Discussion
We demonstrate here that the RPS4/RRS1 R gene pair from Arabidopsis functions in other Brassicaceae plants, B. rapa and B. napus, as well as in N. benthamiana, tomato and cucumber. This provides that transfer of NB-LRR-type R genes confers resistance to multiple pathogens in taxonomically distinct families. The transfer of either RPS4 or RRS1 failed to provide resistance in transgenic plants (Figure 2A, B & 3), thus the success of interfamily transfer is likely due to the “dual R” system.
The number of known pairs of R genes is increasing [11]. For example, the R gene pairs, RPP2A/RPP2 [12], N/NRG1 [13], RPM1/TAO1 [14], Lr10/RGA2 [15], Pi5-1/Pi5-2 [16], Pikm1-TS/Pikm2-TS [17], are only functional when both genes are present and we postulate that some of those pairs may also function in taxonomically distinct families when expressed together. It is also possible that some apparently singular R genes may require a complement to function. The mechanism of how these “dual R” proteins work together is unknown. However, one of the potential functions of the pair is as a negative regulator, as expression of single R genes often leads to inappropriate autoimmune responses in a taxonomically distinct family where no partner is available [18]. In this sense, expression of R proteins along with the corresponding proteins that are the target of pathogen effectors (often called ‘guardee’) may also be able to overcome restricted taxonomic functionality, since loss of the guardees often triggers R protein activation [1], [6].
Because RPS4 and RRS1 specifically react to bacterial effectors AvrRps4 and PopP2 in both Brassicaceae and Solanaceae plants, the recognition mechanism should be highly conserved across species, or built in the R protein pair itself. Recent reports indicate that AvrRps4 directly targets Enhanced Disease Susceptibility 1 (EDS1), but does not interact with RPS4 [19], [20]. EDS1 is a highly conserved key factor in TIR-NB-LRR-mediated ETI and interacts with several TIR-NB-LRR protiens, including RPS4 [21]. However, more recent report by Sohn et al. (2012) showed that AvrRps4 and EDS1 do not directly interacted [22]. Thus, EDS1 can be the direct or indirect target of AvrRps4 in both Brassicaceae and Solanaceae plants. If this is the case, RPS4 and RRS1 could confer broad resistance, as many pathogens would target EDS1 as the key immunity protein. Unlike the AvrRps4-RPS4/RRS1 example, PopP2, a YopJ-like family of cysteine proteases, directly binds to RRS1 [23]. As PopP2-dependent immune response requires RPS4 (Figure 3), it is possible that RPS4 can recognize the interaction between PopP2 and RRS1. In any case, PopP2 produced by Ralstonia could be detected in many plants by introgressing the RPS4/RRS1 cassette.
Perhaps the biggest surprise to us is that the RPS4/RRS1 pair confers immunity against both C. higginsianum and C. orbiculare in taxonomically distinct plant families. This implies that the target (i.e. host immunity-related proteins) of the unidentified Colletotrichum effectors is conserved in Brassicaceae and Cucurbitaceae, and/or that the effectors recognized by RPS4/RRS1 are highly conserved in Colletotrichum. As is the case for AvrRps4 and PopP2, Colletotrichum effectors may directly target EDS1 or RRS1. The sequenced Colletotrichum genomes do not contain apparent AvrRps4 and PopP2 homologs, suggesting that the targeting mechanism should be distinct from the bacterial targeting mechanism [24]. There are a number of potential highly conserved effectors in Colletotrichum spp., and some of them could be detected by the RPS4/RRS1 pair by an unknown mechanism. Isolation of the Colletotrichum effectors would resolve the problem. In any case, RPS4/RRS1 could provide resistance to Colletotrichum species among a number of agronomically important crops and vegetables.
Our study demonstrates that NB-LRR-type R gene-based immunity can be transferred to distantly related species once the right gene pair is identified. Each plant genome contains about 150 to 500 NB-LRR-type R genes [25], [26], some of which are likely to form pairs. As in the case of RPS4/RRS1, some gene pairs might provide a wide range resistance to multiple pathogens. Thus our finding provides a new strategy for creating pathogen-resistant vegetables and crops by a transgenic approach using previously unexploited resource of genetic resistance.