Colletotrichum gloeosporioides strain SMCG1#C (Huang et al., 2019) was used for the wild type (WT). The WT, gene deletion mutants, and the complemented strains were maintained on the potato dextrose agar (PDA) plates at 25°C. Complete medium (CM) was employed to culture fungal mycelia for DNA extraction and protoplast preparation as aforedescribed (Wang et al., 2020).

The CgSAT4 was identified in the C. gloeosporioides genome database¹ using BLASTP with amino acid sequences of Sat4/Hal4 (NM_001178721.1) from S. cerevisiae. Protein sequences were aligned with ClustalX 2.1, and a phylogenetic tree of the CgSat4 and its orthologs from different fungi was generated using MEGA 6.0 (Tamura et al., 2013).

The knockout deletion mutant of CgSAT4 was obtained using the method described by Wang et al. (2020). Firstly, the upstream and downstream flanking regions (ca. 1.2 kb) flanking sequences of CgSAT4 were amplified using the primer sets of 1F/2R and 3F/4R (Supplementary Table S1), respectively. The PCR products were ligated to either side of the hygromycin B phosphotransferase (HPH) cassette to generate the gene replacement fragment using overlap PCR with the primer set OuterF/OuterR (Supplementary Table S1). Secondly, the gene replacement fragments were purified and transformed into the protoplasts of the WT according to the aforedescribed procedure (Yang et al., 2018). Thirdly, the candidate transformants were screened on the TB3 plate and confirmed by Southern blotting as described previously (Wang et al., 2020), with the probe generated with the primer sets CgSAT4_SN_F/CgSAT4_SN_R and F1111/F1112 (Supplementary Table S1), respectively.

For complementation, a 3.0-kb fragment containing the CgSAT4 ORF region and its native promoter (~1,500 bp) was amplified with the primers CgSAT4 -GFP-1F and CgSAT4-GFP-2R (Supplementary Table S1). The PCR products were inserted into the vector pYF11 that was linearized with the restriction endonuclease Xho I using the yeast gap repair approach (Zhou et al., 2011). The resulting fusion construct CgSAT4-GFP was verified by sequencing and transformed into the CgSAT4 gene deletion mutant. The positive complemented strains were screened by GFP signals and Western blotting described by Bruno et al. (2004).

To observe the localization of CgSat4, mycelial plugs of the complemented strain ΔCgsat4/SAT4 expressing fusion protein CgSat4-GFP were cultured in liquid CM medium at 25°C for 24 h. Then, fresh mycelia of the corresponding strains expressing fusion protein CgTrk1-GFP were prepared for fluorescence microscopy observation as described above. Photographs were taken under a confocal laser scanning microscope (Zeiss, Oberkochen, Germany). To evaluate the effect of deletion of CgSAT4 on the localization of CgTrk1, the construct of CgTRK1-GFP was introduced into the WT and the ΔCgsat4 mutant, respectively, as described above. The localization of CgTrk1 in these strains were observed under a confocal laser scanning microscope. The experiment was performed twice.

For vegetative growth assays, the mycelial blocks (6 mm in diameter.) of the WT, the ΔCgsat4 mutant, and the complemented strain were inoculated onto PDA plates, respectively. The plates were kept in an incubator at 25°C. Colony growth kinetics was measured at 5 days post-inoculation. Fruiting bodies were induced on V8 juice agar (V8) plates for 10 days after inoculation at 25°C according to the method of Fang et al. (2018). The experiment was carried out three times, and each treatment had three replicates.

To test the role of CgSat4 on stress resistance, the WT, the ΔCgsat4 mutant, and the complemented strains were inoculated on CM plates supplemented with NaCl (0.7 M), KCl (0.7 M), or LiCl (0.3 M). These plates were kept at 25°C for 4 days. The experiment was carried out three times, and each treatment had three replicates.

To evaluate the effect of CgSat4 on potassium uptake, the WT and ΔCgsat4 mutant were cultured in liquid CM medium containing 7 mM potassium for 2 days as described previously (Yang et al., 2018). The fungal mycelia were harvested and dried in a freeze-dryer. Then, the dried mycelia were digested with H₂SO₄, and mycelial potassium was examined using a flame spectrophotometer (Yang et al., 2018). The experiment was conducted three times, and each treatment had three replicates.

For sporulation, the mycelial blocks (6 mm in dia.) of the WT, the ΔCgsat4 mutant, and the complemented strains were inoculated in the carboxymethyl cellulose (CMC) medium to induce sporulation, and the conidia were collected and counted as described by Fang et al. (2018). The experiment was carried out three times, and each treatment had three replicates.

To induce conidial germination and appressorium formation, the conidial suspensions of the WT, the ΔCgsat4 mutant, and the complemented strains were adjusted to 10⁵/ml, respectively. Ten microliters of conidial suspension of each strain was placed on the glass coverslip (Fisher Scientific, St. Louis, MO, United States) and kept at 25°C. The conidial germination rate of each strain was calculated at 2, 4, and 8 h postinoculation, respectively. The percentage of appressorium formation of each strain was tested at 4, 8, and 12 h postinoculation, respectively. Appressorium turgor pressure was analyzed by the incipient cytorrhysis assay using 1–4 M of glycerol solution, as described by Wang et al. (2020). The experiment was conducted three times with at least 100 structures per replicate.

Onion penetration assays were conducted as described by Wang et al. (2020). Ten microliters of conidial suspension of the WT, the ΔCgsat4 mutant, and the complemented strains was inoculated on the adaxial surface of onion epidermal strips to induce invasive hyphae (IH), respectively. At 24 h post-inoculation, IH was observed under a Zeiss Axio Imager A2M microscope (Carl Zeiss, Jena, Germany). IH was divided into four types (type I, no hyphae penetration; type II, IH with one branch; type III, IH with at least two branches, but having limited expansion; type IV, IH with numerous branching and extensive hyphal growth). The experiment was performed three times, and at least 30 invasive structures were observed in each treatment.

Pathogenicity of the WT, the ΔCgsat4 mutant, and the complemented strain was tested as described by Wang et al. (2020). Conidial suspensions of the WT, the ΔCgsat4 mutant, and the complemented strains were adjusted to 1 × 10⁵ spores/ml, respectively. Five microliters of conidial suspension of each strain was inoculated on healthy leaves of C. lanceolata, Populus × euramericana cv. “Nanlin895” and Liriodendron chinense × tulipifera, respectively. The inoculated leaves were kept in a moist chamber at 25°C, and lesion size was measured at 5 days post-inoculation. DNA was isolated from the L. chinense × tulipifera leaves inoculated by the WT and the ΔCgsat4 mutant, and the fungal biomass in planta was examined using qPCR as described by Yang et al. (2018). The experiment was performed three times, and each treatment had three replicates.

Mycelial plugs of the WT and the ΔCgsat4 mutant were inoculated into 100 ml of CM medium, and shaken at 150 rpm for 2 days at 25°C, respectively. The mycelia of each strain were collected using a layer of Miracloth. Total protein was extracted from the mycelia following the method described by Bruno et al. (2004). Twenty microliters of total proteins were isolated on the SDS-PAGE gel and transferred to a polyvinylidene fluoride (PVDF) membrane using a Bio-Rad imprinting device (Bio-Rad Laboratories, Inc., CA, United States). A primary anti-GFP antibody (GFP labeled mouse monoclonal antibody, Shanghai Antibody Market, China) and a secondary antibody (goat anti-mouse IgG horseradish peroxidase, Shanghai Antibody Market, China) were used to detect GFP. Signal strength corresponding to phosphorylated Hog1 was detected by binding of anti-phosphorylated p38 MAPK (Thr180/Tyr182, rabbit monoclonal antibody) (Cell Signaling Technology, Boston, MA, United States). P38 MAPK antibody (ABMART) was used as the control. The experiment was carried out twice.

To evaluate the effect of the CgSat4 on the potassium transporter CgTrk1 localization and expression, the CgTRK1 coding region, and its native promoter sequence (~1,500 bp) were amplified with the primers CgTRK1-GFP-1F and CgTRK1-GFP-2R. The PCR products were purified and inserted into the vector pYF11 as aforedescribed (Zhou et al., 2011). The fusion construct CgTRK1-GFP was transformed into the WT and the ΔCgsat4 mutant, respectively. The positive candidate strains expressing fusion protein CgTrk1-GFP were screened by GFP signals observation. The expression levels of CgTrk1-GFP in the WT and the ΔCgsat4 mutant were detected using Western blotting with anti-GFP antibody as aforedescribed. The experiment was carried out twice.

The CgTRK1-GFP fusion construct was transferred, respectively, into the wild-type strain and ∆Cgsat4 mutant. The positive transformants were cultured in liquid CM for 48 h. For protein isolation, about 150–200 mg of mycelia were ground into powder in liquid nitrogen and resuspended in 1 ml of extraction buffer [10 mM Tris–HCl (pH 7.5), 150 mM NaCl, 0.5 mM EDTA, 0.5% NP40, 1 mM PMSF, 10 μl of protease inhibitor cocktail (Sigma, United States), and 10 μl of phosphatase inhibitor cocktail 3 (Sigma, United States)]. For the preparation of the phosphatase-treated Cell lysates, the phosphatase inhibitor cocktail was omitted for 2.5 U/ml alkaline phosphatase (final concentration; P6774; Sigma) and the sample was incubated for 1 h with the addition of 1 mM MgCl2 (37°C). The samples were then resolved on a 8% SDS-polyacrylamide gel prepared with 50 μM acrylamide-pendant Phos-tag ligand (Wako, Japan) and 100 μM MnCl₂, according to the instruction provided by the Phos-tag.

To evaluate the effect of deletion of CgSAT4 on the transcription level of CgTrk1, the mycelia of the WT and the ΔCgsat4 mutant were, respectively, cultured in liquid CM for 3 days at 25°C. The mycelia were collected and total RNA was extracted using the TRIzol LS reagent (Invitrogen, Carlsbad, CA, United States). The first-strand cDNA was synthesized and used as the templates of quantitative RT-PCR as described by Yang et al. (2018). The experiment was conducted three times, and each treatment had three replicates.

Data are presented as mean ± SD. Statistical analyses were carried out with the data processing system (DPS) version 9.50 using a one-way ANOVA (p < 0.01).

An orthologue of Sat4 was identified by a BLAST_P search using the Sat4/Hal5 (NM_001178721.1) of S. cerevisiae as the reference to the genome database of C. gloeosporioides.² The orthologue shared a 65% amino acid sequence identity with S. cerevisiae Sat4, which was named CgSat4. Phylogenetic analysis showed that Sat4 anthologies in filamentous fungi have significantly diverged from that of S. cerevisiae. The CgSat4 was most similar to its orthologues from phytopathogenic fungi of Colletotrichum species (Supplementary Figure S1A). CgSat4 contained a low complexity region and a Pfam Pkinase motif at the C-terminus (Supplementary Figure S1B).

A gene deletion mutant ΔCgsat4 was obtained by replacing the open reading frame with the hygromycin phosphotransferase resistance (HPH) gene (Supplementary Figure S1C). Southern blotting analysis confirmed that the CgSAT4 was encoded by a single-copy gene in the WT, which was replaced by the HPH gene in the ΔCgsat4 mutant (Supplementary Figure S1D). A complemented strain ΔCgsat4/CgSAT4 was generated by reintroducing the CgSAT4 encoding region with its native promoter into the ΔCgsat4 mutant.

To investigate the localization pattern of CgSat4 in C. gloeosporioides, the CgSat4-GFP fusion construct was introduced into the ΔCgsat4 mutant to generate the complemented strain ΔCgsat4/CgSAT4. The complemented strain recovered the defects of the CgSAT4 deletion mutant as described above, which indicated the CgSat4 has been properly expressed. Thus, the GFP signals of the complemented strain were employed to determine the localization of CgSat4. The fluorescence microscopy showed that strong GFP signals were distributed in the cytoplasm of the conidia and hyphae (Supplementary Figure S1E). Furthermore, an 80-kDa of the predicted CgSat4-GFP fusion protein was detected using an anti-GFP antibody (Supplementary Figure S1F). The data indicated that CgSat4 was expressed in the cytoplasm of C. gloeosporioides.

To assess the role of CgSat4, the WT, the ΔCgsat4 mutant, and the complemented strain ΔCgsat4/CgSAT4 were, respectively, inoculated on the PDA. The growth kinetics was observed. On the 4th and 5th day, the colony diameter of the ΔCgsat4 mutant significantly reduced compared to the WT and complemented strains (Figures 1A,B).

Fruiting body quantification showed that fruiting bodies produced by the ΔCgsat4 mutant were significantly less than those of the WT and complemented strains (Figures 1C,D). There are no morphological differences in the conidia among the WT, ΔCgsat4 mutant, and complemented strain ΔCgsat4/CgSAT4. However, conidial enumeration showed that conidia produced by the ΔCgsat4 mutant were considerably less than those of the WT and complemented strain (Figure 1E). These results indicated that CgSat4 is required for vegetative growth, fruiting body development, and sporulation in C. gloeosporioides.

In the yeast S. cerevisiae, the Sat4 is involved in potassium influx which is mediated by the Trk1-Trk2 transport system (Mulet et al., 1999; Hirasaki et al., 2011). To evaluate whether the orthologue C. gloeosporioides CgSat4 is required for potassium uptake, the mycelial potassium content of the WT and the ΔCgsat4 mutant was examined. Data showed that the potassium was significantly reduced in the ΔCgsat4 mutant than in the WT (Figure 2A).

The transcription level of CgTrk1 was detected in the WT and ΔCgsat4 mutant by qRT-PCR. The result showed that there was no significant transcription difference of CgTRK1 between the WT and ΔCgsat4 mutant (Figure 2B). However, western blotting analysis showed the expression of CgTrk1 in WT was more than twice as much as that in the ΔCgsat4 mutant (Figure 2C).

The localization pattern analysis also showed that the potassium transporter CgTrk1 was distributed in the plasma membrane in the WT. However, the CgTrk1 was mainly distributed in the vacuoles in the ΔCgsat4 mutant (Figure 2D). We therefore examined the interaction between CgSat4 and CgTrk1 by co-immunoprecipitation (co-IP) and yeast two-hybrid (Y2H) assays. However, the results show that CgSat4 does not interact directly with CgTrk1 (Supplementary Figures S2A,B). We also examined whether CgSat4 regulates the level of CgTrk1 phosphorylation through Phos-tag SDS-PAGE gel electrophoresis. The shifts in the mobility of CgTrk1-GFP from the wild-type strain and ∆Cgsat4 mutant cells had no obvious difference (Supplementary Figure S2C), suggesting that CgSat4 does not regulate the level of CgTrk1 phosphorylation. Combining the aforementioned results, we concluded that CgSat4 may be involved in potassium uptake by regulating the localization of the high-affinity potassium transporter CgTrk1.

In eukaryotic cells, the Hog1 mitogen-activated protein kinase (MAPK) pathway is essential to osmotic stress response (Brewster et al., 1993; Román et al., 2020). As an intracellular osmotic stress regulator, potassium participates in various physiological processes (Pérez-Valle et al., 2007; Kahm et al., 2012). Since the deletion of CgSAT4 resulted in the decrease of mycelial potassium, we hypothesized that decreased potassium concentration may alter the intracellular osmotic pressure and resistance against osmotic stress. To verify this hypothesis, the WT, ΔCgsat4 mutant, complemented strain ΔCgsat4/CgSAT4 were, respectively, inoculated onto the CM plates subjected to osmatic stressors KCl, NaCl, and LiCl, respectively. Compared with the WT and complemented strains, the ΔCgsat4 mutant was more sensitive to the osmatic stressors and displayed a higher growth inhibition rate (Figures 3A,B). The effect of CgSAT4 deletion on the phosphorylation of CgHog1 was further evaluated. The result showed that deletion of CgSAT4 significantly increased the phosphorylation level of CgHog1 compared to the WT (Figure 3C). However, an interaction between CgSat4 and CgHog1 cannot be reproduced (Supplementary Figure S2D).

We found that deletion of CgSAT4 did not affect the conidial morphology. Compared with the WT and complemented strains, the conidial germination rate of the ΔCgsat4 mutant significantly decreased at 2 and 4 h. However, there was no significant difference in the conidial germination among the WT, ΔCgsat4 mutant, complemented strains at 8 h (Figures 4A,B). These data indicate that deletion of the CgSAT4 significantly delayed conidial germination in the early stage.

The data showed that, compared with the WT and complemented strains, appressorium formation rate of the ΔCgsat4 mutant was significantly decreased (Figure 4C). In the ΔCgsat4 mutant, 80% of appressoria failed to melanize, while the WT and the complemented strains formed normal melanized appressoria (Figure 4D). At different concentrations of glycerol, the appressorium collapse rate of the ΔCgsat4 mutant was significantly higher than those of the WT and the complemented strains (Figure 4E). These results indicated that deletion of the CgSAT4 caused the defects in appressorium formation and appressorial turgor pressure.

The onion epidermis penetration assays also showed that the percentages of Types I, II, III, and IV infectious hyphal growth of the ΔCgsat4 mutant were, respectively, 27%, 25.9%, 29.9%, and 21%, compared with 8.6%, 18.2%, 17.5%, and 55.6% in the WT (Figures 4F,G).

To explore the role of CgSat4 in host infection, conidial suspensions of the WT, ΔCgsat4 mutant, and the complemented strains were, respectively, inoculated on the healthy leaves of Populus × euramericana cv. Nanlin895. Five days post-inoculation (dpi), the ΔCgsat4 mutant produced smaller lesions than those of the WT and complemented strains (Figures 5A,B,F). Similarly, compared with the WT and complemented strains, the ΔCgsat4 mutant showed a decreased virulence on Liriopendron chinense × tulipifera, and Cunninghamia lanceolata (Figures 5A,C,D,G,H). Fungal biomass assays using the qPCR method also showed that the biomass of the ΔCgsat4 mutant in planta was significantly reduced than that of the WT (Figure 5E). These data indicated that CgSat4 played a key role in infectious growth and full virulence in C. gloeosporioides.