Truffles: much more than a prized and local fungal delicacy

Abstract

Truffles are hypogeous fungi which live in symbiosis with plant host roots in order to accomplish their life cycle. Some species, such as Tuber magnatum Pico, the ‘white truffle’, and Tuber melanosporum Vittad., the ‘black truffle’, are highly appreciated in many countries because of their special taste and smell. The great demand for the black and white truffles, the increasing attention towards other species of local interest for the rural economy (such as T. aestivum) together with a drop in productivity, have stimulated researchers to develop projects for a better understanding of the ecology of truffles by exploiting the new approaches of environmental microbiology and molecular ecology. Specific primers have been developed to identify many morphologically similar species, the distribution of T. magnatum has been followed in a selected truffle-ground, the phylogeography of T. melanosporum and T. magnatum has been traced, and the microorganisms associated with the truffles and their habitats have been identified.

Introduction

Truffles were prized by ancient Greeks and Romans as an Epicurean delight as far back as the Apicius's legendary banquet in 20 AD, but the scanning of scientific literature on truffles also demonstrates a longstanding love affair between science and the aromatic fungus. The first paper devoted to the nature of truffles appeared in 1564 (Ciccarelli, 1564), while other publications have made them a cult food since 1826 (Brillat-Savarin, 1826). Later the scented truffles were identified as ‘diamonds’ and ‘the best food’ by Hervé This, who was the inventor of molecular gastronomy. What we call a truffle is a hypogeous edible fungus that undergoes a complex life cycle during which the mycelium establishes a symbiotic interaction (ectomycorrhizae, Fig. 1a) with the roots of trees, such as oak, poplar, willow and hazel (Harley & Smith, 1983), and some shrubs, such as Cistus (Fontana & Giovannetti, 1978–79). As a final step, hyphae aggregate and develop a fruiting body – the truffle – which is an ascoma bearing asci and the products of meiotic events, the ascospores (Fig. 1b).

Figure 1.   Tuber magnatum mycorrhizae (a) and ascospores (b).

The true truffles belong to the genus Tuber, one of the few ectomycorrhizal Ascomycetes, even if other Ascomycetes are considered truffles, such as the desert truffles (Kirk et al., 2001). While in traditional classification systems the true truffles were included in the order Tuberales, together with all hypogeous ascomycetes, today they are placed in the order Pezizales. This includes both hypogeous and epigeous fungi, with either saprotrophic or symbiotic lifestyles, but which are all related phylogenetically (O'Donnell et al., 1997; Percudani et al., 1999).

Some species, such as Tuber magnatum Pico, the ‘white truffle’, and Tuber melanosporum Vittad., the ‘black truffle’, are in great demand by the food market in many countries because of their special taste and smell, resulting from a blend of hundreds of volatile compounds (Bellesia et al., 1998; Gioacchini et al., 2005). Several biotic (fungi, yeasts, bacteria, mesofauna, plant host) and abiotic (soil composition, weather such as rain, sunshine and temperature) factors could influence truffle life and enhance or inhibit ascocarp formation (Ceruti et al., 2003). Truffle species present common ecological features such as a relatively wide range of host species and the need for a calcareous soil (pH between 7 and 8), except Tuber borchii tolerating slightly acidic soils. On the contrary, there are important differences in their geographic distribution. While T. borchii and Tuber maculatum are found throughout Europe (Riousset et al., 2001), T. melanosporum is collected in the South and West Europe – Italy, France and Spain – and T. magnatum fruiting bodies have so far been collected in Italy and in the East Europe – Croatia, Slovenia and Hungary – resulting in a limited availability.

At 300–400 Euros per 100 g during the 2004 fruiting season, it is clear why T. magnatum fruiting bodies are one of the most expensive delicacies, together with caviar. From September to January, truffle hunters search for the piquant truffles helped by pigs or trained dogs. The great demand for black and white truffles, the increasing attention towards other species of local interest for the rural economy (e.g., Tuber aestivum) together with their drop in productivity, stimulated researchers to develop projects for a better understanding of the ecology of truffles by exploiting the new approaches of environmental microbiology and molecular ecology.

The aim of this review is to briefly illustrate how many of the open questions raised by truffle biology regarding systematics, diagnosis, population genetics and phylogeography, and morphogenesis, have been addressed recently by researchers, taking advantage of the most advanced techniques of molecular biology.

Morphological vs. molecular identification

Truffles live all over the world, distributed especially in many regions of the northern hemisphere. Around 200 European species, varieties and forms of Tuber have been described by mycologists over the centuries (from the 18th to the 20th). The various events linked to the species and their synonyms are summarized in a monograph of the European species of Tuber where the authors consider only 28 species to be valid (Ceruti et al., 2003).

Truffle fruiting bodies are usually identified from the size and shape of their spores and asci, spore wall ornamentation, structure of the peridium and gleba. These features are generally recognized by specialists; however, sometimes, identification is unreliable. This is the case for species with very similar morphological features, such as T. brumale–T. brumale var. moschatum and T. melanosporum–T. hiemalbum (Riousset et al., 2001).

Biochemical tools, such as one-dimensional gel electrophoresis with total protein (Mouches et al., 1981; Dupréet al., 1985) and isoenzyme analysis (Pacioni & Pomponi, 1991; Gandeboeuf et al., 1994; Urbanelli et al., 1998), were the first to be used to verify the morphological identification of truffles. However, these techniques present several problems: isoenzyme analysis needs high quality material while total protein analysis depends on physiological status and cannot be used with mycorrhizae.

In 1993, our laboratory applied, for the first time, molecular tools for identifying truffles. Multiloci analyses such as RAPD (random amplified polymorphism DNA) and RAMS (random amplified microsatellites sequences), respectively, identified six (Lanfranco et al., 1993) and 11 Tuber species (Longato & Bonfante, 1997).

As multiloci techniques also amplified DNA from plant tissue present in mycorrhizae and from bacteria present in fruiting bodies, single locus analyses predominate, today, for Tuber identification.

Specific primers, designed on the internal transcribed spacer regions (ITS) of ribosomal genes, have been developed to discriminate T. magnatum from the so-called ‘whitish’T. borchii and T. maculatum characterized by less taste and a lower commercial value (Amicucci et al., 1998; Mello et al., 2000) (Fig. 2). As well as for white truffles, specific primers distinguish T. melanosporum from T. brumale and the Asiatic Tuber indicum, both of which are used in food in place of T. melanosporum (Rubini et al., 1998; Douet et al., 2004). Processed foods sold as containing T. melanosporum can be contaminated with up to 5% of other Tuber species (Mabru et al., 2004).

Figure 2.  Internal transcribed spacer regions amplification with specific primers P7/M3 from fruitbody and mycorrhizae of Tuber magnatum and from different Tuber spp. Only T. magnatum shows a 434 bp band. First lane: molecular size marker pUC18 DNA HaeIII digest. (Modified from Fig. 2 in Mello et al., 1999.)

Besides ascocarp identification, the identification of the fungus during its symbiotic phase has been one of the major topics in truffle research. The molecular methods developed for fruiting body identification have also provided diagnostic tools to confirm the occurrence of the desired fungus in the mycorrhizal roots (Stocchi, 1999; Mello et al., 2001; Rubini et al., 2001) where morphological identification is more difficult. This is fundamental in order to follow the fate of inoculated plants.

Although research on the above species was important because of the high commercial profits involved, another challenge has been the controversial taxonomic position of T. aestivum Vittad. with respect to Tuber uncinatum Chatin. These taxa also have a moderate commercial value, and unlike T. magnatum and T. melanosporum, they have a wide geographic distribution. They have been found as far north as Gotland Island, in Sweden. The length of their spore reticulum was considered the most useful morphological characteristic for distinguishing the two taxa. However, some ecological features, geographical distributions and smell and taste are distinctive of the two taxa. These taxa have been separated by total protein analysis (Mouches et al., 1981). After two publications leading to controversial results (Mello et al., 2002; Paolocci et al., 2004), but providing the first ITS sequences of these taxa, a very recent work based on a higher number of samples seems to close the debate: the height of the spore reticulum is not diagnostic, and it is not possible to separate T. aestivum from T. uncinatum (Weden et al., 2005).

The Chinese black truffles (Tuber pseudohimalayense, Tuber sinense and T. indicum) are other cases where species limits are still discussed, as illustrated by Zhang et al. (2005). As for prokaryotes, where assignment of isolates to species is based on phenotypic measures and genome similarity (Gevers et al., 2005), similar difficulties exist for defining the species of fungi (Taylor et al., 1999).

Even if not always possible, the availability of molecular probes has helped to resolve the problem of interspecific differences of truffles, limiting frauds, and answering questions related to their taxonomy but, above all, allowed researchers to investigate the fungus during its symbiotic interaction with the trees in the truffle-ground.

From genetic variability to a hint in the past

Truffles show variation in several traits, including their organoleptic properties, across their geographical range. These variations could be due to environmental and/or genetic factors. In a highly quoted paper, Bertault et al. (1998) claimed: ‘morphological and organoleptic differences seen over the geographical range of the black truffles can probably be explained by environmental variation rather than by genetic factors.’ However, Murat et al. (2004) were successful in revealing a strong geographic pattern for T. melanosporum (F_ST=0.20) using the moderate variations of ITS region of nuclear rRNA gene sequences. A significant association between genetic (F_ST) and geographical distances (Mantel test) was also found when the distance values between populations in western France and the populations of the Rhone Valley were clustered in two groups. On the basis of this significant genetic differentiation and in contrast with previous data (Bertault et al., 1998, 2001), it is suggested that genetic determinants play a role in the organoleptic differences of the black truffle.

Bertault et al. (1998) explained the low level of polymorphism found for T. melanosporum suggesting that it went through a population bottleneck after which new allelic haplotypes have originated in low frequencies. During the maximum expansion of the glaciers, the deciduous forest of Europe was restricted to the Mediterranean coastal zone (Bennet et al., 1991). Climatic and fossil data support the hypothesis that three regions hold the main glacial refuges for the host trees of T. melanosporum: the Iberian and Italian Peninsulas, and the Balkans. Since T. melanosporum is an ectomycorrhizal symbiont of oaks and other temperate deciduous trees, such as Tilia and Corylus species, this symbiotic fungus was likely to have been restricted to some of these regions. Although pedoclimatic factors likely influenced T. melanosporum recolonization patterns, two routes of expansion were proposed for the Perigord truffle in France closely resembling those of Quercus pubescens (Petit et al., 2002): the Rhone valley route and the Atlantic route (Murat et al., 2004, Fig. 3).

Figure 3.  Distribution of the 10 internal transcribed spacer (ITS) haplotypes of Tuber melanosporum in France. The pie chart diameters are proportional to the number of ascocarps analyzed per region. Black lines delimit areas of distribution of the chloroplastic DNA (cpDNA) haplotypes of oaks in France (Petit et al., 2002): haplotype 1 was found in the southern corner of France; haplotype-7 in the Rhone valley, and haplotypes 10–12 in western France. Arrowed lines show potential postglacial recolonization routes for the Perigord truffle: the Atlantic red route and the Rhone valley blue route (adapted from Fig. 2 in Murat et al., 2004).

Like T. melanosporum, T. magnatum has been the subject of studies focused on genetic variability (Lanfranco et al., 1993; Gandeboeuf et al., 1997; Frizzi et al., 2001). Thanks to the finding of a single nucleotide polymorphism in T. magnatum ascocarps, it was possible to identify two haplotypes in a northern Italian population and to trace their spatial and temporal distribution (Mello et al., 2005, Fig. 4).

Figure 4.  Three haplotypes of Tuber magnatum originated by two single nucleotide polymorphisms, indicated by a star in an oligonucleotide sequence (the different colours indicate nucleotides: T, red; C, blue; G, black; A, green). Two haplotypes are present in the northern Italian population (adapted from Fig. 2 in Mello et al., 2005).

Moreover, analysis of samples coming from Italian and Balkan populations suggested a genetic differentiation in T. magnatum over its habitat (Mello et al., 2005), a finding which was further confirmed by Rubini et al. (2005) after an extensive analysis. By polymorphic microsatellites, these authors found that the southernmost and the northwesternmost populations were significantly differentiated from the rest of the populations and probably spread from a refuge of T. magnatum in central Italy during the postglacial expansion.

Unlike the precious black and white truffles, other truffle species show a higher genetic diversity, i.e. T. borchii, T. aestivum and T. maculatum (Lanfranco et al., 1993; Gandeboeuf et al., 1997; Mello et al., 2002; Paolocci et al., 2004). It is interesting to note that these species have a larger geographical distribution than T. magnatum and T. melanosporum (Riousset et al., 2001); for example, T. maculatum is found from France to Russia. It is probable that quaternary climatic modifications had minor effects on them and consequently the bottleneck has been less important for them. However, genetic diversity analyses of these species in their whole geographic distribution are not yet available leaving many question marks over these hypotheses. An additional problem arises from the ambiguity of ascocarp identification in some species: for example the high level of genetic intraspecific variability so far described in T. borchii may be related to uncorrected identifications (A. Zambonelli and A. Mello, pers. commun.).

Truffle productivity and truffle-ground environment

Improving agricultural productivity, through ecosystem management techniques, has been the first goal of humans as they had to face the increasing demand for food. Already in the 16th century, Bruyerin, François I's physician, reported that the cultivation of the black truffle was possible. But, truffle cultivation really began in the 18th century, when Talon planted acorns and harvested truffles near young trees. The production of truffles increased to 1588 tonnes in France at the end of the 19th century (Chatin, 1869). However, after the first world war, production had decreased to reach the present-day value of less than 100 tonnes (Callot et al., 1999).

Research programmes for large-scale mycorrhizal production have been elaborated in southern European countries to increase truffle production (Chevalier, 1994). Their main steps are: production of mycorrhizal roots in controlled conditions, planting out the mycorrhizal seedlings, checking for the presence of introduced Tuber species among the ectomycorrhizal symbionts, and harvest of fruitbodies. Since the 1970s, it has been possible to obtain mycorrhizal seedlings; they produce T. melanosporum after 5–10 years (Le Tacon et al., 1988). At present, more than 80% of the French production of this truffle comes from artificial truffle-grounds (http://www.inra.fr/presse/la_truffe_de_plus_en_plus_rare_et_chere_gare_aux_fraudes). On the other hand, utilization of inoculated seedlings has allowed production of truffles in countries where T. melanosporum is not naturally found, such as Israel, the USA and New Zealand (Ceruti et al., 2003; Hall et al., 2003). At present, T. melanosporum, T. borchii and T. aestivum are collected from artificial truffle-grounds. By contrast, production of T. magnatum in controlled conditions has been rarely reported (and it is not certain whether it depends on successful inoculation in the nursery or on infections established after planting).

Many attempts have been made to isolate truffle mycelia to produce adequate amounts of inocula, but without success because of their slow growth. As it is very difficult to obtain pure mycelial cultures of Tuber spp., truffle-infected plants are usually produced with spore inoculum. As a consequence, especially plants inoculated with T. magnatum spores sometimes became contaminated with unexpected Tuber spp. or other ectomycorrhizal fungi (Amicucci et al., 2001). However, only T. borchii mycelium has been produced in sufficient quantities for research purposes, while T. borchii mycorrhizae have been developed in vitro (Sisti et al., 1998). Pure cultures of other Tuber species have been obtained, opening up the possibility of extending the applications of truffle research (Iotti et al., 2002).

Notwithstanding the progress in truffle research, many questions are still fully open as far as truffles in truffle-grounds are concerned: (i) How abundant are truffle mycorrhizae in a natural truffle-ground? (ii) Can we detect mycorrhizae exclusively in productive zones, or also in nonproductive ones? (iii) What other fungi are present in a truffle-ground? These questions stimulated researchers to investigate the distribution of the symbiotic phase of T. magnatum in a selected truffle-ground where the production of fruiting bodies had been followed for as long as 5 years (Mello et al., 2005). This attempt was essential to know the ecological behaviour of this fungus which cannot be obtained either as a pure culture or from artificial truffle-grounds.

The simplest method to investigate a fungal species during its symbiotic phase is the morphological typing of mycorrhizae in nature. This method, however, is dependent on many factors, such as age, host tree species and environmental conditions. For this reason, morphological typing of ectomycorrhizae is currently supported by molecular methods. In the screening of mycorrhizal tips in the selected truffle-ground, T. magnatum mycorrhizae were found to be very rare: it seems that this fungus invests more in fruiting body formation than in colonization of the roots (Murat et al., 2005). Similar results have also been obtained in another T. magnatum area in central Italy (Bertini et al., 2006). Furthermore, T. magnatum mycorrhizae were present in a nonproductive period for T. magnatum, and in a nonproductive area of the truffle-ground, indicating that there is not a direct linkage between mycorrhizae and fruiting bodies (Murat et al., 2005). All these observations raise questions about the functional role of the truffle's symbiotic phase, suggesting that truffles may be more plastic in their metabolism than expected. They seem to move among different nutritional strategies (saprobic, endophytic and symbiotic) depending on the environment and on the developmental phase of their life cycle. On this issue, a recent report shows that some Tuber species can also be endophytic inside roots of chlorophyllous and achlorophyllous plants such as orchids, suggesting a possible role as a bridge between mycoheterotrohic species and ectomycorrhizal trees (Selosse et al., 2004).

As truffle fruiting bodies are hypogeous it is likely that soil micro-fauna and microorganisms could enhance or inhibit truffle formation (Callot et al., 1999). Many studies have shown that numerous soil microorganisms interact with fungi promoting antagonistic, competitive or synergistic activities (Garbaye & Bowen, 1989; Frey-Klett et al., 1999).

Only a few studies have focused on the fungal biodiversity in truffle-grounds. Luppi-Mosca (1973) identified some fungi which seem common to the truffle environment. In a study of three T. aestivum Italian truffle-grounds, Zacchi et al. (2003) isolated several yeast species, among which, Cryptococcus strains appear to be specific to this habitat. From the activities shown, Cryptococcus humicolus, present on the surface of mature truffles, may contribute to Tuber nutrition during the saprotrophic stage or facilitate fungal ascospore dispersion. Furthermore, Buzzini et al. (2005) found that yeast isolates from T. magnatum and T. melanosporum ascocarps produced some molecules characteristic of the complex aroma of truffles. This suggested that yeasts have a complementary role in contributing to the final Tuber aroma. From these results, it is clear that this role needs to be explored more deeply. Murat et al. (2005), using both morphological and molecular approaches, found that Thelephoraceae, Pezizales and Sebacinaceae were the dominant fungal taxa in the subterranean ECM community in a T. magnatum truffle-ground. While this study provided a snapshot of the fungal community, more studies are surely needed to completely describe both the fungal biodiversity and the interactions with truffles.

In order to characterize the bacterial populations of mycorrhizal fruiting bodies, many studies based on cultivation methods have been performed, leading to the conclusion that Pseudomonas, aerobic spore-forming bacteria and actinomycetes are the most frequently isolated organisms within ascoma or ectomycorrhiza of Tuber spp. (Bedini et al., 1999; Sbrana et al., 2002). Gazzanelli et al. (1999) also assigned to these organisms a potential role in the symbiosis, assuming that certain pseudomonads and bacilli, which express a clear chitinolitic activity, are able to affect ascospore germination within fruiting bodies of T. borchii facilitating ascus opening. Furthermore, the finding for Cytophaga–Flexibacter 16S rRNA gene sequences in T. borchii ascocarps, mycelium and ectomycorrhizae, provides the first evidence that this bacterium could be involved in the entire life cycle of T. borchii (Barbieri et al., 2005).

In conclusion, we can be sure that the microbial biodiversity of truffle-grounds affects their productivity but, at the same time, we are far from knowing the mechanisms involved in this activity. In times of ‘metagenomics’, which is the recovery and analysis by sequencing of the collective genomes of microorganisms in an environment or niche (Daniel, 2005), it will be necessary to compare soil samples coming from areas of different productivity and to highlight differences in the presence of microorganisms.

Conclusions

The increasing attention towards the highly appreciated and commercialized hypogeous ascocarps has led to the point that publications on truffles exceed 1500, so far (Ceruti et al., 2003). The advent of molecular biology techniques as an addition to classical morphology has allowed the development of probes able to identify many truffle species along their life cycle. A website, created and maintained by the Italian Tuber scientific network (http://www.truffle.org), hosts detailed information on various Tuber species together with molecular tools for their identification.

Thanks to polymorphic sites, a reconstruction of the past history of T. melanosporum and T. magnatum has been traced. Their postglacial recolonization from the refuges, resembling that of the host plants, the oaks, is probably the reason for their present geographic distribution.

At the moment we have only patchy information on truffles, as we are not able to answer general questions, such as: how a fruiting body develops from a hyphal net. As truffle fruiting bodies cannot yet be obtained under controlled conditions, our knowledge of the morphogenetic events leading to ascocarp development, is quite limited even if many data are already available (Lacourt et al., 2002; Zeppa et al., 2002). Regarding this, Gabella et al. (2005) clearly demonstrated the presence of an active metabolism in T. borchii fruiting bodies.

The life cycle of truffles is still obscure and the reason for this can be traced to many factors. On the one hand, absence of an experimental system based on spore germination has never allowed the classical breeding of the resulting mycelia. On the other hand, mating-type genes, which determine heterothallism and pseudohomothallism, have never been reported for truffles, unlike for other filamentous Ascomycetes. The T. melanosporum genome sequencing recently launched by Francis Martin (France, pers. commun.) will surely help in decoding the mystery of this fascinating product of nature and enigma for science.

Acknowledgements

We wish to thank F. Martin (INRA) and S. Ottonello (University of Parma) for their long standing collaboration in common truffle projects as well as many Italian researchers involved in the Strategic Program ‘Biotecnologia dei funghi eduli ectomicorrizici’ funded by the National Council of Research. Our research was supported by ‘Commessa Biodiversità’-CNR and CEBIOVEM (DM 17/10/2003, n 193/2003) to P.B.