Phylogenetic position and taxonomy of Kusaghiporia usambarensis gen. et sp. nov. (Polyporales)

ABSTRACT A large polyporoid mushroom from the West Usambara Mountains in North-eastern Tanzania produces dark brown, up to 60-cm large fruiting bodies that at maturity may weigh more than 10 kg. It has a high rate of mycelial growth and regeneration and was found growing on both dry and green leaves of shrubs; attached to the base of living trees, and it was also observed to degrade dead snakes and insects accidentally coming into contact with it. Phylogenetic analyses based on individual and concatenated data sets of nrLSU, nrSSU and the RPB2 and TEF1 genes showed it, together with Laetiporus, Phaeolus, Pycnoporellus and Wolfiporia, to form a monophyletic group in Polyporales. Based on morphological features and molecular data, it is described as Kusaghiporia usambarensis.

The "antrodia clade" was first identified by Hibbett and Donoghue (2001) and currently more than 26 genera are recognised in this clade (Ortiz-Santana et al. 2013). Members in the "antrodia clade" are of economic importance as a source of food, and also of pharmaceutical and biotechnological products. However, it also contains species that are plant pathogens detrimental to forests and forest plantations (Dai et al. 2007;Banik et al. 2010). The "antodia clade" is morphologically diverse and includes species that have resupinate, stipitate or pileatebasidiomata that are either annual or perennial; the hyphal system is monomitic, dimitic or trimitic; the basidiospores are hyaline thin-to thick-walled, subglobose to cylindricaland they cause brown rots (Ryvarden and Melo 2014).
The "antrodia clade" has been widely studied and additional genera have been suggested to belong there. Recent taxonomic and phylogenetic studies, including that of Lindner and Banik (2008), have presented molecular phylogenies of the clade. In a study of Laetiporus and other polypores, Banik et al. (2010) inferred relationships among North American and Japanese Laetiporus isolates; Ortiz-Santana et al. (2013) presented a phylogenetic overview of the "antrodia clade" and Binder et al. (2013) used genomic data and a six-gene data set for evaluating phylogenetic relationships in Polyporales. Further studies include those of Han et al. (2014) in which two new species of Fomitopsis from China were described, and Zhao et al. (2015) used a multi-gene dataset to support the recognition of Fragiliporiaceae, a new family of Polyporales. Han et al. (2016) offered a study of the phylogeny of the brown-rot fungi, including Fomitopsis and related genera, while Justo et al. (2017) revised the phylogeny of Polyporales at family-level.
A mushroom locally known as "Kusaghizi" has a long tradition of being used as food by local communities in the Usambara mountains as first reported by Powell et al. (2013). In a study by Juma et al. (2016), which assessed antioxidant activities of saprobic mushrooms from Tanzania, "Kusaghizi" was included, but neither the study by Powell et al. nor that of Juma et al. reported a scientific name for "Kusaghizi". Here we aim to describe this species and infer its phylogenetic position.

Material
Material of the fungus locally named "Kusaghizi" was collected during the rainy seasons in February 2016 and March 2017 close to the villages Bungu, Buti and Makuri in the Usambara Mountains. These villages are located in the Korogwe District of the Tanga region, Tanzania. The west Usambara Mountains are part of the "Eastern Arc" of ranges in eastern Tanzania, from the Taita Hills in Kenya to the Udzungwa Mountains in southern Tanzania. The samples were examined in a fresh condition for macro-morphological features including colour changes upon cut, bruising and exposure to air. A fruit body was divided into two parts; one was sun dried for 5 days while the remaining part was stored in a freezer at −20°C for further investigations.

Morphological characterisation
Basidioma colours of the holotype were indicated according to Kornerup and Wanschern (1967). Photographs of the fruit body were taken before and after removing it from its substrate. Microscopic characterisation was done from preparations of a rehydrated specimen sectioned with a freezing microtome and stained with Lactic Blue, or treated with 10% KOH and Melzer's reagent.
A total of 40 mature basidiospores were randomly selected and measured. Statistical averages were used to estimate the observed features as follows: A L = mean spore length (arithmetic mean of the length of spores); A W = mean spore width (arithmetic mean of the width of spores); Q = A L /A W ratio; n (a/b) = number of spores (a) measured from given number (b) of specimen. Melzer's reagent was used where IKI+ = Melzer's reagent positive; IKI− = both inamyloid and indextrinoid.

DNA extraction, amplification and sequencing
Total genomic DNA was extracted from both fresh and dried material and kept at −20°C following the protocol of the plant Genomic DNA extraction kit (E. Z.N.A. Fungal DNA Mini Kit Protocols). Diluted samples (10 -1 ) of DNA were used for PCR amplification of the nrLSU, nrSSU, RPB2 and TEF1. Primers LR0R, LR7, LR5 were used for nrLSU (Vilgalys and Hester 1990), and PNS1 and NS41 for nrSSU (Hibbett 1996). PCR conditions for nrLSU and nrSSU were: initial denaturation for 4 min at 95°C, followed by 35 cycles of 1 min at 94°C, 1 min at 54°C, 45 s at 72°C, and a final elongation for 5 min at 72°C. The RPB2 region was amplified using degenerated primers fRPB2-5f and RPB2-7.1R (Matheny 2005). For amplification of TEF1the EF1-983Fand EF1-1567R primers were used (Rehner and Buckley 2005). Touchdown PCR was used with an initial annealing temperature of 66°C following the protocol of Rehner and Buckley (2005). The PCR products were visualised by electrophoresis on 1.5% agarose gels. Products were purified using Illustra ™ ExoStar buffer diluted 10×, following the manufacturer's protocol. Sequencing was carried out by Macrogen.

Data analyses
Sequences from GenBank were selected based on their quality and with an intention of wide coverage of Polyporales and the "antrodia clade" as in Zhao et al. (2015) and Han et al. (2016) respectively. The sequences produced in this study were aligned along with those downloaded from GenBank (Table 1) using MAFFT v. 7 (http://mafft.cbrc.jp/alignment/server/) and manually adjusted using AliView (Larsson 2014). Ambiguously aligned regions were excluded from the analyses. For RPB2 and TEF1 only coding parts of the sequences were used for the analyses. The concatenated data matrix of Polyporalesand the "antrodia clade" contained 4940 and 3760 unambiguously aligned sites respectively. All alignments were based on the nucleotide sequences with each gene analysed separately.
Single-gene analyses were performed to detect significant conflicts among datasets. A conflict among single-locus datasets (nrLSU, nrSSU, RPB1, TEF1) was considered significant if a well-supported monophyletic group, for example posterior probability (PP) ≥ 0.95, was found to be well supported as non-monophyletic when different loci were used. No significant incongruence among the single-gene trees was detected (Supplementary Figures S1A, S1B S1C and S1D), hence the four matrices were concatenated.
Further analyses were carried out after concatenation using Sequence Matrix (Vaidya et al. 2011).
The best-fit model of DNA evolution for the analyses, for both individual codon positions and genes, was obtained using the Akaike Information Criterion as implemented in MrModeltest 2.3 (Nylander 2004). For the Polyporales dataset the GTR+I + G model was employed across sites for nrLSU, nrSSU, and for the 1st and 2nd codon for RPB2. For the 1st and the 2nd codon for TEF1 the model F81 + I + G was applied, while GTR + G was implemented for both the 3rd codon of RPB2 and TEF1. For the "antrodia clade" dataset the GTR +I + G model was employed across sites for nrLSU, nrSSU, for all three codons for RPB2, and the 2nd codon for TEF1. For the 1st codon for TEF1 the F81 + I model was applied while the HKY + I + G model was implemented for the 3rd codon for TEF1. Bayesian Inference was conducted with MrBayes 3.2.6, and branch support was estimated by PP (Ronquist and Huelsenbeck 2003). Four Markov chains were run for 2 runs from random starting trees for 10 million generations, trees were sampled every 100 generations and 25% were discarded as burn-in.
Maximum likelihood estimates were carried out by RAxML v.8.2.10 using the GTR +G + I model of site substitution (Stamatakis 2014). The branch support was obtained by maximum likelihood bootstrapping (MLbs) of 1000 replicates (Hillis and Bull 1993).

Phylogenetic analyses
Analyses were based on a total of 209 sequences representing 201 species of Polyporales, with two russuloid species as out-group. The phylogeny of the Polyporales and the position of the "Kusaghizi" was inferred from four datasets: 36 nrLSU sequences, 25 nrSSU sequences, 26 RPB2 sequences and 22 TEF1 sequences. The Polyporales concatenated dataset (Figure 1) contained 100 sequences of 34 nrLSU, 21 nrSSU, 23 RPB2 and 22 TEF1. Further analyses included members of the "antrodia clade" (Figure 2) containing 145 concatenated sequences of 46 nrLSU, 33 nrSSU, 32 RPB2 and 35 TEF1. Maximum likelihood and Bayesian analyses of these datasets were undertaken, first separately and then also of the concatenated dataset. The analysis of the concatenated Polyporales dataset retrieved a phylogeny with five distinct clades (Figure 1) in addition to Stereum hirsutum and Heterobasidion annosum, as outgroup. Clade annotations follow Zhao et al. (2015). Kusaghiporia usambarensis was found to belong in the "antrodia clade". The annotation of the concatenated phylogeny of the "antrodia clade" (Figure 2) follows Han et al. (2016).

Etymology
Kusaghiporia refers to the sambaa name of the mushroom "Kusaghizi", which means the collector or accumulator, and -poria (Lat.): with pores; usambarensis (Lat.): referring to the Usambara mountain range.

Rot type
Brown rot.

Host
Found growing at the base of the trees Maesopsis eminii and Ficus natalensis.
The BLAST results from GeneBank (NCBI, from 2017-10-16), using blastn with the program "discontiguous megablast" (for cross-species comparison, searching with coding sequences) with Kusaghiporia sequences, showed a highest sequence similarity for all four genes with Laetiporus sulphureus. Based on RPB2 and TEF1 they were: Query cover 99% and Ident. 87%, and Query cover 99%, Ident. 84% respectively. For nrSSU the highest similarity has a Query cover of 91% and Ident. 87%; while for nrLSU the Query cover was 100% and the Ident. 86%. Among the five top scores "Polyporales sp. Kusaghizi", voucher IJV40-2 was found: Query cover only 60% and Ident. 100%. In our opinion the genetic isolation of K. usambarensis as compared to Laetiporus justifies the proposal of a new genus to accommodate the species investigated. The monophyly and strong support of the clade containing K. usambarensis, Laetiporus, Wolfiporia, Crustoderma, Pycnoporellus, and Phaeolus as shown in our phylogeny (Figure 2), also justifies the incorporation of K. usambarensis in Laetiporaceae.

Conclusion
The new genus Kusaghiporia was described based on morphological characters and phylogenetic analyses based on concatenated sequence data from four genes. Kusaghiporia produces large fruit bodies. Together with Laetiporus, Pycnoporellus, Phaeolus, and Wolfiporia it formed a strongly supported clade ( Figure 1) belonging in Laetiporaceae, which is nested in the "antrodia clade". Kusaghiporia is a resource in the local communities of the Usambaras, where it is collected and eaten.