• anastomosis;
  • arbuscular mycorrhiza;
  • genetic exchange;
  • Glomus intraradices;
  • heterokaryon.

Arbuscular mycorrhizal (AM) fungi are remarkably promiscuous in their choice of plant partners. The majority of land plants have these fungal symbionts in their roots, and some of the fungi have been shown to colonize many plant species spanning a wide taxonomic range (Smith & Read, 2008). The fungi may provide mycelial connections between very different plants (Giovannetti et al., 2004). By contrast, interactions between AM fungi were thought to be much more restricted. As in many filamentous fungi, cytoplasmic connections (anastomoses) can sometimes form when hyphae meet, but it seemed that incompatibility mechanisms ensured that this could only happen when the fungi were identical in genotype (Giovannetti et al., 1999). However, a new study shows that anastomoses can form, albeit at a low frequency, between genetically distinct isolates of Glomus intraradices (Croll et al., this issue of New Phytologist; pp. 924–937). The anastomoses allowed nuclei to pass between the fungi and, because there are very few cross-walls in AM fungal hyphae, they could potentially mingle freely and grow in a common cytoplasm. Indeed, when the authors took spores arising from mixed cultures, they found that the mycelium which grew from them had genetic markers from both parents, demonstrating a hybrid origin.

‘This new observation of genetic mingling through anastomosis adds a new process that needs to be incorporated into the story ...’

Fungal hyphae are fragile, and in the soil they are readily fractured by mechanical movement or nibbled through by small animals. Anastomoses provide robustness, allowing broken connections to be repaired and multiple potential routes between distant points in the mycelium. This network-maintenance function is probably the main reason that AM fungi have the ability to make anastomoses, because Croll et al. observed that anastomoses formed more readily between hyphae from spores of the same culture than when genetically distinct isolates met. If genetic exchange were the main advantage, then we would expect the opposite: self interactions should be disfavoured, as they are in the self-incompatibility systems that promote outcrossing in many plants. Nevertheless, genetic exchange that occurs as a by-product of a mechanism which exists for other reasons may have important functional and evolutionary consequences, as in phage-mediated transduction of bacteria, for example.

These observations are relevant to a debate among mycorrhizal researchers as to whether all nuclei in an AM fungus are genetically identical (a homokaryon) or if a number of genetically distinct nuclear lineages co-exist (a heterokaryon). The immediate consequence of nonself anastomosis is that a heterokaryon is formed: populations of genetically distinct nuclei come to share a common cytoplasm. The mixing is not transient, as shown by Croll et al. who found that markers from both partners found their way into the same spore and persisted together for more than a year of subsequent culture. If heterokaryons are the normal state for AM fungi, then anastomosis provides a mechanism that can potentially increase the diversity of nuclei in a mycelium. It is important that there is such a means to replenish diversity because one might expect that genetic drift or selection between nuclei would lead to a decline in nuclear diversity as a fungus grows and sporulates.

If, on the other hand, a homokaryon is the normal state of an AM fungus, then the observations require an altogether more radical interpretation. For genetic markers from different parents to persist in a homokaryon, they must come together into the same nucleus. Croll et al. suggest that some type of parasexual process might occur, in which nuclei fuse and generate a novel recombinant type, although not through conventional meiosis. Parasexuality occurs in some other fungi, and nuclear fusion is not too hard to imagine when the nuclei already share a common cytoplasm. However, it must be emphasized that there is no evidence, to date, either cytological or genetic, for recombination between nuclei in AM fungi.

If nuclei pass through an anastomosis, then one can imagine that mitochondria would also be transferred, although Croll et al. did not investigate this. There is considerable mitochondrial genetic diversity among isolates of G. intraradices, so an anastomosis between them would be likely to generate a heteroplasmic mycelium, with a mixture of genetically different mitochondria. Mitochondrial heteroplasmy might be expected to persist for longer than nuclear heterokaryosis because there are more mitochondrial genomes so genetic drift will be less effective. Surprisingly, the observation is that there is virtually no heteroplasmy in G. intraradices isolated from the field (Raab et al., 2005; Börstler et al., 2008). If anastomosis is constantly creating heteroplasmy, then the processes that restore mitochondrial homogeneity must be relatively rapid. Perhaps there are active segregation mechanisms comparable with those in yeast (Barr et al., 2005; Shibata & Ling, 2007), or perhaps there are genetic bottlenecks in which mitochondrial numbers are greatly reduced. Given that polymorphic mitochondrial markers are available (Raab et al., 2005; Börstler et al., 2008), investigations into the persistence of heteroplasmy seem feasible.

The debate over homokaryosis vs heterokaryosis in AM fungi stemmed from the observation that several genetic variants could be detected in single spores and persisted in cultures. The first observations (Sanders et al., 1995; Lloyd-Macgilp et al., 1996) involved the internal transcribed spacer of the ribosomal RNA genes, which is multicopy in most organisms although concerted evolution usually prevents the individual copies from diverging in sequence. However, it is not just this region that is heterogeneous within AM fungi: there are also co-existing sequence variants for genes that are typically found in single copy (Helgason et al., 2003; Koch et al., 2004). The homokaryotic and heterokaryotic explanations of this phenomenon, and the sometimes conflicting evidence supporting one or the other, have been rehearsed in the literature (Kuhn et al., 2001; Pawlowska & Taylor, 2004; Hijri & Sanders, 2005; Rosendahl & Matzen, 2008; Young, 2008). This new observation of genetic mingling through anastomosis adds a new process that needs to be incorporated into the story, and perhaps offers new possibilities for resolving the question, but it does not in itself provide an answer one way or the other.

These days, the obvious way to address a question of genome organization is to sequence the genome, and a genome project for G. intraradices is indeed under way (Martin et al., 2008). There is no doubt that the genome-sequencing project will eventually lead to a complete assembly of the genome, but even this may not directly answer the heterokaryosis question. The result will presumably be a set of chromosomes, but this will not tell us how these are packaged into nuclei. The results available so far fully confirm the impression that many genes are members of families of related, but distinct, sequences. As a result, the ‘genome space’ is much larger than that of a fungus that gets by with just one variant of each gene. Initial estimates were that each nucleus of G. intraradices had only 16 Mb of DNA (Hijri & Sanders, 2004), comparable with the genome size of yeast, but it is now clear that the amount of distinct sequence is at least an order of magnitude greater than this (Martin et al., 2008). The observation that genetically distinct nuclei can come together through anastomoses (Croll et al.) offers an intriguing interpretation of this result. Suppose that we change our perspective and think of an interconnected mycelium as a subpopulation rather than as an individual (Fig. 1). In this subpopulation there are different nuclear types that mix relatively freely and that occasionally migrate into other subpopulations (mycelial networks) through anastomoses. In this way of thinking, it is the nuclei, not the whole mycelium, that constitute the ‘individuals’. Spore formation is, at most, a mild bottleneck, because AM fungal spores commonly contain hundreds or thousands of nuclei. Spores are ‘lifeboats’ that may carry much of the genetic diversity of the subpopulation through hard times or to new pastures.


Figure 1. A metapopulation concept for arbuscular mycorrhizal fungi. If genetically diverse nuclei persist within a mycelial network (this is not proven), then the assemblage of nuclei can be considered as a subpopulation, with the nuclei as individuals. Anastomoses within a mycelium (coloured circles) provide robust connectivity, while anastomoses between genetically distinct mycelia (as demonstrated by Croll et al., black circles) allow exchange (migration) of nuclei (individuals, coloured spots) between mycelia (subpopulations). Mitochondria, on the other hand, are genetically uniform within each mycelium, but differ between mycelia (Börstler et al., 2008), as indicated by the background colour of the hyphae. Genetically distinct mycelia may differ in important properties, such as host preference.

Download figure to PowerPoint

Understanding the genetic organization of AM fungi has an importance beyond the intrinsic interest of understanding an unusual biological system. Croll et al. measured hyphal density and spore production of the parental isolates and their progeny, and found consistent differences. Genotypic differences mean phenotypic differences, then, and phenotypic differences are certain to mean differences in function. Arbuscular mycorrhizal fungal genotypes differ in the benefit they confer on their hosts (Munkvold et al., 2004) and hence their role in the ecological community. If we are to understand how AM fungi will respond to environmental change, and if we hope to use genetic tools to dissect their physiology or to develop superior inoculants, then we need to piece together the story of their life history. The discovery of nuclear transfer through anastomoses between genetically distinct fungi gives us a significant new piece of the puzzle.


  1. Top of page
  2. References