Environmental Microbiology (2002)

4

(10), 584–594

© 2002 Blackwell Science Ltd

Blackwell Science, LtdOxford, UKEMIEnvironmental Microbiology 1462-2912Blackwell Science, 20024Original Article

A. Mello, A. Cantisani, A. Vizzini and P. BonfanteVariability in black truffles

Received 18 July, 2002; accepted 18 July, 2002. *For correspon-dence. E-mail p.bonfante@csmt.to.cnr.it; Tel. (

+

39) 0116502927; Fax(

+

39) 0116707459.

Genetic variability of

Tuber uncinatum

and its relatedness to other black truffles

Antonietta Mello,

1

Annamaria Cantisani,

1

Alfredo Vizzini

2

and Paola Bonfante

1,2*

1

Istituto per la Protezione delle Piante del CNR, Sezione di Torino and

2

Dipartimento di Biologia vegetale dell’Università di Torino, Viale Mattioli, 25, 10125-Torino, Italy.

Summary

Genetic variability is one of the major survival strate-gies developed by symbiotic fungi. We focused on theectomycorrhizal fungus

Tuber uncinatum

Chatin thatproduces edible ascomata. In order to understand thedegree of its variability and its relatedness to anothermorphologically-similar truffle,

T. aestivum

Vittad.,ascomata of

T. uncinatum

were collected from a sin-gle natural truffle-ground located in the north of Italyand compared with samples from other Italian sites,as well as with

T. aestivum

ascomata from otherEuropean regions.

We used multi-locus approaches, such asmicrosatellite-primed PCR (polymerase chain reac-tion), and single locus markers, such as mitochon-drial and nuclear ribosomal DNA on 30 samples. Theresults demonstrate that the level of genetic polymor-phism among isolates of

T. uncinatum

was higherthan in other

Tuber

species, like

T. melanosporum

.Neighbour-joining analyses were carried out on abinary data matrix on 12 ascomata of

T. uncinatum

and

T. aestivum,

and on 15 internal transcribed spacer(ITS) sequences of these species and 5 from other

Tuber

species. Taken together, they clustered

T. unc-inatum

and

T. aestivum

in two separate groups. Themitochondrial rDNA primers, NMS1 and NMS2, werenot able to differentiate morphologically related andunrelated truffles. Moreover, a pair of primers, inten-tionally designed to differentiate isolates of

T. aesti-vum

and

T. uncinatum

from other

Tuber

species,successfully amplified DNA from all the samples of

T.aestivum

and

T. uncinatum

considered in our analy-sis. In conclusion, different molecular approaches

separate

T. aestivum

and

T. uncinatum

according totheir spore reticulum and their taste and smell.

Introduction

Many of the most important functions of terrestrial eco-systems take place below ground as a product of theinteractions between roots of vascular plants and mycor-rhizal fungi: the resulting associations are in fact the keyplayers in soil ecology (Dahlberg, 2001). In temperate andboreal forests, up to 95% of the short roots form ectomy-corrhizas. These have a beneficial impact on plant growthin natural (Read, 1991) and agroforestry ecosystems(Grove and Le Tacon, 1993). The success of these sym-bioses is mostly a result of the exchange of nutrientsbetween the symbionts (Smith and Read, 1997): the fun-gus gains carbon from the plant while plant nutrientuptake is mediated via the fungus. The establishment ofthe symbiosis allows some fungi to accomplish their lifecycle, i.e. the development of fruiting bodies. Our interestfocuses on some fungi that not only form ectomycorrhizaswith the roots of some higher plants and shrubs but alsohypogeous fruiting bodies, commonly known as truffles(Pegler

et al

., 1993).Truffles belong to the genus

Tuber

, which was for a longtime considered as belonging to the order Tuberales, ataxon including most species of Ascomycetes hypogeousfungi (Dennis, 1981). However, the most recent phyloge-netic trees, based on rDNA sequences, locate this genusinside the order Pezizales, together with Helvellaceae(O’Donnel

et al

., 1997; Percudani

et al

., 1999).The advent of molecular techniques has prompted

investigations on different aspects of truffle biology: first ithas been possible to discriminate among morphologicallysimilar species. For example, specific primers have beendeveloped to discriminate

Tuber borchii

from

T. macula-tum

(Amicucci

et al

., 1998; Mello

et al

., 2000) and

T. mel-anosporum

from

T. brumale

and

T. indicum

(Rubini

et al

.,1998; Paolocci

et al

., 2000).Similar methods allowed us to identify the fungus in the

different steps of its life cycle providing diagnostic tools toconfirm the occurrence of the desired fungus in the myc-orrhizal roots (Stocchi, 1999; Mello

et al

., 2001; Rubini

et al

., 2001). In addition to the studies investigating themechanisms which control truffle development (amongthe others Soragni

et al

., 2001; Polidori

et al

., 2002),

Variability in black truffles

585

© 2002 Blackwell Science Ltd,

Environmental Microbiology

,

4

, 584–594

increasing attention is currently being given to trufflegenetic variability (Bertault

et al

., 1998; 2001). This topicis considered one of the biological keys to understand thepopulation dynamics of fungi, their erratic pattern of dis-tribution and their life cycle. In this regard, the analysis ofthe genetic variability of the ectomycorrhizal

Pisolithus

species has recently shown that evolutionary lineageswithin

Pisolithus

are related to the biogeographical originof the hosts (Martin

et al

., 2002).As experimental material for a genetic analysis of vari-

ability in truffles, we selected

T. uncinatum

Chatin, whosetaxonomic position is controversial. This taxon is morpho-logically related to

T. aestivum

Vittad. The length of theirspore reticulum is considered the most useful morpholog-ical characteristic for distinguishing the two taxa.

As the molecular and biochemical attempts to discrim-inate the two species failed (Pacioni

et al

., 1993; Henrion

et al

., 1994; Gandeboeuf

et al

., 1994; 1997; Urbanelli

et al

., 1998), Chevalier

et al

. (1994) and Riousset

et al

.(2001) suggested that the nomenclature should bechanged. They questioned whether

T. uncinatum

and

T.aestivum

should be considered as two separate speciesor as varieties belonging to the same species. In a recentbook on the hypogeous European fungi, Montecchi andSarasini (2000) proposed two morphological descriptions,the first for

T. aestivum

Vittad., the second one for

T.aestivum

forma

uncinatum

.However, some ecological features, geographical distri-

butions and smell and taste are distinctive of the twospecies, and are carefully described in Chevalier andFrochot (1997).

Tuber uncinatum

is considered to have agreater economic value than

T. aestivum

and projects arebeing developed for the production of mycorrhizal plants,similar to those already developed in France for

T. mel-anosporum

(Chevalier and Grent, 1979).In order to gain more insight into the genetic variability

of

T. uncinatum

we analysed ascomata and mycorrhizasfrom a natural truffle-ground located in Lombardy(Menconico, PV, Northern Italy), as well as ascomata fromother Italian regions. To better investigate the relationshipsof

T. uncinatum

with

T. aestivum

, we also consideredsamples morphologically identified as

T. aestivum,

fromItalian and other European regions.

We selected different genetic markers according to therecommendations by Sunnucks (2000). In particular, weused multi-locus approaches, such as microsatellite-primed PCR, and single locus markers, such as mitochon-drial and nuclear ribosomal DNA.

Results

All the ascomata collected were morphologically charac-terized as

T. uncinatum

or

T. aestivum

depending on thespore reticulum. According to Montecchi and Sarasini

2000) and Riousset

et al

. (2001) when the reticulum of thespores was

<

4

m

m deep, the samples were considered as

T. aestivum

, whereas when it was

>

4

m

m deep the sam-ples were considered as

T. uncinatum

. Whenever thischaracter was intermediate, the samples were classifiedas ‘

aestivum

group’.Thirty ascomata were analysed: 16 ascomata of

T. uncinatum

from the Menconico truffle-ground, four of

T. uncinatum

, five of

T. aestivum

and five belonging to‘

aestivum

group’ were collected from different places inItaly and from France and Spain (Table 1). Two samplesof mycorrhizas obtained by inocula of the seedlings with

T. uncinatum

ascomata were collected in the Menconicotruffle-ground.

Microsatellite analysis

Amplification with the primers (GTG)5, (GAC)5, (AAG)8,(AAC)8 of the inter-repeat sequences between microsat-ellite sequences from 30 ascomata led to the followingresults: (GTG)5 primer gave homogeneous patterns whilethe other three primers revealed consistent polymorphismamong the samples (data not shown). Only about 50% ofbands were shared fragments. On no account could

T.uncinatum

ascomata from different geographical originsbe discriminated.

A distance analysis was made from a binary data matrixon a group of 12 samples. The dendrogram obtained fromthe neighbour-joining (NJ) analysis showed that the threesamples of

T. aestivum

clustered together, as well as thenine samples of

T. uncinatum

(Fig. 1).

ITS analysis

The DNA from all the 30 ascomata of

T. uncinatum

and

T.aestivum

was amplified by the primers ITS1/ITS4 andproduced a band of about 700 bp, while mycorrhizal DNAamplified with ITS1F/ITS4, gave a band of 740 bp (datanot shown).

In order to understand whether the two taxa could bedifferentiated, we sequenced the ITS from 17 ascomataof

T. uncinatum

and

T. aestivum.

A phylogenetic treewas constructed from the alignment of the first 15 ITSsequences listed in Table 1 with those of one

T. melano-sporum,

one

T. borchii

, one

T. magnatum

and two

T.mesentericum

available in Genebank (Fig. 2).

Pithya vul-garis

was used as outgroup. As the forward and reversesequences of samples 443 (440 bp at one end, 180 bp atthe other end) and A4 (427 bp at one end, 181 bp at theother end) were not overlapping, and about 30 bp weremissing, that region was cut in all the samples in order toinclude also

T. aestivum

443 and

T. aestivum

A4 in thealignment (592 bp length).

586

A. Mello, A. Cantisani, A. Vizzini and P. Bonfante

© 2002 Blackwell Science Ltd,

Environmental Microbiology

,

4

, 584–594

Tuber aestivum

and

T. uncinatum

samples cluster in twogroups, supported by 98% and 81% bootstrap respec-tively. Shorter sequences from

T. aestivum

A17 and froma sample belonging to the

‘aestivum

group’ (N 225) werealso used, but in a different alignment where they clus-tered with the sequences of the other

T

.

aestivum

sam-ples (data not shown)

.

Once the two taxa were separatedfrom the molecular point of view, we moved to investigate

the intraspecific variability of

T. uncinatum. First weconsidered ITS sequences from nine ascomata fromMenconico and we found a similarity of 95.00% (608conserved positions out of 640 alignment length) that didnot change, when the sequences from two T. uncinatummycorrhizal samples, coming from the same area (simi-larity of 94.84%), were added (Table 2). For comparison,a search of Genebank found seven complete ITS

Table 1. Samples examined, sites from which they were collected, ITS, mrDNA and 16S rDNA accession numbers.

Species Site ITS mrDNA 16S rDNA

1. T. uncinatum 228 Menconico (Italy) AJ4921992. T. uncinatum 397 b Menconico (Italy) AJ4922003. T. uncinatum 397 c Menconico (Italy) AJ492201 AJ493668 AJ4936724. T. uncinatum 398 Menconico (Italy)5. T. uncinatum 399 Menconico (Italy) AJ492202 AJ4936696. T. uncinatum 400 A Menconico (Italy)7. T. uncinatum 400 B Menconico (Italy) AJ4936738. T. uncinatum 403 Menconico (Italy) AJ4922039. T. uncinatum 404 Menconico (Italy) AJ492204

10. T. uncinatum 411 Menconico (Italy)11. T. uncinatum 412 Menconico (Italy)12. T. uncinatum 413 Menconico (Italy) AJ49220513. T. uncinatum 416 Menconico (Italy) AJ49220614. T. uncinatum 418 Menconico (Italy) AJ49220715. T. uncinatum 419 Menconico (Italy)16. T. uncinatum 420 Menconico (Italy)17. T. uncinatum 224 Unknown (Italy) AJ492208 AJ49367418. T. uncinatum Virg Asti (Italy)19. T. uncinatum 445 Campobasso (Italy) AJ49220920. T. uncinatum 100 Unknown (Italy) AJ492210 AJ49367021. T. aestivum 230 Unknown (Italy) –22. T. aestivum 443 Campobasso (Italy) AJ492211, AJ492212 AJ49367123. T. aestivum A4 Richerenches (France) AJ492213, AJ49221424. T. aestivum A21 Richerenches (France) AJ49221525. T. aestivum A17 Richerenches (France) AJ49221726. ‘aestivum group’ 1 Lecce (Italy) –27. ‘aestivum group’ 2 Lecce (Italy) –28. ‘aestivum group’ 225 Teruel (Spain) AJ49221629. ‘aestivum group’ 442 Campobasso (Italy)30. ‘aestivum group’ 444 Campobasso (Italy)31. Mycorrhizal sample 1 Menconico (Italy) AJ49221832. Mycorrhizal sample 2 Menconico (Italy) AJ49287033. T. magnatum AJ002509a

34. T. borchii TBAJ2510a

35. T. melanosporum AF300827a

36. T. melanosporum AF300826a

37. T. melanosporum AF300825a

38. T. melanosporum AF132501a

39. T. melanosporum AF167097a

40. T. melanosporum AF167096a

41. T. melanosporum U89359a

42. T. indicum AF300822a

43. T. indicum AF300823a

44. T. indicum AF300824a

45. T. indicum U89360a

46. T. indicum U89361a

47. T. indicum U89362a

48. T. mesentericum AF106887a

49. T. mesentericum AF132508a

50. Pithya vulgaris U66008a

51. T. borchii mycelium AJ493676 AJ493675

a. Indicates sequences obtained from Genebank.– indicates sequences not used in alignments. Samples from 1 to 32 considered in our study. The first 30 samples are deposited at HerbariumOrti Botanici Taurinensis c/o Dipartimento di Biologia Vegetale, Università di Torino, as collection of Dr Alfredo Vizzini.

Variability in black truffles 587

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

sequences of T. melanosporum from different geographicareas and six of T. indicum, from China (Table 2). TheITS sequences of T. melanosporum demonstrated anintraspecific similarity of 97.84% whereas those of T. indi-cum of 86.9%.

The relatively high intraspecific degree of polymorphismshown by the group of T. uncinatum from a commonorigin, with respect to that of T. melanosporum, from adifferent origin, prompted us to compare the previoussequences of T. uncinatum with three of T. uncinatum fromdifferent origins (Table 1). The similarity, in this case, was94.08% (Table 2).

Although the two taxa clustered in two groups (Fig. 2)their ITS sequences were not so different as to allowthe design of primers specific for each group (not known).Therefore, a pair of primers to intentionally amplify DNAfrom all samples belonging to T. aestivum and T. uncina-tum was designed. The pair of primers was derived fromthe comparison of 18 ITS sequences (15 already used forthe previous alignment plus two from mycorrhizal samplesand one from sample N. 225) with those of other Tuberspecies. Amplifications with the new primers, UncI/UncII,gave the expected band of about 405 bp in all 30 asco-mata, irrespectively of their morphological identification,

and in the two mycorrhizal samples. They did not amplify14 Tuber species and Sphaerosporella brunnea, antago-nistic to Tuber species (Fig. 3A).

The restriction of the amplified 405 bp fragment from all30 ascomata and two mycorrhizal samples with MspIenzyme showed three different patterns, identified as pat-tern A, B and C (Fig. 3B). The most represented pattern,C, was present in 25 samples, including all T. uncinatum,one T. aestivum, two Tuber of the ‘aestivum group’ and twomycorrhizal samples (Table 1). In contrast, A was presentonly in six samples: three T. aestivum and three Tuber ofthe ‘aestivum group’; and B was only found in one T.aestivum (Table 1). In order to understand the molecularbasis of the allele distribution in all the samples, the ITSsequence of T. uncinatum 228 (genotype C), ‘aestivumgroup’ 225 (genotype A), and T. aestivum A21 (genotypeB), were selected for a detailed analysis.

The ITS sequence of T. uncinatum 228 shows acytosine/thymine substitution at a restriction site of MspI.This leads to the higher band of 249 bp in T. uncinatum228 (genotype C, lane 1) instead of 231 in sample N. 225(genotype A, lane 28) and in T. aestivum A21 (genotypeB, lane 25). The ITS sequence of T. aestivum A21 shows,at another restriction site, a guanine/thymine substitutionthat leads to the lower band of 155 bp (lane 25) instead of136 bp in T. uncinatum 228 (lane 1) and 137 bp in sampleN. 225 (lane 28). Consequently an 18 bp sequence notdetected in agarose gel was present in sample N 225.

Mitochondrial analysis

The DNAs from 30 ascomata were amplified with NMS1/NMS2 primers and gave a band of about 575 bp in all the

Fig. 1. Neighbour-joining tree obtained from microsatellite analysis on 12 samples of T. uncinatum and T. aestivum.

Table 2. Similarities obtained from the alignment of ITS sequences of T. uncinatum.

Samples Sites Similarity

T. uncinatum Menconico (Italy) 95.00%(9 samples, Table 1)

T. uncinatum 22 Unknown (Italy)T. uncinatum 445 Campobasso (Italy)T. uncinatum 100 Unknown (Italy) 94.08%

588 A. Mello, A. Cantisani, A. Vizzini and P. Bonfante

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

samples and a second band of about 620 bp in only threeT. uncinatum (N. 397c, N. 400b, N. 224) as shown inFig. 4A. To understand the identity of the additional bandpresent in the three samples the relative fragment wassequenced. Alignment of the three sequences gave 100%identity and the BLAST search showed significant align-ments with 16S rDNA sequences from Cytophaga-Flexibacter-Bacteroides (CFB) phylogroup. Our goal wasto understand whether the pair of primers was able todifferentiate the two taxa, as it did for various species ofVerticillium and Fusarium (Li et al., 1994). Common frag-ments from three T. uncinatum (N. 397c, N. 399, N. 100)and from one T. aestivum (N. 443) were sequenced(Fig. 4A, lanes: 3, 5, 20, 22; Table 1). While thesequences obtained from the first two samples were com-plete (575 bp), those coming from the other ones werepartial. Alignment of the four sequences by about 200 bpshowed a similarity of 99.20%. A similarity search in theBLAST database showed that the complete sequences

(575 bp) relative to mrDNA from the two T. uncinatum hadsignificant similarities with Aspergillus clavatus mrDNAbut no other Tuber mitochondrial sequences were presentin that database.

After digestion with the enzyme MseI of the bandof about 570 bp, the same restriction patterns wereobserved in all the analysed samples, irrespectively oftheir morphological identity (Fig. 4B).

As a further comparison, NMS1/NMS2 primers werechecked on DNA extracted from the mycelium of T.borchii, a morphologically different truffle species. Cloningand sequencing of the amplified products led to similarresults, revealing bacterial and fungal sequences (datanot shown). While the bacterial sequence showedsignificant alignment with sequence from Cytophaga-Flexibacter-Bacteroides (CFB) phylogroup, alignment ofthe fungal sequence showed a 99.45% identity with mito-chondrial sequences from the two T. uncinatum, N. 397cand N. 399. Taken together, the data suggest that NMS1/

Fig. 2. Neighbour-joining tree obtained from the alignment of the internal transcribed spacer (ITS) region. The sequence of Pithya vulgaris is added as outgroup. Percentage bootstrap support (out of 1000 trials) is shown when it exceeds 55%.

Variability in black truffles 589

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

NMS2 primers do not discriminate morphologically unre-lated truffle species.

Discussion

Population dynamics, as spatial distribution of geneticvariation in a population, has been investigated in someTuber species. The major conclusion reached is that a lowlevel of polymorphism occurs in T. melanosporum and T.magnatum (Longato and Bonfante, 1997; Bertault et al.,1998; Mello et al., 1999; Bertault et al., 2001). Our anal-ysis of the morphologically related T. uncinatum and T.aestivum ascomata led to some interesting and unex-

pected results: (i) the gene tree drawn on 15 partial ITSsequences clusters T. uncinatum and T. aestivum in twoseparate groups; (ii) the dendrogram obtained from themicrosatellite analysis of 12 ascomata showed that sam-ples of T. aestivum clustered together, as well as samplesof T. uncinatum; (iii) the alignment of ITS sequencesrevealed a relatively high degree of polymorphism in T.uncinatum; (iv) a pair of primers successful distinguishedsamples belonging to T. aestivum and T. uncinatum fromother Tuber species; (v) T. uncinatum presents only agenotype whereas T. aestivum samples show three gen-otypes; (vi) the mitochondrial primers NMS1/NMS2 do notallow any discrimination either within T. uncinatum and T.aestivum or between them.

Fig. 3. A. Amplification with UncI/UncII primers of the ITS from: (1) T. uncinatum 228, (2) T. uncinatum 397b, (3) T. uncinatum 397c, (4) T. uncinatum 398, (5) T. uncinatum 399, (6) T. uncinatum 100, (7) T. aestivum 230, (8) T. aestivum 443, (9) ‘aestivum group’ 1, (10) ‘aestivum group’ 225, (11) mycorrhizal sample 1, (12) mycorrhizal sample 2, (13) T. mesentericum, (14) T. melanosporum, (15) T. borchii, (16) T. magnatum, (17) T. ferrugineum, (18) T. maculatum, (19) T. indicum, (20) T. brumale, (21) T. himalayense, (22) T. excavatum, (23) T. macrosporum, (24) T. rufum, (25) T. panniferum, (26) T. foetidum, (27) Sphaerosporella brunnea and (M) pUC18 HaeIII digest. For simplicity, only the positive amplification from 10 out of the first 30 ascomata listed in Table 1 is given.B. RFLP from the UncI-UncII product from: (1) T. uncinatum 228, (2) T. uncinatum 397b, (3) T. uncinatum 397c, (4) T. uncinatum 398, (5) T. uncinatum 399, (6) T. uncinatum 400 A, (7) T. uncinatum 400B, (8) T. uncinatum 403, (9) T. uncinatum 404, (10) T. uncinatum 411, (11) T. uncinatum 412, (12) T. uncinatum 413, (13) T. uncinatum 416, (14) T. uncinatum 418, (15) T. uncinatum 419, (16) T. uncinatum 420, (17) T. uncinatum 224, (18) T. uncinatum Virg, (19) T. uncinatum 445, (20) T. uncinatum 100, (21) T. aestivum 230, (22) T. aestivum 443, (23) T. aestivum A4, (24) T. aestivum A17, (25) T. aestivum A21, (26) ‘aestivum group’1, (27) ‘aestivum group’ 2, (28) ‘aestivum group’ 225, (29) ‘aestivum group’ 442, (30) ‘aestivum group’ 444, (31) mycorrhizal sample 1, (32) mycorrhizal sample 2 and (M) pUC18 HaeIII digest. Genotypes A, B and C are indicated.

590 A. Mello, A. Cantisani, A. Vizzini and P. Bonfante

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

Multi-locus versus single-locus fingerprints

The most efficient genetic markers for population biologyhave been reviewed by Sunnucks (2000). According tothis paper, multi-locus approaches, such as RAPD, AFLP,Microsatellite-primed PCR, can visualize many anony-mous genome fragments simultaneously, but their funda-mental limitation is dominant inheritance, i.e. only oneallele is identified and DNA fragments can be scored onlyas present or absent. By contrast single-locus markers,such as nuclear and mitochondrial ribosomal genes, visu-alize co-dominant inheritance where, each of the two alle-les at a locus of an individual, can be identified.

Multi-locus and single-locus markers were successfullyamplified in truffle samples. ITS-based primers were themost convenient markers to address the problem ofgenetic variability in T. uncinatum and T. aestivum. How-ever, microsatellite analysis was also useful. In fact, ahigh polymorphism was found when primers based on

microsatellite complementary sequences were used. Thismolecular approach revealed that only about 50% of thebands were common among the 30 ascomata analysedsuggesting either that the loci corresponding to thesebands have different lengths in the samples or thatthe primers do not completely match complementarysequences. The variability was nicely correlated with themorphological differences, such as the length of the sporereticulum, as the dendrogram obtained from the microsat-ellite analysis showed that samples of T. aestivum clus-tered together, as well as samples of T. uncinatum.Mitochondrial rDNA is considered a suitable target fortyping inter- and intraspecific differences in fungi (Li et al.,1994; Selosse, 2001). However, in our experiments theNMS1/NMS2 primers amplified a common band for all 30ascomata. Restriction profiles and partial sequencesalignment did not show any variability, raising the questionwhether the primers were suitable for species typing in thegenus Tuber. The very high level of similarity (99.45%)

Fig. 4. A. Amplification with NMS1/NMS2 primers of the mitochondrial rDNA from the following ascomata: (1) T. uncinatum 228, (2) T. uncinatum 397b, (3) T. uncinatum 397c, (4) T. uncinatum 398, (5) T. uncinatum 399, (6) T. uncinatum 400 A, (7) T. uncinatum 400B, (8) T. uncinatum 403, (9) T. uncinatum 404, (10) T. uncinatum 411, (11) T. uncinatum 412, (12) T. uncinatum 413, (13) T. uncinatum 416, (14) T. uncinatum 418, (15) T. uncinatum 419, (16) T. uncinatum 420, (17) T. uncinatum 224, (18) T. uncinatum Virg, (19) T. uncinatum 445, (20) T. uncinatum 100, (21) T. aestivum 230, (22) T. aestivum 443, (23) T. aestivum A4, (24) T. aestivum A17, (25) T. aestivum A21, (26) ‘aestivum group’ 1, (27) ‘aestivum group’ 2, (28) ‘aestivum group’ 225, (29) ‘aestivum group’ 442, (30) ‘aestivum group’ 444. Sequenced bands are indicated by circles (16 SrDNA: lanes 3, 7, 17) and arrows (mrDNA: lanes 3, 5, 20, 22).B. Restriction profiles by Mse I from the above samples except those showing two bands (lanes 3, 7, 17).

and 1

Variability in black truffles 591

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

obtained by alignment of mitochondrial sequences frommorphologically distinct black (T. uncinatum/T. aestivum)and whitish (T. borchii ) truffles, clearly demonstrates thatNMS1/NMS2 do not discriminate Tuber species. There-fore the mitochondrial region between NMS1-NMS2,which is variable enough to discriminate Verticillium andFusarium species (Li et al., 1994), is highly conserved inTuber and does not represent a useful marker. Further-more, although mitochondrial primers were not useful forinvestigating genetic variability, they led to an unexpectedresult. In fact they revealed bacterial sequences, belong-ing to the Cytophaga-Flexibacter-Bacteroides (CFB) phy-logroup, in three samples of T. uncinatum and in T. borchiimycelium and similar to those described in T. borchiimycelium by Barbieri et al. (2000). Irrespectively of itsbiological meaning, the current finding clearly indicatesthat there is a pressing need to reconsider the specificityof NMS1/NMS2 primers for Ascomycetes.

ITS sequence: still a useful region for investigating variability in truffles

Our report adds new information on genetic variability intruffles providing 18 T. aestivum and T. uncinatum ITSsequences and a couple of primers specific for this trufflegroup.

The gene tree relative to the alignment of 15 sequencesdiscriminated between T. aestivum and T. uncinatum asthey clustered in two groups supported by high values ofbootstrap (98% and 81% respectively). Although a fewsequences of T. aestivum were considered, this groupcomprises samples from different geographic origins: twofrom France and one from Southern Italy.

Analysis of nine ITS sequences of T. uncinatum asco-mata from the Menconico truffle-ground revealed that thesimilarity (95%) of this group decreased when sequencesof T. uncinatum from different geographic areas wereadded (94.08%.). The genetic polymorphism is high, com-pared to that shown by seven samples of T. melano-sporum, from a wide geographic area (97.84% similarity),and low, compared to that of six samples of T. indicum,from China (86.90% similarity). Our results, concerningthe low T. melanosporum and the high T. indicum variabil-ity are confirmed by a similarity matrix based on data fromtheir ITS sequences (Mabru et al., 2001). These authorssuggest that the consistent intraspecific variation of T.indicum reflects an ambiguous situation. There is a lackof knowledge about the real origin of T. indicum isolatesand a lack of phylogenetic investigations for other Asiantruffles that are very close to T. indicum. These data, couldchange the value of the intraspecific variability in thisspecies (Mabru et al., 2001). Markers such as RAPD andmicrosatellites confirm that T. melanosporum geneticvariability is low (Bertault et al., 2001) when compared to

RAPD results from T. aestivum/T. uncinatum (Gandeboeufet al., 1997; Bertault et al., 2001).

On the basis of the ITS analysis we conclude that T.uncinatum is more polymorphic than T. melanosporum.These two taxa could be representative of two groups ofblack truffles, as far as concerns their variability.

According to Bertault et al. (2001) the low variability ofT. melanosporum is related to glaciation followed by rapidpost-glaciation demographic expansion. Genetic variabilityin a population is important because it enables the organ-ism to survive ecological and environmental challenges.This can justify the broader diffusion of T. aestivum andT. uncinatum, probably less affected by glaciation, withrespect to that of T. melanosporum: it is likely that thehigher variability of T. uncinatum has an ancient origin.

As we discriminated among T. aestivum and T. uncina-tum we wondered how far they were and whether it waspossible to design specific primers. The ITS sequenceswere not so different to allow the design of primers specificfor each group but drove us to design a pair of primers tointentionally differentiate all samples belonging to T. aes-tivum and T. uncinatum from other Tuber species.

They successfully amplified DNA from all the 30 asco-mata as well as from the mycorrhizas. The availability ofprimers able to amplify such a polymorphic group is surelyimportant for the identification of the two fungal taxa duringtheir symbiotic phase. Mycorrhizas from T. aestivum andT. uncinatum are morphologically similar to those of T.mesentericum, for which specific primers are not currentlyprovided. No amplification was given either from DNAcoming from other Tuber species or from Sphaerosporellabrunnea, a fungus that competes with truffles during myc-orrhiza establishment (Mello et al., 1996; Amicucci et al.,2000). This result is therefore useful in applicative pro-grams where the production of mycorrhizal plants withT. aestivum/T. uncinatum requires the certification of thefungus.

May a genetic analysis have a taxonomic meaning?

One of the major results of our investigation was in thefield of fungal taxonomy: T. uncinatum and T. aestivumcluster in two separate groups. This is a new finding,compared to data from previous investigations performedby Pacioni et al. (1993) and Urbanelli et al. (1998): usingallozyme analysis they assigned the two species to thesame electrophoretic type. In contrast, the use of differenttechniques allowed us to reveal two groups that couldrepresent two species. Furthermore, our genetic analysisprovides new information. Because products of amplifica-tion by UncI/UncII were co-dominant alleles at an individ-ual locus, their distribution was investigated in respect tothe sampling area. UncI-UncII product, digested with MspI enzyme, revealed three genotypes, called A, B and C

592 A. Mello, A. Cantisani, A. Vizzini and P. Bonfante

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

because they overlap those previously found, from anITS1-ITS4 product, by Guillemaud et al., 1996), in a nat-ural population of T. aestivum in France. The genotype C,identified through a cytosine/thymine substitution, is themost common in all the samples, irrespectively of theirmorphological identification. Genotype A has been foundin three T. aestivum samples from Italy and France and inthree samples of the ‘aestivum group’ from Italy andSpain. Only one sample belonging to genotype B hasbeen found, probably because of the small number ofsamples considered in our analysis. Therefore T. uncina-tum shows only a genotype whereas T. aestivum samplesshow three genotypes.

In conclusion, T. uncinatum and T. aestivum can beseparated not only thanks to their spore reticulum, themorphological feature which is considered at the basis oftheir morphological identification, and to their taste andsmell, but also from a molecular point of view. The twogroups could represent two distinct species if the geneticvariability of a broader number of T. uncinatum and T.aestivum samples confirm their separation.

Experimental procedures

Genomic DNA was isolated from samples listed in Table 1and from T. mesentericum, T. melanosporum, T. borchii, T.magnatum, T. ferrugineum, T. maculatum, T. indicum, T. bru-male, T. himalayense, T. excavatum, T. macrosporum, T.rufum, T. panniferum, T. foetidum, Sphaerosporella brunneaas described by Mello et al. (1996).

Microsatellite analysis

Microsatellite-primed PCR were carried out on DNA from all30 ascomata with (GTG)5 primer according to Longato andBonfante (1997), and with (GAC)5, (AAG)8, (AAC)8, accord-ing to Baleiras Couto et al. (1996), except that in the last twocases the annealing temperature was 55∞C. Electrophoresisof the amplified samples was carried out in 2% agarose using1 ¥ Tris borate EDTA (TBE) as running buffer. Bands werescored as presence (1) or absence (0) and used to constructa binary data matrix for a group of 12 samples; the distantmatrix was obtained with PAUP*4.0b10 (PPC/Altivec)(Swofford, 2002) and used for neighbour-joining (NJ)analysis.

ITS analysis

The ITS region from all the 30 ascomata was amplified withthe universal primers ITS1/ITS4 (White et al., 1990) accord-ing to Mello et al. (1996), whereas ITS1F (Gardes and Bruns,1993)/ITS4 were used to amplify DNA from the mycorrhizalsamples, according to Mello et al. (2001). The ITS amplifiedfragments from 12 ascomata of T. uncinatum, five of T.aestivum, three of Tuber belonging to the ‘aestivum group’and two mycorrhizal samples (Table 1), were purified from

agarose gels with a QUIAEX II Gel extraction kit (Quiagen,Promega, Madison, WI). Five sequences were obtained byGenome Express, Grenoble, and one by Genelab, Roma, bycloning, as described by Mello et al. (2001), and subsequentreading with T7 and SP6 primers. Three of these could notbe used in our analysis (Table 1). Seventeen sequences wereobtained directly by reading the purified PCR product withITS1 and ITS4 primers (Sequence Service, Quebec).All sequences were submitted to EMBL database. DNAsequences were aligned using the CLUSTAL X program(Thompson et al., 1997) and manually corrected to optimizealigned sites with GeneDoc (Nicholas et al., 1997).Neighbour-joining analysis, with Kimura 2-parameter modelcorrection for multiple substitutions, was carried out andTreeview was used to design the tree. To investigate nodesupports, 1000 bootstrap replications were performed.Twelve ITS sequences of T. uncinatum from our analysis,seven of T. melanosporum and six of T. indicum deposited inthe Genebank database, under the accession numbers listedin Table 1, have been separately aligned, with Clustal X(at the web site http://www.npsa-pbil.ibcp.fr/). Percentagesimilarity, as conserved positions in each alignment, wasused to compare the similarity of sequences in eachspecies.

Primers specific (UncI: TGGGCCGCCGAAAACTTG,UncII: CTGACGAGATGCCCCGGA) for samples belongingboth, to T. uncinatum and T. aestivum, were designed withinthe ITS and used in PCR at the following conditions: 95∞ for4 min, 27 cycles at 95∞ for 1 min, 59∞ for 45 sec, 72∞ for 45 secand a final extension at 72∞ for 5 min. Electrophoresis of theamplified products was carried out in 1.5% agarose gel using0.5 ¥ Tris acetate EDTA (TAE) as running buffer. The amplifiedproducts were digested with the restriction enzyme Msp Iaccording to Mello et al. (1996) and electrophoresis of thedigested samples was carried out in 2% agarose using TBEas running buffer.

Mitochondrial analysis

Mitochondrial ribosomal DNA (mrDNA) from all 30 ascomatawas amplified with NMS1/NMS2 primers designed in themitochondrial small rRNA region of fungi belonging to theclass Ascomycetes (Li et al., 1994). The reactions wereheated to 94∞ for 5 min, carried out for 35 cycles at 94∞ for45 sec, 57∞C for 45 sec, 72∞ for 1 min followed by a finalextension step at 72∞ for 5 min. Seven sequences (T. unci-natum 397c: higher and lower bands, T. uncinatum 400B:higher band, T. uncinatum 224: higher band, T. uncinatum399, T. borchii mycelium: higher and lower bands) wereobtained by cloning in pGEMt vector and subsequent readingwith T7 and SP6 primers by MWG BIOTECH AG (Table 2).Two direct sequences were obtained with NMS1 primer fromT. uncinatum 100 and T. aestivum 443 (Table 2). Allsequences were submitted to EMBL database. Each of theseven sequences was used as a query in BLAST software(Altschul et al., 1997) to find sequences producing significantalignments.

The amplified products showing a single band weredigested with the restriction enzyme Mse I according toMello et al. (1996) and electrophoresis of the digested sam-

Variability in black truffles 593

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

ples was carried out in 2% agarose using TBE as runningbuffer.

Acknowledgements

We wish to thank Dr Gerarde Chevalier for kindly providingthree of the samples analysed, Mr Francesco Meotto for thecollaboration with the Municipality of Menconico, Dr ClaudeMurat for helping us in the use of PAUP*4.0b10, Dr RiccardoPercudani and Dr Mariangela Girlanda for suggestions in theinterpretation of the phylogenetic trees. The research wasfunded by the National Council of Research in Italy, StrategicProgram: “Biotecnologia dei funghi eduli ectomicorrizici: dalleapplicazioni agro-forestali a quelle agro-alimentari.”

References

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.,Zhang, Z., Miller, W., and Lipman, D.J. (1997) GappedBLAST and PSI-BLAST: a new generation of protein databasesearch programs. Nucleic Acids Res 25: 3389–3402.

Amicucci, A., Zambonelli, A., Giomaro, G., Potenza, L., andStocchi, V. (1998) Identification of ectomycorrhizal fungi ofthe genus Tuber by species-specific ITS primers. Mol Ecol7: 273–277.

Amicucci, A., Guidi, C., Zambonelli, A., Potenza, L., andStocchi, V. (2000) Multiplex PCR for the identificationof white Tuber species. FEMS Microbiol Lett 189: 265–269.

Baleiras Couto, M.M., Eijsma, B., Hofstra, H., Huis in’t Veld,J.H., and van der Vossen, J.M. (1996) Evaluation of molec-ular typing techniques to assign genetic diversity amongSaccharomyces cerevisiae strains. Appl Environ Microbiol62: 41–46.

Barbieri, E., Potenza, L., Rossi, I., Sisti, D., Giomaro, G.,Rossetti, S., et al. (2000) Phylogenetic characteriza-tion and in situ detection of a Cytophaga-Flexibacter-Bacteroides Phylogroup bacterium in Tuber borchii Vittad.ectomycorrhizal mycelium. Appl Environ Microbiol 66:5035–5042.

Bertault, G., Raymond, M., Berthomieu, A., Callot, G., andFernandez, D. (1998) Trifling variation in truffles. Nature394: 734.

Bertault, G., Rousset, F., Fernandez, D., Berthomieu, A.,Hochberg, M.E., Callots, G., et al. (2001) Population genet-ics and dynamics of the black truffle in a man-made trufflefield. Heredity 86: 451–458.

Chevalier, G., and Frochot, H. (1997) La Truffe de Bourgogne(Tuber uncinatum Chatin). Petrarque.

Chevalier, G., and Grent, J. (1979) Application pratique de lasymbiose ectomycorhizienne: production à grande échellede plants mycorhizés par la truffe. Mushroom Sci 10: 483–505.

Chevalier, G., Riousset, G., Dupré, C., and Gandeboeuf, D.(1994) Tuber uncinatum Chat. et T. aestivum Vitt., espècesdifférentes ou simples variétés de la meme espèce? DocMycol 24: 17–26.

Dahlberg, A. (2001) Community ecology of ectomycorrhizalfungi: an advancing interdisciplinary field. New Phytologist150: 555–562.

Dennis, R.W.G. (1981) British Ascomycetes. Cramer, J. (ed.),Vaduz.

Gandeboeuf, D., Dupré, C., and Chevalier, G. (1994) Use ofisoenzyme analysis to differentiate truffles from Europe.Acta Bot Gallica 141: 455–463.

Gandeboeuf, D., Dupré, C., Roeckel-Drevet, P., Nicolas, P.,and Chevalier, G. (1997) Grouping and identification ofTuber species using RAPD markers. Can J Bot 75: 36–45.

Gardes, M., and Bruns, T.D. (1993) ITS primers withenhanced specificity for basidiomycetes-applications tothe identification of mycorrhizae and rusts. Mol Ecol 2:113–118.

Grove, T.S., and Le Tacon, F. (1993) Mycorrhiza in plantationforestry. Adv Plant Pathol 9: 191–227.

Guillemaud, T., Raymond, M., Callot, G., Cleyet-Marel, J.C.,and Fernandez, D. (1996) Variability of nuclear and mito-chondrial ribosomal DNA of a truffle species (Tuber aesti-vum). Mycol Res 100: 547–550.

Henrion, B., Chevalier, G., and Martin, F. (1994) Typing trufflespecies by PCR amplification of the ribosomal DNA spac-ers. Mycol Res 98: 37–43.

Li, K.-N., Rouse, D.I., and German, T.L. (1994) PCR primersthat allow intergeneric differentiation of Ascomycetes andtheir application to Verticillium spp. Appl Environ Microbiol60: 4324–4331.

Longato, S., and Bonfante, P. (1997) Molecular identificationof mycorrhizal fungi by direct amplification of microsatelliteregions. Mycol Res 101: 425–432.

Mabru, D., Dupré, C., Douet, J.P., Leroy, P., Ravel, C.,Ricard, J.M., et al. (2001) Rapid molecular typing methodfor the reliable detection of Asiatic black truffle (Tuberindicum) in commercialized products: fruiting bodies andmycorhizal seedlings. Mycorrhiza 11: 89–94.

Martin, F., Díez, J., Dell, B., and Delaruelle, C. (2002) Phylo-geography of the ectomycorrhizal Pisolithus species asinferred from nuclear ribosomal DNA ITS sequences. NewPhytol 153: 345–357.

Mello, A., Nosenzo, C., Meotto, F., and Bonfante, P. (1996)Rapid typing of truffle mycorrhizal roots by PCR amplifica-tion of the ribosomal DNA spacers. Mycorrhiza 6: 417–421.

Mello, A., Garnero, L., and Bonfante, P. (1999) Specific PCR-primers as a reliable tool for the detection of white trufflesin mycorrhizal roots. New Phytol 141: 511–516.

Mello, A., Vizzini, A., Longato, S., Rollo, F., Bonfante, P., andTrappe, J.M. (2000) Tuber borchii versus T. maculatum:neotype studies and DNA analyses. Mycologia 92: 326–333.

Mello, A., Fontana, A., Meotto, F., and Bonfante, P. (2001)Molecular and morphological characterization of Tubermagnatum mycorrhizas in a long-term survey. MicrobiolRes 156: 1–6.

Montecchi, A., and Sarasini, M. (2000) Funghi IpogeiD’Europa. Trento, Italy: A.M.B. Centro Studi Micologici.

Nicholas, K.B., Nicolas, H.B., Nicholas, J.R., and Deerfield,D.W.I.I. (1997) GeneDoc: analysis and visualization ofgenetic variation. EMBNEW NEWS 4: 14.

O’Donnell, K., Cigelnik, E., Weber, N.S., and Trappe, J.M.(1997) Phylogenetic relationships among ascomycetoustruffles and the true and false morels inferred from 18S and28S ribosomal DNA sequence analysis. Mycologia 89: 48–65.

594 A. Mello, A. Cantisani, A. Vizzini and P. Bonfante

© 2002 Blackwell Science Ltd, Environmental Microbiology, 4, 584–594

Pacioni, G., Frizzi, G., Miranda, M., and Visca, C. (1993)Genetics of a Tuber aestivum population (Ascomycotina,Tuberales). Mycotaxon 7: 93–100.

Paolocci, F., Rubini, A., Riccioni, C., Granetti, B., and Arcioni,S. (2000) Cloning and characterisation of two repeatedsequences in the symbiotic fungus Tuber melanosporumVitt. FEMS Microbiol Ecol 34: 139–146.

Pegler, D.N., Spooner, B.M., and Young, T.W.K. (1993)British Truffles., Kent, U.K.: Whitsstable Litho.

Percudani, R., Trevisi, A., Zambonelli, A., and Ottonello, S.(1999) Molecular phylogeny of truffles (Pezizales: Terfezi-aceae, Tuberaceae) derived from nuclear rDNA sequenceanalysis. Mol Phylogenet Evol 13: 169–180.

Polidori, E., Agostini, D., Zeppa, S., Potenza, L., Palma, F.,Sisti, D., and Stocchi, V. (2002) Identification of differen-tially expressed cDNA clones in Tilia platyphyllos-Tuberborchii ectomycorrhizae using a differential screeningapproach. Mol Genet Genomics 266: 858–864.

Read, D.J. (1991) Mycorrhizas in ecosystems. Experientia47: 376–390.

Riousset, L., Riousset, G., Chevalier, G., and Bardet, M.C.(2001) Truffes d’Europe et de Chine. INRA.

Rubini, A., Paolocci, F., Granetti, B., and Arcioni, S. (1998)Single step molecular characterisation of morphologicallysimilar black truffle species. FEMS Microbiol Lett 164: 7–12.

Rubini, A., Paolocci, F., Granetti, B., and Arcioni, S. (2001)Morphological characterisation of molecular-typed Tubermagnatum ectomycorrhizae. Mycorrhiza 11: 179–185.

Selosse, M.A. (2001) Apport des techniques génétiques etmoléculaires à la connaissance des communautés et des

populations de champignons ectomycorhhiziens. Lejeunia165 Fèvrier.

Smith, S.E., and Read, D.J. (1997) Mycorrhizal Symbiosis,2nd edn., San Diego: Academic Press.

Soragni, E., Bolchi, A., Balestrini, R., Gambaretto, C.,Percudani, R., Bonfante, P., et al. (2001) A nutrient-regulated, dual localization phospholipase A (2) in thesymbiotic fungus Tuber borchii. EMBO J 201: 5079–5090.

Stocchi, V. (1999) Metodi Molecolari Per l’IdentificazioneDelle Diverse Specie Di Tartufo. Urbino: Centro StampaUniversità.

Sunnucks, P. (2000) Efficient genetic markers for populationbiology. Trends Ecol Evol 15: 199–203.

Swofford, D.L. (2002) paup*. Phylogenetic Analysis UsingParsimony (* and Other Methods), Version 4., Sunderland,Massachusetts: Sinauer Associates

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F.,and Higgins, D.G. (1997) The Clustal–windows interface:flexible strategies for multiple sequence alignment aidedby quality analysis tools. Nucleic Acid Res 24: 4876–4882.

Urbanelli, S., Sallicandro, P., De Vito, E., Bullini, L., andBiocca, E. (1998) Biochemical systematics of some spe-cies in the genus Tuber. Mycologia 90: 537–546.

White, T.J., Bruns, T., Lee, S., and Taylor, J. (1990) Amplifi-cation and direct sequencing of fungal ribosomal RNAgenes for phylogenetics. In PCR Protocols. A Guide toMethods and Applications. Innis, M.A., Gelfand, D.H.,Sninsky, J.J., and White, T.J. (eds). San Diego: AcademicPress, pp. 315–322.