Cell-Wall-Degrading Enzymes-Related Genes Originating from Rhizoctonia solani Increase Sugar Beet Root Damage in the Presence of Leuconostoc mesenteroides

Sugar beet crown and root rot caused by Rhizoctonia solani is a major yield constraint. Root rot is highly increased when R. solani and Leuconostoc mesenteroides co-infect roots. We hypothesized that the absence of plant cell-wall-degrading enzymes in L. mesenteroides and their supply by R. solani during close contact, causes increased damage. In planta root inoculation with or without cell-wall-degrading enzymes showed greater rot when L. mesenteroides was combined with cellulase (22 mm rot), polygalacturonase (47 mm), and pectin lyase (57 mm) versus these enzymes (0–26 mm), R. solani (20 mm), and L. mesenteroides (13 mm) individually. Carbohydrate analysis revealed increased simpler carbohydrates (namely glucose + galactose, and fructose) in the infected roots versus mock control, possibly due to the degradation of complex cell wall carbohydrates. Expression of R. solani cellulase, polygalacturonase, and pectin lyase genes during root infection corroborated well with the enzyme data. Global mRNAseq analysis identified candidate genes and highly co-expressed gene modules in all three organisms that might be critical in host plant defense and pathogenesis. Targeting R. solani cell-wall-degrading enzymes in the future could be an effective strategy to mitigate root damage during its interaction with L. mesenteroides.


Introduction
Sugar beets (Beta vulgaris L.) contribute to 55% of the total sugar produced in the United States (https://www.sugar.org/about/us-industry/, accessed on 5 September 2021). Sugar beet biotic stressors, such as fungi and viruses, are some of the major yield constraints that limit sugar production [1][2][3][4]. Resistance to these pathogens is highly limited due to the lack of appropriate genetic resources [1,5]. When available, resistance is usually controlled by a few genes that are being potentially compromised by the evolving pathogens, or if resistance is controlled by many genes (quantitative trait) it becomes highly challenging and time consuming to introgress resistance into the commercial hybrids. In many cases, genetic resistance comes with a trade-off of reduced sugar yield. The future use of cuttingedge molecular biology tools, such as RNA interference (RNAi) and CRISPR-Cas, could help improve host resistance, but this requires the identification of potential target genes in pathogens that are highly critical in controlling pathogenesis and the development of disease symptoms.
Rhizoctonia crown and root rot (RCRR) in sugar beet is caused by the fungus Rhizoctonia solani Kühn, and it is a major problem in the United States and around the world. The fungus can attack the sugar beet at early seedling stages by targeting hypocotyls below the soil causing damping-off, and can attack mature plants at later root development stages. The symptoms are underground roots exhibiting black root lesions, including large

Cell-Wall-Degrading Enzymes Increased Root Rot Symptoms When Combined with Pathogen Inoculations
The data from the two field studies were pooled for analyses (Table 1) since these were not significantly different (p = 0.1887), interactions were not significant (p = 0.1471 to 0.8664), and their variances were homogeneous (p = 0.6036). When compared across all treatments (Table 1, Figure 1), treatments with L. mesenteroides had more (p < 0.0001) rot (40 mm) than treatments without L. mesenteroides (20 mm). When L. mesenteroides and R. solani were inoculated individually, limited rot occurred (13 and 20 mm, respectively), but significantly more rot (33 mm) occurred when combined. In Table 1, when the enzymes were inoculated individually and in combination with L. mesenteroides, rot was higher in all comparisons versus the enzymes without L. mesenteroides (rot with and without L. mesenteroides): pectin lyase (PNL; 57 and 26 mm), polygalacturonase (PG; 47 and 11 mm), and cellulase (CEL; 22 and 4 mm), except for Vicozyme (V; 65 and 58 mm). The water check isolations (n = 4) were negative for both Leuconostoc and Rhizoctonia in both studies. The Leuconostoc checks (n = 4) were positive at high concentrations (solid streak of bacteria) of Leuconostoc in Study 1 and positive in three of four isolations in Study 2, while negative for R. solani. The R. solani checks (n = 4) were positive for R. solani in one of four isolations for both studies. The R. solani checks were frequently invaded by bacteria and were positive at low concentrations (a few isolated colonies in bacterial streak) for Leuconostoc in three of four isolations for both studies. In enzyme-Leuconostoc combination treatments, Leuconostoc was found in 14 of 14 and 13 of 14 isolations (at high concentrations) in Test 1 and Test 2, respectively, while R. solani was not isolated. In enzyme treatments without Leuconostoc, Leuconostoc was found in 6 of 14 isolations in both studies at low concentrations, while R. solani was not isolated.  <0.0001 LSD (α = 0.05) 7 y CEL = cellulase from Aspergillus niger (Sigma product C1184); None = no enzyme; PG = polygalacturonase from Rhizopus (P2401); PNL = pectin lyase from Aspergillus (P3026); and V = Vicozyme L, a commercial multienzyme complex (arabanase, cellulase, B-glucanase, hemicellulase, and xylanase; V2010). The enzyme concentrations were doubled compared to those used in the 2018 and 2019 studies. y None = water; L12311= Leuconostoc mesenteroides strain L12311; and F521 = Rhizoctonia solani strain F521 AG2-2 IIIB. z The two studies were each arranged in a randomized complete block design with eight replications. These studies were analyzed together since the studies were not significantly different (p = 0.1887), interactions were not significant (p = 0.1471 to 0.8664), and their variances were homogeneous (p = 0.6036). p > F was the probability associated with the F value. The means followed by the same letter did not differ significantly based on Fisher's protected least significant difference (LSD; α = 0.05).

Expression of R. solani Pectin Lyase (PNL), Polygalacturonase (PG), and Cellulase (CEL) Genes Were Highly Upregulated during Early Root Infection Stages
As root rot symptoms were increased by the exogenous application of PCWDEs, especially PNL, PG, and CEL when combined with the pathogens, we therefore looked at the expression of R. solani PNL, PG, and CEL genes during root infection in the presence or absence of L. mesenteroides. Expression values of these genes at different time points and treatments were extrapolated from the RNAseq data. Gene expression patterns were investigated to understand their roles in early root pathogenesis besides comparing their expressions to the exogenous enzyme application and root rot symptoms. The two expressed PNL genes at all time timepoints include AG2-2IIIB_02409 and AG2-2IIIB_08245/08246 among eight to nine paralogues ( Figure 2A). Their expression increased over time and AG2-2IIIB_02409 showed the highest expression (Fragments Per Kilobase of transcript per Million mapped reads; FPKM: 15.7) at 3 dpi. Among the three R. solani gene families that code for PCWDEs and were used in this study, PG genes showed the highest expression at early stages of sugar beet root infection by R. solani ( Figure 2B). Out of eight paralogues of R. solani PG genes, expression of AG2-2IIIB_00811 was detected at the earliest timepoint (1 dpi) and was highest at 3 dpi (FPKM: 46.58) in comparison to the other paralogues, namely CEL and PNL genes. Expression of AG2-2IIIB_00811 increased by >4.5-fold and >15-fold at 2 dpi and 3 dpi respectively in comparison to its expression at 1 dpi. The other expressed PG gene paralogues include AG2-2IIIB_02257 and AG2-2IIIB_10535, but their expression was very low (FPKM: 0.09-0.79) and mainly expressed at 3 dpi. Among the 11 CEL genes identified in our study, the expressed paralogues were AG2-2IIIB_04162, AG2-2IIIB_10826, AG2-2IIIB_11320, AG2-2IIIB_03478, AG2-2IIIB_04711, and AG2-2IIIB_02154 ( Figure 2C). In general, their expression was low to moderate. Expression of AG22IIIB_10826 was detected at 1 dpi, and that increased over time and was highest at 3 dpi (FPKM: 8.18). In general, expression of R. solani PG, PNL, and CEL genes were lower in R. solani + L. mesenteroidesinfected sugar beet roots in comparison to R. solani-infected roots only. and (C) cellulase gene families at 1, 2, and 3 dpi. Data are mean ± SE of four biological replicates; * p ≤ 0.05 between mock control (C) and treatments.

Carbohydrates Were Greatly Altered in the Infected Sugar Beet Roots
At 3 dpi, the cellular content of sucrose decreased (vs. control) by 18% in the R. solani and L. mesenteroides-infected roots and showed the highest decrease (27%) when both pathogens infected the roots together ( Figure 3A). Fructose content increased by 500% and 280% when sugar beet roots were infected by R. solani and L. mesenteroides, respectively, in comparison to the control samples ( Figure 3B). This increase was even higher (~560%) when the pathogens co-infected. The cellular content of glucose + galactose increased by 1440% and 900% in the R. solani and L. mesenteroides-infected roots, respectively ( Figure 3C), and co-infection resulted in the highest increase (2146%). Raffinose was present in minute amounts in the control samples only ( Figure 3D).

Percentage of Root N Was Highly Increased upon Infection
Percentage of N in the roots was highly altered depending upon the treatment type ( Figure 4). Sugar beet roots treated with R. solani and/or L. mesenteroides did not show any change in C content except for L. mesenteroides-infected roots, which showed a small increase in C in comparison to the mock samples at 3 dpi ( Figure 4A). Nitrogen content increased (vs. control) by 40% in the R. solani and L. mesenteroides-infected roots, and this increase was even higher (56%) when roots were infected by both pathogens ( Figure 4A). The ratio of C/N decreased by~29% in the R. solani and L. mesenteroides-treated roots, and this decrease was higher (36%) when the two pathogens co-infected ( Figure 4C).

GO and KEGG Analyses of Differentially Expressed Genes
Gene ontology (GO) analysis of sugar beet genes whose transcripts were differentially expressed by pathogen infection at 1 dpi were mainly associated with the plasma membrane, ATP binding, kinase activity, protein phosphorylation, and microtubule-related activities to name a few ( Figure S1A). At 2 dpi, differentially expressed genes were associated with the plasma membrane, chloroplast, cytosol, the integral component of the membrane, and ATP binding, etc., ( Figure S1B). The categories of genes at 3 dpi were like that of 2 dpi, but included other types like protein phosphorylation, serine/threonine kinase, and oxidation-reduction-related, etc., ( Figure S1C).
In R. solani at 1 dpi, differentially expressed genes were associated with the ribosomal subunit, translation, and the tricarboxylic acid cycle to name a few ( Figure S2A). The trend was similar at 2 dpi but included other categories of genes, such as a cellular response to drug-related genes ( Figure S2B). At 3 dpi, differentially expressed genes were mainly related to the ribosome, hyphal growth, intracellular protein transport, and vesicle-mediated transport, etc., ( Figure S2C).
In L. mesenteroides at 1 dpi, differentially expressed genes were primarily associated with penicillin binding (Figure S3A). At 2 dpi, differentially expressed genes included DNA-directed 5 -3 RNA polymerase activity and the cytosolic small ribosomal subunit ( Figure S3B), and at 3 dpi, genes were primarily associated with aldehyde-lyase activity ( Figure S3C).
Pathway enrichment analysis of sugar beet genes at 1 dpi represented genes mainly associated with amino sugar and nucleotide sugar metabolism, starch and sugar metabolism, plant hormone signal transaction, and glutathione metabolism, etc., (Figure 8A). At 2 dpi, differentially expressed genes were mainly associated with the ribosome, glycolysis/ gluconeogenesis, pyruvate metabolism, Arginine (Arg)/Proline (Pro) metabolism, and Alanine (Ala)-Aspartate (Asp)-Glutamate (Glu) metabolism, etc., ( Figure 8B). Some of the differentially expressed genes at 3 dpi were associated with the mitogen-activated protein kinase (MAPK) signaling pathway, amino sugar and nucleotide sugar metabolism, plant hormone signal transaction, starch, and sucrose metabolism, etc., ( Figure 8C). On the other hand, pathway enrichment analysis of R. solani genes that were differentially expressed at 1 dpi were mainly associated with the ribosome, tricarboxylic acid (TCA) cycle, glyoxylate and dicarboxylate metabolism, and oxidative phosphorylation, etc., (Figure 9A). At 2 dpi, differentially expressed genes were mainly associated with the ribosome, glycolysis/gluconeogenesis, pyruvate metabolism, TCA cycle, Arg and Pro metabolism, and Ala/Asp/Glu metabolism ( Figure 9B). At 3 dpi, some additional categories of genes associated with the ribosome, proteasome, phagosome, and oxidative phosphorylation, etc., were observed ( Figure 9C). Based on the types of genes that were differentially expressed in L. mesenteroides, we were unable to perform GO and pathway enrichment analyses. Pathway enrichment analysis of L. mesenteroides differentially expressed genes were primarily associated with the ribosome and selenocompound metabolism at 1 dpi, and RNA polymerase-related at 2 dpi ( Figure S4A,B). At 3 dpi, additional categories of genes related to the ascorbate and aldarate, histidine, phenylalanine, tyrosine, and alkaloid (tropane, piperidine, and pyridine) metabolism were highly evident ( Figure S4C).

WGCNA Analyses of Differentially Expressed Genes
WGCNA analysis was performed using the count data of RNA-sequencing reads mapped to the genomes of all three organisms used in this study. The WGCNA analysis evaluates the pairwise correlation between genes across samples. The correlation coefficient was determined for all genes and a hierarchical clustering tree was constructed using the correlation matrices. Gene modules are represented by different colors and each module consists of several genes. The heatmap and gene cluster dendrogram of sugar beet genes are shown in Figure 10A,B. Highly co-expressed (positive correlation > 0.5; significance of p < 0.05) gene modules in sugar beets ( Figure 10C) in response to R. solani inoculation at 1 dpi were represented by MElightsteelblue1, MEsteelblue, and MEskyblue. At 2 dpi, gene modules detected were MEwhite, MEbrown4, MEdarkmagenta, and MElightcyan, and at 3 dpi, detected modules were MElightyellow, MEtan, MEred, MEviolet, MEbrown, and MEgreen. Sugar beet L. mesenteroides-responsive gene modules at 1 dpi were MEmedi-umpurple3, MEplum1, and MEcyan. At 2 dpi, they were MEpaleturquoise, MEdarkgreen, MEroyalblue, and MEyellowgreen, and at 3 dpi, they were MEdarkturquoise, MEdarkred, MEdarkorange, and MEorange. Sugar beet gene modules responsive to R. solani + L. mesenteroides at 1 dpi were represented by MElightgreen, MEorangered4, and MEdarkslateblue. At 2 dpi, the detected modules were MEbisque4, MEsaddlebrown, and MEdarkgrey, and at 3 dpi, they were MEpink, MEgreenyellow, MEviolet, MEyellow, MEblack, and MEbrown.   The heatmap and gene cluster dendrogram of R. solani genes are shown in Figure 11A,B. The highly co-expressed (positive correlation > 0.5; significance of p < 0.05) R. solani gene modules for the R. solani only inoculation of sugar beets were represented by MEturquoise, MEgrey, and MEblue, and for the R. solani + L. mesenteroides inoculation they were represented by MEturquoise and MEgrey at 3 dpi only ( Figure 11C). The heatmap and gene cluster dendrogram of L. mesenteroides genes are shown in Figure S5A,B. The highly co-expressed (positive correlation > 0.5; significance of p < 0.05) L. mesenteroides gene module at 3 dpi, while infecting sugar beets alone or with R. solani + L. mesenteroides inoculation, was MEturquoise ( Figure S5C), whereas at 2 dpi, the L. mesenteroides gene module for the R. solani + L. mesenteroides inoculation was represented by MEsalmon ( Figure S5C).

R. solani-Derived Plant Cell-Wall-Degrading Enzymes Play a Key Role in Increasing Root Rot
Plant cell-wall-degrading enzymes in phytopathogenic fungi are well known for their critical role in plant pathogenesis [14]. They not only depolymerize complex cell wall polysaccharides, thereby facilitating fungal entry inside the plant cell, but the simpler sugars derived from cell wall degradation can also serve as an easy source of energy during early pathogenesis. Among the different PCWDEs produced by R. solani, pectin-digesting enzymes play a critical role in successful pathogenesis and disease development [15][16][17]. The enzyme PG hydrolyzes pectate by cleaving the α-(1,4)-glycosidic bonds. The other PCWDE used in this study, CEL, targets cellulose present in the cell wall and catalyzes the degradation of cellulose by targeting the β-1,4-glycosidic bonds. Cellulase isolated from R. solani has been shown to act as an elicitor during plant-pathogen interactions, though its enzyme activity is not required for such elicitor activity [18]. Earlier studies involving PCWDE-related genes have demonstrated the role of these enzymes in fungal pathogenicity, including R. solani [15][16][17]. The goal of this study was to delineate precisely the role of R. solani PCWDEs, namely PNL, PG, and CEL, in increasing sugar beet root rot symptoms when in close contact with L. mesenteroides in the soil. We used both exogenously supplied purified enzymes in combination with L. mesenteroides or L. mesenteroides alone, and R. solani + L. mesenteroides inoculations without any PCWDEs. The data presented here suggest that pectin-degrading enzymes such as PNL and PG had more measurable root rot than CEL, with PNL showing the most rot among the three enzymes when individual enzymes were combined with L. mesenteroides inoculation (Table 1; Figure 1). Other enzymes such as xylanase, pectate lyase, and pectin methylesterase were investigated in preliminary studies, but led to less rot when in combination with L. mesenteroides than PNL, PG, and CEL (data not shown). The highest root rot in the case of 'V' was not surprising as this is an admixture of multiple PCWDEs. The enzyme data was further supported by the expression of R. solani PNL, PG, and CEL genes that were highly induced at early stages of root infection by R. solani alone or during co-infection with L. mesenteroides (Figure 2). Though a greater number of CEL genes were expressed in comparison to PNL and PG genes, the overall expression of PNL and PG was higher than CEL, and were highest at 3 dpi.

Altered Cell Wall Metabolism Affects Carbohydrate Contents and Decreases C/N Ratio upon Infection
An increase in simpler carbohydrates, namely fructose and glucose + galactose upon R. solani and L. mesenteroides infections and more so when the pathogens co-infected sugar beet roots (Figure 3), suggest a breakdown of complex cell wall carbohydrates into simpler carbohydrates, thereby promoting pathogenesis [19]. On the other hand, decreases in sucrose content (18%) upon infections by individual pathogens, and even more (27%) during co-infection ( Figure 3A), suggest that the rapid utilization of stored sucrose in roots was an easy source of C by the invading pathogens. Studies on pathogen infections of sugar beet roots during storage, have reported similar findings where sucrose content decreased, and glucose and fructose increased [20]. We also investigated if changes in carbohydrate content during pathogenesis affected total C and N contents in the roots. Total C was primarily unaltered except for the L. mesenteroides-infected roots ( Figure 4A). An increase in total N in the infected roots ( Figure 4B) indicated that there was potentially a high demand for N during fungal and bacterial pathogenesis. This finding was further supported by an increased expression of N metabolism-related genes, both in the roots as well as in R. solani (Figures 8 and 9). As N serves as a building block for amino acids and numerous other nitrogenous compounds, there is a high demand for N, both from the host defense perspective as well as the pathogenesis perspective (reviewed in study [21]). Though N metabolism plays a key role in host plant defense against pathogens, there may be a threshold beyond which additional N availability in plants might contribute to host susceptibility. Other factors, such as the form of N supplied, pathogen type (i.e., fungi (biotroph/necrotrophy) or bacteria), or interacting plant species, etc., can contribute to host plant susceptibility or resistance. From the pathogen perspective, there is a high demand for N at infection sites to maintain successful pathogenesis (reviewed in study [21]). During plant-fungal pathogenic interactions, the majority of the N required by fungi is acquired from host plants in the form of amino acids such as GABA, Gln, and Glu, etc., which are present in higher concentrations (millimolar quantities) in comparison to other amino acids in plant cells [22]. Thus, N starvation can regulate pathogenicity-related genes and thereby control disease development [23]. An increase in N content in the infected roots, and a significantly lower C/N ratio (vs. mock; Figure 4C), may indicate an increase in the N-driven primary metabolism as opposed to the C-driven secondary metabolism, such as the biosynthesis of lignin and cellulose, and their roles in reinforcing physical barriers (reviewed in study [24]). The lower C/N ratio in the infected roots, and increased root damage, implicated an increased host susceptibility upon an altered C/N ratio.

Global Gene Expression Analyses Show Distinct Patterns in Sugar Beet, R. solani, and L. mesenteroides during Complex Interactions
A systems biology approach offers tremendous potential to understand complex biological interactions and identify critical components when multiple organisms interact with each other. In our case, sugar beets growing in the field were root inoculated with R. solani and/or L. mesenteroides and we evaluated global changes in gene expression, both in the sugar beet host and pathogens (R. solani and L. mesenteroides) during early infection stages ranging between 24 hpi and 72 hpi. Our goal was to understand the changes in early gene expression in sugar beets, and the pathogens (fungal and bacterial) that are associated with host response and pathogenesis respectively.
The gene expression data provide insights on pathogen-specific sugar beet responses, including overlaps between fungal and bacterial pathogens in this study. Some of the highly expressed sugar beet genes due to R. solani infection at the earliest time point (1 dpi) included peroxidase 27 (EL10Ac6g15542) and auxin-binding protein ABP19b (EL10Ac8g19059), etc. The higher expression of peroxidase and phytohormone-related genes (such as auxin, gibberellin) observed here possibly indicates their role in host defense at early infection stages ( Table 2; Figure 5A). The similar role of some of these genes against other R. solani strains have been reported in diverse plant species, including soybean [25], potato [26], rice [27], and sugar beet [28,29]. Sugar beet genes were highly upregulated (up to 7-fold) by L. mesenteroides such as lipid-binding protein AIR1B (EL10Ac5g13046) and auxin-repressed 12.5 kDa protein isoform X1 (EL10Ac8g20421), which show bacterial pathogen-specific sugar beet responses (Table 2), whereas other highly upregulated (up to 348-fold) sugar beet genes, such as polygalacturonase inhibitor 1 (EL10Ac3g06968) and auxin-binding protein ABP19b-like (EL10Ac8g19076) highlight the global response of the sugar beet against fungal and bacterial pathogens ( Table 2). An increase in the expression of the sugar beet polygalacturonase inhibitor gene (EL10Ac3g06968) corroborates well with the observation from the exogenous enzyme application study, whereby the PG enzyme had a significant effect on increasing root damage. An increase in R. solani PG genes during early infection stages reiterates the fact that this gene family is highly critical for successful pathogenesis. The future RNAi-mediated targeting of specific members of the R. solani PG gene family, along with other candidates, may be highly effective in improving sugar beet host resistance against both R. solani and L. mesenteroides. Furthermore, designing inhibitors specific to PG, or other PCWDEs such as PNL, may serve as an alternative approach to improve host resistance. Further insights into the candidate genes that were highly co-expressed in the gene modules at the earliest infection stage (1 dpi) revealed a unique sugar beet gene expression against R. solani (ME_lightsteelblue1: heavy-metal-associated (EL10Ac3g05414), methionine-tRNA ligase (EL10Ac8g19256), ATHILA ORF-1 family (EL10Ac4g08425), and U-box domain-containing protein 43 (EL10Ac1g01912), etc.), L. mesenteroides (ME_mediumpurple3: proliferating cell nuclear antigen 2 (EL10Ac1g00531), FAS1 domain-containing protein (EL10Ac5g10640), and retrovirus-related Pol polyprotein (EL10Ac5g11625), etc.), and R. solani + L. mesenteroides (ME_lightgreen: E3 ubiquitin-protein ligase (EL10Ac1g01117), mechanosensitive ion channel protein 10 (EL10Ac4g10289), and organic cation/carnitine transporter 2 (EL10Ac9g22914), etc.). Sugar beet genes that were downregulated by both pathogens include aquaporin TIP2-1 (EL10Ac9g22046) and probably xyloglucan endotransglucosylase/hydrolase (XTH) protein 6 (EL10Ac5g11874), etc. Both genes have been demonstrated to confer host resistance against pathogens in plants [30,31]. The role of aquaporins in host plant resistance against diverse types of pathogens have been reported in several studies [31,32]. In citrus (Citrus sinensis), the CsXTH04 gene was highly upregulated in the citrus bacterial canker-resistant varieties and were induced by phytohormones namely salicylic acid and methyl jasmonate [30]. Downregulation of both genes, during susceptible interaction in our case and evidence from past reports taken together into consideration, might suggest the potential role of these genes in sugar beet host resistance. The future overexpression of these candidates, and an evaluation for resistance, will demonstrate their precise roles in resistance against R. solani and other pathogens.
Gene expression in R. solani showed differential responses during pathogenic interaction with the sugar beet or combined interaction with the sugar beet and L. mesenteroides. The role of R. solani PCWDEs in fungal pathogenesis is well documented [14] and was discussed earlier in the context of the interaction with L. mesenteroides during sugar beet root infection. Besides PCWDEs, some of the other highly upregulated R. solani genes at the earliest time point (1 dpi), both during interaction with sugar beet roots singly or in combination with L. mesenteroides, include ADP, ATP carrier protein (RSOLAG2-2IIIB_02532), 40S ribosomal protein (RSOLAG2-2IIIB_00056, RSOLAG2-2IIIB_02160, RSOLAG2-2IIIB_02528), 60S ribosomal protein L33-B (RSOLAG2-2IIIB_06515), glutamine synthetase-like (RSOLAG2-2IIIB_03242), and uracil permease (RSOLAG2-2IIIB_00651), etc., (Table 3; Figure 6A). Some of them have been functionally characterized in other fungi, demonstrating their critical role in plant pathogenesis. As an example, the mutation of the ADP, ATP carrier protein (transfers ATP from mitochondria to cytoplasm) gene in Verticillium dahliae significantly reduced the fungal pathogenicity of the mutant strain and decreased disease symptoms in tobacco [33]. The highest expression of the ADP, ATP carrier protein gene (RSOLAG2-2IIIB_02532) in our case at 1 dpi will be an exciting candidate for further investigation to determine its role as a pathogenicity factor, functionally validating other uncharacterized genes that were highly expressed at this time point. Some of the highly co-expressed R. solani candidate genes identified at 1 dpi belonged to the module ME_turquoise, including 2-oxoglutarate dehydrogenase, mitochondrial (RSOLAG2-2IIIB_08262), clathrin heavy chain (RSOLAG2-2IIIB_02720), endoglucanase 5 (RSOLAG2-2IIIB_02295), and E3 ubiquitin-protein ligase HUWE1 (RSOLAG2-2IIIB_00065), etc. A few of these genes have been demonstrated to be important pathogenicity factors in other fungi. A T-DNA insertional mutation of the clathrin heavy chain (CHC) gene in Botrytis cinerea showed a significant reduction in fungal pathogenicity during the infection of the bean leaf, cucumber cotyledon, and apple fruit [34]. In another study, the inhibition of serine palmitoyltransferase activity using chemical inhibitors showed antifungal activity against Aspergillus fumigates and Rhizopus oryzae, Candida sp. [35]. Some of the highly expressed R. solani candidate genes such as RSOLAG2-2IIIB_02532, RSOLAG2-2IIIB_02720 identified here could be potential targets for future RNAi implementation in sugar beets to improve host resistance. Rhizoctonia solani genes that were further upregulated (such as RSOLAG2-2IIIB_04036 and RSOLAG2-2IIIB_01383) or downregulated (such as RSOLAG2-2IIIB_01858 and RSOLAG2-2IIIB_01965) in the presence of L. mesenteroides highlight the complex regulation of gene expression during the fungal-bacterial interaction and could be due to low pH in the infected roots resulting from the rapid multiplication of L. mesenteroides [12].
Highly expressed L. mesenteroides genes, such as ATP synthase subunit epsilon (NH16_RS02710), translation initiation factor IF-1 (NH16_RS04570), amino acid ABC transporter permease (NH16_RS02995), and DNA-directed RNA polymerase subunit alpha (NH16_RS04590) (Tables 4 and S5), suggest the importance of these genes in pathogenesis during the early sugar beet root infection stages (Figure 7). Further analysis of candidate genes belonging to the highly co-expressed gene modules is needed, such as those of ME_turquoise, which includes NH16_RS05440 (ADP-ribose pyrophosphatase), NH16_RS06070 (prolyl-dipeptidyl aminopeptidase), NH16_RS03690 (glycosyl transferase), msrA (protein repair), and mgtE (magnesium transporter), etc. Many of these genes mentioned above have been demonstrated to be important bacterial pathogenicity-related factors and could serve as potential targets for future mitigation strategies [36][37][38]. Besides the genes mentioned above, the high abundance of specific L. mesenteroides rRNA genes indicate the role of these genes during early bacterial pathogenesis [39] when rapid multiplication is one of the key events.
Pathway enrichment analysis revealed different strategies employed by sugar beets and R. solani. Genes associated with pathways primarily related to carbohydrates, amino acids, amino sugars, cell membranes, and phytohormones played important roles in sugar beet response (Figure 8). However, in R. solani, at 1 dpi the metabolic pathways were primarily associated with the ribosome, TCA cycle, and oxidative phosphorylation, etc., and at later stages (3 dpi), pathways related to the ribosome, proteasome, and phagosome, etc., played critical roles in pathogenesis (Figure 9). In L. mesenteroides, pathways related to ribosome, selenocompound metabolism and RNA polymerase were predominant during early infection stages (1 and 2 dpi; Figure S4A,B). These data presented here suggest the common role of primary metabolism in all host-plant fungal and bacterial pathogens, besides the role of unique pathways in the pathogens towards maintaining successful pathogenesis and the development of disease symptoms.

Experimental Design, Field Inoculation, and Sample Collection
Two separate studies were conducted in 2020 with 12 (Table 1)  The pathogen strains used in these studies were Leuconostoc mesenteroides strain L12311 (haplotype 11) [8] and R. solani strain F521 anastomosis group 2-2 IIIB (phylogenetic group PG2; 2). The sugar beet cultivar B-7 (Betaseed Inc.; Kimberly, ID) roots were inoculated on the shoulder of the root (where root meets the soil line) using a cork borer to create a 10 mm diameter × 24 mm deep hole. For the non-inoculated water control (mock; M), 0.1 mL of sterile well water was inserted. For the L. mesenteroides treatment, 0.1 mL of a 10 8 cfu mL −1 suspension was inoculated. The L. mesenteroides inoculum was prepared using yeast dextrose calcium carbonate agar (YDC), as described previously [3]. For R. solani treatment, a 2 × 2 mm piece of mycelial mass was placed in the hole with 0.1 mL of sterile well water. The mycelia were produced by growing the fungus for 10 days in potato dextrose broth (product no. 1.00510.0500; EMD Chemicals Inc., Gibbstown, NJ) using a shaker on the bench top at 22 • C. The mycelia had been rinsed with sterile well water prior to use. For the L. mesenteroides + R. solani treatment, 0.1 mL of a 10 8 cfu mL −1 suspension was inoculated along with the 2 × 2 mm mycelial mass. The enzymes were inoculated in a total volume of 0.2 mL at the following rates: CEL (156 units), PG (93 units), PNL (107 units), and V (4 units). The same rate was utilized for the enzyme combination treatments. When the enzymes were evaluated in combination with L. mesenteroides, the enzymes were diluted in 0.1 mL of a 10 8 cfu mL −1 of bacterial suspension instead of sterile well water. After inoculations, the plug was replaced and sealed with petroleum jelly (Unilever, Greenwich, CT). The fields were planted on 20 April, inoculated on 26 August, and evaluated on 16 October. The roots were evaluated for rot by bisecting the root through the inoculation hole and measuring the amount of rot perpendicular to the plug with a ruler. Representative samples were also photographed. To fulfill Koch's postulates, 44 isolations per study were conducted from the following treatments (4 isolations from each for the 4 different checks, 14 isolations from treatments with only water, and 14 isolations from treatments with Leuconostoc suspension) on fungal and bacterial media, as described previously [12].
For RNAseq work, sugar beet root tissues were collected at 24 h, 48 h, and 72 h post-inoculation from mock and infected samples (tissues from the area of infection), flash frozen in liquid N 2 , and stored at −80 • C until the samples were taken out for grinding in a Geno grinder at an ultralow temperature. Ground materials were stored at −80 • C and subsequently used for RNA extraction and carbohydrate analyses.

Carbohydrate Analysis
Extractions of soluble carbohydrates were performed from approximately 100 mg of finely ground sugar beet root tissues (stored in −80 • C) following the method described here. Soluble carbohydrates were extracted in 1 mL of 80% absolute ethanol at 65 • C for 30 min. Following incubation, the samples were cooled for 5 min, vortexed at a medium speed for 2 min, and centrifuged at 13,000 rpm for 8 min to pellet tissues. The supernatants were filtered using a 0.45 µm nylon syringe filter (Pall Corp., Port Washington, NY, USA). A total of 11 soluble carbohydrates were attempted to quantify using a PerkinElmer series 200 HPLC pump and autosampler (PerkinElmer Inc., Waltham, MA, USA) coupled with a Shimadzu RID-10A refractive index detector set at 30 • C (Shimadzu Scientific Instruments Inc., Columbia, MD, USA). Carbohydrates were separated in a Luna NH2 analytical column (heated to 25 • C, 250 × 4.6 mm, 5 µm, 100 Å; Phenomenex Inc., Torrance, CA, USA) surrounded by an NH2 Guard column (4 × 3 mm, Phenomenex Inc.) using an isocratic mobile phase of 80% acetonitrile at a flow rate of 2 mL min −1 . Only four carbohydrates that were detected in the sugar beet roots were sucrose, fructose, glucose + galactose, and raffinose. All carbohydrates excluding sucrose, and glucose + galactose, were quantified using a 5-point external standard curve (0.125-2 mg mL −1 ) except for sucrose, where the range of standard curve was from 2 to 18 mg mL −1 . For quantification of glucose + galactose (the two peaks did not separate), the areas and concentrations of each carbohydrate were added to create a combined standard curve. Analysis of the chromatographs and data processing were performed using PerkinElmer TotalChrom software (version 6.2.1).

Analysis of C and N
Total carbon (C) and nitrogen (N) were measured by high-temperature combustion using a VarioMax CN analyzer (Elementar Americas, Inc. Mt Laurel, NJ, USA) according to the Dumas method [40].

Data Analysis of Enzyme Application and Sugar Beet Root Rot Field Experiments
Normality of the data were assessed using the Univariate procedure, while the homogeneity of variance was determined using Levene s test in SAS (version 9.4; SAS Institute Inc., Cary, NC, USA). The general linear model procedure (Proc GLM) in SAS was used to conduct the analysis of variance. Mean comparisons were conducted using Fischer s protected least significant difference (α = 0.05). Mean comparisons across multiple treatments were conducted using single degree-of-freedom orthogonal contrast statements.

RNA Extraction, Construction of mRNA Libraries and Sequencing
Total RNA extraction was performed using the Plant/Fungi Total RNA Purification Kit (Norgen Biotek Corp, ON, Canada) according to the manufacturer's protocol. The quality and quantity of total RNA were analyzed using Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). Acceptable RIN number for samples was >7.0. High-quality RNA was used for the construction of mRNAseq libraries.
Poly(A) RNA sequencing libraries were constructed according to the protocol described in Illumina's TruSeq-stranded-mRNA sample preparation. Approximately 1 ug of total RNA was used. Ribosomal RNA depletion was performed following the method described in the Ribo-Zero™ rRNA Removal Kit (Illumina, San Diego, CA, USA). Poly(A) mRNAs were purified using oligo-(dT) magnetic beads after two rounds of purification. Poly(A) RNA fragmentation was performed using divalent cation buffer at an elevated temperature. Cleaved RNA fragments were reverse-transcribed to produce cDNAs that were subsequently used to produce U-labeled second-strand DNA. After end repair, 3 adenylation, adapter ligation, and PCR, final libraries were prepared. Quality control and quantification of the sequencing libraries were performed using Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA). Illumina's NovaSeq 6000 sequencing platform at the LC Sciences (Houston, TX, USA) was used for paired-end (150 bp) sequencing following the vendor's protocol.

Differentially Expressed mRNAs and Bioinformatics Analysis
Transcriptome data originating from different samples were merged using Perl scripts to reconstruct a comprehensive transcriptome data. StringTie [43] and edgeR [44] were used to quantify expression levels of transcripts. Expression level of mRNA expression were calculated by measuring FPKM values using the StringTie software. Differential expression of transcripts was estimated using EdgeR-R packages. Differentially expressed genes (DEGs) were identified using a cutoff p-value of < 0.05 and |log2 (fold-change)| > 1.
For annotation of transcripts, Blastx was used against the NCBI database. Gene ontology (GO) analysis was performed by BLAST searching the transcripts to the GO database and calculating gene numbers for each term. Pathway enrichment analysis was performed using the Kyoto Encyclopedia of Genes and Genomes (KEGG; [45]).
The weighted gene co-expression network analysis (WGCNA) was performed using the WGCNA package in R (3.2.2.) [46] and according to the methods described in study [47]. The parameters used for this analysis are, minimum module size (minimum module size for module detection)-30; module membership assignment (kME), minCoreKME (a number between 0 and 1. If a detected module does not have at least minCoreKMESize genes with eigengene connectivity, the module is disbanded. Its genes are unlabeled and returned to the pool of genes waiting for module detection)-0.5; min CoreKMESize (see minCoreKME)-minModuleSize/3; minKMEtoStay (genes whose eigengene connectivity to their module eigengene is lower than minKMEtoStay are removed from the module)-0.3. For each block of genes, the network is constructed. Genes are then clustered using average linkage hierarchical clustering and modules are identified in the resulting dendrogram by the Dynamic Hybrid tree cut. Found modules are trimmed of genes whose correlation with module eigengene (KME) is less than minKMEtoStay. Modules in which fewer than minCoreKMESize genes have KME higher than minCoreKME are disbanded.

Data Availability
The raw data resulting from mRNAseq (BioProject ID: PRJNA791627) were submitted to the NCBI SRA database.

Conclusions
The work presented here shows the precise roles of R. solani-derived PCWDEs, especially the PNL, PG, and CEL in sugar beet root rot that significantly increased in the presence L. mesenteroides. Amongst the different PCWDEs studied here, PNL had the most effect in increasing root rot. This observation was confirmed by using a combination of approaches including the exogenous application of PCWDEs, quantification of carbohydrates and total C/N, and global gene expression analyses in all three systems during their complex interactions. We have identified potential candidate genes and highly coexpressed gene clusters in sugar beet, R. solani, and L. mesenteroides that could potentially be critical in host defense and pathogenesis during the early root infection stages under field conditions. Future RNAi-based strategies to target some of these key host and pathogen genes identified in this work have the potential to improve sugar beet host resistance against both R. solani and L. mesenteroides.