Abstract

   The widely recognized pleiotropic adult plant resistance gene Lr34
   encodes an ATP-binding cassette transporter and plays an important role
   in breeding wheat for enhanced resistance to multiple fungal diseases.
   Despite its significance, the mechanisms underlying Lr34-mediated
   pathogen defense remain largely unknown. Our study demonstrates that
   wheat lines carrying the Lr34res allele exhibit thicker cell walls and
   enhanced resistance to fungal penetration compared to those without
   Lr34res. Transcriptome and metabolite profiling revealed that the
   lignin biosynthetic pathway is suppressed in lr34 mutants, indicating a
   disruption in cell wall lignification. Additionally, we discovered that
   lr34 mutant lines are hypersensitive to sinapyl alcohol, a major
   monolignol crucial for cell wall lignification. Yeast accumulation and
   efflux assays confirmed that the LR34 protein functions as a sinapyl
   alcohol transporter. Both genetic and virus-induced gene
   silencing experiments demonstrated that the disease resistance
   conferred by Lr34 can be enhanced by incorporating the TaCOMT-3B gene,
   which is responsible for the biosynthesis of sinapyl alcohol.
   Collectively, our findings provide novel insights into the role of Lr34
   in disease resistance through mediating sinapyl alcohol transport and
   cell wall deposition, and highlight the synergistic effect of
   TaCOMT-3B and Lr34 against multiple fungal pathogens by mediating cell
   wall lignification in adult wheat plants.

   Key words: Lr34, adult plant resistance, disease resistance, lignin,
   TaCOMT-3B
     __________________________________________________________________

   Sinapyl alcohol is a key lignin precursor in plant cells. This study
   reveals the role of Lr34/Yr18/Sr57/Pm38 in facilitating sinapyl alcohol
   transport and enhancing cell wall lignification, thereby providing
   broad-spectrum resistance to multiple fungal diseases in wheat.
   Additionally, it provides evidence demonstrating that the stripe rust
   resistance conferred by Lr34/Yr18/Sr57/Pm38 can be enhanced by
   incorporating the TaCOMT-3B gene, which is responsible for the
   biosynthesis of sinapyl alcohol.

Introduction

   Approximately 19.8% of the global wheat yield is compromised annually
   due to prevalent diseases, including stripe rust (caused by Puccinia
   striiformis f. sp. tritici), leaf rust (Puccinia triticina), stem rust
   (Puccinia graminis f. sp. Tritici), and powdery mildew (Blumeria
   graminis f. sp. tritici) ([55]Savary et al., 2019). Developing and
   deploying disease-resistant wheat varieties is the most cost-effective
   and environmentally friendly approach to effectively manage crop
   diseases. Wheat disease-resistance genes can be broadly classified into
   two categories based on their expression during the host growth stage:
   all-stage resistance (ASR) and adult-plant resistance (APR)
   ([56]Gustafson and Shaner 1982; [57]Chen and Line 1993). ASR typically
   manifests as strong resistance against pathogens and offers protection
   that is limited to specific races of a pathogen species. However, ASR
   is often undermined by emerging or infrequent pathogen races that
   acquire virulence against deployed resistance (R) genes ([58]Roberts
   and Caldwell 1970). In contrast, some APR genes respond
   non-specifically to pathogen infections. Plants with these APR genes
   display reduced disease severity at the adult plant stage compared to
   susceptible controls, even if both show similar responses to the same
   pathogen at the seedling stage ([59]Ellis et al., 2014). Notably,
   certain APR genes in wheat, such as Lr34/Yr18/Sr57/Pm38 (referred to as
   Lr34), Lr46/Yr29/Sr58/Pm39 (Lr46), and Lr67/Yr46/Sr55/Pm46 (Lr67),
   provide broad-spectrum resistance against multiple diseases
   ([60]William et al., 2003; [61]Spielmeyer et al., 2005; [62]Krattinger
   et al., 2009; [63]Herrera-Foessel et al., 2011; [64]Moore et al.,
   2015).

   The APR gene Lr34, which originated from a Chinese wheat landrace, has
   been extensively used in wheat breeding for over a century
   ([65]Krattinger et al., 2009, [66]2013). Remarkably, this gene was
   present in 85% of the wheat landrace germplasm in China, 97% of which
   was cultivated in the southwest winter wheat region ([67]Yang et al.,
   2008). Lr34 encodes an ABC transporter and confers resistance to both
   biotrophic and hemi-biotrophic fungal pathogens when introduced as a
   transgene into various crops, including maize (Zea mays), rice (Oryza
   sativa), barley (Hordeum vulgare), and sorghum (Sorghum bicolor)
   ([68]Risk et al., 2013; [69]Krattinger et al., 2016;
   [70]Schnippenkoetter et al., 2017; [71]Sucher et al., 2017; [72]Boni
   et al., 2018). Notable observations in barley include increased levels
   of lignin, salicylic acid, and jasmonic acid ([73]Chauhan et al.,
   2015). Additionally, Lr34 has exhibited the ability to enhance drought
   tolerance in rice by facilitating the transport of abscisic acid and
   altering the lipid environment at the plasma membrane ([74]Deppe
   et al., 2018; [75]Krattinger et al., 2019). However, the underlying
   mechanisms by which Lr34 confers resistance to multiple pathogens in
   wheat remain largely unknown.

   ABC transporters are membrane proteins that play diverse roles
   including signal transduction, substrate transportation, protein
   secretion, pathogenesis, and host resistance to pathogens. In
   Arabidopsis thaliana, the ABC transporter AtABCG36/PEN3/PDR8 not only
   interacts with calmodulin to enhance nonhost resistance, but also
   facilitates the transport of indole-3-butyric acids, regulating auxin
   homeostasis in root tips and long-range auxin transport in roots
   ([76]Campe et al., 2016; [77]Aryal et al., 2019). AtABCG36/PEN3/PDR8
   and AtABCG40/PDR12 are involved in the secretion of the phytoalexin
   camalexin, contributing to resistance against Botrytis cinerea ([78]He
   et al., 2019). Moreover, an AtABCG14-induced immune response in
   Arabidopsis improves disease resistance by transporting cytokinin
   ([79]Wang et al., 2017). In tobacco, NpABC1 activates a jasmonic
   acid-induced defense mechanism, thereby enhancing resistance to fungal
   pathogens ([80]Stukkens et al., 2005). These findings highlight the
   diverse and critical roles of ABC transporters in plant defense
   response.

   Previous studies have identified key genes and pathways crucial for
   penetration resistance against pathogens. For example, PEN1 encodes a
   plasma membrane-resident syntaxin (t-SNARE) involved in secretory
   membrane trafficking ([81]Nielsen and Christensen 2012). In
   Arabidopsis, the loss of PEN1 resulted in nearly 90% penetration
   success by the non-adapted pathogen, B. graminis f. sp.
   hordei ([82]Collins et al., 2003). PEN2 encodes an atypical myrosinase,
   and PEN3/AtABCG36/PDR8 is a plasma membran-localized ABC transporter.
   These genes facilitate the transport of tryptophan-derived secondary
   metabolites that contribute to cell wall apposition-mediated defenses
   ([83]Stein et al., 2006; [84]Johansson et al., 2014).

   Lignin plays a key role in plant resistance, and its biosynthesis
   involves a complex network that can be divided into three processes:
   biosynthesis, transport, and polymerization of lignin monomers ([85]Liu
   et al., 2018). Within the cytoplasm, the phenylalanine metabolic
   pathway generates lignin monomers through successive steps such as
   deamination, hydroxylation, methylation, and reduction. Lignin,
   primarily composed of p-coumaryl, coniferyl, and sinapyl alcohol
   monomers, is synthesized in the cytoplasm before being transported to
   the cell wall ([86]Ralph et al., 2019; [87]Vanholme et al., 2019). Cell
   wall lignification has been proposed as a fundamental mechanism for
   protection against fungal invasion ([88]Bacete et al., 2018). The
   formation of cell wall appositions beneath pathogen penetration sites
   provides a mechanical barrier that impedes pathogen invasion
   ([89]Zhang et al., 2012). Epidermal cells strengthen the cell
   wall and secrete antifungal compounds, particularly when encountering
   non-adapted mildew pathogens specific to a crop species, thereby
   preventing pathogen invasion ([90]Hückelhoven and Panstruga 2011).
   Caffeic acid 3-O-methyltransferase (COMT) is crucial for lignin
   biosynthesis and accumulation by catalyzing multi-step methylation
   reactions of hydroxylated monomeric lignin precursors ([91]Zubieta
   et al., 2002; [92]Ma and Xu 2008; [93]Poovaiah et al., 2014).
   Downregulation of the TaCOMT gene results in decreased COMT activity
   and reduced syringyl (S) monolignol content ([94]Ma and Xu 2008;
   [95]Eudes et al., 2017). Specifically, TaCOMT-3D in wheat contributes
   to resistance against sharp eyespot (caused by Rhizoctonia cerealis) by
   promoting lignin accumulation ([96]Wang et al., 2018). Additionally,
   increased lignin content, resulting from the overexpression of other
   genes in the lignin biosynthesis pathway, such as OsPAL1, has been
   shown to confer broad-spectrum resistance to blast (caused by
   Magnaporthe grisea) and bacterial leaf blight (caused by Xanthomonas
   oryzae) in rice ([97]Zhou et al., 2018; [98]Wu et al., 2019). Moreover,
   the transcription factor OsMYB30 induces the expression of 4-coumarate
   coenzyme A ligase (4CL)-encoding genes Os4CL3 and Os4CL5, leading to
   the accumulation of lignin subunits and enhanced resistance to blast in
   rice ([99]Li et al., 2020). In wheat, the expression levels of genes
   encoding phenylalanine ammonia-lyase (PAL) and COMT are significantly
   upregulated following powdery mildew infection, and silencing these
   genes significantly reduces cell wall lignin content and increases the
   penetration rate of B. graminis ([100]Bhuiyan et al., 2009). These
   findings underscore the vital role of cell wall lignification in
   preventing pathogen infection.

   In the present study, we utilized two pairs of Lr34 mutants (lr34) and
   a pair of near-isogenic lines (NILs) to investigate the mechanism of
   Lr34-mediated disease resistance in wheat. Our results showed that
   wheat lines carrying the resistance allele, Lr34res, exhibited thicker
   cell walls and greater penetration resistance. Using in vitro yeast
   assays, we demonstrated that the Lr34res/sus proteins are capable of
   transporting sinapyl alcohol. Furthermore, we confirmed the role of
   TaCOMT-3B in lignin biosynthesis in enhancing disease resistance in
   wheat lines carrying Lr34.

Results

Lr34 prevents fungal penetration by increasing cell wall thickness

   In this study, a pair of near-isogenic lines (NILs) (Avocet+Lr34 and
   Avocet) and two independent lr34 mutant lines (designated as m19 and
   m21) were used to investigate the mechanism of broad-spectrum
   resistance associated with the Lr34 gene in wheat ([101]Krattinger
   et al., 2009). Lr34res and Lr34sus are two alleles present in natural
   wheat populations ([102]Krattinger et al., 2019). The Avocet+Lr34 and
   Avocet lines carry Lr34res and Lr34sus, respectively. The two mutant
   lines, m19 and m21, were obtained through gamma-irradiation deletion,
   resulting in premature termination of Lr34 protein translation
   ([103]Figure 1A). Their corresponding resistant siblings were
   designated as m19 r-sib and m21 r-sib, respectively.

Figure 1.

   [104]Figure 1
   [105]Open in a new tab

   Lr34 confers disease resistance by enhancing cell wall thickness.

   (A) Gene structure of Lr34. Black boxes represent exons; introns are
   denoted by adjoining lines. Mutation sites in the m19 and m21 mutants
   are marked in red. The three sequence polymorphisms between susceptible
   and resistant alleles are highlighted in blue, referencing
   [106]Krattinger et al. (2009).

   (B) Disease severities for stripe rust (mix) in m19 r-sib, m19, m21
   r-sib, m21, Avocet +Lr34, and Avocet at 20 days post inoculation at the
   booting stage in a greenhouse. Bars, 1 cm.

   (C) Expression levels of Lr34res at different plant growth stages.
   Relative expression levels were calculated using the 2^-ΔΔCT method
   with TaGAPDH as the endogenous control. Data are shown as the mean ± SD
   of three biological replicates.

   (D) Infection of the powdery mildew pathogen (Bgt isolate E21) on
   leaves of the three pairs of lines 48 h post inoculation at the booting
   stage. Inoculated leaves were subjected to aniline blue staining for
   callose. Bars, 50 μm.

   (E) Measurement of penetration rates by Bgt (E21) on leaves at the
   booting stage 48 h post inoculation. n = 10 biological replicates. Data
   are represented as mean ± SD. ∗∗P ＜ 0.01. Student’s t-tests.

   (F) Transmission electron microscopy (TEM) images of leaves from the
   three pairs of lines at the booting stages. Bars, 1 μm.

   (G) Determination of cell wall differences between sister lines. n = 20
   biological replicates. Data are represented as mean ± SD. ∗∗P ＜0.01.
   Student’s t-tests.

   Avocet+Lr34, m19 r-sib, and m21 r-sib exhibited disease resistance at
   the adult plant stage, whereas Avocet, m19, and m21 were highly
   susceptible to pathogen infection ([107]Figure 1B). Notably, all lines
   were susceptible at the seedling stage ([108]Supplemental Figure 1A).
   Consistent with its role in mature stages, qRT–PCR analysis revealed
   that Lr34res displayed the highest expression level at the booting
   stage ([109]Figure 1C). Disease assays demonstrated that only
   0.69%–0.81% of the leaf blade area showed successful penetrations in
   lines carrying Lr34res ([110]Figures 1D and 1E), whereas lines lacking
   Lr34res exhibited a significant increase in successful Bgt
   penetrations, ranging from 1.9% to 2.6% ([111]Figures 1D and 1E).
   Callose deposition at the plant cell wall is a known defense response
   to pathogen attack ([112]Jacobs et al., 2003). At 5 days after
   inoculation, the growth of Bgt on leaves of lines carrying Lr34res was
   significantly slower compared to those without the Lr34res allele
   ([113]Supplemental Figures 2A and 2B). Microscopic observations
   revealed that lines carrying the Lr34res allele had significantly
   thicker cell walls than those carrying Lr34sus and the two lr34 mutants
   (m19 and m21) at the booting stage ([114]Figures 1F and 1G). No
   differences in cell wall thickness were detected among the three pairs
   of lines at the seedling stage ([115]Supplemental Figures 1B and 1C).
   From the seedling to booting stage, cell wall thickness increased by
   approximately 30% in lines with Lr34res, whereas no increase was
   observed in lines lacking Lr34res ([116]Figures 1G and
   [117]Supplemental Figure 1C). These findings strongly support the role
   of Lr34res in substantially enhancing penetration resistance and
   preventing pathogen invasion by increasing cell wall thickness.

Cell wall metabolism is compromised in the lr34 mutant

   To investigate the underlying mechanism of Lr34-mediated disease
   resistance, we conducted transcriptome analysis at the booting stages
   of m19 r-sib and m19 lines. A total of 802 differentially expressed
   genes (DEGs) were identified in non-infected samples, with 619
   upregulated and 183 downregulated genes ([118]Supplemental Table 2).
   Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis revealed that these DEGs were highly enriched in the
   phenylalanine metabolism pathway ([119]Figure 2A). Previous studies
   have indicated that lignin, salicylic acid (SA), and jasmonic acid (JA)
   are downstream products of this pathway ([120]Chauhan et al., 2015). We
   therefore specifically examined the DEGs from defense signaling
   pathways and cell wall metabolism. Our results show that most genes
   involved in lignin biosynthesis, as well as SA and JA signaling, were
   largely repressed in the m19 lr34 mutant compared to m19 r-sib
   ([121]Figure 2B). Consistently, qRT–PCR analysis showed significant
   increases in the expression levels of genes encoding PAL, caffeic acid
   3-O-methyltransferase (COMT), 4-coumarate CoA ligase (4CL), and
   cinnamate 4-hydroxylase (C4H) in Lr34-carrying lines at the booting but
   not the seedling stage, validating the RNA sequencing (RNA-seq) results
   ([122]Figures 2C and [123]Supplemental Figure 1D). Interestingly, SA
   and JA contents were similar among the Lr34res-carrying lines (m19
   r-sib, m21 r-sib, Avocet+Lr34), the lr34 mutant lines (m19, m21) and
   Lr34sus (Avocet) ([124]Supplemental Figures 3A and 3B). However, all
   three Lr34res-carrying lines exhibited significantly higher lignin
   content compared to the corresponding mutants or the NIL parent Avocet
   at the booting stage ([125]Figure 2D), with no differences observed at
   the seedling stage ([126]Supplemental Figure 1E). Chemical staining for
   lignin further revealed a marked increase (25%–48%) of stained area in
   the leaf sections of all three Lr34res-carrying lines compared to those
   without Lr34res ([127]Figures 2E and 2F). Therefore, the cell wall
   thickening in wheat lines harboring Lr34res can be attributed to
   increased lignin content.

Figure 2.

   [128]Figure 2
   [129]Open in a new tab

   Transcriptome and metabolite analysis of Lr34 regulatory pathways.

   (A) Kyoto Encyclopedia of Genes and Genomes enrichment analysis for m19
   r-sib and m19 at the booting stage.

   (B) Heatmap of DEGs associated with JA, SA, and lignin pathways in
   non-infected and infected m19 r-sib and m19 plants at the booting
   stage.

   (C) Expression of four lignin biosynthesis genes at the booting stage.
   Relative expression levels were calculated using the 2^−ΔΔCT method
   using TaGAPDH as the endogenous control. n = 3 biological replicates.
   ∗∗P ＜ 0.01. Student’s t-tests.

   (D) Lignin contents in leaves of three pairs of lines at the booting
   stage measured using the thioglycolic acid method. n = 3 biological
   replicates. Data are shown as mean ± SD. ∗P ＜ 0.05, ∗∗P ＜ 0.01.
   Student’s t-tests.

   (E) Staining of lignin sections with phloroglucinol at the booting
   stage. Bars, 100 μm.

   (F) Quantitation of lignin in flag leaf sections at the booting stage.
   n = 10, three biological replicates. Data are shown as mean ± SD. ∗∗P ＜
   0.01. Student’s t-tests.

   (G)Lr34res-mediated lignin biosynthesis pathway. Enzymes and
   metabolites in blue are downregulated in m19; those in black showed no
   change. Metabolites in orange were not detected in m19 (pathway diagram
   adapted from [130]Barros et al., 2015).

   (H) Contents of leaf lignin monomers sinapyl, p-coumaryl, and coniferyl
   alcohol at the booting stage. n = 3 biological replicates. Data are
   shown as mean ± SD. ns, not significant. ∗P ＜ 0.05, ∗∗P ＜ 0.01.
   Student’s t-tests.

   Lignin is primarily synthesized through the polymerization of
   p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) units, which are
   derived from p-coumaryl alcohol, coniferyl alcohol, and sinapyl
   alcohol, respectively, along with other phenolic derivatives
   ([131]Bonawitz and Chapple 2010). Next, we conducted a detailed
   profiling of metabolites involved in the lignin biosynthesis pathway.
   Our results reveal lower levels of key intermediates, including
   cinnamic acid, phenylalanine, p-coumaric acid, sinapic acid,
   sinapinaldehyde, and sinapyl alcohol, in m19 compared to m19 r-sib
   ([132]Figure 2G). To gain further insights, we quantified the three
   lignin monomers, sinapyl, coniferyl, and p-coumaryl alcohol, using
   internal standards as controls across the three pairs of wheat lines.
   The content of sinapyl alcohol was significantly higher in
   Lr34res-carrying lines compared to those without Lr34res, whereas the
   levels of coniferyl alcohol and p-coumaryl alcohol remained unchanged
   ([133]Figure 2H). Fourier-transform infrared (FTIR) spectroscopy has
   been widely used to estimate the chemical composition and functional
   properties of lignin ([134]Yu et al., 2023). Characteristic FTIR
   spectral peaks were observed at 800 cm^−1 to 1800 cm^−1, 2918cm^−1, and
   3378cm^−1 ([135]Supplemental Figure 4) ([136]Yu et al., 2023).
   Consistent with the metabolite profiling, the FTIR spectrum indicate a
   lower percentage of transmittance in Lr34res-carrying lines relative to
   those without Lr34res, suggesting a higher content of lignin functional
   groups in these lines ([137]Supplemental Figure 4). Together, our
   results suggest that the loss of Lr34 compromises lignin biosynthesis,
   potentially through regulating sinapyl alcohol content in cell walls.

The lr34 mutants are hypersensitive to sinapyl alcohol

   Previous studies have demonstrated that ABC transporters are involved
   in stress tolerance by facilitating the transport of various substances
   ([138]Alejandro et al., 2012, [139]Fu et al., 2019b). To determine
   whether sinapyl alcohol could potentially serve as a substrate for
   Lr34, we employed a chemical treatment approach. The addition of 5 mM
   sinapyl alcohol to ^1/[2] Murashige and Skoog medium resulted in
   reduced growth rates of m19, m21, and Avocet compared to the
   Lr34res-carrying lines ([140]Figures 3A and 3B). By contrast, no
   differences in growth rates were observed between the paired lines when
   treated with coniferyl alcohol or p-coumaryl alcohol ([141]Supplemental
   Figures 5A and 5B). The expression of ABC transporters is often
   upregulated in response to substrate treatment ([142]Alejandro et al.,
   2012). We performed qRT–PCR analyses to measure the expression levels
   of Lr34 under treatments with 5 mM of p-coumaryl, sinapyl, or coniferyl
   alcohol. The expression of Lr34 was significantly upregulated upon the
   addition of sinapyl alcohol ([143]Figure 3C), whereas it remained
   unchanged with p-coumaryl alcohol and coniferyl alcohol
   ([144]Supplemental Figures 5C and 5D). These findings suggest that
   sinapyl alcohol may serve as a substrate for Lr34.

Figure 3.

   [145]Figure 3
   [146]Open in a new tab

   Phenotypes for the three pairs of lines differing in Lr34 when treated
   by sinapyl alcohol.

   (A) Phenotypes of seedlings grown on ½ Murashige and Skoog medium
   following treatment with sinapyl alcohol and DMSO. Bars, 1 cm.

   (B) Statistics of seedling length for the materials shown in (A). n =
   15 biological replicates. Data are shown as mean ± SD. ns, not
   significant. ∗∗P ＜ 0.01. Student’s t-tests.

   (C) Lr34res expression in seedlings carrying Lr34res at 0, 4, and 8 h
   after treatment with 5 mM sinapyl alcohol or DMSO (control). Relative
   gene expression levels were calculated using the 2^−ΔΔCT method with
   TaGAPDH as the endogenous control. ns, not significant. ∗P ＜ 0.05, ∗∗P
   ＜ 0.01. Student’s t-tests.

   (D) Disease severity of the seedlings of three pairs of lines 14 days
   after treatment with stripe rust (mix). Bars, 0.5 cm.

   (E) Fungal to wheat biomass ratios in the three pairs of lines. Ratios
   were calculated through the absolute quantification of total cDNA
   contents at 14 dpi using PstGAPDH and TaGAPDH as the internal
   reference, respectively. n = 3 biological replicates. Data are shown as
   mean ± SD. ns, not significant. ∗, P ＜ 0.05. ∗∗, P ＜ 0.01. Student’s
   t-tests.

   (F) Lr34res expression in the seedlings of the three pairs of lines 14
   dpi with stripe rust (mix). Relative expression levels were calculated
   using the 2^−ΔΔCT method with TaGAPDH as the endogenous control. n = 3
   biological replicates. Data are represented as mean ± SD. ∗P ＜ 0.05,
   ∗∗P ＜ 0.01. Student’s t-tests.

   Given that the expression of Lr34 could be induced by sinapyl alcohol,
   we applied sinapyl alcohol exogenously to the leaves of wheat lines
   inoculated with a mixture of Chinese Pst races CYR32, CYR33, and CYR34
   at the seedling stage. Disease assays showed that the lines carrying
   Lr34res exhibited approximately 30% less pathogen accumulation and
   enhanced resistance compared to the control (DMSO-treated). By
   contrast, no significant differences were observed between sinapyl
   alcohol- and DMSO-treated lr34 mutant lines and Avocet ([147]Figures 3D
   and 3E). To investigate whether sinapyl alcohol inhibits fungal growth,
   we conducted growth experiments with Fusarium graminearum (PH-1) in
   potato dextrose agar (PDA) medium. The results indicated a slight
   inhibitory effect of sinapyl alcohol on PH-1 compared to the control
   ([148]Supplemental Figure 6).

   Furthermore, the transcript level of Lr34res was significantly induced
   by sinapyl alcohol treatment at the seedling stage ([149]Figure 3F). We
   also determined the transcript levels of four key genes in the lignin
   biosynthesis pathway—PAL, COMT, cinnamate 4-hydroxylase, and 4CL. All
   four genes were significantly upregulated in Lr34res-carrying lines
   compared to those without Lr34res after sinapyl alcohol treatment,
   whereas no changes were observed under the control condition (DMSO
   treatment) ([150]Supplemental Figure 7). These results
   provide additional evidence supporting the association between
   Lr34-mediated disease resistance and sinapyl alcohol.

Lr34 functions as a sinapyl alcohol transporter

   The lignin biosynthesis pathway has been well established over the past
   decade ([151]Dixon and Barros 2019), with only one characterized lignin
   transporter in Arabidopsis shown to transport p-coumaryl alcohol and
   regulate cell wall thickness ([152]Alejandro et al., 2012). To explore
   whether the Lr34-encoded ABC transporter also transport sinapyl
   alcohol, we initiated our analysis by examining the protein domains of
   LR34. The LR34 protein features two highly conserved nucleotide-binding
   and transmembrane domains ([153]Supplemental Figure 8). Its protein
   structure was predicted using AlphaFold 2 ([154]Supplemental Figure 9).
   To assess the potential interaction between LR34 and sinapyl alcohol,
   molecular docking modeling was performed using Autodock. The results
   revealed that Lr34res has two potential sinapyl alcohol-interacting
   residues, Thr 368 and Ser 1013, within its nucleotide-binding domains,
   with a binding energy of −3.93 kcal/mol ([155]Supplemental Figure 9A).
   Interestingly, molecular docking for Lr34sus identified five potential
   sinapyl alcohol-interacting residues—Asn 335, Leu 337, Leu 365, Thr
   981, and His 1011, also located within the nucleotide-binding domains,
   with a binding energy of −4.44 kcal/mol ([156]Supplemental Figure 9B).

   To experimentally validate whether the LR34 protein could transport
   sinapyl alcohol, we conducted chemical sensitivity assays and efflux
   analyses using yeast mutants of ABC transporters ([157]Alejandro
   et al., 2012). We successfully knocked out eight genes encoding ABC
   transporters in the W303-1a yeast strain, including PDR5, PDR10, PDR11,
   PDR12, PDR15, PDR18, YOR1, and SNQ2. Chemical sensitivity tests were
   performed on each of the eight single-mutant strains in the presence of
   sinapyl alcohol, revealing that only the growth of W303-1a-PDR11Δ and
   W303-1a-PDR15Δ was severely inhibited, indicating sensitivity to
   sinapyl alcohol ([158]Figure 4A). Subsequently, we generated a
   double-mutant strain, W303-1a-PDR11ΔPDR15Δ, which exhibited more
   pronounced growth inhibition in the presence of sinapyl alcohol after
   24 h ([159]Figure 4A). We also tested the inhibitory effects of
   coniferyl alcohol and p-coumaryl alcohol. Our results showed that the
   growth of all three yeast mutants, W303-1a-PDR12Δ, W303-1a-PDR18Δ, and
   W303-1a-PDR5Δ, was inhibited by coniferyl alcohol ([160]Supplemental
   Figure 10A), whereas only W303-1a-YOR1Δ growth was inhibited by
   p-coumaryl alcohol ([161]Supplemental Figure 10B). The growth rate of
   the W303-1a-PDR11Δ PDR15Δ double mutant was decreased to 50% at 16
   h but recovered to 90% at 32 h in the presence of coniferyl alcohol
   ([162]Supplemental Figure 10A). Next, we individually introduced the
   coding sequences of Lr34res, Lr34sus, m19, and m21 into the yeast
   W303-1a-PDR11ΔPDR15Δ double-mutant. Western blot analyses confirmed the
   presence of Lr34res and Lr34sus with an Lr34-specific antibody, whereas
   no signals were detected in yeast strains expressing m19 and m21,
   despite successful RNA transcription for all four sequences
   ([163]Figures 4B and 4C). In the absence of sinapyl alcohol, growth of
   the W303-1a-PDR11ΔPDR15Δ yeast strains was unaffected by the
   expression of Lr34res, Lr34sus, m19, m21, or the empty-vector control
   ([164]Figure 4D). However, in the presence of sinapyl alcohol, the
   growth of transgenic yeast strains expressing m19, m21, and the empty
   vector was significantly inhibited ([165]Figure 4E). Surprisingly, both
   Lr34res and Lr34sus were able to overcome the growth inhibition in the
   yeast double mutant when supplemented with sinapyl alcohol
   ([166]Figure 4E). We hypothesized that the full-length proteins
   generated by Lr34res and Lr34sus were stable in yeast cells, whereas
   the truncated m19 and m21 proteins were likely subject to degradation.
   Time-dependent loading and unloading assays confirmed that yeast cells
   expressing Lr34res or Lr34sus accumulated less sinapyl alcohol compared
   to those expressing m19, m21, or the empty vector ([167]Figure 4F).
   Efflux analysis on preloaded yeasts demonstrated that cells expressing
   Lr34res and Lr34sus efficiently released sinapyl alcohol, whereas the
   release was significantly slower from the W303-1a-PDR11ΔPDR15Δ cells
   expressing m19, m21, or the empty vector ([168]Figure 4G). Results
   of the efflux assay was further validated by expressing AtABCG29, which
   encodes a p-coumaryl alcohol transporter in Arabidopsis ([169]Alejandro
   et al., 2012), in the yeast W303-1a-YOR1Δ mutant ([170]Supplemental
   Figure 10C). Lr34-4A (TraesCS4A01G384800) and Lr34-7A
   (TraesCS7A01G085800), homologous genes of Lr34 ([171]Krattinger et al.,
   2011), were also tested. RT–PCR detected no Lr34-7A transcript,
   suggesting that Lr34-7A is a pseudogene; it was thus excluded from
   further analysis. The yeast assays revealed that neither Lr34-4A nor
   AtABCG29 could transport sinapyl alcohol ([172]Supplemental
   Figure 10D). Therefore, our results indicate that sinapyl alcohol is
   the primary substrate for LR34 in yeast, and both Lr34res and Lr34sus
   are capable of transporting sinapyl alcohol.

Figure 4.

   [173]Figure 4
   [174]Open in a new tab

   Lr34 mediates sinapyl alcohol transport to reduce toxicity sensitivity
   in yeast.

   (A) Growth rate ratios of wild-type yeast (W303-1a), eight independent
   single mutants (W303-1a-SNQ2Δ, W303-1a-PDR18Δ, W303-1a-PDR12Δ,
   W303-1a-PDR10Δ, W303-1a-PDR5Δ, W303-1a-YOR1Δ, W303-1a-PDR11Δ,
   W303-1a-PDR15Δ), and a double-mutant (W303-1a-PDR11ΔPDR15Δ) treated
   with sinapyl alcohol. Data are shown as mean ± SD; n = 3 technical
   replicates.

   (B) Determination of the sizes of yeast cells carrying Lr34res,
   Lr34sus, m19, and m21 by RT–PCR. ScGAPDH was used as the internal
   control. The 5000-bp DNA Ladder (100-5000 bp, Vazyme) was used.

   (C) Western blot analysis of yeast extracts. ACTB (Sangon Biotech) was
   used as the internal control. The 200-kDa Plus Prestained Protein
   Ladder (10–200 kDa, GenStar) was used. The predicted molecular weight
   of Lr34 is about 160 kDa.

   (D and E) Growth of the W303-1a-PDR11ΔPDR15Δ double mutant
   overexpressing Lr34res, Lr34sus, m19, m21 or the empty vector pDR196
   under treatment of SD-Ura (D) and SD-Ura + 3 mM sinapyl alcohol (E),
   respectively. n = 3 biological replicates. Data are represented as
   mean ± SD.

   (F) Accumulation of sinapyl alcohol in the double mutant. Yeast cells
   of the double mutant overexpressing Lr34res, Lr34sus, m19, and m21 or
   the empty vector pDR196 were grown under SD-Ura and then suspended to
   the same optical density when 3 mM sinapyl alcohol was added at time 0.
   n = 3 biological replicates. Data are shown as mean ± SD.

   (G) Efflux rates of sinapyl alcohol in yeast mutant strains. Growth
   curves for each time point were derived from the mean values of three
   independent cultures. n = 3 biological replicates. Data are shown as
   mean ± SD.

   (H) Sinapyl alcohol-d3 content in m19 r-sib and m19 seedlings after the
   treatment. n = 3 biological replicates. Data are represented as mean ±
   SD. ∗∗P ＜ 0.01. Student’s t-tests.

   (I) Determination of Lr34res, Lr34sus, m19, and m21 sizes in wheat flag
   leaves using RT–PCR. TaGAPDH was used as the internal control. The
   5000-bp DNA Ladder (100–5000 bp, Vazyme) was used.

   (J) Western blot analysis of wheat flag leaf extracts. Actin (Sangon
   Biotech) was used as the internal control. The 200-kDa Plus Prestained
   Protein Ladder (10–200 kDa, GenStar) was used. The predicted molecular
   weight of Lr34res is about 160 kDa.

   To verify the transport of sinapyl alcohol by Lr34 in plants, we
   obtained the deuterated labeled sinapyl alcohol (sinapyl alcohol-d3)
   and conducted a chemical treatment on wheat seedlings. Metabolite
   profiling revealed a lower sinapyl alcohol-d3 content in m19 r-sib
   compared with m19 leaves ([175]Figure 4H), indicating an accumulation
   of sinapyl alcohol in the mutants. This suggests impaired Lr34-mediated
   sinapyl alcohol transport in the m19 mutant due to the absence of the
   Lr34 protein. To explore the differences in the transport capabilities
   of sinapyl alcohol between Lr34res and Lr34sus in the yeast system and
   the disease resistance observed only with Lr34res in wheat, we examined
   the mRNA and protein levels of Lr34 in various wheat lines, including
   m19 r-sib, m21 r-sib, Avocet+Lr34, Avocet, m19, and m21
   ([176]Figures 4I and 4J). Surprisingly, unlike in yeast cells where
   both Lr34res and LR34sus proteins were detected, protein signals were
   only observed in m19 r-sib, m21 r-sib, and Avocet+Lr34
   ([177]Figure 4J). As a result, Avocet, which carries the Lr34sus
   allele, was unable to produce functional protein in plants, leading to
   increased susceptibility to diseases. Together, these findings
   highlight the crucial role of Lr34 protein functionality in mediating
   disease resistance in wheat.

TaCOMT-3B mediates sinapyl alcohol biosynthesis and contributes to disease
resistance

   COMT is a key enzyme in the biosynthesis of sinapyl alcohol
   ([178]Fornale et al., 2017) ([179]Figure 5A). Previous studies have
   demonstrated that TaCOMT-3B (TraesCS3B02G612000) is involved in lignin
   biosynthesis in wheat stalks ([180]Fu et al., 2019a). To confirm the
   biochemical function of TaCOMT-3B, we performed an in vitro enzymatic
   assay using 5-hydroxyferulic acid as the substrate. The assay revealed
   that sinapic acid was produced when the purified TaCOMT-3B protein was
   incubated with 5-hydroxyferulic acid, confirming TaCOMT-3B enzymatic
   activity in vitro ([181]Figure 5B). We then tested in vivo enzyme
   activity using the foxtail mosaic virus (FoMV) system. TaCOMT-3B was
   transiently expressed in the leaves of the wheat variety Aikang 58 via
   FoMV, and lignin monomers were quantified. We observed a significant
   increase in sinapyl alcohol content in treated wheat leaves compared to
   the control, whereas coniferyl and p-coumaryl alcohol contents were
   unaffected ([182]Figure 5C). These results provide solid evidence that
   TaCOMT-3B specifically catalyzes the biosynthesis of sinapyl alcohol.

Figure 5.

   [183]Figure 5
   [184]Open in a new tab

   Caffeic acid 3-O-methyltransferase (TaCOMT-3B) mediates Lr34 resistance
   by regulating sinapyl alcohol content.

   (A) Schematic showing the position of caffeic acid
   3-O-methyltransferase (COMT) in the lignin biosynthetic pathway
   (pathway diagram adapted from [185]Barros et al., 2015).

   (B) High-performance liquid chromatography (HPLC) chromatograms of the
   reaction products of recombinant TaCOMT-3B with 5-hydroxyferulic acid
   as the substrate.

   (C) Contents of sinapyl alcohol, coniferyl alcohol, and p-coumaryl
   alcohol in Aikang 58 determined by transient expression of the foxtail
   mosaic virus (FoMV). n = 3 biological replicates. Data are shown as
   mean ± SD. ns, not significant. ∗P ＜ 0.05. Student’s t-tests.

   (D) Relative expression levels of TaCOMT-3B in control and TaCOMT-3B
   knockdown plants. Relative expression levels were calculated using the
   2^−ΔΔCT method with TaGAPDH as the endogenous standard. n = 3
   biological replicates, Data are shown as mean ± SD. ∗P ＜ 0.05.
   Student’s t-tests.

   (E) Relative quantity of sinapyl alcohol in flag leaves at the booting
   stage after silencing via VIGS. n = 3 biological replicates. Data are
   shown as mean ± SD. ∗P ＜ 0.05. Student’s t-tests.

   (F) Lignin contents in flag leaves at the booting stage after silencing
   via VIGS. n = 3 biological replicates. Data are shown as mean ± SD. ns,
   not significant. ∗P ＜ 0.05. Student’s t-tests.

   (G) Response of m19 r-sib and m19 to powdery mildew (E21) at 10 days
   following VIGS treatment at the booting stage. Bars, 1 cm.

   (H) Fungal to wheat biomass ratios in m19 r-sib and m19 silenced
   plants. Ratios were calculated by absolute quantification of total cDNA
   at 10 dpi, using PgtGAPDH and TaGAPDH as the internal reference,
   respectively. n = 3 biological replicates. Data are represented as
   mean ± SD. ns, not significant. ∗P ＜ 0.05. Student’s t-tests.

   (I) Microscopic observation of powdery mildew (E21) growth after VIGS
   treatment at the booting stage. Bars, 25 μm.

   To further investigate the role of TaCOMT-3B in wheat, we employed
   virus-induced gene silencing (VIGS) to silence TaCOMT-3B in wheat
   leaves. Two vectors, one empty (barley stripe mosaic virus [BSMV]:00)
   and another targeting the phytoene desaturase (PDS) gene (BSMV: PDS),
   were used as non-target controls. The qRT–PCR analysis indicated a
   reduction of about 70% in TaCOMT-3B transcript levels in m19 r-sib and
   m19 ([186]Figure 5D). This silencing resulted in approximately a 35%
   decrease in sinapyl alcohol content in lines treated with BSMV
   targeting TaCOMT-3B compared to the controls ([187]Figure 5E).
   Additionally, lignin content was significantly reduced in m19 r-sib
   ([188]Figure 5F). Inoculation of flag leaves with the powdery mildew
   pathogen demonstrated that the silencing of TaCOMT-3B in m19 r-sib led
   to enhanced development of powdery mildew and increased pathogen
   biomass. However, there was no noticeable change in disease severity in
   m19 before or after VIGS treatment at 10 days post infection
   ([189]Figures 5G–5I). Taken together, these results indicate that
   TaCOMT-3B is involved in sinapyl alcohol biosynthesis and contributes
   to disease resistance mediated by cell wall lignification.

Lr34 and TaCOMT-3B synergistically enhance stripe rust resistance

   Given that both Lr34 and TaCOMT-3B influence sinapyl alcohol content in
   plant cell walls, we investigated the genetic interactions between
   Lr34res and TaCOMT-3B in conferring disease resistance. A natural
   population of 312 wheat lines, along with a bi-parental population
   derived from a cross between m21 r-sib and SY95-71, was genotyped using
   gene-specific molecular markers for TaCOMT-3B and Lr34. The m21 r-sib
   parental line carried the Lr34res and TaCOMT-3Ba alleles, whereas
   SY95-71, which carried the Lr34sus and TaCOMT-3Bb alleles, was used for
   genetic analysis. Therefore, the bi-parental population is suitable for
   analyzing the effects of interactions between Lr34 and TaCOMT-3B on
   phenotypes. The stripe rust responses of these lines were assessed
   across six locations over 3 years ([190]Figure 6A). Among the 312
   materials, Lr34 was divided into two genotypes, Lr34res and Lr34sus,
   whereas TaCOMT-3B was divided into two genotypes, TaCOMT-3Ba and
   TaCOMT-3Bb. TaCOMT-3Bb differs from TaCOMT-3Ba by the absence of a
   222-bp segment in the 3′ untranslated region (UTR) ([191]Fu et al.,
   2019a). Based on the genotyping results, the lines were categorized
   into four genotypic combinations ([192]Figure 6B). The results showed
   that lines carrying TaCOMT-3Ba had significantly higher sinapyl alcohol
   content compared to those with the TaCOMT-3Bb allele ([193]Figure 6B).
   Lignin content was notably higher in lines carrying both TaCOMT-3Ba and
   Lr34res, compared to lines carrying either Lr34res or TaCOMT-3Ba alone
   ([194]Figure 6C). Similarly, disease severity was significantly lower
   in lines carrying both Lr34res and TaCOMT-3Ba compared to lines with
   only Lr34res. Conversely, no significant difference in disease response
   was observed in lines carrying only TaCOMT-3Ba or those lacking both
   TaCOMT-3Ba and Lr34res ([195]Figure 6D). Consistently, lines containing
   both Lr34res and TaCOMT-3Ba in the bi-parental population also
   exhibited significantly lower disease severities compared to those with
   each gene alone ([196]Figure 6E). These results suggest that TaCOMT-3Ba
   likely enhances sinapyl alcohol production, thereby aiding the Lr34res
   allele in conferring greater resistance in wheat.

Figure 6.

   [197]Figure 6
   [198]Open in a new tab

   Effects of the interaction between Lr34 and TaCOMT-3B on stripe rust
   response in adult plants assessed using a panel of 312 wheat accessions
   and a bi-parental population.

   (A) Frequency distribution of mean final stripe rust severity across
   six environments over three years in the 312 accession panel and one
   year in the F[4] bi-parental population.

   (B) Relative amounts of sinapyl alcohol in lines with four different
   genotypes.

   (C) Lignin contents in lines corresponding to each genotype combination
   from (B).

   (D) Mean stripe rust severities observed in six environments.

   (E) Stripe rust severities in the same four genotypes from the
   bi-parental population. The number of lines for each gene combination
   is denoted by “n” in (B)–(E). The disease and chemical analysis
   included all lines in the study. Disease evaluation was performed on
   all lines used for sinapyl alcohol and lignin analysis. Different
   letters in (B)–(E) indicate significant differences at P ＜ 0.05
   determined by the Student’s t-test followed by ANOVA.

Discussion

   APR genes have been widely incorporated into breeding programs aimed at
   enhancing crop yield stability by minimizing losses due to pathogens.
   Examples include Lr34 and Lr67 in wheat ([199]Krattinger et al., 2009;
   [200]Moore et al., 2015; [201]Milne et al., 2023), Xa7 in rice
   ([202]Chen et al., 2021), and Rppk in maize ([203]Chen et al., 2022).
   Lr34, the first APR gene cloned in wheat ([204]Krattinger et al.,
   2009), has been shown to provide enhanced resistance against multiple
   pathogens in various plant species ([205]Risk et al., 2013;
   [206]Krattinger et al., 2016; [207]Schnippenkoetter et al., 2017;
   [208]Sucher et al., 2017; [209]Boni et al., 2018). Although it has been
   suggested that LR34 contributes to drought tolerance in rice by
   transporting abscisic acid ([210]Krattinger et al., 2019;
   [211]Braunlich et al., 2021), the biological mechanisms underlying its
   durable and broad-spectrum disease resistance remain largely unknown.
   In the present study, we found that lines with the Lr34res allele
   displayed a reduced penetration rate, higher lignin content, and a
   thicker cell wall compared to those without this allele. We provide
   multiple lines of evidence suggesting that Lr34 encodes a sinapyl
   alcohol transporter, which is involved in plant cell wall
   lignification. Furthermore, we revealed that combining Lr34 (involved
   in sinapyl alcohol transport) and TaCOMT-3B (responsible for sinapyl
   alcohol biosynthesis) enhances disease resistance in wheat. Our
   proposed model shows that Lr34 mediates disease resistance by
   facilitating the transport of sinapyl alcohol, thus enhancing cell wall
   lignification ([212]Figure 7).

Figure 7.

   [213]Figure 7
   [214]Open in a new tab

   Molecular model illustrating how Lr34 confers resistance to multiple
   wheat diseases by facilitating sinapyl alcohol transport and increasing
   cell wall thickness.

   Lr34res-carrying lines are capable of transporting sinapyl alcohol from
   intracellular spaces to the extracellular environment, enhancing lignin
   content and cell wall thickness, thereby conferring resistance.
   Conversely, lines carrying the susceptible Lr34sus allele lack the
   ability to transport sinapyl alcohol, primarily due to pre-termination
   of the translation of the Lr34 protein in wheat. Additionally,
   TaCOMT-3Ba can produce increased levels of sinapyl alcohol, which works
   in conjunction with Lr34res to enhance resistance (this model diagram
   is adapted from [215]Barros et al., 2015).

   Multi-omics approaches have been pivotal in unraveling the mechanisms
   of disease resistance in plants. For instance, the integration of
   metabolomic and transcriptomic data revealed that disease resistance in
   maize is mediated through the activation of defense-related and cell
   death biosynthesis signaling pathways ([216]Li et al., 2022). In this
   study, combined transcriptome and metabolome analyses of lr34 mutants
   revealed the involvement of Lr34 in lignin biosynthesis. RNA-seq
   analysis indicated enrichment of DEGs in pathways related to
   plant-pathogen interactions, mitogen-activated protein kinase
   signaling, as well as diterpenoid and flavonoid biosynthesis. These
   pathways are known to contribute to disease resistance, suggesting that
   Lr34 may be implicated in multiple disease resistance pathways.
   Physiological and biochemical assays demonstrated that wheat lines
   expressing Lr34 exhibited increased cell wall thickness and higher
   lignin content in leaves, aligning with a previous report in barley
   ([217]Chauhan et al., 2015). Interestingly, Lr34 was significantly
   upregulated in three pairs of Lr34res-carrying wheat lines upon
   treatment with sinapyl alcohol, one of the three lignin monomers. This
   observation is consistent with findings in rice, where enhancing the
   contents of lignin subunits G and S was shown to increase leaf cell
   wall thickness ([218]Li et al., 2020). Similarly, our results show that
   higher sinapyl alcohol contents lead to thicker cell walls in leaf
   tissues, underscoring the critical role of lignin in reinforcing the
   structural integrity of plant cells against stressors.

   Cell wall lignification is a well-established fundamental defense
   mechanism against pathogen attacks ([219]Bacete et al., 2018). Studies
   have shown that TaCCoAOMT, a major gene involved in cell wall lignin
   biosynthesis, confers resistance to Fusarium head blight fungus by
   regulating cell wall thickness in wheat ([220]Yang et al., 2021).
   Another gene, TaCOMT-3D, enhances resistance to sharp eyespot fungus in
   wheat ([221]Wang et al., 2018). This study revealed that VIGS-induced
   silencing of TaCOMT-3B, which encodes an enzyme catalyzing the
   synthesis of the lignin monomer sinapyl alcohol, results in a lower
   sinapyl alcohol content and more severe disease symptoms upon pathogen
   infection. The genetic evidence from our study indicates that TaCOMT-3B
   functions synergistically with Lr34 to enhance resistance to stripe
   rust in wheat.

   While the biosynthesis of lignin is well documented, the mechanisms
   governing its transport to the extracellular cell wall remain largely
   unknown. To date, AtABCG29 is the only ABC transporter identified as a
   p-coumaryl alcohol transporter in Arabidopsis ([222]Alejandro et al.,
   2012). Our study provides multiple lines of supporting evidence for a
   role of LR34 as a sinapyl alcohol transporter. Firstly, Lr34res
   alleviates the inhibitory effects of sinapyl alcohol on plant growth
   and endows plants with a higher content of sinapyl alcohol. Secondly,
   both transcriptomic and metabolomic data indicate that genes and
   metabolites involved in sinapyl alcohol biosynthesis were downregulated
   in the lr34 mutant m19, which contains a premature stop mutation in
   Lr34res. The enhanced resistance to sinapyl alcohol stress and the
   increased sinapyl alcohol content in Lr34res lines suggest a potential
   link between Lr34res and sinapyl alcohol. Thirdly, yeast mutant
   analysis and efflux assays indicate that the LR34 protein can transport
   sinapyl alcohol. ABCG transporters, known for their diverse substrate
   selectivity, have been shown to improve disease resistance in plants
   ([223]Devanna et al., 2021). For example, AtPDR12 transports sclareol
   to enhance disease resistance in Arabidopsis, and MtABCG10 is involved
   in the transport of flavonoid precursors that inhibit Fusarium
   oxysporum growth in Medicago truncatula ([224]Campbell et al., 2003;
   [225]Banasiak et al., 2013). Although both Lr34res and Lr34sus were
   found to transport sinapyl alcohol in vitro, only the Lr34res protein
   was detected in wheat. Previous studies have shown that ionic strength
   can affect the solubility and stability of proteins ([226]Duong and
   Gabelli 2014), potentially due to the conversion of histidine (Lr34res)
   to tyrosine (Lr34sus), which reduces the protein’s positive charge.

   Lr34 has been widely utilized in wheat breeding, with more recent
   studies confirming its efficacy against multiple pathogens. A common
   breeding strategy involves combining multiple disease resistance to
   enhance overall crop resistance. Notably, significant additive effects
   have been observed when Lr34 is combined with Lr46, Lr68, Yr30, and
   other minor genes or quantitative trait loci in specific environments
   ([227]Bokore et al., 2021). In contrast, no such effects have been
   observed in other settings ([228]Lillemo et al., 2008). A resistance
   locus for both stripe rust and leaf rust has been consistently
   identified on chromosome 3BL in previous reports ([229]Lin and Chen.,
   2009; [230]Maccaferri et al., 2015; [231]Wu et al., 2020; [232]Jin et
   al., 2022; [233]Vikas et al., 2022). Here, we report that combining
   Lr34 with TaCOMT-3B—both crucial for lignin biosynthesis—significantly
   enhances disease resistance. Therefore, understanding the genetic
   mechanisms underlying disease resistance is essential for improving
   crop resilience through well-targeted crop breeding.

Materials and methods

Plant materials and growing conditions

   In this study, three pairs of wheat lines were utilized, each
   comprising a line carrying the Lr34 resistance allele (m19 r-sib, m21
   r-sib, and Avocet+Lr34) and its corresponding susceptible counterpart
   without the allele (m19, m21, Avocet). Specifically, Avocet+Lr34 was
   designated as Lr34res (resistant), and Avocet as Lr34sus (susceptible).
   m19 and m21 are two independent gamma (γ)-induced mutants with base
   deletions at positions 5035 and 10 620 bp within the Lr34 gene,
   respectively, leading to pre-mature stop mutations that result in
   truncated proteins ([234]Krattinger et al., 2009). The m19 r-sib and
   m21 r-sib lines serve as the resistant sibs for m19 and m21,
   respectively. All lines were cultivated over three cropping seasons
   from 2018 to 2021 in the field at Toluca, State of Mexico, Mexico, and
   at Ezhou in Hubei Province and Tianshui in Gansu Province, China.
   Additionally, a diverse panel of 312 wheat lines was assessed for
   disease severity in the field. The plots were organized in a randomized
   block design with two replicates at each location: Ezhou during
   2019–2021, Tianshui during 2020–2021, and Maerkang for the 2019 growing
   season. A bi-parental genetic population in the F[4] generation,
   derived from a cross between the resistant line m21 r-sib and the
   susceptible line SY95-71, was grown at Ezhou during the 2022–2023
   season. Each line was planted in 1.2-m rows with 0.3 m spacing between
   rows, sowing approximately 100 seeds per row. A mixture of stripe
   rust-susceptible lines—Apav#1, SY95-71, and Mingxian 169—was used as an
   inoculum spreader. Furthermore, the three pairs of Lr34 lines and
   Aikang 58 were cultivated under controlled greenhouse conditions at
   Huazhong Agricultural University. The greenhouse was maintained at
   18°C–22°C with a 22 h light/2 h dark photoperiod to facilitate
   experiments including VIGS, transmission electron microscopy (TEM),
   FoMV, and stripe rust inoculation at the seedling stage.

Inoculation and assessment of disease response

   In the field and greenhouse studies in China, a urediniospore mixture
   of the predominant Chinese Puccinia striiformis f. sp. tritici (Pst)
   races CYR32, CYR33, and CYR34, suspended in light mineral oil (Soltrol
   170), was sprayed onto the spreader rows at both the jointing and
   seedling stages. In Mexico, field trials utilized a mixture of Pst
   isolates MEX 14.191 and MEX 16.04. For powdery mildew experiments in
   the greenhouse, conidiospores of B. graminis f. sp. tritici (Bgt) race
   E21 were brushed onto the test plants. Disease severities were recorded
   using the modified Cobb Scale ([235]Peterson et al., 1948). The growth
   of F. graminearum (PH-1) mycelia was assessed on PDA medium
   supplemented with different concentrations of sinapyl alcohol over a
   period of 4 days at 28°C.

RNA-seq analysis

   During the 2018–2019 crop season in Toluca, the whole-leaf samples,
   pooled from more than six leaves, were collected from m19 r-sib and m19
   at different growth stages from the seedling (Zadoks: growth stage 12)
   to grain filling stage (Zadoks: growth stage 75). The samples were
   immediately frozen in liquid nitrogen and stored in an
   ultra-low-temperature freezer. Three biological replicates were
   analyzed for each stage and the Trizol reagent was used for total RNA
   extraction. RNA libraries were prepared and sequenced by Shanghai
   Majorbio Bio-pharm Technology. For each sample, the NovaSeq 6000
   sequencing system generated more than 11.24 Gb of clean data, with an
   alignment rate ranging from 85.2% to 95.05% against the reference
   sequence from the International Wheat Genome Sequencing Consortium
   (IWGSC) Chinese Spring 1.0 ([236]IWGSC 2018). A total of 91 904
   expressed genes were identified using DESeq^2 for differential
   quantitative gene expression analysis, including 80, 178 known genes
   and 11, 726 new genes. The screening threshold was set to |log^2FC| ≥ 1
   and an adjusted P value of < 0.05.

   Candidate genes identified from the RNA-seq data were subsequently
   validated by real-time qRT–PCR. We used 4 μg of total RNA to obtain
   complementary DNA (cDNA), employing a reverse transcription kit
   (HiScript Ⅲ 1st Strand cDNA Synthesis Kit). The qRT–PCR was performed
   on a CFX96 Touch Real-Time PCR Detection system (Bio-Rad) with ChamQ
   Universal SYBR qPCR Master Mix. Each reaction included 5 μl of ChamQ
   Master Mix, 3.6 μl of double-distilled H[2]O (ddH[2]O), 1 μl of
   1:5-diluted cDNA template, and 0.4 μl of primers (10 μM). Primer
   sequences are provided in [237]Supplemental Table 1.

Microscopy and chemical staining

   Cell wall thickness was assessed in three pairs of wheat lines at both
   the seedling (Zadoks: growth stage 12) and booting (Zadoks: growth
   stage 43) stages. The cell walls of selected leaf veins were observed
   by TEM at the electron microscope facility at Huazhong Agricultural
   University. Leaf samples were cut into small pieces and fixed in 3%
   (v/v) glutaraldehyde in a 50 mM phosphate buffer for 3–6 h at 4°C. The
   samples were then dehydrated in a graded alcohol series, embedded in
   gelatin capsules filled with LR White resin (Sigma-Aldrich, St. Louis,
   MO, USA), and polymerized at 60°C for 48 h. Ultra-thin sections of the
   samples were cut using a diamond knife and mounted on 200-mesh copper
   grids for TEM. The grids were examined with an H-7650 TEM (Hitachi,
   Japan) at 100 kV after staining with uranyl acetate and lead citrate to
   enhance contrast. Powdery mildew-infected leaf samples were immersed in
   fixation solution (methanol:chloroform:acetic acid, 6:3:1) to remove
   chlorophyll. The fixation solution was decanted, and samples were
   placed in 100% ethanol overnight. Subsequent steps involved a gradual
   ethanol rehydration process, where samples were incubated in decreasing
   concentrations of ethanol—90%, 70%, 50%, and 25%—each for at least 2 h.
   Following rehydration, the leaf samples were immersed overnight in
   150 mM K[2]HPO[4] containing 0.05% aniline blue. Finally, the samples
   were rinsed with 150 mM K[2]HPO[4] for 12 h before observation under a
   fluorescence microscope.

   For the analysis of powdery mildew, wheat leaves were clipped 48 and
   72 h after inoculation and decolorized using a 3:1 alcohol:acetic acid
   solution for 2–3 days. The leaves were then stained with a 0.6%
   Coomassie bright blue dye solution (0.6 g R250/100 ml) for about 2 min,
   immersed in water to remove any excess dye, mounted on a glass slide
   with 50% glycerin, and photographed under a microscope.

   The sections were stained with a 1% (w/v) phloroglucinol solution for
   10 min. After absorbing the residual liquid from the surface, 5 M
   hydrochloric acid was applied for another 10 min. Finally, the sections
   were washed with ddH[2]O and sealed for microscopic observations.

   A modified protocol was used for wheat germ agglutinin (WGA)
   ([238]Ayliffe et al., 2011). Briefly, leaves were collected in 10-ml
   tubes. After adding 5 ml of 1 M KOH and 0.05% Silwet L-77, the samples
   were incubated at 90°C for 30 min. The KOH solution was gently decanted
   and the samples were washed twice with 10 ml of 50 mM Tris buffer (pH
   7.5) for 20 min each. Subsequently, the samples were stained with 20
   μg/ml WGA–FITC (Sigma-Aldrich, St. Louis, MO, USA) for 15 min prior to
   microscopy. All WGA–FITC-stained tissues were examined under a
   microscope with GFP light excitation.

Chemical treatments

   Following the method described by [239]Krattinger et al. (2019), wheat
   seeds were sterilized with 70% ethanol for 1 min, followed by 2.6%
   sodium hypochlorite (NaOCl) for 10 min. Subsequently, the samples were
   washed five times for 2 min each with ddH[2]O and plated in Petri
   dishes that contained ½ Murashige and Skoog medium with or without 5 mM
   sinapyl alcohol. Treatments with p-coumaryl alcohol or coniferyl
   alcohol served as controls. The stock solutions were prepared in
   DMSO and equivalent amounts of DMSO were added to the controls. The
   plates were then held at room temperature and seedling growth was
   evaluated after 14 days. The treated lines were measured for length and
   then transplanted to a greenhouse. Inoculation was performed when the
   seedlings reached the first leaf stage. The optimal molecular
   concentration of 5 mM was established based on the results of an
   initial experiment and corroborated by previous studies.

Construction and screening of yeast mutants

   Optimized homologous recombination was used to generate gene-targeted
   knockouts in yeast (Saccharomyces cerevisiae strain W303-1a), following
   the method described by [240]Gardner and Jaspersen (2014). Strains
   W303-1a, pUG6, and pSH65 were purchased from Shanghai Baosai
   Biotechnology. We knocked out eight genes of the ABCG transporter
   family: PDR5, PDR10, PDR11, PDR12, PDR15, PDR18, SNQ2, and YOR1.
   Fifty-base-pair sequences upstream and downstream of the target coding
   sequences were used as homologous sequences for each target gene.
   Knockout constructs were prepared by overlapping PCR, using pUG6 as a
   template and kanamycin (KANMX) as the screening marker. We transformed
   35–50 μl of each construct into yeast cells using the PEG/LiAc
   (polyethylene glycol and lithium acetate mix) method. The
   transformation products were cultured overnight in yeast proteome
   database (YPD) medium and then transferred to YPD medium containing
   500 μg/ml geneticin, using the plate photocopying technique. The
   resultant single-colony yeast mutants were verified by PCR to confirm
   the knockout sequence. Yeast mutants W303-1a-PDR5Δ, W303-1a-PDR10Δ,
   W303-1a-PDR11Δ, W303-1a-PDR12Δ, W303-1a-PDR15Δ, W303-1a-PDR18Δ,
   W303-1a-SNQ2Δ, and W303-1a-YOR1Δ were obtained using this method and
   stored in an ultra-low-temperature freezer. The growth of both W303-1a
   and the mutants was assessed using 3 mM sinapyl alcohol to identify
   mutants with loss of resistance. A second round of knockout was
   performed using the Cre-loxP system. In the yeast mutant
   W303-1a-PDR15Δ, expression of the pSH65 plasmid was induced by
   galactose to produce cyclization recombination (Cre) enzyme to
   excise the middle sequence of the loxP site, thereby conferring
   resistance to W303-1a-PDR15Δ. Homologous recombination was used to
   knock out PDR11. The double mutant yeast strain W303-1a-PDR11ΔPDR15Δ
   was then screened using geneticin.

Heterologous expression of Lr34 in yeast and sinapyl alcohol loading assays

   Full-length cDNA of both Lr34res and Lr34sus was synthesized by
   Shanghai Sangon Biological Engineering Technology & Services Co., Ltd.,
   and the incomplete Lr34res sequences in lines m19 and m21 were
   amplified (refer to [241]Supplemental Table 1 for the primer
   sequences). The four sequences were cloned into the pDR196 vector and
   the resulting constructs were transformed into the mutant yeast
   strain W303-1a-PDR11ΔPDR15Δ using the PEG/LiAc method. Each yeast
   strain was diluted to the same concentration, and 2 μl of each strain
   that overexpressed the Lr34res, Lr34sus, m19, m21, or empty pDR196
   vector was used to inoculate 200-μl of SD-Ura liquid medium with or
   without 3 mM sinapyl alcohol. The cultures were grown in cell culture
   plates.

   The yeast cultures were collected twice using SD-Ura liquid medium
   after active secondary yeast cells reached the logarithmic phase and
   the optical density at 600 nm (OD[600]) stabilized after suspension.
   Sinapyl alcohol was then added in a volume equal to that of the
   suspension. Yeast cells were collected hourly, followed by washing with
   ice-cold water. Glass beads were introduced to lyse the cells. Equal
   volume of yeast liquid was collected and then treated with 3 mM sinapyl
   alcohol for 1.5 h at 4°C, washed, and resuspended in sinapyl
   alcohol-free SD-Ura medium at the initial time point (time 0). Then,
   the yeast culture was collected every 20 min, washed with ice-cold
   water, and lysed using glass beads. The purified cells were then
   filtered through a 0.22-μm aperture filter. The concentration of
   lignin monomers was determined by high-performance liquid
   chromatography.

Measurement of lignin content and composition

   Leaf samples were dried at 105°C for 30 min and maintained at 80°C in
   an oven. The concentration of thioglycolic acid lignin was determined
   according to [242]Fu et al. (2019a). Samples were pulverized using a
   high-throughput tissue-grinding device and the material passing through
   a 100-μm mesh was collected for analysis. For each assay, samples
   (20 mg) were placed into screwcap centrifuge tubes. Then, 2 ml of
   ddH[2]O was added to each tube, which was shaken for 1 min,
   centrifuged, and the precipitate was retained. Subsequently, 1.8 ml of
   methanol was added to each tube, and the samples were incubated at 60°C
   for 20 min before being vacuum-dried. Each sample was then mixed with 1
   ml of 3 M HCL and 0.1 ml of thioglycolic acid and incubated at 80°C for
   10 min. Following centrifugation, the precipitate was washed with
   ddH[2]O, and 1 ml of 1 M NaOH was added to each sample and shaken at
   room temperature for 16 h. The supernatant (1 ml) was mixed with 0.2 ml
   of concentrated hydrochloric acid and incubated at 4°C for 4 h. The
   resulting precipitate was retained, and 1 ml of 1 M NaOH was added,
   followed by shaking to ensure thorough mixing. Then, the mixture was
   diluted 50 times with 1 M NaOH. The absorbance of the diluted solution
   was measured at 280 nm using a microplate reader configured with a
   96-well UV-star microplate. For calibration, 1 M of NaOH was used as a
   blank control. The absorbance value was normalized to the dry weight
   (20 mg) of the sample to determine the relative lignin content in
   leaves.

   Non-inoculated samples collected at the booting growth stage were
   freeze-dried and crushed using a high-throughput tissue grinder.
   Metabolites were determined by MetWare Biotechnology (Wuhan) and the
   metabolome platform at Huazhong Agricultural University. The data were
   analyzed using the targeted metabolomics method ([243]Shi et al.,
   2020). We synthesized deuterated sinapyl alcohol (sinapyl alcohol-d3)
   (MedChemExpress) for detection in the m19 r-sib and m19 samples.

   The flag leaves of wheat at the booting stage were freeze-dried at low
   temperatures and ground into a fine powder. A 6-mg sample was weighed
   and mixed with KBr for tablet formation. The samples were analyzed
   using the Nicolet IS50R spectrometer (Thermo).

Validation of recombinant protein expression and function

   The candidate gene of TaCOMT-3B (TraesCS3B02G612000), was amplified
   using specific primers from the cDNA template of Avocet+Lr34
   ([244]Supplemental Table 1). TaCOMT-3B is located between 829391763 and
   829382973 on chromosome 3B according to the IWGSC Chinese Spring 1.0
   reference sequence ([245]IWGSC 2018). The cDNA was cloned into a
   modified pGEX-6p-1 expression vector (Novagen, Darmstadt, Germany) with
   a glutathione S-transferase (GST) tag. The resulting vector was
   transformed into Escherichia coli BL21 (Weidi Biotechnology, Shanghai)
   for heterologous protein expression. Single colonies were selected and
   cultured in Luria–Bertani liquid medium containing ampicillin until
   A[600nm] of 0.6–0.8 was achieved. Protein expression was then induced
   with isopropyl-β-D-thiogalactoside (with a final concentration of
   0.1 mmol/L) for 16 h at 20°C. Cells were harvested by centrifugation at
   6000 g for 10 min and disrupted using a high-pressure cracker.
   GST-fused target proteins were purified using glutathione Sepharose 4B
   agarose (GE Healthcare, Boston, USA) and confirmed by sodium dodecyl
   sulfate polyacrylamide gel electrophoresis (SDS–PAGE). All purified
   proteins were stored at −80°C. The functional activity of the purified
   protein was assessed in an assay mixture containing 0.2 M Tris–HCl
   (pH = 7.5), 10 mM MgCl[2], 10 mM 5-hydroxyferulic acid, and 0.75 mM
   S-adenosyl methionine, incubated at 37°C for 30-min. Accumulation of
   metabolites was quantified using liquid chromatography–mass
   spectrometry.

FoMV-mediated gene expression

   The FoMV was determined according to [246]Bouton et al. (2018). The
   full-length cDNA sequences of TaCOMT-3B were cloned into a PV101
   vector. The resulting expression construct were introduced into the
   Agrobacterium tumefaciens strain GV3101. Bacterial cultures were
   obtained by inoculating single colonies into liquid Luria–Bertani
   medium supplemented with 20 μg/ml Rifampicin and 50 μg/ml Kanamycin,
   followed by incubation at 28°C for 20 h with constant shaking at
   250 rpm. A. tumefaciens cells were pelleted at 5000 rpm for 10 min at
   24°C and resuspended in infiltration medium containing 0.5 M MES (pH
   5.8), 1 M MgCl[2], and 100 μM acetosyringone. The bacterial suspensions
   were adjusted to an OD[600] of 1 and incubated at room temperature for
   at least 1 h. To initiate infection in N. benthamiana plants, the
   FoMV-vector-containing A. tumefaciens suspension was mixed with an
   equal volume of the pSoup-p19-containing A. tumefaciens suspension. The
   mixture was then pressure infiltrated into the abaxial side of the
   fully expanded leaves of three young seedlings using a needleless
   syringe.

   To initiate infection in wheat plants, leaves of young seedlings were
   rub inoculated using FoMV-containing sap prepared from N. benthamiana
   leaves agroinfiltrated as described above and harvested 3 days
   post-inoculation (dpi). The sap produced by finely grinding
   N. benthamiana leaves in deionized water using a pestle was
   supplemented with 1% (w/v) Celite 545 AW (Macklin) and used for rub
   inoculation of the first two leaves of two-leaf-stage wheat seedlings.
   Five minutes post-inoculation, the leaves were sprayed with tap water
   to remove residual sap and Celite. Subsequently, the plants were bagged
   and maintained under high humidity in darkness for 24 h before they
   were returned to standard growth conditions. The third leaves were
   collected for metabolite determination.

Protein extraction, SDS–PAGE, and western blot

   Total protein extracts were obtained by grinding wheat leaves and yeast
   in liquid nitrogen using a mortar and pestle. Total plant and yeast
   membrane proteins were extracted separately using plant and yeast
   membrane protein extraction kits (Besbio), respectively. Protein
   concentrations were determined using a bicinchoninic acid (BCA) assay.
   Equal amounts of protein (20 μg) were separated on an 8%
   acrylamide/bis-acrylamide gel (Bio-Rad) and blotted to a polyvinylidene
   fluoride membrane (Thermo Fisher). Blots were incubated with a 1:500
   dilution of rabbit monoclonal antibody (anti-Lr34), which was produced
   by GenScript Biotechnology using the peptide H-ESELELASRQRQNGC-OH.
   Signal was detected using enhanced chemiluminescence. The antibody used
   for detection in yeast, anti-ACTB (horseradish peroxidase conjugate)
   mouse monoclonal antibody, was ordered from Sangon Biotech.

BSMV-mediated gene silencing

   Previous studies identified three homologous COMT genes on wheat
   chromosomes 3A, 3B, and 3D, however, only TaCOMT-3B was highly
   correlated with lignin content ([247]Fu et al., 2019a). TaCOMT-3B has a
   222-bp deletion in the 3′ UTR in a subset of wheat collections,
   allowing for its classification into two genotypes ([248]Fu et al.,
   2019a). In this study, we designated the genotype containing the 222-bp
   fragment in the 3′ UTR TaCOMT-3Ba and the one lacking it TaCOMT-3Bb.
   VIGS was performed using a 1:1:1 mixture of α:β:γ (or γ + target) RNAs
   of BSMV on m19 r-sib and m19 ([249]Huang et al., 2017). The target
   sequence, spanning 211 to 525 bp of the TaCOMT-3B coding sequence, was
   synthesized from two oligos with 15-bp overlaps at the ends and
   inserted into an expression plasmid encoding the γ subgenome of BSMV.
   The sequences of these oligos are listed in [250]Supplemental Table 1.
   Plants inoculated with BSMV 00, in which the target gene was not
   silenced, served as controls. The VIGS experiment was conducted on
   fully extended leaves of plants from the jointing stage. The treated
   plants were kept in a dark and humid environment for 2 days and then
   returned to the greenhouse for normal growth. The phenotypes were
   observed after 2 weeks and we observed photobleached flag leaves. The
   expression levels of TaCOMT-3Ba (ID: TraesCS3B02G612000) were measured
   using real-time qRT–PCR with a pair of specific primers amplifying the
   958 to 1147 bp of TaCOMT-3B. We also developed the gene-based molecular
   marker ranging from 95 to 1721 bp of TaCOMT-3B to genotype the wheat
   germplasm collection and genetic population. Primer sequences are
   provided in [251]Supplemental Table 1.

Molecular docking simulation

   AlphaFold2 was used for the prediction of Lr34res and Lr34sus
   three-dimensional structures. The chemical structure of sinapyl alcohol
   was retrieved from the PubChem Compound database. Prior to docking,
   water molecules were removed, and both the sinapyl alcohol ligand and
   the Lr34res and Lr34sus receptors were hydrogenated. Molecular docking
   was performed using Autodock and the results were visualized using
   PyMOL ([252]Seeliger and Groot 2010).

Quantification and statistical analysis

   Statistical analyses were performed using SPSS software. All values are
   presented as means ± SD, and the numbers of samples (n) are indicated.
   Statistically significant differences between control and experimental
   groups were determined by one-way ANOVA using SPSS software.

Accession numbers

   The raw RNA-seq data were deposited in the Sequence Read Archive at the
   China National Genomics Data Center with the BioProject accession
   number CRA018588. Sequence data from this article can be accessed
   through the Ensembl Plants portal
   ([253]http://plants.ensembl.org/Triticum_aestivum/Info/Index) under the
   accession number TaCOMT-3B (TraesCS3B02G61200, Ensembl Plants).

Funding

   This work was supported by the National Natural Science Foundation of
   China (grant nos. 31861143010 and 32372173), the National Key Research
   and Development Program of China (grants 2022YFD1201300 and
   2022YFD1201500), the Fundamental Research Funds for the Central
   Universities (2662020ZKPY005), and the Hubei Hongshan Laboratory
   (2022hspy001, 2021hskf008, and 2022hspy010).

Acknowledgments