The Transcription Factor VdHapX Controls Iron Homeostasis and Is Crucial for Virulence in the Vascular Pathogen Verticillium dahliae

This study demonstrated that VdHapX is a conserved protein that mediates adaptation to iron starvation and excesses, affects microsclerotium formation, and is crucial for virulence of V. dahliae.

I ron is an essential trace element for almost all organisms. As an indispensable cofactor of protein function, redox-active iron participates in multiple cellular processes, such as oxidative phosphorylation and detoxification of oxidative stress (1). However, iron excess is toxic to the cell (2). Thus, iron homeostasis is finely controlled to maintain the subtle balance between uptake, storage, and utilization of iron.
Iron homeostasis is primarily maintained by two transcription factors in filamentous fungi, including the bZip transcription factor, HapX, and the GATA zinc finger transcription factor, SreA, which are interconnected by a negative transcriptional feedback loop, i.e., in Aspergillus nidulans (3,4) and Aspergillus fumigatus (5). During iron starvation, HapX represses SreA-mediated iron utilization pathways and activates siderophore biosynthesis for iron uptake (5)(6)(7). In periods of iron sufficiency, SreA represses HapX expression and siderophore-mediated iron acquisition (3,7). Furthermore, HapX functions by physical interaction with the heterotrimeric CCAA T-binding complex (4). HapX is also critical for iron detoxification by triggering expression of the vacuolar iron importer CccA under conditions of excess iron (7,8). In addition to the core role of HapX in iron homeostasis, deletion of HapX leads to attenuated virulence in A. fumigatus (5), Candida albicans (9, 10), and Cryptococcus neoformans (11). In contrast, HapX of the human-pathogenic dermatophyte Arth-roderma benhamiae is not a virulence determinant (12). Overall, the roles of HapX orthologs in plant vascular wilt fungi have not been well investigated thus far, except in the soilborne fungus Fusarium oxysporum (6).
Vascular wilt caused by Verticillium dahliae is a destructive plant disease that poses a threat to crop production and forest health worldwide (13,14). The fungus infects more than 200 plant species, and an increasing number of new hosts are continually identified (15). In China, apart from the cultivated crop plants, ornamental and landscape plants like smoke trees (Cotinus coggygria) are also infested by V. dahliae (13). The fungus infects its host through the root and colonizes and propagates in xylem vessels (16). Once the plant is infected, no available fungicides can effectively treat the disease (14).
The V. dahliae life cycle comprises three stages, including parasitic and saprophytic stages in xylem and a dormant stage in the soil as long-lived survival structures known as microsclerotia, which play crucial roles in disease spread (14). The xylem transports water and soluble mineral nutrients from the roots throughout the plant. As such, xylem sap is not rich in nutrients and contains lots of organic acids, including a small amount of amino acids (17). During parasitic and saprophytic stages, as anticipated for similar plant pathogens, V. dahliae must employ its iron uptake system to compete for host iron resources.
The relatively recent availability of the V. dahliae genome sequence has greatly accelerated investigations into the mechanisms fundamental to its life cycle and disease progression (18). Multiple genes have been studied for their roles in nutrient uptake and adaptation to adverse circumstances, including iron limitation or excess. For example, FreB, the gene for which encodes a ferric reductase and which is involved in the reduction of the ferric iron to available ferrous iron, has been studied in V. dahliae (19). VdSNF1, the sucrose-nonfermenting protein kinase, was verified as a regulator of catabolic repression and the expression of genes involved in cell wall degradation in V. dahliae (20). Cpc1, a regulator of cross-pathway control, controls amino acid biosynthesis in Verticillium longisporum (21). However, it is unknown how V. dahliae copes with iron limitation and toxicity and whether the fungus coordinates iron homeostasis in response to changing iron levels via conserved iron-regulatory systems found in other fungi. The role of VdHapX in iron adaptation during iron starvation and sufficiency has not been elucidated in V. dahliae.
In this study, we showed that the VdHapX transcription factor is a major regulator of iron homeostasis, allowing adaptation to both iron-depleted and iron-excess conditions. Further, we revealed that VdHapX is crucial for virulence and detoxification to iron or H 2 O 2 excess and affects microsclerotium formation. These results demonstrated a key role of VdHapX in iron homeostasis, virulence, and H 2 O 2 detoxification in V. dahliae.

RESULTS
Loss of VdHapX decreases growth and sporulation. A BLASTP (22) search using the amino acid sequence of HapX of F. oxysporum revealed a single significant match, VDAG_05022, in a reference genome of V. dahliae (23). This match showed significant similarity to HapX of F. oxysporum (63.1% identity) and A. nidulans (56.5% identity). VDAG_05022 encodes a 456-amino-acid protein having a basic-leucine zipper (bZip) domain. Further phylogenetic analysis showed that VDAG_05022 is also highly homologous to other fungal HapX proteins (see Fig. S1 in the supplemental material). Thus, VDAG_05022 was designated VdHapX.
To study the role of VdHapX in V. dahliae, two VdHapX mutant strains (designated ΔVdHapX-11 and ΔVdHapX-12) were identified, and deletion of VdHapX was confirmed in both (Fig. S2). In addition, the wild-type VdHapX copy was ectopically reintroduced into the ΔVdHapX strain, yielding the ΔVdHapX/VdHapX complemented strain (Fig. S2). The ΔVdHapX strain displayed no obvious growth defect on potato dextrose agar (PDA), but the mycelial growth was reduced even under the iron-replete condition (Fig. 1A). Aerial growth was eliminated in the presence of the iron chelator bathophenanthroline disulfonate (BPS) ( Fig. 1A and C). Furthermore, the ΔVdHapX strain showed strikingly reduced radial growth in the presence of elevated iron levels (0.06 mM) or BPS (0.4 mM) compared with the growth of wild-type strain XS11 (Fig. 1B). In iron-starved or iron-replete liquid culture, the ΔVdHapX strain significantly decreased the biomass production to 61% or 70% of XS11, respectively (Fig. 1D). Analysis of conidiation revealed obviously reduced conidiation of the ΔVdHapX strain during iron starvation or sufficiency. Deletion of VdHapX resulted in the production of only 30% of the amount of conidia, compared with XS11 (Fig. 1E). In addition, the ΔVdHapX/VdHapX complemented strain restored the respective phenotypes similar to those of XS11 in all experiments (Fig. 1). These results demonstrated that VdHapX is required for appropriate growth and conidiation under iron-starvation conditions. VdHapX is required for transcriptional regulation of SrbA and SreA. We sought to investigate the role of VdHapX in the presence or absence of iron by reverse transcription-quantitative PCR (RT-qPCR). Transcript levels of VdHapX were significantly upregulated (2-fold changes) in XS11 during iron starvation, indicating that VdHapX is an iron-repressed gene ( Fig. 2A). We next examined the role of VdHapX in the transcriptional regulation of known iron-regulatory genes, such as srbA (VDAG_01557, sterol regulatory element binding protein) and sreA (VDAG_00352, GATA transcription factor). After normalization against the ␤-tubulin gene, Fig. 2B shows that the srbA transcript levels were significantly downregulated in the ΔVdHapX strain compared with those of XS11 under both iron starvation and iron sufficiency but were not affected by iron availability (Fig. 2B), indicating that the induction of srbA is independent of VdHapX. However, expression of sreA was highly upregulated (about 13-fold) during a 45-min shift from iron starvation to iron sufficiency in XS11. Additionally, we found that the sreA transcript levels were significantly upregulated under iron starvation compared with iron sufficiency by VdHapX deficiency (Fig. 2B), suggesting that VdHapX represses SreA under iron-limiting conditions. Further examination revealed that the level of intracellular iron concentration in the ΔVdHapX strain was comparable to that of XS11 (Fig. 2C), suggesting that VdHapX is not required for iron uptake in V. dahliae.
Deletion of VdHapX leads to misregulation of siderophore biosynthesis. To further explore the function of VdHapX in the regulation of siderophore biosynthesis, we monitored the expression of key genes involved in iron uptake. First, we identified putative orthologs of key siderophore biosynthetic genes characterized in Aspergillus except sidG (coenzyme A [CoA]-N 2 -transacetylase) (24). Four homologs of these genes exist in V. dahliae, including the monooxygenase-encoding sidA (VDAG_05313), siderophore nonribosomal peptide synthetases sidC (VDAG_05314) and sidD (VDAG_03964), and the putative siderophore transporter mirB (VDAG_07020). Interestingly, the orthologs of sidA and sidC are clustered together within 21.5 kb in the genomic sequence of V. dahliae (strain VdLs.17), and the intergenic region between the two genes was 4.4 kb. The RT-qPCR analysis showed that the transcript levels of the sidA, sidC, and sidD orthologs were highly upregulated under iron sufficiency compared with those under conditions of iron starvation in XS11. Expression of mirB was highly upregulated under iron starvation compared with the levels observed during iron sufficiency in XS11 (Fig. 3). Also, the transcript levels of the sidA and sidD orthologs were highly increased in the ΔVdHapX strain in comparison to XS11 during iron starvation (Fig. 3); however, Values are the averages from four biological replicates, consisting of three technical replicates each. Error bars represent standard deviations. (C) Total intracellular iron concentration measurment. XS11, ΔVdHapX mutant, and ΔVdHapX/VdHapX complemented strains were grown in iron-depleted MM for 3 days and transferred for 45 min to iron-depleted MM (-Fe) or iron-replete MM (ϩFe). The intracellular iron concentration was determined using an optical emission spectrometer. Error bars represent the standard deviations based on three independent replicates with three technical replicates. Asterisks represent significant differences (P Ͻ 0.01). expression of mirB was decreased in the ΔVdHapX strain, and no differences in transcript levels of sidC were observed between the ΔVdHapX strain and XS11 under iron starvation. In contrast, the transcript level of mirB was slightly upregulated (about 2-fold) in XS11 during iron starvation. These results indicated that VdHapX might regulate siderophore production in a way unique to V. dahliae.
VdHapX is involved in iron detoxification. We next examined potential functional roles of VdHapX in the presence of excess iron. Disruption of VdHapX rendered V. dahliae more susceptible to iron toxicity than XS11 on high-iron medium. Notably, in the presence of 10 mM Fe 2ϩ or Fe 3ϩ , the ΔVdHapX strain displayed a strong growth defect compared with XS11 and the complemented strains (Fig. 4A). Mycelial growth of the ΔVdHapX strain was reduced 3-fold relative to that of XS11 grown on high-iron medium (Fig. 4B). We performed RT-qPCR analyses of a cccA ortholog, which encodes a vacuolar iron importer (8), and found that expression of this ortholog (VDAG_10085) was highly induced (Ͼ300-fold change) under normal iron conditions (0.03 mM Fe 2ϩ or Fe 3ϩ ) in XS11 (Fig. 4C). However, the induction of the cccA ortholog was reduced under high-iron conditions (10 mM Fe 2ϩ or Fe 3ϩ ) in which the ΔVdHapX strain was used as background (Fig. 4D). These data indicated that VdHapX is required for the activity of iron detoxification, mainly via CccA-mediated iron storage.
Apart from cccA, previous reports showed that HapX regulates many genes involved in iron utilization in F. oxysporum (6). We inspected the V. dahliae genome using BLASTP and identified six representative genes related to iron utilization pathways in fungi. These genes included cycA (VDAG_00910), encoding cytochrome c; acoA (VDAG_02332), encoding aconitase; lysF (VDAG_08540), encoding homoaconitase; hemA (VDAG_06343), encoding aminolevulinic acid synthase; VDAG_00564, encoding isopropylmalate dehydratase; and VDAG_04620, encoding dihydroxy acid dehydratase. After a 45-min shift from iron-limiting to iron-replete conditions, deletion of VdHapX impairs the transcriptional activation of VDAG_02332, VDAG_08540, VDAG_06343, and VDAG_04620. However, the transcript levels of VDAG_00910 and VDAG_00564 were significantly increased in the ΔVdHapX strain during iron starvation, but not under iron-replete conditions, in comparison to XS11 (Fig. 5). Together, these results indicate that VdHapX contributes  fected H 2 O 2 sensitivity. Using H 2 O 2 diffusion tests, we showed that the ΔVdHapX strain exhibited modestly increased sensitivity to different concentrations of H 2 O 2 (Fig. 6A). Inhibition zone was determined based on an H 2 O 2 diffusion zone diameter at 4 days postinoculation (dpi). Using a t test, the difference was considered significant at P Ͻ 0.01 (Fig. 6B). To elucidate the molecular underpinnings of H 2 O 2 sensitivity in the ΔVdHapX strain, we examined the expression levels of four orthologs of genes (cat1, VDAG_03661; sod, VDAG_02630; sod_Cu, VDAG_08724; and yap1, VDAG_01588) associated with H 2 O 2 detoxification. With the exception of yap1 (VDAG_01588), transcripts of the other three genes in the analyses were significantly downregulated in the ΔVdHapX strain compared to those in XS11 (Fig. 6C). These data suggested that VdHapX relieves oxidative stress by regulating expression of genes that detoxify H 2 O 2 .
VdHapX positively regulates microsclerotium formation. To examine whether the ΔVdHapX strain was defective in microsclerotium production, we observed microsclerotium formation on solid medium. Four days after inoculation, the ΔVdHapX strain produced melanized microsclerotia similarly to XS11 and the ΔVdHapX/VdHapX strain (Fig. 7A). Furthermore, microscopic examination revealed that the ΔVdHapX strain produced 75% fewer microsclerotia than XS11 (Fig. 7B). The result indicated that VdHapX positively regulates microsclerotium formation.
VdHapX is crucial for virulence. We also sought to determine whether VdHapX contributes to virulence. Compared to XS11 and the ΔVdHapX/VdHapX strain, loss of VdHapX severely attenuated fungal pathogenicity on smoke trees. Smoke tree seedlings inoculated by the ΔVdHapX strain exhibited slight chlorosis at 35 dpi, whereas the seedlings inoculated with XS11 and the complementation strain showed obvious wilt symptoms and reduced plant height (Fig. 8A). We further examined the infection process in more detail, revealing that the ΔVdHapX mutant exhibited delayed hyphal penetration of a cellophane membrane compared to penetration observed in XS11 and the ΔVdHapX/VdHapX strain (Fig. 8B). However, the ΔVdHapX strain can form hyphopo- dia for penetration peg development like those of XS11 and the ΔVdHapX complemented strain (Fig. 8C). A penetration assay on onion epidermis also revealed that deletion of VdHapX delayed formation of invasive hyphae (IH). Compared with XS11 and the complementation strain, IH were not observed on the onion epidermis inoculated with the ΔVdHapX strain at 48 h postinoculation (hpi). However, at 96 hpi, the ΔVdHapX strain could form IH, like those of XS11 and the complemented strain (Fig. 8D). Furthermore, we assayed whether the mutant was able to reach and proliferate within the vascular vessels, and thus, surface-sterilized stem sections were collected at 35 dpi and placed onto PDA. Although the ΔVdHapX strain could be reisolated from stem tissues like XS11 and the complementation strain, the rate of isolation was significantly lower than in XS11 and the complemented ΔVdHapX strain (data not shown). Together, these data demonstrated that VdHapX regulates penetration peg formation and proliferation during the initial colonization of cotton roots.

DISCUSSION
To overcome iron deficiency or excess, pathogenic fungi have evolved sophisticated regulatory mechanisms that mediate resistance and adaptation to iron limitations or excesses. As shown here, the bZip transcription factor VdHapX is required for ion homeostasis under iron-limiting or iron-excess conditions in V. dahliae. Loss of VdHapX affects hyphal growth, microsclerotium formation, conidiation, and adaptation to oxidative stress and virulence and is necessary for adaptation to oxidative stress. Therefore, VdHapX is critically important in the life cycle and for the survival of V. dahliae.
In previous studies, HapX deficiency has resulted in significant reductions in mycelial growth on medium under an iron-limited condition, i.e., in F. oxysporum (6), A. fumigatus (5), C. neoformans (11), Arthroderma benhamiae (12), and A. nidulans (4). Compared with transcript levels observed under conditions of iron sufficiency, HapX and its orthologs are transcriptionally upregulated and repress iron-dependent pathways during iron starvation in fungi (4-6, 11, 12, 25). In agreement with previous studies, strains lacking VdHapX showed a striking reduction in growth on medium with limited iron availability. Importantly, HapX is essential for iron detoxification as well. Consequently, deletion of HapX impaired mycelial growth in V. dahliae and A. fumigatus under with BM during a 7-day incubation period. Photographs of the microsclerotia in a field were taken and converted into an 8-bit grayscale image by using Image J. Using pixels as units in the measurement area, the pixel values of the covered area were measured and the relative ratio was calculated. There were three independent fields in which microsclerotia were enumerated. Error bars represent the standard deviations based on three independent replicates. The asterisk indicates a significant difference at P Ͻ 0.05. conditions of excess iron availability (7), demonstrating that the role of HapX in iron detoxification is conserved. Furthermore, HapX mainly activates the vacuolar iron importer cccA to adapt to high iron, and certain domains of HapX are necessary for the Janus-type transcription factor functions during iron starvation or under high-iron conditions as an activator or repressor (7). Similarly, in our study, knockout of HapX failed to result in accelerated expression of cccA, whose expression was highly upregulated during a shift from iron starvation to iron-rich medium in the wild type. As shown in A. nidulans, HapX functions via binding to a CCAAT motif to target promoter regions (4). Interestingly, the motif is also found in the putative promoter regions of genes regulated by VdHapX in V. dahliae (see Fig. S3 in the supplemental material), suggesting that the HapX binding motif is evolutionarily conserved. The induced expression of cccA when cultured under adequate iron levels was higher (Ͼ5-fold) than that observed under excess iron conditions in the wild type (Fig. 4), indicating that HapX represses or activates cccA depending on the ambient iron availability.
Deletion of HapX in V. dahliae led to activation of the siderophore biosynthesis genes sidA and sidD during iron starvation. Similarly, inactivation of HapX in F. oxysporum led to increased transcript levels of several siderophore genes during iron starvation (6). In contrast to this observation, knockout of HapX in V. dahliae resulted in repression of sidC and mirB under iron-limited conditions. However, there is an additional report of a HapX mutant that exhibited downregulation of genes involved in siderophore biosynthesis under iron-limiting conditions in A. fumigatus (5). As previously shown, transcription factors HapX and SreA are interconnected through a negative-feedback-loop-orchestrated transcriptional regulation of iron homeostasis One-year-old smoke tree seedlings were inoculated and incubated for 10 min with a 10 6 -conidium/ml suspension of XS11, the ΔVdHapX strain, and the ΔVdHapX/VdHapX strain. The smoke tree seedlings were inoculated with distilled water (CK), also for a 10-min incubation period. All seedlings were replanted in soil for 35  and siderophore biosynthesis in fungi, such as in A. nidulans (4) and in A. benhamiae (12). Strikingly, the transcript levels of srbA did not vary between iron starvation and sufficiency, whereas sreA was primarily induced by iron sufficiency as opposed to iron starvation in the wild-type strain. Expression of sreA and srbA was markedly reduced in the ΔVdHapX strain compared with the wild-type strain (Fig. 2). Both SreA and SrbA have been demonstrated to contribute to the activation of siderophore production (26). However, the roles of SreA, SrbA, and HapX in siderophore production of V. dahliae require further characterization, in addition to those of the ferric reductase FreB (19). Interestingly, intracellular iron concentration is not changed in the ΔVdHapX strain compared with the wild-type strain. Potentially, the compartmentalization of iron could be changed in response to increases in external concentrations, though not the total iron concentration. If increased external stimulus (iron) is perceived, signaling from this stimulus may be channeled to trigger decreases in growth processes and not necessarily a simultaneous change in intracellular iron concentration. Nevertheless, data presented in this study indicate a clear role of the V. dahliae HapX in iron homeostasis and regulation of siderophore biosynthesis.
In addition to altered expression of iron-regulatory genes in the ΔVdHapX strain, VdHapX deletion in V. dahliae resulted in decreased growth and conidiation, as well as decreased melanized microsclerotium formation under both iron-limited and ironreplete conditions. In a previous study characterizing transcriptomes during microsclerotium formation, Xiong et al. (27) reported upregulation of genes involved in carbohydrate and protein metabolic processes. In agreement, transcript levels of cccA, acoA, lysF, and hemA were highly downregulated in the ΔVdHapX strain. Not only are these genes involved in amino acid metabolism and respiration, they also have direct roles in vacuolar iron storage and heme biosynthesis, suggesting that VdHapX contributes to microsclerotium formation by also regulating genes related to iron acquisition and metabolism.
Ferrous iron can be rapidly oxidized to produce ferric iron and hydroxyl radicals, both of which may be toxic (28). Ferric reductases are integral membrane proteins involved in the reduction from ferric iron to ferrous iron, and the process is vital for iron uptake (29). Fungal ferric reductases not only play a role iron reduction but also are associated with sensitivity to oxidative stresses. Fungal strains lacking ferric reductases exhibited hypersensitivity to oxidative stress, i.e., in C. albicans (30), A. fumigatus (31), and V. dahliae (19). In our study, the ΔVdHapX strain displayed significantly increased sensitivity to H 2 O 2 . Analyses of the expression of genes associated with H 2 O 2 detoxification also demonstrated that these key genes were downregulated in the mutant. Similarly, deletion of HapX in F. oxysporum led to a suppression of the expression of genes associated with oxidative stress detoxification, such as FOXG_12260 (peroxidase) and FOXG_00142 (cytochrome c peroxidase) (6). Together, our data coupled with these previous observations clearly demonstrate that HapX is required for the irondependent adaptation to oxidative stress in fungi.
The critical role of iron acquisition in virulence has been shown in pathogenic fungi. HapX was shown to be important for virulence of F. oxysporum (6), A. fumigatus (5), C. albicans (10), and C. neoformans (11). As shown here, the ΔVdHapX strains exhibited a clear defect in virulence. Further observations showed that the mutant had delayed penetration into the cellophane membrane and onion epidermis and defective proliferation into the xylem tissues of smoke tree seedlings. However, the ties or interconnectivity between the iron regulatory system, virulence, and H 2 O 2 detoxification are areas that remain to be explored at the genetic and biochemical levels.

MATERIALS AND METHODS
Fungal strains and culture conditions. The wild-type V. dahliae strain XS11 for these experiments was isolated from a smoke tree in Fragrant Hills, Beijing, China (32). The same strain was used for preparation of the gene replacement and complementation strains in this study. The conidia of all fungal strains were stored in 30% glycerol solution at Ϫ80°C. XS11 was cultured on PDA plates, and the transformants of V. dahliae XS11 were grown on PDA plates (1 liter PDA, 200 g potato, 20 g glucose, 15 g agar) supplemented with 50 g/ml Geneticin or 25 g/ml hygromycin as appropriate for selection following the respective gene replacement.
To observe microsclerotium formation, solid BM (1 liter BM, 10 g D-glucose, 0.2 g NaNO 3 , 0.52 g KCl, 0.52 g MgSO 4 ·7H 2 O, 1.52 g KH 2 PO 4 , 3 M vitamin B 1 , 0.1 M biotin, 15 g agar) was evenly applied to the sterile slides. Fresh conidia (0.2 l of 1 ϫ 10 5 conidia/ml) of XS11, the ΔVdHapX strain, and the complemented strain were inoculated onto culture medium for 7 days, and the slide was placed in a sterile petri dish for 5 days at 25°C. Observation and analysis of relative ratio were performed using a microscope (Leica DM 2500) and Image J software.
For plant infection assays, all strains were grown in liquid complete medium (CM; 50 ml 20 nitrate salts, 1 ml 1,000ϫ trace, 10 g glucose, 2 g peptone, 1 g yeast extract, 1 g Casamino Acids, 1 ml vitamin solution) at 25°C with shaking at 200 rpm for 7 days prior to use.
To test the response to peroxide, conidia were uniformly diluted with the PDA medium to 10 6 conidia/ml and spread onto several plates. Sterilized filter paper having a diameter of 5 mm was placed in the center of the plate, and a 10-l droplet of either a 2.5%, 5%, or 10% hydrogen peroxide solution was dropped on the filter paper pieces. Plates were held at 25°C for 7 days prior to observations of the zone of inhibition.
Bioinformatics analysis. The complete sequence of VdHapX was downloaded from JGI (https:// genome.jgi.doe.gov/Verda1/Verda1.home.html) by using BLASTP with the HapX gene of F. oxysporum. The reference genome of V. dahliae (23) was used for BLASTP queries of the V. dahliae genome to identify VdHapX. VdHapX orthologs in other fungi were found in NCBI and JGI. The full-length protein sequences of VdHapX homologs in different fungi were compared, and the phylogenetic tree was constructed with MEGA version 6.0 (33).
Targeted gene knockout and mutant complementation. The entire open reading frame region of the VdHapX gene was replaced with a hygromycin resistance gene cassette constructed using the split-marker method (34). The 5= and 3= flanking sequences of VdHapX were cloned with primers PL906/PL907 and PL908/PL909 and then linked with the hygromycin resistance gene cassette using primers PL906/PL938 and PL939/PL909, respectively. The hygromycin resistance cassette was sequenced by primers M13-F and M13-R. Subsequently, the 5= sequence was connected with the hygromycin resistance gene cassette by fusion PCR with primer PL906/Hy-R, and the 3= sequence was fused to the hygromycin resistance gene cassette by fusion PCR with the primer Yg-F/PL909. Finally, the PCR fragments were sequenced and used for polyethylene glycol (PEG)-mediated protoplast transformation as described by Wang et al. (32). The transformations were screened with 30 g/ml hygromycin B. Mutants were detected by PCR and validated by Southern blotting. The complementation strain was selected with 50 g/ml Geneticin by reintroducing the wild-type copy of VdHapX.
Genomic DNA was extracted by a cetyltrimethylammonium bromide (CTAB) method, and Southern blotting was performed to confirm the deletion of VdHapX by the digoxigenin (DIG) High Prime DNA labeling and detection starter kit I according to the manufacturer's protocol (Roche, Germany). The probe fragment for the Southern blot was amplified from the V. dahliae strain XS11 genomic DNA with primers PL963 and PL907 and labeled with DIG primer. The restriction enzyme KpnI was used to digest the genomic DNA extracted from the wild-type strain and mutant strains.
Gene expression analyses. Total RNA was extracted from the XS11 and the ΔVdHapX strain using TRIzol reagent (Invitrogen) and purified with a PureLink RNA minikit (Ambion). RNA was reverse transcribed using SuperScript III reverse transcriptase (Invitrogen) and oligo(dT) to obtain cDNA used for subsequent experiments. For analyzing the expression of genes regulated by VdHapX in different strains under iron-replete or iron-limited conditions, all strains were grown in liquid MM for 3 days at 25°C and 200 rpm and transferred into liquid MM and liquid MM with 0.03 mM Fe 2 (SO 4 ) 3 , respectively, for 45 min prior to RNA extraction. The mycelia were collected and frozen with liquid nitrogen immediately. Reverse transcription-qPCR was performed with the SuperReal Premix Plus (Tiangen, China) using SYBR green dye and an ABI 7500 real-time PCR system (Applied Biosystems, USA). The genes were tested independently and in triplicate. The results of RT-qPCR were analyzed using the threshold cycle (ΔΔC T ) method (35). The ␤-tubulin gene was used as the internal reference. The XS11 strain in liquid MM was used as a control group for all analyzes. The gene numbers and primers used in the experiments are listed in Table S1 in the supplemental material.
For analyzing the effects of VdHapX gene knockout on reactive oxygen stress, XS11 and the ΔVdHapX strain were grown in liquid MM or liquid MM containing 0.015 mM Fe 2 (SO 4 ) 3 for 3 days at 25°C, respectively. Mycelia were collected and transferred into liquid MM with or without 0.015 mM Fe 2 (SO 4 ) 3 and 1 mM H 2 O 2 after incubation for 45 min. The mycelia were collected for RNA extraction, and the methods of qPCR and data processing were the same as mentioned above. Primer sequences are listed in Table S1.
To detect the effect of deletion of the VdHapX gene on the ability to detoxify high concentrations of ferrous and ferric iron, XS11, the ΔVdHapX strain, and the ΔVdHapX complemented strain were first grown in liquid MM for 3 days at 25°C and 200 rpm. Mycelia were collected and transferred into liquid MM with or without 0.015 and 5 mM Fe 2 (SO 4 ) 3 for 45 min and then collected for RNA extraction. The methods of RT-qPCR and data processing were the same as described above. All primers in this assay are listed in Table S1.
Determination of intracellular iron content. The XS11, ΔVdHapX, and ΔVdHapX/VdHapX strains were inoculated on liquid MM for 3 days at 25°C and 200 rpm and were transferred to fresh MM for 45 min with or without 0.03 mM Fe 3ϩ . The mycelia were collected, rinsed with sterile water, and lyophilized using a vacuum freeze-dryer. The lyophilized mycelia were ground to a fine powder. Two hundred milligrams of the ground hyphae was weighed, placed in vessels of a microwave sample preparation system (Perkin Elmer, USA), and microwave digested for 2 h. The concentration of iron in the solution was determined using an optical emission spectrometer (Perkin Elmer, USA).
Plant infection assays.
One-year-old smoke tree seedlings were used for pathogenicity assays throughout the study. Conidia from XS11, the ΔVdHapX strain, and the ΔVdHapX complemented strain were obtained and adjusted to 10 6 conidia/ml in sterile water. Plant roots were incubated in a 10 6 -conidium/ml suspension for 10 min. Control plants were mock inoculated with distilled water. All of the plants were replanted into sterile soil, placed in a greenhouse, and observed after a period of 35 days.
To observe the ability of V. dahliae mycelium to penetrate and to form hyphopodia, cellophane membranes were overlaid on the MM plates. XS11, ΔVdHapX, and ΔVdHapX complemented strains were inoculated on cellophane for 3 days at 25°C. The cellophane was removed from the surface of the plate following incubation for 2, 4, or 6 days, and the plate was maintained for another day to observe if colonies had grown on the plates. The colonies on the cellophane were rinsed to remove the residual conidia with sterile water, allowing observations of the remaining hyphae.
The onion epidermis was soaked with alcohol and water and placed on the sterilized slides for observations of the hyphal penetration into epidermal cells. Fresh conidia (10 5 conidia/ml) of XS11, ΔVdHapX, and ΔVdHapX complemented strains were inoculated onto the hydrophobic surface of the onion epidermis, and the sterilized slides were placed inside sterile petri dishes to them keep moist at 25°C. These slides were observed every 24 h, and the mycelium was stained with trypan blue staining solution (0.3 ml 1% trypan blue stock, 10 ml lactic acid, 10 ml phenol, 10 ml distilled water [dH 2 O]).