The Race Structure of the Rice Blast Pathogen Across Southern and Northeastern China

Mo Isolate Collection and Maintenance

Panicles exhibiting symptoms of blast disease were collected over a 3 year period, 2006–2008, from fields in each of GD, HN, LN, and HLJ provinces. The methods used to isolate single spores and to store the resulting isolates have been described elsewhere (Feng et al. 2007; Zeng et al. 2009). Each of the four Mo populations generated was represented by 60 monoconidial isolates; the GD population originated from 53 cultivars in 43 counties, the HN population from 51 cultivars in 15 counties, the LN population from 43 cultivars in 16 counties, and the HLJ population from 29 cultivars in 21 counties (Additional file 1: Table S1).

Rice Differential Cultivar Sets, Inoculation and Disease Scoring

Both the CDC and JDC sets were tested for their ability to characterize the 240 Mo isolates. Each differential cultivar was represented by five seedlings, grown in a plastic tray (58 × 38 × 8 cm), as described by Pan et al. (2003). Mo inoculum was also prepared following Pan et al. (2003). Six days after inoculation, the host reaction was recorded using the Pan et al. (1996) scale, where a score of 0–2 was considered to be a resistant reaction and a score of 3–5 as susceptible. Each isolate was tested at least twice in independent inoculations, and the highest score recorded was adopted whenever there was any inconsistency.

Race Coding

Under the CDC system, the identity of each race was defined by an alphanumeric code, which was modified from the international system (Atkins et al. 1967; All China Corporation of Research on Physiological Races of Pyricularia oryzae 1980; Xing et al. 2017). The letters from A to G were sequentially assigned to the seven CDCs, Tetep, Zhenlong 13, Sifeng 43, Dongnong 363, Kando 51, Hejiang 18 and Linjiangxintuanheigu (LTH), with the letter H applied to isolates which were incompatible on all seven CDCs. The races represented within a given alphabetical group were separated from one another by assigning a numerical value derived by summing the codes corresponding to the remaining differentials on which incompatible reactions produced, following by adding 1 to the numerical value (Additional file 2: Table S2). Finally the letter “Z” was inserted at the front of the designator to represent China (Zhongguo). For example, an isolate producing the reaction profile RRSRRRS (R: resistant, S: susceptible) was classified into the ZC group. From C, there were three CDCs, D, E, and F, showing resistance reactions, the race number was as 8 + 4 + 2 = 14 + 1 = 15. Thus this isolate’s race designator was given as ZC15. Under the JDC system, race designators are based on a purely numerical code (Kiyosawa 1984; Additional file 2: Table S2), where, for example, an isolate evoking the reaction pattern SSRSSRRRSSSS is given the designator 433.7 (1 + 2 + 10 + 20 + 400 + 0.1 + 0.2 + 0.4).

Data Analysis

Six parameters were used to characterize the race structure of the four populations, namely the total race frequency f _tr given by (tr/N)*100%, the race diversity index h _rdi given by (N/N-1)(1-σx _i ² ), the population-specific race frequency f _psr given by (psr/tr)*100%, the population-specific race isolate frequency f _psri given by (psri/N)*100%, the dominant race isolate frequency f _dri given by (dri/N)*100%, and the top-three-dominant race isolate frequency f _tdri given by (tdri/N)*100%; in these expressions, N represents the number of the sum isolates of each population, tr the number of races identified in a given population, x _i the frequency of race #i, psr the number of population-specific races, psri the number of population-specific race isolates, dri the number of respective dominant race isolates, and tdri the number of the three predominant race isolates. A cultivar resistance gene frequency f _CR was calculated for each differential cultivar from the expression (cr/N)*100%, where cr represents the number of isolates which proved to be avirulent. Each cultivar was given an overall resistance rating (high: > 85%, intermediate: 60–84%, or low < 60%) based on information provided by local breeders.

Race Diversity

Specific Race Structure

Dominant Race Structure

Resistance Gene Profiling

Similar results were also identified in both northeastern populations (LN vs HLJ), four cultivars, Tetep (88.3% vs 90.0%), Sifeng 43 (98.3% vs 90.0%), Zhenlong 13 (88.3% vs 100.0%), and Dongnong 363 (98.3% vs 100.0%) in the CDC set, were considered as the higher resistance cultivars in both populations (Table 4). Note that the key R genes other than Pi1 carried by such higher cultivars were denied recommending to the local breeding programs, because both Pib and Pik in LN, and Pik in HLJ populations were conveyed the lower resistance in the respective JDCs. Only one JDC, Fukunishiki, expressed the higher resistance in the LN population, referring the key R gene Piz as the promising one in the local breeding program. Together, the key R genes, Pi1 and unknown ones in cv. Zhenlong 13 could be applied in rice breeding programs in both northeastern regions, and Piz and probably Pib (73.3% in HLJ) were as the specific ones for LN and HLJ, respectively.

The CDC Set is Still Suitable to Chinese Populations of Mo

Both the CDC and JDC differential sets were used to explore the race structure of four Chinese Mo populations, allowing conclusions to be drawn regarding race diversity (Table 1), specific race structure (Table 2), and dominant race structure (Table 3). The clearest patterns were obtained from the observation of the host reaction of the members of the CDC set, where race frequency, the race diversity index, the specific race isolate frequency, and the frequency of the three predominant races were fully concordant with one another. On this basis, it was apparent that the southern and northeastern Chinese production regions support two quite distinct populations of the pathogen, with the southern populations being more diverse than the northeastern ones. The biological basis for this difference may well be that a wider range of cultivars is grown in southern as opposed to in northeastern China, which would tend to promote the diversity of the pathogen population (Wu et al. 2014; Zhang et al. 2015). The race structure predicted by the host reaction of members of the JDC set was less informative, even though this set was able to identify a greater number of races (Tables 1, 2 and 3). Following the introduction to China of the JDC set in the early 1980s, numerous attempts have been made to compare the efficacy of the two sets and to test whether their use leads to similar conclusions (Shen et al. 1986; Ling et al. 1989; Lei et al. 2000; Wu et al. 2004; Yang et al. 2004; Shi et al. 2015). The consensus has been that the JDC set is superior, largely because it is more effective in discriminating between Mo races (Shen et al. 1986; Ling et al. 1989; Lei et al. 2000). There might be two missing points: one is that the JDC set consists of 12 cultivars that carry different resistance genes, and the CDC set just seven cultivars that involve unknown R genes, which resulted in a certain difference between the sums of races identified by the two sets in a given population; and secondly, such attempts did not dissect race structures with various aspects and parameters for detailed comparison as shown in the present study. Taking consideration into the disease management and resistance cultivar breeding, the race structures established by a series of consistent, logical, and accessible parameters should be more important to rice pathologists and breeders. The coming question is whether the ideal differential set should take an ingenious balance between capacity and conformity (fitness). That would be addressed by a comprehensive comparison of several differential sets, probably including international and monogenic line sets, in more Mo populations.

A striking outcome of the present investigation was the identification of regionally dominant races (Table 3), which highlights the risk of potential crisis for blast outbreak in the target region (Zhu et al. 2000). This is especially the case for the northeastern region, where the combined frequency of the top three dominant isolates was > 80%.

Some Key Resistance Genes Still Effectively Protect Chinese Rice Production

The management of disease should be bound to the utilization of genetic resistance of the host plants (He et al. 2012; Zhai et al. 2014; Deng et al. 2017; Wang et al. 2017). It is, therefore, required to have a certain and exact information on what resistance gene(s) can be used to preserve and control the regionally specific race structures (Chen et al. 2001; Wang et al. 2017; Xing et al. 2017). The differential cultivars were, in the first place, being evaluated as the useful resistance gene donors for the local breeding programs. Of the cultivars included in the CDC set, only Tetep and Zhenlong 13 showed a broad spectrum of resistance to rice blast in all four production areas (Table 4). Two further cultivars expressed a high level of resistance in three of the four populations (Dongnong 363 in HN, LN, and HLJ; Fukunishiki in GD, HN, and LN). The cultivars Kanto 51, Kusabue, and Tsuyuake, which each carry Pik alleles, proved to be more resistant in the southern regions than those in the northeast, while Sifeng 43 behaved in the opposite manner. Two cultivars were typed as highly resistant in just one province (K1 in GD and K60 in HN). By comparing resistance frequencies of the two differential sets that carry the same key resistance genes, the four genes, Pi1, Pik, Pik-m, and Piz, along with the as yet non-characterized genes harbored by Zhenlong 13, appear to be appropriate for rice breeding programs in either GD or HN, while Pi1 and unknown ones in Zhenlong 13 in both northeastern rice breeding programs. The Pita, Pik-p, Piz, and Pib genes might be the specific ones for GD, HN, LN, and HLJ programs, respectively (Table 4).

As per gene-for-gene principal, resistance gene(s) identified in a given cultivar is depended on not only the resistance gene composition of the host, but also the avirulence genes involved in the isolates being tested. As to the JDCs, the first additional gene, Pik-s, which was carried by the Japanese susceptible cultivar Shin 2, is recognized by its interaction with the Mo isolate Ken Ph-03 from the Philippines (Kiyosawa 1969). The second one, Pish, present in seven of the JDCs, is recognized by a race isolated from the former resistance cultivar. The cultivar Reiho carries Pita-2 (Imbe and Matsumoto 1985). The third one, Pi19, present in 11 of the 12 JDC entries, and is recognized by the Chinese Mo isolate CHNOS58–3-1 (Hayashi et al. 1998). Similarly, three additional genes, namely Pi1, Pi54 (formerly Pik-h) borne by Tetep, and Pik-l by LTH, have been identified using races isolated from either China or India (Rai et al. 2011; Hua et al. 2012; Singh et al. 2015). It was noteworthy that the cultivar Tetep has been recognized as being one of the most important donors of multiple blast resistance genes (Mackill and Bonman 1992; Hua et al. 2012), while LTH is classed as a highly susceptible cultivar for the development of monogenic lines which harbors just a single blast resistance gene (Tsunematsu et al. 2000; Kawasaki-Tanaka et al. 2016). Another striking outcome of the present study is that the CDC set harbors at least five additional blast resistance genes (here assigned the temporary designations α through ε). A comparison of the resistance profiles of Sifeng 43 and BL1, which are both known to carry Pib, showed that the former is resistant to a greater number of the isolates from the LN and HLJ populations. This is presumed to reflect the presence of α. Similarly, the existence of the genes β, γ, and δ was inferred from the resistance profiles of, respectively, Dongnong 363 in LN, Kanto 51 in GD, and Hejiang 18 in both GD and HN (Table 4). The unidentified blast resistance genes harbored by Zhenlong 13 produced a rather similar resistance profile to that of Dongnong 363, except for isolates sampled from GD; the implication was that the blast resistance gene content of Zhenlong 13 is the same as that of Dongnong 363, with the addition of the gene ε, which conditions resistance to some GD isolates. Owning more specific and effective resistance genes in the CDC set that might be another key reason why it was more suitable for Chinese Mo populations, compared with the JDC set (Table 4). Similarly, the abundant resistance genes involving in the international differential cultivars, especially Raminad Str. 3, Zenith, NP125, Dular, and Sha-tiao-tsao, which was also a key determinant for its long-term use at least in America (Wang et al. 2017).

The next step forward will be to confirm such additional resistance genes through genetic approaches. This would make a certain contribution to clarify the gene compositions of the CDCs, and then provide more appropriate gene resources for the local rice breeding programs.