The Arabidopsis RboHB Encoded by At1g09090 Is Important for Resistance against Nematodes

Reactive oxygen species are a byproduct of aerobic metabolic processes but are also produced by plants in defense against pathogens. In addition, they can function as signaling molecules that control various aspects of plant life, ranging from developmental processes to responses to abiotic and biotic stimuli. In plants, reactive oxygen species can be produced by respiratory burst oxidase homologues. Arabidopsis contains 10 genes for respiratory burst oxidase homologues that are involved in different aspects of plant life. Plant pathogenic cyst nematodes such as Heterodera schachtii induce a syncytium in the roots of host plants that becomes a feeding site which supplies nutrients throughout the life of the nematode. In line with this function, the transcriptome of the syncytium shows drastic changes. One of the genes that is most strongly downregulated in syncytia codes for respiratory burst oxidase homologue B. This gene is root-specific and we confirm here the downregulation in nematode feeding sites with a promoter::GUS (β-glucuronidase) line. Overexpression of this gene resulted in enhanced resistance against nematodes but also against leaf-infecting pathogens. Thus, respiratory burst oxidase homologue B has a role in resistance. The function of this gene is in contrast to respiratory burst oxidase homologues D and F, which have been found to be needed for full susceptibility of Arabidopsis to H. schachtii. However, our bioinformatic analysis did not find differences between these proteins that could account for the opposed function in the interaction with nematodes.


Introduction
Reactive oxygen species (ROS) [1,2] such as hydrogen peroxide (H 2 O 2 ), superoxide (O 2 − ) and singlet oxygen ( 1 O 2 ) are unavoidable byproducts of aerobic metabolic processes. They are highly reactive and could cause oxidative damage to DNA, proteins and other molecules of the cell. There are different cellular mechanisms in place to deactivate these damaging ROS molecules. These include enzymatic reactions through catalase, superoxide dismutase, glutathione peroxidase and ascorbate peroxidase [1] but also small antioxidants such as ascorbic acid and glutathione [3]. In addition to these metabolic byproducts, plants also use ROS directly as a defense against pathogens [4]. In this case, ROS are produced by oxidases and peroxidases to generate an apoplastic oxidative burst which The downregulation of RbohB induced by plant pathogenic nematodes indicates a possible function of RbohB in resistance. Although we did not find expression in leaves, one might thus assume that RbohB could be induced in leaves by infection with plant pathogens. We therefore infected the RbohB promoter::GUS line with Alternaria brassicicola, Botrytis cinerea and Pseudomonas syringae pv tomato DC3000. There was no GUS visible after staining at different time points after infection ( Figure  S1). We then infected the promoter::GUS line with second stage juveniles of H. schachtii (Figure 2). Already at 1 dpi, a downregulation of GUS at the infection site was visible, which became more pronounced during later stages of nematode development. Thus, these data confirmed the downregulation of RbohB from transcriptome data. The same picture emerged after infection of the promoter::GUS line with M. incognita. There was again a downregulation of GUS expression, clearly visible at 5 dpi ( Figure 3). There was no staining in galls from 7 dpi onwards.   The downregulation of RbohB induced by plant pathogenic nematodes indicates a possible function of RbohB in resistance. Although we did not find expression in leaves, one might thus assume that RbohB could be induced in leaves by infection with plant pathogens. We therefore infected the RbohB promoter::GUS line with Alternaria brassicicola, Botrytis cinerea and Pseudomonas syringae pv tomato DC3000. There was no GUS visible after staining at different time points after infection ( Figure S1).

Overexpression of RbohB Results in Enhanced Resistance against Nematodes
The downregulation in feeding sites of H. schachtii and M. incognita indicated that expression of RbohB might affect the susceptibility of Arabidopsis roots to these pathogenic nematodes. We therefore produced overexpression lines for RbohB and tested them for resistance against both nematode species. Seedlings of three independent RbohB overexpression lines and a wild type were infected with H. schachtii larvae. All three overexpression lines supported a significantly lower number of female nematodes than the wild type, while the number of male nematodes was reduced

Overexpression of RbohB Results in Enhanced Resistance against Nematodes
The downregulation in feeding sites of H. schachtii and M. incognita indicated that expression of RbohB might affect the susceptibility of Arabidopsis roots to these pathogenic nematodes. We therefore produced overexpression lines for RbohB and tested them for resistance against both nematode species. Seedlings of three independent RbohB overexpression lines and a wild type were infected with H. schachtii larvae. All three overexpression lines supported a significantly lower number of female nematodes than the wild type, while the number of male nematodes was reduced in two of the three overexpression lines ( Figure 4A). The size of syncytia was not affected by the overexpression of RbohB but the size of female nematodes was significantly smaller on all three overexpression lines as compared to the wild type ( Figure 4B). overexpression of RbohB but the size of female nematodes was significantly smaller on all three overexpression lines as compared to the wild type ( Figure 4B).
We also tested two of the overexpression lines with M. incognita. The number of galls was counted at 15 dpi and was approximately 20% lower on the roots of overexpression lines as compared to the wild type ( Figure 5).
Rboh proteins are involved in the production of ROS. We tested the ROS production in roots after infection with M. incognita larvae ( Figure 6). DAB staining showed that there was enhanced ROS production in the two overexpression lines. Since the size of syncytia in the roots of overexpression lines was not significantly different from those in wild type roots, this could indicate that the enhanced resistance of RbohB overexpression lines is due to a direct effect on the nematodes. Infection rate is shown in column sets, with asterisks indicating significant differences (**p < 0.05; ANOVA and LSD). The statistical significance was determined by three independent replicates. Values are means ± SE, n = 15. The bar shows standard error for the mean. Ns, not significant B, Size of female syncytia and female nematodes at 14 dpi. Five syncytia were selected randomly from three independent replicates (total = 15) and the size of syncytia and Infection rate is shown in column sets, with asterisks indicating significant differences (** p < 0.05; ANOVA and LSD). The statistical significance was determined by three independent replicates. Values are means ± SE, n = 15. The bar shows standard error for the mean. Ns, not significant (B), Size of female syncytia and female nematodes at 14 dpi. Five syncytia were selected randomly from three independent replicates (total = 15) and the size of syncytia and associated female nematodes were measured. Data were analyzed for significance difference using ANOVA (p < 0.05) and LSD. Values are means ± SE.
We also tested two of the overexpression lines with M. incognita. The number of galls was counted at 15 dpi and was approximately 20% lower on the roots of overexpression lines as compared to the wild type ( Figure 5).
Rboh proteins are involved in the production of ROS. We tested the ROS production in roots after infection with M. incognita larvae ( Figure 6). DAB staining showed that there was enhanced ROS production in the two overexpression lines. Since the size of syncytia in the roots of overexpression lines was not significantly different from those in wild type roots, this could indicate that the enhanced resistance of RbohB overexpression lines is due to a direct effect on the nematodes.

Overexpression of RbohB Results in Enhanced Resistance against Leaf-Infecting Pathogens
Our expression analysis has confirmed that RbohB is not expressed in leaves (with the exception of leaf primordia). However, the expression of RbohB in our overexpression construct was driven by the 35S CaMV (cauliflower mosaic virus) promoter, which is also expressed in leaves. Since other Rboh genes are involved in resistance against pathogens in the green part of the Arabidopsis plant [11], we tested whether the overexpression of RbohB would also lead to enhanced resistance against bacteria or fungi. Infection of the leaves with Botrytis cinerea spores resulted in much smaller lesions on the overexpression lines as compared to wild type plants (Figure 7). The overexpression lines were

Overexpression of RbohB Results in Enhanced Resistance against Leaf-Infecting Pathogens
Our expression analysis has confirmed that RbohB is not expressed in leaves (with the exception of leaf primordia). However, the expression of RbohB in our overexpression construct was driven by the 35S CaMV (cauliflower mosaic virus) promoter, which is also expressed in leaves. Since other Rboh genes are involved in resistance against pathogens in the green part of the Arabidopsis plant [11], we tested whether the overexpression of RbohB would also lead to enhanced resistance against bacteria or fungi. Infection of the leaves with Botrytis cinerea spores resulted in much smaller lesions on the overexpression lines as compared to wild type plants (Figure 7). The overexpression lines were also more resistant to Pseudomonas syringae (Figure 8), supporting approximately five times less bacteria after 3 days. There was also a phenotypic difference between the overexpression lines and the wild type after infection with P. syringae. The seedlings of the overexpression lines were much larger after infection and also of a darker green color.   Four independent biological replicates were used to collect the data. Data were analyzed for significance difference using ANOVA (p < 0.05) and LSD with different letters "a" and "b" indicating significant difference. Values are means ± SE. B-D: Phenotypic response of the overexpression lines showed less chlorosis and necrosis at 3 dpi as compared to control wild type.

In Silico Characterization of Arabidopsis Rboh Genes and Proteins
Our results showed that RbohB is involved in resistance against plant pathogenic nematodes. On the other hand, it has been reported that RbohD and RbohF are needed for full susceptibility of Arabidopsis to H. schachtii [19]. We have therefore conducted a bioinformatic analysis of the Arabidopsis Rboh genes to explore whether it might reveal differences between these genes/proteins that could explain their different behaviors in the interaction with nematodes.
Arabidopsis has 10 Rboh genes, with three genes on chromosome 1 and one gene on chromosome 3. No Rboh gene was found on chromosome 2, while two and four Rboh genes were mapped on chromosomes 4 and 5, respectively. RbohB and RbohF are located on chromosome 1 while RbohD was found on chromosome 5. In general, the chromosomal positioning of Rboh genes ( Figure 9) represented uneven distribution in the Arabidopsis genome. Four independent biological replicates were used to collect the data. Data were analyzed for significance difference using ANOVA (p < 0.05) and LSD with different letters "a" and "b" indicating significant difference. Values are means ± SE. (B-D): Phenotypic response of the overexpression lines showed less chlorosis and necrosis at 3 dpi as compared to control wild type.

In Silico Characterization of Arabidopsis Rboh Genes and Proteins
Our results showed that RbohB is involved in resistance against plant pathogenic nematodes. On the other hand, it has been reported that RbohD and RbohF are needed for full susceptibility of Arabidopsis to H. schachtii [19]. We have therefore conducted a bioinformatic analysis of the Arabidopsis Rboh genes to explore whether it might reveal differences between these genes/proteins that could explain their different behaviors in the interaction with nematodes.
Arabidopsis has 10 Rboh genes, with three genes on chromosome 1 and one gene on chromosome 3. No Rboh gene was found on chromosome 2, while two and four Rboh genes were mapped on chromosomes 4 and 5, respectively. RbohB and RbohF are located on chromosome 1 while RbohD was found on chromosome 5. In general, the chromosomal positioning of Rboh genes ( Figure 9) represented uneven distribution in the Arabidopsis genome.  (Table S2).
We conducted a phylogenetic analysis by using the Rboh protein sequences, which resulted in a tree with two equal groups, A and B (Figure 10), containing five genes each. RbohG, RbohC, RbohA, RbohD and RbohB were clustered into group A while RbohJ, RbohH, RbohE, RbohF and RbohI were found in group B.   (Table S2).
We conducted a phylogenetic analysis by using the Rboh protein sequences, which resulted in a tree with two equal groups, A and B (Figure 10), containing five genes each. RbohG, RbohC, RbohA, RbohD and RbohB were clustered into group A while RbohJ, RbohH, RbohE, RbohF and RbohI were found in group B.  (Table S2).
We conducted a phylogenetic analysis by using the Rboh protein sequences, which resulted in a tree with two equal groups, A and B (Figure 10), containing five genes each. RbohG, RbohC, RbohA, RbohD and RbohB were clustered into group A while RbohJ, RbohH, RbohE, RbohF and RbohI were found in group B.

Conserved Motif and Domain Analysis of Rboh Proteins
A conserved motif analysis found a total of 15 different motifs ( Figure 11). All members of both groups A and B were common in the presence of motifs 1, 2, 3, 5, 6, 7, 8, 9, 10 and 11, while prevalence of all other motifs was not found in any sequences of either group. Motif 4 was found in all members except RbohI, while motif 12 was present in group A and, in addition, in RbohJ and RbohH from group B. An uneven distribution between both groups was detected for motif 13. A partial difference that could be found between groups A and B was the presence of motifs 14 and 15, occurring only in RbohJ and RbohH of group B. All predicted motifs' widths, sites, E-values, consensus sequences and sequence logos are provided in Table S3 and Figure S2, respectively.

Conserved Motif and Domain Analysis of Rboh Proteins
A conserved motif analysis found a total of 15 different motifs ( Figure 11). All members of both groups A and B were common in the presence of motifs 1, 2, 3, 5, 6, 7, 8, 9, 10 and 11, while prevalence of all other motifs was not found in any sequences of either group. Motif 4 was found in all members except RbohI, while motif 12 was present in group A and, in addition, in RbohJ and RbohH from group B. An uneven distribution between both groups was detected for motif 13. A partial difference that could be found between groups A and B was the presence of motifs 14 and 15, occurring only in RbohJ and RbohH of group B. All predicted motifs' widths, sites, E-values, consensus sequences and sequence logos are provided in Table S3 and Figure S2, respectively. A conserved domain analysis for the occurrence of different characteristic domains in Arabidopsis Rboh proteins was performed using the InterPro online database [21]. As a result, four domains-namely NADPH_Ox, Ferric_reduct, FAD_binding_8 and NAD_binding_6-were found in all 10 Rboh proteins ( Figure S4). In addition, the EF_hand_2 motif for Ca 2+ binding was found at the N-terminus in motif 4, in close proximity to the NADPH_Ox domain, in all Rboh proteins except RbohI, in which motif 4 is missing. RbohA, RbohJ, RbohE and RbohF showed one EF_hand_2 motif while the other Rboh proteins, except RbohI, displayed this motif twice. Several transmembrane regions which anchor the protein to the plasma membrane were also detected in all proteins. These regions were mainly predicted at the start, in the middle and at the end of the Ferric_reduct domain in all polypeptides ( Figure S3). In general, the domain architecture of all Rboh proteins is very similar Figure 11. Conserved motif analyses of Rboh proteins. Phylogenetic tree is classified into group A and B, which are labeled with green and red color boxes, respectively. Motif numbers from 1 to 15 with corresponding colors (see also Figure S2) assigned by the MEME suite tool are indicated at top of each motif. Motifs' heights depict the significance of match, whereas motifs' widths display the number of amino acids in the respective motif. Characteristic domains corresponding to motifs are indicated at the top of the figure. Rboh proteins containing EF_hand_2 motifs are highlighted with orange lines on motif 4. The X-axis shows the number of amino acids.
A conserved domain analysis for the occurrence of different characteristic domains in Arabidopsis Rboh proteins was performed using the InterPro online database [21]. As a result, four domains-namely NADPH_Ox, Ferric_reduct, FAD_binding_8 and NAD_binding_6-were found in all 10 Rboh proteins ( Figure S4). In addition, the EF_hand_2 motif for Ca 2+ binding was found at the N-terminus in motif 4, in close proximity to the NADPH_Ox domain, in all Rboh proteins except RbohI, in which motif 4 is missing. RbohA, RbohJ, RbohE and RbohF showed one EF_hand_2 motif while the other Rboh proteins, except RbohI, displayed this motif twice. Several transmembrane regions which anchor the protein to the plasma membrane were also detected in all proteins. These regions were mainly predicted at the start, in the middle and at the end of the Ferric_reduct domain in all polypeptides ( Figure S3). In general, the domain architecture of all Rboh proteins is very similar ( Figure S4) and could not explain the different functions of RbohB and RbohD/RbohF in relation to H. schachtii.
The majority of conserved motifs corresponded to characteristic domains of Rboh family proteins. Motifs 9 and 10 were located in the region of the NADPH_Ox domain in all proteins and EF_hand_2 motif was represented by motif 4. Motifs 3, 6, 7, 8 and 11 were found in the Ferric_reductase domain. FAD_binding_8 and NAD_binding_6 domains were represented by motifs 2 and 1, respectively. Transmembrane regions (TM) were mostly found at the N-terminus of motif 3.
Our comparison of motifs and domains did not reveal a major difference between RbohB and RbohD/RbohF. We therefore aligned the protein sequences for these ( Figure S5). This clearly showed major differences at the N-terminus, with several larger insertions/deletions. RbohD had three small insertions at the N-terminus compared to RbohB and RbohF. Following these are an insertion in RbohD and RbohF and two insertions in RbohF compared to RbohB and RbohD. There were only small differences in the remaining parts of the proteins, mostly single amino acid changes. The only exception was a 10 amino acid insertion in RbohF, approximately 150 amino acids from the C-terminus.

Expression of RbohB
Our expression analysis using a promoter::GUS line found that RbohB is expressed in roots, which is confirmed by data from Genevestigator [22], as shown in Figure S6. The only exception that we found was expression in leaf primordia. The expression in roots was downregulated in feeding sites induced by plant pathogenic nematodes H. schachtii and M. incognita. Data from a transcriptome analysis of syncytia induced by H. schachtii [20] confirm this and show in addition that most Rboh genes are actually downregulated in syncytia. Downregulation in response to infection with M. incognita is confirmed by data from Genevestigator ( Figure S7). There is no indication that pathogen infection of Arabidopsis leaves leads to induction of RbohB ( Figure S7) and infection of the promoter::GUS line with A. brassicicola, B. cinerea and P. syringae also did not result in any GUS expression, as revealed by staining.
There is currently only one report about Arabidopsis RbohB [12]. The authors reported a role in seed after ripening by comparing an RbohB mutant with a wild type. They found that Arabidopsis RbohB is a major producer of ROS in germinating seeds. Accordingly, they detected the RbohB mRNA in dry seeds and 24 h imbibed seeds using RT-PCR; however, the band from dry seeds was very weak. This might explain why we did not detect expression in seeds with our GUS line. The expression in imbibed seeds is probably due to expression in the emerging root.
Compared to RbohB, RbohD and to a lesser extent RbohF were not only expressed in roots but also in aerial tissues. In line with a role in plant immunity, RbohD was also induced by different PAMPs (pathogen-associated molecular patterns), flagellin, ELF18 and chitin. Both genes were also induced, although at different levels and in different tissues, after inoculating the leaves with pathogenic bacteria and fungi [23]. For RbohB, there are only GeneChip data available which show that this gene is not induced in response to plant pathogenic bacteria and fungi ( Figure S7), as we have also shown here. In response to PAMPs, there might be a very weak induction in response to flagellin and Pep2, but this has to be confirmed ( Figure S8). Thus, RbohB and RbohD/RbohF respond differently in response to plant pathogenic bacteria and fungi as well as to plant pathogenic nematodes. In the case of the latter, while RbohB is strongly downregulated in syncytia induced by H. schachtii, RbohD expression remains at a high level and RbohF is only slightly downregulated (Table S1) [20].

Enhanced Resistance of RbohB Overexpression Lines
The strong downregulation of RbohB in syncytia indicated that RbohB expression might be detrimental to nematode development. We tested this with overexpression lines. We used the CaMV promoter to drive expression of RbohB because this promoter is active in most plant tissues.
A disadvantage of the CaMV promoter is the downregulation in syncytia [24,25]. However, it is still active in the early stages of syncytium development and the possibility to test also against leaf infecting pathogens outweighs this disadvantage [26].
We tested three independent overexpression lines with H. schachtii. All three lines supported significantly less female nematodes compared to the wild type, and two of them also supported significantly less male nematodes. The size of syncytia developing on all three overexpression lines seemed to be smaller but the effect was not significant. On the other hand, the size of female nematodes on all three overexpression lines was significantly smaller than on wild type roots. Thus, the overexpression of RbohB might have a direct effect on the nematodes through the production of ROS. This is supported by the enhanced ROS production that we found on the overexpression lines infected with M. incognita. A RbohB mutant was recently tested by Siddique et al. [19]. They found that there was no significant difference between the mutant and wild type in the number of nematodes at 14 dpi. There was also no difference in the size of the female nematodes and the size of syncytia induced by female nematodes. The reason that the mutant did not show an effect is that the expression of RbohB in syncytia, as shown here and by Szakasits et al. [20], is fast and strongly downregulated.
We also infected two of the overexpression lines with M. incognita. On both overexpression lines, the number of galls was significantly reduced compared to the wild type. In addition to the root infecting nematodes, we tested also leaf infecting pathogens. Although there is no evidence that RbohB might be expressed in leaves in response to pathogen infection, the overexpression lines were also more resistant to B. cinerea and P. syringae. Thus, RbohB has an effect against different kinds of plant pathogens. Although our data indicate a direct effect of the produced ROS, we also cannot exclude that there is an indirect effect via ROS acting as signal for the induction of, for instance, PR genes [27]. A transcriptome analysis of the overexpression lines might indicate whether this is the case.

Opposite Effects of RbohB and RbohD/RbohF on H. schachtii
Our results clearly show that RbohB is downregulated in syncytia to suppress the negative effect on syncytium development, as shown by overexpression. On the other hand, it has been shown that RbohD and to a lesser extent RbohF are needed for full susceptibility of Arabidopsis plants to H. schachtii [19]. How can these opposing effects be explained? We compared the Rboh genes and proteins. A phylogenetic analysis placed the Rboh genes in two groups, having RbohB and RbohD in group A and RbohF in group B. Our bioinformatics analysis of the Rboh proteins found that all 10 Arabidopsis proteins are very similar. All contain the same set of functional domains. There are small differences in the transmembrane domains and the EF_hand-2 domains. However, RbohB and RbohD both have two EF_hand-2 domains while RbohF has only one. Thus, this analysis could not explain the different effects on H. schachtii. We then aligned the protein sequences of RbohB, RbohD and RbohF ( Figure S5). This clearly showed major differences (insertions/deletions) at the N-terminus, but whether these can explain the different functions is not clear. It would be necessary to test this with domain swap experiments.
Given that there are only small differences between all Arabidopsis Rboh proteins, the different functions might be explained by the different expression levels of the Rboh genes. However, this cannot explain the different functions of RbohB and RbohD/RbohF on H. schachtii. Expression of RbohB decreases susceptibility while expression of RbohD and RbohF increases susceptibility to H. schachtii infection. This leaves the possibility that these Rboh proteins cooperate with different interaction partners to achieve their different functions. Indeed, a recent in silico analysis of the Arabidopsis Rboh genes [28] has indicated different pathways with different interacting partners, which could explain the different functions of RbohB and RbohD/F in interaction with H. schachtii. RbohD and RbohF together with RbohA share a set of six common functional partners, all of them kinases, which are not involved with any of the other Rboh proteins. In line with the large number of potentially interacting kinases, there was a large number of potential phosphorylation sites identified, most of them in RbohD. Moreover, most potential phosphorylation sites are located at the N-terminus. Given the differences in protein sequence at the N-terminus, phosphorylation of RbohD and RbohF could explain the importance of susceptibility against H. schachtii in contrast to RbohB. While nothing is known about RbohA, which is downregulated in syncytia (Table S1), RbohD and RbohF are for instance both involved in pathogen defense [15,29], jasmonic acid induced gene expression [30], salinity stress response [16] and stomatal closure [31]. Furthermore, as has already been mentioned, RbohD and RbohF are both important for susceptibility to H. schachtii [19].
The remaining Rboh genes seem to have specific and different functions since they have mostly specific interaction partners [28]. No function has been reported for RbohG. RbohB is involved in seed after-ripening [12], RbohC in root hair tip growth [32], RbohE in tapetum cell death [33] and RbohI in drought stress response in seeds and roots [34]. While the aforementioned genes have specific functions, which might be explained by their specific interaction partners, two other genes, RbohH and RbohJ, have a redundant function. Both are involved in polar root hair growth and pollen tube tip growth [35] while having only one common interacting partner, which, however, is also found as the interacting partner for several other Rboh genes, including RbohB.
While RbohD and RbohF share this set of potential kinase partners, the bioinformatics analysis of Kaur and Pati [28] identified only one specific interaction partner for each of them. Contrastingly, for RbohB, there were five unique functional partners that were not found for any of the other Rboh proteins. They all have different functions and none of them are kinases. However, the CPK6 kinase was identified as a functional partner for RbohB, RbohD and RbohF. Whether one or more of the unique functional partners of RbohB might be responsible for the resistance function of RbohB is not known and would have to be tested experimentally.
It could also be the case that the activity of RbohB and RbohD/RbohF leads to different ROS species, which could result in the observed effects on nematode interaction. However, to our knowledge, nothing is known about this possibility. Kaya et al. [36] have studied the ROS-producing activity of different Arabidopsis Rboh enzymes in a bacterial expression system. They found that RbohB had the lowest activity while RbohH and RbohJ had the highest activity, about 100-fold that of RbohB. RbohD and RbohF were somewhere in the middle. Whether these differences in activity between RbohB on the one side and RbohD and RbohF on the other side can account for the different behavior in response to H. schachtii is not clear. It has also been taken into account that all Arabidopsis Rboh enzymes are activated by phosphorylation [37].

Cloning of Binary Vectors
We used the vector pMAA-Red [38] for overexpression analysis and GUS analysis. For cloning of the overexpression vector, the RbohB coding sequence was amplified from cDNA by PCR. The internal NcoI site was eliminated as follows. The coding sequence was amplified in two parts using primers At1g09090forNco and At1g09090Mrev and At1g09090Mfor and At1g09090revBam, respectively, containing the restriction sites NcoI and BamHI (Table S4). An overlap PCR was then done with both fragments and primers At1g09090forNco and At1g09090revBam. The final PCR fragment was digested with NcoI and BamHI and ligated into the vector pMAA-Red, digested with the same enzymes. The final construct was confirmed by sequencing.
For cloning of a promoter::GUS fusion, we amplified the promoter region by PCR, using as a template Arabidopsis genomic DNA. Primers pRBOHBforKpn and pRBOHBrevNco (Table S4) included restriction sites for KpnI and NcoI, respectively. The PCR fragment was digested with KpnI and NcoI and ligated to the large vector fragment of pMAA-Red, digested with the same enzymes, thus replacing the 35S promoter with the RbohB promoter. The construct was verified by sequencing.

Plant Material and Growth Conditions
Arabidopsis seeds (ecotype Columbia) were surface sterilized for 20 min in 6% (w/v) sodium hypochlorite and subsequently washed three times with sterile water for in vitro growth on either Murashige and Skoog (MS) medium or Knop medium [39]. For seed production, Arabidopsis plants were cultivated on soil in a growth chamber at 25 • C in a 16 h light and 8 h dark cycle.

Arabidopsis Transformation
Binary vectors were introduced into Agrobacterium tumefaciens GV3101 by a freeze-thaw method [40] for transformation of Arabidopsis plants by a modified floral dip method [41]. Transformed seeds were identified as described by [38] and transferred to soil for seed production.
For the promoter::GUS construct, 12 independent transgenic plants were generated and tested for GUS activity to choose a representative line, which was grown further to produce homozygous seeds. For overexpression lines, 12 independent transgenic T2 lines were generated and applied to RT-PCR using the primers described in Table S4 to select the best expressing lines. These were then made homozygous for resistance tests.

Resistance Tests with H. schachtii and M. incognita
H. schachtii was multiplied on mustard (Sinapis alba cv. Albatros) roots in vitro under sterile conditions. Cysts were collected from the mustard stock cultures and J2 larvae were hatched by soaking the cysts in 3 mM ZnCl 2 . Larvae were washed three times in sterile water. They were resuspended in 0.5% (w/v) gelrite (Duchefa, Haarlem, The Netherlands) for infection of Arabidopsis roots.
Arabidopsis seedlings were grown on 0.2 concentrated Knop medium supplemented with 2% sucrose [39] for 12 days. Roots were then inoculated with approximately 50-60 J2 larvae per plant. Three independent experiments were performed with 5 Petri dishes each, with one Petri dish containing approximately 10 seedlings. Female and male nematodes were counted at 14 dpi and the number of males and females per cm of root length was calculated. At 15 dpi, female nematodes and syncytia associated with female nematodes were photographed using an inverse microscope (Axiovert 200M; Zeiss, Hallerbergmoos, Germany) with an integrated camera (AxioCam MRc5; Zeiss, Hallerbergmoos, Germany) and measured according to [42].
M. incognita egg masses were harvested from sterile stock cultures propagated on cucumber roots which were grown on B5 agar medium. The hatching of second-stage juveniles (J 2 ) of M. incognita was stimulated by soaking the egg masses in sterile water for 2-4 days at room temperature in the dark. Freshly hatched J 2 larvae were sterilized by incubation in sterilization solution (0.002% HgCl 2 ) for 5-10 min, followed by 3 to 4 washings with sterile distilled water. After sterilization, the J 2 larvae were suspended in 0.7% gelrite and 12-days-old Arabidopsis plants growing on Knop medium were inoculated with these larvae. The number of galls was counted at 14 dpi.

Pseudomonas Syringae Infection Assay
The infection assay was conducted according to [43], using the strain P. syringae pv tomato DC3000. Arabidopsis plants were grown on half strength MS medium with 0.3% phytagel in a growth chamber at 24 • C with a 12 h light/12 h dark photoperiod for 2 weeks. P. syringae was cultured on King's B medium, and for inoculation, the cells were suspended in 10 mM MgCl 2 at a concentration of 5 × 10 5 CFU/mL. Arabidopsis seedlings were inoculated by flooding the plate with the P. syringae suspension until the seedlings were completely submerged in the inoculum. The P. syringae suspension was removed from the plates and the plates with the inoculated seedlings were put back into the growth chamber. One hour after the inoculation and at 3 dpi, 4 seedlings (only aerial parts) of each line were removed and surface sterilized with 5% H 2 O 2 for 3 min and washed three times with sterile distilled water. The pooled sample of 4 seedlings was homogenized in 10 mL sterile distilled water using a mortar and pestle. Diluted samples were plated onto King's B medium containing rifampicin (50 µg/mL) and colonies were counted after 24 h using proper diluted samples. Data were normalized as CFU/mg using the total weight of inoculated seedlings. Bacterial numbers were evaluated in three independent experiments.

Resistance Test against Botrytis Cinerea
Arabidopsis plants were grown on soil under short day conditions (8 h light, 16 h dark). B. cinerea was grown on potato dextrose media (38 g/liter potato dextrose agar). After two weeks, the spores of the fungus were harvested in sterile water and passed through sterile cotton cloth. Spores were counted with a hemocytometer (Thoma, area 0.0025 mm 2 , depth 0.1 mm, LO-Laboroptik, Lancing, UK). The spore suspension was adjusted to 10 5 spores/mL using 1 2 strength potato dextrose broth. Drops of 5 µL spore suspension were deposited on the upper surfaces of 4-week-old plants on the left or right side to ensure uniformity across all leaves in the visual observations. The plants were covered with plastic coverings to maintain high humidity. The lesion diameter was determined with the help of a meter rod at 5 dpi and a photo of each leaf was taken with a camera (Digital Camera Nikon Coolpix S8200, Nikon, Japan). For each line, 3 plants were used and the experiment was performed in triplicate.

Semi-Quantitative RT-PCR of Overexpression Lines
Total RNA was purified from seedlings using the "NucleoSpin ® RNA Plant" from Macherey-Nagel (Fisher Scientific, Vienna, Austria). A volume of 2 µL of eluted RNA was used for photometric measurement of RNA using NanoDrop (NanoDrop™ 2000c, PEQLAB). Semi-quantitative RT-PCR was conducted using the "One-step Master Mix RT-PCR Kit" (Affymetrix, Santa Clara, California, USA) according to the manufacturer's instructions ( Figure S9). Primers are listed in Table S4.

GUS Reporter Analysis
Histochemical detection of GUS activity was performed according to [44], as described by [45], using X-gluc (Biomol, Hamburg, Germany) in 0.1 M sodium phosphate buffer pH 7.0, 0.1% Triton-X 100, 0.5 mM K 3 [Fe(CN) 6 ], 0.5 mM K 4 [Fe(CN) 6 ] and 10 mM Na 2 EDTA. Plant parts were incubated in X-gluc solution at 37 • C. After staining, chlorophyll was removed from photosynthetic tissues with 70% (v/v) ethanol. The seedlings and different plant parts were stained before and after flowering. Similarly, for GUS staining of syncytia and galls, the infected roots of promoter::GUS plants were incubated with X-gluc at 37 • C. The staining in syncytia, galls and uninfected roots was performed at different time points and photographed under an inverse microscope (Axiovert 200M; Zeiss, Hallerbergmoos, Germany) with an integrated camera (AxioCam MRc5; Zeiss, Hallerbergmoos, Germany).

Sequence Retrieval and Characterization of Rboh Family Genes
Genomic DNA, coding (CDs) and protein sequences of all Rboh genes of Arabidopsis thaliana were retrieved from the Arabidopsis Information Resource (TAIR) website (https://www. arabidopsis.org/) in FASTA format. All other characters of Rboh genes mentioned in Table S2, such as chromosomal number, strand type, number of exons, length of protein, coding (CDs) and genomic DNA (gDNA), gene location on chromosome and isoelectric point (pI) of protein, were also taken from the TAIR website. Grand average of hydropathy (GRAVY) of Rboh proteins was calculated using the Sequence Manipulation Suite online website (https://www.bioinformatics.org/sms 2/protein gravy.html). Signal peptide cleavage site (SPCS) was predicted through SignalP 4.1 Server (http://www.cbs.dtu.dk/services/ SignalP-4.1/).

Chromosomal Mapping and Gene Structure Analyses of Rboh Genes
Chromosomal positioning of 10 Arabidopsis Rboh genes was performed to display their genome-wide distribution. For positioning of the genes on their respective chromosomes, MapChart [46] desktop-based software was used with default parameters. Gene structure analysis was performed to analyze the intron-exon pattern of 10 Rboh genes. For this purpose, online Gene Structure Display Server 2.0 (http://gsds.gao-lab.org/) (Center for Bioinformatics (CBI), Peking University, Beijing, China) was used [47]. Genomic DNA and coding (CDs) sequences of all genes were arranged according to the groups formed by the phylogenetic analysis. Thereafter, arranged gDNA and CD sequences were used to map the untranslated regions, number of exons and number of introns in all sequences.

Phylogenetic Analysis
An evolutionary tree of 10 Rboh protein sequences was constructed using Molecular Evolutionary Genetic Analysis software (MEGA version 7.0, Penn State University, State College, Pennsylvania, USA) [48]. First, all protein sequences were subjected to alignment through the MUSCLE algorithm with default parameters such as gap opening penalty -2.9, gap extension penalty 0 and hydrophobicity multiplier 1.2, and the UPGMBA (Unweighted Pair Group Method with arithmetic mean) clustering method was used. Thereafter, using aligned data, the evolutionary history was inferred using the maximum likelihood method based on the JTT (Jones-Taylor-Thornton) matrix-based model (MEGA version 7.0, Penn State University, State College, Pennsylvania, USA) [49]. Furthermore, initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model and then selecting the topology with superior log likelihood value. From a total of 780 positions, the tree with the highest log likelihood value of -11668.24 was developed using 1000 bootstrap replications, partial gap deletion and 95% site coverage cutoff value.

Conserved Motif and Domain Analysis
To discover conserved motifs in Rboh protein sequences, conserved motif analysis was performed using the online (http://meme-suite.org/) MEME SUITE tool [50]. To display the conserved domains through identified motifs, different parameters were assessed one by one for motif discovery; however, the maximum number of motifs of 15, minimum motif width of 20 and maximum motif width of 220 were finally used. All protein sequences of Rboh genes were arranged according to their clustering in the phylogenetic tree. Conserved domain analysis was performed using the online InterPro database (https://www.ebi.ac.uk/interpro/) with default parameters.

Statistical Analysis
Data regarding the number of nematodes, number of galls, size of female nematodes and syncytia, number of CFU and lesion diameter were analyzed using single factor ANOVA (p < 0.05), while the means comparisons were performed using the least significant difference (LSD) test at a 95% level of confidence.

Conclusions
We have shown that RbohB is a root-specific gene which is strongly downregulated in syncytia induced by H. schachtii in Arabidopsis roots. Overexpression resulted in resistance against different pathogens. In contrast, expression of RbohD/RbohF in syncytia is needed for full susceptibility to H. schachtii. Since the Rboh proteins are very similar, the different behavior of RbohB and RbohD/RbohF might be due to different interaction partners.