Knockout of the lignin pathway gene BnF5H decreases the S/G lignin compositional ratio and improves Sclerotinia sclerotiorum resistance in Brassica napus

Abstract Ferulate‐5‐hydroxylase is a key enzyme involved in the conversion of the guaiacyl monolignol to the syringyl monolignol in angiosperms. The monolignol ratio has been proposed to affect biomass recalcitrance and the resistance to plant disease. Stem rot caused by the fungus Sclerotinia sclerotiorum in Brassica napus causes severe losses in its production. To date, there is no information about the effect of the lignin monomer ratio on the resistance to S. sclerotiorum in B. napus. Four dominantly expressed ferulate‐5‐hydroxylase genes were concertedly knocked out by CRISPR/Cas9 in B. napus, and three mutant lines were generated. The S/G lignin compositional ratio was decreased compared to that of the wild type based on the results of Mӓule staining and 2D‐NMR profiling in KO‐7. The resistance to S. sclerotiorum in stems and leaves increased for the three f5h mutant lines compared with WT. Furthermore, we found that the stem strength of f5h mutant lines was significantly increased. Overall, we demonstrate for the first time that decreasing the S/G ratio by knocking out of the F5H gene improves S. sclerotiorum resistance in B. napus and increases stem strength.

It is a major component of secondary cell walls and plays an important role in plant growth and development, as well as in the defence responses to various pathogens (Dixon, 2001;Dixon et al., 2002;Naoumkina et al., 2010;Zhao & Dixon, 2014). The lignin content approaches that of cellulose in vascular plants, encrusting the polysaccharides (cellulose and hemicelluloses) and perhaps connecting to hemicelluloses; in grasses, covalent bonding of arabinoxylan hemicellulosic polysaccharides to lignin mediated by ferulate is well established (Ralph, 2010). The interaction not only enhances the mechanical strength of plants and prevents cell wall collapse but also prevents toxins in the pathogen from penetrating into the host that would otherwise allow nutrients from the host to be utilized by the pathogen (Boerjan, Ralph, & Baucher, 2003;Ride, 1978;Weng & Chapple, 2010).
The complexity of lignin arises both from the relative proportion of the three major monomeric units from which it primarily derives, and the nature of the various dimeric units that are described by their characteristic inter-unit chemical bonding. The lignin monomer ratio has been identified as an important structural factor affecting biomass recalcitrance (Holwerda et al., 2019;Sakamoto et al., 2020;Yang et al., 2019). Lignin synthesized via the phenylpropanoid pathway initiating from phenylalanine and tyrosine, the three major monolignols, which differ in their methoxylation degrees: no methoxyls in p-coumaryl alcohol, one in coniferyl alcohol and two in sinapyl alcohol. These monolignols are synthesized in the cytoplasm and diffuse or are transported to the cell wall where they are polymerized into lignin, creating the phydroxyphenyl (H), guaiacyl (G) and syringyl (S) units (Boerjan et al., 2003;Ko, Ximenes, Kim, & Ladisch, 2015;Ralph et al., 2004). These lignin units are present at different levels, and their types vary substantially among different plant species (Donaldson, 2001). The syringyl (S) monolignol has a methoxylated C-5 position, whereas the guaiacyl (G) monolignol is unsubstituted at the C-5, allowing it to bond with other monomers/units through carbon-carbon bonds and therefore forming a more condensed polymer (Tobimatsu et al., 2013).
Researchers have found that lignin composition may affect disease resistance in plants. Silencing cinnamyl alcohol dehydrogenase (CAD), caffeic acid O-methyltransferase (COMT) or caffeoyl-CoA Omethyltransferase (CCoAOMT) in diploid wheat caused higher penetration of Blumeria graminis, which revealed the importance of G monolignol biosynthesis in defence against pathogen invasion (Bhuiyan, Gopalan, Yangdou, & John, 2009). Increased deposition of the more condensed G-lignin after inoculation with S. sclerotiorum may present an effective way to prevent pathogen ingress into the vascular system (Eynck, Séguin-Swartz, Clarke, & Parkin, 2012).
Compared to the WT and f5h mutant, the F5H-overexpressing transgenic line significantly reduced the stiffness of the cell walls in the region of the compound middle lamella as determined by contact resonance force microscopy (Ciesielski, 2014). Overexpression of F5H also resulted in a significant increase in the S/G lignin ratio in various plant species, such as poplar, alfalfa and tobacco (Franke et al., 2000;Marita, Ralph, Hatfield, & Chapple, 1999;Reddy et al., 2005;Shuai et al., 2016).
To date, there are no reports on the effect of the lignin monomer ratio on resistance to S. sclerotiorum in B. napus or other Brassica species. Our strategy was therefore to use genetic engineering to target lignin composition and examine the development of resistance to recalcitrant diseases, for example the white mould caused by the fungal pathogen S. sclerotiorum. In this study, to probe the effect of lignin composition and resistance to S. sclerotiorum in oilseed rape, BnF5H was knocked out by CRISPR/Cas9 in B. napus and inoculated with S. sclerotiorum, after which the lignin compositions in the plants were profiled. Notably, our data demonstrated that decreases in the S/G lignin compositional ratio can improve the resistance to S. sclerotiorum in oilseed rape.
2 | RESULTS 2.1 | Genome-wide analysis of the BnF5H gene family in B. napus The BnF5H gene family was clustered into three clades, designated Type I, Type II and Type III. BnF5H-1, BnF5H-4, BnF5H-6 and BnF5H-7 belong to Type I; BnF5H-2 and BnF5H-5 belong to Type II; and BnF5H-3 and BnF5H-8 belong to Type III (Figure 1a). BnF5H grouping into the same branch may have similar protein affinities.
All of the Group I members, BnF5H-1, BnF5H-4, and BnF5H-6, are composed of four exons and three introns, and BnF5H-7 consists of three exons and two introns. Group II members, BnF5H-2 and BnF5H-5, are composed of three exons and two introns, and their genetic structures are similar. In Group III, BnF5H-3 is composed of four exons and three introns, and BnF5H-8 is composed of three exons and two introns. Although there are differences in the number of exons and introns, the general structures are similar ( Figure 1b). These results indicated that members within a single branch had highly similar gene structures, which produced proteins with highly similar functions.
F I G U R E 1 Genome-wide analysis of the BnF5H gene family in B. napus. (a) Phylogenetic analysis of BnF5H proteins from A. thaliana, B. rape, B. oleracea, and B. napus. The protein sequences of eight BnF5Hs with one AtF5H, two BoF5Hs, and two BrF5Hs were used to construct the NJ tree with 1,000 bootstraps, designated Group I, Group II and Group III. (b) The exon-intron structure of the BnF5H genes according to their phylogenetic relationship. (c) The 1,500 bp sequence upstream of the transcriptional start codons was used to analyse cis-regulatory elements using the PLACE database. (d) The conserved motifs of the BnF5H proteins presented according to their phylogenetic relationships. These motifs were identified by MEME, and boxes of different colours represent different motifs [Colour figure can be viewed at wileyonlinelibrary.com] Cis-elements, binding sites for transcription factors, are particularly important in the regulation of gene expression (Liu et al., 2014). To probe the function and potential regulatory mechanisms of BnF5H, the 1,500 bp sequence upstream of the transcriptional start codons was used to analyse cis-regulatory elements using the PLACE database. A total of 10 components are potentially responsive to stress and hormones, including LTR (temperature-responsive element), MBS (involved in drought inducibility), defence and stressresponsive elements (TC-rich repeats), HSE (heat stress), GARE motif (gibberellin response element), TCA element (responsive to salicylic acid), CGTCA motif (MeJA response), ABRE (ABA response element), TGA element (responding to auxin) and ERE (ethylene response) (Figure 1c). These cis-elements suggested that BnF5H is likely to be involved in heat stress, salicylic acid and defence to stress responses.

| Expression profiles of BnF5H in B. napus
The expression profiles of BnF5H in various tissues showed that BnF5H-1, BnF5H-4, BnF5H-6, and BnF5H-7 are the dominant genes expressed in the BnF5H gene family. They are expressed in roots, stems, leaves and pods and highly expressed in hypocotyls, cotyledons, buds, flowers and seeds ( Figure 2 regulating the B. napus lignin processes, and their similar expression was coinciding with their cluster and gene structure. In contrast, the expression of the other four BnF5Hs was extremely low in the tested tissues ( Figure 2).

| Subcellular localization
The subcellular localization prediction of BnF5H family member proteins suggested that BnF5H might be a membrane-localized protein.
To confirm the online prediction, the subcellular localization of BnF5H was carried out by transient expression in tobacco epidermal cells.
Yellow fluorescence was exclusively observed in the plasma membrane of tobacco epidermal cells when the recombinant construct was used, indicating that BnF5H-1 was specifically localized to the plasma membrane ( Figure 3).
To screen the targeted mutagenesis of BnF5H, genomic DNA from 10 T 0 plants harbouring the Cas9-sgRNA construct was extracted for PCR amplification and sequencing. The PCR-amplified products were detected by sequencing analysis of five randomly selected clones from individual transgenic plants. The results revealed that there were insertions (+) or deletions (À) at the desired target sites caused by the CRISPR/Cas9 system, introducing InDels into the BnF5H gene via the nonhomologous end-joining (NHEJ) repair pathway. Among them, the sequence result of transgenic plants indicated that the desired BnF5H-1, BnF5H-4, BnF5H-6, and BnF5H-7 were simultaneously mutated at the target sites or nearby, which included base mutations, deletions or insertions, and we named those lines by KO-7, KO-8 and KO-10 (Figures 4, S1, and S2). To continue screening for inherited targeted mutagenesis in this progenies, 17 T 1 plants from KO-7, KO-8, and KO-10 were randomly selected for sequencing, and all the plants showed the same mutagenesis. The T 2 plants were analysed for the same mutations by sequencing before the lignin chemical composition and inoculation tests ( Figures S3 and S4).
The results indicated that Cas9-sgRNA successfully generated three f5h mutants in the BnF5H gene, which can be inherited in B. napus offspring. S. sclerotiorum in oilseed rape (Wang et al., 2014). In our previous study, a total of 520 lines from different B. napus sources were used to identify resistance to S. sclerotiorum, and the relative susceptibility showed a significant correlation with the S/G ratio (Wei et al., 2017).

| Histochemical studies
We were therefore curious about the effects of lignin monomer composition on the resistance of S. sclerotiorum. Chinese cabbage (Zhang, Yang, & Ma, 2007). Similarly, the defence response in eucalyptus induced the accumulation of G-lignin (Hawkins & Boudet, 2003). The ratio of G/S units increased after inoculation in both resistant and susceptible inoculated tomato plants lines (Eynck et al., 2012). Resistant cotton accumulated more G than Slignin and the G/S increased, but the susceptible materials behaved contrarily upon V. dahliae inoculation (Xu et al., 2011). In tomato, the gall treated by benzothiadiazole (priming agent in plant defence) can increase the lignin pathway gene expression and more G accumulates than S monomer (Veronico et al., 2018). The controversial reports could imply that disease resistance is a typical quantitative trait, and plants with a high G monomer content may not always show more resistance in comparison because of the genetic background difference. Also to be considered, a disease response will enhance G monomer levels upon the infection to counter against the pathogen invasion. In our study, the three f5h mutants with a decreased S/G ratio exhibited increased resistance to S. sclerotiorum in both stems and leaves compared with WT B. napus (Figure 8 and Table S3). Even though the leaf contains less lignin compared with stem, the lignin monomer modification still positively affects Sclerotinia resistance.  Shi et al., 2016), poplar (Huntley, Ellis, Gilbert, Chapple, & Mansfield, 2003;Mansfield, Kang, & Chapple, 2012;Studer et al., 2011), and other hardwoods (Santos, Lee, Jameel, Chang, & Lucia, 2012). One exception reported that there was no significant correlation between saccharification efficiency and 5-hydroxyguaiacyl lignin levels and S/G ratio (Wu et al., 2019). In our work, the genetic manipulation of the monomeric composition of the lignin polymer in B. napus may alter the recalcitrance of lignocellulose during enzymatic digestion, as we found the stem structure was more tightly packed in the f5h mutants, and therefore, saccharification efficiency needs further investigation in the future. Our main aim, however, is to improve the agronomic characteristics of B. napus.
Lodging is a significant problem in crop production because it causes yield loss, poor grain filling, and impedes mechanical harvesting (Berry, Sterling, Spink, Baker, & Ennos, 2004). There are two types of lodging in oilseed rape, root lodging due to root anchorage system failure and the stem lodging as the stem breaks down or buckles. Stem strength is determined by its chemical and biochemical components and their physical structure. The Arabidopsis irx4 mutant, defective in a cinnamoyl-CoA reductase, failed to grow upright because of the reduced lignin content (Jones, Ennos, & Turner, 2001). In Arabidopsis, the dominant repression of lignin pathway transcription factors MYB58 and MYB63 produces the pendent stem phenotype (Zhou, Lee, Zhong, & Ye, 2009). Overexpression of the wheat COMT gene leads to a higher lignin content, improved stem strength, and a lower lodging index (Ma, 2009). The nitrogen and density application can change the lignin composition and results showed that syringyl (S) monomers were the predominant lignin monomeric units responsible for enhancing mechanical strength in wheat (Luo, Ni, Pang, Jin, & Wang, 2019). In this study, the f5h mutants provided us with direct evidence for the stem strength dependence on lignin composition, which might be a significant finding to mitigate lodging problems in oilseed rape.
In conclusion, resistance to S. sclerotiorum correlated with lignin monomer composition in B. napus f5h mutants in which G-units were produced and incorporated into the lignin in higher proportions than S-units. Our findings provide new insights into the disease resistance mechanism of oilseed rape on the basis of lignin chemical composition.

| Plant materials and transformation
The oilseed rape line "Westar" was used as the transformation receptor, and genetic transformation was performed using Agrobacteriummediated methods described previously (Cardoza & Stewart, 2003), followed by hygromycin (25 mg/L) or kanamycin (50 mg/L) selection on Murashige-Skoog (MS) medium. Transgenic lines were verified for each construct by polymerase chain reaction (PCR) with gene-specific primers. The knockout line in the T 2 generation and wild-type plants were grown in an isolated field and managed as usual.

| Analysis of gene expression profiles
To characterize the temporal and spatial expression patterns of BnF5H, we analysed 63 different tissues, which included roots, stems, leaves, flowers, siliques, and seeds from the B. napus cultivar ZS11 at different developmental stages (germination, seedling, bud, initial flowering, and full-bloom stages), using RNA-seq datasets. These transcriptome sequencing datasets were conserved in BioProject ID PRJNA358784. We quantified the gene expression levels according to their fragments per kilobase of exon per million reads mapped (FPKM) values using Cufflinks with default parameters, and a heatmap was drawn using the R package.

| Genome-wide analysis
To elucidate the evolutionary relationships of the BnF5H gene family, the BnF5H protein sequences from A. thaliana, B. napus, B. rapa and B. oleracea were retrieved. Phylogenetic trees were constructed using MEGA 5.1. The BnF5H gene exon-introns were analysed by using the GSDS website to study the BnF5H gene structures. To identify protein motifs, we analysed the full-length protein sequences of BnF5H using MEME software. The 1,500 bp sequence upstream of the transcriptional start codons was used to analyse cis-regulatory elements using the PLACE database.

| Generation of the f5h mutant by CRISPR/Cas9
The full-length DNA sequence of the BnF5H gene family was screened using the Genoscope Brassica napus Genome Browser (http://www.genoscope.cns.fr/brassicanapus/). Cotargeting primer gRNAs (Table S1) of the dominantly expressed BnF5H-1, BnF5H-4, BnF5H-6, and BnF5H-7 were designed using the CRISPR-P 2.0 website (http://crispr.hzau.edu.cn/CRISPR2/). One putative target site located at the second exon of the BnF5H coding sequence was selected to design the sgRNA sequences based on their GC abundance. The sgRNA sequence was 20 bp long and was adjacent to the PAM (NGG) region, which must be located in a conserved region of the four BnF5H members simultaneously and within a functionally conserved domain. To ensure the specificity of the sgRNA sequence, BLAST was performed on the online website (http://brassicadb.org/ brad/). Oligos were designed to specifically target BnF5H, and sgRNA cassettes were assembled into binary CRISPR/Cas9 vectors. Genomic DNA was isolated from the f5h mutant using a plant genomic DNA kit (TIANGEN), followed by PCR amplification using gene-specific primers (Table S1). The BnF5H-1, BnF5H-4, BnF5H-6, and BnF5H-7 PCR products were cloned into the PEASTY-T 1 vector (TransGen Biotech), and at least 20 clones for each mutant line were randomly selected for sequencing.

| Histochemical analysis
Mäule histochemical staining was used to survey and evaluate lignin monomer composition. The stems and roots of f5h mutant and WT plants were sampled at the initial flowering stage, and the silique was sampled 40 days after flowering. The cross section was stained by Mäule reagent (Chapple et al., 1992;Chen et al., 2002). The sections were treated with 1% KMnO 4 for 5 min, rinsed with water, treated with 3% HCl for 2 min, rinsed again with water, and mounted in concentrated 29% NH 4 OH for examination on an OLYMPUS MUX10 microscope.

| Determination of lignin content
The intact stem parts of the WT and KO-7 plants were used to determine lignin content after inoculation, as previously described (Mansfield et al., 2012). The ground samples were solvent-extracted with sequential extractions of water (3 Â 45 ml), 80% ethanol (3 Â 45 ml) and acetone (1 Â 45 ml). Each extraction consisted of a 20-min sonication step followed by a 20-min centrifugation step.

| Measurement of breaking force
The stem strength was measured using plant lodging tester YYD-1 (Hangzhou TOP Instrument Co., Ltd. Hangzhou, China) at mature stage according to the methods of (Luo et al., 2019) as follows. The mature stem was placed on the groove of support pillars with a distance of 10 cm. Aligning the centre position of the stem to the pressure probe. The breaking strength was the peak value when the stem was broken off and its unit was Newtons (N).

| Scanning electron microscopic (SEM) observation
The last branch on the top at mature stage from WT and KO-7 was sliced by hand with same thickness and the cross section first treated with 0.5 M NaOH for 2 hr or 1% H 2 SO 4 for 1 hr, followed by washing with distilled water until reaching pH 7.0. The cross sections were frozen in liquid nitrogen and observed by S-3400 N SEM (Hitachi) at an accelerating voltage of 2 kV.

| In vitro assays for antifungal activity
To determine the resistance of oilseed rape against S. sclerotiorum, 12 detached leaves from the WT and f5h mutant lines both at the five-leaf stage were inoculated, and lesion area was measured. Twelve stems at flowering and maturation time of the WT and the f5h mutant lines were inoculated with S. sclerotiorum. All tests were repeated three times. Both the leaf and stem were inoculated with two inocula, and the S. sclerotiorum resistance test is as follows.
The maintained S. sclerotiorum isolate from (Mei, Qian, Disi, Yang, & Qian, 2010) was cultured on potato dextrose agar (PDA) medium (20% potato, 2% dextrose and 1.5% agar) plates at 22 C in dark, and the 6-mm-diameter mycelia agar plugs punched from the growing margin of 3-day-old culture of S. sclerotiorum placed on the detached leaves or wounded stems gently. The plant S. sclerotiorum resistance was evaluated according to (Mei, Wei, Disi, Ding, & Wei, 2012) with little change. Briefly, the stems from flowering stage or mature stage were cut at the height of 10 cm from the ground, and the two ends were wrapped with polyethylene film to maintain freshness. Two separate wounds on the stem artificially made by 4-mmdiameter puncher were inoculated with the prepared inoculum. The petiole cut site of the detached leaf was wrapped with polyethylene film to keep the fresh. The inoculated leaves and stems were placed in a plastic box, which was covered with moist towels and filter paper in the bottom. To keep the moisture in the box, the top was covered and sealed by polyethylene film. The plastic box was kept in incubator at 22 C in dark. The leaf lesion area and stem lesion length were recorded 2 and 3 days after inoculation, respectively.

| STATISTICAL ANALYSIS
For multiple comparison, significance analysis was performed with one-way ANOVA followed by Tukey's post-hoc tests. Statistical analysis was performed using SPSS Statistics software (version 13.0).
Details of statistical results are in Table S3.

CONFLICTS OF INTEREST
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
The data that support the finding of this study are available in the supplementary materials of this article.