Qip gene in Fusarium oxysporum is required for normal hyphae morphology and virulence

Ribonucleic acid (RNA)-silencing mechanisms exist in many eukaryotes to regulate a variety of biological processes. The known molecular components are related to Dicers, Argonautes and RNA-dependent RNA polymerases. Previous biochemical studies have also suggested that Qip, with an exonuclease domain, facilitates the conversion of duplex small interfering RNAs into single strands. In our study, the Qip gene in Fusarium oxysporum was disrupted using homologous recombination technology. The deletion of the Qip gene resulted in a decrease in colony growth rates but increased the number of branches. Additionally, the ΔQip mutant had a reduced pathogenicity in cabbage. Our results show Qip gene in F. oxysporum is required for normal hyphae morphology and virulence. The mutant will be useful for elucidating the relationship between the RNA-silencing mechanism and hyphal growth and development in F. oxysporum.


Introduction
Fungal small ribonucleic acids (RNAs) of 20-40 nucleotides (nt) are involved in RNA-silencing, which can regulate the expression of target genes and therefore are involved in a variety of biological processes, such as development, antiviral defence and the maintenance of genomic stability (Brennecke et al. 2003;Sijen and Plasterk 2003;Baulcombe 2004;Lu et al. 2005). Based on their RNA precursors and biogenesis mechanisms, these small RNAs can be divided into various types in fungi, including QDE-2-interacting small RNAs (qiRNAs), microRNA-like RNAs (milRNAs), Dicer-independent small interfering RNAs (disiRNAs), small interfering RNAs (siRNAs), long terminal repeat retrotransposon-siRNAs (LTR-siRNAs) and tRNAderived RNA fragments (tRFs) (Jöchl et al. 2008;Lee et al. 2009;Nicolas et al. 2010;Nunes et al. 2011). The maturation of small RNAs in the RNA-silencing pathway is dependent on the RNase III ribonuclease Dicer, RNAdependent RNA polymerases (RdRps) and Argonaute proteins (Lee et al. 2010). Long double-stranded RNA is cleaved into siRNAs by Dicer (Carmell and Hannon 2004). Single-stranded siRNAs guide the RNA-induced silencing complex (RISC), which includes the Argonaute family of proteins, to combine with endogenous messenger RNA and mediate gene silencing (Martinez et al. 2002). Some forms of aberrant RNA are converted into dsRNA by RdRP, which is hypothesized to trigger the initiation of the pathway (Shiu and Metzenberg 2002).
The Qip exonuclease, first identified as a QDE-2 (an Argonaute homologue)-interacting protein in Neurospora crassa, converts duplex siRNA into single strands to activate the RISC (Maiti et al. 2007). Qip is required for the RNA-silencing machinery in quelling, which is vegetative-specific, meiotic silencing by unpaired DNA (MSUD) and normal sexual development (Lee et al. 2010;Xiao et al. 2010). Qip, together with QDE-2 and exosome, mediates the maturation of milR-1 milRNAs (Xue et al. 2012). However, the possible participation of the RNA-silencing machinery and mutant phenotypes modulated by distinct classes of endogenous small RNAs (esRNAs) in this process in Fusarium oxysporum is still unknown.
The filamentous fungus F. oxysporum is an important soil-borne phytopathogen that causes vascular wilt in many economic crops, such as cabbage, tomato, banana and watermelon, worldwide (Armstrong and Armstrong 1981), resulting in incalculable agricultural and economic losses. For effective blight control, research on F. oxysporum's physiological and pathogenic mechanisms is urgent. The completion of genome sequencing provides more information for gene functional studies (Thatcher et al. 2012;Schmidt et al. 2013). In this study, we report the phenotype of a ΔQip mutant of F. oxysporum.

Strains, media and growth conditions
F. oxysporum f. sp. conglutinans wild-type strain A8 (race 1), which is native to Italy, is now maintained at the Beijing Academy of Agriculture and Forestry Science. Microconidial suspensions were stored in glycerol at −80°C until used. This strain was used for the preparation of protoplasts in the targeted gene knockout experiments. The Escherichia coli DH5a strain used for standard cloning steps was grown in Luria-Bertani (LB) medium at 37°C (Hooykaas et al. 1977) supplemented with ammonia benzyl and agar for a solid medium. Potato dextrose agar (PDA, consisting of 20% potato, 2% dextrose and 1.8% agar) containing 200 μg/ml hygromycin B (Roche, Branchburg, NJ, USA) or 200 μg/ml neomycin (Amresco, Solon, OH, USA) was used to select hygromycin-resistant and neomycin-resistant transformants, respectively (Liang et al. 2014). To maintain their phenotypes, transformants were cultured on PDA plates with 100 μg/ml hygromycin B. The wild-type A8 strain and the Qip deletion mutants (ΔQip) were cultured at 28°C.

Generation of the Qip gene disruption cassette and mutant
To construct the Qip gene disruption cassette, 1.3-kb upstream and 1.2-kb downstream sequences of the Qip gene were amplified with the primer pairs Qip-Up-1F/1R and Qip-Down-1F/1R (Table 1), respectively. The PCR products, containing EcoRI-SpeI and HindIII-KpnI restriction sites, respectively, were cloned into pKOV21 using the same sites. The primer pairs Qip-Up-2F/2R and Qip-Down-2F/2R were used to amplify two 0.9-kb fragments, which were used to test for homologous recombination. Wild-type A8 strain protoplasts were prepared as described previously (Talbot et al. 1993). Protoplasts were transformed with the knockout plasmid pKO-Qip using a PEG-CaCl 2 -mediated procedure (Nakayashiki et al. 1999). A schematic diagram of the gene disruption strategy in F. oxysporum is shown in Figure 1. Transformants could grow on the PDA plates supplemented with hygromycin B (200 μg/ml), but could not grow on PDA plates supplemented with neomycin (200 μg/ml) represented putative mutants.

Identification of the Qip deletion mutant
The putative mutants were analysed further by PCR. The primer pairs Qip-Up-3F/3R and Qip-Down-3F/3R were

Identification of mutant phenotypes
To analyse phenotypes, fresh conidia were collected from potato dextrose broth (PDB), washed in sterile distilled water and finally resuspended in sterile distilled water to a concentration of 1 × 10 5 conidia/ml.

Colony morphology and margins of wild-type strain A8 and the Qip deletion mutant
For morphological analyses of the wild-type strain A8 and the Qip deletion mutant, 10 µl (1 × 10 5 conidia/ml) of the suspension was dropped onto the centre of PDA plates and cultured at 28°C for 5 days. Additionally, 1 µl (1 × 10 5 conidia/ml) of the suspension was cultured on PDA plates at 28°C for 36 h. The colony margin of wild-type strain A8 and the Qip deletion mutant were observed using a Zeiss Axiostar Plus microscope (Oberkochen, Germany).
2.5.2. Microscopic examination of mycelial morphology and conidial morphology Conidial suspensions from wild-type strain A8 and the Qip deletion mutant strain were examined using an Olympus BX-51 microscope. Fresh conidia were inoculated on PDB medium at a concentration of 1 × 10 5 conidia/ml and incubated on a shaker at 28°C for 72 h. Then, 5 µl of the culture was placed on glass slides for microscopic observation.

Plants and growth conditions
Disease-sensitive cabbage seeds were covered with sterile gauze and then hydrated in sterile distilled water at~37°C for 60 min. After treatments, seeds were incubated at 28°C for 1 day until germination.

Pathogenicity assays
The root-dip inoculation method was used to assess the virulence of the wild-type A8 strain and its Qip mutant on cabbage in this experiment (Ospina-Giraldo et al. 2003).

Identification of Qip from F. oxysporum
To identify Qip genes in F. oxysporum, we first downloaded all annotated protein sequences of F. oxysporum from NCBI (http://www.broadinstitute.org/annotation/genome/fusarium_group/MultiHome.html), and a total of 12 annotated protein sets derived from 12 F. oxysporum strains were obtained. We used the Qip (QDE-2) protein of N. crassa (id: ABQ45366) as the query to search against the annotated proteins of F. oxysporum. Interestingly, each of the 12 F. oxysporum strains had only one protein with a conserved DEDDh domain near the N-terminus (aa 26-122), which indicates they belong to the 3′-5′ exonuclease domain family, and could be Qip homologues. We further analysed the sequence features of these 12 putative F. oxysporum Qip proteins using a multiple sequence alignment. The amino acid sequences of F. oxysporum Qip protein were conserved, with all 12 being 495 amino acids in length ( Figure 2). Sequence diversity was detected in only 15 loci, suggesting that the functions of these proteins are also conserved. Based on the high sequence similarity among putative F. oxysporum Qip genes, we designed primer pairs to amplify the Qip gene in F. oxysporum f. sp. conglutinans wild-type strain A8 (race 1). As a result, we successfully amplified the Qip gene.

The Qip gene was disrupted in the mutant ΔQip
Hygromycin-resistant, neomycin-sensitive transformants were isolated for PCR testing. PCR using the primer pairs Qip-Up-3F/3R and Qip-Down-3F/3R were used to produce 1.7-kb and 1.3-kb fragments, respectively, and the Qip-1F/1R primers were used to detect the target gene, confirming the mutant was generated by homologous recombination (Figure 3(a)). Genomic DNA from the Qip deletion mutant and wild-type A8 was digested with the PstI restriction enzyme. Southern blots of the digested products were hybridized using the 1238-bp PCR product as the probe (Figure 3(b)). The hybridization bands of the wild-type A8 and Qip deletion mutant were consistent with the expected DNA fragments (3.9 kb and 3.1 kb, respectively, for digestion with PstI). Therefore, the genome of the ΔQip mutant contained only one copy of the hygromycin-resistant gene (Figure 3(c)).

Qip gene is not required for conidial morphology but quantitatively affects hyphal growth
Colonies formed by the ΔQip mutant on PDA plates had slow growth rates (Figure 4). Additionally, the ΔQip mutant's colony edges were phenotypically distinct from those of the wild-type A8, forming relatively compact colonies with more branching (Figure 4). We monitored the growth of the conidiogenic hyphae and noticed that there was no difference in the hyphal morphology between the wild-type and the Qip mutant in the first 16 h. After 16 h, the ΔQip mutant hyphae became more branched than those of the wild-type A8 (Figure 4). The conidial morphology was similar to that of the wild-type A8 (Figure 4). Normal filamentous fungi grow by the delivery of the components for cell wall extension to the hyphal tips (Girbardt 1957;Riquelme et al. 1998). Hyphal branching is generated when the biosynthetic capacity or volume of the cytoplasm in a hyphal tip has reached its maximum (Yarden et al. 1992). Therefore, in the ΔQip mutant, the cessation of hyphal tip elongation generates a signal to form hyphal branches. Numerous mutants have a defect in hyphal tip growth and branching (Perkins et al. 2000), including mutants in signalling pathways (Lee et al. 1998), the cytoskeleton (Xiang and Morris 1999), subunit of the Vacuolar H + (Bowman et al. 2000) and protein kinases (Wang et al. 2011). This result suggests that the Qip gene, a component of the RNA-silencing pathway, was active in other pathways to regulate normal hyphal growth and development. Other RNA-silencing genes, such as Dicer, Argonaute and RdRP, also have similar functions. Dcl-1 mutants accumulated two size classes of siRNAs and displayed reduced growth rates and altered hyphal growth in Mucor circinelloides (Nicolás et al. 2007). Deletion of the CaDCR1 gene in Candida albicans results in a severe slow growth phenotype (Bernstein et al. 2012). In M. circinelloides, Ago-1 was involved in the response to environmental signals in vegetative development (Cervantes et al. 2013). Two Dcr genes are involved in vegetative growth, and the dcr2 and rdr3 genes control reproductive development in Trichoderma atroviride (Carreras-Villaseñor et al. 2013). These data indicated that the Qip gene, together with other RNA-silencing genes, regulates the process of normal hyphal growth and development.

3.4.
Qip is strictly required for the pathogenicity of F. oxysporum in cabbage To assess the involvement of Qip in F. oxysporum's pathogenicity in cabbage, root infection assays were performed. The Qip knockout mutant had a reduced pathogenicity phenotype, significantly different from that of the wild-type A8 in disease-causing ability ( Figure 5). The result confirmed that growth impairment appeared to be significant in pathogenicity. However, we have not clearly shown that the reduced pathogenicity is due to the ΔQip mutation, its reduced parasitic ability or the decrease in pathogenic metabolites. Previous studies showed that a range of pathogenic factors were required for the pathogenicity of F. oxysporum. Mitogen-activated protein kinase (MAPK) and cAMP protein kinase A (cAMP-PKA) signal transduction systems control important steps during plant infection. Some MAPK genes, such as Fmk1, Fga1 and Fgb1, of F. oxysporum play significant roles in the pathogenicity (Di Pietro et al. 2001;Jain et al. 2002Jain et al. , 2003. RNA-silencing pathways regulate the MAPK and the cAMP-PKA cascades to affect virulence in F. oxysporum at the stage of infection. Since F. oxysporum directly penetrates through the roots without appressorium (Lagopodi et al. 2002), the most probable way to gain access to and colonize its hosts is by using its cell wall-degrading enzymes (Ospina-Giraldo et al. 2003). The SNF1 gene, encoding a cell wall-degrading enzyme, was also involved in the process of pathogenicity. Thus, in the ΔQip mutant, the slow growth may be a result of the reduced production of these enzymes, which probably impedes the initial penetration into the host root system. In other cases, the absence of RNAi allows increased transposition activity, which has been suggested as a mechanism to increase virulence (Wang et al. 2010).

Conclusion
The RNA-silencing mechanism regulates eukaryotic development and organization. Though several classes of endogenous small RNAs have been identified in ΔQip mutants of N. crassa, their biological functions still remain unknown. To clarify the function of the Qip gene in F. oxysporum, we first constructed a Qip gene knockout mutant, using homologous recombination, for a deeper analysis of the transcriptome sequencing and the changes in esRNAs. Our results showed that the Qip mutant has a significantly reduced growth rate and that the hyphal morphology was smaller with more branching than that of the wild-type A8. The ΔQip mutant also demonstrated reduced virulence on cabbage plants. These results indicate a role for this component of the RNAi machinery in the control of hyphal development and pathogenicity in F. oxysporum.

Disclosure statement
No potential conflict of interest was reported by the authors.