Ribosomal RNA gene sequence diversity in arbuscular mycorrhizal fungi (Glomeromycota)



    Corresponding author
    1. Research School of Biosciences, University of Kent, Canterbury CT2 7NJ, UK
      *Present address and correspondence: Alia Rodriguez, Institut für Pflanzenbau und Günland (340), Universität Hohenheim, D-70593 Stuttgart, Germany (fax +49 711 459 2297; e-mail aliarodn@uni-hohenheim.de).
    Search for more papers by this author

    1. Research School of Biosciences, University of Kent, Canterbury CT2 7NJ, UK
    Search for more papers by this author

    1. Research School of Biosciences, University of Kent, Canterbury CT2 7NJ, UK
    2. International Institute of Biotechnology, 1/13 Innovation Building 1000, Kent Science Park, Sittingbourne, Kent ME9 8HL, UK
    Search for more papers by this author

*Present address and correspondence: Alia Rodriguez, Institut für Pflanzenbau und Günland (340), Universität Hohenheim, D-70593 Stuttgart, Germany (fax +49 711 459 2297; e-mail aliarodn@uni-hohenheim.de).


  • 1The existence of multiple rRNA sequences within AMF single spores and the high levels of variability found in these symbionts for highly conserved genes with low copy numbers have provoked intense controversy over the genetic structure of the Glomeromycota. Lately, two hypotheses have been proposed to explain the source of such a variation, mainly either at internuclear (multigenomic) or intranuclear (polyploidy) levels.
  • 2Evidence reported by numerous authors suggests that the magnitude of sequence variation, and the presence in single spores of sequences from apparently unrelated taxa within the Glomeromycota, makes the use of one, or very few sequences, from only one or two isolates of AMF insufficient for taxonomic purposes in some species groups.
  • 3A large-scale, collaborative study on genetic variation in AMF across different genes could help to address fundamental questions on the biology and genetic structure of these important symbionts.


The arbuscular mycorrhizas are the most widespread association between plants and symbiotic soil fungi, occurring within the roots of most Angiospermophyta, Pteridophyta, Bryophyta and Coniferophyta. It has been calculated that approximately 80–90% of all land plant families associate with arbuscular mycorrhizal fungi (AMF) (Trappe 1987), so that most terrestrial soils are potentially dominated by this mutualistic association. The fungi support the plants with mineral nutrients and other services and, in turn, receive photosynthates from the autotrophic plant partner.

Arbuscular mycorrhizal fungi (AMF), which are located at the plant–soil interface, are an important link in the chain response of ecosystems to abiotic factors such as nutrient fluxes, water availability and soil pollution. It has become apparent that the association between AMF and vascular plants has decisive consequences for the survival and functioning of plant communities and ecosystems. Key studies on plant community interactions have reported that AMF act as determinants of plant diversity and competition (van der Heijden et al. 1998), and may enhance community productivity (Gange et al. 1993; Klironomos et al. 2000). Thus, although AMF studies have tended to emphasize the nutritional aspects, the multifunctional role that this symbiosis plays in terrestrial ecosystems is now recognized as also being important.

Evidence for the functional diversity of AMF species has emerged in recent years (Smith & Read 1997; Burleigh et al. 2002). Growth responses depend on the particular host/fungus combination (van der Heijden et al. 1998; Klironomos et al. 2000), highlighting the identity of the AMF species as a significant factor governing plant growth in ecosystems. Identification of which species within the Glomeromycota are responsible for the various associations is therefore an important goal in modern plant ecology. However, for various technical reasons, AMF identification in planta has proven difficult in field conditions. One of these difficulties has been to design species-specific primers because the target DNA can vary greatly, even within one individual. DNA-based methodologies have, however, been used for AMF identification under field conditions (Clapp et al. 1999; Helgason et al. 1999; Antoniolli et al. 2000; Pringle et al. 2000), and for taxonomic studies (Simon 1996; Lloyd-MacGilp et al. 1996; Schüßler et al. 2001). The question, however, remains as to whether the genetic variation found within one AMF individual or within one AMF species is too great to allow identification of a species or to make an appropriate phylogeny based on just one or a small number of sequences.

AMF multiple sequences

The polymorphic nature of ribosomal RNA gene sequences in AMF was first described by Sanders et al. (1995). Subsequently more studies reported multiple sequences across both variable (ITS) and conserved regions (SSU and 5.8S) of the rRNA genes of AMF and among species belonging to different genera (Lloyd-MacGilp et al. 1996; Clapp et al. 1999; Helgason et al. 1999; Hosny et al. 1999; Lanfranco et al. 1999; Antoniolli et al. 2000; Pringle et al. 2000; see Table 1). More importantly, in several of these studies multiple sequences have been isolated from single AMF spores, which sometimes have shown no taxonomic correlation (Clapp et al. 1999). These results have sparked interesting theories about the unusual genetic structure of these ancient fungi (Sanders 1999; Kuhn et al. 2001), with the suggestion that the fungi carry multiple genomes located on different nuclei. Another more recent proposal is that the high genetic variation within AMF individuals is due to high ploidy (Pawlowska & Taylor 2004).

Table 1.  Examples of the magnitude of sequence polymorphism found in AMF and the numerous publications in this area. BEG = La Banquee Européenne des Glomales, now known as the International Bank of Glomeromycota (http://www.ukc.ac.uk/bio/beg); SS = single spore from pot culture; MS = multispores from pot culture; fMS = multispores from field; fSS = single spore from field; fR = field root extraction; tSS = single spore from trap culture; ITS = ribosomal internal transcribed spacers; SSU = ribosomal small subunit; LSU = ribosomal large subunit
AuthorsSpeciesCulture codeSourceRegionNo. clones examinedNo. sequences obtained
Sanders et al. (1995)G. mosseaeBEG 12SSITS  3 2
Lloyd-MacGilp et al. (1996)G. mosseaeBEG 25MSITS  3 1
BEG 57MSITS  3 3
BEG 61SSITS  3 2
BEG 55SSITS  2 1
BEG 54SSITS  3 3
Lloyd-MacGilp et al. (1996)G. dimorphicumBEG 59SSITS  4 4
G. fasciculatumBEG 58SSITS  3 2
G. coronatumBEG 28MSITS  3 3
Lanfranco et al. (1999)Gi. margaritaBEG 34MSITS  4 4
 3 × SS   8 7
BEG 343 × SSITS 1211
Clapp et al. (1999)Scutellospora sp.3 × fSSSSU 1212
fMSSSU  6 4
fRSSU  7 7
Helgason et al. (1999)fRSSU 3835
Pringle et al. (2000)A. colossicafSSITS 2422
Antoniolli et al. (2000)G. mosseae3 × fSSITS 2323
Gi. margaritafSSITS  9 9
SSITS  3 3
MSITS  7 7
Clapp et al. (2001)G. coronatumBEG 28TMSLSU 3220
BEG 139MSLSU 4218
BEG 49MSLSU 4123
FO 97MSLSU 4314
G. geosporumBEG 11MSLSU 3514
G. mosseaeBEG 25MSLSU 3813
G. constrictumBEG 130MSLSU 2613
Rodriguez et al. (2001)Entrophospora infrequens6 × tSSLSU21348
fMSLSU 4813
Gigospora roseaBEG 143MSLSU 46 7
Scutellospora heterogamaBEG 40MSLSU 41 8
Acalospora spinosaBEG 10MSLSU 11 6
A. tuberculataBEG 41MSLSU 13 9

Clearly, the issue of multiple ribosomal RNA gene sequences needs further examination before sequences can be used to infer taxonomic status from ecological studies undertaken in the field. As an indication of the amount of variation within Glomus isolates sequences analysed by Rodriguez et al. (2004) have shown that G. etunicatum BEG186 had up to 19% bp differences. This compares with G. mosseae, a much tighter species group, which only has up to 2% of differences (reported in Rodriguez et al. 2004). There have been few studies that have extensively analysed sequences from several isolates of single AMF species (Clapp et al. 2001; Rodriguez et al. 2001, 2004). Schüßler et al. (2001) acknowledge the presence of multiple sequences but dismiss their importance for phylogenetic analyses, stating that the differences between sequences are minor and that they cluster together. This is usually true, if only a few sequences are examined, but, as sample size increases, sequences begin to be found that often cluster with different species (Clapp et al. 2001) or different genera (Rodriguez et al. 2001). For instance, in Glomus constrictum BEG130 (Clapp et al. 2001), where 13 sequences were found after an analysis of 26 clones, 92% clustered together with a published G. constrictum sequence but 8% clustered with published Glomus coronatum sequences. Thus, if only four clones were examined, there would only be a 0.28 chance of detecting a sequence that does not cluster with the majority.

In fact, fewer than four clones are usually analysed from any single isolate or spore, reducing the chances even further. This difficulty is further highlighted by the detection of ribosomal sequences from different genera of AMF, sometimes coexisting at high frequency in single spores, e.g. in Entrophospora infrequens (Rodriguez et al. 2001). This is, however, an unusual fungus that has yet to be isolated in pure culture and yet to be proven to form arbuscular mycorrhizas. Schüßler et al. (2003) have criticized recent work on high genetic variability with single AMF isolates as being due to the sequencing of contaminants. However, this was only proven to be true in two cases from numerous AMF isolates tested and the authors criticized have always taken great care to avoid cross-contamination with other AMF. In the E. infrequens study, the inner ‘endospore’ containing spore cytoplasm was even excised from the ornamented outer wall layer to eliminate potential spore surface contaminants (Rodriguez et al. 2001). We remain confident that this is an unusual fungus within the Glomeromycota and that the other data on ribosomal LSU sequences (Table 1) are robust. These data imply that species boundaries within the Glomeromycota are not always easy, and may even be impossible to establish due to anomalous isolates with slightly different spore morphologies: the result is ‘blurred’ or overlapping species clusters rather than ‘nice’ groupings. In addition, the use of rRNA sequences as molecular markers for tracking AMF in field conditions has clearly limited applicability, and it could only be useful once a substantial data base of sequence variation has been established to prevent erroneous interpretations (e.g. Helgason et al. 1999). Unfortunately, it appears now that the primers used by Helgason et al. (1999) may not be Glomeromycotan specific, as first thought (Schüßler 2003).

AMF genetic structure

Some species of AMF may experience genetic transfer through nuclear exchange and this is probably occurring at low frequency. Spore morphological disparity, indicated in the newly proposed family Diversisporaceae (Schwarzott et al. 2001), could be a reflection of this genetic exchange, resulting in a series of intergrades between examples representing the classic (holotype) species. The traditional concept of species as discrete genetic units may, in AMF, as with many other microorganisms (Bachmann 1998), be flawed. Exchange of nuclei has been inferred to occur in several AMF species from the presence of nuclei at the bridge of anastomosing hyphae in laboratory microcosms (Giovannetti et al. 1999, 2001). Despite the fact that anastomosis has only been reported between isolates of Glomus belonging to the same species, exchange of nuclei through hyphal anastomosis between AMF species cannot yet be ruled out. In their study, Giovannetti et al. (1999) only observed 200 hyphal contacts from the potential of nearly a million anastomoses that could have occurred, assuming an average of 10 m hyphae per cm3 and 3.1 anastomoses per cm of extra-radical hyphae. More recently, Giovannetti et al. (2001) reported the widespread occurrence of anastomoses and nuclear migration in an intact extraradical arbuscular mycorrhizal network.

The magnitude of sequence variation and the presence of sequences from apparently unrelated taxa within the Glomeromycota in single spores (albeit at low frequency) indicates that the use of one or very few sequences from only one or two isolates of AMF as supporting evidence for the erection of new taxa (e.g. Redecker et al. 2000) is dangerous and lends weight to the reservations expressed by Schwarzott et al. (2001). It appears likely that AMF species conceal considerable genetic variation that may represent the mutational history of each nucleus and the population dynamics of these nuclei must be understood before sequence data can be used to determine the dynamics of natural AMF populations and maybe for taxonomic purposes. It seems that many researchers are actively choosing to ignore multi-sequence data, despite the magnitude of genetic variation and the apparent continuum seen between species or possibly genera which have clear implications for ecological studies and taxonomy. Instead of ignoring this variation, new insights into AMF biology could be obtained by documenting it in greater detail for a larger number of AMF species. The presence of sequences from separate taxa even at a low level must be addressed and explained as they have a direct bearing on the integrity of the species concept in AMF. In addition, the clonal fungal material could be used, if they can be cultured, in an in vitro system using transformed roots (e.g. Declerck et al. 1996), to resolve the issue of whether ‘across-taxa’ sequence variation is due to cross-contamination of cultures. Clearly, single sequences are insufficient to define a taxon or even an isolate, as is the use of single isolates of a species.

The high levels of AMF genetic diversity discussed here should also be analysed in light of the wide phenotypic variation in spore morphology observed in these symbionts, their physiological plasticity and adaptation to widely different terrestrial ecosystems. Finally, in view of the magnitude of sequence variation reported in AMF and with the screening potential of techniques such as SSCP (single strand conformational polymorphism) and the wide availability and low cost of sequencing, it should no longer be acceptable to draw taxonomic conclusions on the basis of one or two isolates of species studied. We would, however, make a plea for the funding of an international project supporting a large-scale collaborative study of genetic variation in these ecologically important fungi across a range of different genes. Until the underlying causes of the observed sequence variation can be understood, the interpretation of AMF ecology, taxonomy and phylogenetics remains open to debate. It is still early days in this field and the establishment of entrenched positions between scientists in the field should be avoided.