New species of aquatic chytrids from Oman

Oman is a desert country in the south of the Middle East. Springs and other water sources that harbor aquatic organisms can be separated by hundreds of kilometers. In Oct 2019, we isolated four freshwater aquatic fungi (Chytridiomycota) from benthic detritus baited with pine pollen on a general nutrient medium near Salalah, Oman. Database queries of nuc 28S rRNA (28S) and internal transcribed spacer region ITS1-5.8S-ITS2 (ITS) revealed that one of these strains was Dinochytrium kinnereticum , a recently described algal pathogen from the Sea of Galilee. The other three strains had low molecular identity to available ITS sequences. These unknown strains varied in size and released endogenously swarming zoospores through papillae from mature zoosporangia. Zoospore ultrastructure was consistent with described species in the Rhizophydiales, and molecular phylogenetic results grouped these three strains into a clade in the genus Rhizophydium . We circumscribe these three strains as a sp. nov., thereby expanding the diversity within Rhizophydium described as the new species R. jobii . In doing so, we provide the first report of Chytridiomycota from Oman.


INTRODUCTION
Oman is situated in the southeast corner of the Arabian subcontinent.The Arabian subcontinent was last connected to Africa ~30 Ma, until the eventual formation of the Red Sea (Bohannon et al. 1989).To the north of Oman and Yemen is the Rub' Al Khali Desert (the Empty Quarter), which spans >600 000 square km of arid sand and dunes.The Arabian Sea sits to the south and east of Oman.As a result of its geographic position, the relative isolation of Oman renders it of potential interest for studies of allopatric speciation.Sporadic rainfall and perennial springs fill permanent and ephemeral water sources (wadis), further elevating interests in diversity driven by allopatric speciation, especially among freshwater aquatic organisms.
Oman has a relatively short history of fungal biology and ecology.To date, the majority of mycological research has been conducted within the realm of plant pathology (Al-Nadabi et al. 2018); consequently, the fungal species richness within Oman has historically been inferred from pathogen reports (Al-Bahry et al. 2005).In 2016, the first checklist of fungi in Oman was published (Maharachikumbura et al. 2016), coalescing all previous observations, while expanding that compilation with new observations, including the discovery of novel taxa (Halo et al. 2019).However, there were no flagellated fungi in this checklist.
Chytridiomycota, primarily in the genus Rhizophlyctis, have been detected in high-throughput sequencing surveys throughout the Arabian Peninsula, including within Oman (Abed et al. 2013(Abed et al. , 2019) ) and Saudi Arabia (Moussa et al. 2017).However, only three species of Chytridiomycota were reported between 1940 and 2016 from anywhere in the Middle East: Synchytrium aureum in Palestine, Rhizophydium racemosum in Egypt (Mouchacca 2005(Mouchacca , 2016)), and Dinochytrium kinnereticum from the Sea of Galilee (Leshem et al. 2016).Despite a long history of Chytridiomycota studies in neighboring India (Karling 1964;Dayal 1997) and regular screenings for Synchytrium (Potato Export Guide 2019), to our knowledge, there have been no Chytridiomycota reported within Oman or on the Arabian Peninsula (Mouchacca 2005(Mouchacca , 2016)).As a result, the species richness of Chytridiomycota remains uncharacterized in this region.
To help fill this gap, we conducted a culturing survey within Oman.As a result of these efforts, we report four strains of Chytridiomycota, representing two different species.The first is the second report of Dinochytrium kinnereticum in the Chytridiales; the second, we describe as a novel species, Rhizophydium jobii in the Rhizophydiales.

MATERIALS AND METHODS
Samples and cultures-Benthic detritus was collected from several freshwater springs and ephemerally standing water pools (<0.25 m depth) near Salalah, Oman (17.110°N, 53.995°E) on 10 Oct 2019.Samples were collected in sealable plastic bags and maintained at air temperature until processing (4 h from time of collection).A slurry of detritus and source water was incubated with heat-killed pine pollen at room temperature (~24 C) for 3 d.After 3 d, pollen was drop-streaked onto PmTG agar medium (peptonized milk, tryptone, glucose) containing streptomycin sulfate and penicillin G antibiotics (Barr 1986).After 24 h, individual Chytridiomycota zoosporangia were transferred to a separate PmTG plate to establish axenic cultures, where they were further maintained at room temperature until light microscopy examination and DNA extraction.Specimens were deposited in the Collection of Zoosporic Eufungi at the University of Michigan (CZEUM; czeum.herb.lsa.umich.edu/index.php).
Light microscopy-We incubated new subcultures 3 d prior to analysis and captured images of sporangia and zoospores with a Nikon H600L microscope fitted with a DS-Ri2 camera (Melville, New York).Cells were imaged directly on the agar plate to preserve their structural integrity while not introducing morphological artifacts.Ranges were reported by identifying the largest and smallest nonephemeral (minimum presence of three) cell stages.
DNA extraction, PCR, and sequencing-We extracted DNA from axenic cultures with a DNeasy PowerMax Soil Kit (Qiagen, Hilden, Germany), following the manufacturer's protocols.A portion of nuc 28S rDNA (28S) and the internal transcribed spacer region ITS1-5.8S-ITS(ITS) were amplified with primers ITS4/ITS5 (White et al. 1990) and LR0R/LR5 (Vilgalys and Hester 1990;Cubeta et al. 1991) in Illustra PuReTaq Ready-To-Go PCR Beads (GE Healthcare, Pittsburgh, Pennsylvania).Amplicon purification and Sanger sequencing were conducted commercially by Macrogen (Seoul, Korea).Samples were sequenced bidirectionally to maximize confidence in automated base calls along the length of the amplicon.Sequence results (.AB1 files) were analyzed in MEGA 7.0.26(Kumar et al. 2016) and explored for the presence of ambiguous peaks and poor-quality reads.
Phylogenetic analysis-Nucleotide queries were conducted in GenBank.From these queries, two nucleotide data sets were compiled: the first for Chytridiales and the second for Rhizophydiales.For OAS2, 5, and 6, which allied to the order Rhizophydiales, multiple sequence alignments (MSAs) were created using type strains from this order, additional strains of Rhizophydium, and outgroup strains in the Rhizophlyctidales and Spizellomycetales (TABLE 1).Specifically, we aligned ITS and 28S loci separately using the online version of MAFFT 7 (Katoh et al. 2002(Katoh et al. , 2017)).MSAs were visually inspected and manually edited in AliView 1.26 (Larsson 2014; by end-trimming sequences to equal length, removing insertions that existed in a single sequence, and manually correcting nucleotide misalignments).Loci were concatenated in SeaView 4.6.1 (Gouy et al. 2010).The concatenated alignment was partitioned by locus, as determined by the "greedy" algorithm (Lanfear et al. 2012) of PartitionFinder 2.1.1 (Lanfear et al. 2016).A maximum likelihood (ML) phylogeny was inferred with a GTR+G model of evolution, identified as the best substitution model in PartitionFinder.Bootstrap values were generated from 1000 pseudoreplicates on the concatenated ITS-28S MSA in RAxML 8.0.0 (Stamatakis 2006(Stamatakis , 2014;;Stamatakis et al. 2008).To supplement ML inferences, Bayesian inference (BI) analysis was conducted with MrBayes 3.2.7a(Ronquist et al. 2012) using the GTR+G model with 10 million generations (typically a sufficient number of generations for the standard deviation of split frequencies to fall below 0.01), sampling trees every 100 generations and applying a 25% burn-in to calculate posterior probabilities (PPs).
For OAS3, which was allied to the taxonomic order Chytridiales, the alignment of Leshem et al. (2016) was downloaded from TreeBASE (S17951), expanded with 28S sequence of this strain, and subsequently aligned using MAFFT, and manually edited in AliView as described above.A ML phylogeny was inferred with the GTR+G model of evolution (identified as best substitution model with PartitionFinder) and subsequently bootstrapped after 1000 pseudoreplications in RAxML.Trees and alignments for Chytridiales and Rhizophydiales used in this analysis were deposited in TreeBASE (S26062).
Transmission electron microscopy-Several agar plates of Chytridiomycota strain OAS6 were grown at room temperature for 3 d.Zoospores were harvested by first flooding plates with sterile water and waiting for zoospore release (Letcher and Powell 2005).Once sufficient zoospores were present, they were harvested by pipette and transferred to a 1.5-mL microcentrifuge tube.Zoospores were fixed in a solution of 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 2 h at 22 C, followed by an additional 20 h at 4 C. Fixed zoospores were pelleted by centrifugation at 1000 rpm for 20 min.The pelleted zoospores were washed two times in 0.1 M sodium cacodylate buffer (pH 7.4) and postfixed in 1% osmium tetroxide for 1 h at 22 C.The pellet was washed twice with distilled water and embedded in HistoGel (Richard-Allan Scientific, San Diego, California).The solidified HistoGel containing zoospores was dissected into 1-mm 3 pieces and dehydrated in a graded series (25%, 75%, and 95%) of acetone for 10 min per step.After this graded series, the HistoGel block was rinsed twice in 100% acetone for 20 min.The blocks were then embedded in araldite cy212 epoxy resin and polymerized for 20 h at 60 C. Sections were cut with a diamond knife using a Reichert Ultracut E microtome (Leica, Wetzlar, Germany) and oriented on a copper grid.Sections were poststained for 20 min using uranyl acetate followed by lead citrate for 15 min.Stained sections were examined with a JEM-2100 field emission electron microscope (JEOL, Tokyo, Japan).

RESULTS
Several locations containing standing freshwater were sampled from the Salalah region in southern Oman and subsequently baited with pine pollen.From these efforts, Chytridiomycota were cultivated at only one location.
Zoospore ultrastructure-Of the three strains molecularly allied to Rhizophydium, strain OAS6 was selected for transmission electron microscopy (TEM).The zoospores A hemispherical vesicular area was present between the kinetosome and the ribosomal aggregation (FIG.3D, F).The kinetosome and nonflagellated centriole were nearly parallel and connected by moderately electron-dense fibrillar material in a broad zone of convergence (FIG.3F).After exploring hundreds of sections, OAS6 zoospores appeared to lack a spur over the kinetosome, a microtubule root from the kinetosome to the rumposome, and an electron-dense flagellar plug.Electrondense inclusions were consistently present near the base of the flagellum.
Molecular phylogeny-The ML phylogeny of the Rhizophydiales placed strains OAS2, 5, and 6 in the genus Rhizophydium with strong support (FIG.4).These three strains were monophyletic with 100% bootstrap support, whereas Pateramyces corrientinensis and Staurastromyces oculus grouped sister to Rhizophydium with 76% ML bootstrap support.Within Rhizophydium, OAS strains grouped sister to the morphologically uncharacterized strain MP049, but with poor ML support.BI analysis of the Rhizophydiales produced an identical topology to that by ML inferences, with greater than or equal support for all nodes.In the ML phylogeny of combined ITS-28S of Rhizophydium, the OAS strains grouped with 100% bootstrap support, yet branched sister to a clade containing other uncharacterized strains, MP043 and ARG016, also with poor support (SUPPLEMENTARY FIG. 1).Tests of phylogenetic hypotheses with ML analysis of the expanded alignment from Lesham et al. (2016) placed OAS3 in Dinochytrium kinnereticum with 100% bootstrap support (SUPPLEMENTARY FIG. 2).Etymology: jobii (Latin), named after Job (Ayyūb), an important figure throughout the Abrahamic religions, whose purported Omani tomb is near the collection site of these strains.Diagnosis: Distinguished from Rhizophydium globosum and R. brooksianum by the absence of a prominent laminated spur in TEM analysis of zoospores and by phylogenetic inference.

DISCUSSION
From a single sample in Oman, Dinochytrium kinnereticum was recovered from benthic detritus in an ephemeral freshwater pool.The reisolation of D. kinnerecticum in Oman is not surprising, as it was originally isolated from the Sea of Galilee (Leshem et al. 2016) within the greater Middle East region.Regardless, the reisolation of Dinochytrium kinnereticum in Oman expands the geographic range of this species and suggests that D. kinnereticum is likely widely distributed, aided by nutritional mode plasticity, inferred by its reported facultative parasitism of dinoflagellates.In general, the host range of Chytridiomycota is increasingly difficult to discern due in part to their ability to aerial disperse (Tipton et al. 2019).Consequently, any perceived biogeography could be driven by sampling effort.
Our other three strains, with nearly identical ITS sequences, formed a species-level clade within the taxonomic order Rhizophydiales.The resulting trees were congruent with other taxonomic studies of the Rhizophydiales (Van den Wyngaert et al. 2017).The   Omani strains clustered within Rhizophydium; however, in the expanded Rhizophydiales tree (FIG.4), OAS strains group sister to MP049, whereas in the Rhizophydium tree (SUPPLEMENTARY FIG. 1) they group sister to MP043.In both trees, there was low support for deeper branches within Rhizophydium, underscoring the need for expanded sampling efforts to resolve ambiguous phylogenetic relationships among the geographically diverse strains circumscribed within this genus (Letcher et al. 2006(Letcher et al. , 2008;;Davis et al. 2013).
Many previously described Chytridiomycota taxa are awaiting rediscovery and subsequent assignment of nucleotide information.The majority of Rhizophydium species are morphologically defined, missing type material, and lack molecular annotation (Letcher and Powell 2012).Morphologically, the size range and presence of an exit orifice observed in R. jobii is consistent with Rhizophydium polystomum from New Zealand (Karling 1968), which is among the many species without available sequence data.Although morphologically similar, R. jobii has prominent exit papillae, contrary to Karling's description of R. polystomum that it possesses "barely perceptible" exit papillae.In addition, R. jobii is morphologically congruent with a vague description of R. signyense (Willoughby 1971), originally isolated from soil in Antarctica and described as "saprophytic on chitin."This species, also without sequence data, has branched rhizoids attached at a single point and papillae that deliquesce to liberate zoospores.The morphological congruence of R. jobii with the description assigned to R. signyense hinders the ability to definitively exclude the possibility of R. signyense reisolation.However, the great distance between Oman and Antarctica, as well as R. jobii's ability to grow in the absence of chitin, renders it unlikely to be the same organism described by Willoughby.Overall, the ultrastructural characteristics of R. jobii are congruent with many described species within the Rhizophydiales.Specifically, R. jobii had a hemispherical vesicular area between the kinetosome and the ribosomal aggregation, possessed numerous electron-dense inclusions dispersed throughout the zoospore, and lacked a flagellar plug, consistent with other descriptions of strains within Rhizophydiales.Although morphologically congruent with the Rhizophydiales and molecularly allied to Rhizophydium, R. jobii possessed unique zoospore ultrastructural features that are incongruent with Rhizophydiaceae and the sole genus it circumscribes, Rhizophydium.Specifically, both family and genus are characterized as having zoospores with (i) a laminated curved spur present, (ii) a kinetosome and nonflagellated centriole connection with a wide zone of convergence, and (iii) a microtubule root present (Letcher et al. 2006(Letcher et al. , 2008)).Based on TEM analysis, our strain is similar by possessing a wide zone of convergence between the kinetosome and the nonflagellated centriole; however, it is unique by possessing electrondense vesicles in the posterior of the zoospore, while missing a laminated curved spur and a microtubule root.These unique ultrastructural features support previous hypotheses that Rhizophydiaceae could contain multiple genera (Letcher et al. 2008).As many strains molecularly allied to Rhizophydium remain morphologically uncharacterized, we chose to neither emend the family and genus descriptions nor erect a monotypic genus.If additional Rhizophydium species are found to possess similar ultrastructural features to R. jobii, it would seem appropriate to establish a separate genus.
Ultimately, we believe that these OAS strains represent a new species due to low percent identity to other Rhziphydium sp.strains, low bootstrap support for inclusion of other strains in the species, long branch lengths between R. jobii and other strains, and the unique morphological features of the zoospore.Although relationships within the genus Rhizophydium remain unresolved, these strains expand the known range of Chytridiomycota onto the Arabian Peninsula, while expanding the diversity within the Rhizophydiales.

Figure 4 .
Figure 4. ML tree of Rhizophydiales inferred from analysis of combined ITS and 28S data.Nodes labeled with values (PP/ML bootstrap proportions).The new species, Rhizophydium jobii, is indicated in bold.Members of the Rhizophlyctidales and Spizellomycetales were used as outgroups.