Sampling was conducted on three occasions in total over five afternoons before the tide rising, at three sites (seaward zone, mid zone and landward zone) in the mangrove forest (Leizhou, China, 21°68’N, 110°33′E), where soil samples were collected from 0 to 10 cm below the surface sediments. The samples were collected using a sterilized hand shovel (around 4 cm × 6 cm, with 18 cm handle) and large debris was removed by hand. All the samples from all sites were combined and mixed thoroughly before transferring into sterile 500 mL canning jars and transported to laboratory in an icebox.

Fungal isolation was carried out immediately after arrival, around 30 min after the collection, following Ahumada-Rudolph’s method with some modifications (Ahumada-Rudolph et al., 2016). Briefly, the sediment was suspended in 4 L sterile seawater, homogenized, and left to settle. The supernatant was collected by centrifugation at 100 g for 15 min and stirred well afterwards. Supernatant (2 mL) was diluted in series with sterilized seawater (10–100 folds in 10-fold intervals) with two replicates. An aliquot of 100 μL was transferred to yeast extract peptone dextrose agar (YPD) prepared using sterilized seawater supplemented with ampicillin, kanamycin and streptomycin (final concentration of 50 μg/mL, 50 μg/mL and 25 μg/mL respectively). The aliquot was spread onto the agar using a sterile glass coating rod and the incubation was conducted at 28°C.

Subcultures were performed till individual colonies were obtained. Morphology of individual colonies was verified under the microscope. Spores of isolated fungi were collected using sterilized water with 0.05% Tween 80 and stored at 4°C for up to 1 week. The agar with fungal growth was sliced off from the dish using sterilized scalpel (about 4 mm × 4 mm each piece), submerged in 80% glycerol solution (v/v) and stored at −80°C for 4 months.

Fungal genomic DNA was extracted using the Qiagen DNEasy Plant Extraction kit (Qiagen Inc., Valencia, CA, United States). Primers (ITS1: 5’-TCCGTAGGTGAACCTGCGG-3′ and ITS4: 5’-TCCTCCGCTTATTGATATGC-3′) were used to amplify by polymerase chain reaction (PCR) the whole region of ribosomal internal transcribed spacer (ITS1-5.8 s-ITS2; Korabecna, 2007). The PCR reaction mixture included 10 μL of PCR master mix (Promega, 2X), 1 μL of the forward primer ITS1, 1 μL of the reverse primer ITS4, 1 μL of DNA extraction, and 7 μL of ddH₂O. The thermal cycler was programmed with the following conditions: initial denaturation at 94°C for 4 min, followed by 35 cycles of denaturation at 94°C for 20 s, annealing at 55°C for 20 s, and extension at 72°C for 1 min, concluding with a final extension at 72°C for 10 min. PCR products were purified using the QIA quick PCR purification kit and sent to Sangon Biotech (Shanghai, China) for bidirectional sequencing. The assembled sequences were analyzed using BLAST ¹ against the standard database on NCBI.

Salt tolerance assay was performed on YPD agar supplemented with sodium chloride (NaCl, w/v) at different concentrations (50, 100, 150, 200 and 250 g/L). The spores of isolated fungi were inoculated on the salt-contained agar, cultured at 28°C for up to 10 days, and the fungal growth was monitored.

Real wastewater originating from processes utilized for glucosamine production was used in this study. The wastewater featured with a low pH (0.39 ± 0.05), dark color, fishy smell, and a high Cl⁻ concentration (84.39 ± 1.21 g/L). Other physical properties and chemical composition were not characterized in this study.

Fungi tolerant to salt concentrations equal to or above 100 g/L were adapted gradually into 20, 40, 60 and 80% glucosamine processing wastewater. Peptone, yeast extract and glucose were added into the diluted wastewater to a 20% concentration of the regular YPB (yeast extract peptone dextrose broth) medium. Liquid cultures were performed in 250 mL conical flasks containing 50 mL of growth medium inoculated with 5 × 10⁵ conidia and incubated at 28°C on an orbital shaker at 180 rpm with three individual flasks dedicated for each time point. Conidia collected from low wastewater-containing medium were used for the subsequent culture. Culture supernatants were collected every day from flasks dedicated to each time point by centrifugation at 4,500 g for 30 min. The biomass was weighed after drying and the supernatant was filtered through a 0.22-μm membrane (Millipore, China) at 4°C. The cleared supernatants were then aliquoted in 1.5 mL Eppendorf tubes and stored at −80°C.

Acidity-tolerance assay was performed in liquid culture with 80% wastewater as a medium with different pH values (3.5, 4.0, 4.5, 5.0 and 5.6). The cultivation was performed at 28°C at 180 rpm for 7 days, and the fungal biomass, pH values as well as the Cl⁻ concentration in the wastewater were measured daily.

The fungal strains capable of growing in 80% wastewater (Cl⁻ concentration of 67.512 g/L) were further studied and their Cl⁻ removal capacity was detected by ion chromatography at room temperature using a Metrohm Model 761 Compact Ion Chromatograph (Metrohm, Herisau, Switzerland) with suppressor module, equipped with an ICSep AN2™ analytical column (250 mm × 4.6 mm). The injection volume was 20 μL. The eluent used was a 1.8 mM Na₂CO₃ + 1.7 mM NaHCO₃ mixture and the suppressor regenerating solution was 0.1 M H₂SO₄. The eluent was freshly prepared and filtered through a 20-μm filter before usage. Data acquisition and processing were performed automatically using the integration software MagIC Net™ software 1.1. Standard solutions (GBW (E)082048, China) of known concentrations of Cl⁻ were analyzed in order to create a calibration curve (Supplementary Figure S2).

Fungal cells were stained using a chloride sensitive fluorescent probe MQAE (N-[ethoxycarbonylmethyl]-6-methoxy-quinolinium bromide, Beyotime, China) and then observed with a fluorescence microscope equipped with an imaging system at an excitation wavelength of 360 nm (D-35578 Wetzlar, Leica, Germany).

To observe the distribution of Cl⁻ in fungal cells, MQAE and Congo Red were used to conduct double-staining for the fungal cells. MQAE was loaded first following the manufacturer’s manual. Briefly, the mycelia were washed three times with Krebs-HEPES buffer (0.02 M, pH 7.4), placed onto glass slides immersed in a drop of MQAE (5 mM), incubated for 30 min at 37°C in the dark, and finally washed again using the buffer. The resulting specimens were then stained using 0.01% Congo Red for 10 min, washed with ddH₂O to remove the excess stain, and then covered with 4% paraformaldehyde fix solution and a coverslip. A Leica TCS SP5 confocal laser microscope (Leica Microsystems, Germany) equipped with epifluorescence microscopy (Leica DMI 6000B microscope) was used to observe the specimens at excitation of 360 nm (emission at 460 nm) and 497 nm (emission at 614 nm) separately.

For these experiments, A. piperis was grown on Petri dishes containing yeast peptone dextrose agar (YPD) at 28°C. After a 7-day-culture period, conidia were obtained by washing the plate surface with phosphate-buffered saline (PBS; 10 mM NaH₂PO₄, 10 mM Na₂HPO₄, 150 mM NaCl, pH 7.2) and filtering them through a 40-μm nylon cell strainer (BD Falcon, Franklin Lakes, NJ, United States) in order to remove the hyphal fragments. The conidial cells were counted in a Neubauer chamber. Tenebrio molitor larvae exhibiting clear and uniform color and weighing between 70 and 100 mg were selected for the survival studies (de Souza et al., 2015). Galleria mellonella larvae were maintained and fed as previously described until reaching 200–300 mg in weight (Silva et al., 2018). The survival curves (virulence assay) were performed through injection of different fungal inocula (10², 10³, 10⁴, 10⁵ and 10⁶ conidia/larva). Larvae (10 per each assayed group) were inoculated with fungal conidia using an insulin syringe (10 μL/larva) and incubated at both 28 and 37°C in Petri dishes containing rearing diet. The inoculation was performed by the injection of fungal suspensions into the T. molitor larvae hemocoel in the ventral portion at the second visible sternite above the legs or in the last right proleg of the G. mellonella larvae (de Souza et al., 2015). Larvae inoculated with sterile PBS were used as control groups. Larvae were assessed daily, up to 7 days, to check their survival, being scored as dead when they displayed no movement in response to touch. Survival analyses were determined using the log-rank test and the Kaplan–Meier survival curves (GraphPad Prism 6). The experiment was conducted in two independent experimental sets.

A total of 34 fungal strains were isolated from the wetlands of a mangrove forest, named H1 to H34 (Figure 1), displayed on a dark background showing the diverse morphology of the fungal population isolated from the mangrove. Detailed taxonomy information for these strains can be found in Supplementary Table S1. The fungal strains’ DNA was successfully amplified using the universal primers ITS1 and ITS4. BLAST searches revealed their identities as members of 3 different phyla (Supplementary Table S1), namely Ascomycota, Basidiomycota and Mucoromycota, in which the Ascomycota accounted for the majority (44.1%). The dominating fungal genera identified in this study were Aspergillus (35.3%) and Penicillium (20.6%). Representatives of Trichoderma, Furarium, Mucor, Candida, Amanita and Talaromyces genera were additionally identified. Despite Amanita loosii is generally recognized as an edible mushroom, the strain identified in this study is not recommended for consumption, considering its origin in an industrial waste environment.

An initial screening of these fungal strains for salt tolerance was carried out. The results (Figure 2) showed that 4 strains (H6, H10, H15 and H16) were tolerant to a salinity of 200 g/L, while 8 strains (H1, H4, H8, H12, H13, H14, H22 and H25) tolerant to 150 g/L, 7 strains (H7, H9, H11, H19, H20, H29 and H32) tolerant to 100 g/L, and 10 strains (H2, H3, H8, H23, H24, H26, H28, H30, H31 and H34) tolerant to 50 g/L. Strains H17, H21, H27, and H33 showed tolerance to lower salt levels. However, they were not subjected to further testing because low tolerance was deemed unfeasible for processing shrimp shell effluents, which typically contain salt levels exceeding 50 g/L. The mangrove wetland-derived fungal strains with a salt tolerance level greater than 100 g/L were subsequently gradually adapted into the 80% wastewater and Cl⁻ concentration in the water was measured using ion chromatography. Most of the fungal strains, including those tolerant at 200 g/L NaCl on YPD medium, were not able to grow in the wastewater with pH adjusted to 5.6. Only 7 strains (Figure 2, yellow bars, H4, H6, H7, H12, H16, H18 and H32) adapted to the highest wastewater concentration (80% wastewater). Among these strains, H16 (Aspergillus piperis) removed the highest amount of Cl⁻ (approx. 30%) from the wastewater.

The original pH value of the wastewater was highly acidic (pH 0.39 ± 0.05), in which none of the tested fungal strains survived. A great amount of base (NaOH) was required to neutralize the pH to a level able to support the fungal growth, which would be problematic at an industrial scale. To determine the lowest possible pH for fungal growth, the pH of the wastewater was adjusted to 2.0–5.0, 0.5 intervals and 5.6 to assess viability for strain H16 that had the highest chloride removal. It turned out that H16 could survive down to pH 3.5, with a slow growth rate. Wastewater with pH 5.0 supported the best fungal growth (Figure 3A) and the fungal strain was able to neutralize the acidity of the medium as biomass increased (Figure 3B). Interestingly, while H16 neutralized the acidity of the medium at similar rates at pH 5.0 and pH 5.6, chloride removal was more rapid at pH 5.0 than it was at pH 5.6 (Figure 3C), and the biomass peaked earlier (Figure 3A).

The removal of Cl⁻ was observed to occur in conjunction with fungal exponential phase (Figures 3A,B). When normalized to biomass, chloride reduction varied by days among the tested pH range (Supplementary Figure S1). The fungus apparently tended to reduce more Cl⁻ in a strong acidic environment, as seen from the reduction rates at pH 3.5 and 4.0, which were greater than those at higher pHs. However, the values kept decreasing day by day due to slow growth. pH 4.5, 5.0 and 5.6 brought out smaller Cl⁻ reduction rates at the beginning of the cultivation, but the reductions were enhanced gradually with the highest chloride removal per unit biomass by day 7 at pH 4.5. Exploring the interplays between fungal growth, chloride uptake and medium pH will be an interesting avenue for future research.

The wastewater had a dark color (Figure 4), low pH (1.7 in this study) and a high Cl⁻ concentration (Figure 3C). After adjusting to pH 5.0 using NaOH, strain H16 was able to clarify the broth in pace with the fungal growth, along with the decrease in the Cl⁻ concentration (Figure 4). A visible lightening in the color of the wastewater was also observed, suggesting that other components in the wastewater were remediated by the fungi, but the identification of which was not within the scope of this study. At day 3 post-inoculation, the Cl⁻ concentration in the broth dropped to 17.478 g/L from the original level of 67.512 g/L, corresponding to a 74.11% Cl⁻ removal. Further investigations are needed to lower the Cl⁻ concentration to less than 5 g/L, creating conditions conducive for the growth of most microorganisms. This, in turn, will enable subsequent bio-treatment processes aimed at reducing the chemical oxygen demand (COD) and biochemical oxygen demand (BOD) in the wastewater (Liu et al., 2020).

The Cl⁻ and fungal cells were stained with MQAE (blue) (Figure 5) and/or Congo red (red) (Figure 6) fluorescent stains, and observed using a fluorescent microscope and a confocal laser microscope separately. MQAE is a sensitive chloride ion indicator, and the fluorescence was observed throughout the fungi, including both the hyphae and spores, distributed evenly in the cells, suggesting a strong uptake from the wastewater by strain H16, consistent with the reduction in media Cl⁻ content (Figures 3, 4). Compared to strain H4, which had a reduced chloride clearance ability than H16 (Figure 2), the fluorescence intensity observed was lower (Figure 5). Confocal imaging with Congo Red, which bound to the glucans in the fungal cell wall, showed markedly different outcomes (Figure 6); chloride ions entered H16 cells, while being bound to the exterior of H4 cells. Further studies will be necessary to understand the mechanism behind the uptake of Cl⁻ into the fungal cells.

The pathogenicity of the isolated A. piperis was investigated using T. molitor and G. mellonella virulence models at both 28 and 37°C. The results of the present assay showed that mortality in both insect models was typically dose-dependent, with doses lower than 1 × 10⁵ conidia having little to no effect on larval mortality within 7 days after fungal inoculation (Figure 7). The fungal infection capability was also found to be modulated by temperature, as incubation at 37°C (host temperature) led to more significant larval killing induced by A. piperis in both insect models compared to 28°C (an environmental temperature; Figure 7). It is worth noting that G. mellonella larvae were found to be less sensitive to A. piperis conidial infection than T. molitor larvae under both tested temperatures (Figure 7).