay [39]. The hyphae of the wild type, ∆MoAE4, and ∆MoAE4/MoAE4 strains were cultured in CM media for 5 days and then incubated in the dark in a 1 mg/mL solution of DAB at room temperature for 8 h. The samples were washed with sterile water and observed under a Nikon light microscope. This experiment was performed in triplicate and repeated three times for each strain. Similarly, leaf sheath cells of rice infected by wild type, mutant, and complementation strains were stained DAB at 36 hpi.

Statistical Analysis
All the experiments were performed at least three times. The mean ± SD of the strain diameter, germination rate, and relative expression were determined using SPSS Statistics 22 (IBM, Inc., Armonk, NY, USA). Error bars represent the standard deviation. * indicates a statistically significant difference at p < 0.05. ** indicates a highly significant difference at p < 0.01. *** indicates a highly significant difference at p < 0.001. **** indicates a highly significant difference at p < 0.0001.  . The structure was predicted using the web-based server I-TASSER that shows the ligand binding site residues of transmembrane domain P 129 , A 132 , and F 487 . The active site residue is N 218 .
According to the 3D structures ( Figure 1D and Figure S1C), the conformations with 10 TM helices of MoAE4 were characterized by two inverted repeats that are intertwined to form both core and gate domains, which appears to be similar with that of AE1 or BOR1 [3,10]. The homologous protein sequences and the typical domain patterns reveal MoAE4 is a member of the HCO 3 − transporter AE4 group.

MoAE4 Localizes on Cytomembrane and Tonoplast and Functions in Yeast
Most HCO 3 − transporters function on cytoplasm membrane [16]. To test where MoAE4 occurred, we generated transgenic strains expressing GFP-tagged MoAE4 fusion gene in the wild type of M. oryzae ( Figure S2A). Fluorescent microscopic observation was carried out by using hyphae (6d) and conidia. The lipophilic dye (FM™ 4-64 Dye, AAT Bioquest, USA) was used for observing the cytoplasmic membrane and vacuolar membrane. A strong green fluorescence signal of the MoAE4-GFP protein co-localized with FM4-64 red fluorescence was detected on the cytoplasmic membrane in young hyphae and conidia (Figure 2A,B); interestingly, a strong co-localized yellow fluorescence signal was also detected on the vacuole membrane in hyphae (Figure 2A,B). By comparison, in the wild type or untransformed strains, the background green fluorescence was too weak to be detected. This result suggests that MoAE4 was targeted to cytomembrane and tonoplast.

MoAE4 Transports Cytosolic HCO3 − to Vacuole and Cell outside
To identify the function of MoAE4 in response to NaHCO3, M. oryzae was cultivated under NaHCO3 stress conditions. By using the knockout mutant strain of MoAE4 (ΔMoAE4) and the complemented strain (ΔMoAE4/MoAE4) ( Figure S2B-D), the growth of the tested strains was assessed. When cultivated on complete media (CM) plates at 25 °C without NaHCO3, both the ΔMoAE4 and ΔMoAE4/MoAE4 strains grew at a rate similar to that of the wild type, and their colony morphologies exhibited little difference (Figure Yeast bicarbonate transporters are boric acid tolerant [12][13][14]. Based on the ScBor1/AE4 deletion mutant (∆ScBor1), the complementary yeast strains were created by using MoAE4. As a result, the MoAE4 gene could functionally reverse the defect of the ∆ScBor1 mutant in boric acid tolerance ( Figure 2C), suggesting MoAE4 functions as yeast Bor1. Different from the yeast BOR1, MoAE4 also endowed the NaHCO 3 tolerance in the mutant strains ( Figure 2C), implying MoAE4 may be a HCO 3 − transporter.

MoAE4 Transports Cytosolic HCO 3 − to Vacuole and Cell outside
To identify the function of MoAE4 in response to NaHCO 3 , M. oryzae was cultivated under NaHCO 3 stress conditions. By using the knockout mutant strain of MoAE4 (∆MoAE4) and the complemented strain (∆MoAE4/MoAE4) ( Figure S2B-D), the growth of the tested strains was assessed. When cultivated on complete media (CM) plates at 25 • C without NaHCO 3 , both the ∆MoAE4 and ∆MoAE4/MoAE4 strains grew at a rate similar to that of the wild type, and their colony morphologies exhibited little difference ( Figure 3A,B). When subjected with NaHCO 3 of different concentrations, all strains including the wild type were repressed in mycelial growth under NaHCO 3 stress. Impressively, ∆MoAE4 almost stopped growing at 25 mM of NaHCO 3 , but the wild type and complementary strains were capable of growing even at the concentration of 37.5 mM ( Figure 3A,B), suggesting a role of ∆MoAE4 in tolerance to NaHCO 3 . The expression patterns in response to NaHCO 3 treatment also illustrated this point ( Figure 3C).   To determine the HCO 3 − transport activity of MoAE4, lucigenin (bis-N-methylacridinium nitrate), a chemiluminescent probe, was used as to detect the existence of anion radicals in cells under alkaline conditions [33,34]. After treated with NaHCO 3 solution (50 mM), the tested strains incubated with 10 mM lucigenin were observed with a 470 nm fluorescence microscope. As expected, lucigenin probed HCO 3 − with green fluorescence signals was accumulated in vacuoles of the wild type and ∆MoAE4/MoAE4 strains; however, in ∆MoAE4, green fluorescence signals were only detected in cytoplasm ( Figure 4A,B). Accordingly, intracellular pH value increased significantly in ∆MoAE4 ( Figures 4C,D and S3). These results indicated that MoAE4 functions as a bona fide HCO 3 − transporter.

MoAE4 Is Important for Conidiation and Appressorium Development
Conidiation and appressorium formation were analyzed among the ∆MoAE4, ∆MoAE4/ MoAE4, and wild type strains. The sparse conidiophores with less conidia were observed in the deletion mutant ∆MoAE4, however, both the ∆MoAE4/MoAE4 and wild type produced thick conidiophores and more conidia (Figures 5A-C and S4).
The conidium germination rate of all strains including the wild type was similar at 4-6 h, although ∆MoAE4 appeared to be a little slow in conidium germination at 1-4 h ( Figures 5D and S4B,C). In terms of appressorium formation, ∆MoAE4/MoAE4 had the formation rate similar as the wild type did; but ∆MoAE4 was severely affected ( Figure 5E). As conidia and appressoria are essential factors for disease cycle and infection, MoAE4 is proposed to be involved in pathogenesis.
type produced thick conidiophores and more conidia (Figures 5A-C and S4).
The conidium germination rate of all strains including the wild type was simi 4-6 h, although ΔMoAE4 appeared to be a little slow in conidium germination at ( Figures 5D and S4B,C). In terms of appressorium formation, ΔMoAE4/MoAE4 ha formation rate similar as the wild type did; but ΔMoAE4 was severely affected (Figur As conidia and appressoria are essential factors for disease cycle and infection, MoA proposed to be involved in pathogenesis. The conidia were harvested from the 3-day-old myce lium grown on OMA media, and counted using a hemocytometer for all the three strains. (D) Conidial germination rate Conidial germination was measured on a hydrophobic cover slips and was calculated under the microscope at 1, 2, 3, 4 5, and 6 h post inoculation. (E) Appressorial formation rate. Appressorial formation was measured on a hydrophobic cove slips and was calculated under the microscope at 4 and 6 h per inoculation. The analysis was performed using an inde pendent samples t-test. ns p > 0.05. * p < 0.05. ** p < 0.01. Error bars indicate the mean ± SD from three independent exper iments. conidia were harvested from the 3-day-old mycelium grown on OMA media, and counted using a hemocytometer for all the three strains. (D) Conidial germination rate. Conidial germination was measured on a hydrophobic cover slips and was calculated under the microscope at 1, 2, 3, 4, 5, and 6 h post inoculation. (E) Appressorial formation rate. Appressorial formation was measured on a hydrophobic cover slips and was calculated under the microscope at 4 and 6 h per inoculation. The analysis was performed using an independent samples t-test. ns p > 0.05. * p < 0.05. ** p < 0.01. Error bars indicate the mean ± SD from three independent experiments.

Requirement of MoAE4 for Pathogenicity in M. oryzae
In order to characterize the function of MoAE4 in pathogenic development, pathogenicity assays were carried out using conidia collected from ∆MoAE4, ∆MoAE4/MoAE4, and the wild type. When intact susceptible rice seedlings were spraying-inoculated, at 7 days post inoculation (dpi), some acute expansive disease lesions were observed in rice leaves by the wild type and ∆MoAE4/MoAE4; but no lesions were formed in rice leaves by the ∆MoAE4 (Figure 6). Similarly, when drop-inoculation was assayed, only the wild type and complementary strains still showed pathogenicity.
In order to characterize the function of MoAE4 in pathogenic development, pathogenicity assays were carried out using conidia collected from ΔMoAE4, ΔMoAE4/MoAE4, and the wild type. When intact susceptible rice seedlings were spraying-inoculated, at 7 days post inoculation (dpi), some acute expansive disease lesions were observed in rice leaves by the wild type and ΔMoAE4/MoAE4; but no lesions were formed in rice leaves by the ΔMoAE4 (Figure 6). Similarly, when drop-inoculation was assayed, only the wild type and complementary strains still showed pathogenicity. Leaf sheath infection assays were performed to examine the infection effects of the MoAE4 in rice host. At 12 hpi, most mature (black) appressoria have been formed in the wild type and ∆MoAE4/MoAE4, but less in ∆MoAE4. At 24 hpi, invasive hyphae of the wild type and ∆MoAE4/MoAE4 commenced to branch in rice cells, but the primary infectious hyphae were just formed in ∆MoAE4. At 48 hpi, the majority of invasive hyphae of the wild-type and ∆MoAE4/MoAE4 branched and started entering neighboring cells, but the ∆MoAE4 strains did not due to the defects in appressorium formation ( Figure 6C,D).
To decipher the exact action of MoAE4 during pathogenic development, we defined the three types of infection hyphae according to their developmental morphologies. Then we quantified the proportion of the three types of infection hyphae based on 100 germinated conidia in the inoculated leaf sheath ( Figure 6D). As a result, at 48 hpi more than 60% of inoculated conidia from ∆MoAE4/MoAE4 and wild type formed branched infectious hyphae in one cell, of which about 20% extended to neighboring cells of rice (type II and III), suggesting the requirement of MoAE4 in pathogenesis.

MoAE4 Is Important for H 2 O 2 Tolerance and Clearance Inside or Outside Cells
To address the relationship between MoAE4 and endogenous H 2 O 2 , the wild type and mutant strains were cultured on CM agar supplemented with 2.5 or 5 mM H 2 O 2 at 28 • C for 5 days. As a result, ∆MoAE4 was markedly inhibited in mycelial growth under H 2 O 2 stress, indicating a role of MoAE4 in oxidative stress tolerance ( Figure 7A,B). As rice plant accumulates more H 2 O 2 during pathogen-rice interaction, and MoAE4 expression increases with pathogenic development of M. oryzae, we speculate that MoAE4 is responsible for the clearance of host-derived H 2 O 2 during infection. To test this, DAB staining was used to identify the endogenous ROS accumulated in the cells of rice leaf sheath infected by M. oryzae at 36 hpi ( Figure 7C). In the leaf sheaths inoculated with the ∆MoAE4 strains, more than 60% of the infected cells investigated were stained dark brown; in contrast, less than 20% of the infected cells were stained light brown or colorless as ∆MoAE4/MoAE4 and wild type ( Figure 7C,D), displaying loss of H 2 O 2 scavenging function in ∆MoAE4. Additionally, endogenous H 2 O 2 was measured in M. oryzae more H 2 O 2 accumulated in ∆MoAE4 than in the wild type and ∆MoAE4/MoAE4 ( Figure 7D). These results reveal that MoAE4 is responsible for regulating H 2 O 2 levels exogenous, endogenous, or plant-derived. The above experiments were performed in triplicate and repeated three independent times for each strain. Error bars represent the ± SD of three independently repeated samples, ns p > 0.05. ** p < 0.01.

Discussion
The family of bicarbonate transport proteins are involved in a wide-range of physiological processes in humans and mammals [1,2,42]. Mutation or dysregulation of these transporters results in physiological diseases in humans [43]. Therefore, bicarbonate transporters have attracted medical attention and have been extensively studied. In fungi, such as the unicellular organism yeast, CO2 can diffuse directly out of yeast cells, so the anion exchanger/carbonic anhydrase system and HCO3 − transporter are proposed to be dispensable [32]. Indeed, the AE1 homologue, YNL275w has been verified as an anion transporter just serving for boron detoxification or tolerance in S. cerevisiae [12][13][14]. Bicarbonate gradients modulate growth and colony morphology in A. nidulans [44]. Bioinformatically, other filamentous fungi harbor homologous of mammal HCO3 − transporters. In A. nidulans, there are at least five members of SLC family, but the most homologous SbtB still showed boron transporter, not HCO3 − transporter [16]. Considering the exclusive activity of BOR1 in fungi and plants, YNL275w and homologues should be grouped into the secondary bicarbonate transporter family specific for boron. The bona fide HCO3 − transport

Discussion
The family of bicarbonate transport proteins are involved in a wide-range of physiological processes in humans and mammals [1,2,42]. Mutation or dysregulation of these transporters results in physiological diseases in humans [43]. Therefore, bicarbonate transporters have attracted medical attention and have been extensively studied. In fungi, such as the unicellular organism yeast, CO 2 can diffuse directly out of yeast cells, so the anion exchanger/carbonic anhydrase system and HCO 3 − transporter are proposed to be dispensable [32]. Indeed, the AE1 homologue, YNL275w has been verified as an anion transporter just serving for boron detoxification or tolerance in S. cerevisiae [12][13][14]. Bicarbonate gradients modulate growth and colony morphology in A. nidulans [44]. Bioinformatically, other filamentous fungi harbor homologous of mammal HCO 3 − transporters. In A. nidulans, there are at least five members of SLC family, but the most homologous SbtB still showed boron transporter, not HCO 3 − transporter [16]. Considering the exclusive activity of BOR1 in fungi and plants, YNL275w and homologues should be grouped into the secondary bicarbonate transporter family specific for boron. The bona fide HCO 3 − transport function in fungal YNL275w homologues is actually unknown. In this research, we demonstrated that the MoAE4, as a YNL275w homologue, plays a role in HCO 3 − transport. Importantly, MoAE4 located to cytomembrane and tonoplast promotes conidiation, appressorium formation, and pathogenesis in M. oryzae.
In general, AEs localize to cytoplasmic membrane [10,35]. GFP-tagged Bor1p were detected to localize preferentially to the vacuole and that cells lacking Bor1p have fragmented vacuoles [45]. Recently, kidney anion exchanger 1(kAE1) has been detected on vacuole [46]. These findings suggest Bor1p functions on both cytomembrane and tonoplast. Additionally, in our study, MoAE4 was localized on cytomembrane and tonoplast (Figure 2A; Table  S1). In hyphae, MoAE4 protein was concentrated on tonoplast; but in conidia, a strong fluorescence signal was on cytomembrane, suggesting the different subcellular patterns of MoAE4. This may reflect the specific function of MoAE4 in different developmental stages of M. oryzae.
In addition to the homology with BOR1 in sequences and 3-D structures (Figure 1), MoAE4 rescued the defect of ∆ScBOR1 in boron tolerance, suggesting an authentic BOR1 homologue. However, MoAE4 also endowed the NaHCO 3 tolerance for the mutant strains, for even the wild type failed to grow normally under NaHCO 3 stress ( Figure 2C). Particularly, loss of MoAE4 resulted in the severe mycelial growth inhibition in ∆MoAE4 compared with the wild type and complementary strains ( Figure 3A,B), implying the HCO 3 − transport activity in MoAE4. Based on the lucigenin fluorescence quenching assay, a great amount of HCO 3 − was detected in the cytosol of ∆MoAE4, but not in vacuoles of ∆MoAE4; on the contrary, HCO 3 − was only detected in vacuole in the wild type and ∆MoAE4/MoAE4 (Figure 4), revealing the MoAE4-associated bicarbonate resistant mechanism, by which MoAE4 removes redundant HCO 3 − from cytosol to vacuole and outside cells.
As a pathogenic fungus, conidiation and appressorium formation are key processes for disease cycle and infection [17,18]. MoAE4 loss resulted in the impaired conidiophore formation and then decreased conidial and appressorial productivity ( Figure 5). We speculate that this may be related to the excessively accumulated H 2 O 2 in the MoAE4 deletion mutant ( Figure 7D). After all, ∆MoAE4 became sensitive under H 2 O 2 stress ( Figure 7B,C). At this point, the reduced pathogenicity in ∆MoAE4 could also be partially explained because ∆MoAE4 was exposed to a high hydrogen peroxide stress both in vivo and in vitro ( Figure 7A,E and Figure S6). CO 2 , as a labile molecule, is the oxidation waste product of mitochondrial respiration. In humans, redundant CO 2 must be released, or the equilibrium with HCO 3 − + H + will be disturbed. The ability of HCO 3 − to undergo pH-dependent conversions is central to its physiological role [1]. CO 2 enters the cytoplasm through the membrane and is rapidly hydrated forming carbonic acid (H 2 CO 3 ). This acid is dissociated into H + and HCO 3 − by intracellular carbonic anhydrases [7]. Under normal medium culture conditions, pathogenic fungi such as yeast may not require the anion exchanger/carbonic anhydrase system to help the release of metabolic CO 2 [11]. However, the interaction system between host plant and pathogen forms a whole multicellular organism, which should face a challenge in moving membrane impermeant bicarbonate from inside the cell where it is produced to the environment for disposal. In addition, during interactions between plant and pathogen, respiration from both pathogen and plant must be enhanced to produce available energy [24,[47][48][49]. Therefore, we propose a pathogenic model mediated by MoAE4/MoCA (carbonic anhydrase) system ( Figure 8). Under cultivation conditions, the metabolic CO 2 can be released freely from M. oryzae, and both MoAE4 and MoCA genes are at low levels of expression because of the equilibrium of (CO 2 + H 2 O HCO 3 − + H + ) ( Figure 8A). In the process of invasive hyphae growth, the infected plant cell is a relatively high concentration of CO 2 and low concentration of O 2 microenvironment; and accordingly, the diffusion of fungal CO 2 to the outside (cytosol of plant cell) is hindered. The upregulated MoCA ( Figure S5) will increase the concentration of HCO 3 − , which leads to MoAE4 being upregulated, then MoAE4 transports HCO 3 − to the vacuole or to plant cells ( Figure 8B).
In the model, MoAE4 maintains the homeostasis of intracellular CO2-HCO3 − system, which further ensures the intracellular acid-base balance in cells [50]. According to the importance of the acid-base equilibrium in multiple physiological activities [51], we imply HCO3 -regulated by MoAE4 is a pathogenic signal for blast disease development. Actually, in our recent research, a low dose of sodium carbonate can induce the expression of a large number of disease-related genes in wild type, not in the MoAE4 deletion mutant (unpublished data), revealing a potential drug target for blast disease control. To decipher the regulation mechanism, much work remains to be completed.

Conclusions
In the rice blast fungus, MoAE4 exhibits more sequence and structure homology with AE proteins. Additionally, MoAE4, localized on cytomembrane and tonoplast, possesses boron and NaHCO3 tolerance in yeast. Lucigenin fluorescence quenching assay indicated that MoAE4 has HCO3 − transport activity. Meanwhile, ΔMoAE4 contained more H2O2 than the wild type and complementary strains did, implying a role of MoAE4 in energy metabolism. Importantly, MoAE4 is involved in conidiation, appressorium formation, and pathogenicity in M. oryzae. Overall, MoAE4, as a cytomembrane and tonoplast HCO3 − transporter, promotes pathogenesis of M. oryzae. Based on these results, a pathogenic model mediated by MoAE4 is proposed.

Supplementary Materials:
The following are available online at www.mdpi.com/xxx/s1, Figure S1. Sequence alignment and prediction of MoAE4 structure. Figure S2. The construction strategies for MoAE4 deletion and complementation strains. Figure S3. The green fluorescence signals detection of intracellular pH in conidia. Figure S4. The loss of the MoAE4 gene has a negative effect on sporulation, appressorial development. Figure S5. Transcription abundance of MoCA during disease development. Figure S6. Standard curve line and the endogenous H2O2 of strains hyphae. Table S1. Identifying sub-cellular location. Table S2. Primers used in this study.

Conflicts of Interest:
The authors declare no conflict of interest. In the model, MoAE4 maintains the homeostasis of intracellular CO 2 -HCO 3 − system, which further ensures the intracellular acid-base balance in cells [50]. According to the importance of the acid-base equilibrium in multiple physiological activities [51], we imply HCO 3 − regulated by MoAE4 is a pathogenic signal for blast disease development. Actually, in our recent research, a low dose of sodium carbonate can induce the expression of a large number of disease-related genes in wild type, not in the MoAE4 deletion mutant (unpublished data), revealing a potential drug target for blast disease control. To decipher the regulation mechanism, much work remains to be completed.

Conclusions
In the rice blast fungus, MoAE4 exhibits more sequence and structure homology with AE proteins. Additionally, MoAE4, localized on cytomembrane and tonoplast, possesses boron and NaHCO 3 tolerance in yeast. Lucigenin fluorescence quenching assay indicated that MoAE4 has HCO 3 − transport activity. Meanwhile, ∆MoAE4 contained more H 2 O 2 than the wild type and complementary strains did, implying a role of MoAE4 in energy metabolism. Importantly, MoAE4 is involved in conidiation, appressorium formation, and pathogenicity in M. oryzae. Overall, MoAE4, as a cytomembrane and tonoplast HCO 3 − transporter, promotes pathogenesis of M. oryzae. Based on these results, a pathogenic model mediated by MoAE4 is proposed.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/jof7110955/s1, Figure S1. Sequence alignment and prediction of MoAE4 structure. Figure S2. The construction strategies for MoAE4 deletion and complementation strains. Figure S3. The green fluorescence signals detection of intracellular pH in conidia. Figure S4. The loss of the MoAE4 gene has a negative effect on sporulation, appressorial development. Figure S5. Transcription abundance of MoCA during disease development. Figure S6. Standard curve line and the endogenous H 2 O 2 of strains hyphae. Table S1. Identifying sub-cellular location. Table S2. Primers used in this study.