Resistance to Pseudomonas syringae in a collection of pea germplasm under field and controlled conditions

Authors

  • A. Martín-Sanz,

    1. Instituto Tecnológico Agrario, Consejería de Agricultura y Ganadería de la Junta de Castilla y León, Ctra Burgos, km 119, 47071, Valladolid
    Search for more papers by this author
  • M. Pérez de la Vega,

    1. Área de Genética, Departamento de Biología Molecular, Universidad de León, 24071, León, Spain
    Search for more papers by this author
  • C. Caminero

    Corresponding author
    1. Instituto Tecnológico Agrario, Consejería de Agricultura y Ganadería de la Junta de Castilla y León, Ctra Burgos, km 119, 47071, Valladolid
    Search for more papers by this author

E-mail: camsalco@itacyl.es

Abstract

A total of 242 Pisum accessions were screened for resistance to Pseudomonas syringae pv. pisi under controlled conditions. Resistance was found to all races, including race 6 and the recently described race 8. Fifty-eight accessions were further tested for resistance to P. syringae pv. syringae under controlled conditions, with some highly resistant accessions identified. Finally, a set of 41 accessions were evaluated for resistance to P. syringae pv. pisi and pv. syringae under spring- and winter-sowing field conditions. R2, R3 and R4 race-specific resistance genes to P. syringae pv. pisi protected pea plants in the field. Resistance sources to race 6 identified under controlled conditions were ineffective in the field. Frost effects were also evaluated in relation to disease response. Results strongly suggest that frost tolerance is effective in lowering the disease effects caused by P. syringae pv. pisi and pv. syringae under frost-stress conditions, even in the absence of disease resistance genes, although the highest degree of this protection is reached when frost tolerance and disease-resistance genes are combined in the same genetic background.

Introduction

Protein pea (Pisum sativum) is an important crop used as a protein source for livestock. Injuries caused by winter frost associated with early sowing can increase the incidence of diseases in pea, particularly those caused by bacteria (Schmit et al., 1992; Hollaway & Bretag, 1995; Reeves et al., 1996; Elvira-Recuenco et al., 2003; Hollaway et al., 2007; Martín-Sanz et al., 2011a). The presence of these diseases represents a major constraint to the use of this crop (Lawyer & Chun, 2001; Hollaway et al., 2007; Martín-Sanz et al., 2011a). Pseudomonas syringae is the main pathogen responsible for bacterial diseases in pea, but while bacterial blight is mainly attributed to P. syringae pv. pisi in most countries (Lawyer & Chun, 2001), two pathovars of this species, pv. pisi and pv. syringae, have been described as almost equally frequent in New Zealand, Spain and Australia (Taylor & Dye, 1972; Martín-Sanz et al., 2011a; Richardson & Hollaway, 2011).

The classical symptomatology of pea bacterial blight, caused by P. syringae pv. pisi and pv. syringae, appears initially as elliptical water-soaked areas. Lesions become olive-green, and finally a brown and necrotic area is observed in aerial plant parts. These lesions often encircle the stem and may extend several centimetres and infect both stipules and leaflets. Severe infection of the stem can cause plant death. Bacterial blight is more frequently observed after frost periods (Lawyer & Chun, 2001; Hollaway et al., 2007).

Environmental factors are important in the expression of disease caused by pea bacterial blight pathogens. Freezing stress can act in synergy with the pathogen to cause disease. Not only does frost damage favour entry of the pathogen (Roberts et al., 1995), but the bacteria can also act as an ice nucleate under freezing conditions, increasing frost injury at higher temperatures than would have otherwise occurred (Hirano & Upper, 2000). Young & Dye (1970) and Boelema (1972) observed that frosted pea plants were more susceptible to infection by P. syringae pv. pisi than unfrosted plants, which were only slightly susceptible. Subsequently, Roberts et al. (1995) found that frost and hail damage favours the entry of the pathogen, and Mansfield et al. (1997) observed that disease severity was greater in winter-sown than in spring-sown peas in both winter and spring cultivar types.

Pseudomonas syringae pv. pisi has been found in all countries in which pea production is important (Hollaway et al., 2007). Two races of this pathogen were first described (Taylor, 1972); this was later expanded to six (Taylor et al., 1989), then seven races (Bevan et al., 1995), and recently an eighth race was added (Martín-Sanz et al., 2011a). Race 2 seems to be the most frequent worldwide, although race frequencies vary between regions and according to the predominant pea varieties cultivated and the sowing date (winter or spring) (Hollaway et al., 2007). To date, races 2, 3, 4, 5, 6 and 8 have been found in Spain, with race 4 being the most frequent (Martín-Sanz et al., 2011a). Disease caused by P. syringae pv. syringae also seems to be associated with early sowing in late autumn or early winter (Lawyer & Chun, 2001). Contrasting to P. syringae pv. pisi, no races have been distinguished in P. syringae pv. syringae.

The control of pea bacterial blight is difficult and it is mainly based on the use of pathogen-free seeds, crop rotation, and culture practices minimizing plant lesions which facilitate the infection by the pathogen (Lawyer & Chun, 2001). Chemical control is ineffective against the pathogen in infected fields (Hollaway et al., 2007). The development of resistant cultivars is most likely to provide an effective way to control this disease.

Several race-specific resistance genes to P. syringae pv. pisi have been described (Taylor et al., 1989; Bevan et al., 1995). Resistance to race 6 has been found only in Pisum abyssinicum accessions and in a Spanish P. sativum landrace (Schmit et al., 1993; Elvira-Recuenco & Taylor, 2001). The resistance from P. abyssinicum is race-non-specific, and the introgression of resistance from P. abyssinicum into P. sativum requires long-term breeding approaches. With regard to P. syringae pv. syringae, several pea lines have been described as resistant to this pathovar under controlled (Martín-Sanz et al., 2011a) and field conditions (Richardson & Hollaway, 2011). Thus, more tests in different agroclimatic conditions are needed to ascertain the usefulness of the resistance to breeding programmes.

The aims of this study were: (i) to search for resistance to both P. syringae pv. pisi and P. syringae pv. syringae in pea materials useful for breeding new resistant cultivars, (ii) to test whether the resistance shown under controlled conditions is maintained under field conditions, and (iii) to determine the influence of host frost tolerance in disease expression.

Materials and methods

Assays in controlled conditions

All different Pisum materials will be generally referred to as accessions throughout this paper, except when individual ones are mentioned. The pea cultivars described by Bevan et al. (1995) were used as controls to test gene-for-gene relationships, and the P. abyssinicum accessions JI1640 and JI2202 to test race-non-specific resistance (RNSR) to P. syringae pv. pisi (Elvira-Recuenco et al., 2003) (Table 1). A total of 242 Pisum accessions were tested in the study (Table S1): 68 commercial pea cultivars (mainly protein pea); 67 breeding lines obtained from the Spanish ITACyL collection (48 from its own breeding programme, eight from the Portuguese Estação Nacional de Melhoramento de Plantas and 11 from other sources); 56 Spanish landraces, of which 40 are the Spanish pea core collection (Caminero et al., 2001); 35 accessions from the John Innes Centre Pisum spp. core collection, and 16 accessions from the USDA Pisum spp. core collection. Resistance to P. syringae pv. pisi races 1–7 was tested in these materials under controlled conditions. Among the 242 accessions, a selection of 52 entries of multiple origins and representing the resistance profiles to P. syringae pv. pisi races 1–7 were inoculated with the new race 8. Resistance genes to this pathogen are named as R1 to R6 according to Bevan et al. (1995).

Table 1.   Gene-for-gene relationships between pea cultivarsa and races of Pseudomonas syringae pv. pisi (modified from Bevan et al., 1995)
  1. +: susceptible response; −: resistant response; ?: gene probably present; .: gene absent.

  2. aMaterials include the cultivars used by Bevan et al. (1995) to identify pathogen races and two P. abyssinicum accessions which show resistance to all races; JI numbers refer to passport numbers of John Innes collections.

  3. bRace 8 is the new race described by Martín-Sanz et al. (2011a) on the basis of its differential response pattern to pea tester lines.

 Races/avirulence genes
12345678b
1.......
.2..2.2.
3.3...3.
4..44.4.
....5?...
6?...6?...
.......7?
Cultivar/accessionR Genes
Kelvedon Wonder.......++++++++
Early Onward.2.....+++++
Belinda..3....+++++
Hurst Greenshaft...4.6?.++++
Partridge..34...+++
Sleaford Triumph.2.45?.7?++
Vinco123.5?..+++
Fortune.234..7?+
JI2202Race-non-specific resistance
JI1640Race-non-specific resistance

Resistance to P. syringae pv. syringae was tested in a subset of 58 accessions out of the 242 pea accessions used to test the response to P. syringae pv. pisi (Table S2). The accessions in this subset were firstly selected according to their response to P. syringae pv. pisi to include a representation of each combination of resistance genes and the diversity of the Pisum spp. collection.

The bacterial isolates used in the inoculations are shown in Table 2. They included P. syringae pv. pisi type strains and isolates obtained from pea plants in Spain (Martín-Sanz et al., 2011a). For P. syringae pv. syringae tests, the Spanish P262 isolate was used. All the Spanish isolates were isolated from pea plants in fields showing severe damage by pea bacterial blight.

Table 2. Pseudomonas syringae isolates used to test pea resistance
PathovarRaceIdentificationaOrigin
CultivarCountryYear
  1. aReference isolates from Horticulture Research International, Wellesbourne, Warwickshire, UK. P indicates Spanish isolates.

pisi1HRI-W 299ARondoNew Zealand1970
pisi2P134AthosSpain2004
pisi3P519UnknownSpain2008
pisi4P68MessireSpain2004
pisi5P89IdealSpain2004
pisi6HRI-W 1704BStehgolt (seed)Francia1986
pisi6P123RafaleSpain2004
pisi7HRI-W 2491AUnknownAustralia1976
pisi8P130RafaleSpain2004
syringae P262GraciaSpain2006

Stems and leaves of pea seedlings were maintained, inoculated (at the three-true-leaf stage) and evaluated for P. syringae pv. pisi races according to the procedure described by Elvira-Recuenco et al. (2003). A total of 15 seedlings per accession were separately inoculated with each of the isolates, and two replicates were carried out at different times. The response was evaluated 10 days after inoculation as follows: susceptible, when a clear water-soaked area was observed surrounding the inoculation point; resistant, when a localized necrotic area surrounded the inoculation point (hypersensitive response); incompletely resistant, when a localized necrotic area coexisted with water-soaked areas.

A total of 15 pea seedlings per accession, in three replicates of five seedlings each carried out at different times, were inoculated with P. syringae pv. syringae. The inoculation method was the same as for P. syringae pv. pisi. Each plant inoculated was scored for leaf and stem damage and the average for each accession and organ was calculated. Response was evaluated 10 days after inoculations. For the stem the scale was: 0, small localized necrosis in the infection point; 1, necrotic spots <1 cm in diameter; 2, necrotic lesions at the inoculation point between 1 and 2 cm in diameter; 3, dark brown necrotic depressed lesions more than 2 cm long and reaching the next node in the apical direction; 4, similar to reaction type 3 but lesions were more than 3 cm long with a clear extension into the apical part of the plant; and 5, necrosis of the apical part of the plant (including the apical meristem) from the inoculation point. For the leaves the scale was: 0, small localized necrosis in the infection point to small necrotic spots on <25% of the leaf surface; 1, necrotic spots covering between 25 and 50% of the leaf; 2, necrosis over >50% of the leaf surface; and 3, whole-leaf necrosis. The response on stems was also qualitatively classified as highly resistant (values between 0 and 1), resistant (1–2), moderately resistant (2–3), susceptible (3–4) and highly susceptible (4–5).

Assays under field conditions

Two field assays were carried out during the 2005–06 and 2008–09 seasons. The first field test was performed at Zamadueñas (Valladolid, Spain), in a river valley with fertile soils. This assay will be referred to as ZAMA. The second (2008–09) was carried out at Quintanilla de Trigueros (Valladolid, Spain), in a bleak plateau with cooler winter conditions and will be referred to as TRIG. Time of sowing was evaluated in both assays, with mid-November (referred to as WINTER) and mid-February (referred to as SPRING) sowings.

The ZAMA assay consisted of 18 accessions which had previously shown different responses to P. syringae pv. pisi races under controlled conditions. For each sowing date four trials were conducted: infected with P. syringae pv. pisi races 2, 4 and 6, and a non-infected control. For the TRIG assay two additional trials in each sowing period were added: one infected with P. syringae pv. pisi race 3 and the other with P. syringae pv. syringae (isolate P262). In this assay 41 accessions were evaluated: the same as in ZAMA (except for Coomonte) plus additional accessions representing different resistance gene patterns and others representing variable degrees of resistance to P. syringae pv. pisi race 6 under controlled environmental conditions. Table S3 indicates the final accessions evaluated in each trial, and also includes a genotypic frost tolerance estimation obtained in an environmental chamber test according to the methodology described by Caminero (2002).

For each combination of location × sowing date × bacterial isolate, or trial, a randomized complete block design with three replicates was used for field assays. The experimental unit was a microplot sown with 40 seeds in four 1-m-long rows, with 20 cm between rows. Within each locality and date of sowing, trials (infection treatments) were separated by at least 20 m to avoid contamination among the different bacterial isolates used. Trials were individually infected with P. syringae pv. pisi race 2, race 3, race 4, or race 6, or with P. syringae pv. syringae. Inoculations were conducted by spraying each microplot (at the three-leaf stage) with 750 cL bacterial suspension. To make the suspension, bacteria grown for 2 days on King’s B plates at 25°C were scraped off the plates and suspended in sterile water containing 0·025% (v/v, final concentration) humectant agent manoxol to an OD600 of 0·2, corresponding to a final concentration of 108 CFU mL−1, as determined by dilution plating. Infections were carried out in mid-March for WINTER sowing and at the beginning of April for SPRING sowing. Control trials were sprayed with sterilized water. Data were collected approximately 30 days after inoculation. Bacterial samples were collected to prove that cross-contaminations did not occur.

As field symptoms are more complicated to evaluate than those under controlled conditions, different quantitative measurements, based on plant symptoms, were taken, such as the number of affected plants (incidence) and the amount of infection in affected plants (severity) in each microplot. Incidence was evaluated according to the percentage of diseased plants: 0 = <10%, 1 = 10–25%, 2 = 25–50%, 3 = >50–75% and 4 = >75%. Severity was scored as: 0, very resistant, without symptoms or with small localized necrosis spots on stems or leaves; 1, resistant, small water-soaked spots on leaves and/or stems; 2, moderately resistant, water-soaked spots on leaves and stems covering between 10 and 25% of the plant surface; 3, susceptible, 25–50%; 4, very susceptible, >50–75%; or 5, extremely susceptible, >75% of plant surface covered.

A preliminary analysis of variance of the dependent variables incidence and severity was carried out taking into account a model including the pattern of resistance to P. syringae pv. pisi (pattern of R-genes, see first column in Table 3), genotype (= accession) (nested into pattern) and block as independent variables. For those cases in which significant differences (< 0·05) were detected, least-square means of the patterns of R-genes were classified according to Dunnett’s multiple comparison test (Dunnett, 1955) for comparison with a control (response of the pattern R0: no resistance genes).

Table 3.   Response to Pseudomonas syringae pv. pisi (Ppi) of pea accessions under field conditions
Ppi race Resistance profileIncidenceaSeveritya
  1. R0: represents all the accessions without resistance genes; RNSR: race-non-specific resistance in Pisum abyssinicum (Pa) or P. sativum (Ps).

  2. aFigures indicate least-square means values; *< 0·05.

2TRIG-WINTERR02·892·67
R2 + R31·06*0·83*
R2 + R3 + R40·08*0·08*
R2 + R40·67*0·67*
R32·482·24
R3 + R42·562·67
RNSR-Pa2·502·33
ZAMA-WINTERR02·671·67
R2 + R30·00*0·00*
R2 + R3 + R40·00*0·00*
R32·331·92
R3 + R42·672·50
R42·672·00
RNSR-Pa1·671·00
ZAMA-SPRINGR02·331·50
R2 + R30·75*0·50*
R2 + R3 + R40·22*0·22*
R31·421·42
R3 + R42·001·50
R41·672·00
3TRIG-WINTERR02·893·33
R2 + R31·33*1·22*
R2 + R3 + R41·50*2·00*
R2 + R42·673·00
R31·50*1·37*
R3 + R40·67*0·67*
4TRIG-WINTERR04·082·91
R2 + R33·993·21
R2 + R3 + R43·681·81*
R2 + R44·081·41*
R33·972·92
R3 + R43·482·01*
RNSR-Pa4·083·21
ZAMA-WINTERR02·862·14
R2 + R31·931·28
R2 + R3 + R40·17*0·22*
R32·391·67
R3 + R40·51*0·65*
R41·690·65
RNSR-Pa2·631·44
ZAMA-SPRINGR01·251·03
R2 + R30·880·78
R2 + R3 + R40·00*0·00*
R30·630·53
R3 + R40·13*0·14*
R40·25*0·28*
6TRIG-WINTERR03·172·67
R2 + R33·172·94
R2 + R3 + R43·002·92
R2 + R43·333·00
R33·232·83
R3 + R43·002·78
RNSR-Pa3·673·00
RNSR-Ps3·002·53
ZAMA-WINTERR02·501·33
R2 + R31·831·08
R2 + R3 + R41·44*1·11
R32·001·25
R3 + R41·50*1·33
R42·331·33
RNSR-Pa2·671·67
ZAMA-SPRINGR01·671·33
R2 + R31·251·17
R2 + R3 + R41·221·33
R31·501·33
R3 + R41·331·33
R42·001·00

Frost tolerance influence was analysed in two independent groups of genotypes for each original trial (combination location × sowing date × isolate): those accessions included in R-gene patterns whose response to the disease was not significantly different to R0, and those accessions included in R-gene patterns which significantly reduced disease expression. For each trial and group, a simple linear regression analysis was carried out considering the disease incidence and severity as dependent variables, and the quantitative covariable related to genotypic frost tolerance under controlled conditions as the independent variable.

For the analysis related to P. syringae pv. syringae resistance, a multiple regression stepwise model was used considering disease incidence and severity as dependent variables, and the quantitative genotypic covariables related to the responses in stem and leaves to P. syringae pv. syringae in the growth chamber and the quantitative genotypic covariable related to frost tolerance as independent variables. All statistical analyses were performed with the sas 9·0 package. Table S4 shows the climatic data.

Results

Resistance under controlled conditions

The response to inoculation of the 242 pea accessions with P. syringae pv. pisi under growth-cabinet conditions is shown in Table 4. The response of each accession to P. syringae pv. pisi races (1–7) is shown in Table S1. A set of 44 accessions (18%) were susceptible to all P. syringae pv. pisi races, while the rest showed resistance to at least one of the races. Two P. abyssinicum accessions (JI0130 and JI2385) and nine Spanish pea landraces (ZP0123, ZP0150, ZP0152, ZP0156, ZP0157, ZP0168, ZP0344, ZP1305 and ZP1307) were resistant or partially resistant to all races when inoculated in the stem, but showed susceptibility to some races in the leaves (Table S1).

Table 4.   Number of Pisum accessions showing different combinations of resistance genes to Pseudomonas syringae pv. pisi (Ppi) races 1–7
R GenesaResistance to Ppi racesCultivarsBreeding linesSpanish landracesPisum spp.bAccessions in each category
  1. RNSR: Race-non-specific resistance.

  2. aR gene content is deduced from their interaction with a set of isolates representative of races 1–7, according to Taylor et al. (1989) and Bevan et al. (1995).

  3. bAccessions from the John Innes and USDA collections.

  4. *Possible new resistance genes not previously described.

None116121544
R11 and 710539
R22, 5 and 7303511
R31, 3 and 7303812989
R41, 4, 5 and 710146
R1 + R21, 2, 5 and 700011
R1 + R31, 3 and 700101
R2 + R31, 2, 3, 5 and 714173236
R2 + R41, 2, 4, 5 and 710438
R3 + R41, 3, 4, 5 and 7431412
R2 + R3 + R41, 2, 3, 4, 5 and 7333110
*200011
*1, 2 and 700011
*600404
RNSR1, 2, 3, 4, 5, 6 and 700729
Totals 68675651242

The race-specific R genes present in the accessions were deduced according to the gene-for-gene relationship (Taylor et al., 1989; Bevan et al., 1995). The resistance shown by P. abyssinicum was considered as race-non-specific (Schmit et al., 1993; Elvira-Recuenco et al., 2003). Accessions JI0196 and JI0399 showed a response pattern not previously described and ZP0344 showed partial resistance in the stem to all P. syringae pv. pisi races. Thus, no previously known resistance genes could be assigned to these three accessions. The most frequently found resistance gene was R3, which was present in approximately 65% of the accessions, followed by R2 (29%), R4 (16%) and R1 (4%).

Resistance responses to races 1, 2, 5 and 7 in the stem were always associated with resistance in leaves. Individual R3 and R4 genes conferred complete resistance to races 3 and 4, respectively, in stems but only partial resistance in leaves. However, accessions with the combination R3 + R4 showed complete resistance in both leaves and stems, which would point to some epistatic effect. Accessions with R4 or R3 + R4 genes were only partially resistant to race 5 in both stems and leaves (Table S1).

Resistance to race 6 was the least frequent, and in most cases it was found to be intermediate. The accessions which showed some degree of resistance (complete or intermediate) to race 6 are given in Table 5. They include the 11 accessions which had proved to be resistant to all other previously described races, plus accessions ZP1324 and ZP0109. The responses to the two isolates of race 6 were not always identical (Table 5).

Table 5. Pisuma spp. accessions which showed resistance to Pseudomonas syringae pv. pisi race 6
AccessionTypeCountryResistance to race 6 isolatesStem resistance to other racesR Genesb
P123HRI-W 1704B
  1. W: wild; LR: Spanish landrace; R: resistant, hypersensitive response; I: incomplete resistance, mixture of localized necrotic areas with water-soaked susceptible response; I and R: some plants resistant and some with a mixed response; RNSR1: race-non-specific resistance in P. abyssinicum; RNSR2: possible race-non-specific resistance in P. sativum with intermediate resistance to all P. syringae pv. pisi races; RNSR3: possible summation of race-non-specific resistance and specific resistance genes. R?: hypothetical resistance gene to race 6.

  2. aAccessions JI0130 and JI2385 are Pisum abyssinicum, all others are P. sativum.

  3. bR2, R3 and R4: the major resistance genes described by Bevan et al. (1995).

JI0130WPalestineRR1, 2, 3, 4, 5 and 7RNSR1
JI2385WYemenRI1, 2, 3, 4, 5 and 7RNSR1
ZP0344LRSpainII and R1, 2, 3, 4, 5 and 7RNSR2
ZP0152LRSpainII2, 5 and 7R2 + RNSR3
ZP1324LRSpainII and R1, 3 and 7R3 + R?
ZP0109LRSpainII and R1, 3, 5 and 7R3 + R?
ZP0123LRSpainII and R1, 2, 3, 4, 5 and 7R2 + R3 + R4 + R?
ZP0150; ZP0157LRSpainII1, 2, 3, 4, 5 and 7R3 + RNSR3
ZP0156LRSpainII1, 2, 3, 4, 5 and 7R3 + R4 + RNSR3
ZP0168LRSpainII1, 2, 3, 4, 5 and 7R2 + R3 + R4 + R?
ZP1305; ZP1307LRSpainII and R1, 2, 3, 4, 5 and 7R2 + R3 + RNSR3

The accessions with resistance genes R2 + R4 and R2 + R3 + R4 and the P. abyssinicum accessions JI0130 and JI2385 were resistant to P. syringae pv. pisi race 8 (Table 6). The landrace ZP0344, with an intermediate resistance to races 1–7, also showed an intermediate resistance to race 8. The rest of the accessions evaluated were susceptible.

Table 6.   Response of evaluated Pisum spp. accessions to Pseudomonas syringae pv. pisi race 8
AccessionResponseR genesAccessionResponseR genes
  1. R: resistant; I: intermediate; S: susceptible.

ForrimaxRR2 + R4JI0015SR1
JI2546RR2 + R4SpecterSR2
PI-277852RR2 + R4ZP0204SR2
ZP1328RR2 + R4JI2376SR2
CherokeeRR2 + R3 + R4PI-411143SR2
CoralloRR2 + R3 + R4CartoucheSR3
LincolnRR2 + R3 + R4ChicoSR3
PM29RR2 + R3 + R4ZP0138SR3
PM32RR2 + R3 + R4PI-166084SR3
PM33RR2 + R3 + R4JI2200SR3
JI1829RR2 + R3 + R4MetaxaSR4
ZP1282RR2 + R3 + R4ZP1300SR4
ZP0104RR2 + R3 + R4JI0109SR4
ZP1301RR2 + R3 + R4PI-203064SR4
ZP0123RR2 + R3 + R4IcebergSR2 + R3
ZP0168RR2 + R3 + R4PM28SR2 + R3
JI2385RRNSRZP0840SR2 + R3
JI0130RRNSRJI2105SR2 + R3
ZP0344IRNSRDoveSR3 + R4
GraciaSNonePM11SR3 + R4
PM39SNoneZP0180SR3 + R4
ZP0074SNoneJI1844SR3 + R4
JI0086SNoneJI0241SR3 + R4
PI-166159SNoneJI0196SUnknown
GlotónSR1JI0399SUnknown
ZP1262SR1ZP0109SUnknown

Pea accession responses to inoculation with P. syringae pv. syringae are shown in Table 7 and Table S2. Stems scores indicated that 31% of the evaluated accessions were resistant or highly resistant (scores between 0·2 and 2) and 36% were susceptible (scores between 3·2 and 5). The response in leaves was positively correlated (0·75) with the response in stems. Seven accessions showed scores between 0 and 1 in both stem and leaf (VR in Table 7): Cherokee, Atika, Arthur, ZP0168, Messire, Windham and Kelvedon Wonder. Pisum abyssinicum accessions JI1640 and JI2202 and the pea cv. Melrose were the most susceptible accessions. Some accessions with high scores of 4 in stems reached low scores of 1 in leaves (Forrimax, Nela and Pawnee, and the P. fulvum accession JI1796). Even though Melrose was one of the most susceptible materials in the stem (score of 5), its response in leaves was of resistance (score of 1). Thus, resistance sources to the pathogen were found exclusively in the cultivated pea (P. sativum ssp. sativum), while other Pisum taxa, including wild species, were susceptible, with scores of between 3 and 5. The results point to a lack of any relationship between resistance to P. syringae pv. pisi and to P. syringae pv. syringae. Figure 1 shows examples of the symptomatology caused by P. syringae pv. pisi and pv. syringae in resistant and susceptible responses under controlled conditions.

Table 7.   Number of Pisum accessions showing different levels of resistance to Pseudomonas syringae pv. syringae
Resistance levelNumber of accessionsRange of disease responsea
StemLeaves
  1. VR: very or highly resistant; R: resistant; MR: moderately resistant; S: susceptible; VS: very or highly susceptible.

  2. aAverage value of 15 inoculated plants.

VR70·20–1·000·33–1·00
R111·20–2·001·00–1·00
MR192·20–3·001·00–2·00
S123·20–4·001·00–2·66
VS94·10–5·001·33–3·00
Figure 1.

 Pea stems and leaves inoculated with Pseudomonas syringae pv. pisi and P. syringae pv. syringae strains under controlled conditions. (a) Susceptibility and (b) resistance to pv. pisi race 4. (c) Apical death and foliar necrotic symptoms caused by pv. syringae.

R-genes and RNSR influence on disease occurrence under field conditions

Field trials were carried out to confirm the results under controlled conditions to the P. syringae pv. pisi races 2, 3, 4 and 6, and to P. syringae pv. syringae. Because the spring treatment at TRIG (2008–09) generated a very low general level of disease, these results were not included in the statistical analysis. None of the control trials sprayed with sterilized water showed disease symptoms. All the bacterial samples collected in the infected trials coincided with the isolate inoculated. Table 3 shows the results of the field response to P. syringae pv. pisi races considering the R-gene patterns previously defined under controlled conditions. When trials infected with race 2, 3 and 4 were considered, the resistance patterns carrying R2, R3 and/or R4 (i.e. R2 + R3, R2 + R4 and R2 + R3 + R4), irrespective of the presence of any other R gene, showed significantly less incidence and severity than the R0 pattern. This result was observed in the three assays scored. No significant differences to R0 were observed for patterns lacking R2, R3 or R4 in any of the assays when inoculated respectively with race 2, 3 and 4. These results point to the effectiveness of these genes under field conditions. On the other hand, the RNSR observed in P. abyssinicum (accessions JI0130 and JI1640) under controlled conditions was not confirmed in field conditions against these races.

Resistance to race 6 under controlled conditions was totally ineffective under field conditions. None of the accessions that had shown some degree of resistance when inoculated under controlled conditions had significantly lower infection values than R0 under field conditions. Likewise, no other combinations of R-genes analysed showed significant differences with regard to the R0 control, except two resistance patterns from ZAMA-WINTER for incidence, but not for severity, and for these reason they were not considered.

Relationships between frost tolerance and disease expression

Results concerned with the influence of the frost tolerance covariable on P. syringae pv. pisi disease expression (incidence and severity) are shown in Table 8. The spring assay (ZAMA-SPRING) did not show any significant relationship between frost response and disease expression. However, several significant differences were observed for the two winter sowings. These differences were clearer in TRIG-WINTER, where frost tolerance significantly reduced disease expression in all the trials when the groups of accessions lacking the respective resistance gene to each P. syringae pv. pisi race (2, 3 and 4) were considered. When accessions with the respective resistance genes were considered, it was shown that frost tolerance contributed to reduce disease symptoms in the genotypes with R3 inoculated with race 3. In the case of ZAMA-WINTER, the influence of frost tolerance in the reduction of disease expression was only significant when considering the pea genotypes carrying R4 (≠ R0) infected with race 4. For race 6, frost tolerance significantly reduced disease severity under the harsher winter conditions of TRIG-WINTER.

Table 8.   Influence of frost tolerance on disease expression in pea caused by Pseudomonas syringae pv. pisi (Ppi) and P. syringae pv. syringae (Psy) in field conditions, and relationships between leaf response to Psy under controlled conditions and field resistance
Ppi racePpi resistance patternsEnvironmentDependent variableModel R2Regression modela
  1. aI: incidence; S: severity; FT: genotypic frost tolerance; PSSL: genotypic tolerance to Psy according to leaf inoculations; *: < 0·05; **: < 0·01; n.s.: non-significant.

2Patterns = R0TRIG-WINTERI0·25I = 3·07** − 0·36** FT
S0·33S = 2·92** − 0·38** FT
ZAMA-SPRINGIn.s. 
Sn.s. 
ZAMA-WINTERIn.s. 
Sn.s. 
Patterns ≠ R0TRIG-WINTERIn.s. 
Sn.s. 
ZAMA-SPRINGIn.s. 
Sn.s. 
ZAMA-WINTERIn.s. 
Sn.s. 
3Patterns = R0TRIG-WINTERI0·55I = 3·55** − 0·67* FT
S0·38S = 3·83** − 0·5* FT
Patterns ≠ R0TRIG-WINTERI0·27I = 1·93** − 0·44** FT
S0·23S = 1·82** − 0·39** FT
4Patterns = R0TRIG-WINTERI0·08I = 4·1** − 0·08** FT
S0·21S = 3·26** − 0·19** FT
ZAMA-SPRINGIn.s. 
Sn.s. 
ZAMA-WINTERIn.s. 
Sn.s. 
Patterns ≠ R0TRIG-WINTERIn.s. 
Sn.s. 
ZAMA-SPRINGIn.s. 
Sn.s. 
ZAMA-WINTERI0·37I = 1·05** − 0·57**FT
S0·22S = 0·69** − 0·3*FT
6Patterns = R0TRIG-WINTERIn.s. 
S0·17S = 3·02** − 0·13** FT
ZAMA-SPRINGIn.s. 
Sn.s. 
ZAMA-WINTERIn.s. 
Sn.s. 
Psy EnvironmentDependent variableModel R2Regression model
- TRIG-WINTERI0·38I = 4·13** − 0·19** FT
-  S0·35S = 3·59** − 0·45** FT + 0·51* PSSL

Relationships between field resistance to P. syringae pv. syringae and the covariables frost tolerance and P. syringae pv. syringae resistance under controlled conditions

Resistance was evaluated in TRIG-WINTER (Table 8). The two covariables related to resistance to P. syringae pv. syringae in the growth chamber (in stem and leaves) were not significantly correlated with the response to frost under controlled conditions, supporting independence between them. Frost tolerance was related to significant reductions in disease incidence and severity under field conditions in winter. Leaf resistance response in the growth chamber was also related to a reduction of severity in the field. Stem resistance in the growth chamber did not show any significant influence on disease expression in the field at any step of the stepwise analysis process. Figure 2 shows the response of two cultivars in TRIG-WINTER.

Figure 2.

 Field plots of pea cvs Eiffel (susceptible response) and Cherokee (resistant response) 1 month after the inoculation with Pseudomonas syringae pv. syringae.

Discussion

Previous studies evaluating the response to P. syringae pv. pisi analysed the resistance in pea cultivars and breeding lines in countries such as France (Schmit et al., 1992, 1995), Great Britain (Taylor et al., 1989) or Australia (Hollaway & Bretag, 1995). Cultivars in common between the present and previous works (eight from French studies and four from Great Britain) showed the same responses to P. syringae pv. pisi races. Almost no landraces and wild materials have been evaluated, with the exception of 10 Spanish landraces (Elvira-Recuenco & Taylor, 2001) and the preliminary study of Schmit et al. (1993). Furthermore, in all those studies resistance was evaluated under controlled conditions, thus no data under real field conditions are available. This work affords data on the resistance to P. syringae pv. pisi races (including race 8), but also to the pathovar syringae, in a collection including pea cultivars, breeding lines, landraces and wild materials. Likewise, this work evaluated not only the resistance under controlled conditions, but also under field conditions in Central Spain.

World core collections (JI and USDA) and the Spanish landrace collection showed a higher richness in resistance gene combinations than cultivars and breeding lines (Table 4). Thus, with the exception of two wild P. abyssinicum accessions, resistance to P. syringae pv. pisi race 6 was only found among Spanish landraces (Table 5). This collection holds a high level of genetic variability with many valuable characters for breeding purposes (Martín-Sanz et al., 2011b).

The frequency in which P. syringae pv. pisi races are found is inversely related to the frequency of resistance genes against them among the cultivars sown in the study area (Hollaway et al., 2007). The most frequent race isolated in pea fields in Spain between 2004 and 2008 was race 4 (59%), followed by races 2 (20%) and 6 (12%) (Martín-Sanz et al., 2011a). Considering the 12 cultivars mainly used in Spain in this period (data not shown), the results of the present study partially agree with the relationship pointed out by Hollaway et al. (2007). Only two of those cultivars were resistant to race 4 and five to race 2, which would explain the relative abundance of these races in Spain. However, none of them was resistant to race 6, which is at odds with the low frequency of this race in Spain, while its frequency has been increasing in other countries. The absence of races 1, 3 and 7 is in agreement with the results of Hollaway et al. (2007), as R3 is present in the majority of the cultivars grown in Spain.

Race-specific resistance under controlled conditions for races 2, 3 and 4 was also observed under field conditions. This supports the effectiveness of the corresponding resistance genes (R2, R3 and R4) under real field conditions. In fact, the gene combination R2 + R3 + R4 seems to be the best because it is effective in controlling most P. syringae pv. pisi races (Elvira-Recuenco et al., 2003), including the new race 8, but it is ineffective against race 6. The alleles conferring resistance are dominant (Bevan et al., 1995; Hunter et al., 2001), which facilitates their selection and pyramiding. Genes R3 and R4 are located in the linkage group II and R2 in group VII (Hunter et al., 2001), although they have not been fine-mapped.

The first source of RNSR to P. syringae pv. pisi race 6 was found in P. abyssinicum (Schmit et al., 1993). The two accessions evaluated in the present study, JI0130 and JI2385, also demonstrated resistance to race 6 under controlled conditions. A race-specific resistance was described by Elvira-Recuenco & Taylor (2001) in one (ZP0109) out of 10 Spanish landraces. Among the 56 Spanish landraces tested here, there were 11 accessions (including ZP0109) which showed some degree of resistance to race 6 under controlled conditions. However, none of these were resistant in the field. Elvira-Recuenco et al. (2003) reported that RNSR of P. abyssinicum to race 6 was effective under English field conditions, while it was ineffective in the tests in the present study. This difference could be caused by the different methods of field testing. In the test by Elvira-Recuenco et al. (2003), seeds were germinated and kept in the greenhouse for 3 weeks, then covered for a week in the field before plants were inoculated. In the present study, seeds were directly sown in the field and P. abyssinicum plants showed clear symptoms of cold stress. In fact, the P. abyssinicum accessions evaluated were very sensitive to frost under controlled conditions (Table S3).

Only a few studies have analysed the resistance of pea to P. syringae pv. syringae. Recently, the resistance to this pathogen was evaluated under controlled (Martín-Sanz et al., 2011a) and field conditions (Richardson & Hollaway, 2011). In the present study 58 pea accessions [19 in common with Martín-Sanz et al. (2011a)] were evaluated for their response to a more aggressive isolate of P. syringae pv. syringae in stems and leaves. The response of the pea genotypes common to both experiments was similar. The quantitative and almost continuous response observed (Tables 7 and S2) pointed to polygenic control, similar to that described in common bean (Jung et al., 2003; Navarro et al., 2007). Several cultivars showed high levels of resistance and are therefore potentially useful for breeding, in particular Cherokee, which was also resistant to all P. syringae pv. pisi races, except to race 6. However, a better solution depends on the identification of a useful source of resistance to race 6.

It has been reported that frost damage increases susceptibility to P. syringae (Young & Dye, 1970; Boelema, 1972; Roberts, 1992; Mansfield et al., 1997) and even the resident bacteria can enhance frost damage by acting as ice-nucleating agents (Maki et al., 1974; Hirano & Upper, 2000). Slinkard et al. (1994) found that the threshold temperatures for pea were −2°C (plants able to thrive without appreciable damage) and −6°C (significant yield decrease). According to these data, TRIG-WINTER was a stressful environment to pea plants: 38 days with minimum temperatures of below −2°C among which 7 days were below −6°C (Table S4). ZAMA-WINTER was a milder environment, with only 18 days below −2°C and never reaching −6°C. In ZAMA-SPRING the temperatures were below −2°C on only 4 days. These differences would explain the results related to expression of P. syringae pv. pisi in the different environments. Hollaway et al. (2007) indicated that when pea plants inoculated with this pathogen were maintained at −2°C for 2 h the damage caused by frost reached an incidence of 72%, compared with 21% in non-inoculated controls. This ice-nucleating effect can also explain the greater effect observed in TRIG-WINTER, where the temperature was below −2°C five times after inoculation, while in ZAMA-WINTER temperatures never dropped below 0 after infection.

Likewise, it has been pointed out that frost-tolerant pea materials would be less affected by bacterial diseases than non-tolerant ones (Elvira-Recuenco et al., 2003; Hollaway et al., 2007). The results of the present study agree with this hypothesis but also indicate that the protection effect is related to the presence of particular disease resistance genes. Thus, in the absence of frost stress (i.e. ZAMA-SPRING), there would be no significant difference between frost-tolerant and non-tolerant genotypes, irrespective of the P. syringae pv. pisi race used or the presence or absence of resistance genes against the corresponding pathogen race. When frost stress is present, frost tolerance has been revealed as a mechanism of resistance complementary to the disease resistance genes, as indicated by the examples of race 4 in ZAMA-WINTER or race 3 in TRIG-WINTER (Table 8). However, in harsh environments (e.g. TRIG-WINTER) frost tolerance itself appears to be a mechanism for reducing disease expression, as this effect was detected for all the inoculated races in genetic backgrounds without resistance genes. Interestingly, although no resistance gene has been shown to be effective under field conditions against race 6, frost tolerance seems to be an effective protection for reducing disease expression for this race as well.

With regard to the resistance to P. syringae pv. syringae in field conditions, the present results indicated that frost tolerance also reduces disease incidence and severity. Likewise, in-chamber leaf disease response was positively related to field disease symptoms, indicating that the response to inoculation in leaves in a growth chamber could be a useful method for predicting the response under field conditions. By contrast, stem responses did not show any relationship with disease expression under field conditions.

In conclusion, the resistance to P. syringae in pea tested under controlled conditions is effective under field conditions, except for P. syringae pv. pisi race 6, and frost tolerance in pea generally decreased the disease effects under harsh winter conditions, even in the absence of disease resistance genes. Likewise, resistance to the recently described race 8 of P. syringae pv. pisi was found, although further tests under field conditions will be needed. Thus, pea breeding programmes for cold areas should focus, at least, on joint selection of frost tolerant accessions with a set of genes conferring resistance to the prevalent pathovars and/or races of P. syringae. Further searches to find effective disease resistance sources to P. syringae pv. pisi race 6 under field conditions will be necessary.

Acknowledgements

This research has been supported by Junta de Castilla y León ITACyL 2004/845, INIA RTA 2006-00077-00-00 and Ministerio de Ciencia e Innovación GEN2006-27798-C6-3-E/VEG projects, and by an INIA personal PhD grant to AMS. The authors thank E. Aguado and A. García-Vaquero for technical assistance in the resistance assays. We thank the referees for their valuable comments on the manuscript.

Ancillary