Theoretical and Applied Genetics

, Volume 122, Issue 1, pp 239–249

New slow-rusting leaf rust and stripe rust resistance genes Lr67 and Yr46 in wheat are pleiotropic or closely linked

Authors

  • Sybil A. Herrera-Foessel
    • International Maize and Wheat Improvement Center (CIMMYT)
  • Evans S. Lagudah
    • CSIRO Plant Industry
  • Julio Huerta-Espino
    • Campo Experimental Valle de México INIFAP
  • Matthew J. Hayden
    • Department of Primary IndustriesVictorian AgriBiosciences Center
  • Harbans S. Bariana
    • The University of Sydney Plant Breeding Institute-Cobbitty
  • Davinder Singh
    • CIMMYT, ICRAF House
    • International Maize and Wheat Improvement Center (CIMMYT)
Original Paper

DOI: 10.1007/s00122-010-1439-x

Cite this article as:
Herrera-Foessel, S.A., Lagudah, E.S., Huerta-Espino, J. et al. Theor Appl Genet (2011) 122: 239. doi:10.1007/s00122-010-1439-x

Abstract

The common wheat genotype ‘RL6077’ was believed to carry the gene Lr34/Yr18 that confers slow-rusting adult plant resistance (APR) to leaf rust and stripe rust but located to a different chromosome through inter-chromosomal reciprocal translocation. However, haplotyping using the cloned Lr34/Yr18 diagnostic marker and the complete sequencing of the gene indicated Lr34/Yr18 is absent in RL6077. We crossed RL6077 with the susceptible parent ‘Avocet’ and developed F3, F4 and F6 populations from photoperiod-insensitive F3 lines that were segregating for resistance to leaf rust and stripe rust. The populations were characterized for leaf rust resistance at two Mexican sites, Cd. Obregon during the 2008–2009 and 2009–2010 crop seasons, and El Batan during 2009, and for stripe rust resistance at Toluca, a third Mexican site, during 2009. The F3 population was also evaluated for stripe rust resistance at Cobbitty, Australia, during 2009. Most lines had correlated responses to leaf rust and stripe rust, indicating that either the same gene, or closely linked genes, confers resistance to both diseases. Molecular mapping using microsatellites led to the identification of five markers (Xgwm165, Xgwm192, Xcfd71, Xbarc98 and Xcfd23) on chromosome 4DL that are associated with this gene(s), with the closest markers being located at 0.4 cM. In a parallel study in Canada using a Thatcher × RL6077 F3 population, the same leaf rust resistance gene was designated as Lr67 and mapped to the same chromosomal region. The pleiotropic, or closely linked, gene derived from RL6077 that conferred stripe rust resistance in this study was designated as Yr46. The slow-rusting gene(s) Lr67/Yr46 can be utilized in combination with other slow-rusting genes to develop high levels of durable APR to leaf rust and stripe rust in wheat.

Introduction

Leaf (or brown) rust and stripe (or yellow) rust, caused by Puccinia triticina and P. striiformis, respectively, are important constraints to common wheat (Triticum aestivum) production in most parts of the world. Growing resistant cultivars is the most effective and environmentally sound strategy for controlling rust diseases. Over 60 leaf rust and 40 stripe rust resistance genes are cataloged in wheat (McIntosh et al. 2008). Most of these resistance genes are race-specific and therefore succumb to new variants of the respective pathogen soon after their deployment. In contrast, slow-rusting resistance characterized by slow disease progress in the field despite a compatible host reaction (Caldwell 1968) is known to be durable. Although slow-rusting resistance genes have small-to-intermediate effects when present alone, high levels of resistance have been achieved by combining 4–5 such genes (Singh et al. 2000a). The slow-rusting gene Lr34/Yr18, located on chromosome arm 7DS, has provided durable resistance to leaf rust and stripe rust since the early twentieth century (Dyck 1977; 1987; Singh 1992a).Lr34/Yr18 also confers resistance to powdery mildew (Blumeria graminis) (Spielmeyer et al. 2005), stem rust (Puccinia graminis tritici) (Dyck 1987), and barley yellow dwarf virus (Singh 1993). A morphological characteristic of Lr34/Yr18 is its association with leaf tip necrosis (LTN) in adult plant stage, which can be used to identify it in certain environments (Singh 1992b). Recent cloning of Lr34/Yr18 provided important information on its gene structure, which encodes a putative ATP-binding cassette (ABC) transporter and enabled the development of gene-based markers that facilitated the identification of Lr34/Yr18 in different wheat backgrounds (Krattinger et al. 2009; Lagudah et al. 2009). Another gene that has a dual effect and confers slow-rusting resistance to leaf rust and stripe rust is Lr46/Yr29, which has also been shown to provide durable resistance in gene combinations (Singh et al. 2005).

‘RL6077’ (Tc*6/PI250413) is a near-isogenic line of ‘Thatcher’ that carries a slow-rusting leaf rust resistance gene transferred from accession PI250413 collected in Pakistan (Dyck and Samborski 1979). This gene was confirmed to be present in wheat accessions from Iraq, Afghanistan, and India, based on allelism tests indicating that it is not uncommon in germplasm from that region (Dyck and Samborski 1979; Shang et al. 1986). The leaf rust resistance gene in RL6077 also had an effect on stripe rust and was associated with LTN in the field (Dyck and Samborski 1979; Singh 1992a; Dyck et al. 1994). Similarities between this gene and Lr34 led Dyck et al. (1994) to conclude that the resistance gene in RL6077 was likely to be Lr34, even though it is inherited independently. One of the reasons behind this conclusion was the presence of quadrivalents in pollen mother cells of F1 hybrids of RL6077 and ‘RL6058’ (an Lr34/Yr18 near-isogenic line) in cytological examinations indicating that Lr34/Yr18 could have been translocated in a reciprocal manner to a different chromosome in RL6077 (Dyck et al. 1994). The presence of Lr34/Yr18 in a Thatcher background is known to exhibit increased stem rust resistance to a number of isolates (Dyck 1987). A similar observation of the gene in RL6077 against stem rust was another contributing factor that led Dyck et al. (1994) to postulate the presence of Lr34/Yr18 in RL6077. Until recently, this hypothesis was considered untenable; however, recent analysis of the diagnostic region in RL6077 from the ABC transporter encoded by the Lr34/Yr18 gene showed the absence of Lr34/Yr18 (Lagudah et al. 2009). Consequently, these results have prompted the mapping of the leaf rust and stripe rust resistance gene(s) in RL6077.

In a parallel study conducted with Canadian leaf rust isolates, Hiebert et al. (2010) mapped the leaf rust resistance in RL6077 to chromosome 4D from two populations (Thatcher × RL6077 and RL6058 × RL6077), leading to the designation of this gene as Lr67. In our study, we confirmed their finding using a different cross and a larger population, and investigated the genetic association of Lr67 with stripe rust resistance, thereby providing evidence for a co-segregating gene designated as Yr46 on chromosome arm 4DL. We also show through the comparison of the complete sequence of the ABC transporter gene at the Lr34/Yr18 locus that RL6077 definitely does not carry the haplotype associated with Lr34/Yr18.

We also tested Thatcher, RL6077 (Thatcher + Lr67/Yr46), RL6058 (Thatcher + Lr34/Yr18) and 90RN2491 (Thatcher + Lr34/Yr18 + Lr67/Yr46) against Ug99 stem rust in field tests in Kenya, since increased resistance of Thatcher and Thatcher derivatives in the presence of the Lr34/Yr18 gene against the highly virulent Ug99 strain has been reported (Gavin Vanegas et al. 2008) but the effect of Lr67/Yr46 in Thatcher against Ug99 was unknown.

Materials and methods

Development of mapping populations

An initial F3 mapping population at CSIRO, Canberra, Australia, was generated by crossing RL6077 with a selection of the susceptible cultivar Avocet (Avocet S). This F3 population was used for preliminary screening for leaf rust resistance at Cobbitty, Australia, during 2007; a subset (22 lines) of this population was also evaluated at Cd. Obregon, Mexico, during the 2007–2008 crop season. The F3 population segregated for photoperiod sensitivity, and two different populations—an F5 population with an F4 segregation ratio (hereafter called F4) developed at CIMMYT, and an F4 population with an F3 segregation ratio (hereafter called F3) developed at CSIRO—were subsequently generated using selected photoperiod-insensitive F3 lines. To develop the F4 populations of 148 lines, two photoperiod-insensitive F3 lines that were segregating for resistance were bulk harvested and F4 plants space-planted for individual harvesting. To develop the F3 population of 136 lines at CSIRO, leftover seeds of F3 lines that were segregating according to previous phenotypic evaluations were space-planted and individually harvested. The F4 population developed at CIMMYT was further advanced for two generations, and 148 recombinant inbred lines (RILs) with an F6 genetic ratio were developed.

Field evaluations for characterizing leaf rust and stripe rust resistance

The Avocet × RL6077 F3 and F4 populations from CIMMYT and CSIRO, respectively, were evaluated for leaf rust response in Cd. Obregon, Mexico, during two crop seasons (2008–2009 and 2009–2010). The F4 population was also evaluated for leaf rust at El Batan and for stripe rust in Toluca, Mexico, during the 2009 summer season, and the F3 population was evaluated for stripe rust resistance at Cobbitty, Australia, during the same year. The RILs (F6 Avocet × RL6077 population) were evaluated for leaf rust resistance in Cd. Obregon during the 2009–2010 crop season.

The 148 F4 and the 136 F3 lines of Avocet × RL6077, together with the parents, were grown in Cd. Obregon during the 2008–2009 crop season on different sowing dates (23 November 2008 and 21 January 2009, respectively) and during the 2009–2010 crop season on the same sowing date (18 November 2009). The 148 RILs were sown in Cd. Obregon during the 2009–2010 crop season on 5 December 2009. Plots of 1-m double rows with about 60 plants per line were grown for each population, and spreader rows of the susceptible variety ‘Morocco’ were planted around the experimental area. Morocco was also grown as hill plots on one side of each experimental plot in the middle of a 0.5-m pathway. To initiate the leaf rust epidemics, spreader rows and hills were inoculated using hand-sprayers containing urediniospores of two pathotypes of P. triticina; MCJ/SP and MBJ/SP, suspended in Soltrol Oil (Phillips 66 Co., Bartlesville, OK, USA). In 2008–2009, spreaders and hills around the F4 population were inoculated on 21 and 23 January 2009, whereas spreaders and hills for the F3 population were inoculated on 20 February of that same year. In the 2009–2010 crop cycle, the same inoculation dates (18 and 25 January 2010) were used for all populations. Disease severity on parents was scored according to the modified Cobb Scale where percentage of rusted tissue was visually estimated according to Peterson et al. (1948) and host response to infection was determined according to Roelfs et al. (1992), where for leaf rust, ‘R’ indicated resistant or miniature uredinia surrounded by necrotic tissue, ‘MR’ indicated moderately resistant or smaller to moderate-sized uredinia surrounded by necrotic or chlorotic tissue, ‘MS’ indicated moderately susceptible or moderate-sized uredinia without necrotic or chlorotic tissues, and ‘S’ indicated susceptible or large uredinia without necrotic or chlorotic tissue. Rust response assessments were performed when the susceptible parent Avocet reached 100% leaf rust severity. Lines in each population were scored as homozygous resistant (HR), homozygous susceptible (HS), and segregating (SEG), based on their phenotypic response to leaf rust. In 2009–2010 we evaluated each F4 and F6 line to obtain a score for mean leaf rust severity.

The F4 Avocet × RL6077 population was sown in Toluca and El Batan on 25 and 23 May 2009, respectively. The same plot sizes were used as described earlier for the Cd. Obregon trials. In Toluca, the spreaders consisted of a mixture of six susceptible wheat lines derived from the cross ‘Avocet × Attila’ known to carry the Yr27 stripe rust resistance gene. Mexican P. striiformis isolates Mex96.11 and Mex08.13 were used to inoculate spreaders and hills four times (on 16, 17, 24 and 30 June) to assure the development of the disease epidemic in Toluca. In El Batan, the spreader rows consisted of a mixture of Morocco and two other leaf rust susceptible wheat lines that were inoculated with P. triticina (a mix of pathotypes MCJ/SP and MBJ/SP) on 18 and 22 June 2009. The F4 lines in El Batan were directly inoculated together with the spreaders. The percentage stripe rust severity (Peterson et al. 1948) and the host response to infection (Roelfs et al. 1992) were recorded on the parents, where ‘R’ indicated resistant with necrotic/chlorotic stripes without sporulation, ‘MR’ indicated moderately resistant with necrotic/chlorotic stripes with some sporulation, ‘M’ (or ‘MRMS’) was necrotic/chlorotic stripes with intermediate to abundant sporulation, ‘MS’ indicated moderately susceptible with chlorotic/occasionally necrotic stripes with abundant sporulation, and ‘S’ indicated stripes with or without chlorosis and with abundant sporulation. The F4 lines were scored as HR, HS, and SEG for leaf rust at El Batan and for stripe rust at Toluca, when the susceptible parent Avocet was showing 100S response. The minimum and maximum plant severity response for stripe rust was also recorded for each of the 148 F4 lines at Toluca.

The Avocet × RL6077 F3 population was evaluated for stripe rust resistance under field conditions using a mixture of P. striiformis strains 134E16A+J+Yr27+ and 134E16A+Yr17+ inoculated on spreader wheat rows at Cobbitty, Australia, during 2009. The F3 lines were scored as HR, HS, and SEG response to stripe rust.

Field evaluation of RL6077 against stem rust in Kenya

Wheat genotypes Thatcher, RL6077, RL6058, and 90RN2491—a derivative from the RL6077 × RL6058 cross combining both Lr34/Yr18 and Lr67/Yr46 (Dyck et al. 1994)—were grown in field trials in Njoro, Kenya, in 2009. The prevalent Ug99 derivative race TTKST of P. graminis tritici with virulence for resistance gene Sr24 was used to inoculate spreader rows in the field. Stem rust severity was scored on adult plants using the modified Cobb Scale (Peterson et al. 1948), and host response to infection was evaluated as described in Roelfs et al. (1992), where ‘R’ indicated resistant or miniature uredinia surrounded by necrosis and chlorosis, ‘MR’ indicated moderately resistant or small uredinia surrounded with chlorosis or necrosis, ‘MS’ indicated moderately susceptible or moderate-sized uredinia without chlorosis or necrosis, and ‘S’ indicated susceptible or large uredinia without chlorosis and necrosis.

Comparison of Lr67/Yr46 with other known slow-rusting resistance genes in field and greenhouse tests

Field evaluations and greenhouse tests were conducted on adult plants to investigate the expression of Lr67/Yr46 and compare it with the expression of known slow-rusting resistance genes Lr34/Yr18 and Lr46/Yr29. Plots of single gene lines carrying Lr34/Yr18 (Yr18/3*Avocet) and Lr46/Yr29 (Avocet-YrA*3//Lalbmono1*4/Pavon) were grown in Cd. Obregon during the 2009–2010 crop season and in Toluca in 2009 adjacent to lines from the Avocet × RL6077 populations; leaf and stripe rust responses were recorded when Avocet displayed 100S severity response.

Greenhouse tests were conducted on RL6077 (Lr67/Yr46), RL6058 (Lr34/Yr18), Avocet-YrA*3//LalbMono1*4/Pavon (Lr46/Yr29), Avocet, and two lines from the F3 Avocet × RL6077 population that were HR for the Lr67/Yr46 gene. Eight pots (4 plants/pot) of each genotype were used for the evaluation. Recently emerged, fully expanded flag leaves were inoculated with P. triticina pathotypes MCJ/SP and MBJ/SP, and P. striiformis pathotypes Mex96.11 and Mex08.13. Each race was evaluated separately. Inoculations were conducted by spraying urediniospores suspended in Soltrol oil using an atomizer. Inoculated plants were placed in a dew chamber overnight (P. triticina) or for two nights (P. striiformis) and then transferred to a greenhouse. A data logger (LogTag analyzer, ver. 1.9®) was installed in the greenhouse and programmed to measure the greenhouse temperature every 15 min during the time from when the plants were transferred to the greenhouse after dew exposure to the time when responses were recorded. Minimum, maximum, and average post inoculation temperatures for the leaf rust and stripe rust greenhouse tests were 10, 26, and 18°C, and 10, 25, and 17°C, respectively. Leaf rust infection type was recorded based on a 0–4 Scale (Roelfs et al. 1992) at 11 days post-inoculation, where infection types ‘0’ = no visible symptoms, ‘;’ = necrotic/chlorotic flecks, ‘1’ = small uredinia surrounded by necrosis, ‘2’ = small to medium uredinia surrounded by chlorosis or necrosis, ‘X’ = random distribution of variable-sized uredinia, ‘3’ = medium-sized uredinia without chlorosis, ‘4’ = large-sized uredinia without chlorosis, ‘+’ and ‘−’ were somewhat larger or smaller uredinia than normal for the infection type. Infection type ‘3’ and ‘4’ were considered susceptible while all other infection type was considered resistant. Responses for stripe rust were scored based on a 0–9 Scale (McNeal et al. 1971) at 20 days post-inoculation, where infection type ‘0’ = no visible infection, ‘1’ = necrotic/chlorotic flecks without sporulation, ‘2’ = necrotic/chlorotic stripes without sporulation, ‘3’ = necrotic/chlorotic stripes with trace sporulation, ‘4’ = necrotic/chlorotic stripes with light sporulation, ‘5’ = necrotic/chlorotic stripes with intermediate sporulation, ‘6’ = chlorotic stripes with moderate sporulation, ‘7’ = chlorotic stripes with abundant sporulation, ‘8’ = stripes without chlorosis, moderate sporulation, ‘9’ = stripes without chlorosis and abundant sporulation. Infection types ‘6’ and ‘7’ were considered moderately susceptible, whereas ‘8’ and ‘9’ were considered susceptible, and all other infection type were considered resistant.

Molecular mapping of Lr67/Yr46

Initial genome-wide scans to identify associated simple sequence repeat (SSR) markers were conducted at the Department of Primary Industries, Victorian AgriBiosciences Center, Australia, using a subsample of the original F3 Avocet × RL6077 population. Leaf tissue from the parents, seven HR lines and seven HS lines was harvested and DNA extracted according to the method described by Lagudah et al. (1991b). A collection of SSR markers distributed on all wheat chromosomes was used for screening. Polymerase chain reaction (PCR) products were separated by capillary electrophoresis with an ABI3730xl instrument; allele sizes were determined using GeneScan software, version 3.7 (Applied Biosystems) (Hayden et al. 2008). A total of six SSR markers, Xbarc98, Xbarc288, Xcfd23, Xcfd71, Xwmc48, and Xwmc89 from chromosome 4D differentiated the HR and susceptible genotypes and the parents. To establish genetic linkage and relative marker order with the Lr67/Yr46 locus, markers were first screened on the 148 lines of the Avocet × RL6077- derived F4-lines at CIMMYT’s biotechnology laboratory, and a final linkage map was then generated using the 148 RILs. About 30 seeds per F4 and F6 line were grown in the greenhouse and leaf tissues harvested. DNA was extracted according to a CTAB procedure, and PCR reactions were conducted based on standard methods (CIMMYT 2005) with annealing temperatures according to available information for each marker in graingenes database (http://wheat.pw.usda.gov/GG2/index.shtml). The PCR products were separated on 12% acrylamide gels (29:1), and silver staining was used to visualize the amplification products (CIMMYT 2005). The same markers were also screened in a subset of lines from the F3 Avocet × RL6077 population at CSIRO using the DNA fragment analyzer ABI3730xl. Chromosome 4D short- and long-arm deletion lines, 4DS-01(fragment length FL0.53), 4DS-03 (FL0.67), 4DS-05 (FL0.63), 4DL-09 (FL0.31), 4DL-06 (FL0.38), 4DL-13 (FL0.56), 4DL-12 (FL0.71), and 4DL-14 (FL0.86) (kindly supplied by Dr. TR Endo, Japan), were used to determine the physical location of markers linked closely with Lr67/Yr46.

Analysis of Lr34 ABC transporter and homoeologs in RL6077

The total gene sequence was determined to ensure no additional variants were present at the Lr34/Yr18 locus in RL6077 that could confer Lr67/Yr46 resistance. D-genome-specific primers that amplify the Lr34/Yr18 ABC transporter sequence on wheat chromosome 7D (Krattinger et al. 2009) were used to generate PCR products from RL6077. Full coverage of the corresponding Lr34/Yr18 allele in RL6077 was obtained and compared against the reference Lr34/Yr18 haplotype. Restriction fragment length polymorphism (RFLP) analysis of the homoeologs of the Lr34/Yr18 ABC transporter were investigated using the Lr34/Yr18 cDNA fragment as a probe on membrane filters containing DNA of RL6077, Thatcher and the complete set of nullitetrasomic stocks of Chinese Spring. Genomic DNA of Thatcher and RL6077 were restricted with 14 restriction endonucleases (AccI, BamHI, BgIII, DraI, EcoRI, EcoRV, HindIII, HpaII, NdeI, NcoI, NsiI, PstI, XbaI, XhoI). DNA transfer and hybridization conditions were as described in Lagudah et al. (1991a).

Linkage and statistical analysis

MAPMAKER/EXP version 3.0 (Lander et al. 1987) at minimum log of odds (LOD) of 3.0 was used for the linkage analysis. The Kosambi mapping function was used to calculate the genetic distances. MapChart (Voorrips 2002) was used for drawing the genetic linkage map of the associated markers and Lr67/Yr46. The χ2 test was used to test the goodness of fit of observed segregation with expected ratios for the genotypic classes for all markers and rust response phenotypes.

Results

Characterization of leaf rust and stripe rust resistance in field and greenhouse

Displayed in Table 1 are the field responses of RL6077 (Lr67/Yr46); the two single gene lines carrying Lr34/Yr18 and Lr46/Yr29; the average leaf rust and stripe rust responses of HR lines of the Avocet × RL6077 populations; and the susceptible line Avocet. Since RL6077 was very late maturing, especially in the Cd. Obregon field site due to its day-length sensitivity, responses could only be recorded on the lower leaves. The average response of the HR lines therefore allows a better comparison between Lr67/Yr46 and the two known slow-rusting genes, Lr34/Yr18 and Lr46/Yr29, since the populations were developed from photoperiod-insensitive lines. Lr34/Yr18 displayed a slightly lower leaf rust severity compared with the average response of the homozygous Lr67/Yr46 lines, whereas their stripe rust responses were similar in the field. Lr46/Yr29 showed higher leaf rust and stripe rust responses than Lr67/Yr46 and Lr34/Yr18. The Lr67/Yr46 lines showed strong LTN associated with resistance comparable to the LTN manifested by the Lr34/Yr18-carrying line, whereas the line carrying Lr46/Yr29 showed weaker LTN, as is usually observed in Mexican environments (Fig. S1).
Table 1

Leaf rust and stripe rust responses of RL6077 (RL6077-Lr67/Yr46), HR lines carrying Lr67/Yr46 from the Avocet × RL6077 F4 and F6 populations (average response) (Lr67/Yr46), Yr18/3*Avocet (Lr34/Yr18), Avocet-YrA*3//LalbMono1*4/Pvn (Lr46/Yr29), and Avocet when evaluated in Mexican sites under artificially created epidemics with two Puccinia triticina races at Cd Obregon during 2009–2010 crop cycle and with two P. striiformis races at Toluca in 2009

Genotype

Pathogen (race) and field responsea

P. triticina (MCJ/SP + MBJ/SP)

P. striiformis (Mex96.1 + Mex08.13)

RL6077-Lr67/Yr46

1 MSb

1 Mb

Lr67/Yr46

5 MS

30 M

Lr34/Yr18

1 MS

30 M

Lr46/Yr29

15 MS

40 MSS

Avocet

100 S

100 S

aField responses follow the Modified Cobb Scale (Peterson et al. 1948) and host response to infection as described in Roelfs et al. (1992)

bResponses of flag-1 and flag-2 leaves

In the greenhouse adult plant test with P. triticina races, the Thatcher derivative with Lr34/Yr18 (RL6058) initially displayed somewhat lower responses than lines carrying Lr46/Yr29 and Lr67/Yr46 (Table 2). However, this difference was less evident in a later scoring (results not shown). RL6058 displayed similar, or slightly lower, stripe rust responses compared with RL6077, whereas Lr46/Yr29 showed higher responses with the more aggressive race Mex08.13. Lr34/Yr18 is known to express better in lower temperatures (Dyck and Samborski 1982), whereas the temperature effect on Lr67/Yr46 and Lr46/Yr29 is unknown. The average temperature for both leaf rust and stripe rust greenhouse test was 18 and 17°C, respectively, and use of different (lower and higher) temperature regimes in the future would provide additional information on the comparative expression of these slow-rusting resistance genes. In our study, the stripe rust response of each slow-rusting gene seems to have been influenced by the cultivar background, since the two HR lines from the Avocet × RL6077 population showed higher responses than RL6077, and Thatcher displayed lower responses than Avocet to isolate Mex08.13 of P. striiformis (Fig. S2).
Table 2

Greenhouse infection type responses for leaf rust and stripe rust recorded on flag leaves of RL6077 (RL6077-Lr67/Yr46), two HR F3 lines of the Avocet × RL6077 population (Lr67/Yr46), RL6058 (Lr34/Yr18), Avocet-YrA*3//LalbMono1*4/Pvn (Lr46/Yr29), Avocet and Thatcher after inoculation with two Puccinia triticina (MCJ/SP and MBJ/SP) and two P. striiformis (Mex96.11 and Mex08.13) races

Genotype

Pathogen and infection type responses

P. triticinaa

P. striiformisb

MCJ/SP

MBJ/SP

Mex96.11

Mex08.13

RL6077-Lr67/Yr46

3+

3+

67

7

Lr67/Yr46 (1)

3+

3+

8

78

Lr67/Yr46 (2)

3+

3+

89

78

Lr34/Yr18

3

33+

6

67

Lr46/Yr29

3+

3+

67

9

Avocet

3+

3+

89

9

Thatcher

3+

3+

8

78

aThe infection type responses for leaf rust followed the 0–4 Scale as described in Roelfs et al. (1992) where infection types ‘3’ and ‘4’ are considered high or susceptible, and ‘+’ or ‘−’ are somewhat larger or smaller uredinia, respectively, than normal for the infection type; more than one designation represents a range of infection types. Plants were maintained at minimum, maximum, and average temperatures of 10, 26, and 18°C, respectively, and responses were recorded 11 days post-inoculation

bThe infection type responses followed the 0–9 Scale as described by McNeal et al. (1971), where infection type ‘6 ‘and ‘7’ are considered moderately susceptible, whereas ‘8’ and ‘9’ are considered susceptible; more than one designation represents a range of infection types. Plants were maintained at minimum, maximum and average temperatures of 10, 25, and 17°C, respectively, and responses were recorded 20 days post-inoculation

Distribution of phenotypic categories in Avocet × RL6077 populations

In all leaf rust and stripe rust evaluations in Mexico and Australia, the lines in each population were classified based on their response to the presence and absence of Lr67/Yr46 and grouped into three categories (HR, HS and SEG). The observed ratio for the three phenotypic categories showed a good fit with the ratio expected for segregation at a single locus (Table 3). Evaluations conducted at different sites and during different years helped to determine a final conclusion for each line and to define the category they belonged to for a better marker-trait association.
Table 3

Number of HR, SEG and HS lines in the Avocet × RL6077 F3, F4 and F6 populations and P value from the χ2 tests when tested against segregation for a single gene (Lr67/Yr46)

Population

Disease

Number of observed lines

Number of expected linesa

HR

SEG

HS

HR

SEG

HS

P value

F3

Leaf rustb

34

66

36

34

68

34

0.92

F3

Stripe rustc

27

75

34

34

68

34

0.34

F4

Leaf rustd

56

42

50

56

37

56

0.54

F4

Stripe ruste

57

47

44

56

37

56

0.08

F6

Leaf rustf

67

8

72

69

9

69

0.84

aGenetic ratio expected for segregation of a single gene in F3 (1:2:1), F4 (3:2:3), F6 (15:2:15)

bBased on evaluations conducted in Cd. Obregon, Mexico, in the 2008–2009 and 2009–2010 crop seasons

cBased on evaluation conducted in Cobbitty, Australia, in 2009

dBased on evaluations conducted in Cd. Obregon, Mexico, in the 2008–2009 and 2009–2010 crop seasons and in El Batan, Mexico, in 2009

eBased on evaluation conducted in Toluca, Mexico, in 2009

fBased on evaluation conducted in Cd. Obregon, Mexico, in the 2009–2010 crop season

A majority of lines showed the same response category for leaf rust and stripe rust, indicating that the same gene or closely linked genes conferred resistance to both diseases. The distribution of stripe rust responses for the HR (+Lr67) and the HS (Lr67) leaf rust categories of the Avocet × RL6077 F4-lines is displayed in Fig. 1. The stripe rust responses were clearly lower for the +Lr67 category compared with the −Lr67 category. The phenotypic correlation between the mean severity to leaf rust (data from the 2009–2010 crop cycle in Obregon) and mean severity to stripe rust (Toluca, 2009 crop cycle) for the Avocet × RL6077 F4 lines was high (r = 0.78, P < 0.01).
https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig1_HTML.gif
Fig. 1

Mean stripe rust severity of leaf rust homozygous resistant (LR-HR) and leaf rust homozygous susceptible (LR-HS) F4 lines from the Avocet × RL6077 population

Segregation for additional minor gene(s) was observed for both leaf rust and stripe rust as evidenced by the continuous distribution of lines from the Avocet × RL6077 F4 population for stripe rust severity (Fig. 2), and by the continuous distribution of the Avocet × RL6077 F6 lines in the -Lr67 category for their leaf rust severity (Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig2_HTML.gif
Fig. 2

Distribution of mean stripe rust severity among 148 F4 lines from the Avocet × RL6077 population when tested at Toluca, Mexico, in 2009

https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig3_HTML.gif
Fig. 3

Distribution of mean leaf rust severity among the homozygous resistant (LR-HR) and homozygous susceptible (LR-HS) F6 lines from the Avocet × RL6077 population

Molecular mapping of Lr67/Yr46

Six SSR markers, Xbarc98, Xbarc288, Xcfd23, Xcfd71, Xwmc48, and Xwmc89, from chromosome 4D were identified to be potentially associated with Lr67/Yr46 in the genome-wide scan conducted using the DNA of the parents and the subsample of HR and HS lines from the original F3 population. An initial linkage map was generated using the Avocet × RL6077 F4 population with the consensus leaf rust and stripe rust responses excluding lines whose phenotypic classification remained unclear in the genetic linkage analysis. Additional markers, including Xgwm165 and Xgwm192 in this region of chromosome 4D, were screened on the Avocet × RL6077 F4 population to investigate the trait-marker association. Xwmc48 and Xwmc89 were not linked with Lr67/Yr46, and Xbarc288 was monomorphic in the population but not among the parents (probably due to allele fixation in the two F3 families that were used to generate the advanced populations). Markers Xgwm165, Xbarc98, Xcfd23 and Xgwm192 were scored as dominant markers in the F4 population. The most closely linked markers were Xgwm165 and Xgwm192 based on the initial analysis conducted using the F4 population. The markers that were initially identified to be linked with Lr67/Yr46, Xcfd71, Xcfd23,andXbarc98, were also screened on a subset of the Avocet × RL6077 F3 population, and similar linkage distances were obtained (results not shown). Co-dominant amplification products were detected using the ABI3730xl DNA fragment analyzer; markers Xcfd71, Xcfd23andXbarc98 produced amplification of 231, 229, and 192 bp products, respectively, that were associated with Lr67/Yr46. The corresponding alternate alleles linked with susceptibility and inherited from Avocet were 242, 238, and 189 bp, respectively. The final genetic linkage map was generated using the Avocet × RL6077 F6 RIL population (Fig. 4), and the closest markers, Xgwm165 and Xgwm192, were located at 0.4 cM proximal to Lr67/Yr46. The PCR amplification products of these two markers using two resistant and four susceptible F6 lines from the Avocet × RL6077 population are shown in Fig. 5.
https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig4_HTML.gif
Fig. 4

a Physical map of Lr67/Yr46-associated SSR markers based on ‘Chinese Spring’ chromosome arm 4DL deletion lines, and b Genetic linkage map of Lr67/Yr46 and associated SSR markers of chromosome arm 4DL in Avocet × RL6077 F6 population. Numbers to the right are genetic distances in centiMorgans

https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig5_HTML.gif
Fig. 5

Polymerase chain reaction (PCR) amplification products resolved in 12% acrylamide gels when using Xgwm192 (left) and Xgwm165 (right) for four resistant and two susceptible F6 lines from the Avocet × RL6077 population, and size of bands produced by the size marker (φX174/HaeIII)

Physical mapping of Lr67/Yr46 associated markers using 4D deletion lines

The five markers identified to be associated with Lr67/Yr46 were tested on a set of lines lacking different segments of the 4D short and long arms to establish the physical position of Lr67/Yr46. Amplification products were obtained in all of the lines with deletions in the short arm, whereas one or more lines with deletions in the long arm showed absence of amplification product, confirming that Lr67/Yr46 is present in the long arm (Table 4). The relative marker order of the genetic linkage map was confirmed in the chromosomal deletion bins that span break points between long-arm fraction lengths of 0.3–0.56. The gene Lr67/Yr46 was inferred to be located in a deletion bin distal to the 0.56 chromosomal fraction length break point, based on the location of the most closely linked markers, Xgmw165 and Xgwm192.
Table 4

Comparison of marker genotypes on Chinese Spring chromosome 4D deletion lines

Deletions lines

Response of Lr67/Yr46 associated markers

Line

Ratio intact

Xbarc98

Xcfd23

Xcfd71

Xgwm165

Xgwm192

4DS-01

0.53

+

+

+

+

+

4DS-03

0.67

+

+

+

+

+

4DS-05

0.63

+

+

+

+

+

4DL-09

0.31

4DL-06

0.38

+

+

4DL-13

0.56

+

+

+

4DL-12

0.71

+

+

+

+

+

4DL-14

0.86

+

+

+

+

+

Response to Ug99 stem rust infection

Thatcher, RL6077, RL6058, and 90RN2491 were tested in Kenya. Stem rust scores showed Thatcher with a moderate severity of 25MS, while increased resistance scores of 1MR were observed in RL6058 and 90RN2491, and RL6077 showed a score of 10R-MR. These results indicated that Lr67/Yr46 in a Thatcher background also produces elevated stem rust resistance, as has been found for Lr34/Yr18 against Ug99. However, the additive effect of Lr67/Yr46 + Lr34/Yr18, gene combinations present in 90RN2491, could not be determined because RL6058 with Lr34/Yr18 was already displaying very low disease severity.

ABC transporter sequence and RFLP analysis of RL6077

The complete gene sequence of Lr34/Yr18 allele in RL6077 did not reveal any additional variants, except for the characteristic haplotype of the susceptible allele found in intron 4 and the previously determined variants in exons 11 and 12 (Fig. 6), nor did it reveal any features of the resistant haplotype (Krattinger et al. 2009). The RL6077 sequence is identical to the susceptible haplotypes found in the wheat cultivar Renan (Genbank FJ436985, nucleotide positions 48885–36898) and the diploid D genome, Aegilops tauschii accession AL8/78 (FJ436986, 73713–61747). Thus, Lr67/Yr46 cannot be attributed to the presence of a closely related variant of Lr34/Yr18. RFLP analysis based on 14 restriction enzyme-probe combinations did not reveal any differences between RL6077 and Thatcher (Fig. S3). Thus, at this level of resolution, Lr34 homoeologs did not distinguish the recurrent parent Thatcher from the addition of the Lr67/Yr46 gene.
https://static-content.springer.com/image/art%3A10.1007%2Fs00122-010-1439-x/MediaObjects/122_2010_1439_Fig6_HTML.gif
Fig. 6

Diagrammatic representation of the complete sequence of the Lr34/Yr18 allele from RL6077 compared with the Chinese Spring (CS) haplotype that carries the resistance allele (from Krattinger et al. 2009). The boxes denote exons. The asterisks denote positions of SNPs or indels at intron 4 (T/A), exon 11(±indel TTC) and exon 12 (T/C)

Discussion

From a comparatively larger set of wheat lines, albeit a different wheat population, the present study confirms Canadian reports of adult plant leaf rust resistance at the Lr67 locus located on chromosome 4D (Hiebert et al. 2010). The use of photoperiod-insensitive genotypes selected from the Avocet × RL6077 population enabled field evaluations in diverse field locations without the limitations of variable day length during the growing season. We also provide conclusive evidence of the same gene, or a closely linked gene hereby designated as Yr46, being effective against stripe rust. Hiebert et al. (2010) also reported segregation of stripe rust resistance in F4 lines of Thatcher × RL6077 population and in an F3 population developed from one of the HR F4 lines, ‘H1777’, from the Thatcher × RL6077 population. However, the correlated responses of Lr67 and stripe rust resistance based on phenotypic response was only based on seven F4 lines. Stripe rust resistance based on Australian observations was found to be associated with marker Xcfd71 (Lr67 map location) in the H1777 population, but a comparison with Canadian data on leaf rust response was not made in their study.

A minor QTL for leaf rust adult plant resistance inherited from the wheat line, ND495, was recently reported by Chu et al. (2009) in the same chromosomal region in 4DL. The relationship of this QTL to Lr67 is unknown and further studies to determine whether adult plant stripe rust response cosegregates with the QTL will be required to ascertain whether Lr67/Yr46 is also present in ND495. Two seedling stripe rust resistance genes, Yr22 (Chen et al. 1995) and Yr28 (Singh et al. 2000b), have been reported in this chromosome. Yr28 is associated with marker mwg634 located at the distal end of the short arm of chromosome 4D. Suenaga et al. (2003) identified a QTL for stripe rust resistance in the same chromosome arm at a more distally located position compared with Lr67/Yr46 in bread wheat ‘Oligoculm’, but the associated effect of leaf rust resistance was not detected.

The five markers, Xgwm165, Xgwm192, Xcfd71, Xbarc98, and Xcfd23, identified to be associated with Lr67/Yr46 in this study, positioned this gene to the long arm of chromosome 4D. Based on the physical mapping of the two closely linked markers Xgwm165 and Xgwm192 which map proximal to Lr67/Yr46 the resistance locus/loci is predicted to be located in a deletion bin distal to the break point of 0.57. In the Hiebert et al. (2010) study no recombinants were found in an introgressed region spanning Lr67 and markers Xcfd71, Xbarc98, Xcfd23, Xwmc457 probably due to the small population size evaluated. Furthermore, they suggested from their data that Lr67 is on the distal side of marker Xcfd71 located in the 0.38–0.41 deletion bin. A precise definition of the break points will provide entry points into syntenic regions for comparative genomic analysis to enable a more targeted approach to enrich the region with more markers in refining the position of Lr67/Yr46.

Given the striking similarities between Lr67/Yr46 and Lr34/Yr18 for leaf, stripe, and stem rust responses, we undertook further characterization of the ABC transporter sequence at the Lr34/Yr18 locus and its homoeologs in RL6077 to ascertain whether variant forms of the Lr34/Yr18 locus represent Lr67/Yr46. Because the specific nature of the chromosome translocation present in RL6077 after several backcrosses remains unknown (Dyck et al. 1994), the presence of a duplicated Lr34/Yr18 variant or homoeolog on the translocated chromosome cannot be ruled out. We previously reported only one section of the coding region containing the sequence variants of the Lr34/Yr18 ABC transporter that differentiate resistance and susceptible alleles, with the latter alleles present in RL6077 (Lagudah et al. 2009). With the complete sequence of the gene from RL6077, we now have strong evidence that neither a closely related variant nor a homoeolog of the Lr34/Yr18 is associated with Lr67/Yr46.

The robustness of the closely associated markers Xgwm192 and Xgwm165 for postulating Lr67/Yr46 in a wide range of germplasm, as well as the utility of these markers in a breeding program, needs to be investigated. Since the presence of Lr34/Yr18 has often been deduced by the phenotypic response of adult plants to leaf rust and stripe rust in association with LTN in the field, it is possible that lines have erroneously been attributed to Lr34/Yr18 instead of Lr67/Yr46. However, Kolmer et al. (2008) showed that very few exceptions were found among lines that had these characteristics and lacked the Lr34/Yr18 positive marker allele. Subsequent analysis showed that these lines carried the susceptible Lr34/Yr18 allele based on the diagnostic gene sequence marker (Lagudah et al. 2009); hence, the need to ascertain the presence of Lr67/Yr46 in such genotypes.

Overall, the presence of Lr67/Yr46 gave lower rust severity in association with LTN compared with the susceptible parent in the Avocet × RL6077 population. This allowed us to group lines in each population as HR, HS, and SEG based on their phenotypic response to leaf rust and stripe rust and appearance of LTN. An independent mean severity scoring (%) was in addition made for each line of the population. While the three groups HR, HS, and SEG allowed us to discriminate Lr67/Yr46 in the populations, the mean severity response revealed segregation of additional minor genes in populations, as can be observed in Figs. 1, 2, 3. There were few cases where lines scored as S, showed lower mean severity response than some lines belonging to the R group as can be observed in Fig. 3. Phenotypes of these S lines exhibited severities that were relatively low and lacked LTN compared with the susceptible parent and other S lines in the population. We inferred that these S lines with lower responses most likely carried minor resistance gene(s) in the absence of Lr67/Yr46 (and lack of LTN). Selection of homozygous lines lacking Lr67/Yr46 from the Avocet × RL6077 F6 population that showed relatively lower leaf rust and stripe rust responses can be used to develop new mapping populations for characterization of additional minor gene(s) derived from RL6077.

The expression of Lr67/Yr46 is known to be variable as also has been observed by other authors (Dyck and Samborski 1979) and in some backgrounds relatively high severity levels were observed. A majority of the lines had correlated responses to leaf rust and stripe rust, but there were some lines that fell into different category for leaf rust and stripe rust responses (Table 3). These lines may possibly have been misclassified, from the early generation populations, and hence the use of the F6 population where most of the lines are approximating homozygosity were used for the genetic mapping exercise. Biologically, this is not entirely surprising given the quantitative nature of adult plant slow-rusting phenotypes. This is more so with the stripe rust as the expression of Yr46 was moderate, i.e., severity levels were intermediate.

Additive gene effects have been postulated among gene interactions, whereby up to four or five slow-rusting genes can confer near immunity to rust infection (Singh et al. 2000a). What remains unclear is which specific gene combinations produce optimal additive gene effects when working with fewer genes. In the present study, the level of gene additivity for Lr67/Yr46 + Lr34/Yr18 for enhanced stem rust resistance against Ug99 could not be observed when compared with the effect of Lr34/Yr18 alone because RL6058, which carries Lr34/Yr18, was itself almost free from stem rust. Dyck et al. (1994), investigating the genotype 90RN2491, also failed to observe additive gene effects against leaf rust due to the combined effects of Lr34/Yr18 and what we now know as Lr67/Yr46. Lillemo et al. (2008) examined gene interactions between Lr34/Yr18 and another well-defined dual adult plant rust resistance gene Lr46/Yr29 and found very little evidence for strong additive interaction. However, in each case this could have been due to low disease severity caused by the presence of Lr34/Yr18. Apart from Lr34/Yr18, Lr46/Yr29, and Lr67/Yr46, there is clearly a need to define other slow-rusting resistance loci to enable further characterization of which gene combinations give rise to a stronger synergistic effect against rust infection. Lr34/Yr18 is known to enhance resistance levels when combined with other minor genes, and exploring this phenomenon with Lr67/Yr46 may lead to new gene combinations that allow making more effective use of slow-rusting resistance genes in wheat breeding.

Acknowledgments

We are grateful to Violeta Calvo Salazar, CIMMYT, and Libby Viccars, Sutha Chandramohan, and Lynette Rampling, CSIRO, for their skilled technical support on the molecular studies; to the technical field and greenhouse staff at CIMMYT; to Hanif Miah for the field rust studies at Cobbitty, Australia; to Dr. TR Endo, Japan, for providing the deletion lines used in this study, and to the Grains Research and Development Corporation (GRDC) of Australia for funding the work (Grants CSP00099 and CIM00013). We also thank Alma McNab for the technical editing.

Supplementary material

122_2010_1439_MOESM1_ESM.doc (486 kb)
Fig. S1. Field flag leaf responses to Puccinia triticina (pathotypes MCJ/SP and MBJ/SP) from Mexican site, Ciudad Obregon, in 2010. 1= Avocet S, 2 = Lr34/Yr18 (Yr18/3*Avocet), 3= Lr46/Yr29 (Avocet-YrA*3//LalbMono1*4/Pavon), 4= LrP-7BL (Avocet/Prl), 5= Lr67/Yr46 (Avocet/RL6077), 6= Avocet/Pvn ‘S’ (DOC 287 kb)
122_2010_1439_MOESM2_ESM.doc (1.1 mb)
Fig. S2. Greenhouse flag leaf responses to Puccinia striiformis pathotype Mex08.13.1 = Lr67/Yr46 (Avocet/RL6077), 2 = Lr67/Yr46 (RL6077), 3 = Lr34/Yr18 (RL6058), 4 = Lr46/Yr29 (Avocet-YrA*3//LalbMono1*4/Pavon), 5 = Avocet ‘S’, 6 = Thatcher (DOC 486 kb)
122_2010_1439_MOESM3_ESM.doc (286 kb)
Fig. S3. Thatcher and RL6077 (Thatcher+Lr67/Yr46) probed with an Lr34 cDNA fragment (spans exons 13 to the end of the gene). The 14 lanes shown in the figure corresponds to genomic DNA from the near isogenic pair Thatcher (lanes 1, 3, 5, 7, 9, 11, 13) and RL6077 (lanes 2, 4, 6, 8, 10, 12, 14) cleaved with the restriction endonucleases DraI, EcoRI, EcoRV, HindIII, SacI, XbaI, BglII, NcoI (DOC 1076 kb)

Copyright information

© Springer-Verlag 2010