Phylogenetic affinities of the sequestrate genus Rhodactina (Boletaceae), with a new species, R. rostratispora from Thailand

Abstract Rhodactina is a small sequestrate genus in Boletaceae with two described species, R. himalayensis and R. incarnata. Phylogenetic analyses of a three-gene dataset including atp6, tef1 and rpb2 of Rhodactina species along with selected Boletaceae species showed that all Rhodactina species formed a monophyletic clade, sister to the genera Spongiforma and Borofutus in subfamily Leccinoideae with high support. All of the taxa in the clade have a similar chemical reaction in which basidiospores turn purplish, purplish red to violet or violet grey when in contact with potassium hydroxide. The molecular analyses also showed that all Rhodactina specimens collected from Ubon Ratchathani province, northeastern Thailand, belong to a new species. Morphologically, the new species is different from others by having a markedly prominent hilar appendage and a terminal hilum on its basidiospores. Thus, the new species, Rhodactina rostratispora, is introduced with detailed macroscopic and microscopic descriptions and illustrations.


Introduction
The genus Rhodactina Pegler & T.W.K. Young was first described in 1989 with R. himalayensis Pegler & T.W.K. Young, from northwestern India, as the type species. Typical characters of the genus are a whitish to pinkish puffball like basidiomata lacking both stipe and columella, violet brown to purple brown or pale pink to red hymenophore when mature, combined with purplish to purplish red, dextrinoid basidiospores with longitudinal ridges, lack of both clamp connections and cystidia. The genus was originally classified based on morphological characters in the family Gautieriaceae Zeller as the spore ornamentation was originally viewed as similar to the genera Gautieria Vittad and Austrogautieria E.L. Stewart & Trappe (Pegler and Young 1989). In 2006, the second species, R. incarnata Zhu L. Yang, Trappe & Lumyong was described and the known distribution of R. himalayensis was extended to Chiang Mai Province, northern Thailand. Based on the phylogenetic analyses of atp6 sequences, the genus was moved to the family Boletaceae Chevall (Yang et al. 2006). However, the phylogenetic affinities of Rhodactina within the Boletaceae remained unclear because of very limited taxon sampling. So, at present, there are only two described Rhodactina species, R. himalayensis and R. incarnata (http://www.indexfungorum.org/Names/Names.asp), both of which have been reported to occur in northern Thailand (Chandrasrikul et al. 2011).
Boletaceae diversity seems to be high in Thailand (Chandrasrikul et al. 2011), with several new species described in the last five years (Choeyklin et al. 2012, Halling et al. 2014, Neves et al. 2012, Raspé et al. 2016. During this survey of Boletaceae diversity in Thailand, several Rhodactina collections were made and their morphology and phylogenetic relationships were studied. Phylogenetic analyses were based on three genes: atp6, tef1 and rpb2, which have previously been justified as useful for phylogenetic analyses of Boletales (Kretzer and Bruns 1999, Binder and Hibbett 2006, Hosen et al. 2013, Smith et al. 2015, Orihara et al. 2016, Raspé et al. 2016, Wu et al. 2016. Both morphology and phylogenetic analyses confirmed that all newly collected specimens belong to a new species in the genus Rhodactina. Thus, the third species of Rhodactina, found in Thailand, is described and its phylogenetic affinities are presented in this study.

Specimens collecting
The new Rhodactina specimens were collected and photographed from community forests in Trakan Phuet Phon district, Ubon Ratchathani province, northeastern Thailand, in the rainy season during 2015-2017. The specimens were wrapped using aluminium foil or kept in plastic boxes until return to the laboratory and described within 24 h. The specimens were dried in an electric drier at 45-50 °C. The examined specimens are deposited in the herbaria CMUB and BR (both listed in Index Herbariorum; Thiers, continuously updated).

Morphological studies
The macroscopic description was based on detailed field notes and photos of basidiomata. Colour codes followed Kornerup and Wanscher (1978). Macrochemical reactions (colour reactions) of peridium, hymenophore and microscopic structures were determined using 5 % (w/v) aqueous potassium hydroxide, 28-30 % ammonium hydroxide or Melzer's reagent. Microscopic structures were observed from dried specimens, rehydrated in 5% KOH or 1 % ammoniacal Congo red. For each collection, a minimum of 50 basidiospores and 20 basidia were randomly selected and measured at 1000× with a calibrated ocular micrometer using an Olympus CX31 microscope. Spore dimensions include ornamentation. The notation '(n/m/p)' represents the number of basidiospores n measured from m basidiomata of p collections. Dimensions of microscopic structures are presented in the following format: (a-)b-c-d(-e), in which c represents the average, b the 5 th percentile, d the 95 th percentile and extreme values a and e are shown in parentheses. Q, the length/width ratio, is presented in the same format. Sections of the peridium surface were made radially and perpendicularly to the surface, halfway between the centre and the side of basidiomata. All microscopic features were drawn free hand using an Olympus Camera Lucida model U−DA fitted to the microscope cited above. For scanning electron microscopy (SEM), small fragments of dried hymenophore were mounted directly on to an SEM stub with double-sided tape. The samples were coated with gold for 60 seconds using SPI-Module Sputter Coater, examined and photographed at 15-20 kV with a FIB Quanta 200 3D scanning electron microscope (Thermo Fisher Scientific, United States).

DNA isolation, PCR amplification and DNA sequencing
Genomic DNA was extracted from fresh tissue preserved in CTAB or about 10-15 mg of dried specimens using a CTAB isolation procedure adapted from Doyle and Doyle (1990). The genes atp6, tef1 and rpb2 were amplified by polymerase chain reaction (PCR) technique. For the amplification of atp6, ATP6-1M40F and ATP6-2Mprimers were used (Raspé et al. 2016), with the following PCR programme: 2 min at 95 °C; 5 cycles of 45 s at 95 °C, 60 s at 42 °C, 30 s at 72 °C; 35 cycles of 20 s at 95 °C, 30 s at 55 °C, 30 s+1 s/cycle at 72 °C; 3 min at 72 °C. The primers EF1-983F and EF1-2218R (Rehner and Buckley 2005) were used to amplify tef1 and bRPB2-6F and bRPB2-7.1R primers (Matheny 2005) were used to amplify rpb2. PCR products were purified by adding 1 U of Exonuclease I and 0.5 U FastAP Alkaline Phosphatase (Thermo Scientific, St. Leon-Rot, Germany) and incubated at 37 °C for 1 h, followed by inactivation at 80 °C for 15 min. Sequencing was performed by Macrogen Inc. (Korea and The Netherlands) with PCR primers, except for atp6, for which universal primers M13F-pUC (-40) and M13F(-20) were used; for tef1, additional sequencing was performed with the two internal primers, EF1-1577F and EF1-1567R (Rehner and Buckley 2005).

Alignment and phylogeny inference
The sequences were assembled in GENEIOUS Pro v. 6.0.6 (Biomatters) and introns were removed prior to alignment based on the amino acid sequence of previously pub-lished sequences. All sequences, including sequences from GenBank, were aligned using MAFFT (Katoh and Standley 2013) on the server accessed at http://mafft.cbrc. jp/alignment/server/. Maximum Likelihood (ML) phylogenetic tree inference was performed using RAxML (Stamatakis 2006) on the CIPRES web server (RAxML-HPC2 on XSEDE; Miller et al. 2009). The phylogenetic tree was inferred by a single analysis with three partitions (one for each gene), using the GTRCAT model with 25 categories and three Chalciporus species were used as an outgroup. Statistical support of nodes was obtained with 1,000 bootstrap replicates.

DNA analyses
A total of 127 new sequences were generated and deposited in GenBank (Table 1). The alignment contained 157 taxa spread over the entire family Boletaceae and was 2429 characters long (TreeBase number 21933). The authors could not obtain tef1 and rpb2 sequences from R. incarnata (CMU25116) nor rpb2 sequence from R. himalayensis (CMU25117). The specimens were in relatively poor condition and genomic DNA was highly degraded. The 3-gene phylogram indicated that all selected collections of the new taxon R. rostratispora formed a monophyletic group with high bootstrap support sister to R. incarnata within the Rhodactina clade ( Figure 1) Etymology. From Latin "rostrati-" meaning having beaked prow or a solid projection and "spora" meaning spores, referring to the basidiospores having a markedly prominent and large hilar appendage.
Description. Basidiomata small to medium-sized 0.8-2.5(4.5) cm diam., subglobose to ovoid with a rudimentary elongated basal attachment, with greyish white to pale brown rhizoids at the base and going up along the surface of basidiomata to about half of the height. Peridium surface (outer peridium) fibrillose to arachnoid, off-white to pinkish white (7A2-3 to 9A2), dull, moist, cracked in places. Peridium very thin, 0.1-0.2(0.4) mm thick. Hymenophore cartilaginous, completely enclosed, whitish orange to reddish orange (7A3-4 to 8A5-6) at first becoming orangey red to red (9D-E8 to 10D-E8) with age, then dark red when very old, irregular; Stipe-columella absent. Taste fungoid. Odour absent when young, very strongly fruity alcoholic when old.
In one R. rostratispora specimen (S. Vadthanarat 208), abnormal spores were observed. Those spores were elongated, 21-24 × 4-8 µm, thick-walled, narrowly fusiform to bacilliform, with or without longitudinal ridges, more or less constricted in the middle. They were usually found attached to apparently normal basidia with four sterigmata.

Discussion
Morphologically, the new species R. rostratispora is characterised by its ridged basidiospores having a markedly prominent hilar appendage with a terminal hilum, which is not found in other Rhodactina species (Pegler andYoung 1989, Yang et al. 2006). However, ridged basidiospores having a prominent hilar appendage are found in some other sequestrate Boletaceae in the genus Turmalinea Orihara & N. Maek and Rossbeevera, including T. persicina Orihara, T. chrysocarpa Orihara & Z.W. Ge, T. mesomorpha Orihara, Ro. paracyanea Orihara and Ro. cryptocyanea Orihara. The basidiospores of those species have a long pointed hilar appendage 4.5-6 µm (Orihara et al. 2016) but are not as wide as in R. rostratispora (2.5-5 µm long, 3.5-5 µm wide) and also their hilar appendage lacks a terminal hilum. Macroscopically, those species differ from R. rostratispora in that both Rossbeevera and Turmalinea have basidiomata often turning blue to greenish blue when bruised, which has never been reported in any Rhodactina species (Pegler andYoung 1989, Yang et al. 2006). Moreover, the colour of mature hymenophore of Turmalinea and Rossbeevera species are dark brown or blackish brown (Lebel et al. 2012, Orihara et al. 2016) not red or dark red like in Rhodactina.
The phylogenetic analyses also support the placement of the new taxon in the genus Rhodactina, with R. incarnata being the closest species. The phylogenetic tree also showed that Rhodactina is sister to a clade composed of Spongiforma and Borofutus within the subfamily Leccinoideae, with 100% bootstrap support. According to Wu et al. (2016) In the examination of R. rostratispora, it was found that the hymenophore turned dark purplish to greyish violet with 5% KOH. Interestingly, all of the genera in subfamily Leccinoideae that turn purple to violet with aqueous KOH solution, namely Rhodactina, Borofutus and Spongiforma, are grouped in one clade with 100% bootstrap support. All of the species in the clade share the characteristic of the basidiospores turning more or less purplish, purplish red to violet grey in aqueous KOH solution (Desjardin et al. 2009, Hosen et al. 2013. Spongiforma squarepantsii Desjardin, Peay & T.D. Bruns, which was described from Malaysia, was not included in these analyses, but the original description of this species also mentioned that its basidiospores turn pale lilac grey in 3% KOH (Desjardin et al. 2011). A chemical reaction with KOH was observed not only with basidiospores, but also on the hymenophore (Desjardin et al. 2009). The reaction to 5% KOH has been observed on fresh basidiomata of Borofutus dhakanus Hosen & Zhu L. Yang which is an epigeous species and the only currently known species of this genus. The colour reaction of pileus and pileus context, which turned pinkish blue to purplish blue, was different from that of the stipe and stipe context, which turned yellowish green to olive green. This variation in colour of the reaction to 5% KOH was not mentioned in the original description of the species (Hosen et al. 2013). Therefore, this chemical character is very useful for the identification of boletes belonging to this group. Other taxa that have been reported to show similar colour reactions to KOH and would, therefore, belong to this group, include Austroboletus longipes (Massee) Wolfe, Austroboletus malaccensis (Pat. & C.F. Baker) Wolfe and Austroboletus tristis (Pat. & C.F. Baker) Wolfe (Corner 1972, Horak 2011.
Some basidiomata of R. rostratispora were old when collected, with dark red hymenophore and had a very strong fruity, alcoholic odour. The odour seems to be present in old basidiomata only (S. Vadthanarat 212 and one basidiomata of S. Vadthanarat 406). One possible explanation to the alcoholic smell is that sterigmata are broken from spore release and any remaining cytoplasm in the basidia could leak into the cavities of the hymenophore and be fermented. Fermentation by yeasts might be possible due to the cracking of the peridium, allowing contact of the hymenophore cavities with ambient air. As mammals and marsupials are known to be the main spore dispersal vectors of truffle-like fungi (e.g. Lamont et al. 1985, Cázares and Trappe 1994, Vernes and Dunn 2009, the strong alcoholic smell could facilitate detection and entice consumption of the basidiomata by mammals and thus help spore dispersal. The three Rhodactina species were found only in dipterocarp forest between 100 to 600 m above sea level in India, northern and northeastern Thailand (Pegler andYoung 1989, Yang et al. 2006). They presumably form ectomycorrhizal associations with trees of the genera Dipterocarpus and Shorea (Dipterocarpaceae). However, in the forest where the new species was found, some scattered Eucalyptus trees were also observed. As Eucalyptus species have been reported to be ectomycorrhizal trees (e.g. Giachini et al. 2000, Ducousso et al. 2012, Garrett Kluthe et al. 2016, the Eucalyptus trees found in the forest could also possibly be host of R. rostratispora. However, Eucalyptus is not indigenous to Thailand; several species have been planted since the early 1900s (Luangviriyasaeng 2003). As Rhodactina species seem to be indigenous to Thailand and Eucalyptus not, they are less likely to be ectomycorrhizal partners. Further study is needed, however, to confirm the range of ectomycorrhizal host tree species of R. rostratispora. Borofutus and Spongiforma, the most closely related genera of Rhodactina, are also ectomycorrhizal associates with trees in Dipterocarpaceae. The only known Borofutus species, B. dhakanus is ectomycorrhizal with Shorea robusta (Hosen et al. 2013). As for Spongiforma species, S. thailandica was reported as associated with Dipterocarpus sp. and Shorea sp. in primary forest while S. squarepantsii was reported as associated with unidentified dipterocarp trees (Desjardin et al. 2009, Desjardin et al. 2011 lence in Bioresources for Agriculture, Industry and Medicine, Faculty of Science, Chiang Mai University. OR is grateful to the Fonds National de la Recherche Scientifique (Belgium) for travel grants. The authors are grateful to Dennis Desjardin and Roy Halling for the loan of specimens. The comments of Roy Halling and Roy Watling helped improving the article and are gratefully acknowledged.