- Top of page
- Materials and Methods
- Supporting Information
Arbuscular mycorrhizal fungi (AMF) belong to an ancient fungal phylum, the Glomeromycota. They form obligate symbioses with the majority of land plants (Smith & Read, 1997; Schüssler et al., 2001). The early colonization of land by plants was probably associated with the presence of AMF (Redecker et al., 2000). Nutrient exchange and protection from pathogens are thought to be key elements in the symbiosis (Newsham et al., 1995; Smith & Read, 1997). AMF community composition can determine plant biodiversity and productivity (Van der Heijden et al., 1998). The host range of AMF is thought to be very wide, as most AMF associate with a large number of plant species (Smith & Read, 1997). Morphological characterizations of AMF spores have distinguished < 200 species, although molecular studies suggest that a significantly larger number of species exist (Redecker, 2002). Difficulties in identification and cultivation, coupled with difficulties in obtaining uncontaminated DNA from AMF field samples (Redecker et al., 1999; Corradi et al., 2004), have greatly hindered studies of their ecology and genetic diversity in natural ecosystems.
Although there are many studies focusing on AMF species diversity, only a few studies have used molecular techniques to look at genetic diversity within AMF populations. However, knowledge of such diversity is relevant as (Koch et al., 2006) demonstrated a genetic basis for differential effects on plant growth among isolates of the AMF Glomus intraradices originating from a single field. In glasshouse experiments, the biomass of the host plants Brachypodium pinnatum and Prunella vulgaris varied by up to 33% (Koch et al., 2006). This suggests that, depending on the effect on the plant growth, a given AMF individual in a population could be favoured by certain plant species. Phosphate uptake by the host plant has also been shown to be significantly affected by different isolates of the same AMF species (Munkvold et al., 2004), although within-population variation was not considered. Finally, as field studies identifying AMF associated with particular plant species have only focused on resolving AMF species identities and not intraspecific variations (Helgason et al., 1998; Del Val et al., 1999), a reliable identification of genotypes within and among field populations would allow the study of associations of particular AMF genotypes and plant species.
Several different approaches using molecular markers and different sampling designs have been used to study genetic diversity and structure in AMF populations. Stukenbrock & Rosendahl (2005a) used a hierarchical design to study multilocus genotypes of three Glomus species. Significant genetic structure was found at a small scale, among plots separated by a few metres, whereas among neighbouring field sites, with differing agricultural treatments, no differentiation was detected. Vandenkoornhuyse et al. (2001) used inter simple sequence repeat (ISSR) fingerprints and ribosomal gene polymorphisms to study differentiation among AMF from different sewage treatments in a field. A high degree of diversity was found for two Glomus species and the observed diversity was structured among field plots of different treatments. Using in vitro propagated G. intraradices from a field population, Koch et al. (2004) found high genetic diversity and differentiation among field plots. In the same population, Corradi et al. (2007) found polymorphism in copy numbers of ribosomal genes.
The ideal approach to studying genetic diversity in AMF populations would be to have a marker-based system that can be applied to DNA originating from single spores from field sites. One critical limitation of this approach is the quantity of available DNA for genotyping. While Stukenbrock & Rosendahl (2005b) developed multiplex PCR to reliably amplify three gene introns, the number of loci that can be genotyped is limited. Also, so far, use of multiplex PCR has been unsuccessful in genotyping some AMF spores. Furthermore, the development of reliable markers in intergenic regions remains a major challenge, because of the lack of conserved flanking sequences among related species. Conserved ribosomal gene sequences have been successfully used in resolving deep phylogenetic relationships in AMF (Schüssler et al., 2001) but would be insensitive to within-population variation. Some rapidly evolving regions of ribosomal DNA sequences (e.g. internal transcribed spacer (ITS)) exhibit within-spore variability (Sanders et al., 1995), and these regions are therefore unsuitable as isolate-specific markers. Amplified fragment length polymorphism (AFLP) was successfully used to genotype G. intraradices isolates from one field (Koch et al., 2004). This method allows a large number of loci to be analysed. Nevertheless, a larger quantity of DNA is required for reproducible fingerprints than that contained in an AMF spore and observed loci are difficult to check for homoplasy. Furthermore, genetic markers without sequence-specific primers (such as AFLP or ISSR) can potentially be biased by the amplification of undetected contaminating microorganisms, if material from natural soils or glasshouse experiments is used (Hijri et al., 2002).
To overcome the limitations of DNA quantity per individual and the problems of contaminating microorganisms, a reliable strategy is to obtain single-spore isolates of the fungi from a field site and then put them into an in vitro culture system. Such a culture system has been described (St.-Arnaud et al., 1996) and if initiated with a single spore allows the clonal growth of a large amount of fungal material. Although this approach is extremely labour intensive and time consuming it allows the extraction of large amounts of contaminant-free AMF DNA from each isolate. Additionally, it permits multiple DNA extractions to be made from the same individual, allowing experiments to be properly replicated to control for artifacts. True replication is almost never applied to AMF population or community studies because the amount of DNA per spore allows no replication (but see e.g. Vandenkoornhuyse et al., 2001).
Microsatellites or simple sequence repeats (SSR) are widely used for estimation of population substructures and relatedness among individuals (Queller et al., 1993; Griffiths et al., 1996; Jarne & Lagoda, 1996). SSR are composed of tandemly repeated sequence motifs from one to six nucleotides in length (Tautz & Renz, 1984). Polymorphism at these loci mostly arises through slipped-strand mispairing and subsequent errors during DNA replication (Tautz & Renz, 1984). A large number of SSR loci were found in fungal genomes (Karaoglu et al., 2005). In Neurospora, mutational patterns were studied in detail using cross-species comparisons (Dettman & Taylor, 2004). To our knowledge, only primers consisting of repeat motifs have been used to genotype AMF (Longato & Bonfante, 1997; Vandenkoornhuyse et al., 2001; Douhan & Rizzo, 2003) and suitable regions for primer design in flanking repeats have not been reported.
Genetic diversity of fungal mitochondria has rarely been studied in natural populations. Raab et al. (2005) reported the first sequences of a mitochondrial ribosomal gene in AMF. Knowledge of AMF-specific mitochondrial sequences is critical for the specific amplification of potentially polymorphic sequences in populations, as some species were shown to harbour endosymbiotic bacteria (Bianciotto et al., 1996). To our knowledge, no study has reported mitochondrial diversity in AMF populations.
Given the potential ecological importance of genetic diversity in single populations of AMF, the distribution of genetic diversity across different populations is relevant to the understanding of AMF–host plant interactions. A field population of G. intraradices in Tänikon, Switzerland, was chosen to establish what is, to our knowledge, the largest collection of in vitro cultures of one AMF population. The field site in Tänikon was divided into plots to which different agricultural treatments were applied. During the process of isolation of AMF spores from the field, different host plants were used for an initial round of cultivation. Establishment of in vitro cultures was rarely attempted for AMF, because of the time-consuming procedure required to isolate single spores for propagation under sterile laboratory conditions. Nevertheless, only in vitro cultures provide the necessary quantities of contamination-free DNA for the development of new markers in noncoding regions of the genome or the application of AFLP with appropriate levels of replication. Ten in vitro isolates were initially genotyped using AFLP, revealing strong genetic differentiation among three main genotypes, shown by an average of nearly 50% polymorphic loci among pairs of isolates (Koch et al., 2004). Phenotypic traits measured in 16 isolates revealed large variation in hyphal and spore production (Koch et al., 2004). Genetically different isolates from this field were shown to differentially affect plant growth (Koch et al., 2006), making it possible to test whether host plant preferences exist among genetically different isolates of G. intraradices. Glomus intraradices has a haploid genome (Hijri & Sanders, 2004), facilitating the development of SSR markers.
The aims of the present study were: to use nuclear SSR and mitochondrial markers to study a large collection of in vitro isolates from a single population; to test for associations of isolate genotypes with host plant species used for the establishment of in vitro cultures; and to compare identified genotypes with newly established in vitro cultures from two distant populations. In addition, G. intraradices is the first AMF species for which a whole-genome sequencing project has been initiated (Lammers et al., 2004). The project uses DNA from an isolate of Canadian origin (DAOM181602), allowing us to compare genetic differences within a population with the isolate that is currently being sequenced.
- Top of page
- Materials and Methods
- Supporting Information
We used newly developed SSR and mitochondrial markers to study genetic diversity in a field population of G. intraradices. To our knowledge, these are the first such markers to be established for AMF. In combination, the 13 loci allowed us to identify 17 distinct genotypes within one field population. A comparison with six isolates from two different locations in Switzerland revealed that very similar genotypes can be found in multiple locations. A comparison of ITS sequences of our isolates with sequences of closely related species confirmed that all isolates belong to the genetically diverse species of G. intraradices. In vitro cultivation allowed the extraction of pure fungal DNA in sufficient quantities to control locus specificity and reproducibility of the markers. The identification of multiple repeat motifs and additional indel mutations and single nucleotide polymorphisms within some loci suggests that allele length differences alone do not accurately reflect relatedness (see Supplementary Material). Nevertheless, the combination of the new markers permits a fine-scale genotyping of G. intraradices isolates. To our knowledge, this study shows for the first time the presence of polymorphism in mitochondria of a glomeromycotan species. Based on two introns of the mitochondrial ribosomal large subunit gene, we have shown that the cytoplasm of AMF is not genetically homogenous within a field population. These markers provide a basis on which to study potential associations of mitochondrial genotypes with phenotypic traits of the fungus or environmental factors.
Koch et al. (2004) used a subsample of nine in vitro isolates from the same field population and DAOM181602 to compare genetic diversity using AFLP. Four highly significant monophyletic groups of isolates were identified. This compares with seven distinct genotypes that can be identified among the same isolates based on the newly developed SSR and intron markers. There was a strong and significant correlation between genetic distances estimated by the two genotyping methods (Fig. 5). Interestingly, AFLP indicated genetic differences for pairs of isolates for which no differentiation was detected by SSR and intron markers. However, where large differences were observed between isolates using SSR markers, genetic differences were not detectable with AFLP. This is probably because at large genetic distances AFLP appear to saturate and no longer increase proportionally compared with genetic distances based on SSR and intron markers. Taken together, these results show that AFLP and SSR markers resolve fine-scale differences among isolates within populations of AMF, probably covering large sections of the genome. Nevertheless, SSR markers are advantageous because of their much lower requirements in terms of DNA quantity and the sequence specificity of the primers. Furthermore, no amplification was obtained for these loci in the most closely related species, G. diaphanum, suggesting that these markers are species-specific. Species specificity of markers is particularly important if soil samples are to be analysed that potentially contain a large diversity of AMF species.
Figure 5. Correlation of genetic distances among nine Glomus intraradices isolates from the Tänikon field population and DAOM181602 based on the percentage of different amplified fragment length polymorphism (AFLP) fragments (Koch et al., 2004) and genetic distances based on the 10 newly developed simple sequence repeats (SSR), one nuclear intron and two mitochondrial loci (number of loci that are polymorphic in pairs of isolates). Mantel test: R2 = 0.966, P = 0.0002, 10 000 permutations.
Download figure to PowerPoint
Sequence-specific markers may potentially be used to address some of the remaining questions in AMF genetics concerning heterokaryosis and ploidy. In our study, only single alleles were amplified from each locus, suggesting a haploid genome and genetic uniformity among nuclei within isolates for those markers. Nevertheless, our results do not exclude the possibility that more than one allele might occur at low frequency in an isolate at an undetectable level in some nuclei. Conversely, our markers might not exhibit sufficient polymorphism to discriminate genetically diverse nuclei or different copies of a polyploid genome. Neither of these points can be directly addressed with this data set. Hybridization experiments to determine the presence and absence and numbers of markers per nucleus and the study of loci under higher mutation rates would allow firmer conclusions to be drawn regarding these issues.
The genetic diversity estimated using the seven SSR and nuclear intron polymorphisms is significantly lower than that reported by Vandenkoornhuyse et al. (2001) for field populations of two distantly related Glomus species. Meanwhile, Stukenbrock & Rosendahl (2005a) reported a slightly lower number of genotypes for three different Glomus species. One possible reason for this discrepancy is that different species may vary considerably in their levels of genetic diversity, but another is that markers without sequence-specific primers (such as ISSR or AFLP) were applied to single spores from field samples. Identifying contaminant sequences in spores can be difficult (Hijri et al., 2002). The newly developed SSR markers are likely to be in noncoding regions, as BLAST results do not indicate significant matches in their flanking regions. Markers in noncoding regions are more likely to be selectively neutral than those in intron regions as these might be linked to genes under selection. Nevertheless, the neutrality of any noncoding region in the genome would depend on the occurrence of recombination.
Within the Tänikon field population the genetic diversity was structured among plots, meaning that genotypes were not found at random among the plots. The estimation of sampling efficiency indicates that additional field plots would probably increase the number of genotypes recovered. However, our analysis of genetic structure in the field is potentially biased by the finding that trap host plants affect the genetic diversity of recovered isolates. A further confounding factor in estimating total diversity in a field may stem from the fact that all population genetic studies need to isolate spores for genotyping. Considering the strong differences in hyphal and spore production among isolates of G. intraradices (Koch et al., 2004), a potential bias may arise from the fact that poorly sporulating isolates may be underrepresented.
The comparison of isolates from the Tänikon population with isolates from two geographically distant locations and the isolate DAOM181602 showed that the same or highly similar nuclear and mitochondrial genotypes can be found over a distance of hundreds of kilometres, and different continents in the case of isolate DAOM181602. Considering the high degree of diversity found within the Tänikon population, such a finding is surprising. The isolate DAOM181602, identified by Koch et al. (2004) as belonging to one of the monophyletic groups of the Tänikon field population, can be distinguished from all other isolates based on SSR markers, but not based on mitochondrial markers. Nevertheless, the distance in the minimum spanning network between DAOM181602 (genotype XII) and the most closely related genotypes found in the Tänikon population (II and III; Fig. 2a) does not exceed the average distance between genotypes in the network. Therefore, our sampling across different locations suggests that a significant amount of genetic diversity is already found within one field population and differentiation among field populations, if any, is rather weak. A study of intercontinental diversity of AMF found very similar sequence types in temperate and tropical habitats, also suggesting that AMF may show high local diversity and weak geographic structure (Husband et al., 2002).
Plant species have previously been shown to be associated with distinct AMF species assemblages (Vandenkoornhuyse et al., 2003; Scheublin et al., 2004), with some studies experimentally demonstrating a direct effect of the plant species on the AMF community composition (Johnson et al., 1992; Johnson et al., 2004). Effects on host plant growth have been shown for a subset of AMF genotypes in the Tänikon population (Koch et al., 2006), indicating that there could be fitness benefits to evolving specificity with a given AMF genotype. Our study showed that particular AMF genotypes were significantly more frequently isolated depending on which trap host plant species was used for cultivation. Given the interaction between plant species and AMF genotypes, it is possible that past crop cultivation in the study field site may have had an impact on the abundance of particular genotypes. A more extensive sampling of AMF spores and a broader range of host species would be needed to allow generalizations to be made from our results. As single spores were isolated and transferred from mature trap cultures to establish single-spore cultures, the likelihood of recovering a particular isolate genotype from trap cultures is probably linked to the spore production of that genotype at the time of collection or to the preference of genotypes for a given host. Differential spore production among genotypes could result either from greater plant resources being available to a given isolate or from a competition effect caused by other AMF isolates or species present in the same trap culture. On the basis of our study, it cannot be concluded by which mechanism particular genotypes were favoured for spore production. To our knowledge, this study is the first to report preferences of host plants for particular genotypes within an AMF species, a factor that could influence the co-evolution of specific interactions between AMF genotypes and plants.
Explaining patterns of genetic diversity within AMF species remains a major challenge because of the difficulties in sampling and genotyping these fungi. To infer genetic diversity patterns at the inter-populational or global scale, underlying processes responsible for the genetic diversity at the local level need to be known. A first hypothesis for the distribution of genetic diversity is an isolation-by-distance model, postulating that dispersal of AMF genotypes is restricted in space over generations. A large proportion of offspring spores and hyphal networks would remain within close proximity of the parental mycelium, an assumption likely to be true based on current knowledge of AMF life cycles. If random survival of genotypes (i.e. genetic drift) acts over generations, spatial genetic heterogeneity may arise and be maintained through limited dispersal. Genetic diversity would, therefore, be associated with the geographic scale at which sampling is performed. More distant locations or populations would show greater genetic divergence. This hypothesis is weakened by the findings of this study and those of Koch et al. (2004) because multiple isolates with very similar genotypes were sampled in geographically distant populations. Furthermore, genetic differences within the local population strongly exceeded differences among genetically distant populations. Nevertheless, a large-scale sampling of isolates of different populations in different locations is needed to confirm these findings.
A second hypothesis to explain patterns of genetic diversity concerns adaptation of AMF genotypes to host plants within populations. Taking together the finding of differential phosphate uptake of plants inoculated with different Glomus isolates (Munkvold et al., 2004), the finding of different host plant growth depending on AMF genotypes (Koch et al., 2006), and the results of our study indicating nonrandom association of isolate genotypes and host plants, there appears to be strong evidence for adaptation of AMF genotypes to host plants within populations. Therefore, the diversity of host plants in a particular field would represent a heterogeneous environment for AMF genotypes, based on preferential associations of AMF genotypes and host plant species. Such a process could lead to high genetic diversity within populations. Meanwhile, different AMF populations may not necessarily be genetically differentiated as the presence of similar host plants may have selected for similar AMF genotypes. The study by Koch et al. (2006) and the present study were based on a population sampled from an agricultural field subjected to crop rotation over 20 yr. These conditions may have favoured particular AMF genotypes.
A third hypothesis concerns hyphal fusions, or anastomoses, which were proposed as a potential factor shaping the evolution of genetic diversity in field populations (Sanders, 2002; Giovannetti et al., 2004; Bever & Wang, 2005; Stukenbrock & Rosendahl, 2005a). It was speculated that frequent hyphal fusions among different isolates and eventual genetic exchange may erode genetic differentiation in field populations. While hyphal fusions within populations may play a role in shaping the evolution of genetic diversity, no hyphal fusions were reported among isolates of different geographical origins (Giovannetti et al., 2003). Data from Stukenbrock & Rosendahl (2005a) showed strong linkage disequilibrium between alleles indicating a clonal evolution of genotypes in the three studied Glomus species. Data from Koch et al. (2004) and the present study show both strongly differentiated genotypes of G. intraradices and also genetically close genotypes. Anastomoses might, therefore, have played a role in shaping the genetic diversity of the field population, but the mechanism is clearly not strong enough to homogenize the different genotypes.
It would be possible to adapt SSR markers for amplification of DNA from single spores or root fragments, instead of DNA from in vitro cultures, using nested PCR primers based on flanking sequences of the repeat motifs. This should make it possible to carry out large-scale field studies across populations in order to study global distributions of genetic diversity and host plant associations. This would allow the testing of hypotheses about the evolution of genetic diversity within AMF species. Agricultural applications depending on the development of effective fungal inoculum should consider multiple sampling in field populations to obtain genetically diverse AMF as potential inocula and consider potential associations with particular plant species.