| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dept. of Genetics and Biotechnology, the Danish Inst. of Agric. Sciences, Research Centre Flakkebjerg, DK-4200 Slagelse, Denmark. A.M.D. Jensen's present address: The Danish Centre for Forest, Landscape and Planning, The Royal Veterinary and Agric. Univ., Rolighedsvej 23, DK-1958 Frederiksberg C
b Dep. of Plant Biology, Section for Plant Pathology, The Royal Veterinary and Agric. Univ., Thorvaldsensvej 40, DK-1871 Frederiksberg C., Denmark
c DLF-Trifolium A/S, Post-box 19, Højerupvej 31, DK-4660 St. Heddinge, Denmark
* Corresponding author (amdj{at}kvl.dk)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The benefits conferred to grasses by their fungal symbiont are well documented and may include insect deterrence and drought resistance (Richardson et al., 1993; West et al., 1993; Carrow, 1996; Bush et al., 1997; Morse et al., 2002). However, there may be disadvantages when endophyte-infected grasses are used for fodder, as the occurrence of F. arundinacea endophytes may cause serious problems to livestock, such as fescue toxicosis due to the production of toxic alkaloids (Paterson et al., 1995; Burke et al., 2002; Gadberry et al., 2003). The main toxin responsible for this disorder is thought to be ergovaline (Porter, 1995; Parish et al., 2003).
Epichloë and Neotyphodium endophytes show considerable variation in alkaloid production and are used for improving turf and fodder grass quality (Christensen et al., 1993; Leuchtmann et al., 2000). Nontoxic endophyte-infected F. arundinacea cultivars can be created by artificial inoculation of selected endophytes (Agee and Hill, 1994; Koga et al., 1997; Bouton et al., 2002). F. arundinacea containing endophytes with no adverse effects on animals have been commercialized in recent years (Johnson et al., 1985; Latch et al., 1985; Ahmad et al., 1986; Prestidge et al., 1994; Popay and Wyatt, 1995; Popay et al., 1995; Latch et al., 2000).
In a national research program for the development of endophyte-infected grasses with improved qualities, the first challenge was to investigate endophyte diversity and occurrence to evaluate the variability in sampling. Variation among endophyte-infected grasses for the production of ergovaline as well as microsatellite length polymorphism of endophytes can be used for assessing this diversity (Bony et al., 2001; van Zijll de Jong et al., 2003, Moon et al., 2004). Our hypothesis was that the greatest genetic variation of the endophytes is obtained by sampling a few infected host plants at many locations in a geographically large area as endophytes within a specific location may be clonally related.
The objective of this paper was to determine endophyte infection frequency, ergovaline production, and the genetic diversity of Neotyphodium and Epichlöe endophytes in Festuca spp. sample collections from locations in Italy, Spain, and Denmark. Using microsatellite markers and ergovaline production data in combination with assessment of endophyte infection frequency, the variation within the collected germplasm was evaluated.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Plant Material
In Italy and Denmark, small plants with intact roots were collected. Only seeds were collected in Spain as collection was undertaken during a very dry period, and very few healthy green plants were available. Within 6 d of collection, these were planted in pots (prefabricated sphagnum mix [Pindstrup substrate no. 2 a]) and watered frequently. In Spain, one seed head per plant was collected at each location from approximately 16 to 20 plants. Seeds were sown in water-saturated vermiculite immediately on return to the lab. One plant from each seed head was kept in the greenhouse (temperature 
15°C) for 3 mo under natural light conditions. Supplementary lighting of 100 W m–2 was given from high-pressure sodium lamps (L. UCALOX, Type Lu 400/XO/T/40) from 0800 to 1600 h. Three months after planting, plants were examined for endophyte infection.
Detection of Endophytes
Neotyphodium and Epichlöe infection was diagnosed by examining two tillers of each plant using the immunoblot method according to Gwinn et al. (1998). All fungal endophytes in infected plants were identified as Neotyphodium/Epichlöe on the basis of the specificity of the antibody. Plants that were positive in the tissue print immunoblot test were inspected for the presence of intercellular mycelium by microscopic examination (Latch and Christensen, 1982). Three tillers from each plant were investigated, as described by Latch et al. (1984).
Flow Cytometry
High-resolution DNA analyses were performed on plants of F. pratensis and F. rubra to confirm the species using the Partec CyStain UV precise P no. 05–5002 Kit. Approximately 0.5 cm2 fresh leaf material was chopped into 0.5-mm size with a razor blade in extraction buffer (Partec) filtered through a 50-µm Partec CellTrics no. 04-0041-2317, and stained according to the manufacturer's description. The chromosome number in the nuclei was analyzed using the flowcytometer Partec PA 1 (Partec GmbH, Münster, Germany). DLF-Trifolium varieties were used as internal standards. Due to clear morphological characteristics this analysis was not performed on F. arundinacea.
Material for Ergovaline Analysis
Endophyte-infected plants were screened for the production of ergovaline. For F. arundinacea 29, 78, and 29 plants from Italy, Spain, and Denmark, respectively, were examined. Additionally, 51 endophyte-infected F. rubra plants, 50 from Italy and 1 from Spain, and 31 F. pratensis plants from Italy were analyzed. Due to unintended loss of endophyte-infected plants, a few locations were not represented.
Endophyte-infected plants were cloned to a standard size (5 shoots). Two months after division, two clones of each plant genotype were transferred to a climate chamber for 4 wk where they were grown for 2 wk with a light–dark cycles at 23 and 18°C, respectively, followed by 2 wk at 25 and 20°C, respectively. The twilight period was four steps over a 4-h period where light intensity at plant level accelerated to approximately 900 µmol m–2 s–1. For the following 8 h, light intensity was kept at approximately 900 µmol m–2 s–1. Then light intensity was reduced to darkness over a 4-h period. (high pressure Hg lamp–Osram HQI/BT, 400 W).
Plants were placed randomly in the climate chamber, watered regularly, and never allowed to dry out. After 4 wk in the climate chamber, vegetative plants were harvested in a top and base fraction. The base fraction was the first 6 cm above soil level. Material from the two clones of each plant genotype was pooled before analysis. Samples were frozen, freeze-dried, and milled. The base fraction was analyzed for ergovaline. The screening of endophyte-infected Festuca spp. was a part of a larger screening program including endophyte-infected Lolium perenne L. The temperature regime was chosen as a compromise to obtain lolitrem B and ergovaline production (Ball et al., 1995). The primary aim of this investigation was to test for ergovaline production. Estimation of ergovaline quantity is dependent on plant size, mycelium content, and plant genotype in combination.
Ergovaline Analysis by High Performance Liquid Chromatography (HPLC)
Ergovaline was extracted, analyzed and quantified by high performance liquid chromatography (HPLC) with fluorescence detection (Spiering et al., 2002; Panaccione et al., 2003, modified). Samples of 50 mg in 2-mL screw-capped polypropylene vials were extracted with 1 mL of isopropanol/water/lactic acid (50:49:1) containing 1.11 µg mL–1 of ergotamine tartrate as internal standard. Samples were agitated in a Fast-Prep (FP120 BIO 120 Savant) at speed 5 for 25 s, gently mixed for 1 h on a rotary mixer (low speed), and pelleted by centrifugation for 10 min at 14000 x g. Fifteen µl of the supernatant was subjected to a gradient separation at 40°C on a column of C18 (Waters symmetry C18 cartridge 5 µm 4.6 x 150 mm). Mobil phase A was acetonitrile/aqueous 0.1 M ammonium acetate, 1:3 by volume, and mobile phase B was acetonitrile/aqueous 0.1 M ammonium acetate, 3:1 by volume. The start conditions were 95% A–5% B, a linear gradient to 80% A–20% B at 20 min, a linear gradient to 50% of each A and B to 35 min, linear gradient to 30% A–70% B at 40 min, a linear gradient to 100% B at 45 min and finally a linear gradient to 95% A–5% B at 50 min with a flow rate of 1 mL min–1. Fluorescence detection was performed with excitation at 310 nm and emission at 410 nm. Each plant sample was measured in duplicate. Elution time for ergovaline and ergovalinine was approx. 17 and 33 min., respectively. Ergovaline was quantified using peak area comparison to that of the ergotamine internal standard. The amount of ergovaline is expressed as ppm (µg g–1 d wt).
Isolation of Genomic DNA
Endophyte DNA was isolated from plant tillers. Three or four tillers from each plant were harvested, and wilted leaf tissue was removed. The basal 2 to 3 cm of tiller material was crushed in a leaf press (MERKU, Germany). The resulting plant sap was dissolved in 400 µL extraction buffer (200 mM Tris HCl pH 7.5, 250 mM NaCl, 25 mM NaEDTA, 1% SDS) and incubated for 15 min. Then 400 µL 5 M potassium acetate was added and the samples were vortexed and incubated on ice for 30 min. The samples were centrifuged 10 min (13000 rpm). A 600-µL volume of each sample was transferred to a 1.5-mL centrifuge tube, and DNA was precipitated by adding 600 µL ice-cold isopropanol. The samples were vortexed and incubated on ice for 30 min followed by centrifugation 10 min at 13 000 rpm. DNA pellets were washed in 300 µL 70% ethanol. The air-dried pellets were resuspended in 1.25 mL distilled water.
Microsatellite Primers and Polymerase Chain Reaction (PCR) Amplification
A total of 14 endophyte microsatellite loci were examined in this study. The microsatellites were B1, B4, B9, B10, B11 published by Moon et al. (1999) and NCESTA1AG05, NCESTA1AH01, NCESTA1DH04, NCESTA1FH03, NCESTA1IC04, NCESTA1GA07, NCESTA1AG12, NCESTA1FC04, and NCESTA1DB06 published by van Zijll de Jong et al. (2003). Polymerase chain reaction (PCR) amplifications were performed in 25-µL volumes according to Moon et al. (1999) with the exception that the primers were fluorescently labeled with IRD-800 (LI-COR, Lincoln, NE). Reference strains from F. arundinacea included N. coenophialum isolate Tf27 and e19 (= ATCC 90664), Tf13 and Tf15 (both FaTG-2), and Tf18 (FaTG-3). Reference strains N. uncinatum isolate Fp2 from F. pratensis and E. festucae isolate E90 from F. rubra ssp. rubra were included together with Epichloë typhina (Pers.:Fr.) Tul. isolate E8 and Neotyphodium lolii Latch, Christensen & Samuels isolate AR1 from Lolium perenne (Schardl et al., 1991; Christensen et al., 1993; Moon et al., 1999). All microsatellites were tested on noninfected F. arundinacea DNA as well.
Electrophoresis of Microsatellite Markers
Gel electrophoresis and pattern visualization were performed using a LI-COR model 4000 automated fluorescent DNA sequencer (Middendorf et al., 1992) (LI-COR, Lincoln, NE). Gel dimensions were 25 cm long and 0.25 mm thick. The gel contained 7 M urea and 7.0% SequaGel XR concentrate (National Diagnostics, Atlanta, Georgia). The running buffer was 0.4 x TBE (dilution of 10 x TBE: 0.9 M Tris, 0.9 M boric acid, 0.2 M EDTA). The gel was run at 2000 V constant voltage, and the gel temperature was maintained at 50°C. A 64 well comb was used for lane formation, and 0.6 µL of each sample was loaded. A size ladder was produced by mixing PCR products originating from amplifications of a known DNA sequence of L. perenne. The size ladder included 42, 44, 125, 126, 150, 151, 193, 251, 280, 327, 328, 414, and 551 base pair fragments and was loaded in lanes 1, 32, and 64. In addition, a positive control of an endophyte isolated from a DLF-Trifolium F. arundinacea breeding line was run on each gel as an internal standard.
Microsatellite Data Analysis
The sizes of amplified fragments containing microsatellite alleles were determined by eye relative to the ladder and the positive standard on all gels. Null alleles were regarded as missing values. A genetic dissimilarity was calculated as described by Diwan and Cregan (1997) for each pair of the endophyte-infected Festuca spp. and seven reference isolates from Festuca spp., to obtain a dissimilarity matrix. Similarity between two endophytes within one locus was calculated as the number of common alleles relative to total alleles in the locus observed for the two endophytes. Similarity for two endophytes is the average similarity over all loci and the coefficient of dissimilarity between them is, therefore, 1 – similarity. Two isolates from L. perenne were included as outgroups. A dendrogram was created from the dissimilarity matrix using the unweighted pair-group method with arithmetic average (UPGMA) with the aid of Proc Cluster version 8.2 (SAS Institute, 2001). The matrix of dissimilarities was subjected to an analysis of molecular variance (AMOVA) to estimate variance components within and between locations with the aid of the software package "Arlequin" (Excoffier et al., 1992; Schneider et al., 2000).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Ergovaline Production
Ability to produce ergovaline was investigated in all endophyte-infected F. arundinacea collected in Italy, Spain and Denmark and all infected F. rubra and F. pratensis collected in Italy. Ergovaline analysis was performed for a total of 198 endophyte-infected plants representing 25 ecotype collections (Table 2). The main aim of the experiment was to register the presence or absence of ergovaline in plants. Therefore plant material of the same genotype was pooled before analysis. Quantitative data are not objective because the content of ergovaline depends on plant size, mycelium content, and plant genotype in combination. Because the mycelium was not quantified no statistic analysis was performed on the quantitative data.
|
A remarkable observation is that the Italian and Danish infected F. arundinacea plants were able to produce ergovaline in relatively high quantities after 1 mo in the climate chamber, whereas the infected Spanish plants generally produced much lower amounts of ergovaline under the same conditions. None of 31 infected F. pratensis plants produced ergovaline and only three of the 50 infected F. rubra plants produced detectable quantities of ergovaline.
Microsatellite Polymorphism
A total of 129 amplification products were detected with 14 primer pairs resulting in 79 unique amplification patterns among the endophytes from 121 F. arundinacea, 32 F. pratensis, and 29 F. rubra plants, the seven reference isolates and two outgroup endophytes examined. The number of alleles amplified by each primer pair ranged from 2 to 20, with the microsatellite loci B10 and B11 as the most informative (with 20 alleles each). On average 9.2 alleles were detected for each primer pair. The genetic dissimilarity between all pairs of endophytes ranged from 0.024 to 1.000 with an average of 0.582.
An endophyte from an individual plant is referred to in this article as an isolate. It should be noted that endophyte DNA used for microsatellite analysis was not extracted from isolated cultures but rather from the endophyte-infected tillers of a single plant and therefore contained plant DNA. However, none of the primers amplified plant DNA. Only reference strain DNA originated from pure isolate culture.
It is possible that individual plants collected within a single location may be sibs, and therefore possess the same endophyte isolate, thus only 73 F. arundinacea, 19 F. pratensis, and 11 F. rubra endophyte isolates, having unique banding patterns within each collection site, were included in the cluster analysis (Fig. 1 ).
|
All F. arundinacea endophyte isolates of Group 1 were distinct in that three alleles were detected in at least one microsatellite locus. Furthermore, the isolates divided into subgroups 1a, 1b, and 1c. Subgroup 1a consisted of the F. arundinacea endophytes from all Danish locations, whereas Subgroup 1c consisted of all isolates collected in Spain and Italy. Subgroup 1b consisted of the N. coenophialum reference strains, for which three alleles were detected in three loci. In contrast, the FaTG-2 reference strains (Tf13 and Tf15) of Group 2 had only one or two alleles in all loci. The single FaTG-3 reference strain, Tf18, was highly distinct from the rest of the F. arundinacea isolates, and it also had only one or two alleles in all loci.
All F. pratensis endophytes and the N. uncinatum reference isolate of Group 3 had two alleles in at least 3 loci. All the F. rubra endophyte isolates and the E. festucae reference had only a single allele in at least eight loci and may have a null allele in some of the loci although these were regarded as missing values in the analysis. The N. lolii and E. typhina outgroup control from L. perenne had one allele in all loci except two alleles in locus B4.
The F. arundinacea/N. coenophialum group (Group 1) share 63 to 96% similarity within subgroups and 49% similarity between subgroups. The FaTG-2 reference strains (Group 2) share 34% similarity to Group 1 and only 16% similarity to the FaTG-3 reference isolate Tf18 (Group 5) from F. arundinacea. The F. pratensis isolates (Group 3) are 40 to 66% similar to each other and the N. uncinatum reference strain Fp2. The Italian F. rubra endophytes (Group 6) share 34% similarity, but only 12% similarity to the Spanish F. rubra endophyte and the E. festucae reference strain E90 from F. rubra ssp. rubra (Group 4). All the F. rubra plants included in this investigation were shown to have a chromosome number of 2n = 42 using flow cytometry. The L. perenne endophyte isolates in Group 7 only share 8% similarity to the Festuca endophytes.
Analysis of Molecular Variance
Analysis of molecular variance was performed on the dissimilarity matrix of all the collected F. arundinacea endophytes to estimate the variance components among countries, among locations, and within locations.
The AMOVA analysis performed to quantify the variation within countries compared to the variation between countries demonstrated a larger variation between countries, 61% compared to 39% within countries (Table 3). However, the AMOVA performed on the data of the Spanish and Italian F. arundinacea isolates demonstrated a larger variation within countries, 69% compared to 31% among countries (Table 4).
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the investigation by Romani et al. (2002) 93% of the F. arundinacea accessions collected in Italy were infected. In our investigation 57% were infected (Table 1). However, the collection by Romani et al. (2002) was done not only in the Po valley but also on Sardinia and in Liguria, thus only five of 15 F. arundinacea accessions originated from northern Italy. Our investigation from northern Italy may provide a better estimate of the actual distribution of endophytes in that particular area.
In northern Italy Romani et al. (2002) found the mean infection in F. arundinacea populations to be 72 and 75% in F. rubra. Our investigation in Italy displays a mean infection rate of 35% in F. arundinacea covering rates from 29 to 46%. In F. rubra the infection rate was between 30 and 73 with a mean of 46%. Local geographical differences in the two surveys might account for the variation in infection rate in combination with the number of plants investigated.
In Denmark, all populations showed high levels of infection, with a mean infection rate of 70%. However, only four populations were investigated, all located in Stevns (Table 1). Extending the investigation to include additional populations located throughout Denmark would provide an improved estimate of the actual distribution in Denmark. F. arundinacea endophytes have been reported from other northern European countries (White and Baldwin, 1992; Saikkonen et al., 2000); however, this study presents the first report of Neotyphodium endophytes in F. arundinacea on Danish seminatural grasslands.
Ergovaline Production
All endophyte-infected F. arundinacea plants contained ergovaline. However, a considerable variation in ergovaline levels was measured compared to the endophyte-infected F. rubra and F. pratensis populations which did not produce any ergovaline (Table 2). The different metabolite levels observed after growth under our climate-chamber conditions in the pooled samples might be related to differences in endophyte mycelium content among the different plants, difference in plant size, the fungal strain/plant genotype combination, or differences in the ergovaline biosynthetic pathway (Hiatt and Hill, 1997; Bony et al., 2001). Thus the ergovaline level should be considered as indicative. The plant material was always the same age when the experiment started; however, the mycelium content was not estimated.
It was quite surprising that Danish isolates were able to produce high levels of ergovaline under warm climate chamber conditions, as to our knowledge no cases of fescue toxicosis have been reported in Denmark. However, naturally occurring F. arundinacea is not very widespread for grazing purposes in Denmark, and the Danish climate apparently does not favor ergovaline production. Isolates from Italy were collected in cold mountain areas, and similar to the Danish isolates, exposure to warm climate chamber conditions caused a high production of ergovaline. In contrast, the Spanish isolates, in general, produced only low levels of ergovaline in the warm climate chamber. It is possible that our climate chamber conditions were not sufficient and/or optimized to favor ergovaline production in the Spanish isolates. In their native habitats, the plants usually exist under higher temperatures in contrast to the Danish ones, especially during summer, and may be able to regulate ergovaline production differently. An alternative explanation may be that the Spanish endophyte isolates simply are not able to produce high levels of ergovaline (Vazques de Aldana et al., 2003). The Danish and Italian isolates may be adapted to relatively cooler environments, and the exposure to a heat period in the climate chamber may stress the endophytes and their hosts, which may possibly influence ergovaline production. The tendency of F. rubra to not produce ergovaline is in agreement with Leuchtmann et al. (2000) as they demonstrated that stroma-forming endophytes often are free of alkaloids.
Microsatellite Polymorphism and Cluster Analysis
It has previously been demonstrated that microsatellite-based PCR is useful for fingerprinting and identifying variation among Epichloë and Neotyphodium endophytes in grasses (Groppe et al., 1995; Moon et al., 1999). These methods are powerful as they can be used directly on total DNA extracted from endophyte-infected host plants, without culturing the endophyte from the grass host, a process that can be very time-consuming.
Neotyphodium endophytes of F. arundinacea have been divided into three closely related distinct taxonomic groups, N. coenophialum (Morgan-Jones and Gams), FaTG-2 and FaTG-3, respectively, based on isozyme and DNA sequence analysis (Christensen et al., 1993; Tsai et al., 1994). These groups consist of interspecific hybrids originating from different Epichloë species (Tsai et al., 1994; Moon et al., 2004). N. coenophialum is a three-ancestor hybrid originating from an E. festucae-, E. baconii-, and E. typhina-like ancestors. The FaTG-2 is a hybrid of E. festucae- and Epichloë baconii White-like ancestry, and FaTG-3 is a hybrid of E. baconii- and E. typhina-like ancestry (Moon et al., 2004). The F. arundinacea isolates examined in this study are triallelic in one or more of the microsatellite loci examined and the F. arundinacea triallelic endophytes group closely to N. coenophialum reference strains (Fig. 1, Group 1) sharing at least 49% similarity. No diallelic F. arundinacea endophytes were found in our samples although the FaTG-2 reference strains (Fig. 1, Group 2) are of Spanish origin.
Moreover, within the F. arundinacea group the Danish isolates clearly group separately from the Italian and Spanish isolates. This kind of geographical specialization into "European geographical subgroups" has been demonstrated within other endophytic fungal species (Sieber et al., 1991). A thorough investigation of endophyte-infected F. arundinacea from more European countries will reveal whether the European isolates are separated into different genetic groups.
F. pratensis is predominantly inhabited by N. uncinatum and much rarer by N. siegelii (Craven et al., 2001). Both are interspecific hybrids of E. bromicola-like ancestry together with E. typhina- or E. festucae-like ancestries, respectively (Craven et al., 2001).
The F. pratensis isolates collected from Italy have a maximum of two alleles. It seems like the F. pratensis endophytes divided into two subgroups of which one was more similar to N. uncinatum. Unfortunately N. siegelii reference strains were not included; however, by looking at the published allele size (Moon et al., 2004) none of our F. pratensis endophytes resembled N. siegelii, leading to the conclusion that the F. pratensis endophytes in our study most likely were N. uncinatum.
E. festucae inhabits different species of fine fescue including F. rubra (Leuchtmann et al., 1994). F. rubra is a complex of the three subspecies spp. rubra, commutata Gaudin, and tricophylla Gaudin (Huff and Palazzo, 1998). In this investigation the F. rubra endophyte isolates (Fig. 1, Group 4 and 6) were not very similar to the E. festucae reference strain from F. rubra ssp. rubra. Therefore the chromosomal content of the plants was tested and the results demonstrated that all F. rubra plants had 2n = 42 chromosomes as also F. rubra ssp. commutata and ssp. tricophylla. F. rubra ssp. rubra had 2n = 56 chromosomes (Huff and Palazzo, 1998). The observed dissimilarity in endophyte isolates may be due to subspecies differences of the inhabiting endophyte. However, this is not likely as E. festucae are able to produce ascospores for potential infection of other grass species/subspecies and studies of allozyme variation in 12 loci and 13 AFLP markers showed that E. festucae isolates from different Festuca species or subspecies of the fine fescue types were quite similar (Leuchtmann et al., 1994; Tredway et al., 1999).
Evaluation of Sample Diversity
The geographical region visited in this investigation in southern Europe was chosen on the basis of previous surveys demonstrating the occurrence of endophytes in these areas (Oliveira and Castro, 1997; Zabalgogeazcoa et al., 1999; Romani et al., 2002). Examination of genetic diversity, toxin production, and endophyte infection frequency provided insight for discovering high levels of genetic variation in natural endophyte populations.
The microsatellite analysis demonstrated a large genetic variation between the endophyte isolates, which is also supported by variation in the alkaloid profiles. From the cluster analysis (Fig. 1) it was clear that the largest genetic variation was between the different species of Festuca endophytes. The key question was where and how to sample to find the wide genetic variation within a species.
A rather large genetic difference was observed within the F. pratensis and F. rubra endophytes sampled in northern Italy (Fig. 1). For the F. pratensis endophytes a genetic shift seemed to occur from the first 5 locations (MFI1-MFI5) to the 3 final locations (MFI6-MFI8), whereas the F. rubra endophytes isolates (Fig. 1, RFI14) even from the same location were genetically very different.
The analysis of the Italian and Spanish F. arundinacea samples revealed surprisingly that the genetic variation of the endophytes within a country was much larger than between the countries as shown in the cluster analysis (Fig. 1, Subgroup 1c) and the AMOVA (Table 4). The variation within the single locations in Spain was, in addition, much higher than the variation between the locations (Table 5). As more than 50% of the Italian and Spanish locations of F. arundinacea were infected it seemed to be more important to sample many plants within few distant locations than it was to sample few plants at each of many locations within an area to obtain as much genetic variation as possible. This was also the case for the F. arundinacea endophytes sampled in Denmark, which were clearly genetically different from the southern European isolates and which contained greater genetic variation within rather than between locations. More surveys need to be performed at more geographically distant locations. Therefore a general recommendation for sampling of genetically different endophytes is difficult but from this study it was clear that for the endophyte species included in this study a larger genetic difference exists within each location than would have been expected from the clonally behavior of especially the asexual seed-disseminated endophytes. Thus it seems not to be the case that a single genotype of an endophyte is overtaking an entire field so large samples from each collection site will therefore be an advantage. On the other hand having collections of endophytes from very different climates as the case with the Danish F. arundinacea endophytes compared to the collections from Spain and Italy seems to be important to find new genetic variation.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our hypothesis that the greatest genetic variation of the endophytes was obtained by sampling a few infected host plants at each of many locations in a geographically large area was not supported by this investigation as apparently major genetic differences can be found in a single location, indicating that larger samples at distant sampling sites may be more beneficial for obtaining genetic variation.
Examination of the ecotype collections of F. arundinacea, F. pratensis, and F. rubra, in relation to ergovaline production revealed both highly toxic isolates, and apparently low ergovaline producing isolates. The genetic background is known to play a very important role in the level of endophyte alkaloids (Bouton et al., 2002) and the next step to creating improved pasture grasses will be to find nontoxic isolates and artificially inoculate these low-ergovaline producing isolates into different Festuca spp. cultivars and examine whether toxin production is similarly low, and also determine whether these novel host-endophyte combinations are stable.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 4, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
