Troubles with truffles: unveiling more of their biology


Truffles are the hypogeous fruit bodies of the ascomycetes Tuber spp. They are highly praised and priced gourmet food, and their aroma and taste are known world-wide. But actually, with the exception of ‘specialists’, very few people are aware that the reproductive system of Tuber spp. is a real riddle. This is largely a consequence of the difficulties of growing and the impossibility of mating these symbiotic fungal species under controlled conditions. Molecular markers are now allowing us to look more closely into truffle population genetics and, in turn, into their life cycle and reproduction. Molecular investigations started more than a decade ago, originating from the primary need to reliably type morphologically similar truffle species (Henrion et al., 1994). The focus then shifted to the assessment of intraspecific genetic variability, particularly in species of economic interest such as Tuber magnatum Pico and Tuber melanosporum Vittad., which produce the most appreciated white and black truffles, respectively.

Bertault et al. (1998, 2001) were the first to claim that T. melanosporum experienced a strong genetic bottleneck during the last glaciation, so explaining the apparent absence of phylogeographic signals in T. melanosporum populations. A major implication of this is that the environmental conditions dictate the qualitative differences among truffles of different geographic origin. In addition, embracing the hypothesis that ascocarps result from the fusion of two haploid mycelia, and are therefore to be considered diploid structures, these authors also proposed a selfing reproductive system in Tuber spp. to interpret the absence of heterozygotes found when truffle ascocarps were screened with codominant markers. Bertault et al.'s conclusions and propositions have influenced all the subsequent studies on genetic structure within both T. melanosporum and T. magnatum, with a major impact on sampling strategies and mating models (Frizzi et al., 2001; Murat et al., 2004; Rubini et al., 2004; Mello et al., 2005). Yet, through enlarging the sampling areas and adopting increasingly informative markers, it has emerged that these two species are not so genetically depauperated as was previously thought. Indeed, it is possible to differentiate their populations genetically and track for each species a postglacial expansion pattern that fits nicely with that of most plant species with which these fungi have to establish mutualistic symbiosis (Murat et al., 2004; Rubini et al., 2005).

Even more intriguingly, simple sequence repeat (SSR) studies in T. magnatum have shown the occurrence of an extensive genetic exchange among geographically closed populations, as per two-locus and multilocus linkage disequilibrium analyses (Rubini et al., 2005). This extensive gene flow is hard to reconcile with the absence of any heterozygotic individuals. To resolve this conundrum, we hypothesized that T. magnatum outcrosses and, as most of the ascomycetes, has a prevalently haploid life cycle, with a ‘cryptic’ dikaryotic phase in the ascocarps (Rubini et al., 2005). In this scenario, the absence of heterozygotes results from a sampling bias as the DNA recoverable from ascocarps is contributed largely from the haploid, maternal tissue of the gleba. Conversely, the paternal DNA is not easy to recover because it is present only in the ascospores, and these structures are not usually broken during the DNA extraction process. Direct support for this idea comes from a new strategy allowing the differential recovery and analysis of DNA of pools of ascospores from the DNA of the surrounding gleba within single ascocarps. As expected, in most of the truffles analysed, the SSR patterns of the spores displayed alleles, of clear-cut paternal origin, in addition to those present in the gleba (Paolocci et al., 2006). Furthermore, in the same study we showed that primary mycelia, generated from germinating spores, are very likely homokaryotic as the mycorrhizas resulting from the inoculation of host trees with SSR-genotyped pools of spores were individually haploid. Thus, taken together these data argue against the thesis that truffle mycorrhizas are formed only after heterokaryotic mycelia are established (Lanfranco et al., 1995).

All in all, the SSR-assisted studies of T. magnatum ascocarps and mycorrhizas suggest the prevalence of the haploid phase in the truffle life cycle, a situation that typifies most ascomycetes, and substantiate the view that the fertilization process, and the resulting dikaryotic phase in Tuber spp., are spatially and/or temporally confined in the first stages of ascocarp development (Fig. 1).

Figure 1.

Simple sequence repeat (SSR) patterns and schematic representation of the Tuber magnatum life cycle. (a) Allelic configurations at two SSR loci displayed by the gleba and pools of ascospores from a single T. magnatum ascocarp and by mycorrhizal tips resulting from host-plant inoculation with the same truffle. (b) The T. magnatum life cycle as inferred from SSR studies. (1) Early stages of ascocarp development: the ascocarp primordia probably contain a dikaryotic mycelium (shown as white hyphae) resulting from fertilization, embedded in a network of haploid (homokaryotic) maternal hyphae (shown as grey hyphae). The dikaryotic mycelia develop with the formation of crozier and ascus mother cells (1a) where karyogamy takes place (1b). Karyogamy is shortly followed by meiosis, resulting in the formation of asci containing a variable number of ascospores (1c). (2) Mature ascocarp with asci and spores surrounded by hyphae of maternal origin. At this stage the paternal alleles can only be detected within spores. (3) Ascospores producing homokaryotic primary mycelia. (4) Homokaryotic mycorrhizas resulting from root colonization by primary mycelia. (5) The fertilization process.


These findings in T. magnatum raise some new, enticing questions.

  • 1Can all Tuber spp. outcross?
  • 2Are truffles prevalently outcrossing or heterothallic species?
  • 3What is the morphology of the mating structures in these fungi?

Whilst it is most likely that some other Tuber species, such as Tuber aestivum Vittad. syn. Tuber uncinatum Chatin (Wedén et al., 2004; Wedén, 2004), also outcross, and studies following the approach adopted for T. magnatum would help to assess this, the answer to the second question is more complex. Indeed, despite several attempts, the genes of the mating type have not yet been isolated in these fungi. Once again, population genetics studies, although indirectly, might offer a further opportunity to gain insights into the truffle life cycle. As a matter of fact, it should be possible to estimate the outcrossing rate in these fungi by comparing the expected and observed rates of heterozygosity within populations. Notably, the sampling should concern both the gleba and the spores in each ascocarp. Because each truffle ascocarp results from an independent mating event, that truffles with identical genetic profiles in their gleba represent clones of the same individual (Bertault et al., 2001; Murat et al., 2004) is no longer to be taken for granted.

Despite all these advances in our understanding of the truffle reproductive cycle, one black hole still remains. This is the morphology of the fertilization process. However, the apparent absence of male hyphae in the gleba and the fact that specialized male structures (antheridia) have never been described in these fungi let us argue that the male gamete function may be fulfilled by any detached cells, such as ascospores, hyphal fragments or even mitotic spores (spermatia). Although further studies are needed to shed light on this key point of truffle biology, further support for this hypothesis is provided by the recent and fascinating finding that Tuber spp. produce mitospores (Urban et al., 2004).

Practical implications

Certainly, the discovery of a genetic and phylogeographic structure in T. magnatum and T. melanosporum is likely to have a major impact on attempts to elucidate whether, in addition to environmental conditions, genetic determinants shape the morphology of, and dictate the organoleptic differences within, any given truffle species over its geographical range. However, how this discovery will affect the development of strategies for the cultivation and marketing of these fungi is equally relevant. Our prediction is that the availability of more molecular markers will increasingly make it possible to trace natural truffle populations according to their geographic origin. Far from being a secondary issue, this is of great practical importance for the associations of truffle harvesters and local governments who are actively promoting the economic and social development of rural and marginal areas. At the same time, these results pose intriguing new questions about the potential problems linked to microbial competition and loss of fungal biodiversity. Not least, this is also a relevant ecological problem. Artificial truffle plantations are often established in naturally productive areas to counterbalance the sharp decline in wild truffle harvests (Hall et al., 2003). However, the possible consequences of the deliberate introduction in naturally truffle-producing areas of host trees that have been nursery-inoculated with nonindigenous fungal strains have largely been unexplored.

The notion that T. magnatum, at least, is not an exclusively selfing species is important for successful truffle cultivation. If truffles are the product of a preferentially outcrossing or heterothallic species, the presence of genetically distinct strains or of strains with opposite sexuality at the cultivation site would be the major requirement to allow these fungi to fruit. Thus, we believe that a careful re-evaluation of the procedures for host-plant inoculation is opportune. Such a re-evaluation should promote the presence of as much genetic variability within an artificial truffle plantation as possible. The routine in nursery practice is to inoculate host plants with ascocarps. However, it would be extremely interesting to investigate whether mycorrhizas result from the germination of many or only a few spores or, alternatively, result prevalently from the maternal hyphae of the gleba. Following the same reasoning, the envisaged procedure of large-scale host-plant inoculation using in vitro cultivable individual mycelial strains is not likely to be an advisable practice.

Last but not least, there remains a very attractive hypothesis that still needs to be tested. This is whether, aside from the impact of environmental factors, the underrepresentation of local truffle biodiversity is one of the underlying reasons for unsuccessful production in some artificial truffle plantations.