Extremus adstrictus from a dolomite wall in Poland: the first report outside Mallorca

Most species belonging to the Extremaceae family are rock-inhabiting fungi (RIF), which have a deteriorative potential towards colonized substrate. Extremus adstrictus originally isolated from limestone formations in Mallorca is reported from a dolomite wall in Poland. It is the first non-Spanish documented occurrence of this species. Identification of the strain is supported by morphological and molecular analyses. Sequences of uncharacterized fungal cultures and environmental data are analyzed in order to verify probable distribution of Extremus adstrictus.

The family Extremaceae, where most of the described species are RIF, was introduced in 2014 ( Quaedvlieg et al. 2014) as a result of resolving a clade formerly known as Teratosphaeriaceae II. Currently, Extremaceae accommodates the following genera: Extremus, Petrophila, Saxophila, Staninwardia, Pseudoramichloridium, Vermiconidia (Wijayawardene et al. 2018), Castanedospora, Paradevriesia (Hongsanan et al. 2020;Wijayawardene et al. 2020) and Neohortaea (Delgado et al. 2018). All type species of these genera, except Staninwardia, are sequenced. Most of the species of the Extremaceae family were isolated from rock samples from sites located in Mallorca and Antarctica. The genera Staninwardia, Pseudoramichloridium, Castanedospora and Neohortaea originate from plant, soil and lignite material.
The genus Staninwardia was first introduced with Staninwardia breviuscula from Eucalyptus leaves (Sutton 1971). The second Staninwardia species discovered, S. suttoni, was isolated from Eucalyptus robusta in Australia (Summerell et al. 2006) and remains the only sequenced representative of the genus. The genus Pseudoramichloridium was first introduced in 2009 ( Cheewangkoon et al. 2009) when Pseudoramichloridium henryi was isolated from Corymbia henryi. Simultaneously, originally described in 2007 as Ramichloridium brasilianum, an isolate from forest soil, was recombined and introduced as a second representative of the Pseuchoramichloridium genus. The third species of the genus, Pseudoramichloridium xinjangense, was isolated from soil and described in 2017 (Jiang et al. 2017), but was not sequenced. The genus Castanedospora includes a single species, Castanedospora pachyanthicola, originating from dead leaves of Pachyanthus poiretii and Sabal palmetto in Cuba and the USA (Delgado et al. 2018). The genus Neohortaea accommodates a single species, Neohortaea acidophila, isolated from lignite (Hölker et al. 2004;Quaedvlieg et al. 2014).
The genera Petrophila and Saxophila are each represented by a single species -Petrophila incerta and Saxophila tyrrhenica, respectively, isolated from stone and a stone monument located in the Mediterranean Isola et al. 2016;Crous et al. 2019). Vermiconidia (Crous et al. 2019), originally published as a Vermiconia ) includes four species, Vermiconidia antarctica isolated only from Antarctica, V. calcicola found at various sites in Italy, V. flagrans reported from the Mediterranean and V. foris originating from Italian Alps. All described species and strains of Vermiconidia were isolated from stone substrates. Similarly, the two described species of the genus Extremus are a rock-inhabiting fungi with E. antarcticus isolated from McMurdo Valleys in Antarctica and E. adstrictus from limestone formations in Mallorca (Ruibal et al. 2005;Quaedvlieg et al. 2014;Crous et al. 2019).
The genus Paradevriesia was introduced by Crous et al. (2019) and originally transferred to a new family, Paradevriesiaceae. Paradevriesia is comprised of Paradevriesia compacta from rocks, P. americana from air and P. pseudoamericana from Malus domestica fruit (Crous et al. 2019). The family Paradevriesiaceae is now regarded as a synonym of Extremaceae (Hongsanan et al. 2020;Wijayawardene et al. 2020).
In this work, the strain isolation of Extremus adstrictus from a second location, a dolomitic wall in the center of Kraków, Poland, is reported. Morphological and molecular characteristics of this new specimen are provided.

Materials and methods
Located in southern Poland, the city of Kraków is the second largest city in the country. The climate of Kraków is moderately humid continental with cold winters and warm to hot summers (Grøntoft 2017). Small fragments of a dolomite retaining wall situated in the center of Kraków, Poland ( Fig. 1) were sterilely collected in May 2018 and transferred to small tubes. In laboratory conditions, wall fragments were crushed in a mortar under sterile conditions and scattered on malt extract agar (MEA) medium as inoculum as described in Owczarek-Kościelniak et al. (2020). After 12 weeks of growth on MEA medium at 15°C, colonies were used for morphological description and molecular analyses. The isolated strain was deposited in the culture collection of the Westerdijk Fungal Biodiversity Institute (CBS) and as a dried voucher specimen in fungal collection of the W. Szafer Institute of Botany, Polish Academy of Sciences, Kraków (KRAM F).
Culture characteristics were studied on MEA medium. Measurements and photographs of the colonies were taken using 6 month old cultures. Micromorphological observations were made on 3 month old and 6 month old cultures. Slides were mounted with Shear's medium and observed with a Nikon Eclipse 80i light microscope at a magnification of 1000X. The microscopic structures were measured and photographed using NIS-Elements BR 3.0 imaging software.

Species
Strain Source Country NCBI accession number  (2018). LSU PCR reaction was conducted using the following steps: initial denaturation at 95°C for 2 min, 35 cycles in the following order 95°C -35 s, 56°C -1 min and 72°C -1 min, and the final elongation in 72°C for 10 min. Exo-BAP kit (Eurx, Poland) was used for enzymatic purification of amplicons. Bidirectional sequencing was performed at Macrogen Europe B.V. (Amsterdam, The Netherlands). Reads were assembled and trimmed in Geneious Prime® 2020.0.4. Generated sequences were deposited at the NCBI's Gen-Bank nucleotide database (Table 1).
BLASTn query was performed in order to verify fungus identity and to find sequences of the closest relatives. Sequences showing high similarity with newly generated sequences, as well as sequences of other related species were downloaded from GenBank and aligned with the MAFFT algorithm (Katoh et al. 2005) as implemented in Geneious Prime® 2020.0.4. In Geneious Prime® 2020.0.4, a dataset of three concatenated loci, ITS, LSU and RPB2, was prepared. Polychaeton citri CBS 116435 was used as an outgroup. The best partitioning model was determined separately for each loci by PartitionFinder 2.1.1 (Lanfear et al. 2016). Analyses were performed for each loci and for the concatenated datablock at the CIPRES Science Gateway (Millet et al. 2010) using maximum likelihood (ML) analyses using RAxML (Stamatakis 2014) with 1,000 bootstrap replicates and the Bayesian Inference (BI) using MrBayes (Ronquist et al. 2012) in two concurrent runs of four chains for 2,000,000 generations. Final phylogenetic trees were prepared with FigTree 1.4.3. Estimates of the average evolutionary divergence over sequence pairs within groups were calculated in MEGA X (Kumar et al. 2018).

Results and discussion
The MEA cultures of Extremus adstrictus W3 strain from Kraków were mostly consistent with the original type strain description (Fig. 2). Colonies on MEA grew slowly, reaching up to 12.5 mm in diameter after 2 months of growth. Colonies were oval, compact, black with distinct margins and embedded in the medium, reverse was black. Hyphae were septate, sparsely branched, pale brown to brown, and 1.0-4.5 μm wide. Conidia were intercalary, one-septate, brown, several in chains, rarely single, and 6.0-9.5 × 2.5-4.5 μm. Chlamydospores were produced singly and were brown, globoid to ovoid, growing intercalary or apically, one to two, rarely three-celled, and 6.0-11.0 × 5.0-10.0 μm. Colony diameter formed by the isolate from Poland was smaller than colony diameter of type strain of E. adstrictus. Furthermore, chlamydospores were not reported in the original description of the species .
Successfully amplified ITS and LSU loci from new Extremus adstrictus W3 strain were 514 bp and 575 bp for ITS and LSU, respectively. BLASTn searches confirmed a close affinity of the analyzed sequences to the sequences of the type strain of E. adstrictus, showing 98.27% (8 bp difference) and 100% identity for the ITS and LSU loci, respectively. Overall, 38 ITS sequences from the Extremaceae were downloaded from the Gen-Bank. Extremus adstrictus and E. antarcticus similarity was checked in a distance matrix of the aligned ITS datablock. Type strains of these species were 95.09% identical. The number of Extremaceae sequences used in phylogenetic tree reconstruction was reduced to 31 by the elimination of the sequences which similarity to the E. adstrictus was lower than 95.09%. Using Mega X, the average evolutionary divergence over sequence pairs within E. adstrictus group was calculated using the Tajima-Nei model. The rate variation among sites was modeled with a gamma distribution. The divergence within the group was 0.02, whereas in the Petrophila and Saxophila group it was 0.06 and 0.00, respectively.
The reconstructed phylogenetic tree of concatenated ITS, LSU and RPB2 datablocks (Fig. 3) confirms affinity of the isolate from Kraków to the genus Extremus and the identification of the Polish strain W3 as E. adstrictus is well supported. Several sequences of uncharacterized cultures and environmental sequences showed affinity to sequences of Extremus adstrictus obtained from type culture and the strain W3 from Kraków that is presented in supplementary files (Fig. S1, Table S1). Considering morphological characteristics of the type strain and isolate W3 from Kraków, it seems probable that most of the unclassified cultures also represent E. adstrictus.
The sequence from culture TRN80 (from limestone) and the sequence from an uncultured Devriesia clone 10S50C15 (from soil) form an unsupported clade between Extremus adstrictus and E. antarcticus (Fig. S1). The sequence from the alpine soil, fungal sp. MKOTU91, also has an inconclusive position, visible here as basal to the previous sequences. The specific affinities of these three sequences are unresolved.
Among analyzed sequences related to Extremus adstrictus, only one sequence, from uncultured fungus clone 4248_135 closest to isolate from Kraków, does not originate from the stone material, but from irrigation water from a pond in Lithuania (Marčiulynas et al. 2020). All uncharacterized cultures from the TRN collection originate from limestone material in Spain, TRN433 from the Central Mountain System and the remainder from Mallorca (Ruibal et al. 2005 Ruibal et al. 2018;Owczarek-Kościelniak et al. 2020). More black, slow growing fungi were reported from the Mediterranean area (Marvasi et al. 2012;Edigi et al. 2014;Isola et al. 2015;De Leo et al. 2019).
Although fungi do not raise significant interest, their role in stone biodeterioration is important and undeniable. Recent studies have proven that some RIF species possess corrosive properties (Isola et al. 2013;Breitenbach et al. 2018;Gerrits et al. , 2021, thus awareness of their colonization on man-made structures and objects of cultural heritage needs to be raised and further studies should be conducted.

Acknowledgements
This study was supported by the statutory funds of the W. Szafer Institute of Botany, Polish Academy of Sciences, Kraków. Figure S1. Maximum likelihood consensus tree of a concatenated ITS-LSU-RPB2 loci of Extremaceae. Numbers above branches indicate bootstrap support values (ML) and posterior probabilities (BI). The scale bar represents the expected changes per site. T -type strain; eT -ex-type strain. Download file Table S1. Additional fungal ITS sequences used in phylogenetic reconstruction. T -type strain; eT -ex-type strain. Download file