Tuberpulchrosporum sp. nov., a black truffle of the Aestivum clade (Tuberaceae, Pezizales) from the Balkan peninsula

Abstract Knowledge on the diversity of hypogeous sequestrate ascomycetes is still limited in the Balkan Peninsula. A new species of truffle, Tuberpulchrosporum, is described from Greece and Bulgaria. Specimens were collected from habitats dominated by various oak species (i.e. Quercusilex, Q.coccifera, Q.robur) and other angiosperms. They are morphologically characterised by subglobose, ovoid to irregularly lobed, yellowish-brown to dark brown ascomata, usually with a shallow basal cavity and surface with fissures and small, dense, almost flat, trihedral to polyhedral warts. Ascospores are ellipsoid to subfusiform, uniquely ornamented, crested to incompletely reticulate and are produced in (1–)2–8-spored asci. Hair-like, hyaline to light yellow hyphae protrude from the peridium surface. According to the outcome of ITS rDNA sequence analysis, this species forms a distinct well-supported group in the Aestivum clade, with T.panniferum being the closest phylogenetic taxon.


Introduction
The genus Tuber F.H. Wigg. (Ascomycota, Pezizales, Tuberaceae) is globally famous and historically appreciated for the production of hypogeous ascomata, known as 'truffles'; several of them are highly prized due to their unique aroma and culinary value. Moreover, the genus is known for the symbiotic ectomycorrhizal associations that its members form with several gymnosperm and angiosperm forest-tree species as well as with orchids (Riousset et al. 2001;Selosse et al. 2004;Mello et al. 2006;Trappe et al. 2009). Furthermore, truffles are also important for serving as a primary or supplementary source of nutrition for soil micro-fauna and several mammals (Hanson et al. 2003;Trappe and Claridge 2010;Schickmann et al. 2012).
A continuous interest in the study of this particular group has resulted in several recent reports on new Tuber species from various parts of the world (e.g. Crous et al. 2017;Fan et al. 2015; Guevara-Guerrero et al. 2018;Piña Páez et al. 2018). It is estimated that their number ranges between 180 and 220 (Zambonelli et al. 2016) nested in 11 major phylogenetic clades (Bonito et al. 2013). In particular, the Aestivum clade is composed of species associated with a large spectrum of host plants and are reported to occur in the Old World, i.e. Europe, North Africa and/or Asia (Jeandroz et al. 2008;Bonito et al. 2013;Payen et al. 2014). Indicative examples are T. aestivum Vittad. (the type species of the genus), T. panniferum Tul. & Tul., T. malenconii Donadini, Riousset, G. Riousset & G. Chev. and T. mesentericum Vittad., as well as T. sinoaestivum Zhang & Liu recently described from China (Zambonelli et al. loc. sit.;Zhang and Chen 2012). The morphologically diverse and economically important species T. magnatum Picco also forms part of this clade (Bonito et al. 2010a;. Although Tuber diversity is well documented in Europe (Bonito at al. 2010a, Ceruti et al. 2003, Jeandroz et al. 2008, the south-eastern part of the continent and especially the Balkan Peninsula was until recently poorly investigated. Indicative of this fact is that, by the end of the last century, only three Tuber species had been recorded in Greece (Zervakis et al. 1999). However, during the last two decades, an ever increasing interest in the collection of truffles led to a remarkable increase in the number of pertinent records (e.g. Diamandis and Perlerou 2008;Konstantinidis 2009;Agnello and Kaounas 2011;Alvarado et al. 2012a,b;Gyosheva et al. 2012); thus, to date, 15 Tuber spp. are reported from Greece. Similarly, only two Tuber spp. had been recorded in Bulgaria by the end of the last century; however, this number is fast-growing during the last few years and 14 species are currently known to exist (Dimitrova and Gyosheva 2008;Gyosheva et al. 2012;Lacheva 2012;Nedelin et al. 2016;Assyov and Slavova 2018). Regarding adjacent countries, 12 truffle species were reported to occur in Serbia, including one recently described (Marjanović et al. 2010;Milenković et al. 2015), while six Tuber spp. were recorded in Montenegro, five in FYROM and four in Albania (Pacioni 1984;Marjanović et al. 2010).
In the frame of this work, several truffle specimens originating from north and central continental Greece and from Bulgaria were studied with respect to their morphology and phylogenetic relationships to other Tuber taxa and a new species is hereby proposed.

Sampling and Morphological characterisation
Specimens used for this study were collected during 2008-2017 from north and central Greece (Regions of Epirus, Thessaly, Eastern Macedonia and Thrace, Western Greece and Attica), as well as from Bulgaria (Regions of Eastern Stara Planina and Black Sea coast). Specimens are deposited in the fungaria of the Laboratory of General and Agricultural Microbiology (Agricultural University of Athens, ACAM), of the Institute of Biodiversity and Ecosystem Research (SOMF) and the authors' personal collections. Macroscopic characters such as size, peridium surface texture, colour and odour were observed in fresh ascomata. Colour coding and terminology is derived from the "Flora of British Fungi -Colour Identification Chart" (Royal Botanic Garden Edinburgh 1969).
Microscopic characters were examined by hand-cut sections on fresh and dried material, using a Zeiss Axioimager A2 microscope under bright field and Differential Interference Contrast (DIC) and an AmScope T360B. Microphotographs were taken with the aid of a mounted digital camera (Axiocam). Microscopic observations were performed in water, 3% (w/v) potassium hydroxide (KOH) and Melzer's reagent. To assess the ascospore size, a minimum of 30 mature ascospores from each type of asci (2 to 8-spored) were measured and dimensions are provided as (minimum) average ± standard deviation (maximum); quotient (Q), i.e. length divided by the width, was calculated for each ascospore and the median value (Qm) is given. For scanning electron microscopy (SEM), ascospores were scraped from the hymenial surface and mounted on aluminium foil, which was then fixed on a microscope holder and sputter-coated with gold. Observations were performed in JEOL JSM-5510.

DNA sequencing and Phylogenetic analyses
Total genomic DNA was extracted from herbarium specimens using the Nucleospin Plant II DNA kit (Macherey and Nagel, Germany) following the manufacturer's protocol with minor modifications. The internal transcribed spacer (ITS) region of nuclear ribosomal DNA (nrDNA) was amplified using the primer combination ITS1/ ITS4 (White et al. 1990). Polymerase chain reactions (PCR) were performed in 50 μl containing 50 ng DNA template, 0.25 μM of each primer, 0.2 mM of each dNTP, 1× HiFi Buffer (Takara BIO INC., Japan) and 1 U HiFi Taq DNA polymerase (Takara BIO INC., Japan). Conditions for PCR amplification were as follows: 94 °C for 5 min, followed by 35 cycles of 94 °C for 30 sec, 50 °C for 30 sec and 72 °C for 1 min, with a final extension at 72 °C for 10 min. PCR products were purified using Invitrogen Pure-Link kit (Thermo Fisher Scientific, Korea) and were submitted for sequencing to Ce-MIA SA (Larissa, Greece). DNA sequences were then visualised, manually edited and assembled using UGENE (Okonechnikov et al. 2012). Validated sequences, generated in this study, were deposited in GenBank under the accession numbers MK113975 to MK113982 (Table 1). Moreover, the percent sequence identity was estimated by using ClustalOmega (Sievers and Higgins 2018) through the EMBL-EBI portal. A total of 62 Tuber ITS rDNA sequences were used for phylogenetic analysis by including eight sequences of T. pulchrosporum sp. nov. and 54 sequences from GenBank (nine of them representing type specimens) which correspond to 31 Tuber taxa mainly of European distribution (Table 1). Choiromyces alveolatus (Harkn.) Trappe (AF501258, EU697268) was used as the outgroup. Sequence alignment was performed through the online version of the multiple sequence alignment programme MAFFT v7 (Katoh and Standley 2013) by applying the Q-INS-I strategy and alignments were inspected and manually adjusted at misaligned sites by using MEGAX (Kumar et al. 2018). The pertinent matrix was deposited in TreeBASE under the accession number 23587.
Phylogenetic relationships of taxa were inferred by using maximum likelihood (ML) and Bayesian Inference (BI) through the CIPRES portal (www.phylo.org; Miller et al. 2010). ML analysis of the ITS dataset was conducted by RAxML v8.2 (Stama-takis 2014) with 1,000 bootstrap replicates and search for the best-scoring ML tree. BI analysis was performed by MrBayes v3.2.1 (Ronquist et al. 2012) and the General Time Reversible + Gamma (GTR+G) model was selected as the best model under the Akaike Information Criterion (AIC) implemented in MrModeltest v2.3 (Nylander 2004). To estimate posterior probabilities, 20,000,000 Markov chain Monte Carlo (MCMC) simulation generations were run in two parallel independent runs of four chains, one cold and three heated, with trees sampled every 1,000 generations and the first 25% of trees were omitted as burn-in. A 50% majority rule consensus tree was built and visualised with iTOL (Letunic and Bork 2016). Clades with bootstrap support (BS) ≥ 70% and Bayesian posterior probability (PP) ≥ 95% were considered as significantly supported.  Diagnosis. Ascomata 0.6-7(-10) cm in diam., subglobose, ovoid to irregularly lobed, usually with shallow basal cavity, surface with fissures and small, dense, almost flat trihedral to polyhedral warts, yellowish-brown to dark brown. Ascospores 25.0-37.0 × 18.2-25.6 μm in (1-)2-8-spored asci, ellipsoid to subfusiform on average, Qm=1.4, crested to incompletely reticulate. Hair-like, hyaline to light yellow-brown hyphae protruding from peridium surface.
Etymology. "pulchrosporum" refers to the uniquely distinct/impressive ornamentation of the ascospores.
Distribution and ecology. Hypogeous, in soil, appearing solitary or in small groups from March to June, under Quercus sp., Q. coccifera or Q. ilex L. or under Carpinus sp. or in mixed stands of Quercus sp. and Pinus nigra J.F. Arnold or of Q. ilex and Pinus halepensis Miller or of Quercus robur L., Corylus sp., Carpinus sp. and Acer sp. It seems to be rather common in continental (northern and central) Greece, while it also occurs in the regions of Eastern Stara Planina and the Black Sea coast of Bulgaria.  Phylogenetic aspects. The resultant ITS sequence data comprises of 64 sequences which were aligned at 780 sites, 738 of which represent the ITS1-5.8S-ITS2 region, i.e. between the end of the SSU motif (CATTA) and the beginning of LSU motif (TAGGG) (Bonito et al. 2010a). ML and BI analyses yielded similar tree topologies and only the tree inferred from the Bayesian analysis is presented (Fig. 5). The morphologically variable genus Tuber is monophyletic (BS: 100%, PP: 1.00) and several lineages are revealed; for the purposes of this study, the following highly supported clades were included: Aestivum, Excavatum, Gennadii, Gibbosum, Latisporum, Maculatum, Macrosporum, Melanosporum, Puberulum, Regianum, Rufum, Tumericum (=Japonicum).
According to the phylogenetic analysis performed, T. pulchrosporum belongs to the Aestivum clade. All eight sequences of this new taxon form a distinct highly supported subclade (BS: 100%, PP: 1.00). Greek specimens possessed almost identical ITS sequences (99.8 -100%) and so did Bulgarian samples, whereas the comparison between collections from the two countries resulted in sequence identity values of 98.13 ± 0.08%. In total, intraspecific sequence identity values for T. pulchrosporum exceeded 98% (i.e. 98.05 -100%). The new species is sister to T. panniferum (BS: 100%, PP: 1.00); the respective sequences demonstrated low sequence identity (73.21 -75.08%) further evidencing their distinct taxonomic status.

Discussion
The molecular analysis evidenced that the eight sequences representing T. pulchrosporum are grouped within the Aestivum clade by forming a distinct terminal group supported with high BS and PP values. The closest phylogenetic relative of T. pulchrosporum is T. panniferum Tul. & C. Tul., i.e. a Mediterranean species with analogous ecological preferences (Jeandroz et al. 2008). T. panniferum also exhibits a rather similar macromorphology characterised by a brownish pubescent peridium, absence of pyramidal warts and ascomata often bearing a cavity, although the tomentum is much more prominent, exhibiting thus a felted appearance. However, the microscopic features of the two species are clearly different. In T. panniferum, the ornamentation consists of isolated spines never exceeding 3 μm in height, while the peridial surface is covered by rusty brown hyphae which form a dense cottony mass (Montecchi and Sarasini 2000;Riousset et al. 2001;Moreno-Arroyo et al. 2005).
By morphology alone, T. pulchrosporum is easily distinguishable within the Aestivum clade since no other species produces ascospores bearing such a uniquely crested ornamentation. The more distant T. aestivum (Wulfen) Spreng. (including T. uncinatum Chatin) and T. sinoaestivum J.P. Zhang & P.G. Liu could be distinguished macroscopically thanks to their blackish peridial surface with prominent pyramidal warts and ascospores bearing a complete reticulum. Ascospores of T. mesentericum Vittad. show some affinity in their outline to those of T. pulchrosporum but they clearly possess a much more reticulate network; moreover, the peridial surface is black with pyramidal warts as in T. aestivum.
Although phylogenetically more distant, some other species with asci containing 1-8 ascospores may superficially resemble T. pulchrosporum. Hence, T. regianum Montecchi & Lazzari, the recently described T. magentipunctatum Z. Merényi, I. Nagy, Stielow & Bratek and T. bernardinii Gori, all belonging to the Regianum clade (Zambonelli et al. 2016;Crous et al. 2017), possess a reddish-brown to brown peridial surface with dense and rather flat warts as in the case of T. pulchrosporum. However, they all produce ascospores with pointed spines which are connected to form a complete reticulum. Ascomata of T. malenconii Donadini, Riousset, G. Riousset & G. Chev and T. pseudoexcavatum Y. Wang, G. Moreno, Riousset, Manjón & G. Riousset also show a macroscopic resemblance to T. pulchrosporum, with their rough indistinctly warty peridial surface (black for the former and brown for the latter), often with a similar basal cavity as well. However, ascospores of both T. malenconii and T. pseudoexcavatum have short spines, basally/broadly connected, exhibiting a more or less regular reticulum (Donadini et al. 1979;Manjón et al. 2009). Therefore, the unique type of ornamentation of T. pulchrosporum ascospores clearly distinguishes it from all species with similar macroscopic appearance.