Harorepupu aotearoa (Onygenales) gen. sp. nov.; a threatened fungus from shells of Powelliphanta and Paryphanta snails (Rhytididae)

Abstract

A cleistothecial fungus, known only from the shells of giant land snails of the family Rhytidae, is described as a new genus and species within Onygenales, Harorepupu aotearoa gen. sp. nov. Known only from the sexual morph, this fungus is characterized morphologically by a membranous ascoma with no appendages and ascospores with a sparse network of ridges. Ribosomal DNA sequences place the new species within Onygenales, but comparison with the known genetic diversity within the order linked it to no existing genus or family. It is the first species of Onygenales reported from the shells of terrestrial snails. This fungus has been listed as Critically Endangered in New Zealand and has been previously referred to as ‘Trichocomaceae gen. nov’ in those threat lists.

Introduction

Few fungi have been reported from the shells of terrestrial snails compared to aquatic snails (Řßhová et al. 2014). In a survey of fungi associated with empty shells of Cepaea hortensis, Řßhová et al. (2014) reported 27 species, mostly common soil fungi. They found few potentially keratinolytic species and concluded that the fungi they detected were likely to be accidental colonisers rather than specialist shell decomposing fungi. Snail shells have a layer of calcium carbonate covering a core of conchiolin, a keratin-like compound very resistant to decay (Ormsby et al. 2006, Goffer 2007).

Říhová et al. (2014) mentioned a report on the NZFungi database (http://nzfungi2.landcareresearch.co.nz/) of a species of Trichocomaceae reported from shells of Powelliphanta and Paryphanta species in New Zealand. These snails are members of the family Rhytididae (Mollusca; Gastropoda; Pulmonata), the thick shells of which are composed almost entirely of conchiolin with only thin outer layers of calcium carbonate (Ormsby et al. 2006). Hitchmough (2002) listed this fungus as ‘Undescribed genus, Trichocomaceae’ and accorded it a Nationally Critical threat status. The same fungus has been mentioned in Department of Conservation reports (e.g. Anon. 2007, Miller & Holland 2008).

The tentative NZFungi identification of the fungus on Powelliphanta and Paryphanta as Trichocomaceae was based on the macroscopic appearance of the ascomata and ascospore morphology. An asexual morph has not been observed. Trichocomaceae is a family in Eurotiales, some species of which have sexual morphs similar to those of Onygenales, the two orders being most easily distinguished morphologically by their asexual morphs (Currah 1994). Currah (1994) notes that amongst these fungi, the keratin degrading species are restricted to the families Onygenaceae and Arthrodermataceae within Onygenales. Of these two families, the fungus on Powelliphanta and Paryphanta is morphologically similar to Onygenaceae sensu Currah (1985). Two fungi reported from cultures derived from Cepaea shells by Řßhová et al. (2014) were identified using DNA sequences as Onygenales, but the sequences for these are not available.

In this paper we describe the fungus associated with Powelliphanta and Paryphanta shells as a new genus within Onygenales incertae sedis, its phylogenetic position being based on SSU, ITS and LSU sequences. We compare it with the known genetic diversity within the order.

Materials and Methods

Morphology

Fungarium specimens were rehydrated in 3% KOH and the hymenial elements examined microscopically in either 3% KOH or 3% KOH mixed with Lugol’s iodine solution. Vertical sections were cut at a thickness of about 10 µm using a freezing microtome and mounted in lactic acid. Material for scanning electron microscopy (SEM) was obtained by placing a mass of dried ascospores onto carbon tape on a stub, then sputter coating with gold. Photomicrographs taken on a Jeol Neoscope JCM-5000 (Landcare Research). Specimens have been deposited in PDD.

Molecular analyses

For DNA extraction, three separate extractions were done from three different ascomata from PDD 105262. DNA was extracted and amplified using a REDExtract-N-Amp Plant PCR Kit (Sigma-Aldrich, USA), following the manufacturer’s protocol except that the ascomata were ground in 30 µL extraction solution with a plastic pestle. Amplification primers for ITS were ITS1F and ITS4 (White et al. 1990; Gardes & Bruns 1993), for LSU were LR0R and LR5 (Bunyard et al. 1994; Vilgalys & Hester 1990), and for SSU were NS1 and NS4 (White et al. 1990).

Additional sequence data of SSU, LSU and ITS were downloaded from GenBank (Table 1). Sequences of each gene were aligned with MAFFT 7.122b (Katoh & Standley 2013) and trimmed with BioEdit (Hall 1999). Alignments were deposited in TreeBASE (www.treebase.org/treebase/), study accession number 17085. Molecular phylogenies were constructed using Bayesian inference (BI) and maximum likelihood (ML). To select the most appropriate model of sequence evolution, jmodeltest 2.1.1 (Darriba et al. 2012; Guindon & Gascuel 2003) was applied on each alignment (ITS, SSU, LSU). The GTR + I + G model was selected for ITS, SSU, and LSU according to the Akaike information criterion (AIC). The SSU and LSU matrices were concatenated with SeaView (Gouy et al. 2010). BI analyses were performed with MrBayes 3.2 (Ronquist & Huelsenbeck 2003). Three independent Markov chain Monte Carlo (MCMC) runs were performed simultaneously. Each MCMC ran for 3 × 10⁶ generations for the SSU+LSU analysis and the ITS analysis, sampling every 500 generations until convergence (standard deviation of split frequency < 0.01). The first 25% of trees were discarded as burn-in while the remaining trees combined with a 50% majority rule consensus. ML analyses were performed with phyML 3.0 (Guidon et al. 2010) running inside SeaView (Gouy et al. 2010) with the following options: GTR model; aLRT branch support; empirical nucleotide equilibrium frequencies; optimized invariable site; optimized across site rate variation with 8 rate categories; NNI tree searching operations; BioNJ starting tree with optimized tree topology.

Table 1 Species, culture, orvoucher numbers, and GenBank accession numbers of isolates used in the phylogenetic analyses.

Taxonomy

Harorepupu P.R. Johnst., H.D.T. Nguyen, D.C. Park, & Hirooka, gen. nov.

MycoBank MB811561

Etymology: From the Māori words harore = fungus, and pūpū = snail (fem.).

Diagnosis: Ascomata globose, sessile, membranous, solitary or in small, confluent groups; asci subclavate, wall undifferentiated; ascospores hyaline, oblong-elliptic, ornamentated with anastomosing ridges.

Type: Harorepupu aotearoa P.R. Johnst. et al. 2015.

Harorepupu aotearoa P.R. Johnst., H.D.T. Nguyen, D.C. Park, & Hirooka, sp. nov. MycoBank MB811562

(Fig. 1)

Fig. 1

Harorepupu aotearoa. A. Ascomata on shell, arrows indicate groups of ascomata on host shell (PDD 105262). B. Detail from A. C. Ascoma with wall breaking to expose powdery mass of yellow spores inside (PDD 74629). D. Asci in 3% KOH plus Lugols iodine (PDD 74629). E. Ascomatal wall in vertical section (PDD 89035). F. Surface of ascoma (PDD 105262). G–I. Ascospores under light microscope, at three planes of focus (PDD 105262). J. Ascospores under SEM (PDD 105262). Bars: A, B = 10 mm; C = 0.5 mm; D, G–J = 10 pm; E–F = 20 µm.

Etymology: The species epithet is from the Māori word for the country of origin.

Diagnosis: Ascomata 0.8–1.2 mm diam, white to pale yellow; asci 13–16 × 7.5–8.5 µm, 8-spored; ascospores 4.2–5.4 × 2–3.1 µm (average 4.8 × 2.6 µm), oblong-elliptic, ends rounded, sparse network of narrow, ridge-like ornamentations.

Type: New Zealand: Nelson: Golden Bay, Wainui Falls Tr., on empty shell of Powelliphanta sp., 16 May 2014, P.R. Johnston FUNNZ 2014/0999 (PDD 105262 — holotype).

Description: Ascomata 0.8–1.2 mm diam, globose, sessile, membranous, surface slightly woolly but with no distinctive appendages, white to pale yellow; Opening by irregular cracks, revealing the dry, powdery, bright yellow spore masses inside; wall 80–100 µm thick, comprising tightly tangled hyphae 4–6 µm diam, walls thin, mostly hyaline, outer 3–4 rows of cells sometimes with pale yellow walls; outermost layers of hyphae sometimes with ends free; peridial appendages lacking. Asci 13–16 × 7.5–8.5 µm, clavate with a narrow, foot-like base and rounded apex, wall thin, undifferentiated, 8-spored, contents orange-brown in Lugol’s iodine. Ascospores 4.0–5.5 × 2–3 µm (average 4.8 × 2.6 µm), oblong-elliptic, ends rounded, ornamented with sparse network of narrow, anastomosing ridges, hyaline to pale yellow, 0-septate. Asexual morph not seen.

Additional specimens examined: New Zealand: Nelson: vic. Karamea, Kohaihai, Nikau Walk, on empty shell of Powelliphanta sp., 11 May 1994, P.R. Johnston (PDD 74629); vic. Karamea, Oparara Basin, Moria Gate Track, on empty shell of Powelliphanta sp., 10 May 2006. T. Atkinson FUNNZ 2006/1066 (PDD 92048); vic. Westport, Charming Creek Walkway, on empty shell of Powelliphanta sp., 10 May 2006 A. Wilson FUNNZ 2006/0160 (PDD 89035). Northland: Waipoua Forest, on empty shell of Paryphanta sp., 2001, E. Horak, (PDD 74625).

Results

DNA sequences from all three ascomata from PDD 105262 were identical. They have been accessioned as GenBank KP683349, KP683350, and KP683351.

Phylogenetic analyses with the combined SSU + LSU sequences was performed to determine the higher taxonomic placement of Harorepupu aotearoa. After removing ambiguously aligned regions, SSU and LSU alignments were both 1300 base pairs long and contained a total of 257 (20%) and 421 (32%) parsimony informative characters respectively. Both the BI analysis (Fig. 2) and ML analysis (not shown) placed H. aotearoa in an isolated position in Onygenales. To determine whether we could place it in a well-supported family in Onygenales, we then performed phylogenetic analyses of the ITS region, with an alignment of 876 base pairs in length that contained 456 (52%) parsimony informative characters. H. aotearoa is sister to Nannizziopsiaceae but lacking strong statistical support, where the aLRT branch support was only 0.74 in the ML analysis (data not shown) and the posterior probability is only 0.58 in the BI analysis (Fig. 3). All phylogenetic analyses show that H. aotearoa represents an isolated lineage within Onygenales.

Fig. 2

50% majority rule consensus tree from Bayesian inference analysis of SSU and LSU sequences. Posterior probabilities greater than 0.7 shown above the edges. Taxa labelled EX are represented by sequences from ex-type cultures; bold type indicates the type species of genera.

Fig. 3

50% majority rule consensus tree from Bayesian inference analysis of ITS sequences. Posterior probabilities greater than 0.7 shown above the edges. Taxa labelled EX are represented by sequences from ex-type cultures; bold type indicates the type species of genera.

Discussion

Although Harorepupu aotearoa has never been grown on artificial media, we obtained DNA sequence data from dried specimens. Our comprehensive LSU and SSU phylogenetic tree showthat this fungus is a member of Onygenales and that is distantly related from any recognized onygenalean fungi. In our ITS tree, H. aotearoa was sister to the Nannizziopsiaceae clade but with low support in the BI analysis. The family Nannizziopsiaceae was described by Stchigel et al. (2013) on the basis of D1/D2 phylogenetic data, host range, morphology, and colony odour. Based on sexual morphology, historically taxonomically important for the group, species in Nannizziopsiaceae differ from our fungus in having ascomata with peridial appendages and ascospores that appear smooth under the light microscope (Currah 1985). The future discovery of additional species of Harorepupu, and of any asexual morph, could help clarify its position within the order. For now, however, we prefer to treat it as incertae sedis within the order rather than introduce a new family name for this single genus.

The biology of Harorepupu aotearoa is not understood, but as all collections are on empty shells of members of the family Rhytididae, it may be restricted to this substrate. If this is the case, threats to the snail population will present a threat to the fungus population. At present, with increased predation and disturbance resulting in larger numbers of dead Rhytididae shells on the forest floor, this fungus may temporarily be more common than usual.

Members of the family Rhytididae are distributed across many regions linked geologically to Gondwana. Although Harorepupu is at present known only from New Zealand, additional material, and perhaps more species, may be expected on the shells of these snails in other regions.