Allophlebia, a new genus to accomodate Phlebia ludoviciana (Agaricomycetes, Polyporales)

Allophlebia is proposed as a new genus in Meruliaceae based on morphological characters and molecular data. The genus, so far monotypic, is typified by Peniophora ludoviciana and the new combination A. ludoviciana is proposed. The type species is characterized by a resupinate basidioma, a monomitic hyphal system with clamp connections, two types of cystidia (leptocystidia and metuloids), clavate basidia, and hyaline, thin-walled and ellipsoid basidiospores. A phylogeny for Allophlebia and related taxa was inferred from ITS and nLSU rDNA sequences and new information on the geographic distribution of A. ludoviciana is provided.


Introduction
Phlebia Fr. (Polyporales, Meruliaceae) was described by Fries in 1821 and intended for species with a hymenium composed of irregular veins and ridges. Fries (1828) pointed to P. radiata as the most typical member of his new genus and this species is now generally accepted as the type (Donk 1957). Species in Phlebia sensu lato usually have resupinate basidiomata that are ceraceous to subgelatinous in fresh specimens, and with a membranous, firm ceraceous, corneous, or coriaceous consistency when dried. The hymenial surface varies from smooth, tuberculate, odontioid, merulioid to poroid. The hyphal system is monomitic, rarely dimitic, with hyphae clamped and embedded in a more or less evident gelatinous matrix. Cystidia can be present or absent; basidia are clavate, narrow, with a basal clamp and disposed in a dense palisade; and basidiospores are allantoid to ellipsoid, smooth, thin-walled, IKI−, and CB− (Eriksson et al. 1981;Bernicchia and Gorjón 2010). All species analyzed are saprobes on decaying wood (Nakasone 1990).
The original concept for Phlebia was considerably broadened along the years (Donk 1931, 1957, Nakasone 1991, 1996, 1997, 2002, Nakasone and Burdsall 1984. However, this wide concept for Phlebia proved to be polyphyletic (Larsson et al. 2004, Binder et al. 2013, Floudas and Hibbett 2015, Justo et al. 2017. Several genera have been introduced or resurrected to accommodate different species of Phlebia, e.g., Cabalodontia Piątek, Crustodontia Hjortstam & Ryvarden, Cytidiella Pouzar, Hermanssonia Zmitr., Jacksonomyces Jülich, Mycoacia Donk, Mycoaciella J. Erikss. & Ryvarden, Phlebiopsis Jülich, Scopuloides Hjortstam & Ryvarden, and Stereophlebia Zmitr. Other Phlebia species have been moved to other genera, most notably to Crustoderma Parmasto and Skvortzovia Bononi & Hjortstam. After such removal and transfer of species and after adjustments for synonyms, the genus still holds around 100 species, many of which are based on names for which there are no modern interpretation (www.mycobank.org). According to molecular data, P. radiata together with many other Phlebia species belong in Meruliaceae in Polyporales (Justo et al. 2017), while a few are recovered in Hymenochaetales (Larsson et al. 2006). During studies of corticioid fungi from northeast Brazil, specimens of Phlebia ludoviciana (Burt) Nakasone & Burds. were collected. Molecular phylogenetic analyses showed that this species could not be placed in any of the corticioid genera already described. Thus, the aims of this paper were to describe a new genus for P. ludoviciana and to discuss the geographical distribution of this species. Specimens were identified based on macro-(measures, texture, consistency, shape, and color of the basidiomata) and micro-morphology and sections of the basidiomata were checked with 3% potassium hydroxide solution (KOH), stained with 1% aqueous phloxine. Melzer's reagent and Cotton Blue were used to analyze, respectively, dextrinoid and amyloid (IKI+/IKI−), and cyanophilous (CB+/CB−) reactions of the microstructures. Presence/absence of sterile structures and basidiospores was noted and measurements of at least 20 of them were taken, when possible (Hjortstam et al. 1987;Watling 1969). The material was deposited in the Herbarium Pe. Camille Torrend (URM), Departmento de Micologia (UFPE), and in the Herbarium of the University of Oslo (O).

DNA extraction, PCR amplification, and sequencing
Basidiomata fragments (30-50 mg) were removed, placed in tubes of 1.5 ml, and stored at -20°C until DNA extraction. The method of DNA extraction followed Goés-Neto et al. (2005) and the reaction mix and parameters for PCR reactions of the ITS and LSU regions followed Smith and Sivasithamparam (2000), using the primer pairs ITS4-ITS5 and LR0R-LR5, respectively (White et al. 1990;Moncalvo et al. 2000;Lima-Júnior et al. 2014). The purification of PCR products was done with ExoSAP-IT™ PCR Product Cleanup Reagent (Thermo Fisher Scientific, USA), following the manufacturer's recommendations. The samples were sequenced at the Plataforma Tecnológica de Genômica e Expressão Gênica do Centro de Biociências (CB), UFPE, Brazil, or sent to Stab Vida Lda (Madan Parque, Caparica, Portugal). The cycle sequencing was carried out with the same primers used for PCR reactions (Moncalvo et al. 2000). All new sequences were deposited in GenBank (National Center for Biotechnology Information, Bethesda, MD, USA).

Phylogenetic analyses
The 2.0 Staden Package software was used for analyses and edition of electropherograms (Bonfield et al. 1995). These sequences were subjected to BLASTn search in NCBI to recover similar sequences from GenBank and used in the dataset to establish phylogenetic relationships (Table 1). Each gene region was aligned with the MAFFT v.7 online server using default settings (http://mafft.cbrc.jp/alignment/server/), then improved manually using MEGA 7.0 and combined to form the concatenated dataset (Kumar et al. 2016).
The ITS and LSU regions were first analyzed independently (data not shown). Since no important topological differences were detected, the regions were combined into a single matrix for the final analyses. The models of evolution were obtained from MEGA 7.0 (Kumar et al. 2016) and confirmed in TOPALi v2.5 (Milne et al. 2008) for each dataset. Phylogenetic analyses and tree construction were performed using maximum likelihood (ML) and confirmed in Bayesian algorithm (BA). ML analysis was performed using MEGA 7.0 (Kumar et al. 2016) with 5000 bootstrap replications and based on GTR + G + I model. BA analyses were run in TOPALi v2.5 (Milne et al. 2008) with 5×10 6 generations, also based on GTR+G + I model. Statistical support for branches was considered informative with Bayesian posterior probabilities (BPP) ≥0.95 and bootstrap (BS) values ≥70%. The trees were visualized with FigTree (Rambaut 2014) and the final layout was made in Adobe Illustrator CS6.

Results
Five specimens were sequenced (URM 93082, URM 93251, URM 93329, O-F-110340, O-F-110341), generating five ITS and four LSU sequences (Table 1). These were combined with ITS and LSU sequences selected through BLAST searches against GenBank.
No strongly supported topological conflict was detected among the datasets analyzed (ITS, LSU, and ITS+LSU). Thus, only the combined analysis is presented here, performed mainly with ITS sequences since only that region is available for some key specimens. The combined dataset included 174 sequences (116 ITS and 58 LSU) and comprised 2138 characters including gaps. Climacocystis borealis (Fr.) Kotl. & Pouzar and Junghuhnia nitida (Pers.) Ryvarden were selected  The results of the phylogenetic analyses generated from ML and BA showed similar tree topologies and small or insignificant differences in statistical support values. Thus, the ML tree with bootstrap support values (BS) and posterior probabilities (PP) from the BA analysis was used to show the results of this study (Fig. 1).
The newly generated sequences were placed in a strongly supported clade (BS 99%, PP 0.99) with several samples of A. ludoviciana previously deposited in GenBank. Other sequences at GenBank identified differently also grouped in the same clade. The A. ludoviciana clade was phylogenetically separated from the clade representing Phlebia s.s, and from other described genera (Fig. 1).

Discussion
When combining Peniophora ludoviciana to Phlebia, Nakasone et al. (1982) grouped this species with P. brevispora, P. subochracea, and P. subserialis in section Leptocystidiophlebia Parmasto based on morphology and culture characteristics. Our results show that P. ludoviciana is phylogenetically close to P. subochracea, while P. brevispora and P. subserialis are distantly related, both f r o m e a c h o t h e r a n d f r o m P . l u d o v i c i a n a a n d P. subochracea (Fig. 1). Floudas and Hibbett (2015)   identified as P. subserialis were placed in three different clades, one corresponding to A. ludoviciana and sister to P. subochracea, one close to P. nothofagi and P. fuscoatra, currently belonging to Mycoacia, and the last one belonging to the Phanerochaete clade and provisionally identified as Phanerochaete 'krikophora'. Justo et al. (2017) recovered P. ludoviciana (FD-427, reported as Phlebia sp. in GenBank) in a clade with P. subochracea I (HHB8715) reported as Phlebia cf. subserialis in GenBank), both representing A. ludoviciana and sister to P. subochracea II. In our study, the Allophlebia clade is phylogenetically separated from the Phlebia s.s. clade (BS=87/PP=0.96) as well as from other genera in Meruliaceae and from other sequenced species of Phlebia recovered outside Meruliaceae. It is strongly supported as a monophyletic group (99) (Fig. 1), and in accordance with the recommendations by Vellinga et al. (2015). The new genus may also include Fungal sp. (TP2) f r o m T h a i l a n d ( K l o m k l i e n g e t a l . 2 0 1 4 ) and P. ochraceofulva (FBCC295) from Sweden (Kuuskeri et al. 2015), but they represent isolates without vouchers, which prevents morphological studies.
The five sequences of A. ludoviciana generated in our study clustered with two sequences from the USA and French Guiana and  originally collected on Salix humboldtiana in Argentina, is characterized by membraneous basidiomata and one kind of cystidia, viz. strongly encrusted metuloids projecting beyond the hymenium (Rajchenberg and Wright 1987). The type of P. subserialis is from France and sequences from there and other European countries, as well as one sequenced specimen from India (Table 1), are distantly placed in the phylogenetic tree (Fig. 1). Phlebia subserialis has narrower leptocystidia (3-4 μm), lacks encrusted cystidia, and has longer, suballantoid basidiospores [6-7(-8) × 2-2.5 μm] (Bernicchia and Gorjón 2010). It is unclear why this species has been confused with A. ludoviciana. One reason could be that some early mycologists established an opinion that the two cystidia types in A. ludoviciana are just a single type in different stages of development (Rogers and Jackson 1943). Specimens of P. subserialis reported in the Americas should be reevaluated (Nakasone et al. 1982). Grammothelopsis puiggarii is a species characterized by large, angular pores (1-2 per mm), large, dextrinoid, thick-walled basidiospores and dextrinoid skeletal hyphae (Rajchenberg and Wright 1987). This species is currently placed in Polyporaceae and cannot possibly be confused with A. ludoviciana. The sequences named G. puigguarii are most likely the result of contamination or sequencing mistakes.
The specimens of A. ludoviciana studied by Nakasone et al. (1982) were all collected on dead wood of various angiosperm tree species. The specimens sequenced by us and by earlier studies were also all collected on decaying angiosperm wood. The environmental sequences of A. ludoviciana in GenBank were mostly generated from living tissue of angiosperm plants representing the genera Elaeis (oil palm), Hevea, Polylepis, Phragmites, Rubia, and Solanum. Sequences were also generated from the rhizosphere of Broussonetia, from Nyssa rail ties, dry grassland soil, air, and from nests of Atta and Cyphomyrmex ants (Fig. 1). This information adds to the growing body of evidence indicating that basidiomycetes with a saprophytic lifestyle may serve also other ecological functions (Pinruan et al. 2010;Martin et al. 2015).