Yang Liu, Xin Zhang, Guixin Yuan, Dongli Wang, Yangyang Zheng, Mengqi Ma, Liwei Guo, Vijai Bhadauria, You-Liang Peng, and Junfeng Liu

PNAS November 2, 2021 118 (44) e2110751118

Significance

In this study, we generated a mutant of the rice nucleotide-binding and leucine-rich repeat (NLR) immunity receptor RGA5 by engineering its heavy metal–associated domain that recognizes the noncorresponding Magnaporthe oryzae Avrs- and ToxB-like effector AvrPib and confers resistance in transgenic rice to the blast fungus isolates with AvrPib, which is known to trigger blast resistance in rice cultivars carrying the R gene Pib, albeit by unknown mechanisms. Thus, this work demonstrates that integrated domain-containing plant NLR receptors can be engineered to confer resistance to pathogens carrying avirulence effectors that trigger plant immunity by unknown mechanisms, thereby providing a practical approach for developing multilines and cultivars with broad race spectrum resistance.

Abstract

Plant nucleotide-binding and leucine-rich repeat (NLR) receptors recognize avirulence effectors directly through their integrated domains (IDs) or indirectly via the effector-targeted proteins. Previous studies have succeeded in generating designer NLR receptors with new recognition profiles by engineering IDs or targeted proteins based on prior knowledge of their interactions with the effectors. However, it is yet a challenge to design a new plant receptor capable of recognizing effectors that function by unknown mechanisms. Several rice NLR immune receptors, including RGA5, possess an integrated heavy metal–associated (HMA) domain that recognizes corresponding Magnaporthe oryzae Avrs and ToxB-like (MAX) effectors in the rice blast fungus. Here, we report a designer rice NLR receptor RGA5^HMA2 carrying an engineered, integrated HMA domain (RGA5-HMA2) that can recognize the noncorresponding MAX effector AvrPib and confers the RGA4-dependent resistance to the M. oryzae isolates expressing AvrPib, which originally triggers the Pib-mediated blast resistance via unknown mechanisms. The RGA5-HMA2 domain is contrived based on the high structural similarity of AvrPib with two MAX effectors, AVR-Pia and AVR1-CO39, recognized by cognate RGA5-HMA, the binding interface between AVR1-CO39 and RGA5-HMA, and the distinct surface charge of AvrPib and RAG5-HMA. This work demonstrates that rice NLR receptors with the HMA domain can be engineered to confer resistance to the M. oryzae isolates noncorresponding but structurally similar MAX effectors, which manifest cognate NLR receptor–mediated resistance with unknown mechanisms. Our study also provides a practical approach for developing rice multilines and broad race spectrum–resistant cultivars by introducing a series of engineered NLR receptors.

See: https://www.pnas.org/content/118/44/e2110751118

Fig. 5.

RGA4/RGA5^HMA2 confers specific resistance in transgenic rice to the blast fungus carrying AvrPib. (A) Transgenic lines of Nipponbare-expressing RGA4/RGA5^HMA2 and the monogenic Lijiangxintuanheigu (LTH) line K14 (carrying Pib) form resistant lesions after infection only by the transgenic blast fungus strain DG7-AvrPib but not by DG7 and DG7-AVR-Pia. Similarly, transgenic Nipponbare lines of RGA4/RGA5 and the LTH K1 line (carrying Pia) form resistant lesions only by the blast fungus strain DG7-AVR-Pia but not by DG7 and DG7-AvrPib. In contrast, Nipponbare develops susceptible lesions after infection by all three strains. RGA5 and RGA5^HMA2 were independently cotransformed with RGA4 into Nipponbare. The T1 generation seedlings were used for inoculation. Inoculation was performed by spotting three 10-μL droplet of conidial suspension (10⁵ conidia/mL) onto the detached leaves of 4-wk-old rice seedlings. Images of two representative leaves of different lines were taken 4 d after inoculation. The numbers 1 and 2 represent two independent transgenic rice lines. (B) Box-and-whisker plots show lesion areas on the infected rice leaves from different rice lines in A inoculated with different isolates. Each sample from two lines was conducted with three independent biological replicates. Disease areas of each lesion were measured with ImageJ after 4 d of inoculation. The statistical analysis was conducted using an estimation method. Means along with SD were calculated from at least nine lesions of three independent seedlings for each rice line. (C) Biomass of the rice blast fungus M. oryzae MoPot2 in relation to the rice ubiquitin gene. Relative fungal growth was calculated as a ratio (MoPot2/OsUbq) to reflect the amplification efficiency. The transgenic blast fungus strains were labeled on the bottom. Asterisks represent statistically significant differences in the expression levels of MoPot2 at P < 0.05. Significant differences were determined using t test.