Aestipascuomyces dupliciliberatus gen. nov, sp. nov., the first cultured representative of the uncultured SK4 clade from Aoudad Sheep and Alpaca

We report on the isolation of the previously uncultured Neocallimastigomycota SK4 lineage by two independent research groups from a wild aoudad sheep rumen sample (Texas, USA) and an alpaca fecal sample (Baden-Württemberg, Germany). Isolates from both locations showed near identical morphological and microscopic features, forming medium-sized white filamentous colonies with a white center of sporangia on agar roll tubes and a heavy biofilm in liquid media. Microscopic analysis revealed monocentric thalli, and spherical polyflagellated zoospores with 7– 20 flagella. Zoospore release occurred through an apical pore as well as by sporangial wall rupturing, a duality that is unique amongst described AGF strains. Isolates were capable of growing on a wide range of mono-, oligo-, and polysaccharides substrates. Phylogenetic assessment based on the D1-D2 large rRNA subunit (D1-D2 LSU) and internal transcribed spacer-1 (ITS-1) regions demonstrated high sequence identity (minimum identity of 99.07% and 96.96%, respectively) between all isolates; but low sequence identity (92.4% and 86.7%, respectively) to their closest cultured relatives. D1-D2 LSU phylogenetic trees grouped the isolates as a new monophyletic clade within the Orpinomyces-Neocallimastix-Pecoramyces-Feramyces-Ghazallamyces supragenus group. D1D2 LSU and ITS-1 sequences from the obtained isolates were either identical, or displayed extremely high sequence similarity to sequences recovered from the same Aoudad sheep sample on which isolation was conducted, as well as several sequences recovered from domestic sheep and few other herbivores. Interestingly, members of the SK4 clade seem to be encountered in animals grazing on summer pasture. We hence propose accommodating these novel isolates in a new genus, Aestipascuomyces (derived from the Latin word for “summer pasture”), and a new species, A. dupliciliberatus. The type strain is Aestipascuomyces dupliciliberatus strain R4.


Introduction
The herbivorous gut harbors a wide range of bacterial, archaeal, protozoan, and fungal communities that collectively mediate the transformation of plant biomass into fermentable sugars and short-chain fatty acids (SCFA) [1]. Within such complex assemblages, members of the anaerobic gut fungi (AGF, phylum Neocallimstigomycota) remain the most enigmatic [2,3]. During the last few decades, an increased understanding of the AGF diversity, ecology, and metabolic capabilities has been accumulating, and it is now broadly agreed that AGF play an integral role in the anaerobic degradation of recalcitrant lignocellulosic material [4,5], through hyphal penetration of plant material and production of a wide array of polysaccharide-degrading enzymes [1,2,6].
Culture-independent diversity surveys have clearly demonstrated that AGF diversity is much broader than previously inferred from culture-based approaches. Such studies have identified several novel yet-uncultured lineages, mostly through the use of the ITS-1 and D1/D2 LSU regions as phylogenetic markers [20][21][22][23][24]. Despite multiple recent efforts to isolate and characterize novel AGF lineages [11][12][13][14][15], many candidate genera remain uncultured. A recent study combining amplicon-based diversity survey with isolation efforts suggested that the success of isolation of an AGF taxon is positively correlated to its relative abundance in a sample and negatively correlated to the sample evenness [24]. Further multiple culture-based [11,15], and culture-independent [20,24] studies have provided evidence that poorly sampled animal hosts harbor a wide range of hitherto uncharacterized AGF taxa. Based on these observations, we adopted two strategies to isolate novel AGF taxa: A targeted sequence-guided isolation strategy, where samples harboring relatively high proportions of yet-uncultured genera are prioritized for AGF isolation efforts, and a sampling strategy targeting animals from which no prior isolation efforts have been reported.
Intriguingly these efforts, driven by two different hypotheses, and sampling different animals (a wild aoudad sheep and a zoo-housed alpaca) from two different geographical locations (Texas, USA, and Baden-Württemberg, Germany) have yielded almost identical strains of a hitherto uncultured AGF lineage (SK4, originally identified in samples from New Zealand).
This study demonstrates the global distribution of AGF lineages across multiple continents, suggests that some yet-uncultured AGF genera are not refractive to isolation given the right sampling and isolation conditions, and highlights the value of implementing a sequenceguided culturing approach as well as directing isolation efforts to poorly sampled animals.

Materials and Methods
Samples. Fresh fecal and rumen contents were collected in several sterile 50-ml falcon tubes from a wild aoudad sheep (Ammoragus lervia) during a hunting trip in Sutton County, Texas, USA in April 2018. Fecal samples were collected from an Alpaca (Vicugna pacos) at the Karlsruhe Zoo, Germany in August 2019. Tubes were filled completely to ensure the absence of oxygen. Aoudad sheep samples were stored on ice and transferred to the laboratory within 24 hours, where they were either directly utilized for DNA extraction or stored at -20 0 C. Alpaca fecal samples were stored at room temperature until the next day where they were used for isolation.
Isolation. The aoudad sheep sample exhibited a relatively high abundance (76.6%) of the yet-uncultured SK4 lineage in a prior study [24], and hence was chosen for targeted enrichment and isolation. Isolation efforts were conducted on fecal, as well as rumen samples.
Rumen samples used in the isolation process were stored unopened at -20ºC. Fecal samples were opened once in an anaerobic chamber (Coy laboratories, Grass Lake, Michigan, USA) to obtain 0.5 gram for use in culture-independent diversity survey efforts, then stored at -20ºC. Isolation efforts were conducted 22 months post sample collection and DNA extraction. Samples were enriched in autoclaved rumen fluid-cellobiose (RFC) medium [25] for 24h at 39ºC. Enriched tubes were serially diluted into anaerobic rumen fluid medium (RF) supplemented with either 0.1% w/v cellulose or a (1:1) mixture of cellobiose and switchgrass (0.1% w/v), and an antibiotics mixture of 50 μg/mL kanamycin, 50 μg/mL penicillin, 20 μg/mL, streptomycin, and 50 μg/mL chloramphenicol. Following enrichment, serial dilutions up to 10 -5 were performed, and the dilution tubes were incubated for 3 days at 39ºC. Dilutions showing visible signs of growth (change in the color of cellulose, clumping and floating of the switch grass, and production of gas bubbles) were used to prepare roll tubes [26] using RFC medium with 2% agar. Roll tubes were incubated for 2-3 days at 39ºC, after which single colonies were transferred into RFC medium. Roll tube preparation and colony picking were repeated at least 3 times to ensure the purity of the obtained isolates. Obtained isolated are being maintained via bi-weekly sub-culturing into RFC media. Cultures are stored on agar medium for long-term storage as previously described in [25].
Inoculated serum bottles were then incubated for 7 days at 39 °C in the dark. Fungal growth was monitored by light microscopy and serum bottles with signs of anaerobic fungal growth were then used to inoculate roll tubes followed by incubation for 4 days at 39 °C in the dark.
Single colonies were transferred into fresh medium. Roll tube preparation and colony picking were repeated at least 3 times to ensure the purity of the obtained isolates.

Morphological characterization.
For aoudad sheep isolates, both light and scanning electron microscopies were utilized to observe various microscopic features at different growth stages. For light microscopy, fungal biomass was collected from an actively growing 2-3d old culture in RFC medium. Fungal biomass was stained with lactophenol cotton blue for examination of various thallus features including: hyphae, sporangia, zoospores, and other specific microscopic structures as previously described in [13][14][15]. For nuclear localization, samples were stained with DNA-binding dye 4, 6 diamidino-2-phenylindole (DAPI, final concentration of 10 μg/ml), followed by incubation in the dark for 10 min at room temperature. All light microscopy examinations were conducted using an Olympus® BX51 microscope (Olympus, Center Valley, Pennsylvania) equipped with a Brightline DAPI high contrast filter set for DAPI fluorescence and a DP71 digital camera (Olympus, Center Valley, Pennsylvania). Sample preparation and fixation for scanning electron microscopy was conducted as previously described in [13]. The prepared samples were then examined on a FEI Quanta 600 scanning electron microscope (FEI Technologies Inc., Hillsboro, Oregon, United States).
For Alpaca isolates, light microscopy was performed using a Nikon® Eclipse E200 with a DFK 23U274 camera (Imaging Source®), while fluorescence microscopy (to visualize DAPI staining) was performed using a Zeiss® Axio Imager Z1 at an excitation wavelength of 353 nm. Differential interference contrast microscopy (DIC) was used for generating image overlay.
Substrate utilization. Growth of the type strain (R4) obtained from aoudad sheep was assessed by replacing the cellobiose in RFC medium with glucose, xylose, mannose, fructose, glucuronic acid, arabinose, ribose, galactose, sucrose, maltose, trehalose, lactose, cellulose, xylan, starch, inulin, raffinose, polygalcturonate, chitin, alginate, pectin, peptone, or tryptone at a final concentration of 0.05% w/v [13,14]. To assess substrate utilization in the Alpaca isolate, the strain was grown in defined rumen-free media adapted from [12] with omission of clarified rumen fluid and addition of trace metal (prepared according to [27]) and vitamin solution (prepared according to [28]) and replacement of the cellobiose with 0.05% of hemicellulose, xylan, starch, crystalline cellulose, inulin, chitin, pectin, cellobiose, maltose, trehalose, lactose, sucrose, glucose, xylose, mannose, fructose, arabinose, ribose, galactose, or glucuronic acid, or 0.5% of wheat straw. The ability of a strain to utilize a specific substrate was considered positive if it exhibited viable growth on the tested substrate after four successive transfer events [13,14,18]. All results were compared to substrate-free medium.  [14,15] using the following PCR protocols: For the aoudad sheep samples: Initial denaturation at 94ºC for 5 min followed by 39 cycles of denaturation at 94º C for 1 min, annealing at 55º C for 1 min, and elongation at 72º C for 2 min, and a final elongation step at 72º C for 10 min. For the alpaca samples: Initial denaturation at 98ºC for 30 sec followed by 30 cycles of denaturation at 98º C for 10 sec, annealing at 62º C for 30 sec, and elongation at 72º C for 90 sec, and a final elongation step at 72º C for 2 min. PCR amplicons were cloned into TOPO-TA cloning vector (USA) (Life Technologies®, Carlsbad, CA), or PCR Cloning Kit (NEB) following the manufacturers' instructions, and were Sanger-sequenced at the Oklahoma State University DNA sequencing core facility (22 clones from 5 aoudad sheep strains), or Eurofins Genomics (14 clones from 1 alpaca strain). For every clone sequence obtained, the ITS-1, and the D1/D2-LSU regions were extracted in Mega7 [29] by trimming using the sequence of the ITS1 reverse primer MNGM2, and the sequence of the LSU forward primer NL1, respectively. The trimmed sequences were aligned to anaerobic fungal reference ITS-1 and D1/D2-LSU sequences using MAFFT v7.471 [30] and the alignments were manually curated in BioEdit [31]. The refined alignments were used to construct maximum likelihood trees to assess the phylogenetic position of the obtained sequences using IQ-TREE v2.0.3 [32]. The best model was selected using ModelFinder [33] and 1000 ultrafast bootstraps [34] were applied. Gonopodya prolifera was used as the outgroup (NR_132861 for ITS-1, JN874506 for 28S).
To assess the ecological distribution of this novel lineage, we queried the trimmed ITS-1 sequences against a manually curated Neocallimastigomycota ITS-1 database encompassing all known cultured genera, as well as yet-uncultured taxa previously identified in culture-independent studies [20,21,23,24,35,36] using blastn. Hits with significant sequence similarity (>87%) were evaluated by insertion into ITS-1 phylogenetic trees. We also queried the D1/D2 LSU dataset generated in our prior effort [24], and hits with >93% sequence similarity were further evaluated by insertion into D1/D2-LSU phylogenetic trees.
Data and culture accession. Clone sequences are deposited in GenBank under accession numbers MW019479-MW019500 for the Aoudad sheep strains R1-R5, and MW049132-MW049145 for the Alpaca strain A252.

Isolation.
Five rumen isolates (R1-R5) were obtained from a single wild aoudad sheep in Texas, USA. Concurrently one isolate, A252, was obtained from fecal samples of Alpaca in Baden-Württemberg, Germany. These six isolates were not phylogenetically affiliated with any of the previously cultured genera. Preliminary morphological and microscopic characterization as well as phylogenetic analysis showed identical attributes for strains R1-R5 and only minimal differences between R strains and strain A252. One isolate (strain R4) was chosen as the type strain for detailed characterization. Below, we present detailed characterization of the putative novel genus morphology and phylogenetic affiliation, highlighting differences between R4 and A252 when appropriate.
3.2. Colony morphology and liquid growth pattern. On solid media, strain R4 formed circular, white filamentous colonies with a white center of sporangia (Figure 1a). Colony size ranged from 2-5 mm. In liquid media, strain R4 produced a heavy fungal biofilm-like growth that loosely attaches to the tube's glass surface (Figure 1b).
• Zoospore release. Zoospore release in strain R4 was achieved through two mechanisms, either from an apical pore (Figure 2o) as previously observed in Feramyces [14],) or through rapturing of the sporangial wall (Figure 2p) as commonly observed in Neocallimastix [16].
To our knowledge, the simultaneous utilization of both mechanisms by a single strain has not been previously reported in other AGF taxa. Sporangial walls either stayed intact (Figure 2o) or completely disintegrated after zoospore discharge (Figure 2q).

Substrate utilization.
Strain R4 utilized a wide range of substrates as the sole carbon and energy source. These included monosaccharides, e.g., glucose, fructose, mannose, xylose, and glucuronic acid, but not arabinose, galactose, or ribose. Strain R4 was able to metabolize and vigorously grow on all disaccharides tested including cellobiose, lactose, maltose, sucrose, and trehalose. Among the polymers tested, strain R4 was able to grow on cellulose, xylan, starch, inulin, and raffinose, but not alginate, chitin, pectin, polygalacturonate, peptone, or tryptone.
On the other hand, strain A252 grew on polysaccharides including wheat straw, hemicellulose, xylan, starch, inulin, but did not grow on chitin, pectin, or crystalline cellulose.
The disaccharides cellobiose, maltose, lactose, and sucrose supported the growth of strain A252, but trehalose was not utilized. Strain A252 was capable of utilizing the monosaccharides glucose, xylose, and fructose, but not mannose, arabinose, ribose, galactose, or glucuronic acid. On the other hand, the obtained isolates showed a slightly higher ITS-1 length heterogeneity (196-200 bp; average 197.5 bp), within-strain divergence between copies (0-4.38%), as well as inter-sequence divergence between strains (0-5.84%). ITS-1 phylogeny (Figure 4b) placed the obtained isolates close to the genus Feramyces. Blastn against our custom ITS-1 database identified 1327 sequences with 87% sequence similarity. All hits were affiliated with the SK4 clade (originally identified in domesticated sheep and red deer samples in NZ [21,22]).

Phylogenetic analysis and ecological distribution.
The majority of hits were from the same wild Aoudad sheep samples from which the US isolates were obtained (n=1311), domesticated sheep (n=5) previously reported in NZ [21], as well as oryx, blackbuck deer, horse, miniature donkey, mouflon, and elk (n=11). Analysis of all available SK4-affiliated sequences obtained from prior studies [21,22,24] and the current study indicates a clade ITS-1 sequence divergence range of 0-13.2%, with two welldefined subclades. Interestingly, divergent ITS-1 sequences originating from one isolate routinely clustered within both clades (Figure 4b), precluding equating subclades to two distinct species and highlighting the difficulty associated with species-level OTU assignment using ITS-1 data in the Neocallimastigomycota.
Notably, it seems that members of the SK4 clade exhibit higher abundance when animals graze on summer pasture. For example, in New Zealand's domesticated sheep, SK4 was only identified as part of the AGF community when the animals were grazing on summer, but not winter, pasture [21,22]. In addition, while the exact feed of other animal hosts harboring the SK4 lineage (e.g. aoudad sheep, oryx, blackbuck deer, horse, miniature donkey, mouflon, and American elk) is not available [24], all the above samples were collected during summer months (between April and October), suggesting a potential relationship between the enrichment of SK4 in the AGF community and the season feed type.    . Sequences were aligned in MAFFT [30] and manually curated in BioEdit [31] . Curated alignments (LSU: 677 characters 209 sequences; ITS: 295 characters 126 sequences) were used to construct ML-trees using IQTREE with the predicted models TN+F+R2 (28S rDNA) or HKY+F+G4 (ITS) and -bb 1000. Bootstrap values are shown for nodes with more than 70 % bootstrap support. Background color indicates the origin of the isolate (blue: Texas, USA; green: Baden-Württemberg, Germany).

Figure Legends
Neocallimastix-Pecoramyces-Feramyces-Ghazallamyces supragenus clade [24,37]. All members in this clade are characterized by the production of polyflagellated zoospore, with the notable and peculiar exception of the genus Pecoramyces, which produces monoflagellated zoospores. This suggests an acquisition pattern of zoospore polyflagellation at ~ 46.3 Mya (the most current estimate of this clade emergence per [37]), followed by a recent loss and reverting to zoospore monoflagellation for the relatively recently-evolved genus Pecoramyces (current estimates of emergence at 19.1 Mya, [37]). Similarly, all members of this supragenus clade form monocentric thalli with the exception of Orpinomyces genus that is known to develop polycentric thalli, also suggesting that the development of polycentric thalli is a recent independent event that happened multiple times in the Neocallimastigomycota tree (for example with the emergence of Orpinomyces, Anaeromyces, and Cyllamyces). The closest cultured representatives of the SK4 clade are the genera Feramyces and Neocallimastix. While the three genera share similar morphological and growth patterns (e.g. polyflagellated zoospores, and monocentric thalli development), they exhibit several distinct macroscopic and microscopic features. For example, members of the SK4 genus produce zoospores with 7-20 flagella, as opposed to 7-16 for Feramyces [14] and 7-30 for Neocallimastix [16]. Additionally, SK4 members produce terminal sporangia, while the Feramyces genus members produce terminal, pseudo-intercalary, and sessile sporangia [14]. Also, and perhaps most notably, members of the SK4 genus show two zoospore release mechanisms; either through an apical pore or via rupturing of the sporangial wall. On the other hand, the majority of Neocallimastix genus members are known to release zoospores through complete rupturing and lysis of the sporangial wall ( Figure 25 in [16], with only a few exceptions (e.g. [38,39]), while Feramyces members release zoospores through apical pores (Figure 2x in [14]). To our knowledge, the dual zoospore release mechanism has not been encountered before in any of the cultured AGF genera members and hence is highly characteristic of the SK4 genus.
Within the microbial world, a large fraction remains uncultured. This is more commonly encountered within the bacterial and archaeal domains, although a similar pattern has been suggested with Fungi [40][41][42][43]. Within the anaerobic fungal phylum Neocallimastigomycota, multiple putative novel genera were identified in culture-independent studies [20,21,35].
Failure to obtain these taxa in pure culture could be attributed to several reasons. First, some AGF taxa are extremely fastidious and might require special nutritional and culturing requirements, and hence would evade isolation using routinely utilized isolation and enrichment protocols [1,44]. Second, some AGF taxa might exhibit a very limited ecological distribution pattern and could be confined to few phylogenetically-related animal hosts.
Indeed, many novel genera recently isolated appear to be of limited distribution, being observed only in very few samples from which they have been successfully isolated (e.g. Aklioshbomyces from white-tailed deer, Ghazallomyces from Axis deer, Khyollomyces (AL1) in the Equidae [15]). We argue that, in addition to mere presence, the relative abundance of the target lineage in the sample could be an important determinant for isolation success in the AGF. Our recent efforts [24] suggest that while some AGF genera are generalists, present in low abundance in a large number of samples and are often readily recovered from these samples, e.g. Orpinomyces, and Anaeromyces; others show a clear correlation between the success of their isolation and their relative abundance within a sample, especially in samples where one or a few lineages make up the majority (>90%) of the AGF community.
While information is currently lacking on the AGF community in the Alpaca sample that was used for isolation in Germany, we believe that the success of obtaining a cultured SK4 representative was largely dependent on its presence in high relative abundance in the samples used for isolation. Therefore, this study clearly demonstrates the value of the sequence-guided isolation strategy that was employed here, whereby samples are initially prescreened using culture-independent approaches followed by targeting promising samples exhibiting a high proportion of novel/wanted genera for isolation efforts using a wide range of substrates, sample types, and growth conditions. Evidently, this approach will unfortunately involve storing the samples at -20 0 C for a certain amount of time to allow for sequencing and data analysis to be conducted. Nevertheless, while some AGF taxa might not survive prolonged freezing, we have been successful in recovering isolates from samples stored frozen, especially when tubes were unopened, or at least where repeated freezing and thawing cycles were avoided, and where tubes were filled to the top with little to no room for air [14].
Based on morphological, physiological, microscopic, and phylogenetic characteristics, we propose accommodating these new isolates into a new genus, for which the name Aestipascuomyces (from aesta, latin for summer, and pascui, latin for pasture, to indicate the apparent enrichment of the clade during animal feeding on summer pasture) is proposed. The type species is Aestipascuomyces dupliciliberatus (to indicate the two zoospore release mechanisms exhibited by members of the clade), and the type strain is Aestipascuomyces dupliciliberatus strain R4. Obligate anaerobic fungus that produces globose polyflagellated zoospores (7-20 flagella).