Abstracts

GENETIC DIVERSITY AMONG ISOLATES OF Stemphylium solani FROM COTTON

Y.R. MEHTA

Instituto Agronomico do Paraná-IAPAR, Cx. Postal 481, Londrina, PR, Brazil, e-mail: yrmehta@pr.gov.br Fax: (43) 376-2101

(Aceito para publicação em 13/06/2001)

MEHTA, Y.R. Genetic diversity among isolates of Stemphylium solani from cotton. Fitopathologia Brasileira 26:703-709. 2001.

ABSTRACT

The fungus Stemphylium solani causes leaf blight of tomato (Lycopersicon esculentum) in Brazil. In recent years, severe epidemics of a new leaf blight of cotton (Gossipium hyrsutum) caused by S. solani occurred in three major cotton-growing Brazilian states (PR, MT and GO). Molecular analysis was performed to assess the genetic diversity among the S. solani isolates from cotton, and to verify their relationship with representative S. solani isolates from tomato. Random amplified polymorphic DNA (RAPD) markers and internal transcribed spacers of ribosomal DNA (rDNA) were used to compare 33 monosporic isolates of S. solani (28 from cotton and five from tomato). An isolate of Alternaria macrospora from cotton was also used for comparison. RAPD analysis showed the presence of polymorphism between the genera and the species. The A. macrospora and the S. solani isolates from cotton and tomato were distinct from each other, and fell into separate groups. Variation by geographic region was observed for the tomato isolates but not for the cotton isolates. Amplifications of the ITS region using the primer pair ITS4/ITS5 resulted in a single PCR product of approximately 600 bp for all the isolates. Similarly, when amplified fragments were digested with eight restriction enzymes, identical banding patterns were observed for all the isolates. Hence, rDNA analysis revealed no inter-generic or intra-specific variation. The genetic difference observed between the cotton and the tomato isolates provides evidence that S. solani attacking cotton in Brazil belongs to a distinct genotype.

Additional Key words: Gossypium hirsutum, Lycopersicon esculentum, Stemphylium leaf blight, Molecular genetics, RAPD.

RESUMO

Diversidade genética entre isolados de Stemphylium solani de algodoeiro

Stemphylium solani causa manchas foliares em tomateiro (Lycopersicon esculentum). Em anos recentes, epifitias severas de uma nova doença do algodoeiro (Gossipium hirsutum) causada por S. solani ocorreu em três estados brasileiros (PR, MT e GO). Análise molecular foi realizada para verificar variabilidade entre os isolados de S. solani de algodoeiro e ao mesmo tempo verificar sua relação com isolados representativos de S. solani de tomateiro. Ensaios de polimorfismo de DNA amplificado ao acaso (RAPD) e regiões internamente transcritas de DNA ribossomal (rDNA) foram utilizados para analisar 33 isolados monosporicos de S. solani (28 de algodoeiro e cinco de tomateiro). Um isolado de Alternaria macrospora de algodoeiro também foi incluido para comparação. Análise de RAPD demonstrou a presença de polimorfismo entre os gêneros e espécies. Os isolados de A. macrospora e S. solani de algodoeiro e tomateiro eram distintos, e formaram grupos separados. Variação de acordo com a região geográfica foi encontrada para os isolados de tomateiro, mas não para os isolados de algodoeiro. A amplificação da região de ITS utilizando-se o par de primers ITS4/ITS5 resultou num produto de PCR de aproximadamente 600 pb para todos os isolados. Da mesma forma, quando os fragmentos amplificados foram digeridos com oito enzimas de restrição, padrões idênticos de bandas foram observados para todos os isolados. Desta forma, a análise de rDNA não revelou diferença intergenérica ou intra-específica. A separação de isolados de algodoeiro com os isolados de tomateiro sugere que eles pertencem a um genótipo diferente de S. solani atacando algodoeiro no Brasil.

INTRODUCTION

Cotton (Gossypium hirsutum L. or G. barbadense L.) cultivars are susceptible to several foliar pathogens, including Alternaria alternata (Fr.) Keissler and A. macrospora (Zimm.). In recent years, severe epidemics of a new leaf blight caused by Stemphylium solani Weber. occurred in three major cotton-growing Brazilian states (Paraná, Mato Grosso and Goiás), and especially in the State of Paraná, causing up to 100% yield losses in some commercial fields (Mehta, 1998). Since S. solani also causes leaf blight of tomato (Lycopersicon esculentum L.) in Brazil, it was not clear whether the cotton pathogen was the same as the tomato pathogen. Considering S. solani as a new pathogen of cotton in Brazil, it was also not clear whether it was introduced with cotton germplasm imported from other countries (Butler & Jones, 1956; Kurozawa & Mussi, 1995; Mehta, 1998). In recent years, Stemphylium leaf blight of tomato (grey leaf spot of tomato) has become a rare disease due to the introduction of tomato hybrids highly resistant to this disease. Additionally, tomato fields are normally protected against other fungal diseases by periodical fungicidal sprays.

Testing over 400 seed lots of cotton of different cultivars collected from three major growing areas of the Brazilian states was conducted during 1998 and 1999 using the "blotter test" (Mathur & Cunfer, 1993) and a semi-selective agar medium (Mehta, unpublished data). None of the seed lots tested showed any infection or contamination by S. solani or any other Stemphylium species, leading to speculation that S. solani is not transmitted by seed.

In previous cross inoculation tests, Stemphylium isolates from cotton were highly aggressive to cotton, tomato, potato (Solanum tuberosum L.) and blue lupinus (Lupinus angustifolius L.), whereas an isolate of S. solani from tomato was highly aggressive on tomato and potato but less aggressive on cotton (Mehta, 1998). Although gross morphological characteristics and the conidial dimensions were similar, conidia of cotton isolates were slightly shorter and narrower, smooth-walled in contrast to the verrucose walls of S. solani from tomato, and showed abundant germination from cells of the conidia throughout the conidium body - a feature that has not been reported before for S. solani (J.C. David, IMI, England, personal communication; Mehta 1998). Subsequently, Brogin & Mehta (1997) reported that S. solani isolates from cotton produced a cultivar-specific phytotoxin, which was not produced by S. solani isolates from tomato. Because such morphological and pathological differences were not sufficient to distinguish the isolates at the specific level, it was believed that the cotton isolates belonged to a new strain of S. solani.

Control of Stemphylium leaf blight of cotton by fungicidal sprays is not cost effective (Mehta & Oliveira, 1998). The disease can be economically controlled by the use of resistant cultivars (Cavalcanti et al., 1999). However, success in breeding for disease resistance depends on the knowledge of the diversity of the pathogen population. Molecular analysis can be applied to estimate genetic diversity of the pathogen population so that the germplasm of national importance is tested for resistance against all different pathogen genotypes occurring across the diverse cotton growing areas of the State.

In recent years, random amplified polymorphic DNA (RAPD) and ribosomal DNA (rDNA) analyses have been extensively employed in phylogenetic and taxonomic studies of fungi (Achenback & Patrick, 1996; Achenback et al., 1997; Machado et al., 1997; Ouellet & Seifert, 1993; Williams et al., 1990). The internal transcribed spacer (ITS) region of the rDNA has been useful in distinguishing relationships at the species level (Achenbach et al., 1997). Kusaba & Tsuge (1995) studied phylogeny of Alternaria fungi known to produce host-specific toxins on the basis of variation in the internal transcribed spacers of ribosomal DNA, and suggested that Alternaria pathogens that produce host-specific toxins are pathogenic variants within the species A. alternata. As no previous molecular analyses of S. solani from any plant species were encountered in the literature, the present study was undertaken to assess the genetic diversity among the S. solani isolates from cotton, and verify their relationships with representative S. solani isolates from tomato, utilizing rDNA and RAPD analyses.

MATERIALS AND METHODS

Cultural characteristics and pathogenicity tests

Isolation of S. solani was made on potato dextrose agar (PDA) from diseased leaves from over 12 commercial cotton cultivars grown at agro-climatically different locations (Table 1). Monosporic cultures were made and stored on PDA slants for further use. An isolate of A. macrospora from cotton from the State of Mato Grosso was used for comparison, since it produces symptoms somewhat similar to those produced by S. solani (Mehta, 1998). Mycelial plugs of 0.5 cm diameter cut from the margins of eight-day-old colonies were placed in the centre of plates containing 20 ml of PDA. Colony diameter and colour, formation of sectors and production of pigment were recorded on three replicate plates after ten days incubation at 25 ^oC with 12 h fluorescent light/dark periods.

Pathogenicity tests of all the fungal isolates were made on the susceptible cotton cultivar IAPAR-71. Fungal colonies grown on three plates per isolate were used for inoculum preparation. Maximum sporulation was obtained within 10-12 days on PDA containing 50 g (fresh weight) each of macerated young green leaves of "mucuna" (Styzolobium atterrimum Piper et Tracy) and lupinus per litre and hence were used for inoculum preparation. Ten-day-old colonies were gently scraped from PDA plates with 20 ml of sterile distilled water per plate. A suspension of mycelial fragments and conidia (9,000 conidia ml^-1) plus one drop of Tween 20 was prepared. Fifteen-day-old seedlings of cotton cv. IAPAR 71 were inoculated with a soft paint brush (Mehta, 1998).

Five to six seedlings were inoculated with each isolate, and uninoculated plants were maintained as controls. Inoculated seedlings were incubated in a dew chamber for 48 h at 21 ^oC and then transferred to a glasshouse, where the day and night temperatures means during the test period were 25 and 18 ^oC, respectively. Disease symptoms were evaluated eight days after inoculation on leaves that showed the most severe symptoms by using a visual scale of 0 to 5, where 0 =no disease; 1 =minute pinhead size spots, less than 5% leaf tissue diseased; 2 =small brown to dark brown necrotic lesions with no chlorosis, 5-25% diseased; 3 =necrotic lesions with chlorosis, 26-50% diseased; 4 =necrotic lesions with chlorosis, up to 10 cm diameter, usually coalescing, 51-75% diseased; and 5 =lesions coalescing, >76% diseased, and/or with desiccated and abscised leaves (Mehta, 1998).

DNA extraction

Thirty-three monosporic isolates of S. solani (28 from cotton collected during 1995-96, two from tomato from the State of São Paulo, collected in 1993, and three from tomato from the State of Goiás, collected in 1995) were analysed with RAPD and rDNA (Table 1). An isolate of A. macrospora from the State of Paraná, collected in 1997, from cotton was also included for comparison. All fungal cultures were grown in potato dextrose broth on a rotary shaker for four days. Mycelium was filtered through Whatman No. 1 filter paper, washed twice with TE buffer (10 mM Tris-HCl; 1 mM EDTA, pH 8.0) and total DNA extracted as described by Raeder & Broda (1985). DNA was quantified by gel electrophoresis as well as by a DyNa Quant 200 Fluorometer (Pharmacia). DNA samples were diluted in TE buffer to give a concentration of 20 ng mL^-1 and stored at - 21 ^oC for further use. Samples that showed degradation of DNA in electrophoresis gels were discarded and DNA extraction repeated. RNA was eliminated by addition of RNAase (1.0 mg/ml).

RAPD analysis

PCR reactions were performed in 25 ml volumes containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2 mM MgCl₂, 200 mM of each dATP, dCTP, dGTP and dTTP, 5 picomoles of primer, 100 ng of genomic DNA and one unit of Taq polymerase (Pharmacia, USA). Negative controls without DNA were included in all the reactions. Of 33 decamer primers (Operon Technologies Inc., Alameda, CA) tested using four arbitrarily selected S. solani isolates from cotton, eight were informative and were selected for the present studies (Table 2). Selection of primers was solely based on the generation of relatively larger numbers of bands (whether monomorphic or polymorphic), in order to avoid any bias in the selection process. Amplification was performed on a thermal cycler (MJ Research, Inc. Watertown, MA, USA), according to the following program: 94 ^oC for 3 min followed by 40 cycles of 94^o C for 1 min, 40 ^oC for 1 min, 72 ^oC for 2 min, and a final extension at 72 ^oC for 5 min. The amplification products (25 ml) were electrophoresed in 1.4% agarose gels with TBE running buffer, stained with ethidium bromide and were scanned into a computer imaging file using a Kodak EDAS 120 digital camera. All RAPD reactions were repeated once to confirm the presence of some faint bands and only reproducible bands were scored. RAPD data were analysed considering the presence or the absence of bands. A similarity matrix was analysed by the unweighted pair group method (UPGMA in the NTSYS-pc, version 1.8) using Jaccard's coefficient. The similarity matrix and dendrogram were calculated using both the monomorphic as well as polymorphic fragments.

Amplification of ITS region

Sequences for the primers ITS4 and ITS5 region were 5´-TCCTCCGCTTATTGATATGC-3´ and 5´-GGAAGTA AAAGTCGTAACAAGG-3´, respectively (White et al., 1990). Amplification was performed in a thermal cycler (MJ Research, Inc. Watertown, MA, USA) using the following program: 94 ^oC for 3 min followed by 35 cycles of 94 ^oC for 1 min, 55 ^oC for 1 min, 72 ^oC for 2 min and a final extension at 72 ^oC for 5 min. Amplifications were performed in 50 ml volumes containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2 mM MgCl₂, 200 mM of each dATP, dCTP, dGTP, dTTP, 25 pmol of each primer, 100 ng of genomic DNA, and 1 unit of Taq polymerase. Amplification products (5 ml of a 50 ml reaction) were electrophoresed in 1.4% agarose gels with TBE running buffer, stained with ethidium bromide and photographed to verify the fragment size. The amplified fragments were digested with eight arbitrarily selected endonucleases (Bam II, Bgl II, Dra I, EcoR I, Hind III, Hinf I, Taq I, Xho II) in a 20 ml volume containing 1 ml of the restriction enzyme. Digestion was performed as per the instructions of the manufacturer and fragments were separated by electrophoresis in 2% agarose gels. All amplifications and digestions were repeated once to verify the reproducibility of the results.

RESULTS

Cultural characteristics and pathogenicity tests

Stemphylium solani is a very variable fungus. A range of different colony characteristics of S. solani were found among the cotton and the tomato isolates grown on PDA (Table 1). Colonies were slow growing on PDA, reaching a maximum of 66 mm in diameter 10 days after incubation and were dark brown to black. Mycelium was velvety, cottony or immersed. Some of the isolates produced a yellow pigment in the medium that turned deep red with age.

All the cotton isolates were pathogenic to the susceptible cotton cv. IAPAR-71, whereas the tomato isolates showed some variation. The first disease symptoms appeared four days after inoculation as small, sunken, circular to irregular spots on inoculated plants. Of the five tomato isolates, two isolates from the State of Goiás were non-pathogenic to cotton. The A. macrospora isolate used for comparison was highly aggressive on cotton.

RAPD analysis

No banding pattern was observed in the negative controls without DNA. Of the 58 amplified products generated with the eight primers and 34 fungal isolates, 30 were polymorphic. RAPD analysis showed the presence of polymorphism between the isolates, especially according to their host origin. The tomato isolates from the State of São Paulo (isolate Nos. 1 and 2) for example, were clearly distinct from the other three tomato isolates from the State of Goiás, as well as from the 28 cotton isolates (Figures 1A and 1B). Amplification with primer OPA-02 resulted in a marker band OPA-02₇₀₀, which was specific only for the tomato isolates from the State of Goiás, whereas the marker band OPA-02₁₄₀₀ was specific for the tomato isolates from São Paulo, as well as one A. macrospora isolate. One marker band OPY-02₅₀₀ was found to be exclusive to two tomato isolates from the State of São Paulo and to one tomato isolate from Goias (isolate No. 32). The specificity of such bands was confirmed in at least two repetitions. As expected, the A. macrospora isolate was completely distinct from S. solani isolates whereas the cotton isolates of S. solani mostly formed two major groups (Figure 2). At the 0.75 similarity coefficient, about 86% of the cotton isolates fell into one large group, whereas the rest of the isolates were separated into a second group at about 0.6 similarity coefficient.

Amplifications of the ITS region

Amplifications of the ITS region using primer pair ITS4/ITS5 resulted in a single PCR product of approximately 600 bp for all the isolates and in all the repetitions conducted with fresh DNA samples. When amplified fragments were digested with eight restriction enzymes, identical banding patterns were observed for all the S. solani isolates as well as for one isolate of a closely related genus Alternaria (A. macrospora). Hind III showed two bands of approximate sizes of 400 bp and 200 bp, Hinf I showed one band of 300 bp and another of 150 bp (probably a double restriction fragment) (Figure 3). Taq I showed one band of 250 bp (probably a double restriction fragment) and an other of 100 bp. EcoR I showed only one band of 300 bp (probably a double restriction fragment), whereas Xho II, Bam II and Dra I, showed a single band of 600 bp indicating that the fragments were not digested. Thus the amplification of ITS region revealed no inter-generic or intra-specific variation.

DISCUSSION

RAPD analysis is extremely powerful and can separate individuals having intra and inter-specific variability. It gives more comprehensive information regarding the genetic variability among the pathogen populations as it is based on the entire genome of an organism (Achenback et al., 1997). The separation of cotton isolates from tomato isolates on the basis of RAPD data provides evidence that the cotton leaf pathogen may belong to a distinct genotype of S. solani specific to cotton in Brazil, and suggests that it probably did not originate from the tomato leaf blight pathogen. As leaf blight of cotton is a new disease in Brazil, almost no information is available on the epidemiology and the biology of S. solani, either from cotton or from tomato. The perfect state of most of the Stemphylium species belongs to Pleospora (Fr.) (Simmons, 1969). The perfect state is formed under prolonged cold conditions (4 ^oC) that do not prevail in Brazil, and consequently, there is no report about the occurrence of the perfect state of this fungus, either from cotton or from tomato. Thus, the variation between the isolates from these two crops found by RAPD analysis, is probably not due to the sexual hybridisation. Although in the present study only one isolate of S. solani from the State of Mato Grosso and one from the State of Goiás were included, compared with 28 isolates from the State of Paraná, differences between the cotton isolates from the three States were not seen. As a result of the introduction of several tomato hybrids highly resistant to Stemphylium leaf blight, our efforts to include a larger number of representative isolates from tomato, including some from the State of Paraná, were unsuccessful. Consequently, differences between the tomato isolates must be treated with caution, as they may in fact represent geographical variation and not host specialization.

Analyses of restriction site differences in the ITS region were not sufficiently sensitive to detect differences among the populations and pathotypes of S. solani attacking two different crops. Given the taxonomic controversy due to the conidial differences between isolates pathogenic on cotton and tomato, restriction site homogeneity provides support for both as members of the same species. However, given the lack of variability observed in Alternaria, complete clarification will only be achieved following sequencing of ITS region. Similar results were obtained with Fusarium solani (Mart.) Appel & Wollenw. Emend. Snyd. & Hans., causing sudden death syndrome of soybean [Glycine max (L.) Merrill] (Achenback et al., 1997) and Mycosphaerella pinodes Berk. & Blox. and Phoma medicaginis Pass., causing Ascochyta blight of pea (Pisum sativus L.) (Onfroy et al., 1999).

Stemphylium solani from cotton is a slow-growing fungus. Some isolates do not sporulate on common culture media like PDA and, hence, the process of identification is delayed. The use of RAPD analysis offers further possibilities for rapid diagnosis of this pathogen.

In pathogenicity studies conducted previously on 48 cotton cultivars, Cavalcanti et al. (1999), using 14 monosporic isolates of S. solani, originating from different cultivars and locations, observed a very small mean square for the genotype-isolate interactions (0.2) as compared with the genotype and isolate main effects (1.8 and 9.5, respectively). This indicates that the differences among the isolates were probably due to the aggressiveness and not due to differences in virulence. The molecular analysis presented herein also revealed relatively little variation among the S. solani isolates from cotton, suggesting that using a number of different isolates may not be necessary when screening cotton germplasm against this pathogen. Such results might help in selecting appropriate breeding strategies against this new cotton pathogen. Results presented here may also form the basis for further studies using larger samples to address the taxonomic questions related to S. solani populations originating from different hosts.

ACKNOWLEDGEMENTS

Thanks are due to C. Kurozawa and C.A Lopes for providing some fungal strains from tomato. T.K.K. Tame, J.A. Oliveira, A. Souza, A. J. Souza and E.H. Ota provided technical assistance. Thanks are also due to R.N.G. Miller for his valuable suggestions in the manuscript.

LITERATURE CITED

Publication Dates

History