Molecular phylogeny and morphology of Pseudobaeospora cyanea

Pseudobaeospora cyanea, a rare basidiomycete belonging to an underinvestigated genus, is currently recorded from only three localities of the Iberian Peninsula. Moreover, to date no sequences of this rare species have been deposited in GenBank. In this paper a new collection from NW Italy is reported with detailed morphological descriptions and iconography. The first ITS and LSU sequences for the species are provided and uploaded to GenBank, and the taxonomic placement of P. cyanea within the genus is discussed.


Introduction
The genus Pseudobaeospora Singer, circumscribed in 1942 to accommodate the type species Pseudobaeospora oligophylla (Singer) Singer, currently includes 30 species (http://www. mycobank.org/ accessed February 2021) of small white-spored agarics characterized by small thick-walled dextrinoid basidiospores (Bas 2002). The ecology is unclear, although suggested to be presumably saprotroph (Arauzo 2011). Its current accommodation in the family Tricholomataceae, with Leucopaxillus as a sister taxon, is supported by Desjardin et al. (2014) and by Sánchez-Garcia and Matheny (2017), although this placement is questioned by Wu et al. (2017). The genus Pseudobaeospora has been overlooked by taxonomists for a long time and until 1995 only two species were known in Europe (Arnolds et al. 2003), which were synonymized a few years later (Roniker and Moreau 2007). Several new European species have been described since then, mostly on a macro and micromorphological basis (Bas 2002), leading to more than 20 species being currently known for Europe (Adamčík and Jančovičová 2011), most of which are rare and only known from very few locations. The coverage of genus Pseudobaeospora in GenBank and UNITE is currently insufficient, with only seven species identified (about 20% of the total of described species) having available sequences in the database (https://www.ncbi.nlm.nih. gov/genbank/ and https://unite.ut.ee/search.php#fndtn-panel1 accessed February 2021).
Pseudobaeospora cyanea Arnolds, Tabarés & Rocabruna is a species described in Spain, based on macro-and micro-morphological analyses of a collection from Vidreres, Catalonia. The holotype specimens were collected in early November on Mediterranean hills (200 m a.s.l.) with Pinus pinaster Aiton, Arbutus unedo L., and Erica arborea L. (Arnolds et al. 2003). The species has been reported again in late October 2007 from two localities of the Basque Country (N Spain), around 600 m a.s.l., in the litter of Chamaecyparis lawsoniana (Murray) Parl. plantations (Arauzo 2011).
In the present study, a new collection of P. cyanea from a locality near Genoa (NW Italy) is reported. Morphological and molecular analyses of this collection are carried out, with the aim of increasing the knowledge about distribution, genetics, and phylogenetic relationships in this poorly known genus.

Morphological analysis
The specimens were identified through macro-morphological observations and evaluation of micro-morphological features. The herbarium specimens were prepared with a dryer and deposited in the mycological herbarium of the "Giacomo Doria Civic Museum of Natural History" (GDOR M3986).
The dried specimens were rehydrated in pure water, and the microscope slides were mounted with Congo red. More slides were prepared with Melzer's reagent and cotton blue, to observe the dextrinoid and cyanophylic reactions of the basidiospores. The slides were observed at 100× magnification with a Leica DM 500 binocular optical microscope. For basidiospores and other structures, at least 20 individuals were measured.

Sequence alignment and phylogenetic analysis
The BLASTN algorithm was used to compare the sequence obtained in the present work against the GenBank database. The sequence was then aligned with the other Pseudobaeospora sequences currently available on GenBank and UNITE, with the addition of Xerula pudens (Pers) Singer (Physalacriaceae) for rooting purposes, using the MUSCLE tool in the MEGA 7 software (Pennsylvania State University, PA, USA). Then a phylogenetic tree was inferred by maximum likelihood with 500 bootstrap replicates, using MEGA 7.

Habitat and ecology
The specimens were collected on 6 December 2016, D. Gisotti & F. Boccardo (GDOR M3986) in the locality of Pegli, Genova, 44°25'53.4"N, 8°48'34.2"E, at an elevation of 95 m, in an area of shrub-like Mediterranean vegetation with Pinus pinaster, Arbutus unedo, Erica arborea, Cistus salvifolius L., and Quercus ilex L., on poor acidic soil with serpentine bedrock. The basidiomata are gregarious, growing in the needle litter of P. pinaster. The species is reported to be presumably saprotrophic (Arnolds et al. 2003).

Sequencing and phylogenetic analysis
The sequences obtained from the specimen were uploaded with accession number MT271829 for ITS, and MT889638 for LSU, representing the first entry for this species in GenBank. The BLAST comparison of the ITS and LSU sequences obtained from our specimens did not show high similarity against any of the sequences contained in GenBank. In particular, the comparison of the ITS sequence showed the percent identity against other Pseudobaeospora sequences to be very low; the closest

Discussion
The species is highly distinctive in terms of both macro-and micro-morphological features: the combination of pale gills, vivid bluish-purple pileus, green reaction of the pileipellis to KOH and the presence of cheilocystidia readily separates P. cyanea from other European species (Arnolds et al. 2003). Our observations are fully consistent with the original description, allowing for a confident identification of our collection. All three previous findings were in Spain (Arnolds et al. 2003;Arauzo 2011); the type specimens were collected in November on Mediterranean hills with Pinus pinaster, Arbutus unedo and Erica arborea (Arnolds et al. 2003). This Italian report, similar in habitat and season of growth, widens remarkably the known area of occurrence of this species, raising questions about its possible presence in other areas of Western Europe in which suitable habitats are present.
The original identification of the species is based on macro-and micro-morphological features, and no genetic data are available yet from the holotype specimens and from the material of the Basque collections. Since there are very few sequences of Pseudobaeospora available it is difficult to establish the taxonomic position of this species within the genus. The phylogenetic tree based on ITS (Fig. 3) suggests that P. cyanea is closely related with P. lilacina; while the LSU tree (Fig. 4) suggests that P. cyanea is part of a clade that also comprises Pseudobaeospora lilacina, P. pyrifera, and P. wipapatiae. The difference between the two trees is due to the fact that the deposited ITS and LSU sequences do not always come from the same samples. Other taxa that show morphological affinity with the species, such as P. dichroa Bas, P. pallidifolia Bas, A. Gennari & Robich, P. jamonii Bas, Lalli & Lonati, and P. laguncularis Bas, could not be included in the phylogenetic tree because no sequences are currently available on the public database for any of them.
The species mentioned above share with P. cyanea several morphological features, like the coloured basidiomata, the greenish to lilac reaction of the pileipellis to KOH (except P. lilacina), the non hymenidermoid nature of the pileipellis (except P. wipapatiae) and the presence of clamp connections (Bas 2003;Schwarz 2012;Desjardin et al. 2014;Wu et al. 2017). Considering the intra-generic morphogroups proposed by Bas, based on basidioma coloration and micro-morphological features, P. cyanea fits in the "Pyrifera group", that includes P. pyrifera, P. jamonii, and P. laguncularis, grouped by the presence of cheilocystidia (Bas 2003). The species characterized by the absence of cheilocystidia, like P. lilacina, P. deckeri, P. dichroa, and P. pallidifolia, can be placed in the similar "group" (Bas 2003). Although these groups are deemed probably artificial by Bas himself, our molecular investigation indeed supports a rather close relationship between P. cyanea and P. pyrifera. This is in contrast with Voto's intra-generic arrangement, which separates these two species. Indeed, Voto (2009) placed P. cyanea in sect. Anistoderma Voto and P. pyrifera in sect. Pseudobaeospora Singer, based exclusively on the differences in the structure of the pileipellis.
The genus Pseudobaeospora includes several new species described in the recent past with a controversial position within the Tricholomatoid clade of Agaricales (Desjardin et al. 2014;Wu et al. 2017). Many unanswered questions also remain on its ecology, the distribution of its species, and their phylogenetic relationships. The scarcity of species with available genetic data is a liability to the definition of the phylogeny and intra-generic arrangement of Pseudobaeospora, and it is advisable to promote the sequencing of more species in the future.