Rediscovery and ITS-barcoding of Entonaema cinnabarinum (Xylariales, Ascomycota) from Bulgaria

: Entonaema is a peculiar and morphologically easily recognised fungal genus with disjunct and predominantly tropical-subtropical distribution. Entonaema cinnabarinum is typified on material from Australia and is the sole known species in Europe. It was first reported on this continent three decades ago, based on a collection from Bulgaria and remains so far one of the utmost rare European fungi with merely five sites known on the continent. After diligent search, the Bulgarian Entonaema was rediscovered in the area where it was first collected, and new sites were also found. Two ITS rDNA sequences of Bulgarian collections were obtained, appearing to be the first verified accessions of European origin. In the phylogenetic inference they appear closely related to an accession from South Korea, dubbed E. splendens . The outcomes of the phylogenetic analysis confirm the self-standing status of Entonaema, but its precise affiliation within the order Xylariales remains to be assessed further. Morphological characterisation, ample macroscopic and microscopic illustrations, as well as SEM images of ascospores of the new Bulgarian findings are included.


Introduction
The genus Entonaema A. Möller is easily recognised in the field due to the azonate, filled with liquid and often conspicuously coloured ascostromata (Rogers, 1981;Stadler et al., 2008a). Its members have a mostly subtropical and tropical distribution (Rogers, 1981;Stadler et al., 2008a;Fedosova, 2012). The latest treatment, based on polyphasic approach, recognises six species across the globe (Stadler et al., 2008a).
Entonaema cinnabarinum (Cooke & Massee) Lloyd was described from Australia (Cooke, 1887) and is by far the only known European member of the genus (Rogers, 1981;Stadler et al., 2004Stadler et al., , 2008a. It was first reported on this continent from Bulgaria by Benkert (1993), who found it several years earlier at the estuary of Kamchia River. Shortly after, another finding from the same area was described and illustrated by Laessøe (1997). Apart of the Bulgarian site, merely four more findings are known in Europe, namely in France, Hungary, Russia and Spain (Stadler et al., 2004;Rubio Domínguez & Menédez Valderey, 2011;Fedosova, 2012;Fintha et al., 2019). A sequence of this species was obtained by Triebel et al. (2005) from a culture of a French collection. It was however found later that the specimen from which this sequence was derived, was contaminated by a Daldinia species . Two other sequences (AM292043, AM292044) initially released as E. cinnabarinum by Šrůtka et al. (2007) were also reassessed later to belong to a species of Daldinia Historia naturalis bulgarica 45 (2023) Ces. & De Not. (Pažoutová et al., 2013). Reference sequences of E. cinnabarinum are thus not available in public databases at present.
Being aware of the Bulgarian finds of Benkert and Laessøe, the first author searched continuously during his field trips to obtain fresh collections of this interesting species and indeed such were found in the autumn of 2014 at distant sites on the Bulgarian Black Sea coast. Additional observations were conducted in 2015, 2021 and 2022, and further specimens were collected in 2021 and 2022. Having in hands abundant material, we sought to obtain ITS rDNA sequences as to provide phylogenetic data and detailed morphological description of those rare collections. The results are communicated in this paper.

Materials and methods
The collections were photographed in-situ. Air-dried voucher specimens are deposited in the Mycological Collection of the Institute of Biodiversity and Ecosystem Research of the Bulgarian Academy of Sciences (SOMF) or in the personal collection of M. Slavova (referred to as 'MSL'). The description is based solely upon the materials studied in this paper. Colours in the description refer to the chart of Kornerup & Wanscher (1978) as far as possible; colour names from this chart are followed by respective codes in parenthesis, beginning with the prefix 'K&W'. Colours lacking such codes are vernacular notations rather than chart entries.
Study of microscopic features was held with Am-Scope T360B light microscope, equipped with Am-Scope MU900 digital camera. All microscopic observations were done on dried material. Tap water was preferred as a mounting medium for slides preparation, but KOH (10%), IKI and Melzer's reagent were also used for testing colour reactions or to better visualise microscopic structures. Cotton blue in lactophenol was employed (with all due caution) to observe the germination slit of the spores. Ascospores were always measured in water on slides from spore deposits on stromatal surface. Measurements were taken with Piximetre v. 5.10 on digital images. Scanning electron microscopy (SEM) was performed with JEOL-JSM-5510. Samples for SEM were obtained from spore deposits on stromatal surface of specimen SOMF30878. These were transferred onto double-sided adhesive tape and sputter-coated with gold prior to observation.
Sequence of the rITS region of one of the specimens was obtained following the protocol presented in details in Bozok et al. (2020). Forward and reverse reads were produced. These were visualised, assembled into consensus sequences and edited upon necessity in Sequencher version 5.4.5 (Gene Codes, Ann Arbor, MI, USA). The sequencing of another specimen was outsourced to ALVALAB (Spain). The protocol communicated by the laboratory was already presented in another recently published work (Assyov 2022). The new sequences were deposited in Gen-Bank. Their accession numbers are listed after the respective voucher specimens in the text below.

Analysis of rITS sequences
Two sequences were successfully obtained from specimens from distant Bulgarian localities. They are identical in pairwise comparison. The initial BLAST search retrieved as closest match a sequence from  Tul.. The two sequences from the Bulgarian specimens of E. cinnabarinum group together with the sequence KR673521, labelled as E. splendens, forming a fully supported clade, which is sister to similarly fully supported clade, containing two North American sequences, identified as Entonaema liquescens Möller. Those two lineages further cluster together with an accession of an unidentified Australian species of Entonaema into a statistically supported clade (branch support 0.86), which is related to the clades containing sequences of Diatrypaceae and Xylariaceae. Further on, the previously mentioned French sequence of culture of E. cinnabarinum, contaminated with Daldinia childiae J.D. Rogers & Y.M. Ju (cf. Stadler et al., 2014) groups with another sequence identified as E. cinnabarinum (KJ572193) and with a sequence labelled as Entonaema sp. (AB495010) forming a statistically supported clade (branch support 0.89), which nests deeply into a clade containing sequences of various species of Daldinia, itself placed into the Hypoxylaceae. This former clade also includes two sequences, originally labelled as E. liquescens. In respect of those two accessions, it must be noted that Stadler et al. (2020) and Wibberg et al. (2021) (Stadler et al., 2008a).

Discussion
It was shown by the phylogenetic analysis that the sequences of the Bulgarian Entonaema collections cluster together in a well-supported clade with an Asian sequence, originally identified as E. splendens. The studied here specimens however, could not be identified with E. splendens, which at present is believed to be conspecific with E. liquescens, based on HPLC profiles (Stadler et al., 2008a). A few characteristic traits of this latter species are not consistent with the morphology of our specimens, namely the dull yellow or olive yellow colouration at maturity, the greenish bruising on handling, and the green subsurface granules (Stadler et al., 2008a). The use of E. splendens for the sequence KR673521  is thus very likely a misapplication, but its existence is important as it presents an additional evidence for the wide transcontinental distribution of E. cinnabarinum. Following the identification key in Stadler et al. (2008a) our specimens would key out confidently to E. cinnabarinum due to the orange extractable in KOH pigments and the orange to red subsurface granules. Further on, the outcomes of the analysis are congruent with the conclusion of  that the previously available sequences labelled as E. cinnabarinum in fact belong to a species of the genus Daldinia. As already mentioned above, Stadler at al. (2020) and Wibberg et al. (2021) also questioned the authenticity of culture ATCC 46302 of E. liquescens, based on genomic data, which failed to reveal presence of genes related to the assembly of mitorubrin-type compounds. The latter work also found that phylogenetic reconstructions place sequences of this culture into Daldinia. In our tree the reference ITS sequence KY610389, originally accessioned as E. liquescens, clusters similarly in a well-supported clade containing species of Daldinia. We thus provide the first genuine sequences of E. cinnabarinum of European origin. While the phylogenetic analysis in the present paper supports the self-standing status of the genus Entonaema, its phylogenetic position remains not precisely resolved as for the limitations of analysis of nrITS sequences and the relatively limited number of taxa, included in the dataset used for the present study. Our inference places the sequences of Entonaema somewhat close to some members of the family Xylariaceae, while they appear not that closely related to the species of Hypoxylaceae, included in the analysis. However, it was previously shown that E. cinnabarinum, E. globosum and E. liquescens contain pigments of mitorubrin/rubiginosin-type, not detected in studied species of Xylariaceae, while they lack xylarals that are present in a number of species of Xylaria (Stadler et al., 2004(Stadler et al., , 2008a. They are also known to lack binaphtalenes that are evidenced in some studied members of Hypoxylaceae (Stadler et al., 2008a), but on the other hand, mitorubrin and rubiginosins are not unknown in some groups of Hypoxylon (Stadler et al., 2008b;Wibberg et al., 2021). This seems consistent with our phylogenetic inference, which resolved Entonaema as a wellseparated lineage, but does not correspond to the inferred closer relation to the members of Xylariaceae. Studies employing other DNA-regions will hopefully establish the precise position of the genus Entonaema within the order Xylariales in future.
On account of the morphological features of the Bulgarian collections, they are essentially consistent with the contemporary descriptions in Stadler et al. (2008a) and Fedosova (2012). One difference, which attracted our attention, was the germ slit of the ascospores. This proved to be difficult to observe, but it was visualised in lactophenol cotton blue, Melzer's reagent, and occasionally in KOH and water. Where seen, it seemed to run almost through the entire spore length, while Stadler et al. (2008a) and Fedosova (2012) described it as 2/3 of the ascospore. We are at the moment uncertain as to how conclusive this difference may be. Further on, Fournier & Magni (2004) described the colours in the French collections as "sienna, rust, bay to dark brick", later affirmed also in Stadler et al. (2008a). The Bulgarian specimens deviate from this colour pattern, often featuring notable orange, pinkish and reddish tinges (Fig. 2). In addition, it seems that we report for the first time observations in SEM of the ascospores of an Entonaema-species, which appear to be smooth. It may be nevertheless worthy performing SEM-studies on other taxa of the genus, as it is now well-known that the ascospore ornamentation is a valuable taxonomic character in the related genus Daldinia .
Previous research of pigments in Entonaema already pointed out the existence of some differences between the profiles of the holotype of E. cinnabarinum on one hand and the studied by the authors Bulgarian and French collections, on another (Stadler et al., 2004). Regrettably, for the moment we could not conduct an HPLC study of the Bulgarian specimens and a single sequence of an unidentified Entonaema from Australia exists on GenBank (KP311453). It is apparent that the lack of typederived sequences is at present hampering the research in the genus Entonaema. Further DNA studies, ideally of the holotype of E. cinnabarinum, or of representative, topotypic Australian collections, congruent with the type, are essential and could provide more conclusive answers about the taxonomic status of the European populations of Entonaema. There is no doubt that further effort to obtain sequences from more European collections will be also valuable.
The so far known habitats of E. cinnabarinum in Bulgaria are all situated in immediate vicinity to the coast of the Black Sea. The majority of the sites are in seasonally flooded riparian forest with exception of two, which are in mixed broadleaf non-flooded forests, but still not far from riparian habitats. The fungus is invariably confined to F. angustifolia, occurring on fallen trunks and large branches, almost always with bark still present on them. The findings of Benkert (1993) and Laessøe (1997) were also reported to have been related to this tree. Production of stromata apparently spans at least from July until November (Benkert, 1993;Laessøe, 1997; and the observations in this paper). The recent Hungarian finds were also documented as occurring on F. angustifolia (Fintha et al., 2019). The known French collections of E. cinnabarinum, were found on F. excelsior L., but also on Acer negundo L. and Platanus sp. (Fournier & Magni 2004). The recent Spanish findings, in contrast, were documented on Corylus avellana L. (Rubio Domínguez & Menédez Valderey, 2011). The collection of Fedosova (2012) from the eastern coast of the Black Sea was found on a dead trunk of an unidentified broadleaf tree. Judging from the habitat requirements of E. cinnabarinum in Bulgaria, we anticipate that it may also occur in similar riparian habitats, which are present in coastal areas in the neighbouring Romania and Türkiye. The reporting of possible future well-documented findings in those two countries is of undoubted interest.

Acknowledgments
The work of the first author was supported by the project "Phylogeny, distribution and sustainable use of fungi". Thanks are due to Dr Pablo Alvarado (ALVALAB, Spain) for obtaining one of the sequences used in this work. The authors extend their thanks to the two reviewers, whose comments and suggestions helped to substantially improve the manuscript.