Target of rapamycin controls hyphal growth and pathogenicity through FoTIP4 in Fusarium oxysporum

Abstract Fusarium oxysporum is the causal agent of the devastating Fusarium wilt by invading and colonizing the vascular system in various plants, resulting in substantial economic losses worldwide. Target of rapamycin (TOR) is a central regulator that controls intracellular metabolism, cell growth, and stress responses in eukaryotes, but little is known about TOR signalling in F. oxysporum. In this study, we identified conserved FoTOR signalling pathway components including FoTORC1 and FoTORC2. Pharmacological assays showed that F. oxysporum is hypersensitive to rapamycin in the presence of FoFKBP12 while the deletion mutant strain ΔFofkbp12 is insensitive to rapamycin. Transcriptomic data indicated that FoTOR signalling controls multiple metabolic processes including ribosome biogenesis and cell wall‐degrading enzymes (CWDEs). Genetic analysis revealed that FoTOR1 interacting protein 4 (FoTIP4) acts as a new component of FoTOR signalling to regulate hyphal growth and pathogenicity of F. oxysporum. Importantly, transcript levels of genes associated with ribosome biogenesis and CWDEs were dramatically downregulated in the ΔFotip4 mutant strain. Electrophoretic mobility shift assays showed that FoTIP4 can bind to the promoters of ribosome biogenesis‐ and CWDE‐related genes to positively regulate the expression of these genes. These results suggest that FoTOR signalling plays central roles in regulating hyphal growth and pathogenicity of F. oxysporum and provide new insights into FoTOR1 as a target for controlling and preventing Fusarium wilt in plants.

plant pathogens (Dean et al., 2012), and one of the most difficult plant diseases to control because the spores of F. oxysporum can survive in soil for over 10 years (Husaini et al., 2018). The proximity of roots induces the dormant chlamydospores to germinate and initiate infection. Elongated hyphae start to adhere to the plant root surface and penetrate the roots through wounds or root tips. Ultimately, invasive hyphae reach the xylem vessels and proliferate, causing disease (Berrocallobo & Molina, 2008;Di et al., 2003). In the process of phytopathogenic fungi infecting the host, the plant cell wall is a major physical barrier that phytopathogenic fungi must overcome by producing a variety of cell wall-degrading enzymes (CWDEs) that allow fungi to invade host tissues through the degradation of plant cell wall components. The pathogen F. oxysporum secretes various CWDEs, such as pectinase, cellulase, and β-glucosidase, to degrade the plant cell wall (Calero-Nieto et al., 2007;Jonkers, 2009;Ospina-Giraldo et al., 2003). As a result, pectin can obstruct the vasculature in plants, thereby preventing water absorption, resulting in plant wilt and death ; free monosaccharide and oligosaccharides that originate from the plant cell wall are used for fungal growth and reproduction .
Target of rapamycin (TOR) is an evolutionarily conserved Ser/Thr protein kinase in eukaryotes. It is well known that the TOR signalling pathway regulates cell growth and proliferation in response to nutrients, energy, and stresses (De Virgilio & Loewith, 2006;Dobrenel et al., 2016;Yuan et al., 2013). Two TOR genes were first identified by screening rapamycin (RAP)-insensitive mutants in budding yeast (Saccharomyces cerevisiae) (Heitman et al., 1991b;Kunz et al., 1993).
However, only a single TOR gene has been identified in Arabidopsis thaliana, most animals, and humans (Menand et al., 2002;Sabers et al., 1995). The TOR protein contains five conserved regions: HEAT repeats, a FAT domain, an FRB domain, a kinase domain, and an FATC domain (Schmelzle & Hall, 2000;Wullschleger et al., 2006). TOR forms two structurally and functionally distinct protein complexes: TOR complex 1 (TORC1) and TORC2 (Loewith et al., 2002). Among eukaryotic species, TORC1 contains the TORC1 regulatory subunits KOG1 (known as RAPTOR in mammals) and Lethal with SEC13 protein 8 (LST8), which regulates cell growth and metabolism in response to nutrient and energy requirements (Wang & Proud, 2009). TORC2 possesses two TORC2-specific subunits, AVO3 (known as RICTOR in mammals) and AVO1 (known as SIN1 in mammals), which controls spatial cell growth and survival by regulating cytoskeletal structure and polarity (De Virgilio & Loewith, 2006;Wullschleger et al., 2006). Despite functional importance, little is known about the TOR signalling pathway in F. oxysporum.
In rapidly growing cells, ribosome biogenesis is a major energyconsuming process that accounts for a significant proportion of total transcriptional output (Warner, 1999). The regulation of ribosome biogenesis occurs primarily at the transcriptional level and involves all three nuclear RNA polymerases (Iadevaia et al., 2014;Martin et al., 2006). TORC1 positively regulates several steps in ribosome biogenesis, including ribosomal RNA transcription and the synthesis of ribosomal proteins and other components required for ribosome assembly and biogenesis (Ben-Sahra et al., 2013; Chauvin et al., 2014;Tsang et al., 2010). TORC1 interacts directly with Sfp1 and phosphorylates Sfp1 to regulate ribosome biogenesis (Lempiainen et al., 2009). Sfp1 is a transcriptional activator of ribosome biogenesis genes (Fingerman et al., 2003;Marion et al., 2004). Under optimal growth conditions, Sfp1 is phosphorylated by TORC1 and then binds to the promoters of ribosome biogenesis genes to promote their expression in the nucleus. By contrast, nutrient depletion results in its relocalization from the nucleus to the cytoplasm in yeast (Lempiainen et al., 2009). The AGC-family kinase SCH9 (known as S6K in mammals) is another downstream regulatory factor of TORC1. TORC1 phosphorylates and activates SCH9 to regulate ribosome protein synthesis (Chauvin et al., 2014;Iadevaia et al., 2014;Magnuson et al., 2012).
RAP is a macrolide immunosuppressant produced by Streptomyces hygroscopicus. It mimics nutrient limitation to arrest cell growth and proliferation. RAP specifically binds to FK506 binding protein of 12 kDa (FKBP12), which interacts with the FRB domain of TOR to form a ternary complex (Heitman et al., 1991b;Loewith et al., 2002). The resulting complex prevents TOR from associating with its scaffold protein RAPTOR and phosphorylating its substrate proteins (Aylett et al., 2016;Hara et al., 2002), which hinders TOR protein activity and results in irreversible arrest of the cell cycle at the G1 phase (Heitman et al., 1991a).
RAP can inhibit TORC1, but TORC2 is insensitive to RAP (Loewith et al., 2002). The TORC2-specific subunit RICTOR plays indispensable roles in TORC2 function (Gaubitz et al., 2016;Wullschleger et al., 2005). A recent crosslinking study of TORC2 in budding yeast demonstrated that the C-terminus of RICTOR occupies the FRB domain of TOR kinase, preventing the RAP-FKBP12 complex from binding to the FRB domain of TOR kinase in TORC2, which makes TORC2 insensitive to RAP (Gaubitz et al., 2015). Recent studies have also revealed that treatment with ATPcompetitive TOR protein kinase inhibitors, including Torin1, Torin2, Ku-0063794 (KU), and AZD-8055 (AZD), can result in different effects on both TORC1 and TORC2 than RAP treatment (Chresta et al., 2010;Garcia-Martinez et al., 2009), as TOR is directly and specifically targeted by the ATP-binding pocket of the TOR kinase domain, suppressing the functions of both TORC1 and TORC2 complexes (Benjamin et al., 2011).
Potato is a major staple food worldwide, but potato Fusarium wilt and dry rot diseases caused by F. oxysporum are global challenges for potato production. The TOR signalling pathway plays critical roles in regulating mycelial growth and virulence in fungi, and inhibition of TOR activity significantly reduces mycelial growth and pathogenicity in Botrytis cinerea, Verticillium dahliae, and Fusarium graminearum (Li et al., 2019;Xiong et al., 2019;Yu et al., 2014). In this study, we functionally characterized the conserved FoTOR signalling pathway in the regulation of mycelial growth and pathogenicity of F. oxysporum. RNA sequencing (RNA-seq) analysis showed that FoTOR signalling plays vital roles in multiple intracellular and extracellular processes, including ribosome biogenesis and cell wall degradation. Additionally, FoTOR1 interacting protein 4 (FoTIP4), a new component of the FoTOR signalling pathway, was identified and characterized. Our findings also show that FoTIP4 plays an important role in the regulation of ribosome biogenesis and cell wall degradation in F. oxysporum. Our results suggest that FoTOR1 may serve as a promising target for controlling and preventing Fusarium wilt caused by F. oxysporum in plants.

| The TOR signalling pathway is conserved in F. oxysporum
In order to identify evolutionarily conserved TOR signalling pathway components in F. oxysporum, BLASTp analysis of the F. oxysporum f. sp. lycopersici genome database (http://fungi.ensem bl.org/Fusar ium_oxysp orum/Info/Index ?db=core) using Schizosaccharomyces pombe TOR signalling pathway components as reference was performed. We found that putative homologous gene sequences encoding key components of TORC1, including TOR, KOG1, and LST8, were present in the genome of F. oxysporum, and putative homologues encoding specific components of TORC2, such as AVO3 and AVO1, were also found in the F. oxysporum genome (Table 1) (Table S1).
We found TOR kinase expansion in eight out of 14 sequenced F. oxysporum strains, with six strains containing one copy, seven strains containing two copies, and one strain containing three copies.
Phylogenetic analysis showed that the orthologous copies of the TOR kinase (indicated as Core TOR) form a monophyletic group and nine TOR paralogues (indicated as LS TOR) clustered together ( Figure S1); this is in agreement with a previous study (DeIulio et al., 2018).
To verify the interactions between TOR and other key components of the TOR signalling pathway, we performed yeast twohybrid (Y2H) assays. We found that FoTOR1, rather than FoTOR2, interacted with FoKOG1 (Figures 1e and S2). Interestingly, neither FoTOR1 nor FoTOR2 interacted with FoAVO3 in Y2H assays, implying that there may be no functional TORC2 in F. oxysporum.
To further determine FoTOR function, FoTOR2 deletion mutants (ΔFotor2) were created (Figure 1f). Morphological analyses showed that hyphal growth of the ΔFotor2 mutants was similar to that of the wildtype F. oxysporum strain (Figure 1g), indicating FoTOR2 is dispensable for hyphal growth. However, all hygromycin-resistant transformants of FoTOR1 deletion mutants were ectopic mutants and we failed to get a null mutant. These results suggest that the deletion of FoTOR1 in F. oxysporum may be lethal, indicating that FoTOR1 is an essential and central regulator of the FoTOR signalling pathway in F. oxysporum.

| TOR inhibitors inhibit mycelial growth of F. oxysporum
RAP is a well-known TOR-specific inhibitor. It specifically binds to FKBP12 to form a gain-of-function complex, which negatively regulates TOR kinase activity (Heitman et al., 1991b). To test whether RAP has an inhibitory effect on TOR in F. oxysporum, we first as-  (Chresta et al., 2010;Garcia-Martinez et al., 2009;Montane & Menand, 2013;Xiong et al., 2017). Furthermore, the germination rate and production of spores were reduced upon RAP and Torin1 treatment, and the relative transcript levels of sporulation-related genes were significantly decreased upon FoTOR inhibition in F. oxysporum ( Figure   S3). In addition, we tested the sensitivity of the ΔFotor2 mutants to RAP and Torin1. The results showed that the growth of the ΔFotor2 mutant was similar to that of the wildtype F. oxysporum strain upon RAP and Torin1 treatment ( Figure S4), indicating that TOR inhibitors inhibit mycelial growth of F. oxysporum in a FoTOR2-independent manner. Combined treatment with RAP and Torin1 exerted more obvious growth inhibitory effects than treatment with RAP or Torin1 alone ( Figure 2e). The IC 50 value of a single drug (RAP 10 nM, Torin1 1 μM) was significantly reduced when F. oxysporum was subjected to combined treatment (Figure 2f), implying that potential synergistic effects can be generated by combining RAP with Torin1. Next, a computer-simulated affected fraction (Fa)-combination index (CI) curve was assessed using CompuSyn software. A synergistic effect (CI < 1) was observed when hyphae were treated with a combination of RAP + Torin1 (Figure 2g). These results suggest that RAP and Torin1 inhibit hyphal growth by simultaneously targeting the TOR signalling pathway in F. oxysporum.

| Deletion of FoFKBP12 leads to insensitivity to RAP in F. oxysporum
We found that RAP effectively inhibited hyphal growth at a low concentration ( Figure 2). It was previously reported that FKBP12 mediates the inhibitory effects of RAP on TOR (Heitman et al., 1991b). Therefore, we analysed the sequences of FoFKBP12 in the F. oxysporum genome. A single copy of an FKBP12 orthologue (FOXG_08379, named FoFKBP12) encoding a protein with 59% similarity to SpFKBP12 was found (Table 1). Sequence alignment and phylogenetic analysis showed that FoFKBP12 is evolutionarily conserved among species (Figure 3a,b). Amino acids known to be involved in the formation of the RAP inhibitory ternary complex were conserved in the FoFKBP12 sequence ( Figure 3a). In order to test the ability of FoFKBP12 to bind RAP and TOR, we generated FoFKBP12 deletion mutants (ΔFofkbp12) by a homologous recombination gene deletion strategy ( Figure S5a,b). Morphological analyses showed that hyphal growth of the mutant was similar to that of wildtype F. oxysporum on potato dextrose agar (PDA) (Figure 3c), suggesting that FoFKBP12 is dispensable for hyphal growth. The RAP sensitivity test showed that the ΔFofkbp12 mutant was insensitive to RAP, but the sensitivity to RAP was restored in the complemented strain (ΔFofkbp12 + FoFKBP12) ( Figure 3c). As expected, the wildtype, ΔFofkbp12, and A previous study showed that Arabidopsis has adapted an evolutionary mutation in the FKBP12 gene, resulting in loss of its ability to bind RAP (Sormani et al., 2007). In order to further confirm the function of FoFKBP12, FoFKBP12 overexpression Arabidopsis transgenic lines were generated. All FoFKBP12 transgenic lines displayed sensitivity to RAP, which reflected shorter primary root length, smaller cotyledons, and lower fresh weight compared with wildtype Arabidopsis (Figure 3d,e), and this observation was consistent with observations in the ScFKBP12 overexpression line (BP12-2) in Arabidopsis (Ren et al., 2012). We tested the ability of FoFKBP12 to bind RAP in an S. cerevisiae (Y2HGold strain) FKBP12 deletion mutant (ΔScfkbp12).
The wildtype yeast strain was sensitive to RAP while the ΔScfkbp12 mutant was RAP-resistant. FoFKBP12 restored the sensitivity to RAP in the ΔScfkbp12 + FoFKBP12 yeast strain ( Figure 3f). We analysed the interaction between FoTOR1 and FoFKBP12 by Y2H assays. For each experiment, one pair of Y2H plasmids was cotransformed into the ΔScfkbp12 mutant strain. FoFFKBP12 was unable to interact with FoTOR1 without RAP treatment. By contrast, when the medium was supplemented with 1 μg/ml RAP, FoFKBP12 interacted strongly with FoTOR1 ( Figure 3g). These results indicate that FoFKBP12 mediates the inhibitory effects of RAP on FoTOR activity in F. oxysporum.

| FoTOR1 is a key regulator of ribosome biogenesis and CWDEs in F. oxysporum
The TOR signalling pathway integrates various extracellular and intracellular signals, such as growth factors, nutrients, energy, and other environmental cues, regulating multiple cellular processes, including ribosome biogenesis, protein synthesis, autophagy, and metabolic processes ( Rapamycin-insensitive companion of TOR (RICTOR) Translation initiation factor 2 subunit alpha (eIF2α) Representative seedlings are shown. The experiment was repeated three times. Bar = 1 cm. (e) Fresh weight and root length of FoFKBP12 overexpression transgenic Arabidopsis lines exposed to RAP (5 μM). Each column represents the average of 10 seedlings. The data are presented as the mean ± SD of n = 3 independent experiments. **p < 0.01 compared with wildtype plants (Student's t test). (f) FoFKBP12 restored the sensitivity to RAP in the yeast FKBP12 mutant (ΔScfkbp12). Strains growth on yeast-peptonedextrose (YPD) medium with (+) or without (−) 1 μg/ml RAP at 28 °C for 3 days.  (Table S2). Among the upregulated genes, the GO terms organonitrogen compound metabolic process (GO: 1901564) and organonitrogen compound biosynthetic process ( GO: 1901566) were highly enriched (Figure 4c, Table S2). Among the downregulated genes, the GO terms transporter activity (GO: 0005215) and transmembrane transporter activity (GO: 0022857) were the most significantly enriched (Figure 4d, Table S2).  (Table S3). These ribosomal core proteins combine with small nucleolar RNAs to form small nucleolar ribonucleoproteins that play a crucial role in ribosome biogenesis by guiding the processing and modification of preribosomal RNAs. To further confirm these observations, 10 randomly selected genes involved in ribosome biogenesis were selected for quantitative reverse transcription PCR (RT-qPCR) analysis (Figure 4f). They were all downregulated, which was consistent with the RNA-seq data.

Kyoto
Additionally, we tested the expression levels of ribosome biogenesis genes in ΔFotor2 and ΔFofkbp12 lines treated with RAP. RAP had no effect on the expression levels of ribosome biogenesis genes in the ΔFofkbp12 mutant, while these genes were downregulated in the ΔFotor2 line treated with RAP ( Figure S6a). The RNA-seq results support the previous observations that TOR suppression by RAP inhibits hyphal growth, suggesting that the FoTOR signalling pathway positively regulates ribosome biogenesis and vegetative growth in F. oxysporum.
In the process of host infection, F. oxysporum secretes a large number of CWDEs to degrade plant cell walls, which facilitates invasion and colonization. The mRNA levels of CWDEs, including cellulases, xylanase, and pectinases, were changed upon TOR inhibition ( RAP had no effect on the expression levels of CWDE genes in the ΔFofkbp12 mutant ( Figure S6b). These results indicate that FoTOR1 plays an important role in the regulation of CWDEs in F. oxysporum.

| FoTOR1 interacts with FoTIP4 to regulate ribosome biogenesis in F. oxysporum
To further identify new effectors of the FoTOR signalling pathway, we used the Y2H system to screen for interacting proteins of FoTOR1 in a cDNA library of F. oxysporum. The results revealed putative FoTOR1 interacting proteins including FoTIP1/FoKOG1, FoTIP2/ FoLST8, and FoTIP3/FoFKBP12. Importantly, a novel FoTOR1 interacting protein 4 (FoTIP4) was detected in F. oxysporum (Figure 5a).

Sequence alignment revealed that FoTIP4 is homologous to ScSFP1
with low amino acid sequence similarity (18%). SFP1 is a C2H2-type zinc finger transcription factor that plays an essential role in the

| FoTOR signalling regulates the expression of CWDEs through FoTIP4 in F. oxysporum
The FoTOR signalling pathway is a key regulator of the expression of CWDEs in F. oxysporum, as shown in Table S4. To determine whether FoTIP4 mediates the effects of the FoTOR signalling pathway in the regulation of CWDEs, we examined the pathogenicity of ΔFotip4 mutant strains. In infection assays with potato leaves and tubers, the lesion symptoms of ΔFotip4 mutant strains were reduced by approximately 50% compared with those of the wildtype F. oxysporum strain (Figure 7a,b). Additionally, a cellophane penetration assay was performed to verify invasive growth of ΔFotip4 mutant strains.
Efficient penetration was detected in the wildtype and complemented (ΔFotip4 + FoTIP4) strains, but not in ΔFotip4 mutant strains

| D ISCUSS I ON
TOR is an evolutionarily conserved protein kinase that regulates cell growth and metabolism in response to growth factors, hormones, cellular energy status, and nutrient abundance (Ma & Blenis, 2009;Saxton & Sabatini, 2017). In yeast and animals, TOR is engaged in two large complexes: TORC1 and TORC2.   Figure 1). Furthermore, deletion of FoTOR1 may be lethal, while deletion of FoTOR2 has no effect on hyphal growth in F. oxysporum. Phylogenetic analysis showed that FoTOR1 protein was clustered as the core TOR group, while FoTOR2 protein was clustered into the TOR paralogues group; these results were consistent with a previous study (DeIulio et al., 2018). KOG1/ RAPTOR functions as a scaffold coupling TOR to substrates in S. cerevisiae and animals (González & Hall, 2017 In S. cerevisiae and animals, the TOR signalling pathway has been studied in detail (Cornu et al., 2013;De Virgilio & Loewith, 2006;Saxton & Sabatini, 2017;Wullschleger et al., 2006).
RAP and Torin1 can effectively inhibit mycelial growth of F. oxysporum in a dose-dependent manner (Figure 2). In this study, we found that the FoFKBP12 deletion mutant was resistant to RAP ( Figure 3). Resistance of FKBP12 deletion mutants to RAP has been reported in other fungi, including B. cinerea, F. graminearum, Fusarium fujikuroi, and Mucor circinelloides (Bastidas et al., 2012;Melendez et al., 2009;Teichert et al., 2006;Yu et al., 2014). (c) Relative transcript levels of ribosome biogenesis-related genes in wildtype F. oxysporum, ΔFotip4 mutants, and the complemented strain. The data are presented as the mean ± SD of n = 3 independent experiments. **p < 0.01 compared with wildtype F. oxysporum (Student's t test). (d) Electrophoretic mobility shift assay showed that FoTIP4 binds to the promoter of FoSIK1 (FOXG_12883) containing the RRPE sequence. The symbols − and + represent absence and presence, respectively. (e) RAP caused FoTIP4 relocalization to the cytoplasm from the nucleus. Hyphae carrying a GFP-tagged FoTIP4 protein were grown for 3 days in potato dextrose broth, RAP (1 μM) was added, and fungi were incubated for 12 hr. GFP and 4′,6-diamidino-2-phenylindole (DAPI) fluorescence were observed under a confocal laser scanning microscope  (Li et al., 2019;Xiong et al., 2019;Yu et al., 2014).

Consistent with the
In this study, we identified a novel FoTOR1 interacting protein, FoTIP4, in a yeast library screening ( Figure 5

| Fungal strains and culture conditions
The F. oxysporum strain was isolated from Chongqing local potato with dry rot (Fusarium wilt) and was verified through sequencing of the internal transcribed spacer (ITS

| Construction of vectors for gene deletion and complementation
The primers used to amplify the flanking sequences or coding sequence of each gene are listed in Table S6. Gene deletion and complementation of F. oxysporum were carried out as described previously (Luo et al., 2016). Agrobacterium tumefaciens AGL-1 was used to transform the conidia of F. oxysporum. A. tumefaciensmediated transformation was performed as described previously (Maruthachalam et al., 2011). Randomly selected transformants were transferred to fresh PDA with hygromycin.

| Construction and screening of a Y2H library
Total RNAs were extracted from mycelium of F. oxysporum using the RNAprep Kit (TIANGEN). The first-strand cDNA was obtained using 1-2 μg of total RNA as template. This first-strand cDNA was then used as template for low-cycle (c.20) long-distance PCR (LD-PCR) amplification to generate 3-6 μg double-stranded cDNA (dscDNA).
These dscDNAs were purified using CHROMA SPIN TE-400 columns (Takara) to obtain DNA for library construction. The cDNA library for Y2H screening was fused with the GAL4 activation domain of the pGADT7 vector as prey using the Matchmaker Gold Yeast Two-Hybrid System (Clontech). The FoTOR1 (FOXG_18412) gene was fused with the GAL4 DNA-binding domain in pGBKT7 to ensure that there was no autoactivation and toxicity, and the FoTOR1 fusion protein was used as bait to identify interacting proteins. Co-transformation of the library cDNA (prey) and the plasmid pGBKT7-FoTOR1 (bait) into Y2HGold yeast strain allowed interaction between prey and bait.

| Yeast two-hybrid assay
In order to construct plasmids for Y2H analyses, the coding sequences of genes were amplified from the cDNA of F. oxysporum.
The genes were inserted into the yeast GAL4 binding domain vector pGBKT7 or the GAL4 activation domain vector pGADT7 (Clontech). Pairs of Y2H plasmids were co-transformed into S. cerevisiae Y2HGold following the PEG/LiAc transformation protocol (Clontech). Transformants were grown at 28 °C for 5 days on synthetic medium lacking leucine and tryptophan (SD−Leu−Trp).
Then yeast colonies were transferred to synthetic medium lacking histidine, leucine, tryptophan, and adenine (SD−His−Leu−Trp− Ade) containing 40 µg/ml Xα-Gal and 200 ng/ml aureobasidin A as described in the manual (Clontech). The pair of plasmids pGBKT7-53 and pGADT7-T was used as a positive control. The pair of plasmids pGBKT7-Lam and pGADT7-T was used as a negative control. Three independent experiments were performed.

| Fluorescence microscopy
The full-length FoTIP4 encoding sequence was subcloned downstream of the FoTIP4 promoter in Gateway entry vector p8GWN to

| Electrophoretic mobility shift assay
The full-length Kit (Beyotime). The same unlabelled DNA fragment was used as a competitor, while the RRPE or PAC box within a probe changed into AAAAAAAA was used as a negative control. The EMSA was performed using the electrophoretic mobility shift assay kit (Beyotime) according to the manufacturer's instructions.

| Transcriptome sequencing and analysis
Hyphae of F. oxysporum were grown for 4 days in PDB at 27 °C with shaking at 160 rpm, treated with 1 µM RAP or DMSO (as a control), and incubated for 12 hr. Total RNA of F. oxysporum mycelium was isolated using the RNAprep Pure Plant Kit (TIANGEN). For each treatment, three independent biological replicates were performed.
An Illumina HiSeq 2000 platform was used to sequence the cDNA library, and 100-bp paired-end reads were generated. The clean reads were mapped to the F. oxysporum reference genome using TopHat2 software. Cufflinks and Cuffdiff were used to assemble the mapped reads and identify DEGs, respectively. GO enrichment (corrected p value < 0.05) of the DEGs was performed using GOseq software.

| RT-qPCR
Total RNA of F. oxysporum mycelium treated for 12 hr in PDB containing DMSO or RAP (1 μM) was isolated using the RNAprep Pure Plant Kit (TIANGEN). Relative transcript levels were assayed by one-step real-time PCR analysis using the CFX96 real-time PCR system (Bio-Rad). Real-time PCR primers were designed using Primer Premier v. 5.0 (details are presented in Table S6). FoEIF1α was used as an internal control. The data are presented as the mean ± SD of three independent experiments.

| Combination index value measurement
CI values were used to quantitatively measure the interaction between Torin1 and RAP. The interaction is categorized as synergism (CI < 1), additive effects (CI = 1), or antagonism (CI > 1) (Chou, 2006

| Pathogen inoculation and cellophane invasion assays
Pathogen inoculation was performed by point inoculation on the surface of potato leaves and tubers with conidia of wildtype F. oxysporum, ΔFotip4 mutants, and the complemented strain (ΔFotip4 + FoTIP4) (10 7 conidia/ml) as described previously (Thatcher et al., 2009). Inoculated leaves and tubers were cultured on moist filter paper at 27 °C in a short daylight condition for 4 days. Each strain was inoculated on at least 10 potato leaves or tubers every time. The cellophane invasion assay was performed as described previously (Prados Rosales & Di, 2008). Each experiment was repeated at least three times.

ACK N OWLED G EM ENTS
This work was supported by grants from the National Natural L.L., T.Z., R.D., and M.R. wrote the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.