These authors contributed equally to this work.
Tuber aestivum and Tuber uncinatum: two morphotypes or two species?
Version of Record online: 9 JAN 2006
FEMS Microbiology Letters
Volume 235, Issue 1, pages 109–115, June 2004
How to Cite
Paolocci, F., Rubini, A., Riccioni, C., Topini, F. and Arcioni, S. (2004), Tuber aestivum and Tuber uncinatum: two morphotypes or two species?. FEMS Microbiology Letters, 235: 109–115. doi: 10.1111/j.1574-6968.2004.tb09574.x
- Issue online: 9 JAN 2006
- Version of Record online: 9 JAN 2006
- Received 22 January 2004, Revised 12 March 2004, Accepted 8 April 2004
- Tuber spp.;
- Gene genealogy;
- EF 1-α
Tuber spp. are ectomycorrhizal fungi that establish symbioses with shrubs and trees. Because of their different smell and taste, Tuber uncinatum and Tuber aestivum are two truffle morphotypes with a different market value, but whether or not T. uncinatum and T. aestivum are different taxa is still an open debate among mycologists. In order to identify molecular keys characterizing both T. aestivum and T. uncinatum morphotypes, ITS/RFLPs analyses were carried out on a large collection of samples from all over Italy and from other European countries, followed by a study of the phylogenesis of ITS, β-tubulin and EF 1-α genes, on representative samples. The present study provides compelling evidence that: (i) T. uncinatum and T. aestivum belong to the same species, (ii) neither morphotype presents a specific molecular fingerprint, but they may even share identical alleles at any of the loci analysed; (iii) T. aestivum is most likely under a selfing reproductive mode. Our findings suggest that ecological, rather than genetic causes may account for differences in sporal morphology, taste and smell between T. aestivum and T. uncinatum truffles.
Tuber is a major genus of ectomycorrhizal fungi within Pezizales producing hypogeous fruit bodies, known as truffles. Some species belonging to this genus are economically important because they produce edible truffles with a distinctive flavour. Ascocarps of Tuber species are generally distinguished by their phenotypic characteristics, mainly based on morphological features of asci and spores. Fruit bodies belonging to Tuber aestivum Vittad., Tuber uncinatum Chatin or to the black Asian truffle species Tuber indicum Cooke and Massee and Tuber himalayense Zhang and Minter, generally display a large set of sporal configurations associated with quite pronounced molecular variability [1–5]. The presence in these Tuber spp., of strains exhibiting a continuum of morphological traits between two taxa, often makes it difficult to differentiate between varieties and species. Therefore, whether T. uncinatum is a bona fide species or only a subspecies of T. aestivum is still an open debate among mycologists [1,6–9]. Species differentiation is not merely a taxonomical issue, but has profound ecological and economical implications, such as marketing of both fruit bodies and nursery-inoculated host plants, which are largely used in reforestation programs for growing truffle species.
The present study focuses on extensive molecular analyses of fruit bodies which, on the basis of preliminary macro- and microscopic observations, had been ascribed to either T. aestivum or T. uncinatum. The aim was to verify whether phylogenetically important markers can allow species/strain discrimination between these two morphotypes and test T. aestivum/T. uncinatum samples for the presence of clonal or recombining individuals.
2Material and methods
2.1Sample sources and DNA isolation
Morphological analyses were performed on fresh fruit bodies, by using distilled water as mounting media for microscopic analysis. T. aestivum and T. uncinatum characterization was based on a number of parameters previously reported [10,11]. In Table 1 are listed all the truffles examined, their collection sites and time. Genomic DNA was isolated from ascocarps basically according to Paolocci et al. .
|Sample name||RFLP pattern||Haplotype||Collection site||Country||Collection date||Morphotype|
2.2PCR amplification, RFLP and sequence analysis of ITS
The universal ITS1/ITS4 primers were derived from White et al. . PCR amplification was carried out as reported by Paolocci et al. , but with the addition in the PCR mixture of 10% (v/v) of DMSO. ITS was amplified from a total of 61 T. aestivum/T. uncinatum ascocarps.
About 30 ng of each PCR product was restricted using the endonucleases MspI, MseI and TaqI (Biolabs) according to the suppliers' instructions. The restricted fragments were separated on a 2% agarose gel. Double-strand sequence analyses were carried out as previously described .
2.3Amplification and sequencing of the β-tubulin and EF 1-α genes
The forward primer Bt2a, derived from Glass and Donaldson  and the reverse primer Btspect (5′-GTCGGGGAACTCTTCACGGATCTTRGAG-3′) designed by aligning truffle β-tubulin genes cloned in our laboratory (unpublished results), were used to amplify a homologous β-tubulin gene fragment on the T. aestivum/T. uncinatum samples.
The primers EFtFw: 5′-TACAA(A/G)TGCGGTGGTAT(T/C)GACAAGCG-3′ and EFtBw 5′-GACGCATGTCACGGACGGCGAAACG-3′ were designed on the conserved regions of the EF 1-α gene in Ascomycetes.
PCR amplification was carried out in GeneAmp PCR system 9700 using 10–20 ng of target DNA. The final aqueous volume of 50 μl contained 10 mM Tris–HCl, pH 9.0, 50 mM KCl, 1.5 mM MgCl2, 200 μM of each dNTP, 10 pmol of each primer and 2.5 U Taq Polymerase (Amersham Pharmacia Biotech). The cycling protocol was as follows: an initial denaturation at 94 °C for 90 s, followed by 25 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s and extension at 72 °C for 45 s (β-tubulin) or for 1 min (EF 1-α) for each cycle; the last cycle was followed by a final extension at 72 °C for 7 min. The β-tubulin and EF 1-α genes were amplified from 18 fruit bodies considered in the ITS sequencing analysis. In addition, for phylogenetic purposes, the same genes were also amplified from Tuber mesentericum, Tuber melanosporum and Tuber magnatum.
2.4Sequence alignments and statistical analyses
The ITS, β-tubulin and EF 1-α sequence alignments were obtained using ClustalX software .
Sequence statistics, nucleotide diversity and distance-based analyses were performed using MEGA version 2.1 . Phylogenetic trees were obtained using the Neighbor-joining method based on the two-parameter distance model of Kimura . All gaps were excluded from phylogenetic analyses by selecting the “complete deletion” option in MEGA. Bootstrap tests were performed using 1000 replicates.
Congruence for the gene genealogies was inferred from maximum parsimony analysis (MP) by applying the permutation tail probability test (PTP)  and the incongruence length difference/partition homogeneity test (ILD/PHT)  as implemented in PAUP 4.0b10 . Only parsimony informative characters were considered, and gaps were treated as new states. To give each locus equal weight, the informative characters (4, 2, 22 for β-tubulin, EF 1-α and ITS, respectively) were weighted reciprocally as reported by Kroken and Taylor .
3.1Morphological analysis of ascocarps
Differentiation between T. aestivum and T. uncinatum was achieved by morphological observation [10,11]. In agreement with Gross , those of the T. aestivum/T. uncinatum samples displaying an average 2 μm high sporal reticulum were ascribed to the T. aestivum morphotype, whereas those with a sporal reticulum averaging about 4 μm in height were ascribed to T. uncinatum (Table 1). However, several fruit bodies showed an intermediate sporal reticulum height. They are marked in Table 1 with (b) to indicate their uncertain classification.
3.2Amplification and RFLP analysis of the ITS
Positive and highly reproducible PCR amplification of DNA from all the specimens examined was achieved by adding 10% (v/v) DMSO to the reaction mixture, suggesting the presence of a G/C rich region in the ITS of, at least, some truffles. Length polymorphism was detected between T. aestivum/T. uncinatum and T. mesentericum ITS fragments, which were about 700 and 685 bp long, respectively.
RFLP analysis of the ITS was initially performed using the enzyme MspI, by which Guillemaud et al.  and Mello et al.  were able to identify three different RFLPs classes within their collections. Conversely, within our 61 samples this endonuclease revealed only two of the three MspI patterns reported in the literature (Table 1, Fig 1(a)). The majority (85.2%) of T. aestivum/T. uncinatum ascocarps showed pattern 2 (MspI-2), whereas pattern 1 (MspI-1) was observed in the remaining 14.8%, most of which collected from the same region. Moreover, truffles showing the MspI-2 pattern displayed a further polymorphism when RFLP analysis was performed with the enzymes TaqI and MseI (Figs. 1(b) and (c)). The TaqI restriction enzyme revealed two different patterns in the T. aestivum/T. uncinatum samples, accounting for 31.1% and 68.9%, respectively; similarly, MseI RFLP analysis revealed two patterns in the ratio 26.2–73.8%. The combination of three restriction enzymes therefore allowed us to discover more ITS/RFLP classes than had been identified previously [3,9]. Overall, this analysis revealed the presence of at least four different ITS classes, or haplotypes, named A, B, C and D, within our T. aestivum/T. uncinatum samples (Table 1 and Fig. 1), with D being the most frequent ITS haplotype (68.9%). Considering the additional MspI pattern identified in previous studies [3,9], all the ITS T. aestivum/T. uncinatum genetic variability exhibited by truffles from different European collection sites can be grouped into five RFLP classes. Relationships between the RFLP classes and either morphological classification or truffle harvesting time and collection site did not emerge.
3.3ITS sequence analysis
ITS sequencing was carried out on truffles selected from within the four RFLP classes according to the following criteria: morphotype, harvest time and collection sites (Table 1). For example, sequences were performed on two truffles, E61 and E32, which showed the same RFLP pattern, but were ascribed one to T. aestivum and the other to T. uncinatum based on morphological criteria and ripening times.
ITS sequences obtained from T. aestivum/T. uncinatum were aligned with those of T. mesentericum, T. magnatum, T. melanosporum and T. indicum, selected as reference species for phylogenetic analyses. The Neighbor-joining tree (Fig. 2(a)) showed that all the brown truffles , grouping T. aestivum/T. uncinatum and T. mesentericum fruit bodies, were well separated from the black truffles T. melanosporum and T. indicum, but had the most similarity with T. magnatum. Within the T. aestivum, T. uncinatum and T. mesentericum spp., the phylogenetic tree showed two major clusters corresponding to the T. aestivum/T. uncinatum and T. mesentericum complexes, respectively. Despite the high number of samples, the T. aestivum cluster was quite uniform (average nucleotide diversity=0.75%). The ITS tree topology supported the classification of all the ITS haplotypes revealed by RFLP analysis, while confirming the impossibility of splitting the T. aestivum and T. uncinatum samples into two separate groups.
3.4Amplification and sequence analysis of the β-tubulin and EF 1-α genes
In order to better assess the degree of heterogeneity among and within the Tuber spp. considered in this study, a genomic fragment relative to β-tubulin and EF 1-α genes was PCR amplified on the 18 samples previously selected for ITS sequence analysis as well as on T. mesentericum, T. magnatum and T. melanosporum. Direct sequence analysis confirmed that only one amplicon per individual was obtained without sequence ambiguities. The identity of amplicons was confirmed by database search. The alignment of truffle β-tubulin sequences showed four different β-tubulin alleles within T. aestivum/T. uncinatum. Fig. 2(b) shows the Neighbor-joining tree derived from the alignment of the β-tubulin sequences considering both coding and non-coding regions. Overall, the T. aestivum/T. uncinatum cluster is quite homogeneous (average nucleotide diversity=0.27%). The high degree of nucleotide identity of the coding region of the EF 1-α genes from different filamentous Ascomycetes allowed us to design the primer pair EFtFw/EFtBw and successfully amplify a fragment of this target gene in truffle species. All samples showed a 1.3-kbp long amplicon, with a protein-coding region split into four exons, as evidenced by the alignment with genomic and cDNA sequences of different fungal species. In truffle species, introns are conserved with respect to both number and position. Conversely, Tuber spp. showed a different intron organization from other filamentous fungi (data not shown). The T. aestivum/T. uncinatum sequences showed an extremely high conservation rate (average nucleotide diversity=0.17%) and provided us with only two parsimony informative characters for phylogenetic concordance analysis (Fig. 2(c)).
In sum, all the trees relative to the loci analysed showed that: (1) brown truffles are phylogenetically closer to the white truffle T. magnatum than to the black truffle T. melanosporum; (2) T. mesentericum is the most likely T. aestivum sister taxon and (3) identical alleles are shared by samples exhibiting the T. aestivum or the T. uncinatum morphotype.
The sequencing of the three loci considered in this study on 18 T. aestivum samples provides us with a number of parsimony informative characters to test for congruence of gene genealogies.
We assessed congruence of gene genealogies using PTP and ILD/PHT tests. MP analysis of the 10,000 randomized data sets produced by the PTP test resulted in trees ranging from 268 to 360 steps. The observed tree length (132 steps) fell outside the range of tree lengths for the randomized data sets (P=0.0001), allowing us to reject the null hypothesis of panmixia. On the other hand, all the trees resulting from MP analysis of the 10,000 randomized data sets produced by the ILD/PHT, were 132 step long like the sum of the original tree length for each locus, allowing us to accept the null hypothesis of clonality (P=1.00).
The results of both statistical tests and the finding that no individual was found to be heterozygous at the loci analysed seem to corroborate the hypothesis of a strict self-fertilizing propagation system in T. aestivum, as previously reported on the basis of allozyme analyses [2,7]. The present data are however not sufficient to completely rule out the hypotheses of a long-term strict inbreeding and pseudohomothallism, since both reproductive modes could not generate random associations of alleles within and between loci .
The complexity of the truffle life cycle combined with the difficulties to grow these fungi and the impossibility to mate them in controlled conditions, have been major obstacles to the understanding of the mating and of the propagation system of these hypogeous fungi. Hence, the species concepts are overwhelmingly phenotypic, but caution must be used when grouping truffles into species or strains/varieties based only on fruit body morphology alone, as most of the widely distributed truffle species include ascocarps with a high degree of polymorphism and divergent morphological traits. Only molecular-assisted screening of a large fruit body collection for each species can provide evidence about the reproductive system and species delimitation in these fungi. In the present study, therefore, capillary morphological and molecular analyses were carried out on T. aestivum and T. uncinatum, two widely distributed truffle morphotypes, to assess their genetic relationship. This is the first report on Tuber spp., where the sequences of the rDNA spacer regions have been contrasted with other genes, specifically the β-tubulin and EF 1-α genes, to identify population variation at the DNA sequence level. Recently, Mello et al. , differently from other authors [1,6–8], claimed successful differentiation of T. uncinatum from T. aestivum by means of ITS and ISSR analyses. For a deeper understanding of the genetic variability within these taxa, more than 60 samples from Italy, Istria and France, morphologically ascribed to either T. aestivum or T. uncinatum, were initially analysed by means of ITS/RFLPs. Although the RFLPs are co-dominant markers, they showed only one ITS allele on each of the truffles analysed, in accordance with Guillemaud et al. . None of the ITS classes correlates either with the phenotypic ascocarp traits (e.g., the sporal reticulum height) or with the truffle ripening time. Also, no correlation between ITS polymorphism and ascocarp origin was found. Taken together, these data lend further support to the thesis that molecular-assisted analyses do not corroborate species distinction between T. aestivum and T. uncinatum. Differently from Mello et al. , our large-scale sample screening allowed us to detect more than a single ITS haplotype within T. uncinatum morphotypes, and, most importantly, T. aestivum samples sharing the same ITS haplotypes with T. uncinatum (Table 1). Therefore, the ITS locus is unlikely to be a diagnostic marker for differentiating T. aestivum from T. uncinatum. In addition, the overall genetic dissimilarity existing within the ITS of T. aestivum/T. uncinatum samples, as per this study, can be confidently confined within intraspecific variability when comparing the sequences of T. aestivum and T. mesentericum. Despite the fact that the latter species is a T. aestivum sister species, and some authors  even regard it as a T. aestivum variety, two clearly divergent clusters split the T. aestivum/T. uncinatum and the T. mesentericum complex apart. When the recently published T. uncinatum ITS sequences  were added to our ITS phylogenetic tree, they were found to be dispersed among the T. aestivum B, C and D classes (data not shown).
The same conclusion can be drawn by analysing the β-tubulin and EF 1-α phylogenetic trees (Fig. 2). Likewise, these analyses not only showed that T. uncinatum and T. aestivum are monophyletic, but also that they can share identical β-tubulin and 1-α elongation factor alleles.
Although multiple alleles for each of the three loci considered were revealed, no individual was found to be heterozygous at any locus. Furthermore, PTP and ILD/MH statistical analyses proved congruence among ITS, β-tubulin and EF 1-α gene genealogies. Molecular approaches different from gene genealogies, such as allozyme in T. aestivum[2,7] and microsatellites in T. melanosporum prompted the authors to reject the hypothesis of heterothallism and put forward a homothallic or even an exclusive selfing propagation system in truffles. All the data herein reported point to a very close mating system in T. aestivum. Yet, more data and analyses are needed to assess whether no random associations of alleles within and between loci result indeed from primary homothallism, pseudohomothallism or even from a strict inbreeding on a long-term scale [24,27].
In conclusion, the data obtained with the three co-dominant markers used in this study strongly argue against the thesis that T. uncinatum and T. aestivum cluster into two separate groups . On the contrary, these data lead us to suggest that specific, though still unclear, soil and climatic conditions may induce T. aestivum to fruit under different conditions, which, in turn, could affect the flavour and aroma of the truffle and the sporal morphology of ascocarps, thereby giving rise to the T. aestivum and T. uncinatum morphotypes. Should this thesis be proved correct, there would be no point in continuing to use distinct marketing approaches for T. aestivum and T. uncinatum inoculated plants. This study paves the way for innovative and more ecology-oriented research approaches to determine whether environmental conditions alone may be sufficient to induce a given mycelial strain to fruit as either T. aestivum or T. uncinatum, since single truffle grounds or even single host plants producing both morphotypes were reported .
The cloning of more genomic loci and their analysis using a larger sample from all over Europe and specimens from truffle grounds producing both T. aestivum and T. uncinatum morphotypes, are now in progress in our laboratory. The expected results might further strengthen the hypotheses of a strict selfing propagation system in T. aestivum, and of an ecological pressure as the leading force that shapes T. aestivum plasticity.
The research was funded by the Italian National Research Council, Progetto Strategico “Biotecnologie dei funghi ectomicorrizici: dalle applicazioni agro-forestali a quelle agro-alimentari”. Contribution no. 45 from the Institute of Plant Genetics, Division of Perugia.
- 1991) Genotypic patterns of some Italian populations of the Tuber aestivum–T. mesentericum complex. Mycotaxon 42, 171–179., (
- Rubini, A., Paolocci, F., Arcioni, S. and Granetti, B. (2001) Analisi morfologica comparativa delle spore e caratterizzazione molecolare di Tuber indicum Cooke e Massee e Tuber melanosporum Vitt. In: Actes du Ve Congrès International “Science et culture de la truffe et des autres champignons hypogés comestibles”. pp 94–101. Aix en Provence, France.
- 1993) Genetics of a Tuber aestivum population. Mycotaxon 7, 93–100., , , (
- 1978) L'espèce Tuber aestivum Vitt. I. Definition. Mushroom Sci. 10, 957–975., , , (
- Gross, G. Zu den europäischen Sippen der Gattung Tuber. In: Atlas der Pilze des Saarlandes, Teil 2: Nachweise, Ökologie, Vorkommen und Beschreibungen (Derbsch, H., Schmitt, J.A., Eds.), Vol. 3, (1987) 79–99. Verlag der Delattinia, Saarbrücken.
- White, T.J., Bruns, T., Lee, S., Taylor, J. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: PCR Protocols. A Guide to Methods and Applications (Gelfand, M.A., Sninski, D.H., White, T.J., Eds.), pp.315–322 Academic Press, San Diego, CA.
- Swofford, D.L. (2002) Paup. Phylogenetic Analysis Using Parsimony (and other method), Version 4. Sinauer Associates, Sunderland, MA.
- 1979) The order, families, and genera of hypogeous Ascomycotina (Truffles and their relatives). Mycotaxon 9, 297–340.(