Académique Documents
Professionnel Documents
Culture Documents
Received 3 August 1999; received in revised form 9 September 1999; accepted 13 September 1999
Abstract
Phylogenetic relationships among truffle species from Europe and China were investigated through parsimony analysis of
the ITS sequences. Three major clades were obtained among the species analysed. The so-called white truffles appeared
polyphyletic since Tuber magnatum was grouped with brown truffles and not with the other white species (T. maculatum,
T. borchii, T. dryophilum, T. puberulum). The black truffles investigated in this study, T. brumale, T. melanosporum, T. indicum
and T. himalayense, were grouped in an independent clade. The Përigord black truffle T. melanosporum and the Chinese black
truffles T. indicum and T. himalayense, were very closely related. The delimitation of these species was estimated by a distance
analysis on several isolates collected from different geographic areas. In spite of intraspecific variations of the internal
transcribed spacers (ITS) sequences, T. melanosporum and the Chinese black truffles can be unambiguously attributed to
distinct taxa. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights
reserved.
0378-1097 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 9 9 ) 0 0 4 7 4 - 7
or Tuber brumale Vitt. var brumale Vitt. and T. bru- T. indicum. Several black tru¥es have been described
male Vitt. var. moschatum F. de B. [7,9^11] is still in Asia (T. indicum Cooke and Massee, Tuber hima-
controversial. layense BC Zang and Winter, Tuber pseudohima-
The di¤culty in di¡erentiating some species and/or layense [24], Tuber sinense K Tao and Liu) and the
varieties in the Tuber taxa can be due to the overall distinction between them is sometimes tricky.
similarity of morphological characters and the fact In this study, a phylogenetic investigation of the
that descriptions have been somewhat approximate. genus Tuber was completed by a subspeci¢c popula-
The separation of T. aestivum and T. uncinatum into tion analysis in order to limit incorrect phylogenetic
two species, for example, is the result of contradic- interpretations due to intraspeci¢c variations. Parsi-
tory taxonomical interpretations. These two species mony inferences were applied on the ITS1-5.8S-ITS2
formerly separated by their morphological and eco- sequences. ITS regions have already been proven to
logical di¡erences [8,12^14] are now, after biochem- be useful in the phylogeny of closely related species
ical and molecular analyses, considered as a complex [18], although intraspeci¢c ITS polymorphism was
of strains in which intermediate forms make up a shown in several fungi [25^28]. ITS polymorphism
continuum between the two extreme forms of the of several samples of T. melanosporum and of some
species [8,9,15]. This heterogeneity within one species Chinese tru¥es collected in di¡erent regions was
underlines the need to integrate population dynamics then analysed by polymerase chain reaction-restric-
analyses in phylogenetic studies. However, in system- tion fragment length polymorphism (PCR-RFLP).
atics, morphological description referred to only one
holotypus, and molecular phylogenetic analyses are
usually based on a unique sequence for each species. 2. Materials and methods
The interspeci¢c phylogeny of organisms is con-
fronted with the problem of species delimitation. In 2.1. Fungal isolates and sequences
the case of the genus Tuber, phylogenetic approaches
based on isozyme polymorphism [10,16] and RAPD The samples used in our assays, their geographic
patterns [9,17] used several samples for each species origin and the GenBank accession numbers are listed
to estimate the intraspeci¢c variation of the charac- in Table 1. The tru¥es from the di¡erent geographic
ters. In spite of their interest, these analyses are areas were identi¢ed in this laboratory.
sometimes contradictory as the characters considered
were the result of a mix of inter- and intraspeci¢c 2.2. DNA extraction and PCR ampli¢cation
variations or could be not correlated at all [18^20].
The problem of species delimitation was intensi¢ed DNA extraction from ascocarps was performed on
with the importation into Europe of black tru¥es 200 mg of gleba. The samples were ground in sand
from China. Macroscopic characters of the fruit with 1 ml of extraction bu¡er (0.2 M Tris-HCl pH 8,
bodies and microscopic observation of ascospores 0.02 M EDTA pH 8, 1.4 M NaCl, 2% CTAB, 0.2%
do not allow them to be easily distinguished from L-mercaptoethanol) [29]. After centrifugation, 20 Wg
the famous black tru¥e Tuber melanosporum, espe- of proteinase K and 10 Wg of RNAse A were added
cially when the ascocarps are immature [21]. Com- to the supernatant and left for 5 min at 62³C. After
parison of T. melanosporum and Tuber indicum with centrifugation, 10 min at 8000 RPM, 2 ml of 5 M
RAPD markers led to the conclusion that the two potassium acetate (pH 6) were mixed to the super-
taxa are distinct [9]. However this work also reported natant. After 10 min at 4³C, proteins were pelleted
a high polymorphism among Chinese tru¥es, like- by centrifugation, 15 min at 8000 RPM. The super-
wise observed with microsatellite and internal tran- natant was gently mixed with half a volume of
scribed spacers (ITS) sequence analyses [22,23]. It chloroform and centrifuged for 15 min at low speed.
was unclear whether this polymorphism was related The supernatant was then mixed with 0.7 vol. of
to geographic variations as proposed by Paolocci et isopropanol, 10 min at 4³C. DNA was pelleted by
al. [23] and Longato and Bonfante [22], or if it re- centrifugation (15 min, 8000 RPM) and washed
£ected a misidenti¢cation of several taxa taken to be twice with 70% ethanol. The air-dried pellet was re-
suspended in 100 Wl of TE bu¡er (10 mM Tris-HCl France). The cycling conditions consisted of an ini-
and 1 mM EDTA pH 8). The purity and quantity of tial denaturation step at 94³C for 5 min, 1 min at
the DNA were checked by spectrophotometry at 52³C and 1 min at 72³C (30 cycles). The ampli¢ed
260, 280 and 320 nm. DNA was visualized by electrophoresis on 1.5%
The ITS 1F and ITS 4 primers used to amplify agarose gels.
the ITS sequence have been described by Gardes For sequencing, the PCR products were puri¢ed
and Bruns [29]. The ampli¢cations used 25 Wl using the QIAquick kit (QIAGEN S.A., France) ac-
DNA template solution in 50 Wl of reaction mixture. cording to the manufacturer's protocol. The DNA
The ¢nal solution contained 25 WM each of dATP, templates were mixed with PRISM dye deoxy termi-
dCTP, dGTP and dTTP (Euromedex, France), nator cycle sequencing kit. Samples were loaded on a
0.5 WM of each primer and one unit of thermostable 4% polyacrylamide, 48 cm gel, using an Applied Bio-
polymerase (Red Hot, AB-Fischer Scienti¢c, systems automated sequencer (model 373 Perkin El-
Table 1
Origin of the fungal material and sequences used in this study
Species Isolate Origin GenBank accession number
T. melanosporum Vitt. Tm3 Lot (France)
Tm5 Huesca (Spain)
Tm6 Lot (France)
Tm12 Lot (France)
Tm13 Vaucluse (France) AF132501*
Tm14 Vaucluse (France)
TmM2 unknown (EMBL) Y09790
TmA1 L'Aquila (Italy) U89359
Fig. 1. Cladogram obtained with 18 aligned ITS1-5.8S-ITS2 sequences of di¡erent species of tru¥es. The ITS sequence of P. vulgaris (As-
comycotina, Discomycetes, Pezizales) is added as an outgroup. The consensus tree resulted from 500 bootstrap replicates on 329 sites (117
informative), while 406 steps were required for maximum parsimony. Partial ITS sequences of T. aestivum and T. excavatum were added
in independent analyses (*: based on the comparison of ITS1-5.8S sequences. **: based on the comparison of 5.8S-ITS2 sequences) and
the resulting groups are noted in the tree.
mer/Applied Biosystems Division, Foster City, CA, at the recommended optimal temperature. For each
USA) PCR product, 5 restriction enzymes were used:
MseI, NlaIV, RsaI, HinfI and DdeI (MBI Fermen-
2.3. DNA restriction tas-Euromedex, France). These enzymes were chosen
to cut inside the most variable regions of the ITS1
Ten microlitre of ampli¢ed DNA was digested for and ITS2 sequences as revealed by restriction maps.
RFLP analysis using the conditions de¢ned by the The restriction digests were then electrophoresed on
manufacturer. The solutions were incubated for 3 h 3% agarose gels (Euromedex, France).
2.4. Sequence alignment, phylogenetic inference and T. aestivum (data not shown). The position of these
distance analysis species in the phylogeny estimated with the two trees
is reported in Fig. 1. T. excavatum appears related to
The sequences were aligned with the software Sea- T. ferrugineum. However their relationships with
View [30] (algorithm ClustalW) and the phylogenetic clades I or II is not clearly established (low bootstrap
inferences performed with Phylo_win [30] (algorithm value). T. aestivum constitutes a monophyletic group
DNApars from PHYLIP, v.3.5c, J. Felsenstein) on a with T. uncinatum and a bootstrap value of 100 val-
UNIX system. Bootstrap analyses were made of idates this clade.
500 replicates using the bootstrap and consensus op- Clade III in Fig. 1 is composed of morphologically
tions of Phylo_win. The ITS sequence of Pithya vul- similar species (T. melanosporum, T. brumale and the
garis was used to root the tree. The data from re- Chinese tru¥es). At least two sequences for each
striction analyses were analysed using Restsite [31] species were used in this clade. Fig. 1 shows that
and Neighbor (option Neighbor-Joining algorithm) no di¡erences were found between the ITS sequences
from PHYLIP. As the restriction map for each spe- of the two varieties of T. brumale described (var.
cies was known, the results were analysed in terms of brumale and var. moschatum). T. melanosporum
presence or absence of site (s option of Restsite). The forms an independent taxon distinct from the Chi-
dendogram was built with the software NJ Plot nese tru¥es analysed. This distinction is strongly
available on PHYLIP. supported by a bootstrap value of 100. Inside the
taxon with the Chinese tru¥es, T. indicum and
T. himalayense are separate. In addition, a third tax-
3. Results on appears (Ti38), close but distinct from the
T. himalayense clade.
3.1. Phylogeny of Tuber taxa
3.2. Distance analysis of T. melanosporum and some
ITS regions were ampli¢ed and sequenced. After tru¥es from China
alignment of the sequences, a consensus tree was
built consecutively with a 406-step parsimony proce- The unrooted dendogram in Fig. 3 was built with
dure (Fig. 1). The ITS sequence of T. brumale being the data obtained by the PCR-RFLP approach (Fig.
longer due to a duplication of a part of the ITS1 2). Among the ¢ve enzymes used (HinfI, MseI,
region (880 versus 600^700 bp for the other Tuber NlaIV, DdeI and RsaI), only one (MseI) presented
species), the alignment step was crucial to perform three intraspeci¢c polymorphic pro¢les for T. mela-
any phylogenetic analysis. Of the 19 Tuber species nosporum. Two enzymes, MseI and NlaIV, exhibited
compared, 329 sites were selected giving 117 infor-
mative sites. Fig. 1 shows that most of the clades
were supported by high bootstrap values. The three
major clades (clades I, II, III) were obtained inde-
pendently of the number of species included in each
one (data not shown).
Since the T. aestivum and Tuber excavatum ITS
sequences obtained from GenBank were truncated,
their alignment with the full sequences of the other
species presented in Fig. 1 was not possible. Inter-
mediate steps were required to position these two
species on the ¢nal tree (Fig. 1). A ¢rst tree was built Fig. 2. Electrophoregram obtained after restriction of the ITS se-
quence of T. melanosporum with the enzymes HinfI (lane 2),
after alignment with T. aestivum and ITS1-5.8S re-
MseI (lane 3), NlaIV (lane 4), DdeI (lane 5), RsaI (lane 6). Lanes
gions of all the other species except T. excavatum, a 8^12 : restriction pro¢les of T. indicum obtained with Hinf1,
second one resulted in alignment of 5.8S-ITS2 re- MseI, NlaIV, DdeI and RsaI, respectively. Lanes 1, 7 and 13:
gions of all species plus T. excavatum but without 100-bp ladder.
Fig. 3. Distance analysis between several isolates of T. melanosporum, T. himalayense and T. indicum, based on PCR-RFLP pro¢les of
ITS regions. The unrooted tree was built with the Neighbor-Joining algorithm. The scale corresponds to the relative genetic distance.
¢ve di¡erent restriction patterns for T. indicum and for the theory of a complex of species with ecological
no polymorphism for T. himalayense. All the en- and morphological varieties [8]. Clade II is composed
zymes used showed a high interspeci¢c polymor- of the so-called white tru¥es. The species Tuber
phism. The genetic distances given on the dendo- borchii, Tuber maculatum and Tuber puberulum,
gram are relative and only relevant to the enzyme which make it up, have morphological similarities.
system used in our analyses. In these conditions, However, since Tuber magnatum, the famous Pied-
three groups are separate, with intraspeci¢c poly- mont white tru¥e, is grouped with brown tru¥es in
morphism appearing less than interspeci¢c polymor- clade I, the white-tru¥e group appears polyphyletic.
phism. These three separate groups correspond to The presence of T. magnatum in the clade of
the three taxa analysed: T. melanosporum, T. indicum brown tru¥es, the proximity of T. excavatum and
and T. himalayense. T. ferrugineum, two species with alveolate and echi-
nulate ascospores, and the mix in clade III of asco-
spores that are echinulate (T. brumale, T. melanospo-
4. Discussion rum), more or less reticulate (T. indicum, T.
himalayense) and perhaps alveolate (Ti38), suggest
The cladogram on Fig. 1 shows that three evolu- that ascospore ornamentation and colour of the as-
tionary groups may be considered in the Tuber ge- cocarp gleba and peridium are not phylogenetically
nus. The clades obtained are signi¢cant, being sup- informative. The homoplasmy of the `ascospore or-
ported by high bootstrap values. In clade I, the namentation' character is supported by observations
homology of T. uncinatum and T. aestivum argues of spore ontogenesis showing that alveolate and echi-
nulate ornamentations have the same ultrastructural ci et al. [37]. Ti80 appears to be included in the clade
origin [32,33]. Similar conclusions have been drawn with our T. indicum samples. However, Ti20 is in the
for other fungal groups [34]. However, this does not clade with T. himalayense samples, and we propose
decrease the value of these characters as determina- that this isolate is actually T. himalayense, not
tion keys. T. indicum. This result con¢rms the observations of
The distance analysis performed with the PCR- Paolocci et al. [37] on the morphology of the Ti20
RFLP approach showed a polymorphism within tru¥e which di¡ered from the other T. indicum mor-
the ITS sequence of T. melanosporum. This is in phology used in their study. This also explains the
agreement with Henrion et al. [35] who observed very high divergence within the T. indicum species
polymorphic sites in the ITS 2 region using AluI reported by these authors and the absence of the
and HinfI restriction enzymes. We obtained a similar clear distinction they observed between T. melano-
result with a di¡erent enzyme complex, supporting sporum and T. indicum.
the conclusion that a polymorphism higher than we In spite of intraspeci¢c polymorphism of ITS se-
described in our distance analysis may be found in quences, the species T. melanosporum, T. indicum
the ITS region of T. melanosporum. Although slight, and T. himalayense clearly formed three separate
the intraspeci¢c polymorphism observed of the usu- groups (Fig. 3). The distance analysis supports the
ally constant ITS sequences is in contrast with the results obtained with the phylogenetic tree (Fig. 1)
very low diversity obtained with length polymor- where the high bootstrap value for each node indi-
phism analyses using RAPD and mirosatellites cates that the three taxa were unambiguously di¡er-
[9,36]. The use of a co-dominant technique like ent. In addition, sample Ti38 from Paolocci et al.
PCR-RFLP to analyse variability should allow de- [37] was included in our parsimony analysis. We ob-
tection of higher degrees of polymorphism than a served that this sample was distinct from T. indicum
dominant technique. In the case of T. aestivum, and close to T. himalayense. It would be interesting
ITS polymorphism and the high variability of to further investigate the delimitation of this taxon in
RAPD patterns [9,15,23] could be related to the het- order to determine if it corresponds to a variety of
erogeneity of the morphology of this complex taxon. T. himalayense or to a di¡erent but closely related
Such an interpretation cannot be made for T. mela- species, like T. pseudohimalayense [24].
nosporum. In addition, we did not ¢nd any correla- Our study has combined a molecular phylogenetic
tion between ITS polymorphism and the ascocarp investigation with a population analysis to clarify the
origins. This result has already been reported by systematics of Tuber species. It illustrates that intra-
Bertault et al. [36]. speci¢c analyses must be carried out to develop a
The PCR-RFLP analyses were also applied to the more reliable classi¢cation of the species based on
Chinese tru¥es in order to detect any intraspeci¢c DNA comparison, especially with the ITS. Although
polymorphism. Of the two species analysed, T. hima- these sequences may be useful to di¡erentiate T. mel-
layense and T. indicum, only the latter exhibited a anosporum, T. indicum, T. himalayense and Ti38
high polymorphism of its ITS sequence. As for [37,38], the polymorphism we observed must be tak-
T. melanosporum, this polymorphism cannot be cor- en into account before primers are designed in re-
related to any geographical origin. In another ITS gions of the ITS where no variation occurs.
sequence analysis, Paolocci et al. [37] reported an
even greater polymorphism than we observed. Mi-
crosatellite investigations also showed high polymor- Acknowledgements
phism [22], whereas the RAPD approach only re-
vealed low genetic variability [9]. These di¡erent We are very grateful to Pebeyre S.A. (Cahors,
analyses cannot be correlated considering the tech- France) and to the Consiel Rëgional de Midi-Pyrë-
niques used. Two ITS sequences from GenBank in- nëes (France) who supported this work, and to Dr
cluded in our analyses were obtained from ascocarps M. Kulifaj (Pebeyre S.A., Cahors, France) for pro-
(Ti80 and Ti20) determined as T. indicum by Paoloc- viding the tru¥es used in this study.
[34] Roux, C., Almaraz, T. and Durrieu, G. (1998) Phylogeny of [37] Paolocci, F., Rubini, A., Granetti, B. and Arcioni, S. (1997)
some smuts fungi based on ITS sequences analysis. C. R. Typing Tuber melanosporum and Chinese black tru¥e species
Acad. Sci. 321, 603^609. by molecular markers. FEMS Microbiol. Lett. 153, 255^
[35] Henrion, B., Chevalier, G. and Martin, F. (1994) Typing truf- 260.
£e species by PCR ampli¢cation of the ribosomal DNA [38] Rubini, A., Poalocci, F., Granetti, B. and Arcioni, S. (1998)
spacers. Mycol. Res. 98, 37^43. Single step molecular characterization of morphologically sim-
[36] Bertault, G., Raymond, M., Berthomieu, A., Callot, G. and ilar black tru¥e species. FEMS Microbiol. Lett. 164, 7^
Fernandez, D. (1998) Tri£ing variation in tru¥es. Nature 394, 12.
734.