Introduction
Leaf rust caused by Puccinia triticina Eriks., is less damaging compared to stripe rust (caused by P. striiformis West.) and stem rust (caused by P. graminis f. sp. tritici) but due to its frequent and widespread occurrence it causes greater annual losses in bread wheat (T. aestivum L. em. Thell). Development and deployment of host genetic resistance is the most effective and economical approach to combat this disease. Although more than 78 leaf rust resistance genes (Lr1 to Lr78) identified from cultivated wheat and its progenitor and non-progenitor species have been catalogued (McIntosh et al., Reference McIntosh, Dubcovsky, Rogers, Morris and Xia2017; Kolmer et al., Reference Kolmer, Bernardo, Bai, Hayden and Chao2018), leaf rust losses are occurring continuously worldwide as most of these genes deployed by breeders condition a hypersensitive reaction and are amenable to frequent breakdowns due to emergence of new virulences.
Resistance to a particular disease is attributed by genetic control, race-specificity and durability. The durability of resistance is related to the activation of haustorium formation-based resistance mechanism(s) (Niks and Rubiales, Reference Niks and Rubiales2002; Singh et al., Reference Singh, Huerta-Espino and William2005; Madrid et al., Reference Madrid, Rubiales, Moral, Moreno, Millán, Gil and Rubio2008). Based on histological observations two types of resistance i.e. pre-haustorial and post-haustorial have been described in diploid wheat (Niks and Dekens, Reference Niks and Dekens1991). Pre-haustorial resistance prevents the formation of haustoria by the fungus despite the normal haustorium mother cells (HMCs) formation and a papilla often induced at the site of attempted cell wall penetration (Heath, Reference Heath1981; Jacobs, Reference Jacobs1989b; Niks and Dekens, Reference Niks and Dekens1991). Pre-haustorial resistance is very common in non-host interactions as it reacts non-specifically to the different pathotypes (Heath, Reference Heath1977, Reference Heath, Groth and Bushnell1985), thus is expected to be long lasting (Niks and Dekens, Reference Niks and Dekens1991). This type of resistance does not allow the formation of hypersensitivity reaction or sporulating pustules, thus, imposing less selection pressure on the pathogen. In general, race-specific hypersensitivity resistance is post-haustorial.
Wheat wild relative, T. monococcum, has high levels of resistance to the wheat leaf rust fungus, P. triticina. Pre-haustorial and post-haustorial resistance to leaf rust has been studied and reported in various T. monococcum accessions (Niks and Dekens, Reference Niks and Dekens1991; Jacobs et al., Reference Jacobs, Pretorius, Kloppers and Cox1996; Anker and Niks, Reference Anker and Niks2001; Lind, Reference Lind, Lipman, Maggioni, Knupffer, Ellis, Leggett, Kleijer, Faberova and Blanc2005). Serfling et al. (Reference Serfling, Templer, Winter and Ordon2016) reported pre-haustorial resistance in Einkorn (T. monococcum accessions) against P. triticina.
One of the leaf rust-resistant recombinant inbred lines, designated as RIL101, derived from a cross involving A-genome wild wheat germplasm accessions T. boeoticum acc. pau5088 and T. monococcum acc. pau14087 at Punjab Agricultural University, Ludhiana showed a high level of resistance to various diseases including leaf rust, and was used to transfer rust resistance to the hexaploid wheat background. RIL101 was crossed to leaf rust susceptible T. aestivum cv WL711 using T. durum cv N59 as a bridging parent (Singh et al., Reference Singh, Chhuneja, Ghai, Kaur, Goel, Bains, Keller, Dhaliwal, Buck, Nisi and Solomon2007). The inheritance of leaf rust resistance from wild relatives has been worked out in the backcross populations derived from the cross N59/RIL101//3*WL711 and leaf rust resistance was found to be effective throughout the plant life (All stage resistance: ASR) (Sandhu, Reference Sandhu2013). The present investigation was carried out to determine the mechanism of resistance in the T. monococcum and T. boeoticum parental accessions used to transfer leaf rust resistance to bread wheat background.
Materials and methods
Plant material
The plant material used for studying the mechanism of resistance for leaf rust consisted of diploid wheat wild relatives i.e. T. monococcum acc. pau14087 (hereafter referred to as T. monococcum), T. boeoticum acc. pau5088 (hereafter referred to as T. boeoticum), Tb5088/Tm14087 derived RIL101 and leaf rust susceptible T. aestivum cv WL711. The detailed information on T. monococcum and T. boeoticum accessions used in these studies and crossing scheme to generate BC1F8 introgression lines has been presented in Singh et al. (Reference Singh, Chhuneja, Ghai, Kaur, Goel, Bains, Keller, Dhaliwal, Buck, Nisi and Solomon2007) and Chhuneja et al. (Reference Chhuneja, Kaur, Garg, Ghai, Kaur, Prashar, Bains, Goel, Keller, Dhaliwal and Singh2008). Seedling and adult plant stage screening of these introgression lines identified leaf rust resistance gene(s), which was effective at all growth stages.
Screening against leaf rust pathotypes
The plant material was screened against different P. triticina pathotypes at the adult plant stage at the Institute of Plant Biology, University of Zurich, Zurich. Three P. triticina pathotypes including two Swiss pathotypes (97512-11, 97512-19) and a Mexican pathotype (TCB/TD) showed a prominent reaction on these germplasm lines and were selected for further studies. The avirulence-virulence formula for TCB/TD is avirulent to Lr 3ka, 9, 10, 11, 16, 17, 19, 21, 24, 25, 27 + 31, 29, 30, 32, 33 and virulent to 1, 2a, 2b, 2c, 3, 3bg, 13, 14a, 14b, 15, 18, 20, 23, 26, 28. Swiss pathotype 97512-19 is avirulent to Lr 10. The germplasm lines T. monococcum, T. boeoticum, RIL101 and WL711 were inoculated with rust at Zadoks growth stages 37–39 when most of the flag leaves developed using a fine mist of P. triticina urediniospores suspended in Tween 20 (as a surfactant). Infected plants were covered with a polythene hood to maintain high humidity and transferred to a growth chamber with a temperature of 16 °C and 90% humidity for 24 h in the dark. The climate conditions of the chamber were then changed to 25 °C/20 °C for 16 h/8 h day/night and 70% humidity. Leaf samples inoculated with each pathotype were harvested on the third, fifth, seventh, tenth and twelfth-day post-inoculation, and fixed for further analysis. For each genotype, three inoculated leaves were sampled. Disease severity was recorded as the percentage of leaf area covered by rust, following a modified Cobb's scale (Peterson et al., Reference Peterson, Campbell and Hannah1948) and infection type viz. 0 = immune, TR = traces of resistance, MR = moderately resistant, MS = moderately susceptible and S = susceptible.
Sample preparation
The samples for fluorescent microscopic analysis were prepared according to Moldenhauer et al. (Reference Moldenhauer, Moerschbacher and Van der2006). The flag leaf sections (±3 cm long) were fixed and cleared in ethanol: trichloromethane (3:1) solution containing 0.15% trichloroacetic acid for 24 h. The samples retrieved from fixative were washed with 50% ethanol and incubated in 0.5 M sodium hydroxide at 90 °C in an incubator for 30 min. After washing twice with water, the leaf samples were treated with 0.1 m Tris HCl buffer (pH 5.8) for 30 min. For staining, the leaf samples were soaked in 0.1% Uvitex 2B (Ciba-Geigy) for 5 min, rinsed four times (10 min each) with water and finally with 25% Glycerol for 15 min. The material was then stored in 50% glycerol at 4 °C until further use.
Microscopic observation
The Uvitex stained leaf samples of T. monococcum, T. boeoticum, RIL101 and WL711 were analysed under a fluorescent microscope for studying the leaf rust infection against two pathotypes i.e. TCB/TD and 97512-19 that showed differential phenotypic expressions. About 2 cm of leaf segments within the 2–10 cm region from tip was excised and examined using a Nikon Labophot fluorescence microscope equipped with epifluorescence optics. A UV-1A filter (excitation filter 330–380 nm, barrier filter 420 nm) was used to visualize fungal structures by their light blue fluorescence. The endophytically growing fungal infection structures in the leaf were observed using a confocal laser scanning microscope (Leica Microsystems) when excited with UV-laser beams at 351 and 364 nm and scanned with filter settings at 400–500 nm. The necrotic mesophyll cells were visualized with laser excitation at 514 and 543 nm and scanned with filter settings at 560–680 nm. The infection units in each leaf segment were recorded for four parameters (pre-stomatal exclusion of fungus, aborted penetration of infection units, early abortion and fully established mycelia colonies) as studied by Jacobs et al. (Reference Jacobs, Pretorius, Kloppers and Cox1996). Each leaf segment was screened for all the cells having infection structures which varied from 10 to 50.
Results
Phenotypic analysis
The phenotypic evaluation of T. monococcum and T. boeoticum at the adult plant stage showed resistance against the P. triticina pathotypes tested in this study, although the level of resistance varied in the two wild wheat accessions. T. monococcum produced nearly immune and immune reaction against TCB-TD and 97512-19, respectively. T. monococcum showed flecking with 5 MR type of reaction against pathotype 97512-11 (Fig. 1, Table 1). In T. boeoticum, the terminal disease severity was 10–20 MR against TCB-TD and 97512-11; however, a susceptible reaction was observed against 97512-19. In response to 97512-19, T. boeoticum showed medium to large size pustules surrounded by large chlorotic areas with more than 40% of leaf coverage (Table 1). RIL101 also showed a disease score of 10–20 MR to pathotypes TCB-TD and 97512-11 whereas gave a susceptible reaction against 97512-19. WL711, which is susceptible to Indian Pt pathotypes, exhibited a moderate type of susceptible reaction with some chlorosis against all the three pathotypes (Fig. 1, Table 1).
Adult plant leaf rust reaction (left to right) of T. monococcum (Tm), T. boeoticum (Tb), RIL101 and WL711 against (a) Pt pathotype TCB/TD; (b) Pt pathotype 97512-19 and (c) Pt pathotype 97512-11.
Leaf rust severity in T. monococcum, T. boeoticum, RIL101 and WL711 against three Puccinia triticina isolates used in the study
Disease severity recorded as infection type viz. 0 = immune, TR = traces of resistance, MR = moderately resistant, MS = moderately susceptible and S = susceptible (as modified Cobb's scale).
Microscopic investigation of haustorium formation
The progress of the infection process for rust pathotypes TCB/TD and 97512-19 was observed in T. monococcum, T. boeoticum, RIL101 and WL711 (Figs. 2 and 3).
Development of infection units of Puccinia triticina pathotype TCB-TD in T. monococcum (Tm), T. boeoticum (Tb), RIL101 and WL711 at 3, 5, 7, 10 days post-infection under Nikon Labophot fluorescence microscope. The growing fungal structures (SP = spore, GT = germ tube, AP = appressorium, IP = infection peg, SSV = sub stomatal vesicle, IH = infection hypha, HMC = haustorial mother cell, SH = secondary hyphae, MC = mycelial colonies) are depicted in blue (excitation 351/364 nm, filter 400–500 nm) and the necrotic mesophyll cells (HCN = host cell necrosis) are depicted in red (laser excitation 514/543 nm, filter 560–680 nm).
Development of infection units of Puccinia triticina pathotype 97512-19 in T. monococcum (Tm), T. boeoticum (Tb), RIL101 and WL711 at 3, 5, 7, 10 days post-infection under Nikon Labophot fluorescence microscope. The growing fungal structures (SP = spore, GT = germ tube, AP = appressorium, IP = infection peg, SSV = sub stomatal vesicle, IH = infection hypha, HMC = haustorial mother cell, SH = secondary hyphae, MC = mycelial colonies) are depicted in blue (excitation 351/364 nm, filter 400–500 nm) and the necrotic mesophyll cells (HCN = host cell necrosis) are depicted in red (laser excitation 514/543 nm, filter 560–680 nm).
Mechanism of infection against P. triticina pathotype TCB/TD:
On infection with TCB/TD, there was no pre-stomatal exclusion observed in any of the lines tested at 3 dpi as all the germinated spores formed appressoria over the stomata (Fig. 2). A few cells in the parental accessions (T. monococcum, T. boeoticum) and RIL101 showed aborted penetration of the fungi. T. monococcum significantly differed from T. boeoticum for the presence of infection sites with early aborted structures (Fig. 2). T. boeoticum had an almost negligible early abortion with the presence of numerous HMCs per infection site. Secondary infection hyphae were also observed in some of the infection sites. When observed at 5 dpi, secondary infection with HMCs accompanied by host cell necrosis (HCN) was very frequently recorded in T. monococcum. Growing hyphae in T. boeoticum and RIL101 did not exhibit any hypersensitive response at this stage. Two days later (7 dpi) HCN was also present at some sites in T. boeoticum and RIL101, but the growth of hyphal colonies was not much affected. Compared to this, HCN had spread restricting the growth of fungal structures in T. monococcum. There was complete cell death at the infection site and thus no mycelial colonies were developed. This difference was more significantly noticed at 10 dpi. The high hypersensitive response was observed at the site of infection in three genotypes (T. monococcum, T. boeoticum and RIL101) at this stage. Fully established mycelial colonies were recorded in T. boeoticum and RIL101 which were rather more pronounced in T. boeoticum at 12 dpi (data not given). No fungal growth occurred beyond the necrotic areas in T. monococcum at 12 dpi (data not given). The susceptible check, WL711, had numerous HMCs at third day post-infection which fully developed into mycelial colonies at further stages post-infection despite some necrosis at the fifth- and seventh-day post-infection.
Mechanism of infection against P. triticina against pathotype Pt 97512-19
Infection with pathotypes Pt 97512-19 also showed a similar pattern of fungal growth in the leaf samples as observed in TCB/TD (Fig. 3). The hyphal growth progressed at a very slow rate in T. monococcum. T. monococcum developed 1-3 HMCs 3 dpi which increased at the second sampling stage. Although, the HMC counts increased in some infection sites in T. monococcum but no further growth of the fungus was observed in the leaf segments sampled until 10 dpi. The infection sites at 7, 10 and 12 dpi were accompanied by necrosis of the cells. On the other hand, in T. boeoticum HMC count were more than six at the first sampling stage followed by complete hyphal network formation on the fifth day of infection. The fungal growth increased in T. boeoticum forming colonies of hyphae without any HCN. The infection process in RIL101 was similar to T. monococcum with 1–3 HMCs at the first sampling stage. But it behaved more like T. boeoticum in the later stages with complete hyphal formation at 7 dpi. However, some HCN was observed at 10 dpi. The susceptible check, WL711 showed the presence of numerous HMCs with infection hyphae at third day post-infection which developed into mycelial colonies in later stages. Some HCN was also observed at 5 and 7 dpi.
Discussion
Diploid wild wheat parental accessions (T. monococcum and T. boeoticum), RIL101 and WL711 were examined phenotypically and microscopically for resistance or susceptibility, and classified based on phases of development of the infection units. Disease reaction of these lines indicated that T. monococcum and T. boeoticum have different levels of resistance against Swiss and Mexican pathotypes. T. monococcum and T. boeoticum are closely related diploid wild wheat species but their host-status for the wheat leaf rust fungus is different (Anker et al., Reference Anker, Bantjer and Niks2001). The present study results also suggests that T. monococcum and T. boeoticum accessions have different genes conferring resistance to leaf rust. The gene (s) from T. monococcum is/are conferring a high level of resistance against both the pathotypes tested. T. boeoticum and RIL101, on the other hand, has the gene which is showing moderate resistance to these pathotypes. The low level of resistance in RIL101 could be due to the fact that this RIL carries leaf rust resistance gene from T. boeoticum. However, it was not feasible to determine how many leaf rust resistance genes are present in T. monococcum and T. boeoticum as the RIL population from this cross is completely resistant against all of the pathotypes that were tested (data not published).
Different fungal structures were observed when the development process of two rust pathotypes TCB/TD and 97512-19 were studied in T. monococcum, T. boeoticum, RIL101 and WL711. All the fungal growth stages including spore germination to sporulation can occur within a 7–10 days period at optimum and constant temperatures. The uredospore germinates forming a germ tube that continues to grow along the leaf surface until a stoma is reached or the endogenous spore reserves are depleted (Dickinson, Reference Dickinson1969). On reaching the stoma, the appressorium is formed within 24 h of infection. The penetration peg from the appressorium enters the stoma and a sub stomatal vesicle is formed in the intercellular space within-host leaf. Thereafter, infection hyphae begin to grow from this sub-stomal vesicle towards the mesophyll cells (Allen, Reference Allen1926). The haustorial mother cell (HMC) develops against the mesophyll cell and subsequently, haustoria is formed in the host cell. Secondary infection hyphae are produced, which come in contact with other host cells resulting in additional HMC and haustoria, thus forming a branched network of fungal mycelium (Allen, Reference Allen1926).
T. monococcum showed a post-haustorial type of resistance to the Mexican pathotype TCB-TD and pre-haustorial resistance against the Swiss pathotype 97512-19. The first sampling was carried out at 3 dpi and hence the initial fungal penetration in the plant material was not studied. The resistance mechanisms operating in hosts and non-host genotypes do not affect the spore germination, appressorium formation, stomata penetration and sub stomatal vesicle formation (Jacobs, Reference Jacobs1989a). Lee and Shaner (Reference Lee and Shaner1984) also observed no significant difference in pre-penetration stages between the susceptible and slow leaf rusting wheat genotypes.
Histological observations revealed a significant difference in the development of the two P. triticina pathotypes (TCB/TD and 97512-19) in T. monococcum. The development of pathotype TCB/TD in T. monococcum was initially slow but haustoria mother cells soon developed into hyphae. However, further fungal growth was completely restricted by necrosis of the host cell, thus suggesting that the gene (s) in this accession confer a hypersensitive type of resistance to leaf rust. On the other hand, the resistance to 97512-19 in T. monococcum appeared to be pre-haustorial as the HMCs did not form any haustoria or primary hyphae till 10 dpi which are essential for effective infection. Moreover, necrotic areas surrounding HMCs often restricted the further growth of fungus, thus inhibiting colonization. This differential development of two pathotypes in T. monococcum could be due to different genes conferring specific resistance to these pathotypes.
The mechanism of resistance in T. boeoticum against both TCB/TD and 97512-19 was post-haustorial, i.e. a hypersensitive response elicited after the formation of a haustorium. Although the resistance response in T. boeoticum was quite similar for the two pathotypes, the level of resistance varied against these pathotypes. The HCN was observed more frequently in the infection sites tested with TCB-TD. Despite HCN, the mycelial colonies were successfully formed in T. boeoticum at 7 dpi in TCB-TD and at 5 dpi in 97512-19. Against 97512-19, only small necrotic areas relative to large colonies were measured in T. boeoticum. These results confirm phenotypically observed moderate resistance to P. triticina pathotypes TCB-TD and 97512-19 in T. boeoticum. Serfling et al. (Reference Serfling, Templer, Winter and Ordon2016) also observed weak post haustorial resistance reaction with weak signs of autofluorescence 3 dpi in the phenotypically susceptible T. boeoticum accession 36554 against the isolate under study. The dissimilarity in host−pathotype interaction could be the reason for the higher proportion of infection sites with HCN in TCB-TD than in 97512-19. Kloppers and Pretorius (Reference Kloppers and Pretorius1995) also observed a significant difference in the number of colonies for South African pathotypes tested against Lr37 in RL6081. Similarly, in a study by Bender et al. (Reference Bender, Pretorius, Kloppers and Spies2000), infection sites with HCN were more frequent in the lines with Lr12 and Lr13 against pathotype UVPrt2 compared to UVPrt13.
As in T. boeoticum, the infection sites with established colonies were observed in Tb5088/Tm14087 derived RIL101 but a hypersensitive response was more frequent, thus indicating a post-haustorial mechanism of resistance. T. aestivum cv WL711, susceptible to most Indian pathotypes, showed some hypersensitive response despite of complete colonization at very early stages of infection. Terminal disease severity recorded in WL711 against TCB/TD (virulent to Lr13) was 40 MS (Table 1, Fig. 1a). HCN was observed in WL711 at fifth- and seventh-day post-infection, which further showed fully developed colonies at tenth-day post-infection (Fig. 2).
A few necrotic cells in WL711 could be attributed to stress due to different environmental conditions or host−pathogen interaction or due to the presence of Lr13 gene (Agarwal et al., Reference Agarwal, Saini and Sharma2003). As suggested by Samborski et al. (Reference Samborski, Kim, Rohringer, Howes and Baker1977), HCN in the compatible host−pathogen interactions may result from a general stress reaction. Susceptibility is not necessarily always associated with the absence of necrotic cells at infection sites (Bender et al., Reference Bender, Pretorius, Kloppers and Spies2000).
The mechanism of rust resistance in diploid wild wheat ranges from pre-haustorial resistance without necrosis to post-haustorial resistance with frequent necrosis. Niks and Dekens (Reference Niks and Dekens1991) reported high levels of pre-haustorial resistance in two out of three accessions of T. monococcum and post-haustorial resistance in a T. boeoticum accession. Among the 152 resistant T. monococcum accessions reported by Anker and Niks (Reference Anker and Niks2001), only three accessions were found to have a high percentage of pre-haustorial resistance to P. triticina pathotype Felix. Lind (Reference Lind, Lipman, Maggioni, Knupffer, Ellis, Leggett, Kleijer, Faberova and Blanc2005) screened 348 accessions of T. monococcum, obtained from different gene banks, for leaf rust resistance and detected 13 accessions (3.7%) with pre-haustorial resistance. Serfling et al. (Reference Serfling, Templer, Winter and Ordon2016) reported pre-haustorial resistance in T. monococcum accession with complete inhibition of generation of haustoria and weak post-haustorial resistance in T. boeoticum accession. Our histology study reported both pre-haustorial and post-haustorial resistance in T. monococcum accession pau14087 and post-haustorial resistance in T. boeoticum accession pau5088 against both the pathotypes, thereby justifying the differential defence response to leaf rust pathogen in the two wild wheat accessions.
T. monococcum acc. pau14087 conferring a pre-haustorial kind of resistance to one of the two pathotypes tested could prove a durable source of resistance in rust resistance breeding. Pre-haustorial resistance in T. monococcum can be used to replace the currently widely used single gene and race-specific resistance against leaf rust in wheat which is at constant risk of being overcome by the emergence of new virulent races. The present study describing the development of fungal structure and host−pathogen response may assist in understanding the molecular genetic aspects of the genes transferred from wheat wild relatives (T. monococcum and T. boeoticum) into the hexaploid wheat background. Thus, the leaf rust resistance genes in T. monococcum and T. boeoticum will be different with different mechanisms of resistance. We have already initiated the independent transfer of leaf rust resistance from T. monococcum and T. boeoticum into hexaploid wheat (T. aestivum) using T. durum as a bridging species. Once transferred independently, a pyramiding of the two leaf rust resistance genes through marker-assisted selection may result in durable resistance, because of their different mechanisms of infection.
Acknowledgements
This work was carried under Indo-Swiss collaboration in Biotechnology and Monsanto's Beachell Borlaug Fellowship administered by Agri Life Research – Texas A&M University, USA. The Financial support provided by the Department of Biotechnology, Ministry of Science and Technology, Government of India, Swiss Agency for Development and Cooperation, under Indo-Swiss collaboration in Biotechnology and Monsanto's Beachell Borlaug Fellowship (AKR) is gratefully acknowledged. We are thankful to the Institute of Plant Biology, UZH for providing the rust inoculums and technical assistance provided by the Center for Microscopy and Image Analysis, UZH.
