Insights & Perspectives

No jacket required – new fungal lineage Think again defies dress code

Recently described zoosporic fungi lack a during trophic phase

Timothy Y. James1) and Mary L. Berbee2)

Analyses of environmental have provided tantalizing evidence for species). Ranging from unicellular organ- ‘‘’’ or ‘‘cryptomycota’’, a of mostly undescribed and deeply isms to some of the largest and most long- lived of all organisms [3, 4], the remain- diverging aquatic fungi. Here, we put cryptomycota into perspective through ing fungi are hiding all around us and consideration of , the only clade member growing in culture. This is modern approaches to studying diversity timely on account of the publication in Nature of the first images of uncul- and communities are beginning to reveal tured cryptomycota from environmental filtrates, where molecular probes the true phylogenetic diversity of the revealed non-motile cyst-like structures and motile spores, all lacking typical group. After a full decade of progress in fungal chitinous cell walls. Current studies of Rozella can complement these understanding fungal diversity using environmental DNA community studies, fragmentary observations from environmental samples. Rozella has a fungal- we now realize that most of the fungi in specific synthase and its resting sporangia have walls that appear to the environment do not actually match contain chitin. Cryptomycota, including Rozella, lack a cell wall when absorb- those specimens from herbarium cabi- ing food but like some other fungi, they may have lost their ‘‘dinner jacket’’ nets and culture collections that were through convergence. Rather than evolutionary intermediates, the cryptomy- used to build the fungal tree of life [2, 5]. Most fungal sequences from environ- cota may be strange, divergent fungi that evolved from an ancestor with a mental DNA studies can be assigned to a nearly complete suite of classical fungal-specific characters. described class or even genus [6–8], but some represent unknown taxa on deeply Keywords: diverging branches [9–12]. When line- .chitin synthase; cryptomycota; evolution; Rozella; rozellida ages known exclusively from environ- mental DNA sequences cannot be assigned to a phylum, they challenge Introduction form the traditional basis for detection our understanding of the biodiversity and classification, they appear only and phylogeny (breadth and depth), and even characteristics of fungi. A Fungi lead hidden lives briefly in the life of a . Add to this that many groups of fungi are difficult or recent breakthrough by Jones et al. [13] Fungi lead cryptic lives by growing inside impossible to obtain in pure culture [1, 2], on the diversity and characteristics of theirfoodsource,andiftheyemergeitis and the result is that by most estimates one such enigmatic lineage, named only to reproduce as mushrooms, cups, less than 10% of all fungi have been ‘‘cryptomycota’’, raised the possibility or other spore-producing structures. observed and formally described (100 that the lines dividing fungus from the Although the reproductive structures thousand out of an estimated 1.5 million protozoan soup from which they evolved [14] may be fuzzier than appreciated. Specifically, Jones et al. concluded that DOI 10.1002/bies.201100110 the widespread group cryptomycota were intermediate between fungi and 1) Department of Ecology and Evolutionary *Corresponding authors: ancestral protists. In their words: Biology, University of Michigan, Ann Arbor, MI, Timothy Y. James USA E-mail: [email protected] ‘‘Co-staining with cell wall markers dem- 2) Department of Botany, University of British Mary L. Berbee onstrates that representatives from the Columbia, Vancouver, BC, Canada E-mail: [email protected]

94 www.bioessays-journal.com Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc...... Insights & Perspectives T. Y. James and M. L. Berbee

clade do not produce a chitin-rich cell or they have been inaccurately assigned to form the primary (basal-most) branch wall during any of the life cycle stages to other phyla. LKM11-related sequences on the fungal tree. Lara et al. [22] coined observed and therefore do not conform to were recovered in essentially all the name ‘‘rozellida’’ for the clade. hn again Think thestandardfungalbodyplan’’[13]. environmental DNA surveys of fresh- Because the trophic phase of Rozella water aquatic ecosystems, such as the lacks a cell wall and may have retained Our objectives here are to review the pH 2.0 Rio Tinto in Spain [16], anoxic the ability to phagocytose host cyto- current state of knowledge of cryptomy- sediments in lakes [17], and among plasm, Lara et al. suggested that rozel- cota with a focus on its culturable picoeukaryotes (<5 mM) in lake water lida might be parasites positioned on species in Rozella. We provide evidence [18]. The LKM11 sequences were also the most deeply diverging branch of showing that, although the cryptomy- described from terrestrial and marine Fungi that retained ancestral cota may have diverged early from other systems, including the rhizosphere sur- protistan characteristics. fungi, they have the capacity to make rounding aspen roots [19], anoxic Jones et al. [13] opened the door to a chitin-rich walls. While the cryptomy- coastal sediments [20], and deep-sea broader view of the environmental clade cota do not conform to the ‘‘standard sediments [21] (Fig. 1). These obser- by adapting state-of-the-art cytological fungal body plan’’, derived adaptation vations suggested that LKM11, like fungi and nucleic acid probing techniques to to intracellular parasitism should be in (chytrids), repro- directly observe cells [36]. They coined considered as an alternative expla- duce with motile spores and thus thrive the name cryptomycota for the group, to nation to retention of ancient, inter- in freshwater as well as in soil and highlight its cryptic nature and its char- mediate characters. marine ecosystems. By 2010 over 30 cul- acters, which they, like Lara et al., inter- ture-independent environmental DNA preted as intermediate between fungi Cryptomycota: Unicellular studies documented the presence of and ancestral protists (Box 2). Jones bodies linked to LKM11 either as an early-diverging fun- et al. used tyramide signal amplifica- environmental lineages gal clade or as a close relative to the tion-fluorescent in situ hybridization fungi [21]. Because no member had ever [36] (TSA-FISH) to identify cells of cryp- Cryptomycota were first detected as been seen, the group remained an tomycota in filtrates from multiple sour- DNA sequences occurring in mesocosms enigma. ces, including pond water from the of non-sterile water from Lake The first breakthrough on the place- campus of University of Exeter. The Ketelmeer, the Netherlands [15]. The ment of the LKM11 clade was the demon- cryptomycota cells took three forms. cloned sequences were loosely related stration of a relationship to the aquatic Most stained with antibodies to a-tubu- to fungi, but showed only a distant genus Rozella with robust phylogenetic lin, appearing similar to the uniflagel- match to any of the known sequences support [22]. Rozella is an internal para- lated zoospores of chytrids. Some lacked at the time and the group was named site, primarily of water molds [23], that flagella, having either encysted or lost after one of the clones, LKM11. After this was once classified in the order their flagellum during preparation. The discovery, the following decade of (Chytridiomycota) third cell type appeared to be attached environmental DNA surveys uncovered [24] (Box 1, Fig. 2). However, James to other cells, sometimes to diatoms, sequences related to LKM11 numerous et al. [25] demonstrated clearly (with and Jones et al. hypothesized that this times. The sequences have variously statistical support) that the older classi- was a parasitic or saprotrophic associ- been called ‘‘fungi’’, a ‘‘novel clade’’, fication was wrong and Rozella diverged ation. A key conclusion of the paper was that none of the many cryptomycota cells could be stained for the presence of chitin or cellulose with calcofluor freshwater white or wheat germ agglutinin. The (35)% wall-less cryptomycota and the genus Rozella, having drawn their origin from the primary node on the fungal tree of life, have prompted us to critically con- freshwater sider the characteristics that both define sediment fungi and led to their dramatic success (28%) sea water as dominant terrestrial forms. (4%)

marine sediment Cryptomycota and the (15%) evolutionary origins of a rhizosphere, soil or rock (17%) chitinous wall

Figure 1. Proportion of habitats from 43 environmental DNA studies in which cryptomycota The cell wall of most fungi consists of b- have been detected. Data from Table 1 of the supporting information of Jones et al. [13]. These data do not reflect the frequency of surveys reporting cryptomycota in the various 1,3 glucans, chitin, mannans, and gly- habitat types, because cryptomycota are absent from many marine environmental DNA stud- coproteins, with chitin microfibrils play- ies but are very common in freshwater studies. ing a major role in tensile strength and

Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc. 95 T. Y. James and M. L. Berbee Insights & Perspectives .....

Box 1 chondria were inside a parasitic vacuole [33]. During reproduction of the parasite, the host displays hypertro- What are the Rozella parasites? phy and its transcriptome or proteome is somehow co- As Rozella is the only member of the cryptomycota for opted into making the septa/cell walls that the parasite which the morphology and life cycle have been described, uses to produce zoosporangia (Fig. 2E). The septa are we briefly review what is known about the genus. Rozella required to develop the pressure needed for forcible dis- consists of obligately biotrophic endoparasites that can charge of zoospores [23]. The walled segments of the host only be grown in dual culture with their hosts [23]. Hosts may also be converted into the parasite’s pigmented and include aquatic molds in the Oomycota and Fungi. One thick-walled resting sporangia (Fig. 2F).

Think again species, R. coleochaetis, has been reported from the The source and the chemical composition of the para- green alga [26]. Experimental inoculations site’s resting sporangial wall are unknown. We stained a have suggested that the host range of each species culture of R. allomycis using methods similar to those is limited, with most evidence pointing to either used by Jones et al. [13]. As expected [34], calcofluor species- or genus-level host specificity [27–29]. The white stained the cell walls of , the host, provid- best-known species is R. allomycis, a parasite of the ing evidence for chitin (Fig. 2H). Mature resting spores of common ‘‘model’’ water mold genus Allomyces Rozella did not stain, indicating that the outer ornamen- (). R. allomycis is also relatively com- tation lacked these polysaccharides (Fig. 2J). However, mon, occurring on 2 of the 43 Allomyces isolates reported the immature, unpigmented resting spore walls in Rozella by Wolf [30]. The infection begins with posteriorly unifla- did contain chitin or cellulose judging from their strong gellate, wall-less zoospores of the parasite swimming to staining with calcofluor white (Fig. 2J). In combination with an uninfected host (Fig. 2A). The spores attach to the host, the detection of a chitin synthase homolog in the retract their flagellum, and form a cyst on the surface of R. allomycis genome, these data suggest that the parasite the host cell (Fig. 2B). The cysts begin to develop a cell can make its own chitinous resting sporangium. Some wall and form a penetration tube. A vacuole forms at the parallels can be drawn between the injection of naked posterior end of the cyst (2C and D), and the parasite protoplasts of Rozella into the host through enlargement cytoplasm is injected into the host through a wall that is of a vacuole, and the mechanism by which the protoplasm apparently weakened by the parasite [31]. Once inside the of a microsporidian spore is rapidly ejected into a host cell host, R. allomycis then grows as a wall-less form that through expansion of a posterior vacuole [35]. This mode feeds on the host cytoplasm [32]. A naked thallus may of infection may reflect an evolutionarily conserved mech- have the advantage of being able to proliferate through anism, determinable perhaps if the phylogenetic relation- the host mycelium, allowing the parasite to squeeze ship between and Rozella, demonstrated in through the occasional partial septa. The naked thallus at least one phylogenetic study [25], can be rigorously may also facilitate of the host’s cytoplasm, tested. as suggested for R. polyphagi in which the host’s mito-

structural integrity [39]. Chitin syn- Compared with walls from most other Most fungal genomes (except thases are widely distributed among organisms, walls of fungi are distinctive Schizosaccharomyces) encode at least . Several of the divergent in three ways. Their synthesis involves one chitin synthase from each of two protists with recently the combined action of an exceptionally deeply diverging divisions (Table 1, sequenced genomes from the ‘‘Origins large number of chitin synthases; they Fig. 3). Further, the newly sequenced of Multicellularity’’ project [40] have are continuously remodeled to permit genomes of early diverging fungal phyla chitin synthases (Fig. 3, Table 1). active growth; and they surround fungal have twice the number of chitin Insects have two chitin synthases cells that are actively taking up synthases as the better-known (Table 1) [41]. Even humans have hya- nutrients [44–46]. Possibly related to (Table 1). In the fungal luronan synthases that produce hyalur- the complexity and distinctive charac- model systems in Ascomycota, where onic acid but are homologs to chitin teristics of the fungal wall, the gene chitin synthases are best characterized, synthases (Table 1) [42]. Like chitin, hya- duplications that gave rise to the oldest paralogous proteins differ in timing and luronic acid is secreted to the outside of of the fungal chitin synthases are more location of activity [47]. Among the chi- cells, but it becomes a component of ancient than the divergences of the tin synthases, the division 2 genes form vitreous humor in the eye and synovial fungi themselves [44] (Fig. 3). a monophyletic group known only from fluid in joints [43]. The wide phyloge- While chitin alone cannot define fungi and microsporidia [48]. Class IV netic distribution of chitin synthases fungi, the presence of division 2 chitin enzymes from division 2 have usually together with evidence (Fig. 2J) of a wall synthases, and especially chitin syn- been found to synthesize the bulk in Rozella’s young spores suggest that thases with a myosin (Fig. 3), of the chitin in walls [45, 47]. cryptomycota, or their recent ancestors, along with transport of chitosomes Microsporidia have only one chitin syn- have or had a chitinous wall at some life along the cytoskeleton, may be unifying thase, a division 2, class IV chitin syn- history stage. characters for most of Kingdom Fungi. thase for spore wall production that is

96 Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc...... Insights & Perspectives T. Y. James and M. L. Berbee hn again Think

Figure 2. Life cycle of Rozella allomycis, a parasite of Allomyces. Photos are of strain CSF55 are important in polar growth in other isolated from Hattiesburg, MS, USA. A: Posteriorly uniflagellate zoospores. Note the refractive fungi [49–51]. Chitin synthases must be lipid sac and slipper shape. B: Aggregation on host hyphae. C: Posterior vacuole observed correctly targeted to the plasma mem- at end of cyst as contents are injected into host cell. D: Empty cysts (c) on host hypha; note brane [52]. In filamentous fungi, chitin germ tube visible on one cyst and injected young amoeboid thallus appearing inside the host synthases, packed in vesicles called chi- near cysts. E: Early stages of zoosporangium formation; note septa (s) produced by the host that separate the parasite’s zoosporangia (z). Five zoosporangia are shown, and the terminal tosomes, are transported along a cyto- one displays a discharge papillum (p). F: Developing parasite resting sporangia delimited by skeletal highway to the hyphal tip. In host septa. Immature resting sporangia lack brown pigment. G: Mature, thick-walled resting Ustilago, an analysis of mutants and of sporangia. H: Resting sporangia in G stained with calcofluor white (marker for chitin and localization of fluorescently tagged cellulose). Host hyphae and cross walls, but not parasite resting sporangia, stain. I: Resting proteins showed that the myosin sporangia removed from host cells; an immature and mature sporangium is shown. J: domain is essential for exocytosis of Immature wall of resting sporangium (from I) stains with calcofluor white, but the pigmented mature wall does not, possibly because final wall layers mask the inner polysaccharides. its chitin synthase, and possibly also Scale bar ¼ 5 mm in A and B; 10 mm in C, D, I, and J; 20 mm in E–H. for its short range transport along the actin cytoskeleton to the hyphal tip [51]. The earliest diverging fungi are mostly recognizable even though the highly JN646249), confirming that the parasite like Rozella, with the main body grow- reduced genomes of microsporidia is capable of producing a chitinous wall ing isotropically to form a rounded evolved very quickly, erasing most at some stage of its life cycle. globule, rather than elongating at a traces of ancestral relationships [48] narrow hyphal tip. Surprisingly, even (Fig. 3). We predict that, at the least, a A fungus-specific solution to Chytridiomycota with globular bodies recognizable ortholog to the division 2 targeting chitin synthesis to sites have several paralogs of chitin syn- class IV fungal gene will be detected by of active growth thases with myosin domains (Fig. 3, sequencing a genome of Rozella or other Table 1). Like the Chytridiomycota, cryptomycota. Preliminary results from Remaining to be determined is whether Rozella may require polar wall depo- draft genome sequencing of R. allomycis the Rozella genome has the distinctive sition for spatial orientation at specific indeed have revealed a division 2 syn- fungal-specific chitin synthases with an life history stages, for example, when thase (GenBank accession number N-terminal myosin head domain that producing specialized apical exit papil-

Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc. 97 T. Y. James and M. L. Berbee Insights & Perspectives .....

cysts are parasitic requires further Box 2 investigation, but if they are, they What should we call the clade? may – as in Rozella –injecttheirpro- This essay revolves around a possible new fungal phylum that as yet lacks a toplasm into a host cell. formal name. It was first tagged as LKM11, the code name for a DNA sequence If they originated from the basal clone. Not intending formal naming, Jones et al. [13] proposed cryptomycota node of the Fungi, cryptomycota have and Lara et al. [22] proposed rozellida as provisional names for the group. Here had ample time for diversification. we use ‘‘cryptomycota’’ to emphasize that the clade is fungal but its diversity How successful have cryptomycota is largely cryptic. We could equally well have adopted the name rozellida to been over this time period, and are emphasize the connection with Rozella. In the spirit of rationalizing nomen- they as common as fungi in their pre-

Think again clature, new higher level names for fungal taxa are, where possible, based on ferred environments? Using an argu- and typified by their first described genus. Newly rationalized names receive ment based on phylogenetic branch wide support [37] appearing in GenBank and in standard references such as lengths, Jones et al. suggest that cryp- the Dictionary of Fungi [38]. As a formal name, Cryptomycota would be tomycota radiated to become nearly as unacceptable because Cryptomyces is already used for a genus in diverse as all other fungi, although this Ascomycota. So, if established as a new phylum, the group could be called estimate could be biased if the crypto- ‘‘Rozellomycota.’’ mycota, like some other intracellular parasites, have unusually rapid rates of substitution [61]. The cryptomycota are primarily aquatic and are largely lae for zoospore escape, or when form- could also represent secondary loss of a absent in studies of airborne fungal ing germ tubes to penetrate host cells wall, an adaptation to intracellular para- particles [62–64]. Phylogenetic (Fig. 2D). sitism. Other pathogenic fungi in analyses suggest that other fungal A strikingly different mechanism that normally produce chitinous walls phylaoriginatedonlandorinfresh- of localization of chitin synthases can also grow inside a host cell as a water [65]. Finding cryptomycota in a evolved in Saprolegnia, an wall-less trophic form. Wall-less trophic cold methane seep [66] and deep-sea (Straminopila). Superficially similar to forms are found in Blastocladiomycota sediments [21] justifies raising the Fungi, Saprolegnia is related instead to [55], Entomophthoromycotina [56], possibility that they first diversified diatoms and brown algae (Fig. 3) and microsporidia [57], and Beauvaria in the sea, like many and pro- cellulose, not chitin, is the main con- (Ascomycota [58]). Chitin triggers a tists, but unlike most other fungi. In stituent of its walls. Along with strong innate immune response from general though, most extant cryptomy- other Straminopila including diatoms, animals and plants [52], and its absence cota prefer freshwater habitats; they Saprolegnia nonetheless has genes during intracellular parasitism suggests decreased in abundance along a sal- sharing sequence motifs with fungal convergent adaptation to avoid host inity gradient in a salt marsh in division 1 chitin synthases (Table 1). detection. These examples set the expec- Rhode Island [67], their frequency is However, the lack of chitosomes in tation that stage-specific suppression of low in the open oceans [13], and they Saprolegnia suggests that it evolved chitin synthases is relatively easy. areabsentfromsomedeep-seasurveys an alternative way to shuttle chitin syn- [68, 69]. Environmental DNA studies of thases to the hyphal tip [53]. Two of its picoeukaryote communities in lakes six chitin synthase paralogs have How much of the have documented 2:1 to 3:1 (fungi: microtubule interacting and sorting cryptomycota life cycle cryptomycota) [18, 70, 71]. However, domains, which are not present in chi- in at least one study that utilized tin synthases of any true fungus [54]. and ecology do we know TSA-FISH rather than PCR-generated While the origin of chitin synthases is clone libraries to survey eukaryotic ancient, the mechanisms for subcellu- The life cycle stages detected in uncul- groups in French lakes, cryptomycota lar localization appear to have evolved tured cryptomycota are also found in were found to be more abundant [72]. independently. Rozella but the converse is not true. Such studies of fungi and cryptomy- Zoosporangial and resting stages are cota in filtered water may be the aquatic Evolution of fungus-specific known only from Rozella.Anopen analog of the many studies that have chitin synthases: Loss is easy, question is whether Jones et al. cap- sampled airborne propagules of fungi gain was rare tured a trophic phase among the [73]. These studies may provide a motile, unattached cysts, or among skewed picture of the community as The cryptomycota lack cell walls in some the attached cells that they filtered number of propagules produced may stages. Unlike most fungi, Rozella species from pond water. The unattached cysts reflect reproductive strategies and lack walls when taking up nutrients may be able to phagocytose cells such phenology more than population cen- and may use phagotrophy rather than as bacteria or picoplankton, or they sus size [63]. Resolution of the complete absorptive nutrition across a cell wall. may be a transitional, amoeboid crawl- life cycle of cryptomycota should come The lack of a wall during feeding may ing phase, as observed in members from additional studies utilizing frac- be a primitive character retained by of Ichthyosporea () tionation of environmental samples or the common ancestor of all fungi, but [59, 60]. Whether or not the attached direct sampling of each of the many

98 Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc...... Insights & Perspectives T. Y. James and M. L. Berbee

Table 1. All sequenced fungal genomes have numerous paralogous genes with motifs characteristic of chitin synthases, and copies of the deeply diverging division 1 and division 2 genes are present in all fungal phyla hn again Think Any paralogs Presumed functional motifsc Gene and source with motifs (phylum or higher group, genus, No. of related to Substrate Substrate Catalytic accession numbera of example) paralogs transport?b binding binding base Processivity Processivity

Chitin synthase division 1 consensus T(MY)NE DXGT LAEDRIL QRRRW (S/T)WG Ascomycota, Aspergillus, 4 No TYYNE DAGT LAEDRIL QRRRW SWG ANID_04367 Blastocladiomycota, Allomyces, 10 No TMYNE DVGT LAEDRIL QRRRW SWG AAMAG 10750.1 Chytridiomycota, Batrachochytrium 3 No TMYNE DVGT LAEDRIL QRRRW SWG BDEG 08256.1 Chytridiomycota, Spizellomyces 6 No TMYNE DVGT LAEDRIL QRRRW SWG SPPG 04845.2 Zygomycota, Rhizopus 6 No TMYNE DVGT LAEDRIL QRRRW SWG RO3G_16230.3 Microsporidia, Encephalitozoon None No Straminopila (not Fungi), Saprolegnia 6 Yes; 30 microtubule TMYNE DVGT LAEDRIL QRRRW SWG SPRG_02074.2 interacting and sorting domain MIT PF04212 Chitin synthase division 2 consensus (T/P)(A/C) DADT LGEDR(YFE)L Q(R/G)RRW (S/T)WG Y(S/T)E Ascomycota, Aspergillus, 3 Yes PAYTE DADT LGEDRYL QRRRW SWG ANID_06318.1 Blastocladiomycota, Allomyces, 25 Yes; 50 myosin PCYTE DSDT LGEDRYL QRRRW SWG AMAG_07719.1 cd00124; also note 50 oligopeptide transporter protein in AMAG_15310.1 Chytridiomycota Batrachochytrium, 11 Yes; 50 myosin PCYTE DADT LGEDRYL QRRRW SWG BDEG_03361.1 cd00124 Chytridiomycota Spizellomyces 11 Yes; 50 myosin PCYTE DADT LGEDRFL QRRRW SWG SPPG 03441.2 cd00124 ‘‘Cryptomycota’’ Rozella GenBank na Unknown TCYSE DADT LGEDRYL QRRRW SWG JN646249 Zygomycota Rhizopus 20 Yes; 50 myosin PCYTE DADT LGEDRYL QRRRW SWG RO3G_17187.3 cd00124 Microsporidia, Encephalitozoon 1 No TCYSE DADT LGEDRYL QRRRW SWG GenBank XP_965977 Straminopila (not Fungi), Saprolegnia None No Outgroups Choanoflagellata (unicellular 1 No PCYNE DCGT LAEDRFL QRRRW TWG opisthokont) Salpingoeca PTSG 01414.1 Choanoflagellata Salpingoeca PTSG 130 SAM domain TMYNE DADI MGEDRWL QRRRW SWG 01542.1 Filasporea (unicellular opisthokont) 1 No ADFDN DGDV LGEVP-L QRRRW RWG CAOG 03353.2 Apusozoa (unicellular) Thecomonas 1 No PNVTL DGDT MGEDRWL QRKRW SWG AMSG_12058.2 Arthropoda, Drosophila, GenBank 2 Yes; 30 transmembrane TMWHE DGDI QGEDRWL QRRRW SWG NP_730928 amino acid transporter domain, pfam03845 Chordata, Homo, hyaluronan 4 No SAYQE DSDT FGDDRHL QQTRW GWG synthase, GenBank AAH35837 a Sequences are from the Broad Institute http://www.broadinstitute.org/ unless otherwise noted. b From comparisons with the Conserved Domain Database [79]. c Division-specific sequence motifs are from Choquer et al. [80]. We found fungal-specific chitin synthases with BLAST searches using as our queries the conserved amino acid domains from division 1 Chs2p gene (DVGTRL...HDVSWG, GenBank NP_009594.1) from Saccharomyces cerevisiae and division 2 class V gene from Aspergillus nidulans (HHHIRN...DDFSWG, GenBank AAB05797.1), with e-30 as a cutoff for fungi, e-9 for outgroup organisms. To find the motifs, we aligned sequences using MUSCLE [81]. Outside of fungi and animals, gene function is unknown and homology with the first substrate binding domain is uncertain.

Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc. 99 T. Y. James and M. L. Berbee Insights & Perspectives .....

Agaricus the notoriously unstable microsporidia Puccinia [25] raises the possibility that the basal Ustilago Schizosaccharomyces position of the entire cryptomycota Aspergillus clade may be an artifact. Phylogenetic

Ascomycota Saccharomyces error can result from violations of stand- Glomus ard molecular evolution models, such as Rhizopus heterotachy (variation in site-specific Basidiobolus zygomycota 䇾 䇿 DNA substitution rates over time) [76]. Olpidium Fungi Coemansia A multilocus phylogeny using sequen- Entomophthora ces from complete genomes of Rozella, onts onts

Allomyces Blastocladiomycota k Think again and ideally other cryptomycota, would Physoderma help to place these key taxa. No Neocallimastix matter what its position, cryptomycota Spizellomyces Chytridiomycota opistho Batrachochytrium will undoubtedly remain a highly diver- Rozella gent group of fungi with characters that NAMAKO-36 cryptomycota help illuminate early fungal evolution. LKM11 However, the jury is still out about Lily_stem_CM2 how early these fungi diverged from Microsporidium Microsporidia Encephalitozoon all others. Nuclearia protists Homo Drosophila (fly) animals Conclusions Monosiga Sapingoeca protists; etc. The unveiling of cryptomycota over the Capsaspora last decade has revealed a prominent branch of ubiquitous and diverse organ- Abeoforma protists; Ichthyosporea Ichthyophonus isms that straddle the divide between Sphaeroforma fungi and the opisthokont protozoa Physarum from which they evolved. Studies of slime molds Dictyostelium these organisms from environmental Phaeodactylum Straminopila samples will continue to be important Saprolegnia Arabidopsis green plants in illuminating habitat and host relationships, while cultured isolates Chitin synthases Body type during feeding Motile dispersal? Any chitin synthase homolog Filament ous, has wa ll of Rozella serve as cryptomycota’s most Division 1 Unicellular yeast, has wall Single posterior flagellum tractable representatives for experimen- Division 2, no myosin Unicellular, not yeast, has wall tal and genomic analysis. Our detection Unicellular, not yeast, no wall amoeboid of the chitinous wall of Rozella required Division 2 with myosin domain staining of immature resistant sporan- Figure 3. Diagrammatic tree of fungi showing that the evolutionary divergence of many chitin gia that have yet to be detected among synthase paralogs predates all phyla of fungi, except, perhaps, the cryptomycota. We environmental samples hybridizing to hypothesize that the ancestor of cryptomycota was able to make a chitinous cell wall cryptomycota probes. Showing that because many animals and all other fungi have chitin, calcofluor white stains Rozella’s resting Rozella produced one of the fungal- spores (Fig. 2J), and the genome of Rozella has at least one chitin synthase. This tree places specific chitin synthases required taxa mentioned in this review in a commonly accepted phylogenetic context. We coded taxa genomic analysis, as will the reconstruc- for gene presence based on comparisons of genes from sequenced genomes (Table 1), and then predicted that taxa lacking sequenced genomes would share the chitin synthases of tion of its full suite of chitin synthases, their closest relatives. now a work in progress. We predict that future analyses of environmental genomes will also reveal chitin syn- possible hosts that lurk in the sediments – that the cryptomycota originated from thases from other cryptomycota. The in which cryptomycota have been found the first divergence in the fungal tree of explosion of diversity of true fungi on using the developed nucleic acid probes life. But is this phylogenetic position land coincided with the evolution of [13, 72, 74]. correct? The earlier analyses of James polar tip growth facilitated by a semi- et al. [25] with the two available protein rigid cell wall. However, the semi-rigid coding loci in addition to ribosomal wall predated terrestrial fungi and is a Beware of error and DNA showed the position of Rozella to shared character of the cryptomycota uncertainty be generally consistent with Jones et al. and all other fungi. [13]. However, tree topology tests did Keeping in mind that their phyloge- not rule out alternative positions of netic diversity rivals that of the rest of Up until this point, we have accepted – Rozella among the fungi [75]. This, the fungi, the varied characteristics and based only on the ribosomal DNA locus along with the clustering of Rozella with ecological capabilities of the cryptomy-

100 Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc...... Insights & Perspectives T. Y. James and M. L. Berbee

cota will offer surprises. Major additions 9. Porter TM, Schadt CW, Rizvi L, Martin AP, 27. Held AA. 1974. Attraction and attachment of et al. 2008. Widespread occurrence and phy- zoospores of parasitic chytrid Rozella allomy- to the tree of life are far from unique to logenetic placement of a soil clone group cis in response to host-dependent factors. mycology, and the vastness of protistan adds a prominent new branch to the fungal Arch Microbiol 95: 97–114. hn again Think diversity is only recently becoming tree of life. Mol Phylogen Evol 46: 635–44. 28. Karling JS. 1942. Parasitism among the chy- widely appreciated [77]. New lineages 10. Schadt CW, Martin AP, Lipson DA, Schmidt trids. Am J Bot 29: 24–35. SK. 2003. Seasonal dynamics of previously 29. Shanor L. 1942. A new Rozella of the poly- are sprouting across the tree, such as unknown fungal lineages in tundra soils. sporangiate series. J Elisha Mitch Sci S 58: the uncultured rappemond algae [78], Science 301: 1359–61. 99–101. demonstrating that our understanding 11. Vandenkoornhuyse P, Baldauf SL, Leyval 30. Wolf FT. 1941. A contribution to the life history C, Straczek J, et al. 2002. Evolution – exten- and geographic distribution of the genus of global biodiversity has been skewed sive fungal diversity in plant roots. Science Allomyces. Mycologia 33: 158–73. towards organisms that can be cultured 295: 2051. 31. Held AA. 1973. Encystment and germination or that are macroscopic. Because 12. Takishita K, Miyake H, Kawato M, of the parasitic chytrid Rozella allomycis on these are the early days of research into Maruyama T. 2005. Genetic diversity of host hyphae. Can J Bot/Rev Can Bot 51: microbial in anoxic sediment 1825–35. uncultured cryptomycota, Lara et al. [22] around fumaroles on a submarine caldera 32. Held AA. 1980. Development of Rozella in and Jones et al. [13] are extrapolating floor based on the small-subunit rDNA phy- Allomyces – a single zoospore produces from fragmentary evidence. However, logeny. Extremophiles 9: 185–96. numerous zoosporangia and resistant spor- 13. Jones MDM, Forn I, Gadelha C, Egan MJ, angia. Can J Bot/Rev Can Bot 58: 959–79. by opening a window on uncultured et al. 2011. Discovery of novel intermediate 33. Powell MJ. 1984. Fine structure of the cryptomycota, their work is leading forms redefines the fungal tree of life. Nature unwalled thallus of Rozella polyphagi in its to stimulating new hypotheses that 474: 200–3. host Polyphagus euglenae. Mycologia 76: 14. Steenkamp ET, Wright J, Baldauf SL. 2005. 1039–48. challenge classical concepts of fungal The protistan origins of animals and fungi. Mol 34. Aronson JM, Machlis L. 1959. The chemical life cycles, ecological niches, and evol- Biol Evol 23: 93–106. composition of the hyphal walls of the fungus utionary trajectories. 15. van Hannen EJ, Mooij WM, van Agterveld Allomyces. Am J Bot 46: 292–300. MP, Gons HJ, et al. 1999. Detritus-dependent 35. Franzen C. 2004. Microsporidia: How can development of the microbial community in an they invade other cells? Trends Parasitol 20: experimental system: Qualitative analysis by 275–9. Acknowledgments denaturing gradient gel electrophoresis. Appl 36. Not F, Simon N, Biegala IC, Vaulot D. 2002. This work was supported in part by NSF Environ Microbiol 65: 2478–84. Application of fluorescent in situ hybridization DEB-0732984 Assembling the Fungal 16. Amaral-Zettler LA, Gomez F, Zettler E, coupled with tyramide signal amplification Keenan BG, et al. 2002. Eukaryotic diversity (FISH-TSA) to assess eukaryotic picoplankton Tree of Life and by a Natural Sciences in Spain’s River of Fire. Nature 417: 137. composition. Aquat Microb Ecol 28: 157–66. and Engineering Research Council of 17. Slapeta J, Moreira D, Lopez-Garcia P. 2005. 37. Hibbett DS, Binder M, Bischoff JF, Canada Discovery grant to Berbee. The extent of protist diversity: Insights from Blackwell M, et al. 2007. A higher-level phy- molecular ecology of freshwater eukaryotes. logenetic classification of the Fungi. Mycol Proc R Soc Lond Ser B: Biol Sci 272: 2073–81. Res 111: 509–47. 18. Lefe` vre E, Bardot C, Noel C, Carrias JF, 38. Kirk PM, Cannon PF, Minter DW, Stalpers References et al. 2007. Unveiling fungal zooflagellates JA. 2008. Ainsworth & Bisby’s Dictionary of as members of freshwater picoeukaryotes: the Fungi. Wallingford, UK: CAB International. 1. Fortin JA, Becard G, Declerck S, Dalpe Y, Evidence from a molecular diversity study in 39. Bowman SM, Free SJ. 2006. The structure et al. 2002. Arbuscular mycorrhiza on a deep meromictic lake. Environ Microbiol 9: and synthesis of the fungal cell wall. root-organ cultures. Can J Bot/Rev Can Bot 61–71. BioEssays 28: 799–808. 80: 1–20. 19. Lesaulnier C, Papamichail D, McCorkle S, 40. Ruiz-Trillo I, Burger G, Holland PWH, King 2. Hibbett DS, Ohman A, Glotzer D, Nuhn M, Ollivier B, et al. 2008. Elevated atmospheric N, et al. 2007. The origins of multicellularity: A et al. 2011. Progress in molecular and CO2 affects soil microbial diversity associated multi-taxon genome initiative. Trends Genet morphological taxon discovery in Fungi and with trembling aspen. Environ Microbiol 10: 23: 113–8. options for formal classification of environ- 926–41. 41. Qu MB, Liu TA, Yang J, Yang Q. 2011. The mental sequences. Fungal Biol Rev 25: 38–47. 20. Dawson SC, Pace NR. 2002. Novel kingdom- gene, expression pattern and subcellular 3. Stajich JE, Berbee ML, Blackwell M, level eukaryotic diversity in anoxic environ- localization of chitin synthase B from the Hibbett DS, et al. 2009. The fungi. Curr Biol ments. Proc Natl Acad Sci USA 99: 8324–9. insect Ostrinia furnacalis. Biochem Biophys 19: R840–5. 21. Nagano Y, Nagahama T, Hatada Y, Res Commun 404: 302–7. 4. Smith ML, Bruhn JN, Anderson JB. 1992. Nunoura T, et al. 2010. Fungal diversity in 42. Jing W, DeAngelis PL. 2003. Analysis of the The fungus Armillaria bulbosa is among the deep-sea sediments – the presence of novel two active sites of the hyaluronan synthase largest and oldest living organisms. Nature fungal groups. Fungal Ecol 3: 316–25. and the chondroitin synthase of Pasteurell- 356: 428–31. 22. Lara E, Moreira D, Lopez-Garcia P. 2010. amultocida. Glycobiology 13: 661–71. 5. Lutzoni F, Kauff F, Cox CJ, McLaughlin D, The environmental clade LKM11 and Rozella 43. Lehninger AL. 1975. Biochemistry. 2nd edi- et al. 2004. Assembling the fungal tree of life: form the deepest branching clade of fungi. tion. New York: Worth Publishers. Progress, classification, and evolution of sub- Protist 161: 116–21. 44. Ruiz-Herrera J, Gonza´ lez-Prieto JM, Ruiz- cellular traits. Am J Bot 91: 1446–80. 23. Held AA. 1981. Rozella and Rozellopsis: Medrano R. 2002. Evolution and phyloge- 6. Jumpponen A, Jones KL. 2009. Massively Naked endoparasitic fungi which dress up netic relationships of chitin synthases from parallel 454 sequencing indicates hyperdiverse as their hosts. Bot Rev 47: 451–15. yeasts and fungi. FEMS Yeast Res 1: 247–56. fungal communities in temperate Quercus mac- 24. Barr DJS. 1980. An outline for the reclassifi- 45. Munro CA, Gow NAR. 2001. Chitin synthesis rocarpa phyllosphere. New Phytol 184:438–48. cation of the , and for a new order, in human pathogenic fungi. Med Mycol 39: 7. O’Brien HE, Parrent JL, Jackson JA, the Spizellomycetales. Can J Bot/Rev Can Bot 41–53. Moncalvo JM, et al. 2005. Fungal community 58: 2380–94. 46. Bulawa CE. 1993. Genetics and molecular- analysis by large-scale sequencing of 25. James TY, Kauff F, Schoch C, Matheny PB, biology of chitin synthesis in fungi. Annu Rev environmental samples. Appl Environ et al. 2006. Reconstructing the early evolution Microbiol 47: 505–34. Microbiol 71: 5544–50. of the fungi using a six gene phylogeny. 47. Roncero C. 2002. The genetic complexity of 8. Porter TM, Skillman JE, Moncalvo JM. Nature 443: 818–22. chitin synthesis in fungi. Curr Genet 41: 367– 2008. Fruiting body and soil rDNA sampling 26. Sparrow FK, Paterson RA, Johns RM. 1965. 78. detects complementary assemblage of Additions to the phycomycete flora of the 48. Ruiz-Herrera J, Ortiz-Castellanos L. 2010. Agaricomycotina (Basidiomycota, Fungi) in a Douglas Lake region. V. New or interesting Analysis of the phylogenetic relationships and hemlock-dominated forest plot in southern fungi. Pap Mich Acad Sci Arts Lett 50: 115– evolution of the cell walls from yeasts and Ontario. Mol Ecol 17: 3037–50. 23. fungi. FEMS Yeast Res 10: 225–43.

Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc. 101 T. Y. James and M. L. Berbee Insights & Perspectives .....

49. Tsuizaki M, Takeshita N, Ohta A, Horiuchi gen. nov., sp nov., and Abeoforma whisleri, 71. Lefranc M, Thenot A, Lepere U, Debroas D. H. 2009. Myosin motor-like domain of the gen. nov., sp nov.) from invertebrate digestive 2005. Genetic diversity of small eukaryotes in class VI chitin synthase Csmb is essential to tracts represent an undescribed clade within the lakes differing by their trophic status. Appl its functions in Aspergillus nidulans. Biosci unicellular opisthokont lineage Ichthyosporea Environ Microbiol 71: 5935–42. Biotechnol Biochem 73: 1163–7. (Mesomycetozoea). Protist 162: 33–57. 72. Lepere C, Masquelier S, Mangot JF, 50. Abramczyk D, Park C, Szaniszlo PJ. 2009. 61. Itoh T, Martin W, Nei M. 2002. Acceleration Debroas D, et al. 2010. Vertical structure of Cytolocalization of the class V chitin synthase of genomic evolution caused by enhanced small eukaryotes in three lakes that differ by in the yeast, hyphal and sclerotic morpho- mutation rate in endocellular symbionts. their trophic status: A quantitative approach. types of Wangiella (Exophiala) dermatitidis. Proc Natl Acad Sci USA 99: 12944–8. Isme J 4: 1509–19. Fungal Genet Biol 46: 28–41. 62. Amend AS, Seifert KA, Samson R, Bruns 73. Lacey J. 1996. Spore dispersal – its role in 51. Treitschke S, Doehlemann G, Schuster M, TD. 2010. Indoor fungal composition is geo- ecology and disease: The British contribution Steinberg G. 2010. The myosin motor domain graphically patterned and more diverse in to fungal aerobiology. Mycol Res 100: 641–60. of fungal chitin synthase v is dispensable for temperate zones than in the tropics. Proc 74. Mangot JF, Lepere C, Bouvier C, Debroas vesicle motility but required for virulence of Natl Acad Sci USA 107: 13748–53. D, et al. 2009. Community structure and Think again the maize pathogen Ustilago maydis. Plant 63. Fierer N, Liu ZZ, Rodriguez-Hernandez M, dynamics of small eukaryotes targeted by Cell 22: 2476–94. Knight R, et al. 2008. Short-term temporal new oligonucleotide probes: New insight into 52. Lenardon MD, Munro CA, Gow NAR. 2010. variability in airborne bacterial and fungal the lacustrine microbial food web. Appl Chitin synthesis and fungal pathogenesis. populations. Appl Environ Microbiol 74: Environ Microbiol 75: 6373–81. Curr Opin Microbiol 13: 416–23. 200–7. 75. Sekimoto S, Rochon DA, Long JE, Dee JM, 53. Leal-Morales CA, Gay L, Fevre M, 64. Frohlich-Nowoisky J, Pickersgill DA, et al. 2011. A multigene phylogeny of Bartnicki-Garcia S. 1997. The properties Despres VR, Poschl U. 2009. High diversity Olpidium and its implications for early fungal and localization of Saprolegnia monoica chitin of fungi in air particulate matter. Proc Natl evolution. BMC Evol Biol, in press. synthase differ from those of other fungi. Acad Sci USA 106: 12814–9. 76. Lopez P, Casane D, Philippe H. 2002. Microbiology 143: 2473–83. 65. Jones EBG, Sakayaroj J, Suetrong S, Heterotachy, an important process of protein 54. Guerriero G, Avino M, Zhou Q, Fugelstad J, Somrithipol S, et al. 2009. Classification of evolution. Mol Biol Evol 19: 1–7. et al. 2010. Chitin synthases from Saprolegnia marine Ascomycota, anamorphic taxa and 77. Stoeck T, Behnke A, Christen R, Amaral- are involved in tip growth and represent a Basidiomycota. Fungal Diversity 35: 1–187. Zettler L, et al. 2009. Massively parallel tag potential target for anti-oomycete drugs. 66. Nagahama T, Takahashi E, Nagano Y, sequencing reveals the complexity of PLoS Pathog 6: e1001070. Abdel-Wahab MA, et al. 2011. Molecular evi- anaerobic marine protistan communities. 55. Bland CE, Couch JN. 1985. Structure and dence that deep-branching fungi are major BMC Biol 7: 72. development. In Couch JN, Bl CE, eds; The fungal components in deep-sea methane 78. Kim E, Harrison JW, Sudek S, Jones MDM, Genus Coelomomyces. Orlando: Academic cold-seep sediments. Environ Microbiol 13: et al. 2011. Newly identified and diverse Press. p 23–80. 2359–70. plastid-bearing branch on the eukaryotic tree 56. Humber RA. 1989. Synopsis of a revised 67. Mohamed DJ, Martiny JBH. 2011. Patterns of life. Proc Natl Acad Sci USA 108: 1496– classification for the Entomophthorales of fungal diversity and composition along a 500. (Zygomycotina). Mycotaxon 34: 441–60. salinity gradient. Isme J 5: 379–88. 79. Marchler-Bauer A, Lu SN, Anderson JB, 57. Keeling PJ, Fast NM. 2002. Microsporidia: 68. Bass D, Howe A, Brown N, Barton H, et al. Chitsaz F, et al. 2011. CDD: A Conserved Biology and evolution of highly reduced intra- 2007. Yeast forms dominate fungal diversity in Domain Database for the functional cellular parasites. Annu Rev Microbiol 56: 93– the deep oceans. Proc R Soc Lond Ser B: Biol annotation of proteins. Nucleic Acids Res 116. Sci 274: 3069–77. 39: D225–9. 58. Pendland JC, Hung SY, Boucias DG. 1993. 69. Le Calvez T, Burgaud G, Mahe S, Barbier G, 80. Choquer M, Boccara M, Gonc¸calves IR, Evasion of host-defense by in vivo-produced et al. 2009. Fungal diversity in deep-sea Soulie´ M-C, et al. 2004. Survey of the protoplast-like cells of the insect mycopath- hydrothermal ecosystems. Appl Environ Botrytis cinerea chitin synthase multigenic ogen Beauveria bassiana. J Bacteriol 175: Microbiol 75: 6415–21. family through the analysis of six euascomy- 5962–9. 70. Lefe` vre E, Roussel B, Amblard C, Sime- cetes genomes. Eur J Biochem 271: 2153–64. 59. Whisler HC. 1962. Culture and nutrition of Ngando T. 2008. The molecular diversity of 81. Edgar RC. 2004. MUSCLE: Multiple Amoebidium parasiticum. Am J Bot 49: 193–9. freshwater picoeukaryotes reveals high sequence alignment with high accuracy 60. Marshall WL, Berbee ML. 2011. Facing occurrence of putative parasitoids in the and high throughput. Nucleic Acids Res 32: unknowns: Living cultures (Pirum gemmata plankton. PLoS One 3: e2324. 1792–7.

102 Bioessays 34: 94–102,ß 2011 WILEY Periodicals, Inc.