Nucleotide-Binding Leucine-Rich Repeat Genes CsRSF1 and CsRSF2 Are Positive Modulators in the Cucumis sativus Defense Response to Sphaerotheca fuliginea

Cucumber powdery mildew caused by Sphaerotheca fuliginea is a leaf disease that seriously affects cucumber’s yield and quality. This study aimed to report two nucleotide-binding site-leucine-rich repeats (NBS-LRR) genes CsRSF1 and CsRSF2, which participated in regulating the resistance of cucumber to S. fuliginea. The subcellular localization showed that the CsRSF1 protein was localized in the nucleus, cytoplasm, and cell membrane, while the CsRSF2 protein was localized in the cell membrane and cytoplasm. In addition, the transcript levels of CsRSF1 and CsRSF2 were different between resistant and susceptible cultivars after treatment with exogenous substances, such as abscisic acid (ABA), methyl jasmonate (MeJA), salicylic acid (SA), ethephon (ETH), gibberellin (GA) and hydrogen peroxide (H2O2). The expression analysis showed that the transcript levels of CsRSF1 and CsRSF2 were correlated with plant defense response against S. fuliginea. Moreover, the silencing of CsRSF1 and CsRSF2 impaired host resistance to S. fuliginea, but CsRSF1 and CsRSF2 overexpression improved resistance to S. fuliginea in cucumber. These results showed that CsRSF1 and CsRSF2 genes positively contributed to the resistance of cucumber to S. fuliginea. At the same time, CsRSF1 and CsRSF2 genes could also regulate the expression of defense-related genes. The findings of this study might help enhance the resistance of cucumber to S. fuliginea.


Introduction
Cucumber (Cucumis sativus) is one of the largest protected vegetable cultivation in China. Cucumber powdery mildew caused by S. fuliginea is an important disease of cucumber that seriously affects cucumber's yield and quality. Cucumber powdery mildew has some characteristics, such as wide distribution, short incubation period and so on [1]. It reduces cucumber yield by 20%-50%, or even 0% [2]. At present, planting resistant cultivars is the most important and effective way of controlling cucumber powdery mildew. It is vital to clone key genes related to the infection of S. fuliginea and explore its mechanism of action so as to improve the resistance of cucumber to S. fuliginea.
Producing toxins to destroy the host's defense system and make plants susceptible to diseases is one way in which pathogens infect plants [3]. To control pathogens, plants have formed an innate immune system [4,5]. This immune system has two layers: the first layer is called pathogen-associated molecular patterns (PAMPs)-triggered immunity (PTI), which is the recognition of PAMPs by pattern-recognition receptors on the cell surface; the second layer is called effector-triggered immunity (ETI), which is the recognition of pathogenic factors within the bacteria. This process mainly triggers a hypersensitive response (HR), CsRSF1/CsRSF2-overexpressing cucumber plants indicated that CsRSF1/CsRSF2, acting as positive regulators, were required for host resistance response to S. fuliginea.

Gene Cloning and Sequence Analysis of CsRSF1 and CsRSF2
According to the sequence of the target gene found in the Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html, last accessed on 9 April 2021), the fulllength CDS sequences of CsRSF1/CsRSF2 were cloned from the experimental materials of B21-a-2-1-2 and B21-a-2-2-2. The comparison of the sequences showed no difference in CDS sequences of CsRSF1/CsRSF2 between the B21-a-2-1-2 variety and B21-a-2-2-2 variety. The full CDS sequence of CsRSF1 was 2457 bp, which encoded 818 amino acid (AA) residues with the molecular weight of 93.59 kDa and the theoretical iso-electric point (pI) of 6.05 ( Figure S1). The full CDS sequence of CsRSF2 was 2466 bp, which encoded 821 AA residues with a molecular weight of 94.25 kDa and a theoretical pI of 6.58 ( Figure S2). In addition, the chromosomal distribution showed that CsRSF1 was located on chromosome 6 (scaffold00927), and CsRSF2 was located on chromosome 4 (scaffold01000) ( Figure S3).
The analysis of the protein sequence showed that CsRSF1 contained an N-terminal RPW8 domain, an NB-ARC domain, and an LRR domain at the C-terminus ( Figure S4A). Similarly, CsRSF2 also had an N-terminal RPW8 domain, an NB-ARC domain, and an LRR domain at the C-terminus ( Figure S4B).

Subcellular Localization and Expression Patterns of CsRSF1 and CsRSF2
Bioinformatics prediction indicated that the CsRSF1 protein was localized in the cytoplasm and cell membrane, and the CsRSF2 protein was localized in the cytoplasm. To confirm this, the p35S:GFP-CsRSF1/CsRSF2 fusion vector was constructed ( Figure 1A), and the p35S:GFP-CsRSF1/CsRSF2 and the control p35S:GFP vectors were separately introduced into tobacco leaves. Confocal microscopic examination showed that CsRSF1 and CsRSF2 proteins were both located in the cytoplasm and cell membrane. In addition, the CsRSF1 protein was also located in the nucleus ( Figure 1B).
The expression patterns of CsRSF1 and CsRSF2 in different organs (leaves, cotyledons, stems and roots) of susceptible and resistant cultivars were analyzed by qRT-PCR ( Figure 2). The results showed the expression patterns of CsRSF1 and CsRSF2 were different in various organs in the resistant and susceptible cultivars. Among the resistant cultivars, CsRSF1 exhibited the highest expression in roots, followed by leaves and cotyledons, but it was weakly expressed in stems ( Figure 2A). Nevertheless, CsRSF1 in the susceptible cultivar was relatively highly expressed in leaves, followed by stems, roots, and cotyledons ( Figure 2B). Moreover, CsRSF2 in the resistant cultivars was significantly expressed in cotyledons but barely expressed in stems, leaves, and roots ( Figure 2C). In the susceptible cultivars, CsRSF2 was significantly expressed in cotyledons and stems, followed by leaves and roots ( Figure 2D). This result indicated CsRSF1 and CsRSF2 genes showed varied expression patterns among organs in cucumber. The resistance of CsRSF1 and CsRSF2 may be different in various organs in the resistant and susceptible plants.

Response of CsRSF1 and CsRSF2 to Exogenous Substances
The transcript levels of CsRSF1 and CsRSF2 were determined in the susceptible and resistant cultivars at the two-leaf-one-heart stage after exogenous treatment with ABA, MeJA, SA, ETH, GA, and H 2 O 2 to investigate the expression patterns of CsRSF1 and CsRSF2 genes in response to exogenous substances. The result indicated that the regulation of CsRSF1 and CsRSF2 was different between resistant and susceptible varieties of cucumber after exogenous treatment with MeJA, SA, ETH, and H 2 O 2 . In addition, CsRSF1 and CsRSF2 may be involved in phytohormone-related signaling pathways ( Figure 3). Compared with the control plants, the levels of CsRSF1 transcripts showed an upward trend at first and then a downward trend, which reached the highest level 12 h after ABA treatment in the resistant and susceptible cultivars ( Figure 3A,B). The levels of CsRSF2 transcripts decreased 12 h, 24 h, and 48 h after GA treatment in the resistant and susceptible cultivars ( Figure 3C,D). Therefore, the results indicated that CsRSF1 might be involved in the ABA signaling pathway, while CsRSF2 might be involved in the GA signaling pathway. The expression patterns of CsRSF1 and CsRSF2 in different organs (leaves, cotyledons, stems and roots) of susceptible and resistant cultivars were analyzed by qRT-PCR ( Figure 2). The results showed the expression patterns of CsRSF1 and CsRSF2 were different in various organs in the resistant and susceptible cultivars. Among the resistant cultivars, CsRSF1 exhibited the highest expression in roots, followed by leaves and cotyledons, but it was weakly expressed in stems ( Figure 2A). Nevertheless, CsRSF1 in the susceptible cultivar was relatively highly expressed in leaves, followed by stems, roots, and cotyledons ( Figure 2B). Moreover, CsRSF2 in the resistant cultivars was significantly expressed in cotyledons but barely expressed in stems, leaves, and roots ( Figure 2C). In the susceptible cultivars, CsRSF2 was significantly expressed in cotyledons and stems, followed by leaves and roots ( Figure 2D). This result indicated CsRSF1 and CsRSF2 genes showed varied expression patterns among organs in cucumber. The resistance of CsRSF1 and CsRSF2 may be different in various organs in the resistant and susceptible plants.

Expression Patterns of CsRSF1 and CsRSF2 in the Resistant/Susceptible Cucumber Cultivars after S. fuliginea Inoculation
qRT-PCR was used to detect the transcription levels of CsRSF1 and CsRSF2 in the susceptible and resistant cultivars after inoculation with S. fuliginea ( Figure 4). The comparison of the two cultivars showed that CsRSF1 expression was induced at 6 h and peaked at 12 h in the susceptible cultivar. In the resistant cultivar, CsRSF1 expression was largely induced at 3 and 6 h. The CsRSF1 expression in the resistant cultivars were higher than that in the susceptible cultivars between 0 h and 9 h ( Figure 4A). However, the CsRSF2 expression reached a peak at 3 h in the susceptible cultivar. CsRSF2 expression increased in the initial phase and reached a peak at 6 h in the resistant cultivar ( Figure 4B). This finding suggested that CsRSF1 and CsRSF2 were induced upon S. fuliginea infection in the two cultivars. The expression level in the leaf was normalized as 1. Expression analysis of candidate genes using the 2 −ΔΔCt method. Data are means ± SD from three biological replicates per cultivar. Different letters (a,b) above the bars indicate a significant difference determined by Student's t-test ( p < 0.05).

Response of CsRSF1 and CsRSF2 to Exogenous Substances
The transcript levels of CsRSF1 and CsRSF2 were determined in the susceptible and resistant cultivars at the two-leaf-one-heart stage after exogenous treatment with ABA, MeJA, SA, ETH, GA, and H2O2 to investigate the expression patterns of CsRSF1 and CsRSF2 genes in response to exogenous substances. The result indicated that the regulation of CsRSF1 and CsRSF2 was different between resistant and susceptible varieties of cucumber after exogenous treatment with MeJA, SA, ETH, and H2O2. In addition, CsRSF1 and CsRSF2 may be involved in phytohormone-related signaling pathways ( Figure 3). Compared with the control plants, the levels of CsRSF1 transcripts showed an upward trend at first and then a downward trend, which reached the highest level 12 h after ABA treatment in the resistant and susceptible cultivars ( Figure 3A,B). The levels of CsRSF2 transcripts decreased 12 h, 24 h, and 48 h after GA treatment in the resistant and susceptible cultivars ( Figure 3C,D). Therefore, the results indicated that CsRSF1 might be involved in the ABA signaling pathway, while CsRSF2 might be involved in the GA signaling pathway.   were compared with the mock control, which was set to 1. Expression analysis of candidate genes using the 2 −ΔΔCt method.
Data are means ± standard deviations (SD) from three biological replicates per cultivar. The asterisks indicated a significant difference (Student's t-test, * p < 0.05 or ** p < 0.01).

Expression Patterns of CsRSF1 and CsRSF2 in the Resistant/Susceptible Cucumber Cultivars after S. fuliginea Inoculation
qRT-PCR was used to detect the transcription levels of CsRSF1 and CsRSF2 in the susceptible and resistant cultivars after inoculation with S. fuliginea ( Figure 4). The comparison of the two cultivars showed that CsRSF1 expression was induced at 6 h and peaked at 12 h in the susceptible cultivar. In the resistant cultivar, CsRSF1 expression was largely induced at 3 and 6 h. The CsRSF1 expression in the resistant cultivars were higher than that in the susceptible cultivars between 0 h and 9 h ( Figure 4A). However, the CsRSF2 expression reached a peak at 3 h in the susceptible cultivar. CsRSF2 expression increased in the initial phase and reached a peak at 6 h in the resistant cultivar ( Figure 4B). This finding suggested that CsRSF1 and CsRSF2 were induced upon S. fuliginea infection in the two cultivars.  Data are means ± standard deviations (SD) from three biological replicates per cultivar. The asterisks indicated a significant difference (Student's t-test, * p < 0.05 or ** p < 0.01).

Silencing of CsRSF1 and CsRSF2 Impairs Host Resistance to S. fuliginea
TRV-based virus-induced gene silencing (VIGS) was used to knock down CsRSF1/CsRSF2 transcripts in Xintaimici to explore whether CsRSF1/CsRSF2 was required for resistance of cucumber to S. fuliginea ( Figure 5). A 3′-terminal fragment specific in CsRSF1 and a 5′-terminal fragment specific in CsRSF2 were inserted in the antisense orientation into the vector pTRV2, and the pTRV2:CsRSF1/CsRSF2 recombinant plasmids were generated ( Figure 5A). Subsequently, TRV:CsRSF1/CsRSF2 viruses were injected into cucumber cotyledons. After inoculation with TRV:CsRSF1 or TRV: 00 (as a control) viruses for 5 days and TRV:CsRSF2 or TRV: 00 (as a control) viruses for 7 days, the mild chlorotic mosaic symptoms of TRV appeared in the cotyledons of the infected Xintaimici plants ( Figure 5B). The TRV: 00 infections did not significantly affect the expression of CsRSF1/CsRSF2 in Xintaimici plants. Meanwhile, the transcript level of CsRSF1/CsRSF2 was markedly reduced 0.5-0.75/0.2-0.5 in TRV:CsRSF1/CsRSF2-infected (CsRSF1/CsRSF2silencing) plants ( Figure 5C). These results indicated that CsRSF1 and CsRSF2 were successfully transiently silenced in cucumber cotyledons. The expression of CsRSF2 in resistant and susceptible varieties were inoculated with S. fuliginea. Expression analysis of candidate genes at 0, 3, 6, 9, 12 and 24 hpi (hours post-inoculation) using the 2 −∆∆Ct method. The expression level in B21-a-2-2-2 was the mock control as 1. Data are means ± standard deviations (SD) from three biological replicates per cultivar. The asterisks indicated a significant difference (Student's t-test, * p < 0.05 or ** p < 0.01).

Silencing of CsRSF1 and CsRSF2 Impairs Host Resistance to S. fuliginea
TRV-based virus-induced gene silencing (VIGS) was used to knock down CsRSF1/ CsRSF2 transcripts in Xintaimici to explore whether CsRSF1/CsRSF2 was required for resistance of cucumber to S. fuliginea ( Figure 5). A 3 -terminal fragment specific in CsRSF1 and a 5 -terminal fragment specific in CsRSF2 were inserted in the antisense orientation into the vector pTRV2, and the pTRV2:CsRSF1/CsRSF2 recombinant plasmids were generated ( Figure 5A). Subsequently, TRV:CsRSF1/CsRSF2 viruses were injected into cucumber cotyledons. After inoculation with TRV:CsRSF1 or TRV: 00 (as a control) viruses for 5 days and TRV:CsRSF2 or TRV: 00 (as a control) viruses for 7 days, the mild chlorotic mosaic symptoms of TRV appeared in the cotyledons of the infected Xintaimici plants ( Figure 5B).
The TRV: 00 infections did not significantly affect the expression of CsRSF1/CsRSF2 in Xintaimici plants. Meanwhile, the transcript level of CsRSF1/CsRSF2 was markedly reduced 0.5-0.75/0.2-0.5 in TRV:CsRSF1/CsRSF2-infected (CsRSF1/CsRSF2-silencing) plants ( Figure 5C). These results indicated that CsRSF1 and CsRSF2 were successfully transiently silenced in cucumber cotyledons. Further, the cotyledons of TRV-infected plants were inoculated with S. fuliginea to evaluate the defense role of CsRSF1/CsRSF2 ( Figure 6). At 7 dpi with S. fuliginea, large areas of white powder (a symptom of powdery mildew disease) appeared on the cotyledons of CsRSF1/CsRSF2-silencing plants, while the areas of white powder in control and TRV: 00-treated plants ( Figure 6A,B) were smaller. Moreover, many hyphae and branch spores invaded CsRSF1/CsRSF2-silencing plants, but only a few initial germ tubes and appressoria were observed in control and TRV: 00-injected cucumbers under the microscope ( Figure 6A,B). In addition, the DI of the CsRSF1-silencing plants was 32.41; the DI of the control and TRV: 00-injected cucumbers were 22.22 and 24.44, respectively (Table  1); the DI of the CsRSF2-silencing plants was 36.11; and the DI of the control and TRV: 00injected cucumbers was 16.68 and 17.77, respectively ( Table 2). These results indicated that the silencing of CsRSF1/CsRSF2 impaired resistance to S. fuliginea, and CsRSF1 and Further, the cotyledons of TRV-infected plants were inoculated with S. fuliginea to evaluate the defense role of CsRSF1/CsRSF2 (Figure 6). At 7 dpi with S. fuliginea, large areas of white powder (a symptom of powdery mildew disease) appeared on the cotyledons of CsRSF1/CsRSF2-silencing plants, while the areas of white powder in control and TRV: 00-treated plants ( Figure 6A,B) were smaller. Moreover, many hyphae and branch spores invaded CsRSF1/CsRSF2-silencing plants, but only a few initial germ tubes and appressoria were observed in control and TRV: 00-injected cucumbers under the microscope ( Figure 6A,B). In addition, the DI of the CsRSF1-silencing plants was 32.41; the DI of the control and TRV: 00-injected cucumbers were 22.22 and 24.44, respectively (Table 1); the DI of the CsRSF2-silencing plants was 36.11; and the DI of the control and TRV: 00-injected cucumbers was 16.68 and 17.77, respectively (Table 2). These results indicated that the silencing of CsRSF1/CsRSF2 impaired resistance to S. fuliginea, and CsRSF1 and CsRSF2 were required for the cucumber defense response to S. fuliginea infection.    2.6. CsRSF1 and CsRSF2 Overexpression Improved Resistance to S. fuliginea in Cucumber CsRSF1/CsRSF2 was cloned into a pRI101-AN-GFP vector, and a GFP:CsRSF1/CsRSF2 plasmid was generated to further investigate whether CsRSF1/CsRSF2 was vital in cucumber disease resistance (Figure 7). The PCR analysis showed that the introduced genes could be detected in transient overexpression plants by chimeric primers between the CsRSF1/CsRSF2 and GFP fusion genes ( Figure 7A). Protoplast location experiments showed that the GFP signals were detected in the CsRSF1/CsRSF2-GFP co-expression regions, confirming the success of the transient overexpression of CsRSF1/CsRSF2 in vivo ( Figure 7B). In addition, the qRT-PCR assay demonstrated that the expression of the CsRSF1/CsRSF2 genes increased by 3-5.5/20-23 compared with that in the GFP: 00-injected cucumber cotyledons ( Figure 7C). These results indicated that CsRSF1 and CsRSF2 were transiently overexpressed in cucumber cotyledons successfully. CsRSF1/CsRSF2 was cloned into a pRI101-AN-GFP vector, and a GFP:CsRSF1/CsRSF2 plasmid was generated to further investigate whether CsRSF1/CsRSF2 was vital in cucumber disease resistance (Figure 7). The PCR analysis showed that the introduced genes could be detected in transient overexpression plants by chimeric primers between the CsRSF1/CsRSF2 and GFP fusion genes ( Figure 7A). Protoplast location experiments showed that the GFP signals were detected in the CsRSF1/CsRSF2-GFP co-expression regions, confirming the success of the transient overexpression of CsRSF1/CsRSF2 in vivo ( Figure 7B). In addition, the qRT-PCR assay demonstrated that the expression of the CsRSF1/CsRSF2 genes increased by 3-5.5/20-23 compared with that in the GFP: 00-injected cucumber cotyledons ( Figure 7C). These results indicated that CsRSF1 and CsRSF2 were transiently overexpressed in cucumber cotyledons successfully. After inoculation with S. fuliginea, the CsRSF1/CsRSF2-overexpressing plants exhibited significantly enhanced resistance to S. fuliginea compared with the control and GFP: 00-injected plants (Figure 8). At 7 dpi with S. fuliginea, large areas of white powder were present on the cotyledons of the control and GFP: 00-treated plants, while the areas of After inoculation with S. fuliginea, the CsRSF1/CsRSF2-overexpressing plants exhibited significantly enhanced resistance to S. fuliginea compared with the control and GFP: 00-injected plants (Figure 8). At 7 dpi with S. fuliginea, large areas of white powder were present on the cotyledons of the control and GFP: 00-treated plants, while the areas of white powder in CsRSF1/CsRSF2-overexpressing plants were smaller ( Figure 8A). Furthermore, microscopic observation indicated that the hyphae abundance of S. fuliginea was low on the inoculated base sheaths of the CsRSF1/CsRSF2-overexpressing plants than on the sheaths of the control and GFP: 00-injected plants ( Figure 8B). The finding provided supporting evidence that CsRSF1/CsRSF2 overexpression increased resistance to hyphae development in S. fuliginea. The average disease indexes of the CsRSF1/CsRSF2-overexpressing plants infected with S. fuliginea were 24.88 and 17.33, whereas those of control and GFP: 00-injected plants were 34.59 and 32.00 (Table 3). These results indicated that CsRSF1 and CsRSF2 positively regulated cucumber resistance to powdery mildew caused by S. fuliginea. white powder in CsRSF1/CsRSF2-overexpressing plants were smaller ( Figure 8A). Furthermore, microscopic observation indicated that the hyphae abundance of S. fuliginea was low on the inoculated base sheaths of the CsRSF1/CsRSF2-overexpressing plants than on the sheaths of the control and GFP: 00-injected plants ( Figure 8B). The finding provided supporting evidence that CsRSF1/CsRSF2 overexpression increased resistance to hyphae development in S. fuliginea. The average disease indexes of the CsRSF1/CsRSF2-overexpressing plants infected with S. fuliginea were 24.88 and 17.33, whereas those of control and GFP: 00-injected plants were 34.59 and 32.00 (Table 3). These results indicated that CsRSF1 and CsRSF2 positively regulated cucumber resistance to powdery mildew caused by S. fuliginea.

CsRSF1 and CsRSF2 Modulated Expression Levels of Defense-Related Genes
Chitinase (HM015248), CuPi1 (U93586.1), and PR-1a (AF475286) are three well-known S. fuliginea-related defense genes [33][34][35][36]. The expression patterns of the aforementioned three defense-related genes were analyzed in CsRSF1/CsRSF2-overexpressing and CsRSF1/ CsRSF2-silenced cucumber plants as well as the control plants to examine whether CsRSF1 and CsRSF2 regulated these defense-related genes (Figure 9). The results showed that the transcriptional levels of CuPi1 and PR-1a decreased in CsRSF1-silenced plants compared with the TRV: 00-infected control plants. Simultaneously, the transcriptional level of Chitinase significantly increased, whereas the transcriptional levels of Chitinase and PR-1a increased in CsRSF1-overexpressing plants compared with GFP: 00, while the transcriptional level of CuPi1 decreased. However, the transcriptional levels of CuPi1 and PR-1a in CsRSF2-silenced plants were significantly higher than those in the TRV: 00-infected control plants. Only the transcriptional level of Chitinase decreased, whereas the transcriptional levels of Chitinase, CuPi1, and PR-1a significantly increased in the CsRSF2-overexpressing plants compared with GFP: 00. The results indicated that the CsRSF1/CsRSF2-induced defense pathway might be related to different defense-related genes.

Discussion
Powdery mildew is a common and widespread plant disease of considerable agronomic relevance. Powdery mildew outbreaks often result in severe harvest losses [37]. Cucumber powdery mildew resistance is a particularly genetically complex trait, which often involves multiple genes, but the genetic mechanisms of cucumber resistance are still poorly understood. Recently, 67 NBS-LRR resistance gene homologs have been predicted in cucumber, some of which may be related to the resistance of powdery mildew [38]. Zhang et al. [39] found four QTLs of cucumber powdery mildew, which are pm5.1, pm5.2, pm5.3, pm6.1. The pm5.2 locus has the highest contribution rate, and four NBS resistance genes (BGICucGB009775, Csa009587, Csa009605, Csa009602) were predicted in this region.
NBS-LRR is a class of proteins important in pathogen recognition and defense response signal transduction [40,41]. However, the role of NBS-LRR gene resistance to powdery mildew, which has been studied less and is mainly concentrated in wheat. For example, the wheat NBS-LRR gene Pm21 has broad-spectrum resistance to wheat powdery mildew [42]. However, in cucumber-S. fuliginea interactions, the role of the NBS-LRR gene is unclear. In this study, CDS regions of CsRSF1 and CsRSF2 were cloned from the cucumber leaves and their functions between cucumber-S. fuliginea interactions were identified. The expression of CsRSF1 showed a trend of increase at first and then decreased in resistant and susceptible cultivars challenged with S. fuliginea. The expression of CsRSF2 was different in resistant and susceptible cultivars challenged with S. fuliginea. In the early stage of S. fuliginea infection, the expression of CsRSF1 and CsRSF2 was higher in resistant cultivars than in susceptible cultivars. These results indicated that both CsRSF1 and CsRSF2 were involved in cucumber response to S. fuliginea infections, having a positive regulatory role.
The subcellular localization of a protein is strongly related to its biological function in plants. The NBS-LRR protein, as important intracellular receptors, can directly or indirectly recognize and interact with Avr protein products inside the cell to generate a defense response [23]. The barley CNL protein MLA10 can interact with WRKY (transcription repressor) and MYB6 (transcription activator) to activate the defense response in the nucleus [24,25]. Similarly, Arabidopsis CNL proteins RPS5, RPS2 and RPM1 recognize the pathogen effectors and interact with protein kinases and activators to generate defense response in the plasma membrane [43,44]. Therefore, the subcellular localization of NBS-LRR proteins is especially important in the plant's innate immune system. In this study, the subcellular localization of 35S::CsRSF1-GFP and 35S::CsRSF2-GFP fusion proteins in tobacco epidermal cells showed that both CsRSF1 and CsRSF2 were localized in the cytoplasm and plasma membrane; mean-while, CsRSF1 was also localized in the nucleus. CsRSF1 and CsRSF2 had potential functions in various parts of the cell.
Many studies verify a gene's function by using silencing and overexpressing plants. Research showed that the tomato NBS-LRR gene SLNLC1 silences attenuated tomato resistance to Stemphylium lycopersici [45]. The overexpression and silencing of the NBS-LRR gene CaRGA in chickpeas resulted in increased resistance and sensitivity to Fusarium wilt [46]. Currently, the application of stable transgenic technology to cucumber is still difficult. Hence, an experimental method was used for the transient agroinfiltration of cucumber cotyledons to obtain transient CsRSF1 and CsRSF2 silencing and overexpression plants [47]. In this study, inoculation with S. fuliginea resulted in the typical whitish appearance of powdery mildew disease on CsRSF1 and CsRSF2 silencing plants and also caused them to exhibit greater hyphal abundances and a lower DI than control plants. This result is contrary to CsRSF1 and CsRSF2 overexpression plants. The functional analysis showed that CsRSF1 and CsRSF2 overexpression significantly increased the resistance of cucumber to S. fuliginea, while CsRSF1 and CsRSF2 silencing significantly compromised resistance to S. fuliginea. Different regulatory proteins might activate different kinds of genes in cucumber defense response to S. fuliginea [34]. Chitinase has a positive regulatory effect on the resistance of cucumber to S. fuliginea. SA signaling pathway induces PR gene expression, which can enhance resistance to a broad range of pathogens [48]. Therefore, PR gene expression is a sign of plant disease resistance. PR-1a is one of the PR genes. In this study, the expression of PR-1a was lower on CsRSF1 silencing plants than control plants, higher on CsRSF1 overexpression plants than control plants. The expression of Chitinase was lower on CsRSF2 silencing plants than control plants, higher on CsRSF2 overexpression plants than control plants. CsRSF1 positively modulated the expression of PR-1a, while CsRSF2 positively modulated the expression of Chitinase.
Plant hormones are essential for plants to respond to biological and abiotic stresses. They are important signal molecules that can participate in the regulation of plant defense signal transduction pathway and plant defense gene expression [49,50]. Several studies have shown that plant hormone are involved in the NBS-LRR protein-mediated defense responses [51].
ABA is a positive or negative regulator in the process of plant defense against various pathogens [52,53]. Exogenous application of ABA enhances rice resistance to Cochliobolus miyabeanus [54]. In addition, the application of ABA increases the susceptibility of Arabidopsis plants to Pseudomonas syringae [55]. JA helps to regulate the plant defense response to insect and microbial pathogens [56]. Niu [57] pointed out that JA was a signal molecule involved in wheat powdery mildew resistance. Exogenous application of MeJA to Arabidopsis thaliana helps protect it against gray mold [58,59]. SA is vital in plant defense; it is one of the key signaling molecules involved in activating pathogen resistance [60,61]. The Arabidopsis NBS-LRR gene RRS1-R-mediated resistance to Ralstonia solanacearum depends on SA. Plant infection by pathogens can promote ET production and in turn induce the expression of a series of genes related to the disease course [62,63]. ET signaling also interacts with JA and SA signaling pathways to help plants resist the invasion of pathogenic bacteria [64][65][66][67]. The RCY1 gene, which encodes the Arabidopsis CC-NBS-LRR protein to confer resistance to the cucumber mosaic virus, requires SA and ET signal transduction mechanisms [68]. GA promotes plant growth by stimulating the degradation of the negative growth regulator DELLA protein. DELLA protein controls the plant's immune response by regulating the balance of JA and SA signals, indicating that GA is essential in response to pathogens' invasion [69]. Reactive oxygen species (ROS) production is excessive in the early stages of plant-pathogen interactions. Appropriate levels of ROS can not only promote cell wall enhancement and plant antitoxin production but also have a signaling role in plant defense responses [70]. ROS signaling is involved in the resistance of transgenic cotton (expressing endochitinase gene) to Rhizoctonia solani [71]. The wheat NBS-LRR gene TaRCR1 maintains ROS's stability by regulating the transcription of ROS-scavenging enzyme gene CAT1 and peroxidase gene POX2. TaRCR1 then regulates the expression of defense-related genes to enhance the resistance to Rhizoctonia cerealis [70]. In this study, cucumber leaves were sprayed with different exogenous substances, such as ABA, MeJA, SA, ETH, GA, and H 2 O 2 . The qRT-PCR analysis showed that both CsRSF1 and CsRSF2 genes could be induced and expressed to different degrees. These results suggested that CsRSF1 and CsRSF2 may be involved in the crosstalk of exogenous substances against S. fuliginea. In addition, the expression of the CsRSF1 gene was upregulated and then downregulated at all treatment time points and reached the peak 12 h after ABA treatment. After GA treatment, the expression of the CsRSF2 gene was downregulated at all treatment time points. The results indicated that CsRSF1 might be associated with the ABA signaling pathway in defense responses. CsRSF2 might be involved in defense responses of the GA signaling pathway. However, their mechanism of action is currently unclear.
The rice NBS-LRR resistance protein Pit interacts with the GEF OsSPK1 to activate OsRac1 and trigger rice immunity [72]. The Gossypium hirsutum NBS-LRR gene GhDSC1 modulates defense responses of ROS accumulation and expression of JA-regulated defense genes [73]. By now, the regulatory mechanisms of NBS-LRR gene resistance to powdery mildew and it activates the defense response by which signaling pathway are unclear, especially in cucumber. Our study lay a foundation for exploring regulatory mechanisms of NBS-LRR gene resistance to powdery mildew in cucumber.

Pathogen Growth and Inoculation
S. fuliginea was collected from the naturally infected cucumber plants of the Vegetable Institute of the Liaoning Academy of Agricultural Sciences (Shenyang, China). The pathogen was cultured on leaves of potted cucumber and amplified at 25 • C in the laboratory. Spores of S. fuliginea were harvested from highly infected cucumber leaves by washing leaves with sterile water containing 10 µg·mL −1 SDS and filtering with sterile gauze. The spore concentration was adjusted to 30-40 spores per field of vision (10 × 10 times under microscope). Then the spore suspension was sprayed evenly on cucumber leaves without S. fuliginea using a throat sprayer [74].

Full-Length CDS Cloning and Sequence Analysis
Primers (F and R) were designed according to the sequences of the target genes found on the Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html, accessed on 16 July 2018). Primer design was conducted using Primer 5 software (Premier, Mississauga, Ontario, Canada). The gene-specific primers are listed in Table S1. Total RNA was extracted from the leaves of B21-a-2-1-2 and B21-a-2-2-2 using an RNAprep pure plant kit (Tiangen Biotech, Beijing, China) and synthesized into cDNA using the FastQuant cDNA (Tiangen Biotech, Beijing, China). The following PCR program was used: 98 • C (10 s), 56 • C (15 s), and 68 • C (3 min) for 35 cycles. The reaction was conducted in a 20 µL volume containing 5 µL of 5× PrimeSTAR GXL buffer, 2 µL of dNTP mixture, 0.5 µL of each primer, 2 µL of genomic DNA, 0.5 µL of PrimeSTAR GXL DNA polymerase, and 8.5 µL of doubly distilled water. DNAMAN software (Lynnon Biosoft, San Ramon, CA, USA) was used for sequence alignment. Bioinformatics software was used for sequence analysis.

Subcellular Localization
The coding regions of CsRSF1 and CsRSF2 were amplified using PrimeSTAR GXL DNA polymerase (Takara, Dalian, China). The PCR amplification procedure was as follows: 98 • C (10 s), 56 • C (15 s), and 68 • C (3 min) for 35 cycles. The PCR products were detected by agarose gel electrophoresis and ligated to the 3 end of the GFP coding region without the stop codon in the p35S:GFP vector, generating the fusion constructs p35S:GFP-CsRSF1 and p35S:GFP-CsRSF2. The recombinant plasmids p35S:GFP-CsRSF1, p35S:GFP-CsRSF2, and p35S:GFP were separately transformed into Agrobacterium tumefaciens strain EHA105 and then centrifuged after overnight culture. The precipitate after centrifugation was cultured in induction medium (10 mmol·L −1 2-(4-morpholino) ethanesulfonic acid (MES), pH 5.7, 10 mmol·L −1 MgCl2 and 200 mmol·L −1 acetosyringone (AS)), collected and diluted to OD 600 = 0.6, and then infiltrated into Nicotiana benthamiana leaves. The plants were incubated at 25 • C for 3 days after injection, and point GFP signals were observed and photographed using a laser confocal fluorescence microscope (Leica, TSC SP8, Solms, Germany) with an excitation wavelength of 488 nm and 505 to 530 nm bandpass emission filter.

Tobacco Rattle Virus (TRV)-Mediated CsRSF1/CsRSF2 Gene Silencing
The pTRV2:CsRSF1 and pTRV2:CsRSF2 recombinant constructs were generated by subcloning a 233 bp-specific fragment within the 3 region of the sequence of CsRSF1 (from 2225 to 2457 nucleotides in CsRSF1 CDS sequence) and a 289 bp-specific fragment within the 5 region of the sequence of CsRSF1 (from 1 to 289 nucleotides in CsRSF1 CDS sequence) in the antisense orientation into the pTRV2. The gene-specific primers are listed in Table S1. Then, the pTRV2:CsRSF1 and pTRV2:CsRSF2 recombinant constructs were introduced into different A. tumefaciens EHA105 aliquots. A 12.5 mL culture of different A. tumefaciens (pTRV2:CsRSF1, pTRV2:CsRSF2, and pTRV2, pTRV1) strains were grown overnight at 28 • C in the YEP medium supplemented with 50 mg·L −1 of rifampicin and 50 mg·L −1 of kanamycin. Then, each 100 µL overnight culture was inoculated into 12.5 mL portions of the YEP medium with the aforementioned antibiotics and cultivated at 28 • C until the culture reached selected optical densities of OD 600 = 0.8-1.0. The samples were supplemented with 10 mmol·L −1 MES, 10 mmol·L −1 MgCl 2 , and 200 µmol·L −1 AS until the OD 600 = 0.4 for each construct. The induced A. tumefaciens EHA105 strains carrying different pTRV2-derived vectors (pTRV2, pTRV2:CsRSF1, and pTRV2:CsRSF2) were mixed with the pTRV1 A. tumefaciens strain EHA105 in a ratio of 1:1. Then, the samples were co-infiltrated into fully expanded cotyledons of cucumber plants using a 1 mL syringe. The plants were placed in a room at 22 • C with a 16 h light and 8 h dark photoperiod for growth.

CsRSF1 and CsRSF2 Overexpression Transformation
The full CDS sequences of the CsRSF1 and CsRSF2 genes were subcloned into a modified pRI101-AN-GFP vector with a GFP tag, resulting in the transformation vector pRI101-AN-GFP:CsRSF1 and pRI101-AN-GFP:CsRSF2. The cloning site was between BamHI and SalI. The gene-specific primers are listed in Table S1. The constructed vectors and pRI101-AN-GFP were transformed into A. tumefaciens strain EHA105. The transformation method was the same as described earlier.

Quantitative Reverse-Transcription PCR (qRT-PCR)
Total RNA was extracted from cucumber euphylla, and cotyledons using an RNAprep pure plant kit (Tiangen Biotech, Beijing, China) and was used for cDNA synthesis using a fasting RT Kit cDNA (Tiangen Biotech, Beijing, China). The primers were designed using Primer 5 software (Premier, Mississauga, ON, Canada), and the PCR reaction quality was estimated based on melting curves. Three biological replicates of cDNA were used for three repeated experiments. The actin gene of cucumber was used as the internal reference.  (Table S2).

Coomassie Brilliant Blue R250 Staining and Microscopy
Coomassie brilliant blue R250 staining was used to observe the S. fuliginea hyphae on the cotyledons with the control of TRV:00, TRV:CsRSF1, TRV:CsRSF2, GFP: 00, GFP:CsRSF1, and GFP:CsRSF2 transgenic plants on day 7 with S. fuliginea. The infected cotyledons of control and transgenic plants were immersed in a destaining solution (containing 0.225 g trichloroacetic acid, 150 mL of absolute ethanol, and 50 mL of acetone) for 30 min at 70 • C. The samples were then stained with Coomassie brilliant blue R250 solution (containing 0.225 g trichloroacetic acid, 0.9 g Coomassie brilliant blue R250, and 150 mL of methanol) for 2 min at room temperature. After the samples were washed with sterile water, they were immersed in 20% glycerol and photographed. The fungus was visualized under a light microscope (Nikon Ts2, Tokyo, Japan).

Disease Index
The typical symptoms of pathogen whiteness were measured in six disease severity ratings from 0 to 9, where 0 denoted no symptom, 1 denoted white lesions accounting for less than 5% of the entire inoculated leaves, 3 denoted white lesions accounting for 5%-25% of the entire inoculated leaves, 5 denoted white lesions accounting for 26%-50% of the entire inoculated leaves, 7 denoted white lesions accounting for 51%-75% of the entire inoculated leaves, and 9 denoted white lesions accounting for more than 75% of the entire inoculated leaves. Disease index (DI) was calculated using the following formula: DI (%) = [(N × D)/(H × T)] × 100, where N is the number of leaves with the respective disease rating, D is the disease rating, H is the highest disease rating, and T is the total number of observed leaves [75]. No fewer than 60 cucumber leaves were surveyed for the DI evaluation.

Data Processing and Statistical Analysis
Data were the mean ± standard deviation of three biological replicates per cultivar. The 2 −∆∆Ct method was used to calculate the relative expression levels of the target genes. Standard errors of deviation were assessed using the STDEVA function in Excel. Statistical significance was analyzed using the Student's t-test (p < 0.05 or p < 0.01) via SPSS software (SPSS 22.0, Nlinedown, Guangdong, China).

Bioinformatics Analysis Content and Tools Website
The primary structure was analyzed by using ProtParam. The website is http://web. expasy.org/protparam/, accessed on 9 April 2019 . The domain was analyzed by using SMART and InterProScan. The websites are http://smart.embl-heidelberg.de/, accessed on 9 April 2019 and ftp://ftp.ebi.ac.uk/pub/soft-ware/unix/iprscan/5/, accessed on 9 April 2019.

Conclusions
Cucumber CsRSF1 and CsRSF2 genes acted as positive modulators in the response of cucumber to S. fuliginea infections, and their regulatory mechanisms might be different. They participated in the defense response to S. fuliginea by regulating the expression of certain defense-related genes. In summary, this research provided new insights into the important roles of CsRSF1 and CsRSF2 in the resistance to S. fuliginea. This research laid a good foundation for exploring the molecular mechanism of cucumber resistance to S. fuliginea and breeding resistant varieties.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/ijms22083986/s1, Figure S1: Nucleotide and deduced amino acid sequences of cucumber CsRSF1 CDS encoding a protein. Figure S2: Nucleotide and deduced amino acid sequences of cucumber CsRSF2 CDS encoding a protein. Figure S3: Chromosome location of NBS-LRR superfamily in the cucumber. Figure S4: The structural domain of CsRSF protein. Table S1: The primers of PCR used in this study. Table S2: The primers of qRT-PCR used in this study.