Root-knot nematodes (RKNs) from the genus Meloidogyne are a class of sedentary, endozoic plant parasites with a wide host range (Mitkowski and Abawi, 2003). RKNs (particularly Meloidogyne incognita) cause significant economic damage in the range of hundreds of billions of dollars annually (Jones et al., 2013). M. incognita infected host plants may induce a series of disease resistance reactions that ultimately inhibit both invasion and life cycle completion of nematodes (Abad and Williamson, 2010). The plant immunity response typically consists of two stages: pathogen-associated molecular pattern-triggered immunity (PTI) and effector-triggered immunity (ETI) (Jones and Dangl, 2006). PTI requires pattern recognition receptors (PRRs) located in either the cytomembrane or cytoplasm (Boller and He, 2009; Dodds and Rathjen, 2010). Adaptive pathogens have been reported to interfere with host PTI via small secretory proteins known as effectors (Block et al., 2008; Göhre and Robatzek, 2008; Cunnac et al., 2009). Several plant species have evolved specific resistance genes (R genes) that are able to recognize particular effectors, thus facilitating ETI (Zipfel et al., 2006; Tsuda and Katagiri, 2010). Resistance associated with ETI has been shown to be stronger and more enduring than PTI associated resistance (Glazebrook, 2005; Torres et al., 2006; Underwood et al., 2007). R gene encoded R proteins play a pivotal role in detecting pathogens and in initiating immune responses (Jacob et al., 2013). Plant R proteins are typically composed of three major domains: a coiled-coil (CC) or a Toll/interleukin-1 receptor (TIR) domain, a nucleotide-binding (NB) domain, and a leucine-rich repeat (LRR) domain (Takken and Goverse, 2012). The R gene structure is highly conserved in both plants and mammals (Padmanabhan et al., 2009); however, plants possess considerably more R genes, suggesting that plant R genes are more pathogen-specific and potentially facilitate more diverse and complex immune responses than those of mammals.

R genes for RKNs have been isolated from various plants. The first isolated RKN R gene was Mi-1 (Vos et al., 1998), which originated from tomato (Solanum lycopersicum) and conferred race-specific resistance to RKN (Meloidogyne spp.) and potato aphids (Macrosiphum euphorbiae) (Milligan et al., 1998). Subsequently, RKN R genes have either been cloned or mapped from cotton (Gossypium hirsutum; GHNTR1) (Zhang et al., 2015), potato (Solanum tuberosum; Rmc) (van der Voort et al., 1999), pepper (Capsicum annuum; Me) (Celik et al., 2016), and myrobalan (Prunus cerasifera, Ma) (Claverie et al., 2011). These RKN R genes contain NBS-LRR domains and typically exhibit reduced expression levels in the absence of pathogens. However, RKN R gene expression significantly increases in response to RKN infection (Zhang et al., 2015). The plant hypersensitive response (HR) is a disease resistance mechanism that occurs in response to an incompatible plant–pathogen interaction and is typically accompanied by programmed cell death (Karrer et al., 1998). Mi-1 in tomato induces the HR 12 h after RKN infection, blocking the formation of feeding sites and therefore inhibiting RKN development (Melillo et al., 2006). However, the molecular mechanism underlying Mi-1-facilitated RKN resistance remains unclear. Furthermore, due to the species specificity of R genes, use of Mi-1 to engineer RKN resistance in heterologous plants did not result in stable RKN immunity. While transgenic Mi-1 expressing eggplant (Solanum melongena) varieties displayed RKN resistance, no aphid resistance could be found (Goggin et al., 2006). Furthermore, Mi-1 expressing Arabidopsis thaliana and tobacco (Nicotiana benthamiana) exhibited no pathogen resistance (Williamson and Kumar, 2006). These results indicate that R gene induced resistance signaling pathways vary considerably between different plant species (Martin et al., 2003; Bonardi and Dangl, 2012). Plants with a longer lifespan than herbaceous plants (such as perennial fruit trees) have a higher risk of RKN infection, consequently warranting a more comprehensive, and thus more complex, mechanism of RKN resistance. The identification of novel RKN R genes from related plant species may assist in the engineering of stable and enduring RKN resistance.

Studies using model plant species revealed that R proteins require the co-chaperone protein complex HSP90-SGT1-RAR1 to achieve correct folding, maturation, and stabilization (Huang et al., 2014). The heat shock protein 90 (HSP90) is a highly conserved protein in higher plants, is a suppressor of the G2 allele of skp1 (SGT1), and is required for mla12 resistance (RAR1) (Kadota and Shirasu, 2012). HSP90 is involved in the assembly, maturity, and stabilization of numerous critical signaling proteins in eukaryotic cells, and typically exists as a homodimer (Csermely et al., 1998; Hubert et al., 2003). SGT1 regulates important biochemical processes, such as ubiquitination and kinetochore assembly, and maintains the stability of the R protein structure in the absence of pathogens (Austin et al., 2002). RAR1 was first characterized in barley (Hordeum vulgare), and while rar1 mutant plants appear morphologically normal, they are less resistant to fungal pathogens, despite the presence of the R gene MLA12 (Bieri et al., 2004). Both HSP90 homodimer and SGT1 interact with a number of R proteins, and consequently, loss of function of HSP90, SGT1, or RAR1 will reduce the functionality of R proteins such as MLA1, MAL6, Rx, RPM1, RPS5, Mi-1, and I-2, thus compromising plant immunity (Lu et al., 2003; Bieri et al., 2004; Holt et al., 2005; Bhattarai et al., 2007; Botër et al., 2007; Van Ooijen et al., 2010). Recently, the relationship between the HSP90-SGT1-RAR1 co-chaperone protein complex and RKN resistance has been further characterized. While RKN resistance is not affected in Mi-1-engineered tomato in response to RNA interference induced reduction of SlRAR1 gene expression, reduced expression of SlHSP90 and SlSGT1 results in increased numbers of galls and egg mass in plants exposed to M. javanica RKNs (Bhattarai et al., 2007). This suggests that HSP90 and SGT1 also have a function in RKN resistance. Most studies that examine the relationship between R genes and the HSP90-SGT1-RAR1 co-chaperone protein have been based on herbaceous model plant species, and comparable research for woody plants is still missing.

The Xinjiang wild myrobalan (Prunus sogdiana) is distributed throughout native forests in the eastern region of the Tianshan Mountains (Esmenjaud et al., 1996; Wang et al., 2006; Kang et al., 2008). We collected P. sogdiana seeds from the Xinjiang Yili native forest in 2005 and used these seeds from 2007 onward to cultivate more than 150 individual P. sogdiana plants at the Shangzhuang experimental station of the China Agricultural University (CAU). Among the resulting plants, we found several individuals that displayed complete resistance to M. incognita and repeated this experiment six times, which lasted 6 years (Xiao et al., 2010; Li et al., 2011; Liu et al., 2013; Chen et al., 2015; Qiu et al., 2016). Via homologous cloning of genes from highly resistant P. sogdiana individuals based on reported RKN R gene sequences, we identified a series of NB domain fragments that respond to RKN invasion. Full-length R gene cDNA sequences were obtained via 3′- and 5′- RACE, revealing PsoRPM2 as the underlying R gene. To characterize the observed M. incognita resistance conferred by this gene, we created transgenic heterologously expressing PsoRPM2 lines of the RKN-susceptible tobacco variety W38 (Nicotiana tabacum cv. W38). Analyses of the interactions between PsoRPM2, PsoHSP90-1, and PsoSGT1 provided insight into the underlying mechanism of P. sogdiana RKN immunity.

Several methods for the prevention and control of RKN disease have been developed due to the severity of resulting damage to the agricultural industry; however, it remains a mainstream choice to seek naturally RKN-resistant R genes in plants. Previous studies have shown that the R gene initiates the identification of pathogen invasion and maintains a resting state if the pathogen is invaded (Jacob et al., 2013). This could reduce the cost of labor and pesticides compared to human intervention. However, the detection of novel R genes in plants often requires prior genomic data or prolonged hybridization to obtain resistance-related molecular markers. All this is more difficult to achieve in woody plants, especially in wild fruit tree germplasm resources. Fortunately, previous research has found that most of the R genes are highly conserved in their NB-ARC domains (Jacob et al., 2013). This conserved domain can be utilized as a molecular marker (Gupta and Rustgi, 2004), which makes it possible to clone novel R genes from plants without prior genomic data such as in the case of the Xinjiang wild myrobalan P. sogdiana. A number of resistant genes have been obtained via this RGAs homologous cloning method, such as MbR4 (Lee and Lee, 2005), Lr1 (Cloutier et al., 2007), and GHNTR1 (Zhang et al., 2015). This method is fast and causes low cost; however, it still has several shortcomings. For example, some R genes with considerable variability in the NB-ARC conserved region cannot be cloned via homologous (Liu et al., 2014). The PsoRPM2 gene can rapidly and significantly increase its expression in response to M. incognita invasion and thus enhance the RKN resistance of transgenic tobacco lines, suggesting that it is likely involved in disease resistance, which is similar to previous studies on the Mla gene (Halterman et al., 2003) and the GHNTR1 gene (Zhang et al., 2015) against powdery mildew and RKN, respectively.

The PsoRPM2 protein belongs to the TIR-NB-ARC-LRR subfamily, which is similar to Ma (Claverie et al., 2011), and the Mi-1 and the GHNTR1 proteins belong to the CC-NB-ARC-LRR subfamily (Milligan et al., 1998; Zhang et al., 2015). The difference between CC and TIR domains may affect the R gene function during plant immunity (Shirasu, 2009). The NB-ARC domain in the middle of R proteins has been suggested to act as an elicitor, which may maintain an intramolecular interaction with the LRR domain during the inactive protein state (Tameling et al., 2002; Steinbrenner et al., 2015). The NB-ARC domain structure of the PsoRPM2 protein was most similar to the ced-4 protein (Yan et al., 2004), which was obtained via three-dimensional structure analysis (confidence 100%, coverage 96%). The ced-4 protein mediated apoptosis (Yan et al., 2004), indicating that the NB-ARC domain may achieve the nematode resistance signal of PsoRPM2. This indicated that the NB-ARC domain of PsoRPM2 may induce RKN resistant signaling. The number of repeats of the LRR domain differs between different nematode proteins and its 3D structure was predicted to have a horseshoe structure (Takken and Goverse, 2012). This domain of PsoRPM2 may be associated with pathogen recognition and promotion of disease resistance as previously reported (Krasileva et al., 2010).

The gene function of wild-planted resources is relatively difficult to determine because the genetic transformation systems of these wild germplasms are largely immature or require several years of hybridization. Therefore, it forms a viable alternative approach to use a model plant material without nematode resistance to analyze candidate RKN-resistant genes. In our PsoRPM2 transgenic tobacco lines, L1–4 were associated with a remarkable reduction of M. incognita susceptibility (Figure 5). However, the level of RKN resistance varied between the four PsoRPM2-expressing transgenic lines. L1 displayed almost no nematode infection symptoms and has the highest expression of the four transgenic lines (Figure 3B), while the remaining three transgenic lines exhibited slight signs of nematode infection (Figures 5D,E). The expression changes of PsoRPM2 during nematode infection (0–5 dpi) in transgenic tobacco lines were not obvious (Figure 3B), suggest the 35S promoter may not infected by RKN infection. However, even the most RKN resistance L1 did not reach the resistance level of resistant P. sogdiana (Li et al., 2011). This demonstrated that xenogeneic transgene method has several shortcomings: several of the extreme examples of heterologous genes may be affected by the host plant immune system during gene expression or protein translation, or the foreign gene cannot properly match the local immunity signaling pathway. An extreme example is that the tomato Mi-1 gene is neither resistant to RKN nor to the potato aphid when it is expressed in the eggplant (Goggin et al., 2006). A further example is that host genetic background and dose-dependency may influence the Mi-1 RKN resistance in tomato varieties (Jacquet et al., 2005). In combination with the expression of the PsoRPM2 gene in resistance P. sogdiana and the resistance to RKN of transgenic tobaccos (i.e., L1, Figures 3B, 5), we found that the higher expression of the PsoRPM2 gene usually induced stronger RKN resistance, which is consistent with a previously reported observation for Mi-1.2 transgenic tomatoes (Milligan et al., 1998). In addition, we found that the PsoRPM2-expressing transgenic tobacco lines all exhibited early flowering, which implies that PsoRPM2 is also involved in plant reproductive development.

Previous studies have shown that the R protein is not only a single promote element in the immune systems of herbaceous model plants, but also requires assistance of the chaperone complex to promote disease resistance (Shirasu, 2009; Tran et al., 2017). With the help of genomic data of P. mume, we found that PsoRPM2 and its co-chaperones HSP90, SGT1, and RAR1 could also have interacted (Figures 6, 7), suggesting that the cooperative relationship of this R protein and its co-chaperone complex is also present in woody fruit tree. The sequence conservation of these chaperone proteins of different plants suggests that their function may be important for the immunity and physiological processes of plants (Kadota et al., 2010). A lack of HSP90, SGT1, or RAR1 proteins affects the resistance of many R genes including the RKN-resistant gene (Hubert et al., 2003; Bieri et al., 2004; Holt et al., 2005; Botër et al., 2007; Van Ooijen et al., 2010), while the SGT1 protein assists the HSP90 protein to maintain a safe R protein level, thus preventing an autoimmune response (Hahn, 2005; da Silva et al., 2007). Similar to Arabidopsis, multiple HSP90 homologues exist in P. mume and P. sogdiana; however, only the PsoHSP90-1 protein interacts with the PsoSGT1 protein (Figure 6). This unique manifestation is consistent with an effect observed in Arabidopsis (Takahashi et al., 2003). However, it is worth noting that HSP90.2 and HSP90.3 proteins are still involved in pathogen resistance in Arabidopsis (Krishna and Gloor, 2001), even though they do not interact with the SGT1-RAR1 protein, indicating redundancy in the function of the HSP90 protein. This suggests that HSP90-2 and HSP83 may also be involved in the disease resistance process of P. sogdiana. The PsoHSP90-1 and SGT1 genes are able to form dimers, which is consistent with the findings of previous reports (Nyarko et al., 2007; Röhl et al., 2013). Several R proteins can form homodimeric (Mestre and Baulcombe, 2006; Bai et al., 2012) or heterodimeric interactions (Williams et al., 2014) thus triggering immunity. However, in this study, no interaction was observed in the PsoRPM2 protein (Figure 6B), and this complex variation in the plant immune system may be associated with a different disease resistance pattern of the R gene. In the limited study of chaperone protein complexes in RKN-resistance publications, the normal function of Mi-1 requires the interaction of Hsp90-1 and Sgt1-1 proteins, whereas RAR1 had no effect on RKN resistance (Bhattarai et al., 2007). PsoRPM2-HSP90-1-SGT1-RAR1 may also possibly have formed a complex in P. sogdiana, and the resistant mechanism of the PsoRPM2 gene may be similar to other R genes that require assistance from the co-chaperone protein complex (Kadota et al., 2010).

In summary, our study showed that the PsoRPM2 gene enhanced the plant RKN-resistance and interacted with its chaperone protein complex. Future work should focus on the resistance mechanism of the PsoRPM2-HSP90-1-SGT1-RAR1 complex during RKN infection and should develop stable genetic transformation systems in P. sogdiana.

Conceived and designed the experiments: XZ and JH. Performed the experiments: XZ, KX, and HC. Analyzed the data: XZ and KX. Wrote the manuscript: XZ, KX, and JH.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.