Variation In Aggressiveness And Aflp Among Alternaria Solani Isolates From Indonesia

Alternaria solani is a necrotroph fungus that causes three-phased diseases in tomato. Management of the pathogen by using resistant cultivars requires knowledge on the aggressiveness and genetic diversity of the fungus. The aims of this study were to isolate A. solani from major tomato and potato producing areas in Indonesia and to study their aggressiveness and genetic variability. Twenty two A. solani isolates were recovered from early blighted tomato and potato in Central and West Java. A. alternata was also isolated from tomato leaves in West Java and North Sumatra, indicating that early blight in Indonesia may be caused by more than one Alternaria species. Resistance tests of four tomato genotypes to selected A. solani isolates revealed that local isolates were more aggressive in inciting early blight and stem lesion than an imported isolate from USA. This implies that introduced breeding materials must be tested to local isolates to obtain effective resistance genes. Cluster analysis based on amplified fragment length polymorphism (AFLP) obtained from EcoRI+AG and MseI+C primer amplification separated 28 local and Taiwan isolates from the US isolate, which was coincided with aggressiveness separation between the local isolates and the US isolate. Three clusters of AFLP genotypes which did not associate with geographic origin were observed among tropical isolates. The low genetic diversity among the Indonesian isolates suggests clonal population structure with wide distribution. Successful local tomato breeding requires the availability of local A. solani collection with well-characterized aggressiveness level and molecular diversity to obtain effective resistance genes.

from EcoRI+AG and MseI+C primer amplification separated 28 local and Taiwan isolates from the US isolate, which was coincided with aggressiveness separation between the local isolates and the US isolate. Three clusters of AFLP genotypes which did not associate with geographic origin were observed among tropical isolates. The low genetic diversity among the Indonesian isolates suggests clonal population structure with wide distribution. Successful local tomato breeding requires the availability of local A. solani collection with well-characterized aggressiveness level and molecular diversity to obtain effective resistance genes.

INTRODUCTION
The necrotroph fungus Alternaria solani (Ellis & Martin) Sorauer incites early blight of foliage, collar rot of seedlings or stem lesion of adult plants of tomato (Solanum lycopersicum) and lesion on tomato fruit (Chaerani and Voorrips 2006). On potato (Solanum tuberosum) the fungus also infects tubers during storage in addition to causes early blight disease (Odilbekov et al. 2016). Early blight is the most common and destructive phase of the disease syndromes because it can cause premature defoliation which weakens plants and exposes fruits to sunscald injury (Chaerani et al. 2007). Early blight symptoms on foliage are characterized by dark brown to black lesions with concentric rings, which produce a "target spot" effect (Sherf and MacNab 1986). Necrotic lesions first affect older leaves and spread upward as plants become mature.
In Indonesia early blight epidemics was reported for the first time in the early 1900s to severely devastated potato fields in the highlands of West Java and later spread to potato fields at lower elevations causing considerable yield losses (Chaerani 2006). The disease soon spread to potato fields in the highlands of North Sumatra in 1920s. At present early blight is one of the common tomato and potato diseases in Indonesia, especially on tomato grown in lower altitudes, where late blight is less thriving (Chaerani 2006).
A. solani is difficult to control since it has a short life cycle, is primarily air borne, has long survival ability in plant debris and soil, and has a wide solanaceaous host range (Chaerani and Voorrips 2006). Host resistance is the most economically and environmentally sound approach over fungicidal spray to restrict early blight incidence in the field, but tomato and potato cultivars with high level of early blight resistance and desirable horticultural traits have not been identified (Gardner and Panthee 2012;Odilbekov 2015;Adhikari et al. 2017).
Effectiveness of host resistance to A. solani is greatly influenced by the genetic variation of the pathogen (Odilbekov et al. 2016). Previous studies revealed considerable variability among isolates with respect to morphology, physiology, biochemistry, aggressiveness and genetics (Bonde 1929;Henning and Alexander 1959;Petrunak and Christ 1992;Weir et al. 1998;Castro et al. 2000;Martinez et al. 2004;van der Waals et al. 2004;Varma et al. 2006;Lourenço et al. 2009;Lourenço Jr et al. 2011;Weber and Halterman 2012;Meng et al. 2015;Kumar et al. 2016;Odilbekov et al. 2016). Great variability in A. solani is unexpected, since the sexual stage of this pathogen is not known (Meng et al. 2015). The occurrence of differential aggressiveness among isolates has led to the conclusion for the presence of physiological races (Bonde 1929). However, the race concept in A. solani has not been widely accepted due to lack of characterized differential varieties (Chaerani and Voorrips 2006;Gannibal et al. 2014;Adhikari et al. 2017) Genetic variation within A. solani populations has been assessed using various molecular markers which included random amplified polymorphic DNA (RAPD) (Weir et al. 1998;Varma et al. 2006;Kumar et al. 2008;Lourenço Jr et al. 2011;Weber and Halterman 2012;Leiminger et al. 2013), random amplified microsatellites (RAMS) (van der Waals et al. 2004), amplified fragment length polymorphism (AFLP) (Martinez et al. 2004;Lourenço Jr et al. 2011;Odilbekov et al. 2016), and simple sequence repeats (SSRs) (Meng et al. 2015), as well as sequence analyses of A. solani genome portion (Lourenço et al. 2009;Weber and Halterman 2012). Among different types of molecular markers, AFLP marker has become one of the major methods of choice for studies of genetic diversity because it does not require prior knowledge of genomic sequence and generates a large number of polymorphic loci in a single experiment (Martinez et al. 2004;Odilbekov et al. 2016). The highly polymorphic nature of AFLP marker makes it especially useful for differentiating clonal lineages of fungi that reproduce asexually (McDonald 1997).
AFLP DNA analysis combines the power of restriction fragment length polymorphism (RFLP) with the flexibility of PCR-based technology (Vos et al. 1995). The technique is based on a multistep process which includes DNA digestion with a rare and frequent cutter, fragment ligation to adaptors, pre-selective PCR amplification, selective PCR amplification, and fragment separation. The use of two selective primers under stringent PCR conditions gives the AFLP technique reproducible results unlike RAPDs which use one arbitrary decamer primers.
Previous AFLP analyses of A. solani isolates from tomato and potato revealed great diversity among exotic and local isolates in Cuba (Martinez et al. 2004), whereas clonal population with high genetic variability was observed among Brazilian isolates (Rodrigues et al. 2008;Lourenço Jr et al. 2011). Exotic isolates from China, Turkey, Greece and USA tended to group separately from isolates from Cuba and Brazil. Both studies found genetic differentiation between isolates from tomato and potato. The combination of molecular markers with pathogenicity assays represents a useful tool in pathogen population genetics (Bouajila et al. 2007). However, little information is available on the association of A. solani genotypic diversity with aggressiveness. Van der Waals et al. (2004) found a relatively high variability in RAMS profiles among potato A. solani isolates from South Africa, but failed to obtain apparent correlation with isolate aggressiveness.
Despite its widespread distribution in Indonesia, intensive collection and genetic studies of the pathogen has not been pursued before. The knowledge of genetic diversity of this pathogen is useful for improvement of early blight disease management using resistant cultivars and for selection of resistance genes in breeding program. The objective of this research was thus to reveal the variability among A. solani collected from major tomato and potato producing areas in Indonesia based on aggressiveness of selected isolates and AFLP analysis.

Fungal Cultures
Tomato and potato leaves with typical early blight lesions collected from main tomato and potato producing areas in 10 localities of West and Central Java, and North Sumatra were transported in a cooler box or packed in paper bags containing silica gels and brought to the laboratories of Indonesian Ornamental Crops Research Institute (IOCRI) and Indonesian Vegetables Research Institute (IVEGRI) for isolation.
A. solani was isolated by using either monospore isolation or diseased tissue plating technique on 1.5% water agar (WA). Purification was done by transferring hyphal tips to potato dextrose agar (200 g potato, 20 g dextrose, 15 g agar, and dH 2 O to 1 liter). Cultures were maintained on V8 juice agar medium (200 ml V8 juice, 3 g CaCO 3 , 20 g agar, and dH 2 O to 1 liter) (Ribeiro 1978) at 21-22 o C in a 12-h diurnal period of fluorescent light. For each isolate, single-spore cultures were established on V8 juice agar medium. Species identification was performed morphologically based on colony appearance and conidial characteristics (Ellis and Gibson 1975).

Aggressiveness Test of Selected Isolates
Inoculation experiments were done at Plant Research International (PRI) glasshouse. Three A. solani isolates selected to represent different geographic localities in West Java (Cianjur, Sukabumi and Lembang) and one isolate from USA contributed by Dr. E. Cobb (Department of Plant Pathology, Cornell University, Ithaca, New York, USA) were compared for their aggressiveness on four tomato genotypes. A. solani sporulation requires special conditions which include mycelial wounding, exposure to florescent light and partial desiccation (Barksdale 1969;Rodrigues et al. 2010a). Sporulation was induced by following Barksdale (1969) procedure.
Aerial mycelia were scraped off from 7 to 10-day old cultures using a scalpel and the plates were returned to the same incubation conditions with the bottom of the Petri dishes facing down and leaning on inverted lids. After 48 h spores were removed using an L-shaped glass rod in the presence of 5 ml sterile dH 2 O. Seeds of three with the four genotypes planted in two alternating rows. Boxes were placed on a bench lined with a wet mat. Leaves were sprayed with A. solani spores at a rate of 10 3 ml -1 tap water until run-off using a glass atomizer. Immediately after spraying, the boxes were covered with a transparent lid and placed on a bench lined with a wet mat. The bench was covered with a transparent plastic forming a tunnel (Figure 1a and c). Intermittent misting for 15 min at 45-min intervals to maintain high humidity inside the tunnel was supplied from a humidifier ( Figure 1b). After 24 h the lids were removed and each side of the plastic tunnel was opened.
During the day the humidifier was turned off for 8 h to allow the plant surface to dry. The tunnel was closed during the night. The glasshouse received periodic misting for 16 h to maintain the relative humidity from 40% to 72% during the day and 85% to 100% during the night. The temperatures in the glasshouse were maintained at 20-27 o C during the day and 16-24 o C during the night. Treatments were replicated three times. A same experimental set up was performed to investigate the aggressiveness of the four isolates in inciting stem lesion severity. Spores (10 3 ml -1 in tap water) were sprayed onto plant stem base until run-off.
Early blight severity on each leaf was recorded four times at 3, 7, 10 and 14 days after inoculation (DAI) on a scale of 0 to 5, where 0 = no visible lesions on leaf; 1 = up to 10% leaf area affected; 2 = 11 to 25%; 3 = 26 to 50%; 4 = 51 to 75%; and 5 = more than 75% leaf area affected or leaf abscised (Vakalounakis 1983). Leaves that were not completely unfurled during the inoculation were not assessed. The disease scales were converted into percentage of early blight index (PEBI) for each plant using the following formula (Pandey et al. 2003): PEBI = (sum of all ratings × 100)/(no. of leaves sampled × maximum disease scale). PEBI was used to calculate the area under the disease progress curve (AUDPC) using the formula: where R i is the PEBI at the ith observation, t i is time (days after planting) at the ith observation, and n is the total number of observations (Nash and Gardner 1988). Stem lesion severity was recorded twice at 10 and 14 DAI using a modified scale from Barksdale and Stoner (1977), where 0 = no lesions, 1 = lesion size 1 to 2 mm, 2 = lesions between 2 mm and 1 cm, and 3 = lesion size equal to or more than 1 cm. Stem lesion index and AUDPC were calculated using the same formulas as described. AUDPC values were analyzed as 4×4 factorial experiments in a randomized completely block design.

DNA Isolation
Genomic DNA was extracted using DNA extraction kit (Puregene-Biozym BV) from mycelia which were scraped off from 7 to 10-day old cultures grown on V8-juice agar medium in 9-cm diameter Petri plates. DNA concentration was estimated by 1% agarose gel electrophoresis and compared to that of lambda DNA.

AFLP Analysis
AFLP analysis was performed at the PRI following the procedure described in Vos et al. (1995). Briefly, genomic DNA (50 ng) from each isolate was digested with the six-base restriction enzyme EcoRI and the four-base restriction enzyme MseI, then the digested fragments were ligated with EcoRI-biotine and MseIbiotine adaptors in the presence of T4 DNA ligase. The resulting primary template was diluted and pre-amplified with EcoRI primer (E00) and MseI primer (M00). The pre-amplification product was diluted and used as the secondary template for selective amplification with EcoRI+AG and MseI+C primers which contained two and one selective bases, respectively, at their 3' end after the core sequences. The EcoRI primer was radio labeled with 32 P (Invitrogen).
AFLP fragments were then separated on a 6% denaturing polyacrylamide gel run with a constant power of 70 W for 2 h 45 min using OWL sequencing gel system. DNA fragment sizes were estimated by comparing migration distance to that of a 30-330-bp AFLP ladder which was also radio labeled with 32 P. The gels were placed into a cassette and exposed to Kodak X-OMAT AR film overnight. AFLP fragments between 50 and 330 bp were scored manually by eye for the presence (1) or absence (0) for each isolate. Only bands that could be scored unambiguously were recorded. Four A. solani isolates obtained from Ir. Gunarto (East-West Seed Indonesia), two from Dr. L.L. Black (AVRDC, Taiwan), and the US isolate were included in the AFLP analyses. In the case where isolates showed cultural variation, each variant was also run in the gel analysis to check whether morphological variation can be differentiated at molecular level. Two A. alternata isolates obtained in the surveys (see Results and Discussion) were also included as comparison in the analysis. Several genetic studies showed that AFLPs were able to distinguish fungal species within the same genus regardless of the primer combination used (Casasnovas et al. 2013;Oviedo et al. 2013).

Phylogenetic Analysis
The resulting binary data describing the presence or the absence of AFLP fragments at all scored loci were used to generate a similarity matrix based on simple matching coefficient (Sneath and Sokal 1973). The using the sequential agglomerative hierarcical nested (SAHN) cluster analysis and a dendrogram was constructed using the unweighted pair-group method of arithmetic averages (UPGMA) module available in the software. All calculations were done with the aid of the software NTSYSpc (v2.01, Exeter Software, Setauket, NY, USA).

Fungal Isolation
Twenty two A. solani isolates were successfully recovered from infected tomato and potato leaves collected from nine localities surveyed in West and Central Java (Table 1). Tomato leaf samples collected in Lembang (West Java) and North Sumatra yielded A. alternata isolates. Both Alternaria species could presence in the same field and even on the same leaf, and therefore the necrotic spots caused by A. solani can be confused with those caused by A. alternata (Tymon 2014;Odilbekov et al. 2016). Generally, lesions caused by A. alternata are smaller, darker, sunken and can be numerous in number compared to those caused by A. solani (Tymon 2014). Our results, where surveys were mostly done on tomato fields, showed that A. solani was more commonly detected than A. alternata. On the contrary, Tymon (2014) detected A. alternaria more often than A. solani in the potato fields and concluded that the former species is the dominant causal agent of early blight potato in the Pacific Northwest of USA.
A. alternata was also identified among early blighted potato foliage in Germany, the Netherlands and Poland (Latorse et al. 2010).
Both species showed similar cultural morphology on V8 juice agar medium, i.e. grey olive to black colonies with cottony and velvety mycelia, but A. solani could clearly be distinguished from A. alternata by the formation of large, dark, obclavate and muriform conidia with beaks (Figure 2d and e) which were born singly on simple conidiophores. On the contrary, A. alternata produced beakless, small conidia which were born in chains on conidiophores and readily sporulated in culture in complete darkness without special treatment.
Recently, molecular identification of the causal agent of early blight disease has been possible through the use of species-specific primers (Edin 2012;Chowdappa et al. 2014). These primers allowed large field assays for rapid and sensitive detection of A. solani simply by PCR amplification of DNA samples from early blighted leaves without the need to isolate the fungus. Three types of A. solani culture on V8 juice agar medium were observed: abundant sporulating type which was characterized by compact mycelia with distinct radial zonation in the medium (Figure 2a), non-sporulating type which produced white, sterile and massive aerial mycelia (Figure 2b), and intermediate type (Figure 2c). Mixed culture types which appeared as growth sectoring could presence in the same Petri dish. Only the sporulating type section was transferred to a fresh medium when subculturing it because sporulating ability may be lost during repeated subcultures (Bonde 1929).
We observed cultural variability among monospore cultures of parental isolates (isolate 16, 17, and 19) as was reported by Bonde (1929) and later by Henning and Alexander (1959). In spite of that, AFLP analysis revealed identical banding patterns among descendants of each of those parental isolates (Figure 3), indicating that such anonymous genome portions were not informative for differentiating cultural variation. A similar result was reported by Rotondo et al. (2012) among single-spore isolates of small-spores Alternaria from apple, where morphological variation was not in accordance with AFLP groupings.

Aggressiveness Test
Necrotic spots with yellow halo on leaves were started to visible on 3 DPI. Necrotic spots incited by aggressive isolates on susceptible tomato genotypes enlarged and coalesced (Figure 4c) which often turned leaves into dark color leading to premature defoliation (Fig 4d), Indonesian isolates were more aggressive than that of the US isolate in causing both disease syndromes (Table 2 and 3). The US isolate was even unable to incite stem lesion (Table 3). Although aggressiveness tests were done on representative isolates, the results showed that local isolates were more aggressive than that of the imported one. Repeated subculturing of the US isolate may account for its reduced aggressiveness as was encountered by Weber and Halterman (2012) in their experiments for A. solani reference isolate. Contrary to Gannibal et al. (2014) finding that isolates from tomato is more specialized and associated with tomato, our results showed that isolate 10 which was isolated from potato leaves in Sukabumi (West Java) was the most aggressive in causing tomato early blight and stem lesion.

Low to varying degrees of aggressiveness among
A. solani were reported by several researchers. For example, Shahbazi et al. (2010) observed varying    Table 1. Different susceptibilities to A. solani isolate were observed among tomato genotypes (P<0.05). The resistance degree of the tomato genotypes to early blight was consistent with the claimed resistance level, except for HRC90.145 which turned out to be susceptible to Indonesian isolates. HRC90.145 was previously rated as moderately resistant to A. solani (Poysa and Tu 1996). This implies that introduced breeding materials must be tested to local A. solani isolates to obtain an effective resistance in Indonesia. The resistance to stem lesion phase among the tomato genotypes tested was in accordance to the resistance to early blight. Previous classical genetic and genetic mapping studies found that resistance of tomato to both disease syndromes may be found in the same source, but may also segregated independently in other resistance sources (Chaerani and Voorrips 2006;Chaerani et al. 2007).
Varying levels of aggressiveness among A. solani isolates even among isolates originating from different germ tube tips from the same conidium were repeatedly reported (Bonde 1929;Neergaard 1945;Stall 1958), but consistent host-specific reactions of isolates which indicates the presence of physiological races could not be demonstrated (Castro et al. 2000). It is hardly possible to identify tomato and potato germplasm with race-specific resistance since results of genetic studies on resistance to A. solani derived from different donors point to the same conclusion that resistance to the pathogen is expressed quantitatively that is controlled by minor genes (Chaerani et al. 2007;Odilbekov 2015;Adhikari et al. 2017).

Isolate (host and origin)
Tomato genotypes  transformed before statistical analysis. transformation before statistical analysis.

AFLP Polymorphism and Phylogenetic Analysis
By using two and one selective bases for EcoRI and MseI primers, respectively, a total of 40 fragments were obtained. Nine polymorphic fragments in the size range of 165-325 bp were scored between A. solani and A. alternata (Figure 3), but only 5 polymorphic fragments were detected within A. solani isolates. Theoretically, the use of a two-nucleotide extension for the EcoRI primer along with a single nucleotide extension with the MseI primer should generate more fragments and potentially more polymorphic markers (Kothera et al. 2003 A. solani were clearly separated from A. alternata at the similarity level of about 0.65 ( Figure 5). Isolate subdivision with respect to temperate and tropical origin was observed at 0.71 coefficient similarity where the US isolate formed a distinct cluster. Within the tropical isolates, no subdivision based on geographic origin was observed. Three distinct genotypes were observed at 0.91 coefficient of similarity with one branch contained only isolate 88 obtained from Purwakarta (West Java). The other two branches contained 12 and 14 isolates, each with identical AFLP banding patterns.
Studies in USA, Cuba, Brazil, India and China using isozyme and molecular markers also found no location effect or no clear differentiation of A. solani collected from various regions in those countries (Petrunak and Christ 1992;Weir et al. 1998;Martinez et al. 2004;Varma et al. 2006;Kumar et al. 2008;Lourenço et al. 2011;Leiminger et al. 2013;Meng et al. 2015). Isolates from different agroclimatic conditions were found to be more genetically similar than isolates in close proximity to one another (Kumar et al. 2008). The observed low genetic diversity in our studied population suggests clonal population structure and is consistent with the asexual reproduction mode of the pathogen. The wide distribution of such population structure could be shaped by similar host adaptation processes on varieties with narrow genetic background. Most commercial tomato cultivars in Indonesia were developed from few highyielding lines introduced from Taiwan (Chaerani 2006 Several other genetic diversity studies of fungal pathogen also found no clear association of molecular groupings with degree of virulence or aggressiveness of the pathogen (Kothera et al. 2003;Atallah et al. 2011;Rotondo et al. 2012). This indicates that such anonymous genomic regions were not suitable for fine discriminations of virulence or aggressiveness.
Host subdivision could not be addressed in our study because of the low sample number from potato. Isolates obtained from contiguous tomato and potato fields such as in Margahayu, Lembang, could possibly originate from a tomato field infected with an adapted lineage to this host. Other studies in the USA, Brazil and Cuba found population subdivision according to the host of origin as detected by RAPD and/or AFLP markers (Weir et al. 1998;Martinez et al. 2004;Lourenço et al. 2011) as well as by sequence analyses of ITS region of rDNA and the genes encoding glyceraldehyde-3phosphate dehydrogenase (gpd), and Alt a1 (Lourenço et al. 2009).
Re-identification of Alternaria species and A. solanilike from both hosts by morphological analysis, host plant association studies, and sequence analyses of the Alt a1, gpd and/or calmodulin genes confirmed that early blight of tomato and potato are caused by distinct Alternaria species (Simmons 2000;Rodrigues et al. 2010b;Latorse et al. 2010;Gannibal et al. 2014).
Currently, there are four other Alternaria species beside A. solani that are associated with early blight disease: A. alternata was identified from potato in Germany, the Netherlands, Poland (Gannibal et al. 2014), and USA (Tymon 2014); A. tomatophila was identified from tomato in Russia (Gannibal et al. 2014) and Brazil (Rodrigues et al. 2010b); and A. cretica and A. grandis were known to cause early blight on tomato and potato, respectively, in Brazil (Rodrigues et al. 2010b). Our surveys found two isolates of A. alternata from early blighted tomato leaves, one of them was presence in the same area with A. solani. A thorough sampling and identification of early blight pathogen from tomato and potato in Indonesia could lead to a better understanding whether the disease is indeed caused by more than one Alternaria species. Considering the almost continuous cropping systems, geographic proximity of tomato and potato fields in many producing areas, and polycyclic infection throughout the year in the absence of harsh environmental conditions in Indonesia, it is interesting to study the population dynamics of A. solani and changes in its epidemiology. Since host plant assays were performed on few isolates, tests on a wider A. solani collection would provide a better understanding on the aggressiveness spectrum of the pathogen. The use of one primer combination was sufficient to separate the tropical isolates from that of the temperate isolate and to discriminate aggressiveness level among selected isolates, but was powerless in discriminating intra-variation among local isolates. Genetic analysis with more AFLP primer combinations and the use of SSR markers (Meng et al. 2015), would reveal more genetic variation. The information obtained from this study should be useful for selection of appropriate type of A. solani resistance and resistance genes in local tomato breeding programs.

CONCLUSION
Twenty two A. solani isolates were successfully recovered from early blighted tomato or potato leaves originated from nine localities in Central and West Java. Two isolates of A. alternata were also identified from tomato fields in West Java and North Sumatra. Host plant assays of selected A. solani isolates showed that local isolates were more aggressive than that of the temperate isolate from USA in inciting both early blight and stem lesion symptoms. AFLP analysis with EcoRI+AG and MseI+C primer combination separated the US isolate from local and Taiwan isolates, which partly coincided with low aggressiveness of the former isolate. Overall, low genetic diversity among local and tropical A. solani isolates were observed, which suggests clonal population structure with wide distribution.

ACKNOWLEDGMENTS
We are grateful to Dr. Sjaak van Heusden for advice in DNA isolation; and Aart Hermsen and Dirk Geurtsen for plant care in the glasshouse. This research was a part of the first author's PhD studies financially supported by the Royal Netherlands Academy of Arts and Sciences (KNAW) in the framework of the Scientific Programme Indonesia-Netherlands.