Evaluation of different methods for the characterization of carrot resistance to the alternaria leaf blight pathogen (Alternaria dauci ) revealed two qualitatively different resistances
C. Boedo
ab, R. Berruyer
c, M. Lecomte
b, S. Bersihand
a, M. Briard
b, V. Le Clerc
b, P. Simoneau
aand P. Poupard
a*
aIFR Quasav 149, UMR PaVe´ A77, UFR Sciences, Universite´ d’Angers, 2 boulevard Lavoisier, 49045 Angers Cedex 01;bIFR Quasav 149, UMR GenHort 1259, Agrocampus Ouest - Institut National d’Horticulture et de Paysage, 2 rue Le Noˆtre, 49045 Angers Cedex 01; andcIFR Quasav 149, UMR GenHort 1259, UFR Sciences, Universite´ d’Angers, 2 boulevard Lavoisier, 49045 Angers Cedex 01, France
Alternaria leaf blight (ALB), caused byAlternaria dauci, is one of the most damaging foliar diseases of carrot worldwide.
The aim of this study was to compare different methods for evaluating levels of carrot resistance to ALB. Three techniques were investigated by comparison with a visual disease assessment control:in vivoconidial germination, a bioassay based on a drop-inoculation method, andin plantaquantification of fungal biomass by quantitative PCR (Q-PCR). Three carrot culti- vars showing different degrees of resistance toA. dauciwere used, i.e. a susceptible cultivar (Presto) and two partially resis- tant genotypes (Texto and Bolero), challenged with an aggressive or a very aggressive isolate ofA. dauci. Both partially resistant genotypes produced a higher mean number of germ tubes per conidium (up to 3Æ42±0Æ35) than the susceptible one (1Æ26±0Æ18). The drop-inoculation results allowed one of the partially resistant genotypes (Bolero, log10(S+1) = 1Æ34±0Æ13) to be distinguished from the susceptible one (1Æ90±0Æ13). By contrast, fungal growth measured by Q-PCR clearly differentiated the two partially resistant genotypes with log10(I) values of 2Æ77±0Æ13 compared to the susceptible cultivar (3Æ65±0Æ13) at 15 days post-inoculation. This result was strongly correlated (r2= 0Æ91) with the disease severity index scored at the same date.
Data obtained with the different assessment methods strongly suggest that the Texto and Bolero genotypes have different genetic resistance sources.
Keywords:Daucus carota, in vivoconidial germination, necrotrophic fungus, plant resistance, real-time quantita- tive PCR
Introduction
Alternarialeaf blight (ALB) of carrot (Daucus carota) is responsible for economically significant fungal dam- age worldwide in all carrot growing areas. Losses caused by ALB can reach 40–99% (Vintal et al., 1999; Ben-Noon et al., 2001). Alternaria dauci, the necrotrophic causal agent of the disease, is one of the most destructive foliar pathogens of this crop (Farrar et al., 2004). The symptoms may be confused with those of other pathogens such as Cercospora carotae and Xanthomonas hortorum pv. carotae (Gaube et al., 2004). Leaves of severely diseased crops appear burned as a result of coalescence of dark brown
lesions, sometimes surrounded by a chlorotic halo, under favourable conditions (Farraret al., 2004). ALB is controlled through integrated use of pathogen-free seeds, crop rotations, fungicide applications and partially resistant cultivars. However, none of these measures is completely efficient under high disease pressure. Partially resistant carrot cultivars provide various levels of disease tolerance (Simon & Strand- berg, 1998), but breeders are still searching for culti- vars showing higher levels of resistance to ALB. The first analysis of the genetic architecture of carrot ALB resistance was recently published (Le Clerc et al., 2009).
Field and greenhouse screening using a disease rating scale is the routine procedure for the identification of resistant genotypes (Pawelecet al., 2006; Guginoet al., 2007). It is widely used by breeders and Groupe d’Etude et de Controˆle des Varie´te´s et des Semences (GEVES), the
*E-mail: [email protected] Published online 6 January 2010
organization responsible for registering new cultivars in the French and European catalogues. However, this method is time consuming, expensive and may be affected by uncontrollable environmental conditions. Also, when it comes to evaluation of symptom development, it is dif- ficult to differentiate between phenotype classes that have intermediate levels of resistance to ALB (Cadot et al., 2002). A powerful phenotyping tool would be advanta- geous for carrot breeders in characterizing host resistance toA. dauci.
Since the first report by Bo¨hmet al.(1999), real-time quantitative polymerase chain reaction (RT-Q-PCR) assays have been used to an increasing extent to quantify the level of plant infection by pathogens. The method is based on measurement of the intensity of a fluorescent signal which is proportional to the amount of DNA gen- erated during PCR amplification (Rasmussen et al., 1998). This method has already proved to be a valuable tool for accurate quantification of fungal pathogen bio- mass in plant extracts (Berruyeret al., 2006). A prelimin- ary study demonstrated the feasibility of this method for quantifyingA. dauciin carrots (Boedoet al., 2008).
The effects of different levels of resistance of potato cultivars toin vivogermination ofAlternaria solaniconi- dia were evaluated (Ditaet al., 2007): conidia were able to germinate in several cultivars tested regardless of the resistance level, but significantly higher germination rates were noted in the middle and upper plant parts of a sus- ceptible cultivar than in those of a resistant one. Resis- tance evaluation based on the germination of fungal conidia has proven useful for some pathosystems, but has never been applied toA. dauci.
This study was conducted to assess the feasibility of several methods to improve the evaluation of carrot resis- tance to ALB. Three carrot cultivars showing different levels of resistance to ALB were used: Presto, a susceptible cultivar (Ben-Noonet al., 2001), and Texto and Bolero, two partially resistant cultivars. Several techniques were compared: in vivo conidial germination, a greenhouse bioassay on non-detached leaves and, finally, the quanti- fication of fungal biomass by Q-PCR. Visual assessment of disease symptoms was used as a control method.
Materials and methods
Plant and fungal material
Carrot seeds of cvs Presto, Texto and Bolero (Vilmorin) were pretreated with a fungicide containing two active ingredients [iprodione (dicarboximide, 3 g commercial product kg)1) and thiram (dimethyldithiocarbamate, 3Æ5 g active ingredient kg)1)]. Plants were grown as previ- ously described (Boedo et al., 2008). Two isolates of A. dauci collected in France in 2000 from naturally infected carrots (P2 from Gironde; A2 from Maine et Loire), were used. Isolate P2 was rated as aggressive and isolate A2 as very aggressive. Isolates were grown in Petri dishes on carrot juice agar (CJA) (200 mL Jokercarrot juice, 20 g agar L)1) and then incubated for 20 days under
a 16-h photoperiod (Sylvania, GRO-LUX, F36W⁄GRO- T8 lamps) at 24±2C for production of conidia.
Standard inoculation of plants and visual assessment In the greenhouse, young plants were inoculated at the 6- to 8-true-leaves stage with a conidial suspension adjusted to 1·104conidia per mL. Conidial suspensions were prepared according to Pawelecet al. (2006) using sterile water amended with 0Æ05% Tween 20 (Sigma).
Control plants were sprayed with 0Æ05% Tween 20 in dis- tilled water. Inoculated and control plants were incu- bated under plastic covers for 2 days with day⁄night temperatures of 24C. Boxes and plastic covers were sprayed three times a day for 2 days with sterile water in order to maintain a high relative humidity. After removal of plastic covers, the day⁄night temperatures were main- tained at 24C. The relative humidity was maintained at 80% using an automatic greenhouse misting system. For visual assessment, samples consisted of the second (f2) and third (f3) true leaves from five inoculated plants. Five replications were made for each sampling date and each cultivar⁄isolate combination. Disease severity (DS) was rated, from 10 to 25 days post-inoculation (d.p.i.), using a 0–5 scale (adapted from Pawelecet al., 2006) according to the percentage of leaf area damaged (0 = no visible dis- ease damage, 1 = <5%, leaf area damaged, 2 =‡5–20%
leaf area damaged, 3 = 20–40% leaf area damaged, 4 = 40–60% leaf area damaged, 5 = >60% leaf area damaged).
Measurement ofin vivoconidial germination Conidial germination ofA. dauciisolate P2 was moni- tored in the three cultivars. For each carrot cultivar, the third leaf (f3) from 10 different plants was sampled at 1 d.p.i. Leaves were individually cut into pieces and discol- oured in equal volumes of ethanol 100% acetic acid for at least 24 h, then cleared in lactophenol [37% glycerin (w⁄v), 19% lactic acid (w⁄v) and 21% phenol (w⁄v) in bi-distilled water] for 24 h. Subsequently, leaves were stained in cotton blue (0Æ5% methyl blue in lactophenol) for 30 min and thoroughly rinsed with water. Specimens were mounted in Entellan (Merck) and observed under a light microscope (Leica AF 6000). Conidia in which ger- mination tubes were present, regardless of length or num- ber of germ tubes per conidium, were considered germinated. The number of germ tubes produced by coni- dia was also recorded for each cultivar. One hundred con- idia (10 per leaf) were counted. Values were expressed as a percentage of germinated conidia and as a mean number of germ tubes per conidium (±SE).
Material sampling and DNA extraction for real-time PCR assays
Samples consisted of the second (f2) and third (f3) true leaves from five inoculated plants collected 0, 4, 10, 15, 21 and 25 d.p.i. or from five control plants collected
0 and 25 days after water spraying. Five replications were made for each sampling date and each cultivar⁄isolate combination. Plant samples and DNA extractions were treated as previously described (Boedoet al., 2008). For Q-PCR assays, DNA concentrations were measured by fluorometric assay (Turner Design Inst. 700) and all DNAs were diluted to a final concentration of 10 nglL)1 in order to minimize the effect of PCR inhibitors. This concentration was determined according to the results of preliminary experiments. Fungal DNA from pure cul- tures on carrot juice medium was extracted according to Mo¨lleret al. (1992). DNA concentrations of mycelium or of non-inoculated carrot leaves were determined by fluo- rometry in order to plot calibration curves and assess the specificity of the Q-PCR assay.
Real-time PCR for in planta fungal quantification The biomass of fungal DNA for each sample was deter- mined by Q-PCR usingSYBRGreen chemistry on an ABI prism 7000 SEQUENCE DETECTION SYSTEM software (Applied Biosystems) as described in Boedoet al. (2008).
The following specific primers were used for glyceralde- hyde-3-phosphate dehydrogenase (gpd) fungus gene amplification: gpd-sense (5¢-AAGCTTCCCCAAGCA CTCACAA-3¢) and gpd-antisense (5¢-CTGCGTTCT GCAGCTGTAGAGA-3¢). Each run included a no-tem- plate control in order to rule out test reagent contamina- tion. Q-PCR products were tested for the presence of non-specific products by analysis of the melting curve withDISSOCIATION CURVES SOFTWARE(Applied Biosystems).
Serial dilutions of pure genomic DNA fromA. dauciwere used to plot a calibration curve, which was used to quan- tify the fungal DNA content in each DNA sample (assay).
To ensure that Q-PCR-inhibitors potentially present in plant DNA extracts would not induce significant variations in the Q-PCR results, theYpr10*Md.bgene (GenBank Accession No. AY026908) encoding a patho- genesis-related PR-10 protein fromMalus domesticawas used as an external control in the qPCR assay, as previ- ously described (Boedoet al., 2008).
Fluorescence data analysis and cycle threshold (Ct) cal- culation were performed using three replicates of the Q- PCR assay for each sample. The fungal genome copy num- ber (Nf) was calculated from Ct values and calibration curves using anA. daucihaploid genome size of 30 Mbp (Akamatsuet al., 1999). Plant genome copy number (Np) was set at 12 500 copies per PCR reaction, corresponding to a DNA quantity of 25 ng, a carrot haploid genome size of 980 Mbp (Bennett & Smith, 1976) and the diploid nature of carrot. An infection ratio I = (Nf⁄Np)·100 was calculated as previously described (Berruyeret al., 2006).
Drop inoculation
Drop-inoculation experiments were performed on young greenhouse-grown plants as follows: an inoculation chamber was set up around the third true leaf (f3) without detaching it from the plant. The inoculation chamber con-
sisted of a 90-mm-diameter Petri dish containing an 85- mm-diameter filter paper disc. A slot was cut in the dish for the petiole, and a weighted paperclip was used to pre- vent the leaf from touching the Petri dish lid. Typically, moisture build up was fast in the chamber and remained high throughout the experiment. In cases where low mois- ture was observed, 1 mL of water was added to the disc.
Inoculation was performed using a conidial suspen- sion prepared from cultures of isolate P2 according to Pawelecet al. (2006) and adjusted to different concentra- tions (200, 1000 and 4000 conidia per mL). For each cul- tivar, 5-, 6- or 8-week-old plants were used. A recorded number (Nd, c. 40) of 5-lL drops of inoculum were deposited on the leaf using a micropipette. The exact concentration (C, expressed in CFU mL-1) of the inocu- lum suspension was checked by plating conidia on CJA and colony counting. This enabled the number of viable conidia deposited on each leaf during inoculation, Nc = C · Nd · 0Æ005, to be calculated. After inoculation, day⁄night temperatures were maintained at 21C in greenhouse conditions. At 7, 9 and 13 d.p.i., the number of lesions (Ns) was counted on each leaf, while taking the fact that large lesions sometimes emerge as a combi- nation of several smaller ones into account. The lesion number per viable conidium ratio (S) was calculated as follows: S = (Ns⁄Nc)·100. The experiment was repeated three times. Each time, two to four inoculations were performed for each combination of cultivar, conid- ial concentration and inoculation time.
Statistical analysis
For thein vivogermination of conidia assay, statistical analyses (Kruskal-Wallis test) were performed using
STATGRAPHICSsoftware (StatPoint Technologies Inc.). For the visual disease scoring results,in plantafungal quanti- fication by Q-PCR and for the drop-inoculation assay, statistical analyses were performed usingR2.6.1 software (R Development Core Team, 2005). The homoscedastic- ity of DS, S, log10(S+1), I ratio and log10(I) was checked by the Breusch-Pagan test. S and I did not show homoscedas- ticity, whereas log10(S+1) and log10(I) did. A linear model from DS, log10(S+1) and log10(I) was made, considering repetition as a block treatment. Residuals were found to be normally distributed (normal Q-Q plot) in each case.
Results
Measurement ofin vivoconidial germination Alternaria dauciconidial germination was observed on leaf surfaces of the three carrot cultivars 1 d.p.i. The per- centage of conidial germination was not significantly dif- ferent from 100% in any of the tested cultivars (data not shown). When considering the mean number of germ tubes per conidium (Fig. 1), the difference between the three cultivars was significant (P<0Æ05): the number was lower on cv. Presto (1Æ26±0Æ18) than cv. Bolero (2Æ80±0Æ24), and highest on cv. Texto (3Æ42±0Æ35).
Fungal quantification by real-time PCR
Quantification of fungal biomass by Q-PCR was achieved in cvs Presto, Texto and Bolero from 0 to 25 d.p.i. with A. dauci isolate P2 or A2. Regardless of plant cultivar or fungal isolate, log10(I) increased
overall throughout the experiment. With isolate A2, there were no significant differences between the three cultivars in terms of fungal biomass at 0, 5, 10 and 25 d.p.i. (Fig. 2a). The isolate A2, biomass was higher in Presto than in Texto or Bolero at 15 d.p.i. At 21 d.p.i., Presto and Bolero were significantly more colonized by isolate A2 than Texto (P<0Æ05). At 15 and 21 d.p.i., the biomass of isolate P2 was significantly higher in the susceptible Presto than in the partially resistant Texto and Bolero (Fig. 2b). Bolero was much more colonized by the fungus than Texto at 0, 5 and 25 d.p.i. Overall, the biomass of isolate A2 biomass was higher than that of isolate P2.
Figure 2Fungal growth quantification over an infection time course from 0 to 25 days post-inoculation (d.p.i.) withAlternaria dauci isolates A2 (a) or P2 (b) in susceptible carrot cv. Presto or partially resistant cvs Bolero and Texto. Results were expressed as an infection ratio [I = (fungal genome copy number⁄plant genome copy number)·100]. Data are means and SEs of infection ratios calculated from five replications. At 15 d.p.i., mean infection ratios for A2 and P2 are shown in both (a) and (b) to aid comparison of isolate aggressiveness.
(a)
(b)
(c)
Figure 1Development ofAlternaria dauciconidia (isolate P2) on the leaf surface of susceptible carrot cv. Presto (a), or partially resistant cvs Bolero (b) and Texto (c) at 1 day post-inoculation. Leaf samples were stained with methyl blue and observed by light microscopy. A, appressoria-like structure; C, conidium; GT, germ tube.
In this experiment, regardless of theA. dauciisolate used, 15 d.p.i. was the best time to differentiate the sus- ceptible cultivar from the partially resistant ones. With isolate A2, log10(I) values at 15 d.p.i. reached 4Æ27, 3Æ56 and 3Æ46 in cvs Presto, Bolero and Texto, respectively.
For the same cultivars inoculated with P2 conidia, the values were 3Æ65, 2Æ96 and 2Æ77, respectively,, with a standard error of 0Æ13.
Visual disease assessment
For each cultivar, DS was visually assessed between 10 and 25 d.p.i. on leaves collected for quantification of fun- gal biomass by Q-PCR. With isolate A2, the DS index at 10 d.p.i. was higher in cv. Presto than in cv. Bolero (Fig. 3a). At 15 d.p.i., the visual assessment allowed Pre- sto to be differentiated from the two other cultivars. At 21 d.p.i., Presto and Texto each had a higher DS index than Bolero. At 15 d.p.i., with isolate P2, results similar to those obtained using isolate A2 were obtained (Fig. 3b):
the DS index was significantly higher for Presto than for the two other cultivars (P<0Æ05). At 21 d.p.i., the DS index for cv. Bolero was significantly lower than for Pre- sto. As with the Q-PCR results (Fig. 2), the best discrimi- nation between the susceptible cultivar and the partially resistant cultivars was obtained at 15 d.p.i. At that time, correlation between DS and log10(I) was high (r2= 0Æ91).
Drop inoculation
Drop inoculation was tested in cvs Presto, Texto and Bolero using three different inoculum concentrations and three different inoculation dates (Fig. 4). Mean S values ranged from 0Æ002 to 78. As expected, the total number of lesions per leaf was greater when a more concentrated inoculum was used (data not shown). Nevertheless, S generally decreased when the spore concentration was greater, perhaps because of competition effects. S was also lower on older plants. No significant discrimination between the partially resistant and susceptible cultivars was noted at 4000 conidia per mL, or on 5-week-old plants. At 1000 conidia per mL, the most discriminating results were obtained from 6-week-old plants: irrespec- tive of incubation time, cv. Texto displayed the highest S.
Although S was higher for the susceptible Presto than for the partially resistant cv. Bolero, this difference was not significant. At 200 conidia per mL, the most discriminat- ing results were obtained 13 d.p.i. from 6-week-old plants: log10(S+1) was significantly lower in Bolero (1Æ34) than in Presto (1Æ90) and Texto (1Æ72, SE = 0Æ13). Similar results were obtained at 9 d.p.i. from 8-week-old plants inoculated at 200 conidia per mL.
Discussion
The methods for the evaluation of carrot resistance againstA. daucidescribed here were different from ear- lier studies. Previously Dugdaleet al. (2000) used a spe- cific bioassay on mature detached leaves where
chlorophyll levels were used as indicators of infection because chlorophyll concentrations declined during dis- ease progress. The decrease in chlorophyll content was greater in susceptible cv. Fancy than in wild carrot, suggesting that wild carrot expressed partial resistance to A. dauci. The use of this criterion could be interesting for the assessment of carrot cultivar resistance to ALB, but this method is time consuming and not suitable for analy- sing numerous samples. Pawelec et al. (2006) tried to develop a rapid and reliable technique based on a detached leaf assay or a hypocotyl assay for screening
Figure 3Visual disease assessment over an infection-time course from 10 to 25 days post-inoculation (d.p.i.) withAlternaria dauci isolates A2 (a) or P2 (b) in susceptible carrot cv. Presto or partially resistant cvs Bolero and Texto. Disease index was determined on leaves collected for the fungal growth quantification study (see Fig. 2) and was based on a 0–5 scale according to the percentage of leaf area damage. Data are means and SEs of the disease index calculated from five replications. At 15 d.p.i., mean disease index for A2 and P2 is shown in both (a) and (b) to aid comparison of isolate aggressiveness.
carrot plants showing resistance to ALB. Three cultivars showing different levels of resistance toA. dauciwere used: Presto (susceptible), Bolero and the inbred line B5280 (partially resistant). In that study, neither the detached leaf assay nor the hypocotyl assay revealed sig- nificant differences among the tested cultivars, so these assays are not reliable for evaluating carrot resistance to ALB. More recently, Kra¨meret al.(2009) developed a digital image analysis system to test resistance against several carrot fungal pathogens, includingA. dauci. Preli- minary results using a set of 10 carrot accessions sug-
gested that the system could be useful as a screening method for evaluation of new resources of ALB resistance in carrot.
In the present study, several methods were compared with the aim of developing new phenotyping tools for characterizing host resistance toA. dauci.First,in vivo germination of conidia in carrot leaves was investi- gated. A higher mean number of germ tubes per conid- ium was determined for the two partially resistant cultivars by comparison with the susceptible one.
The penetration of A. dauci in the epidermis of the
(a) (b)
Figure 4Visual disease assessment after drop inoculation withAlternaria dauciisolate P2 on susceptible carrot cv. Presto or the partially resistant genotypes Bolero and Texto. (a) Symptom number ratio [S = (number of lesions⁄number of conidia)·100] calculated 7, 9 or 13 days post-inoculation (d.p.i.) of leaves from 5-, 6- or 8-week-old plants with 5-lL droplets containing 200, 1000 or 4000 conidia per mL. Data are means and SEs of lesion number ratios. (b) Scans of drop-inoculated leaves (1000 conidia per mL) from cvs Presto, Bolero and Texto taken 13 d.p.i.; inoculation performed as in (a) on 6-week-old plants.
susceptible cultivar seemed to be rapidly successful, as at 1 d.p.i. a single germ tube per conidium was appar- ently sufficient for leaf penetration. For the two partially resistant cultivars, several germ tubes (around three per conidium) were produced over the same time, which could be interpreted as several attempts by the fungus to penetrate the leaf epidermis. A recent qualita- tive study using scanning electron microscopy demon- strated that leaf penetration byA. dauciwas successful on Presto, Bolero and Texto genotypes, regardless of cultivar resistance level: for all three cultivars, direct penetration of the fungus with or without ‘appressoria- like’ structures through the intercellular junctions was observed 1 d.p.i. and subcuticular hyphae were present 10 or 15 d.p.i. (C. Boedo, unpublished data). Few pre- vious studies have considered the number of fungal germ tubes produced in susceptible vs. resistant plant genotypes. The results reported in the present paper differed from the data obtained with powdery mildew in melon (Floris & Alvarez, 1996), in which the devel- opment of the first conidial germ proceeded equally well in all melon genotypes tested, second and third germ tubes were observed in a susceptible cultivar, but not in the control resistant lines. The differences observed between theA. dauci–carrot and the powdery mildew pathosystems could be caused, at least partly, by the necrotrophic vs. biotrophic nature of the respec- tive fungal pathogens. Interestingly, the mean number of germ tubes per conidium allowed cv. Texto, which had a significant higher value, to be differentiated from cv. Bolero. These two carrot cultivars originated from different genetic resistance sources (G. Simon, Vilmorin, France, personal communication).
When using the drop-inoculation method, cv. Bolero exhibited a significantly lower lesion number ratio than the susceptible genotype, whereas cv. Texto could not be distinguished from cv. Presto. Similar results were obtained with the visual disease assessment method at 21 d.p.i.: a significantly lower disease index was recorded for Bolero than for Presto. Overall, both methods revealed an intermediate position of Texto after measurement of disease index (Fig. 3) or lesion number ratio determined after plant inoculation with 200 conidia per mL (Fig. 4a), although the differences between the genotypes were not always significant.
The results described above: (i) suggest that Bolero could be more resistant than Texto, and (ii) confirm that different genetic resistance sources are present in Bolero and Texto. The drop-inoculation method had some advantages over the control method based on visual disease assessment, such as the requirement of less plant material and a more objective disease-scoring criterion. Based on these results, however, it could not be definitively concluded whether drop inoculation could be used as a valuable phenotyping tool to dis- criminate between carrot cultivars. Experiments are underway to test the method on a larger panel of car- rot genotypes with various degrees of resistance to A. dauci.
In a previous paper (Boedo et al., 2008), standard experimental conditions were defined forin plantaquan- tification ofA. dauciand a significant difference in fungal biomass (isolate P2) between cvs Presto and Texto was shown based on preliminary results. In the present work, for both isolates (A2 and P2, quantification ofA. dauci biomass by Q-PCR clearly differentiated the susceptible carrot cultivar from the two partially resistant genotypes at 15 d.p.i. This main result was highly correlated (r2= 0Æ91) with the DS index scored at 15 d.p.i. by the visual assessment method, also differentiating Bolero and Texto from Presto. Q-PCR analyses indicated that the mean A. daucibiomass in Bolero was generally intermediate (except at 0 and 5 d.p.i.) between that observed in the two other genotypes. Texto appeared to be more resistant than Bolero on the basis of the significant differences noted at 21 d.p.i. (isolate A2) and 25 d.p.i. (isolate P2).
This assumption was not supported by the data obtained using the drop-inoculation or visual-assessment methods.
Symptom development and fungal biomass are indeed not always strictly correlated (Thommaet al., 1999). As it is more objective than visual assessment, the sensitive and robust Q-PCR method has already proved its use- fulness for quantifying plant colonization by pathogens.
This technique was successfully used to distinguish res- ponses of genotypes in several pathosystems [Phytoph- thora capsici⁄pepper (Silvar et al., 2005); Verticillium albo-atrum⁄alfalfa (Larsen et al., 2007); V. dahliae⁄
potato (Atallahet al.,2007)]. The results of the present study strongly suggest that the Q-PCR method could also be of great interest for carrot breeders seeking to discriminate cultivars for resistance againstA. dauci.
In conclusion, a difference in the degree of resistance was noted between cvs Bolero and Texto on the basis of some results obtained through determination of the lesion number ratio (drop inoculation) or Q-PCR quanti- fication of fungal biomass: the combination of different methods revealed two qualitatively different resistances in carrot againstA. dauci. Two of the methods developed here, i.e. counting the number ofin vivofungal germ tubes per conidium and measuring in planta fungal growth, provided a clear-cut distinction between the sus- ceptible cultivar and the two partially resistant geno- types. These methods could potentially be used as an alternative or to complement the standard visual disease scoring system; their validation as phenotyping tools will be evaluated using a larger set of carrot cultivars with varying degrees of resistance to the fungus.
Acknowledgements
The authors would like to thank A. Suel and B. Hamon for their technical assistance, R. Gardet and J. Granger for their help in greenhouse experiments, and D. Manley for reviewing the English. This work was financed by the French General Board for Companies (Cre´ation Varie´- tale Potage`re FCE project, 2007–2010). CB was granted a fellowship by ANRT (CIFRE) and Clause Vegetable Seeds.
References
Akamatsu H, Taga M, Kodama M, Johnson R, Otani H, Kohmoto K, 1999. Molecular karyotypes forAlternariaplant pathogens known to produce host-specific toxins.Current Genetics35, 647–56.
Atallah ZK, Bae J, Jansky SH, Rouse DI, Stevenson WR, 2007.
Multiplex real-time quantitative PCR to detect and quantify Verticillium dahliaecolonization in potato lines that differ in response toVerticilliumwilt.Phytopathology97, 865–72.
Ben-Noon E, Shtienberg D, Shlevin E, Vintal H, Dinoor A, 2001.
Optimization of chemical suppression ofAlternaria dauci, the causal agent of alternaria leaf blight in carrots.Plant Disease 85, 1149–56.
Bennett MD, Smith JB, 1976. Nuclear DNA amounts in
angiosperms.Philosophical Transactions of the Royal Society of London, Series B274, 227–74.
Berruyer R, Poussier S, Kankanala P, Mosquera G, Valent B, 2006.
Quantitative and qualitative influence of inoculation methods onin plantagrowth of rice blast fungus.Phytopathology96, 346–55.
Boedo C, Le Clerc V, Briard Met al., 2008. Impact of carrot resistance on development of theAlternarialeaf blight pathogen (Alternaria dauci).European Journal of Plant Pathology121, 55–66.
Bo¨hm J, Hahn A, Schubert Ret al., 1999. Real-time quantitative PCR: DNA determination in isolated spores of the mycorrhizal fungusGlomus mosseaeand monitoring ofPhytophthora infestansandPhytophthora citricolain their respective host plants.Journal of Phytopathology147, 409–16.
Cadot V, Boulineau F, Gue´nard M, Olivier V, Molinero-Demilly V, 2002. Setting up a resistance test toAlternaria dauciof carrot by inoculation in the open field, as part of registering varieties in the National French Catalogue of Vegetable Species. In: IVe`me Rencontres de Phytopathologie⁄Mycologie, Journe´es J. Chevaugeon, Aussois, France.
Dita MA, Brommonschenkel SH, Matsuoka K, Mizubuti ESG, 2007. Histopathological study of theAlternaria solani infection process in potato cultivars with different levels of early blight resistance.Journal of Phytopathology155, 462–9.
Dugdale LJ, Mortimer AM, Isaac S, Collin HA, 2000. Disease response of carrot and carrot somaclones toAlternaria dauci.
Plant Pathology49, 57–67.
Farrar JJ, Pryor BM, Davis RM, 2004.Alternariadiseases of carrot.Plant Disease88, 776–84.
Floris E, Alvarez JM, 1996. Nature of resistance of seven melon lines toSphaerotheca fuliginea.Plant Pathology45, 155–60.
Gaube C, Dubourg C, Pawelec A, Chamont S, Blancard D, Briard M, 2004. Bruˆlures foliaires parasitaires de la carotte.Alternaria daucisous surveillance.PHM Revue Horticole454, 15–8.
Gugino BK, Carroll JE, Widmer TL, Chen P, Abawi GS, 2007.
Field evaluation of carrot cultivars for susceptibility to fungal leaf blight diseases in New York.Crop Protection26, 709–14.
Kra¨mer R, Nothnagel T, Rabenstein F, 2009. Comparison of resistance response of carrot accessions (Daucus carotaL.) to Alternaria dauciandA. radicina. In:Proceedings of the 33rd International Carrot Conference. Anaheim, California, USA: PM-103.
Larsen RC, Vandemark GJ, Hughes TJ, Grau CR, 2007.
Development of a real-time polymerase chain reaction assay for quantifyingVerticillium albo-atrumDNA in resistant and susceptible alfalfa.Phytopathology97, 1519–25.
Le Clerc V, Pawelec A, Birolleau-Touchard C, Suel A, Briard M, 2009. Genetic architecture of factors underlying partial resistance toAlternarialeaf blight in carrot.Theoretical and Applied Genetics118, 1251–9.
Mo¨ller EM, Bahnweg G, Sandermann H, Geiger HH, 1992.
A simple and efficient protocol for isolation of high molecular weight DNA from filamentous fungi, fruit bodies, and infected plant tissues.Nuleic Acids Research22, 6115–6.
Pawelec A, Dubourg C, Briard M, 2006. Evaluation of carrot resistance toAlternarialeaf blight in controlled environments.
Plant Pathology55, 68–72.
R Development Core Team, 2005.R: a Language and Environment for Statistical Computing Version 2.2.1.
Vienna, Austria: R Foundation for Statistical Computing, http://
www.R-project.org.
Rasmussen JB, Morrisson T, Herrmann M, Wittwer C, 1998.
Quantitative PCR by continuous fluorescence monitoring of a double strand DNA specific binding dye.Biochemica2, 8–11.
Silvar C, Diaz J, Merino F, 2005. Real-time polymerase chain reaction quantification ofPhytophthora capsiciin different pepper genotypes.Phytopathology95, 1423–9.
Simon PW, Strandberg JO, 1998. Diallel analysis of resistance in carrot toAlternarialeaf blight.Journal of the American Society for Horticultural Science123, 412–5.
Thomma BPHJ, Eggermont K, Tierens KFMJ, Broekaert WF, 1999. Requirement of functional ethylene-insensitive 2 gene for efficient resistance ofArabidopsisto infection byBotrytis cinerea.
Plant Physiology121, 1093–102.
Vintal H, Ben-Noon E, Shlevin E, Yermiyahu U, Shtienberg D, Dinoor A, 1999. Influence of rate of soil fertilization on alternaria leaf blight (Alternaria dauci) in carrots.
Phytoparasitica27, 193–200.
