open 3 access freely available online © plosi o

OPEN 3 ACCESS Freely available online © PLOSI o -

Picomonas judraskeda Gen. Et Sp. Nov.: The First Identified Member of the Picozoa Phylum Nov., a Widespread Group of Picoeukaryotes, Formerly Known as 'Picobiliphytes'Ramkumar Seen ivasan1*, Nicole Sausen1, Linda K. Med I in2, Michael M elkonian11 D epartm en t o f Botany, C ologne Biocenter, University o f Cologne, Cologne, G erm any, 2 M arine Biological Association o f th e UK, The Laboratory, The Citadel, Plym outh, United Kingdom

AbstractIn 2007, a novel, putatlvely photosynthetic plcoeukaryotlc lineage, the 'plcoblllphytes', w ith no known dose eukaryotic relatives, was reported from 18S environmental clone library sequences and fluorescence In situ hybridization. Although single cell genomics later showed these organisms to be heterotrophlc rather than photosynthetic, until now this apparently widespread group o f plco-(or nano-)eukaryotes has remained uncultured and the organisms could not be formally recognized. Here, we describe Picomonas judraskeda gen. et sp. nov., from marine coastal surface waters, which has a 'p lcoblllphyte ' 18S rDNA signature. Using vital m itochondrial staining and cell sorting by flow cytometry, a single cell- derived culture was established. The cells are blflagellate, 2 .5-3.8x2-2.5 pm In size, lack plastlds and display a novel stereotypic cycle o f cell m otility (described as the "jum p, drag, and skedaddle"-cycle). They consist o f tw o hemispherical parts separated by a deep cleft, an anterior part that contains all major cell organelles Including the flagellar apparatus, and a posterior part housing vacuoles/vesicles and the feeding apparatus, both parts separated by a large vacuolar cisterna. From serial section analyses o f cells, fixed at putative stages o f the feeding cycle, It Is concluded that cells are not bacterlvorous, but feed on small marine colloids o f less than 150 nm diameter by fluld-phase, bulk flow endocytosls. Based on the novel features o f cell motility, ultrastructure and feeding, and the ir Isolated phylogenetic position, we establish a new phylum, Picozoa, for Picomonas judraskeda, representing an apparently widespread and ecologically Im portant group o f heterotrophlc picoeukaryotes, formerly known as 'plcoblllphytes'.

C ita tio n : Seenivasan R, Sausen N, Medlin LK, M elkonian M (2013) Picomonas judraskeda G en. Et Sp. Nov.: T he First Identified M em ber o f th e Picozoa Phylum Nov., a W idespread G roup o f P icoeukaryotes, Formerly Known as 'Picobiliphytes'. PLoS ONE 8(3): e59565. doi:10.1371/journal.pone.0059565

E d ito r: Ross Frederick Waller, University o f M elbourne, Australia

R eceived July 8, 2012; A c cep ted February 19, 2013; P u b lish ed M arch 26, 2013

C o p y rig h t: © 2013 Seenivasan e t al. This is an open-access article d istribu ted u n d er th e te rm s o f th e Creative C om m ons A ttribution License, w hich perm its unrestricted use, distribution , and reproduction in any m edium , provided th e original au th o r and source are credited .

F u n d in g : This p ro jec t w as su p p o rted by G erm an Research F oundation (DFG) g ran t no. ME 658/27-1. The funders had no role in s tudy design , da ta collection and analysis, decision to publish, o r p repara tion o f th e m anuscrip t.

C o m p e tin g In te rests : The au tho rs have declared th a t no com peting in terests exist.

* E-mail: rseeniva@ uni-koeln.de

Introduction

M icrobial p lankton plays a pivotal role in global biogeochem ical processes and provides am enities an d services that are essential to m ankind’s existence, including food production , rem ediation of waste and regulation of aspects o f the clim ate in the biosphere [1— 4], Picoplankton, organism s that can pass through filters o f 2 pm [5] o r 3 pm [6], are w idespread and dom inate the biom ass in the aquatic environm ent [7-9]. A lthough initially focusing on phototrophic picoprokaryotes, the use o f m olecular techniques to directly analyze gene sequences in natural samples was also applied to very small eukaryotes in three seminal studies in 2001 that revealed an unexpectedly large and novel diversity o f picoeukaryotes [10-12]. Picoeukaryotes are now know n to be ubiquitous in surface waters o f all oceans and are likely the most ab u n d an t eukaryotes in the sea [6,13-15]. W hen defined as cells < 3 pm , picoeukaryotes consist o f the num erically m ore abundan t phototrophic and the less ab u n d an t hetero trophic picoeukaryotes (especially in oligotrophic coastal sites; [16]). T h e latter are generally regarded as bacterial grazers [13] a lthough some are

herbivores feeding on picoplanktonic cyanobacteria o r photosynthetic picoeukaryotes [17], and others are parasites [18],

A large fraction o f the in-situ diversity o f picoeukaryotes has not been studied in culture and these organism s are often regarded as ‘uncu ltu rab le’ [15]. Thus, various m ethods to address the function o f uncultured picoeukaryotes have been developed in recent years. O ne approach has been the application o f specific oligonucleotide probes for fluorescence in-situ hybridization (FISH) often coupled with flow cy t°m etry [19-22]. Furtherm ore, direct sequencing of com m unity D N A or R N A (m etagenomics, m etatranscriptom ics) reveals gene repertoires and m etabolic functions [23], and targeted m etagenom ics has recently been applied to uncultured p icoeukaryotes [24,25]. A nother approach has been to stim ulate the growth o f heterotrophic picoeukaryotes by incubation in unam ended seawater in the dark resulting in the identification o f previously unknow n organism s [26-28], Finally, sorting o f single picoeukaryotic cells [29,30] has recently opened possibilities for singe cell genomics (SCG) by whole genom e am plification and next- generation sequencing o f sorted cells thus coupling m olecular identification to m etabolic functions [31-33].

PLOS ONE I www.plosone.org 1 March 2013 | Volume 8 | Issue 3 | e59565

mailto:[email protected]

http://www.plosone.org

Picozoa, a New Phylum of Picoeukaryotes

M ost o f these m ethods have also been applied to a widely distributed group of uncultured picoeukaryotes th a t represent a deep evolutionary b ran ch w ithout clear affinities to o ther eukaryotes. Initially described from cold and po lar waters using 18S rD N A clone libraries an d FISH , as a group o f picoeukaryotic phycobilin-containing algae with affinities to cryptophytes and katablepharids and nam ed ‘picobiliphytes’ [34,35], these organisms were subsequendy also found in subtropical an d tropical surface waters using the same m ethods [36]. T h e latter authors, however, reported th a t ‘picobiliphytes’ were o f nanoplanktonic ra th e r than picoplanktonic size, an d thus renam ed them ‘bilip hytes’; they also concluded th a t in tropical eddy-influenced surface waters ‘biliphytes’ contributed abou t 30% o f the phyto- planktonic biomass. ‘(Pico)biliphyte’ sequences were also detected in surface waters o f o ther oceans, such as the South-East Pacific O cean [37], the South C hina Sea [38], the Ind ian O cean [39], and the brackish Baltic Sea [28]. A nother study using the same F IS H probes originally applied to identify ‘(pico)biliphytes’ on filtered samples from the N orth Pacific faded to co-localize phycoerythrin with the hybridized cells an d concluded th a t these picoeukaryotes w ere likely no t obligate photoau to trophs b u t ra ther m ixotrophs or phagotrophs and the presence o f orange (phycoerythrin) fluorescence in such cells could represent ingested prey (e.g. picocyanobacteria, [40]). This uncertain ty o f w hether ‘(pico)biliphytes’ w ere auto trophic o r hetero trophic was dispelled by a study th a t used single cell genomics on three ‘(pico)biliphyte’ cells (from the three clades identified by N ot et al. 2007, [34]) isolated by FACS (fluorescence-activated cell sorting) w ith a lysosomal m arker [41], T h e authors did no t find evidence o f plastid D N A or o f nuclear-encoded p lastid-targeted proteins in the partially sequenced genom es o f the three cells and concluded that ‘(pico)biliphytes’ were likely heterotrophic. Furtherm ore, they found b o th viral and bacterial D N A associated with the cells, w hich led them to conclude th a t ‘(pico)biliphytes’ m ay feed on bacteria and large D N A viruses, a lthough viral infections as well as passive a ttachm en t o f bacteria and viruses to the cells could no t be excluded. A lthough all these studies provided im portan t inform ation abou t this ‘uncu ltu red ’ group o f hetero trophic pico- or nanoeukaryotes, the cells themselves have never been seen alive nor studied in detail by light and electron microscopy.

H ere we used a fluorescent m itochondrial m arker to isolate single ‘(pico)biliphyte’ cells by FACS and established a first culture o f these organisms. A detailed light and electron m icroscopic study based on this culture allowed us to describe a new genus and species [Picomonas judraskeda gen. nov., sp. nov.), an d to formally erect a new phylum (Picozoa phylum nov.) for these enigm atic picoeukaryotes. T he cells displayed m any behavioral and structural features that, to o u r knowledge, have no t yet been observed in any o ther eukaryotes and presum ably relate to the cells’ specific feeding strategy. W e verify th a t cells are hetero trophic and lack plastids, and conclude th a t the species studied feeds on small-size m arine colloids (< 150 nm) using a feeding strategy th a t we describe as a fluid-phase, bulk flow m echanism .

Materials and M ethods

Culturing and Fluorescence-activated Cell SortingA bout 600 ml o f surface seawater (5 m depth) was collected

from the N o rth Sea (Helgoland Roads, 54°11 'N , 7°54 'E , Germ any) in August 2008 and successively filtered through 10 pm and 2 pm Isopore m em brane filters (Millipore). F rom the filtrate, 50 ml was centrifuged a t 4000 g in Falcon centrifuge tubes for 10 m in a t 15°C (Multifuge 3SR+, Heraeus). Cells from the supernatan t were collected on a 0.2 pm m em brane filter and

directly transferred into a culture flask contain ing 0.1 pm Filter- Sterilized Sea W ater (FSSW), the inconspicuous pellet was resuspended with FSSW . Both samples were incubated a t 15°C in a 14 /1 0 hour L /D cycle w ith 9 pm ol*m 2*s 1 p h o ton fluence rate. Cells were inoculated a t regular intervals (every two weeks) into fresh FSSW . Presence o f ‘(pico)biliphytes’ was tested by am plification o f nuclear-encoded small subunit rD N A (18S rDNA) by Polym erase C hain R eaction (PCR) w ith ‘(pico)biliphyte’- specific prim ers (PIC O B I01F and P01IT S1R , T able SI). In addition, the universal 18S eukaryote-specific reverse rD N A prim er (1055R, T able SI) was used along with P IC O B I01F for secondary am plification if required, and am plicons were confirm ed by D N A sequencing. Cycle sequencing was perform ed with BigDye® T erm in ato r V I .3 (Applied Biosystems) a t the Cologne C entre for Genom ics (University o f Cologne).

A ‘(pico)biliphyte’-positive enrichm ent culture was p repared for fluorescence-activated cell sorting by add ing 20 nM M itoTracker® G reen FM (M7514, Invitrogen) to 10 ml o f sample an d incubated a t 15°C in the dark for 15-20 m in. Cells were sorted w ith a FA C SvantageSE (Becton Dickinson, NJ) using an A rgon laser at 488 nm . Cells with high green fluorescence and forw ard scatter were targeted and sorted directly into P C R tubes for further confirm ation of ‘(pico)biliphytes’ by P C R am plification and DN A sequencing. Single cell sorting was perform ed for the targeted region (Fig. 1; positive to ‘(pico)biliphytes’) into 24-well plates and plates were incubated a t 15°C as described above. Single cell- derived cultures w ere rechecked for the presence o f ‘(pico)biliphytes’ by P C R with universal eukaryotic rD N A prim ers (IF, BR, T able SI) and confirm ed by sequencing. Cells were regularly (usually every 10-14 days) transferred into 10 ml o f fresh FSSW in tissue culture flasks (25 or 50 ml) with 20 pi o f soil extract [42] and m onitored by light m icroscopy using phase contrast optics.

Fluorescence MicroscopyFor fluorescence m icroscopy (FL), cells (from a 10 day old

culture) were fixed with 1.25% (v/v, final concentration) glutaraldehyde (GA) in culture m edium for 30 m in on ice. Fixed cells were directly p laced on a glass slide p re-coated w ith poly L- lysine (P8920, Sigma) and allowed to settle for 30 m in a t room tem perature. For FL, cells w ere p robed with MitoTrackei® R ed C M X R os (Invitrogen, D arm stadt, G erm any) for 10 m in. Cells were w ashed with saline e thanol (22 ml o f 100% ethanol, 5 ml o f deionized w ater, 3 ml o f 25X SE T [3.75 M N aC l, 25 m M E D TA , 0.5 M T ris -H C l and p H 7.8]). D A PI (4', 6-diam idino-2- phenylindole; 2 p g /m l final eone.) was used for nuclear staining.

Electron MicroscopyFor transm ission electron microscopy, a cell suspension (10 ml)

was fixed with a m ixture o f 0.32% freshly p rep ared p a ra form aldehyde (PFA), 0.125% glutaraldehyde (GA) and 0.01% of osm ium tetroxide (v/v, final concentration), and incubated on ice for 30 m in. T h e fixed cells w ere subsequently transferred to 1.5 ml centrifuge tubes (pre-coated with dichlorodim ethylsilane (40410, Fluka, Buchs, Switzerland)) and pelleted a t 4000 g for 10 m in. at 15°C (this step was repeated until a visible pellet was seen). T he pellet was washed an d centrifuged twice with 500 pi o f FSSW . Bovine serum album in (50% BSA, A6063-Sigma) was added to coat the pellet. T h e supernatan t was rem oved and 50 pi o f 1.25% GA (v/v) in FSSW was overlaid to fix the pellet w ith BSA, and incubated on ice for 30 m in. T h e translucent pellet was carefully transferred to a new 1.5 ml centrifuge tube for dehydration. Before dehydration, the salinity o f the pellet was slowly reduced by 50% and 25% o f FSSW w ith 12.5% and 22.5% o f ethanol in w ater (v /v final eone.) respectively. D ehydration, em bedding and




Figure 1. Identification and isolation of '(pico)biliphytes' by fluorescence-activated cell sorting {MitoTracker’ Green FM [FL1] vs. Forward Scatter [FSC]) and PCR am plification using specific primers. 1A. Approximately 100 cells from each of three regions with higher fluorescence/forw ard scatter (R1-R3) w ere sorted into PCR tubes. Cells marked in 'red ' in the cytogram corresponded to '(pico)biliphytes' by PCR amplification and sequencing. 1B. Sorted cells used for primary amplification with 'picobiliphyte'-specific primers (PICOBI01F/P01ITS1R) and reamplified with PICOBI01F and 1055rev. Amplicons in region R1 w ere confirm ed by DNA sequencing to correspond to '(pico)biliphytes'. Arrow 650 base pairs; M 1 Kb ladder. doi:10.1371/journal.pone.0059565.g001

serial sectioning were perform ed as described in G eim er and M elkonian, 2004 [43]. A total o f 52 cells were serially sectioned (m inim um num ber o f 20 sections pe r cell o f 60 nm thickness each), o f w hich 8 cells were completely sectioned. E lectron m icrographs were taken w ith a Philips C M 10 transm ission electron m icroscope using a G atan ORIETS T E M C C D cam era an d Digital M icrograph software. Further analysis o f electron m icrographs was carried out in Photoshop (Adobe CS4, version 11.0.2). For 3 D -m odeling o f P. judraskeda cells, images from serial sections were initially aligned w ith A dobe Photoshop (CS4), an d later im ported into IM O D 4.1.10 [44] software. T h e final 3 D -graphical anim ation o f P. judraskeda cell (Movie S2, M ovie S3) was created by B lender v.2.63 m odeling software (h ttp ://w w w .b len d e r.o rg /).

For scanning electron m icroscopy; cells were fixed with 1 % PFA and 1.25% GA (v/v) for 30 m in in culture m edium . Fixed cells were directly p laced on a glass slide, coated w ith poly L-lysine and allowed to settle for 30 m in. Cells were gradually washed to decrease the salinity with 100%, 50% and 25% (v/v) FSSW followed by dehydration as in M artin -C ereceda et al. (2009) except that the last critical po in t drying step was with liquid C 0 2 [45]. Cells were coated with gold layer o f 40 11111 thickness. T he SEM images w ere taken in Q u a n ta ™ 250FEG im ager (FEI G m bH , Frankfurt, Germ any).

Phylogenetic AnalysesT h e P. judraskeda full length rD N A operon was obtained by

prim er walking. Am plification was carried out with the com bination o f eukaryotic universal ribosom al prim ers and Picomonas- specific prim ers (IF -P01IT S1R ; PIC O B I01F-L 52R ; P01ITS1F- 28S_1433rev; P01ITS2F-28S_1433rev; 28S_PicoD15F-28S_2933rev; 28S_PicoD21F-28S_2933rev; 28S_PicoG 4For-28S 3356rev, T able SI). O verlapping prim ers were designed if necessary for the am plification. T h e detailed list o f prim ers used in this study is given in T able S I. T he com plete nuclear ribosom al operon o f Picomonas judraskeda has been deposited at N C B I G enbank with ace no JX 988758 . Furtherm ore, we constructed ‘(pico)biliphyte’-specific clone libraries as given in M edlin et al. 2006 [46], using taxon-specific prim ers (PIC O B I01F an d L52R, T able SI) to study the diversity o f ‘(pico)biliphytes’ in the same habitat. In addition, one environm ental sequence was obtained for

clade ‘B P2’ from same hab ita t by P C R using P IC O B I02F and L 52R prim ers (Table SI).

Overall 191 environm ental picozoan 18S rD N A sequences retrieved from the G enbank database (h ttp ://w w w .n cb i.n ln i.n ih . g o v /g en b an k /; Accessed 2012 July), nine new environm ental sequences (ace no. JX 9 8 8 7 5 9-JX 988767), an d the sequence (ace noJX 988758) obtained from the cultivated strain P. judraskeda were subjected to phylogenetic analyses. Prior to the analyses, the m anual alignm ent process was guided by the conserved secondary structure o f the 18S rR N A obtained from o ther known eukaryotes[47] using SeaView 4.3.2 as an alignm ent editor. Sequence regions that h ad uncertain alignm ent positions w ere post-aligned by M fold[48], Because o f some shorter environm ental sequences (lacking the 5 '-p a rt o f the gene), only 1253 (from a total o f 1798 bp o f the 18S rD N A from our strain) unam biguously aligned characters could be used for the phylogenetic analyses (the alignm ent is available upon request). T ree topology (m axim um likelihood, ML) was calculated by using R A xM L PT H R E A D S version 7.2.6 [49], T o determ ine the best tree topology, 20 M L trees w ere com puted starting from 20 random ized m axim um parsim ony starting trees. T h e G T R + I+ T evolutionary m odel was used and 1000 m axim um likelihood bootstrap replicates were calculated by R A xM L. In addition, PAETP 4.0 B10 [50] was applied to calculate 1000 N eighbour Jo in in g (NJ) bootstrap replicates. For NJ, the G T R + I+ T evolutionary m odel was selected by M odeltest v3.7 [51]. Bayesian interferences w ere calculated from M rBayes with the covarion option by using two runs w ith four M arkov chain M onte Carlo (M CM C) chains over 3,000,000 generations, and the first 1,000,000 generations were discarded as “b u rn -in ” [52,53]. B ootstrap and Bayesian posterior probabilities values less than 60% an d 0.80 respectively were considered as “ negligible support.” T w o taxon samples (M S584-5, M S584-11) were reassem bled by contig assembly obtained from the Single-cell Am plified G enom e data o f Yoon et al. (2011). Accession num bers for all sequences are depicted on the tree (Fig. S2).

Nomenclatural ActsT h e electronic version o f this docum ent does not represent a

published work according to the In ternational C ode of Zoological N om enclature (ICZN), and hence the nom enclatural acts


http://www.blender.org/

http://www.ncbi.nlni.nih



contained in the electronic version are no t available under that Code from the electronic edition. Therefore, a separate edition of this docum ent was p roduced by a m ethod th a t assures num erous identical and durable copies, an d those copies were simultaneously obtainable (from the publication date noted on the first page o f this article) for the purpose o f providing a public an d p e rm anen t scientific record, in accordance with Article 8.1 o f the C ode. T he separate print-only edition is available on request from PLoS by sending a request to PLoS O N E , 1160 B attery Street Suite 100, San Francisco, CA 94111, U SA along with a cheque for SIO (to cover p rin ting an d postage) payable to “Public L ibrary of Science.”

In addition, this published work and the nom enclatural acts it contains have been registered in ZooB ank (h ttp :/ /z o o b a n k .o rg /), the p roposed online registration system for the IC Z N . T he ZooB ank LSIDs (Life Science Identifiers) can be resolved and the associated inform ation viewed th rough any standard web browser by appending the LSID to the prefix “ h t tp : / /zo o b a n k .o rg /.” T he LSID for this publication is: urn:lsid:zoobank.org:pub:009138ED - 3F29-4351-894D -A D 2E9C A C 28D 0.

Results

Isolation and Cultivation o f Picomonas judraskeda sp. novSeaw ater samples collected betw een Ju ly an d O ctober in 2007,

and August 2008 from H elgoland R oads (Germany) and subsequendy filtered th rough 10 pm and 2 pm m em brane filters, contained D N A from ‘(pico)biliphytes’ as dem onstra ted by P C R am plicons using prim ers specific for ‘(pico)biliphytes’ (results not shown). ‘(Pico)biliphyte’- positive samples were regularly transferred into 0.1 pm filter-sterilized seawater (FSSW) and the presence o f ‘(pico)biliphytes’ m onito red by P C R and DN A sequencing. These ‘enrichm ents’ were used for flow cytom etry and sorting. W e em ployed a m itochondrial m arker (M itoTracker® G reen FM) and obtained a single cell-derived culture verified by rD N A sequencing using ‘(pico)biliphyte’-specific an d universal eukaryotic ribosom al D N A (rDNA) probes (Fig. 1). Investigation of the culture by light and electron m icroscopy verified th a t it contained only a single species o f eukaryote, here described as Picomonas judraskeda gen. nov., sp. nov. for w hich we erect a new phylum (Picozoa phylum nov.). In addition, the culture contained several types o f m orphologically distinguishable bacteria . In general, we found th a t FSSW , w ith very low n u trien t concentrations an d the presence o f bacteria (an axenic culture o f P. judraskeda could no t be established) were essential for m aintain ing successful growth, w hich was nevertheless slow and sometimes unpred ictable, depending on the source o f seawater (seawater from H elgoland, the C anary Islands and South Africa was tested), the season during w hich the seawater was collected (seawater from H elgoland R oads collected betw een O ctober and Ju n e generally did no t support grow th o f the organism), as well as on recovery from a lag phase o f several days after serial transfer. T he m axim um cell density o f P. judraskeda observed in culture was 3 0 - 40 cells/m l (cells were identified by phase contrast m icroscopy based on their peculiar swim m ing behavior while focusing through the whole w ater colum n o f the tissue flask). T h e organism was m aintained for 3 years in culture until it died in Ju n e 2012 after unsuccessful serial transfers.

Light Microscopy and Cell MovementT h e cells o f P. judraskeda sp. nov. are m ostly stationary, floating

in the w ater colum n. T h e oblong cells vary in length betw een 2.5—3.8 pm , their w idth being 2-2 .5 pm (n = 20). Larger cells (lengths > 4 pm an d widths > 3 pm) w ere only observed by LM in

chem ically fixed preparations (Fig. 2A,B). T h e cells are biflagellate with a long (12-14 pm) an d a short (7-9 pm) flagellum inserted laterally n ear the m iddle o f one o f the long sides o f the cell (Fig. 2A- C). T his cell side is term ed ventral, the opposite dorsal. T h e long flagellum is arb itrarily term ed the an terio r flagellum, the short flagellum the posterior flagellum. E ach cell consists o f two nearly hem i-spherical parts separated by a deep cleft (Cl, Fig. 2C). T he an terio r p a rt (AP) contains the nucleus and the single m itochondrion (Fig. 2B), w hereas the posterior p a r t (PP) varies in size (presum ably depending on the trophic status o f the cell o r the stage o f the feeding cycle, see below), w hich accounts for the observed relatively large variations in cell length.

T h e cells o f P. judraskeda exhibit a unique m ode of motility : After extended periods o f rest, a stereotypic pa tte rn is initiated consisting o f a rap id short-distance ju m p o f the cell into the an terio r direction (approxim ately 3 -5 pm), im m ediately followed by a slower dragging cell m ovem ent in the opposite (posterior) direction. T h e drag m ovem ent lasted for abou t 2 s, during w hich the cells travelled a distance o f approxim ately four times their cell length. T his ju m p /d ra g - cycle was observed from two up to five times consecutively after w hich the cells suddenly ‘shot’ away (a behavior term ed ‘skedaddle’ here, a colloquial term m eaning to m ove away rapidly). D uring skedaddling, the cell travelled a distance o f m inim um 50 pm before it becam e im m otile again and until the next ju m p /d ra g cycle was observed (Movie SI). D uring resting periods o f the cell, the posterior flagellum (PFj was held closely adjacen t to the ventral cell surface curving a round the posterior end o f the cell, often visible under the m icroscope and mostly imm otile, while the an terio r flagellum extended from the cell surface revealing an undulatory wave patte rn . T h e behavior o f the flagella during the ju m p /d ra g /sk ed a d d le cycle was no t studied in detail b u t apparently involved flagellar re-orientation and rapid undulatory m ovem ents o f one or bo th flagella. O ccasionally late stages o f cell division have been observed; the daughter cells violently separate from each o ther using the skedaddle type m ovem ent for cell separation (not shown).

Electron MicroscopyT h e ultrastructure o f P. judraskeda has revealed several highly

unusual features th a t collectively, to the best o f our knowledge, have no t been described before in any o ther eukaryotic cell. T he cell is distinctly separated into two parts. T h e an terio r p a r t (AP) contains alm ost all cell constituents o f a typical eukaryote cell: T he large hem ispherical nucleus occupies nearly h a lf the volum e o f the an terio r cell p a r t an d in its spherical part, the nuclear envelope is appressed to the plasm a m em brane (Fig. 2D,E; Fig. SI; M ovie S2). T h e interphase stage o f the nucleus appears to have a considerable am oun t o f condensed chrom atin , w hich is uncom m on in unicellular organisms.

T h e single m itochondrion with tubu lar cristae (Fig. 3A,C) is located posterior to the nucleus in the ventral region o f the an terio r p a rt (Fig. 2D,E; M ovie S2). Its shape resembles th a t o f a sofa/couch th a t extends from near the left to the right surface o f the cell (length ~ 1 .6 pm , Fig. 3A), w ith its high back (height — 1 pm) oriented tow ards the cell’s center, the end touching the nuclear envelope (Fig. 2D). T h e seating area narrow s towards the ventral region (Fig. 2D) and term inates n ear the ventral surface at the level o f the basal body o f the posterior flagellum (Fig. 2D,E). T h e m itochondrion displays two additional highly distinctive structural differentiations:

(1) T h ere are two m em brane-bound inclusions w ith electron- dense, g ranular contents located a t specific positions inside the m itochondrion (edsm l, edsm2, Fig. 3). T h e first (edsm l) is


http://zoobank.org/

http://zoobank.org/.%e2%80%9d




LUri I



Figure 2. A Picom onas cell. 2A. Differential interference contrast of a chemically fixed cell. Inset show s phase contrast im age of a live cell from tissue culture flask p ho tog raphed with an inverted m icroscope (Scale bar 5 ¡im). 2B. Fluorescence and phase contrast overlay, nucleus (blue), m itochondrion (red). 2C. SEM image. 2D. A longitudinal section th rough a cell in th e plane of th e flagella, viewed from th e cell's left. 2E. A 3 D serial section reconstruction of th e cell dep icted in 2D. AF/PF (an te rio r- /p o s te rio r flagellum); AP/PP (anterior/posterior part of th e cell); G (Golgi body); M (mitochondrion); MB ('microbody'); N (nucleus); t rn,tr2 (distal [tr2] and proximal [tri] flagellar transitional regions); P (posterior digestive body); Cl (cleft separating th e anterior from th e posterior part of th e cell); vc (vacuolar cisterna). doi:10.1371/journal.pone.0059565.g002

edsm 2 edsm 2

edsm 2

ed sm l

edsm 2

Figure 3. Ultrastructure of the mitochondrion of a Picom onas cell. Single m itochondrion with tubular cristae, m em brane bound (single and double) electron dense material (edsm l and 2 respectively) and regular projections of th e o u ter envelope m em brane. 3A. a double m em brane bound edsm 2 near th e ventral surface; serial sections (1-4) of th e edsm 2 displayed som e tube-like structures in th e lumen (thin lines); also note th a t the edsm 2 is a branched structure. 3B. A m em brane invagination of edsm 2 into th e m itochondrion th a t reveals continuity betw een th e m itochondrial envelope m em branes and th e tw o m em branes encircling th e edsm 2. 3C. Oblique section of a cell from dorsal right to ventral left with edsm l and edsm 2 displayed. The edsm l (on th e left) is positioned betw een th e ou ter (black arrow) and inner (white arrow) mitochondrial envelope m em brane and th e edsm 2 (on th e right) with a double m em brane (higher m agnifications of th e tw o edsm s on th e right. A large num ber of short cylindrical m em brane protrusions term ed 'mitovilli' (arrow heads) extend from th e ou ter mitochondrial m em brane tow ards posterior digestive body (P), they term inate in granular m aterial covering th e posterior digestive body. A vacuolar cisterna (vc) in th e 'cleft' region separa tes th e anterior from posterior part of th e cell and is only absen t (i.e. contains a large hole) in th e area of th e mitovilli (3C). Towards th e right are show n higher m agnifications of tw o serial sections from th e cell dep icted in 3C revealing details of th e mitovilli-posterior digestive body junction. F (feeding apparatus with basket fibers; G (Golgi body); M (mitochondrion); N (nucleus); P (posterior digestive body); vc (vacuolar cisterna). Numbers at th e to p right indicate th e section num ber of a series. Scale bar: 100 nm. doi:10.1371/journal.pone.0059565.g003




located a t the high back end o f the m itochondrion, positioned w ithin the in ter-m em brane space w hich is thereby dilated (Fig. 2D; Fig. 3C, inset). A ppropriate ly sectioned, this inclusion appeared to be located in the m atrix enclosed by a single m em brane (the inner envelope m em brane). Because the edsm l could be followed th rough 7 serial sections (a cell longitudinally sectioned from left to right in Fig. SI), we assum e th a t the shape o f this inclusion is cylindrical (diam eter 150-200 nm , length m inim um 400 nm). T h e second (edsm2) is located in the ventral, left p a rt o f the m itochondrion an d is enclosed by two m em branes (Fig. 3A,C, inset). This inclusion also has contact with the m itochondrial envelope and in one fortuitous section, we were able to dem onstrate the continuity betw een the two m em branes enclosing the edsm 2 and the m itochondrial envelope (Fig. 3B), i.e. the edsm2 is a tubular invagination o f the cytoplasm into the m itochondrial m atrix. Serial sections revealed th a t the edsm 2 is a b ran ch ed structure; its contents sometimes appeared to contain small tubules ra th e r th an granules as in the edsm l (Fig. 3A, sections 1-4).

(2) T h e lower side o f the m itochondrion adjacent to the posterior p a r t o f the cell (the bo ttom of the seating area) displayed ano ther highly unusual specialization. O ver an a rea o f ~ 1 x0 .6 pm (Fig. 3A,C; M ovie S2), regularly-spaced projections extend from the outer m itochondrial envelope m em brane tow ards the posterior p a rt (PP) for 50 -6 0 nm (n = 30) after w hich they term inate in an electron dense g ranu lar a rea that is in contact w ith the m em brane of a large vacuole (the posterio r digestive body, see below). T hese projections, term ed here “ mitovilli,” are abou t 20 nm wide (narrow er near their base) and are spaced a t 40 nm (Fig. 3C, sections 41, 42). F rom this we calculate th a t the specialized surface o f the m itochondrion displays abou t 300 mitovilli (Movie S2).

T h e single Golgi body is located betw een m itochondrion, nucleus and basal body of the an terio r flagellum closely associated (resting on) w ith the seating area o f the m itochondrion (Fig. 2D,E; Fig. 3C; M ovie S2). It consists o f 5 -7 Golgi cisternae (often artificially inflated caused by suboptim al chem ical fixation), the cis-face oriented tow ards the nucleus, the trans-face tow ards the basal body. Tw o sheets o f rough endoplasm ic reticulum extend from the nuclear envelope posteriorly along the left an d right sides o f the an terio r p a r t o f the cell, respectively (Fig. 4A section 13; M ovie S2).

T w o large, spherical m icrobodies (MB; w ith granular contents, bounded by a single m em brane) o f 600-800 nm diam eter are located in the dorsal region o f the an terio r p a rt o f the cell, besides the m itochondrion (Fig. 2D,E; Fig. 4A; Fig. S I, sections 26, 29, 32; M ovie S2).

T h e an terio r p a r t o f the cell is separated from the posterior p a rt by a large, horizontally-oriented, plate-like vacuolar cisterna (vc, Fig. 2D,E; Fig. 3C; Fig. S I; M ovie S2), th a t was sometimes observed to be dilated (Fig. S I, likely an artifact o f chem ical fixation). A t its m argin, the vacuolar cisterna is appressed to the plasm a m em brane (Fig. 3C; Fig. SI), except n ear the ventral cell surface w here space is left for passage o f three m icrotubular flagellar roots into the posterior p a rt o f the cell (see below (Fig. 6 ; Fig. 7)). T h e vacuolar cisterna displays a large hole in the region w here the mitovilli abu t the posterior p a r t o f the cell (Fig. 3C sections 41, 42; Fig. SI sections 9-15).

T h e posterior p a rt o f the cell consists o f num erous vesicles and vacuoles o f different sizes and electron density b u t lacks ribosom es and endoplasm ic reticulum , w hich are confined to the an terio r part. In addition it contains the cytostom e an d feeding basket (see

below). T h e lengths (and volumes) o f the posterior p a rt (PP) were found to vary considerably am ong sectioned cells (from an terio r to posterior: 0 .6 -1 .5 pm , com pare e.g. Fig. 2 with Fig. S I, section 13), w hereas the lengths o f the an terio r p a r t (AP) o f the cells were quite constant (1.9-2 pm , b o th n = 1 0 ) . Tw o regular structures were identified in the posterior p a rt o f the cell:

T h e first structure is a large vacuole (“posterior digestive body,” P) th a t is located in the ventral region o f the posterior p a rt o f the cell adjacent to the vacuolar cisterna an d the mitovilli (Fig. 2D,E; Fig. 3C; M ovie S2). Its shape is basically a short b a r w ith the long axis extending from the left to the right cell side (~ 1 pm; Fig. 4A,B), its dep th (from ventral to dorsal) 600-700 nm and its height 400-500 nm . T h e m em brane surface o f this vacuole is often highly irregular w ith invaginations (Fig. 3C). Som etim es these invaginations contain vesicles (Fig. 4B, section 5). T he contents o f this vacuole consist o f irregular, fluffy m aterial o f m edium electron density and num erous vesicles o f various sizes with electron-translucent contents (the occasional denser vesicles/ tubules presum ably represent the invaginations described above, Fig. 2D; Fig. 3C).

T h e second conspicuous structure in the posterior p a r t (PP) o f the cell is the cytostom e/feeding basket th a t together com prise the feeding apparatus (‘F ’, Fig. 4). T he feeding apparatus is located ventro- posterior to the posterior digestive body (P) and consists o f a basket o f abou t 50 -6 0 parallel runn ing fibers (diam eter o f a single fiber: 15 nm , 30 nm repeat structure from one fiber to the next) o f varying lengths th a t extend perpendicu lar to the an terio posterior cell axis, from the ventral surface w here they are a ttached to the plasm a m em brane for up to 1.2 pm enclosing the cytostom e, towards the dorsal region w here they term inate (Fig. 4B,C,F; M ovie S2; M ovie S3). In addition to these m ajor fibers, there is a fine netw ork of very th in fibers th a t interconnect adjacen t fibers irregularly (not shown; Fig. 4F). T he basket is open towards the an terio r and dorsal regions o f the posterior p a rt (PP), b u t closed tow ards the posterior end o f the cell (Fig. 4A -C ) where it sometimes extends into a short tail-like projection o f the cell (Fig. 4A section 1 and Fig. 4D,E). T ow ards the dorsal end, the feeding basket gradually widens (up to 500 nm at its dorsal end). T h e slit-like cytostom e (‘cy’, see Fig. 6C) is form ed w here the ‘side walls’ o f the basket fibers connect to the plasm a m em brane on the ventral surface (the fibers are anchored in electron-dense ribbons a t the plasm a m em brane; Fig. 4C, sections 13, 15-18; white arrow head) and extends in the anterior-posterior d irection for a length of abou t 1 pm , its w idth being on average 150 nm (n = 10; Fig. 4 B - E; M ovie S2; M ovie S3). In serial sections o f some cells it was observed th a t the anterior-m ost p a r t o f the feeding basket bisected the overlying posterior digestive body (Fig. 4B, section 11). In these cells (non-feeding stage), the feeding basket appeared to contain mostly small vesicles and m ultivesicular bodies (Fig. 4B). In o ther cells, possibly during active feeding, the feeding basket contained a single, large vacuole th a t extended well beyond the basket into the dorsal region o f the posterior p a r t (PP) to reach the dorsal plasm a m em brane (Fig. 4C, sections 15-18; the length of this vacuole could be up to 1.2 pm). T his vacuole contained irregular-shaped, fluffy m aterial o f m edium electron density (very rarely a vesicle was seen w ithin this vacuole; Fig. 4A, section 7). W e term this vacuole the “food vacuole (FV).” T h e food vacuole extended ventrally into the narrow parts o f the basket (Fig. 4C , sections 16-18), w here it was sometimes observed to be associated with smaller vesicles (Fig. 4C, section 17). T h e an terio r end o f the cytostom e (and thus the feeding basket a t the plasm a m em brane) is linked to the posterior ends o f two m icrotubular flagellar roots (P rl, Pr2, see below for description of the flagellar apparatus; Fig. 4B, section 2 and Fig. 4C , sections 1, 3; M ovie S2) th rough the two electron-




pnw.

Figure 4. Feeding apparatus. 4A. Longitudinal sections of a Picomonas cell from ventral to dorsal with th e feeding apparatus in cross section. Sections 2 to section 7 depict a recently form ed food vacuole inside th e 'basket' of th e feeding apparatus. 4B & C. Cross sections th rough tw o cells of Picomonas from anterior to posterior; sections begin in th e central part of th e cell, posterior to the basal bodies. 4B. Cell, in the non-feeding mode, w ithout a food vacuole within th e basket. 4C. Cell during active feeding with a large food vacuole within the basket (black triangles); th e food vacuole contains irregularly-shaped, 'fuzzy' material, presum ably taken up by endocytosis th rough th e cytostom e (white triangles). Two rows of fibers representing the left (LF) and right (RF) margins of the basket accom pany the food vacuole. 4D. An SEM im age of Picomonas visualizing th e left side of the cell. Note th a t th e posterior flagellum has been shed at th e tr2; w hite triangles indicate th e cytostom e region of the feeding apparatus. 4E. Longitudinal section th rough the basket near th e cytostom e. It shows approx. 60 rows of fibers arranged in parallel (white arrows depict fibers in the right margin of the basket). 4F. 3D m odel of th e feeding apparatus with rows of fibers in parallel arrangem ent (white arrows) forming a basket (arrangem ent of the cell as in 4D). Note th a t th e fibers are in terconnected by thin filaments. The basket is open tow ards th e to p and right (i.e. tow ards the anterior and dorsal direction of th e cell respectively), while it is closed at th e bo ttom (the posterior end of the cell). On th e left side of th e basket (representing th e cell's ventral surface) the fibers are attached to th e plasma m em brane thus forming th e narrow, slit-like cytostom e. AF/PF (an te rio r-/p o s te rio r flagellum); G (Golgi body); LF/RF (left and right row of fibers of th e basket); M (mitochondrion); MB (microbody); N (nucleus); P




(posterior digestive body); Ar2 (anterior m icrotubular flagellar root 2); Pr1 (posterior m icrotubular flagellar root 1); Pr2 (posterior m icrotubular flagellar root 2); FV, (food vacuole); vc (vacuolar cisterna); cy (cytostome). Numbers a t the to p right indicate the serial section. Scale bar: 200 nm; except Fig. 4A, section 2 (100 nm). doi:10.1371/journal.pone.0059565.g004

dense ribbons a ttached to the plasm a m em brane that m ediate this connection. W e also no ted that the posterior flagellum is located along the ventral cell surface in close proxim ity to the cytostom e (200-400 nm distance) runn ing alm ost parallel to the cytostom e slit (Fig. 2E; Fig. 4C!, sections 1,3,16; Fig. S I, section 7; M ovie S2).

T h e two flagella em erge from the ventral cell surface of the cell, close to the ju n ctio n o f the Golgi body and the ventral tip o f the m itochondrion (Fig. 5A; Fig. S I; M ovie S2). Both flagella have a sm ooth surface an d lack appendages (hairs, scales) w hen observed by T E M and SEM (Fig. 5A,B; see also Fig. 2C). T heir tips form short hair-points o f on average 0.6 pm length (n = 5; not shown). T h e flagellar bases form an angle o f 120-140" (n = 5). W hen viewed from the ventral side, the an terio r flagellum projects towards the celFs left, its base form ing an angle o f about 40" to the anterio-posterior cell axis (Fig. 5A; Fig. 6 Cl ; M ovie S2). T he

posterior flagellum deviates only slightly (~ 10- 2 0 ") from this axis projecting tow ards the cell’s right an d extending to the cell’s posterior. T h e two basal bodies are displaced against each o ther by about one basal body diam eter, in that the an terio r basal body is located m ore ventrally th an the posterior basal body (Fig. 5A; Fig. 6A,C; M ovie S2). Except for a single proxim al connecting fiber (Fig. 5A, section 3) b o th basal bodies w ere no t in terconnected by fibrous structures.

T h e basal body and the flagellar transitional region display several unusual features: Basal bodies are relatively short (360- 380 nm) an d m ost o f their lum en is filled w ith electron dense m aterial (Fig. 5A,B). This m aterial seems to have th in connections to the A-tubules o f the m icrotubular triplets (Fig. 5B, section 11) and to the proxim al transitional p late (Fig. 5C!, section 3). O ften the electron dense m aterial consists o f a longer, distal p a rt o f high

Figure 5. Flagellar apparatus. 5A. Longitidinal section o f Picomonas from ventral to dorsal (sections are oblique from anterior part). Anterior flagellum (AF) and posterior flagellum (PF). The anterior flagellum locates close to the ventral surface followed m ore interiorly by the posterior flagellum. The basal bodies of both flagella are connected by a proximal connecting fiber (pcf) and exhibit tw o transitional regions (the proximal is term ed 't r i ' and the distal 'tr2'). 5B, C. Consecutive serial cross sections and tw o serial longitudinal sections of the anterior flagellum. Serial sections of 5B correspond to 5C, d eno ted in 5B_a-j a t th e left lower end. a. The axonem e with 9outer d oub le t and 2 central pair m icrotubules, b. The distal trasitional region (tr2) is involved in flagellar shedding (* indicates electron dense material near ou te r doublets), f. the central pair of m icrotubules orginate. h. the transitional region 1 (tri), C-tubules added (arrowhead indicates a m icrotubulular triplet and arrow indicates a m icrotubular doublet. i,j. cross sections th rough basal body with m icrotubular triplets arranged in the clockwise direction, th e basal body lum en is filled with electron dense material. AF/PF (anterior—/posterio r flagellum); trq, tr2 (distal [tr2] and proximal [tri] flagellar transitional regions); Ab/Pb (anterior—/posterio r basal body); pcf (proximal connecting fiber); G (Golgi body); N (nucleus); Pr2 (posterior m icrotubular flagellar root 2). Num bers a t the to p right indicate the serial section. Scale bar: 100 nm. doi:10.1371/journal.pone.0059565.g005




tr2

Apb

Ppb

PFtr2

tr2(2) Ar1

(2) Ar2

Pr2 (6)

Figure 6. Basal apparatus and probasal bodies. 6A. Electron m icrograph show s consecutive serial sections o f Picomonas judraskeda from left to right (Sections are angled abo u t 30 deg ree tow ards left from central axis). Each probasal body diverges from its parental basal body a t right angles, both probasal bodies are oriented tow ards th e ventral surface o f the cell and are attached to the plasma m em brane. 6B. 3D- schem e of the arrangem ent of basal bodies with probasal bodies. 6C. A schem atic p resentation o f the flagellar basal appara tu s seen from th e ventral surface of the cell with proximal parts o f the axonem es, basal bodies and four m icrotubular flagellar roots (for details see text). Numbers in brackets indicate the num ber o f m icrotubules in each root. AF/PF (anterior—/posterio r flagellum); trq,tr2 (distal [tr2] and proximal [tri] flagellar transitional regions); Ab/Pb (anterior—/posterio r basal body); A pb/Ppb (anterior/posterior probasal body); pcf (proximal connecting fiber); Ar1/Ar2 (Anterior m icrotubular flagellar roots 1 and 2); Pr1/Pr2 (posterior m icrotubular flagellar roots 1 and 2); G (Golgi body); N (nucleus); M (m itochondrion); cy (cytostome); P (posterior digestive body). Numbers at the to p right indicate the num ber o f th e serial section. Scale bar: 200 nm. doi:10.1371/journal.pone.0059565.g006

electron density (260-290 nm in length) and a shorter proxim al p a rt o f lower electron density (60-80 nm in length), separated by an electron- translucent space o f about 30 -5 0 nm in length (Fig. 5A, sections 2, 3). A cartw heel has no t been observed in either o f the basal bodies (but m ay have been obscured by the proxim al electron dense material). A nother unusual feature o f the flagellar bases is the presence o f two transitional plates in the proxim al part o f each flagellum (Fig. 5A,C, (trl& tr2)). T h e spacing betw een the transitional plates was usually a round 380-400 nm bu t occasionally could be as short as 320 nm (Fig. 5A,C). Both transverse plates appeared to be structurally identical, however, in our fixation, the distal transverse plate (tr2) constricted the flagellum an d axonem e (Fig. 5A,C). At this site the flagellum seems to be shed because we often observed flagellar stumps term inating w ith the tr2 (see also Fig. 4D , PF; Fig. 6A section 7). Interestingly, the central pa ir o f m icrotubules originate n ear the t r i and extend th rough the tr2 (Fig. 5A, section 2 and Fig. 5C, section 2).

D uring interphase, the basal bodies generally do not app ear to be associated with probasal bodies, bu t in one cell (likely in p repara tion for cell division), we observed probasal bodies associated w ith each basal body (Fig. 6A,B, (Ppb, Apb)). Both probasal bodies extended from their paren tal basal bodies at right angles, and interestingly were oriented towards the same side, closely associated with the ventral surface o f the cell (Fig. 6A, B).

Serial sections revealed the presence o f four m icrotubular flagellar roots, two associated w ith each basal body. W e have given the roots descriptive term s (A rl, Ar2, P r l , Pr2; Fig. 6 Cl ; M ovie S2), because we did no t study basal body developm ent during the cell cycle. T h e A rl originates at the an terio r basal body, consists o f two

m icrotubules and runs towards the cell’s an terio r left (Fig. 6A, section 5; see also Fig. 7A, section 45). It accom panies the nuclear envelope and term inates before reaching the an terio r end o f the cell (not shown; M ovie S2). T h e Ar2 originates a t the anterior basal body (near its proxim al end; Fig. 7A, section 47) and also consists o f two m icrotubules, runs beneath the ventral cell surface to the cell’s posterior passing the m itochondrion and en tering the posterior pa rt (PP) o f the cell, w here it term inates n ear to the an terio r tip o f the cytostom e just opposite to the posterior flagellum (Fig. 4C, section 1; Fig. 6 Cl ; Fig. 7A, section 47; M ovie S2). T h e P r l originates at the left surface o f the posterior basal body n ear its proxim al end (Fig. 6A, sections 6 , 7), consists o f two m icrotubules, runs along the ventral surface o f the cell to the cell’s posterior, passes the ventral surface o f the m itochondrion , and enters the posterior p a rt (PP) o f the cell w here it term inates near the an terio r edge of the left side wall o f the cytostom e (Fig. 4B, section 2 an d Fig. 4C sections 1, 3; Fig. 6A, section 6 an d Fig. 6 Cl ; Fig. 7A, section 54; M ovie S2). T h e Pr2 originates near the right surface o f the posterior basal body (Fig. 6A, section 8 and Fig. 6 Cl ; Fig. 7A, sections 47, 49). It is a b ro ad root consisting o f 6 m icrotubules (Fig. 7B, sections 49, 50, and Fig. 7D, sections 5-7; M ovie S2). It runs close to the ventral surface o f the cell towards the cell’s posterior, passes the ventral surface o f the m itochondrion, is located in a shallow depression o f the latter (Fig. 7D, sections 5, 6), then traverses the vacuolar cisterna (Fig. 7C). T h e Pr2 takes a p a th parallel to the P r l extending into the posterior p a rt (PP) o f the cell (Fig. 6A, section 8) and term inates n ear the an terio r end of the cytostom e near its right side wall (Fig. 4B, section 2 and Fig. 4C, sections 1, 3; Fig. 6 Cl ; M ovie S2). All three posteriorly




PLOS ONE I www.plosone.org March 2013 I Volume 8 I Issue 3 I e59565



Figure 7. Flagellar root system in Picom onas judraskeda . 7A. Non-consecutive serial sections from dorsal to ventral (sections are oblique to cell's right). Each basal body is connected to tw o m icrotubular flagellar roots. The anterior root 1 (Arl ) runs anteriorly to cell's left. The Ar2 runs posteriorly, a t the right side o f the cell and passes th e cleft (cl). The o th er tw o flagellar roots originate from th e posterior basal body and extend tow ards the posterior part o f the cell. One of the posterior flagellar roots (Prl ) runs on th e left side of th e cell. The o ther broader posterior flagellar root (Pr2), runs betw een the Ar1 and P rl. 7B. The Pr2 with 6 m icrotubules (arrowheads) obliquely sectioned. 7C. A cell with the Pr2 passing the vacuolar cisterna and m itochondrion. 7D. Consecutive serial cross sections th rough th e Pr2 located in a depression of the m itochondrion. AF/PF (anterior—/posterior flagellum); Ab/Pb (anterior—/posterio r basal body); Ar1/Ar2 (Anterior m icrotubular flagellar roots 1 and 2); Pr1/Pr2 (posterior m icrotubular flagellar roots 1 and 2); G (Golgi body); M (m itochondrion); vc (vacuolar cisterna); P (posterior digestive body); CMT (secondary cytoplasmic m icrotubule). Numbers a t the to p right indicate the num ber o f th e serial section. Scale bar: 200 nm. doi:10.1371/journal.pone.0059565.g007

oriented roots (Ar2, P r l and Pr2) run parallel to each o ther into the posterior p a rt (PP) o f the cell, spaced 0.5 pm (from Ar2 to P r l; Fig. 4C, section 1; Fig. 6 C; M ovie S2). T h e P r l an d Pr2 run for abou t 400 nm into the posterior p a rt o f the cell, w here the m icrotubules term inate n ear the plasm a m em brane (the Ar2 seems to term inate before the o ther roots). A t this position, electron dense ribbons associated w ith the plasm a m em brane extend from the P r l an d Pr2 an d continue alongside the left (from P rl) and right (from Pr2) side wall o f the cytostom e to the cell’s posterior end (Fig. 4B,C; Fig. 6C; M ovie S2). A few additional cytoplasmic m icrotubules (C M T in Fig. 7A, section 53) originate a t an angle o f abou t 45° from the proxim al region o f the P r l , to extend to the cell’s left, an d ru n for an unknow n distance along the left surface of the an terio r p a r t o f the cell close to the vacuolar cisterna (not shown).

T h e cells are covered only by the plasm a m em brane w ith no scales or glycocalyx being discernible. O ften, rod-shaped bacteria were encountered apparendy physically a ttached to the plasm a m em brane by their ends. T hese bacteria were (ultra)structurally in tact and could associate with any p a rt o f the plasm a m em brane (except flagella) (not shown).

Molecular PhylogenyA m olecular phylogenetic analysis o f a broadly sam pled taxon

set (104 taxa o f eukaryotes, excluding only Excavata) using a data set consisting of 18S rD N A , 5.8S rD N A and 28S rD N A (4461 aligned characters) could no t position the Picozoa in one o f the know n eukaryotic supergroups. N either m onophyly o f the Flacrobia [54] n o r the previously reported sister group relation o f Picozoa w ith the telonem ids [41] could be substantiated in this analysis (results no t shown). T o explore the genetic diversity o f the Picozoa, phylogenetic analyses o f 201 partial nuclear-encoded SSU rD N A sequences o f Picozoa (with the exception of P.

judraskeda all others derived from environm ental clone libraries) were perform ed (Fig. 8). T h e tree revealed a high genetic diversity w ithin the Picozoa (for a tree w ith all accessions, see Fig. S2). O f the three m ajor clades previously reported in the Picozoa (‘biliphyte’ clades ‘BP 1 -3 ’; [36]) only one (‘BP2’) was m onophy- letic in our analyses. W e tentatively identify 12 novel clades (‘P I P IS ’, for Picozoa clades 1-13) in the Picozoa (clade P4 positioned w ithin the “ 15 unresolved tax a” in ‘B P1’ is currently uncertain and no t depicted in Fig. 8) th a t were well-supported by bootstrap and Bayesian posterior probability values (except for clades 10 and 11; Fig. 8). In addition to clades 1-13, m any sequences rem ained w ithin unresolved assemblages, presum ably because of the low num ber o f inform ative sites available in these partial SSU rD N A sequences. Picomonas judraskeda was positioned in the well-supported clade P3 whose m em bers had alm ost identical sequences and included also “ clone j j 11” from our environm ental clone library o f the same hab ita t (Fig. 8). Seven o ther o f o u r environm ental clones o f Picozoa were positioned in clade P2, w hereas one clone (“ Picozoa Fie”) was a new genotype in clade ‘B P2’ (Fig. 8).

Taxonomic Summary

Phylum Picozoa phylum novD ia g n o s is . Fleterotrophic, m arine protists o f picoplanktonic

size (cells m ay pass th rough a 3 pm m em brane filter) mostly characterized by either o f two signature sequences in the nuclear- encoded SSU rD N A , 5'G C G T G A T G C CAA AAT C C G 3' (PICOBIOl) or 5 'ATA T G C C C G T C A AAC C G T 3' (PICOBI02).

Class Picomonadea class novD ia g n o s is . W ith the characteristics o f the phylum .

Order Picomonadida ord. novD ia g n o s is . W ith the characters o f the phylum . T ax a are

characterized by the signature sequence 5 'G C G T G A T G C CAA AAT C C G 3 ' (PICO BIOl) th a t represents a m olecular synapo- m orphy o f the order. T h e m ost inclusive clade containing taxa with the sequence accession num bers G U 822951 (clade P 6), H Q ß65255 (clade P5) and Picomonas judraskeda (JX988758; clade P3), b u t no t H Q 868810 (clade P 8) an d JX 9 8 8 7 6 7 (‘B P 2 j. T his is a b ranch-based definition in w hich all o f the specifiers are extant. T h e Picom onadida includes a t least clades P 1 -P 6 plus several hitherto unresolved taxa.

Family Picomonadidae fam. novD ia g n o s is . W ith the characters o f the order. T h e m ost

inclusive clade contain ing taxa w ith the sequence accession num bers H Q 868687 (unresolved), EU 368015 (clade PI), EU 368029 (unresolved), D Q 222877 (clade P2) and Picomonas

judraskeda (JX988758; clade P3), b u t no t GU 822951 (clade P 6) and H Q ß65255 (clade P5). This is a b ranch-based definition in which all o f the specifiers are extant. T h e P icom onadidae includes a t least clades P 1 -P 3 plus several hitherto unresolved taxa.

Type Genus. Picomonas gen. nov

Genus Picomonas gen. novD ia g n o s is . Cells are biflagellata w ith a long and a short

flagellum inserted laterally. E ach cell consists o f two nearly hem ispherical parts separated by a cleft. T h e an terio r p a r t contains the flagellar apparatus, nucleus, endoplasm ic reticulum , single Golgi body an d m itochondrion with tubular cristae, w hereas the posterior p a r t is variable in size an d contains the feeding apparatus and num erous vesicles and vacuoles. T h e an terio r and posterior parts o f the cell are separated by a single, large, vacuolar cisterna th a t leaves only p a r t o f the m itochondrion in direct contact with the posterior part. T h e cells exhibit a unique m ode o f motility: After extended periods o f rest, a stereotypic p a tte rn is initiated consisting o f a rapid, short-distance ju m p , im m ediately followed by a slower, dragging cell m ovem ent in the opposite direction. This pa tte rn m ay be repeated several times before cells finally show an extrem ely fast and extended m ovem ent away from their original




96/99/1.00

89/61/1.00-

87/84/1.00-15 unresolved seq.

rEU368019; FS04G188 01Aug05 5m -EU368018; FS04G190_01Aug05_5m EU368003; FS01AA1 l_01Aug05_5m l EF539143;TH10.59

JX988759T2S1 W T SDP 071

Picozoa clone #11;-FJ349634; 051011 HQ867521; SHAH506 HQ869166; SHBA495 HQ869754; SHBF736 FJ431633; RA080215T.063 HQ866250; SGSP723 DQ222878; RA001219.38

Picomonas judraskeda ; JX988758

O .Û Û

coÛ .

o■ Ö

o

48 unresolved seq.

78/74/1.00

82/-/0.97-J

- / - / l .00

i____

. • 1 3 i -

- /

HQ867354; SHAC64885/82/-1------ HQ867337; SHAC623

^H Q 867420; SHAC744DQ060524; NW414.27

- / - / 0.91

AY343928; He000803-7 99/100/1.00

00/0.85-93/63/1.00

14 unresolved seq.

85/93/-

COCLCO

-/63/0.88

-/-/0.80 99/99/1.00

99/92/1.00

rDQ222874; HE001005.148'

CNIÚ _m

Picozoa He; JX9887670.05

substitutions/site




Figure 8. Unrooted randomized accelerated maximum likelihood (RAxML) phylogenetic tree of partial nuclear SSU rDNA of Picozoa. The phylogeny is based on 1253 aligned characters o f the SSU rDNA and includes 201 sequences o f Picozoa. Most sequences are da tabase entries derived from clone libraries (nine environm ental sequences genera ted from a sam ple taken a t Helgoland Roads and one sequence from Picomonas judraskeda are new sequences; accession num bers of the newly determ ined sequences are p resen ted in Fig. S2). Bootstrap values >60% and posterior probabilities >0.80 for the three m ethods o f analyses used (RAxML/NJ by PAUP/MrBayes) are show n on the respective branches. Branches in bold show maximal (100% /1.00) support. Labeling of clades ('BP1-3') followed [Cuvelier e t al. 2008], 12 novel clades ('PI —P I3') are recognized (for details see Results). The sequence o f Picomonas judraskeda (in bold) was positioned in clade 'P3' (shaded). doi:10.1371/journal.pone.0059565.g008

position (term ed ‘skedaddle’). T h e genus represents the m ost inclusive clade containing the type species, P. judraskeda ((JX988758; clade P3), b u t no t taxa w ith the sequence accession num bers H Q 868687 (unresolved), EU 368015 (clade PI), EU 368029 (unresolved), an d D Q 222877 (clade P2). This is a branch-based definition in w hich all o f the specifiers are extant. T h e genus Picomonas includes a t least all taxa in clade P3.

Type Species. Picomonas judraskeda sp. nov

Picomonas judraskeda sp. novD ia g n o s is . C haracters o f the genus. T h e oblong cells vary in

length betw een 2 .6 -3 .8 pm , their w idth is 2 -2 .5 pm . T h e longer flagellum m easures 12-14 pm ; the shorter flagellum 7 -9 pm . T he longer (anterior) flagellum is oriented tow ards the anterior; the shorter (posterior) flagellum towards the posterior end of the cell. T h e nucleus is hem ispherical, and over the spherical p a r t o f its surface is appressed to the p lasm a m em brane. T h e feeding apparatus essentially consists o f a basket o f abou t 50-60 parallel runn ing fibers (30 nm repeat) o f varying lengths th a t extend from the ventral surface w here they are a ttached to the plasm a m em brane for up to 1.2 pm tow ards the dorsal p a r t w here they term inate. T h e basket is open tow ards the cell’s an terio r and dorsal parts, b u t closed tow ards the posterior end o f the cell. T he slit-like cytostom e is form ed a t the ventral cell surface w here the ‘side walls’ o f the basket fibers connect to the plasm a m em brane and extends in the anterior-posterior d irection for a length of abou t 1 pm , its w idth being abou t 150 nm .

T ype h a b ita t. M arine plankton.T ype lo ca lity . Surface w ater (5 m depth) from H elgoland

R oads, (54°11 'N , 7°54'E), N orth Sea, G erm any.T ype m a ter ia l. T h e nam e-bearing hapantotype is a block of

resin-em bedded cells for electron m icroscopy (prepared from a single cell-derived culture established from the original natural sample) deposited a t the C ulture Collection o f Algae a t the U niversity o f Cologne (CCAC; h ttp ://w w w .ccac.u n i-k o e ln .d e /), G erm any, un d er the designation P IC O .H .001. A parahapan to- type block from the same fixation is deposited a t the C C A C under the designation P IC O .H .002 . T h e culture was subsequently lost and is no longer available.

E tym ology . N am ed after its unique stereotypic m ode o f cell motility, w hich consists, in succession, o f a short fast ju m p (ju-) into the an terio r direction, a slow drag (-dra-) into the opposite direction and an extrem ely fast and extended m ovem ent o f the cell away from its original position (skedaddle; -skeda). A lthough colloquial in the English language and o f unknow n origin, the nam e ‘skedaddle’ m ay be derived from the G reek ü K sS aap ô ç (skedasmos,“ dispersion”).

Discussion

Rationale for the New Protist Phylum PicozoaT h e hetero trophic protist Picomonas judraskeda gen. nov., sp. nov.,

described here as a m em ber o f a new protist phylum , the Picozoa, has apparen tly no t yet been studied before; the living cell and its m orphology by light and electron m icroscopy were unknown.

G ene sequences obtained from environm ental clone libraries, however, previously identified a unique pico- o r nanoplanktonic eukaryotic lineage th a t has b ro ad therm al an d geographic distribution and becam e know n under the nam es ‘picobiliphytes’ [34] o r ‘biliphytes’ [36]. Originally envisaged as a novel photosynthetic lineage w ith affinities to katab lepharids/cryp tophytes [34,36], recent w hole-genom e shotgun sequence da ta o f three ‘(pico)biliphyte’ cells sorted by FA CS from an environm ental sample, did no t find any evidence o f p lastid D N A or o f nuclear- encoded p lastid-targeted genes in these genomes, and concluded th a t ‘(pico)biliphytes’ were likely heterotrophic [41], All previous phylogenetic studies using environm ental sequence com parisons, however, agreed th a t these organism s com prise a genetically unique and diverse novel eukaryotic group to be delineated a t a high taxon level. In the recently p roposed revised classification of eukaryotes, ‘(pico)biliphytes’ have been placed into “ Incertae sedis E ukaryota” [55] and denoted as “ Poorly characterized, known only from environm ental samples, an d no species o r genera described.” W e established a single cell-derived culture o f a ‘(pico)biliphyte’ and characterized it by light an d electron m icroscopy. O u r results support the conclusion of Y oon et al. (2 0 1 1 ) th a t these organism s are heterotrophic because no plastids were found. In addition, we revealed a set o f highly unusual behavioral and structural features o f the cells th a t to the best o f our knowledge have no t yet been reported for any o ther eukaryotic cell. A m ong these features we list: (1) flagellate cells exhibit a stereotypic pa tte rn o f motility consisting o f three phases, “jum p , d rag an d skedaddle,” (2) each cell is separated into two parts o f alm ost equal size, an an terio r p a rt containing the co m p artm en ts / organelles typical o f a eukaryotic cell, an d a posterior p a rt that consists exclusively o f vacuoles/vesicles and the feeding apparatus.(3) A single, large vacuolar cisterna physically “ seals” b o th parts o f the cell except for a specialized region in w hich regular projections o f the outer m itochondrial envelope, term ed “ mitovilli,” m ediate d irect contact betw een bo th cell parts. (4) a feeding apparatus consisting o f a large basket o f fibers th a t term inate a t the ventral cell surface thereby defining the boundaries o f a long, slit-like cytostom e, w hich allows form ation o f a large food vacuole containing only particles o f less than 150 nm in size, (5) finally, three o f the four m icrotubular flagellar roots en ter the posterior p a rt o f the cell, being closely spaced, and term inate near the an terio r end o f the cytostome.

W e feei th a t these features together w ith their unresolved position in the eukaryotic phylogenetic tree justify the recognition o f this w idespread group of m arine pico- o r nanoplanktonic protists a t the phylum level. W e recognize, however, th a t we have investigated only one representative species an d d o n ’t know to w hat extent the structural features described for Picomonas judraskeda apply to o ther m em bers o f the phylum . In the diagnosis o f the phylum , we therefore m ake only lim ited use o f the newly discovered structural features, although we believe th a t m ost o f them will eventually be found to be diagnostic for the phylum w hen m ore m em bers have been studied (for the choice o f the nam e Picozoa, see below). T h e phylogenetic affinities o f the Picozoa m ust rem ain uncertain a t present. W e found no evidence for their p lacem ent in the novel eukaryotic supergroup H acrobia


http://www.ccac.uni-koeln.de/



[54], w hich itself has been recently called into question [56] and was also no t recovered in our analysis (unpublished observations). O th e r groups with w hich the Picozoa have previously been affiliated, nam ely Telonem a, katablepharids/cryp tophytes or Plantae [34,36,40,41,56], in our analyses using the rR N A operon did no t show affiliations with the Picozoa, w hich rem ained in an unresolved position (unpublished observations). W e also conclude th a t the structural features th a t characterize Picomonas judraskeda b ear little resem blance to the cell structures in katab lepharids/ cryptophytes, T elonem a or Plantae [54,57-60].

Are Picozoa Picoeukaryotes?Picoplankton was originally defined as those organisms whose

cell size lies betw een 0.2 an d 2 pm [5]. A lthough initially specified as a size fraction < 2 pm , there has never been full agreem ent abou t the operational definition o f the term picoplankton [61]. Because m ost field studies, especially those addressing small eukaryotic plankton, have been conducted using 3 pm filters, it has becom e custom ary to extend the size range of organism s regarded as picoplankton to those th a t pass filters o f pore size 3 pm , i.e. the size fraction < 3 pm [6,15]. In lieu o f cultures and direct m easurem ents o f cell sizes, cell size fractionation th rough isopore- type m em brane filters an d epifluorescence m icroscopy using autofluorescence (in case o f photosynthetic cells), often in com bination w ith fluorescence in-situ hybridization (FISH) served as a surrogate for cell sizes in ‘uncu ltu red ’ organisms. These approaches, however, are no t w ithout pitfalls: Initial studies reported th a t ‘(pico)biliphyte’-specific sequences could be recovered after filtering seawater th rough either 3 pm [34] o r 2 pm [36] filters suggesting th a t these cells were indeed of ‘p ico’-size. Epifluorescence m icroscopy including F IS H and using either whole seawater [36] o r cells sorted according to phycoerythrin autofluorescence [34] w ithout pre-filtration th rough 2 -3 pm pore size filters, however, re tu rned signals th a t depicted irregularly- shaped particles o f larger sizes (3.5-6 pm ; [34,36]). These results leave room for in terpretation . Fragile cells are know n to break during filtration [62], i.e. ‘(pico)biliphyte’ D N A in samples filtered th rough 2 or 3 pm filters could have originated from larger cells th a t did no t survive the filtration procedure intact. O n the o ther hand, larger fluorescing particles on filters do no t necessarily represent only the target cells; they could correspond to cell aggregates o r reflect trophic (i.e. phagotrophic) interactions betw een different partners. T h a t the phycoerythrin-fluorescing particles have m isled authors o f earlier investigations to conclude th a t Picozoa w ere photosynthetic (and hence ‘picobiliphytes’ or ‘biliphytes’) is now obvious (Yoon e t al. 2011 an d this study), bu t was already h in ted a t in a study th a t reported absence o f phycoerythrin fluorescence in N orth Pacific samples th a t contained ‘biliphyte’ cells using F IS H probes [40].

W h at do our results contribute to the question o f w hether Picozoa are picoeukaryotes or no t o r phrased differently, w hat can we conclude abou t the ‘real’ dim ensions o f these cells? First, we can conclude th a t no t only D N A b u t live cells o f Picozoa can pass th rough 2 pm filters since we established a culture o f P. judraskeda from samples pre-filtered in this way. T hus, operationally defined, the Picozoa are picoeukaryotes. H ow ever, filter thresholds should no t be overin terpreted as flexible cells m ay pass th rough pores that are smaller th an their actual cell size [63]. O u r m easurem ents o f the cell sizes o f live P. judraskeda cells, and o f cells fixed and em bedded for transm ission electron m icroscopy revealed th a t cell lengths in interphase cells were quite variable ranging from 2 .5 -3.8 pm (whereas cell widths were m uch less variable, see Results). T h e variability in cell lengths am ong cells is m ainly caused by the variable size o f the posterior cell p a rt (PP) th a t contains the feeding

apparatus and the system o f ‘digestive’ vacuoles/vesicles an d m ight refer to either different phases o f a feeding cycle or a starvation state o f the cell (see below). Newly divided cells a re also considerably smaller th an average cells. I t is obvious th a t the smaller cells o f P. judraskeda are ^ 3 p m (2.5 x2 pm) and could pass th rough bo th 2 o r 3 pm filters intact, especially as P. judraskeda cells are naked providing additional plasticity to the cells. O n average, cell lengths m easured in thin-sectioned cells were abou t 5% less th an those m easured in live cells, w hich is to be expected because o f m inor cell shrinkage during dehydration. T h e apparently larger cell sizes observed by LM in chem ically fixed cells are likely caused by swelling due to osm otic im balances in the fragüe cells when fixed with low glutaraldehyde concentrations. W e therefore conclude th a t cells o f Picomonas judraskeda fall into the picoplanktonic cell-size range, bo th operationally by passing through a 2 pm filter, and by direct m easurem ent o f cell sizes o f live cells. W e therefore use the prefix ‘pico’ in the nam e o f the new phylum ; although we recognize th a t the cell dimensions o f P. judraskeda are a t the upper lim it o f the p icoplankton and during feeding transcend the bo rder to the nanoplankton.

Feeding Behavior and Food SourcesW hereas the hetero trophic na ture o f the Picozoa is now beyond

doubt, their m ode of feeding rem ains essentially unknow n. K im et al. (2011) speculated th a t phycoerythrin fluorescence in Picozoa m ay have been the result o f phagotrophic feeding o f Picozoa on cyanobacteria, e.g. Synechococcus spp., whereas Y oon et al. (2011) discussed the possibüity th a t the reported plastid an d nucleom orph [34] m ay have “ come from a kleptoplastid or cryptophyte alga cap tured as food” by Picozoa. In their study on single-cell am plified genomes o f three individual cells o f Picozoa isolated from seawater Y oon et al. (2011) discovered high abundances o f specific single-stranded and double-stranded D N A viruses as well as D N A from m arine bacteria o f the Bacteriodetes, P roteobacteria, and Firm icutes groups. Furtherm ore, the three cells differed with respect to the associated viruses and bacteria. T h e authors concluded th a t they had studied “ com plex biotic interactions am ong previously uncharacterized m arine m icroorganism s,” and regarding one cell, a “virus infection cap tured in-situ,” although they did no t rule ou t passive a ttachm en t o f viral and bacterial D N A to the surface o f the cells. In conclusion, Yoon et al. (2011) suggested th a t Picozoa m ight feed on Proteobacteria, Bacteriodetes and large D N A viruses [41].

D o our electron m icroscope observations shed light on the feeding behavior an d the likely food source of the Picozoa? O ne of the m ost unusual structural features o f P. judraskeda is the subdivision o f the cell into two parts, an an terio r p a r t housing alm ost all cell constituents an d a posterior part, containing w hat we have identified as a digestive system including the feeding apparatus. Interestingly bo th parts are separated by a large vacuolar cisterna th a t leaves larger spaces only for in teraction of the posterior digestive body with the single m itochondrion and for passage o f three m icrotubular flagellar roots th a t presum ably position the cytostom e (see Results). W e serially sectioned 52 cells, b u t never encountered an in tact o r recognizable bacterium within the putative food vacuole inside the feeding apparatus, in the posterior digestive body or in any o ther p a r t o f the cell. Because we fixed a growing culture, we w ould expect to encounter bacteria, if they were a suitable food source for P. judraskeda. A lthough we canno t exclude the possibüity th a t a specific bacterium , th a t was no t p resent in our bacterized culture, could serve as a selective food item for P. judraskeda, there are o ther reasons to believe th a t P.

judraskeda does not, an d in fact cannot, feed on bacteria (feeding of P. judraskeda on Synechococcus spp. can be excluded, because we




never encountered ch lorophyll/phycoerythrin autofluorescence in sorted cells th a t were shown to yield ‘(pico)biliphyte’-specific DN A after P C R am plification with the respective primers). W e apparently fixed cells for electron m icroscopy a t different stages in the feeding cycle as evidenced by the presence or absence of a food vacuole w ithin the feeding basket (see Results). W e m easured the w idth o f the cytostom e (the region m arked by the a ttachm ent o f the fibers o f the feeding basket to the plasm a m em brane) at these different stages in the feeding cycle an d found th a t it d id not change, always being a ro u n d 150 nm . W e conclude th a t the slitlike cytostom e is a rigid structure th a t cannot take up particles that are larger than its w idth, thus excluding bacteria. T his m ay also explain the peculiar motile behavior o f the cells, w hich is very unlike th a t o f bacterivorous nanoflagellates [64,65]. W e propose th a t P. judraskeda feeds on particles smaller than 150 nm th a t are taken up by a fluid phase, bulk flow m echanism . This generates a single food vacuole o f enorm ous size (it can be estim ated th a t the m em brane a rea o f a large food vacuole corresponds to abou t 3 0 - 40% o f the total plasm a m em brane surface o f the posterior p a rt o f the cell) arguing for rap id m em brane turnover. A t present, we can only speculate abou t the force(s) th a t initiate the fluid phase, bulk flow m echanism . T h e close association o f the posterior flagellum with the cytostom e slit (see Results) m ay indicate th a t flagellar motility (perhaps during the d rag m ovem ent) could be involved. R egard ing the possible food source, we note th a t the contents are irregular aggregates o f ‘fuzzy’ m aterial th a t resem ble < 1 2 0 nm m arine colloids, w hich are dispersed widely in seawater [66] and m ay contain lipopolysaccharide m aterial o f bacterial m em branes [67]. T h e Picozoa m ay thus be specially adap ted to exploit < 1 2 0 nm m arine colloids as a food source. T h e ‘skedaddle’ m ovem ent could then be envisaged no t prim arily as a phobic response to escape predators, b u t ra th e r as a m echanism to explore new food resources once grazing a t a specific location has depleted resources. T h e abundance an d spatial distribution of m arine colloids m ay also explain the relatively low num ber o f Picozoa th a t we observed in our culture o f P. judraskeda (30-40 cells/m l) as well as the com parable cell num bers reported in na tura l samples using F IS H (55 cells/m l in N o t et al. 2007 [34] and up to 300 cells/m l in Cuvelier et al. 2008 [36]). This also suggests th a t filter-sterilized seawater (0.1 pm filters) could have been the m ajor source o f m ost o f the colloidal food particles necessary to support grow th o f P. judraskeda, although a con tribution by the bacterial population is also likely. W e assum e th a t the unusual structures observed in the m itochondrion (mitovilli and the electron dense inclusions in the in term em brane space) m ay also be involved in the processing o f specific, perhaps mostly lipidie molecules derived from small colloidal food particles. T h e two prom inen t m icrobodies could be involved in the degradation o f fatty acids derived from such lipidie molecules.

C ould Picozoa perhaps feed on viruses as well as suggested by Y oon et al. (2011)? Viruses constitute the m ost ab u n d an t group of nucleic acid-containing particles in the ocean and up to 108 virus particles per milliliter have been recorded in productive coastal surface waters [68]. A lthough filtration o f na tu ra l seawater th rough 0.1 pm filters w ould likely exclude the larger size class o f m arine viruses, the smaller size class (30-60 nm), w hich is 4- to 10-fold m ore abundant, w ould easily pass th rough a 0.1 pm filter [69]. I f we assume th a t P. judraskeda feeds on < 1 5 0 nm particles by a fluid-phase bulk flow m echanism s, then it is likely th a t small viruses, such as circular single-stranded D N A viruses (Nanoviridae, C ircoviridae; [70]) w ould be taken up as well. This m ight explain their prevalence in the single-cell genom e am plification of Picozoa [41], although virus particles or D N A a ttached to the surface o f a cell o r even co-sorted with such a cell (given the high

num ber o f viral particles p resent in seawater, a sorted droplet o f on average 10 picoliters [71], could still contain one or two co-sorted virus particles) should no t be dismissed. A lthough we did not recognize viral particles inside food vacuoles o f P. judraskeda, we do no t exclude the possibility th a t Picozoa take up small size-class viruses during feeding. W hether these are digested as proposed by Y oon et al. (2011) o r exocytosed unaltered during the feeding cycle needs further investigation. W e note, however, th a t the large vacuolar cisterna separating the an terio r from the posterior p a r t o f the cell w ould be ideally positioned to p reven t access o f endocytosed viral particles to the cell’s nucleus.

ConclusionD uring the last decade, culture-independent m olecular surveys

based on rD N A clone libraries, phylogenetic analyses, and fluorescence in-situ hybridization have revealed num erous novel, high-ranking picoeukaryotic (< 3 pm) lineages in the oceans. This new knowledge is rapidly altering our understand ing o f m arine m icrobial food webs, and the biogeochem ical significance of m arine protists [72]. A lthough culture-independent techniques have been essential for the discovery o f picoeukaryotic biodiversity, for an understand ing of the biology o f the organism s involved, they should be com plem ented by studies o f the respective organism s in culture. H ere, we provided evidence th a t a genetically diverse and apparently w idespread group o f p icoeukaryotes in the w orld’s oceans, hitherto know n as ‘picobiliphytes’ or ‘biliphytes’, an d here form ally described as Picozoa phylum nov., displays highly unusual structural and behavioral characteristics th a t m atch its isolated position in the eukaryotic phylogenetic tree. Based on the characteristics described for Picomonas judraskeda gen. nov., sp. nov., we conclude th a t Picozoa are hetero trophic and feed on small (< 1 5 0 nm) particles by a novel fluid-phase, bulk flow uptake m echanism . Further studies on o ther m em bers o f the Picozoa are needed to substantiate this conclusion. W e strongly recom m end th a t m ore effort should be m ade to cultivate the vast ‘uncu ltu red’ diversity o f eukaryotic m icrobes in the sea.

Supporting Information

Figure SI E le c tr o n m ic r o g r a p h s o f n o n -c o n se c u t iv e lo n g itu d in a l s e r ia l s e c t io n s th r o u g h a P ic o m o n a s j u d r a s k e d a c e ll f r o m it s le f t to r ig h t su r fa c e . A com plete reconstruction o f the cell shows the absence of a plastid. N um bers in the upper right corners o f the respective m icrographs correspond to the num ber o f the serial section. A hem ispherical nucleus (N) occupied a large volum e o f the an terio r p a r t (AP) o f the cell. A pproxim ately, 60% o f the surface of the nuclear envelope is closely associated with the plasm a m em brane. A single m itochondrion (M) w ith tubular cristae is positioned n ear the ventral surface o f the cell (by definition the surface from w hich the flagella em erge).The Golgi com plex (G) consists o f a single Golgi body characteristically located in an an terio r groove of the m itochondrion betw een the nucleus and the flagellar apparatus. Tw o basal bodies (Ab, Pb) and the flagellar basal apparatus are located n ear the Golgi body an d the m itochondrion a t the ventral surface o f the cell. Tw o prom inen t ‘m icrobodies’ (MB) are located n ear the dorsal surface in the A P of the cell. T h e posterior p a rt o f the cell (PP) contains the ‘feeding ap p ara tus’ (cytostome (cy)/ feeding basket) and num erous vacuoles/vesicles. A large vacuolar cisterna (vc) separates the an terio r from the posterior p a r t (AP/PP) o f the cell.(TIF)

Figure S2 U n r o o te d r a n d o m iz e d a c c e le r a te d m a x im u m l ik e l ih o o d (RAxM L) p h y lo g e n e t ic tr e e o f p a r t ia l n u c le a r




SSU rD N A o f P ic o z o a , a c c e s s e d f r o m G en b a n k in J u ly 2012. T h e sequence (ace no JX 9 8 8 7 5 8 ) from Picomonas judraskeda as well as o ther newly determ ined sequences (ace no JX 988759- JX 988767) are highlighted in the tree. B ootstrap values > 6 0 % and posterior probabilities > 0 .8 0 for the three m ethods o f analyses used (R A xM L /N J by PA U P/M rB ayes) are shown on the respective branches. Branches in bold show m axim al (100% /I.00) support. Clades m arked as ‘P I - P I 3’, predicted 12 novel picozoan clades supported by b o o tstrap /posterio r probability values (‘clade’ P4 currently rem ains unresolved). All sequences in the tree are shown w ith their accession num bers followed by the sample designation as provided in the database.(EPS)

T ab le SI P r im e r s u s e d fo r a m p lif ic a t io n a n d s e q u e n c in g o f n u c le a r rR N A o p e r o n .(TIP)

M ovie SI M o v e m e n t o f a c e l l o f P ic o m o n a s ju d r a s k e d a .(AVI)

References1. A z a m F, F en ch e l T , F ield J G , G ray J S , M ey er-R e il LA , e t al. (1983) T h e

ecological ro le o f w a te r co lu m n m icrobes in th e sea. M a r E co l P ro g r Ser 10:2 5 7 -2 6 3 .

2. G io v an n o n i SJ, S tingl U (2005) M o lecu lar d iversity a n d ecology o f m icro b ia l p lan k to n . N a tu re 437: 3 4 3 -3 4 8 .

3. G io v an n o n i SJ, V e rg in K L (2012) S easonality in o c e an m ic ro b ia l com m unities. Science 335: 6 7 1 -6 7 6 .

4. S tro m SL (2008) M ic ro b ia l ecology o f o cean b iogeochem istry : a c o m m u n ity p erspective. Science 320: 1 0 4 3 -1 0 4 5 .

5. S ie b u rth J M , S m e ta c e k V , L e n z J (1978) P e lag ic ecosystem s tru c tu re : h e te ro tro p h ic co m p a rtm e n ts o f th e p lan k to n a n d th e ir re la tio n sh ip to p lan k to n size fractions. L im n o l O c e a n o g r 23: 1256—1263.

6. V a u lo t D , E ik rem W , V ip rey M , M o re a u H (2008) T h e diversity o f sm all eu k ary o tic p h y to p la n k to n ( < 3 pm ) in m arin e ecosystem s. F E M S M ic ro b io l R ev 32: 7 9 5 -8 2 0 .

7. P a r te n s k y F, F less W R , V a u lo t D (1999) P ro c h lo ro c o c c u s , a m a r in e p h o to sy n th e tic p ro k ary o te o f g lobal significance. M ic ro b io l M o l Biol R ev 63: 1 0 6 -1 2 7 .

8. S can lan D J, O strow sk i M , M a z a rd S, D u fre sn e A, G arc z a re k L, e t al. (2009) E co log ical genom ics o f m arin e p ico cy an o b ac te ria . M ic ro b io l M o l B iol R ev 73:2 4 9 -2 9 9 .

9. S to ck n er J G , A n tia N J (1986) A lgal p ico p la n k to n f ro m m arin e a n d fresh w ater ecosystem s: A m ultid iscip linary p erspective . C a n J F ish A q u a t Sei 43: 2 472— 2503.

10. D iez B, Pedrós-A lió C , M a ssa n a R {2001) S tu d y o f gene tic d iversity o f eukaryo tic p ico p la n k to n in d iffe ren t ocean ic reg ions b y sm all-su b u n it rR N A g en e c lon ing a n d sequencing . A p p l E n v iro n M ic ro b io l 67: 2932—2941.

I I . L ó p e z -G a rc ía P , R o d r ig u ez -V a le ra F , Pedrós-A Jío C , M o re ira D (2001) U n e x p e c te d d iversity o f sm all euk ary o tes in deep -sea A n ta rc tic p lan k to n . N a tu re 409: 6 0 3 -6 0 7 .

12. M o o n -v a n d e r S taay SY, D e W a ch te r R , V a u lo t D (2001) O c e a n ic 18S rD N A sequences f ro m p ico p la n k to n rev ea l u n su sp ec ted eu k ary o tic diversity . N a tu re 409: 6 0 7 -6 1 0 .

13. J ü rg e n s K , M assan R (2008) P ro tis ta n g raz in g o n m arin e b a c te rio p la n k to n . In K irc h m a n D L (Ed) M ic ro b ia l Ecology o f th e O cean s. W iley , N e w Y o rk N Y ,383—441.

14. W o rd e n A Z , N o t F (2008) Eco logy a n d diversity o f p icoeukaryo tes . I n K irc h m a n D L (Ed) M ic ro b ia l Ecology o f th e O cean s . W iley , N ew Y ork , N Y , 1 5 9 -1 9 6 .

15. M assan a R (2011) E u k ary o tic p ico p la n k to n in surface oceans. A n n u R ev M ic ro b io l 65: 9 1 -1 1 0 .

16. M assan a R , B alagué V , G uillou L , Pedrós-A lió C (2004) P icoeukaryo tic d iversity in a n o lig o tro p h ic coasta l site s tu d ied b y m o lecu la r a n d c u ltu ra l ap p ro ach es . F E M S M icro b io l E co l 50: 2 3 1 -2 4 3 .

17. S h e rr EB , S h e rr B F (2002) S ignificance o f p re d a tio n b y p ro tists in aq u a tic m ic ro b ia l fo o d w ebs. A n to n ie v an L eeu w en h o ek 81: 293—308.

18. C h a m b o u v e t A, M o rin P, M a rie D , G uillou L (2008) C o n tro l o f tox ic m arin e d inoflagellate b loom s b y seria l p a ras itic killers. Science 322: 1254—1257.

19. B iegala IC , N o t F, V a u lo t D , S im o n N , C u rie M (2003) Q u a n tita tiv e assessm ent o f p ico eu k ary o tes in the n a tu ra l e n v iro n m en t us in g taxon-specific o ligonucleo tid e p ro b es in assoc ia tion w ith T S A -F IS H {tyram ide signal am plification - f lu o rescen t in situ hybrid izatio n ) a n d flow cy tom etry . A p p l E n v iro n M icro b io l 69: 5 5 1 9 -5 5 2 9 .

20. N o t F , S im on N , B iegala IC , V a u lo t D (2002) A p p lica tio n o f flu o rescen t in situ h y b rid iz a tio n co u p led w ith ty ra m id e signal am p lificatio n (F IS H -T S A ) to assess eu k ary o tic p ico p la n k to n com position . A q u a t M ic ro b E co l 28: 157—166.

Movie S2 Animation o f three dimensional view of Picomonas judraskeda.(AVI)

Movie S3 Animation o f three dimensional structure of the feeding apparatus o f Picomonas judraskeda.(AVI)

A cknow ledgm entsRS wishes to thank the Alfred-Wegener-Institute for Polar and Marine Research (Bremerhaven, Germany) for hosting initial experiments and the R R Z K of the University o f Cologne for providing computational resources (CHEOPS).

Author ContributionsConceived and designed the experiments: RS LK M M M . Performed the experiments: RS. Analyzed the data: RS NS M M . C ontributed reagents/ materials/analysis tools: LK M M M . W rote the paper: RS LKM MM.

21. N o t F, L a tasa M , M a rie D , C a r io u T , V a u lo t D , e t al. (2004) A single species Micromonas pusilla (P rasinophyceae) d o m in a tes th e eu k ary o tic p ico p la n k to n in the w este rn Eng lish C h a n n e l. A p p l E n v iro n M ic ro b io l 70: 4064—4072.

22. M a n g o t J F , L ep é re C , B ouv ier C , D eb ro as D , D o m a iz o n I (2009) C o m m u n ity s tru c tu re a n d d y n am ics o f sm all euk ary o tes ta rg e te d b y n e w o ligonucleo tide p robes: n ew insig h t u b ti th e lacu strin e m ic ro b ia l fo o d w eb. A p p l E n v iro n M ic ro b io l 75: 6 3 7 3 -6 3 8 1 .

23. L ó p e z -G a rc ía P , M o re ira D (2008) T ra ck in g m ic ro b ia l b io d iversity th ro u g h m o lecu la r a n d g en o m ic ecology. R es M icro b io l 159: 67—73.

24. G uvelier M L , A llen A E , M o n ie r A , M c G ro w J P , M essie M , e t al. (2010) T a r g e te d m e ta g e n o m ic s a n d eco lo g y o f g lo b a lly im p o r ta n t u n c u ltu re d eu k ary o tic p h y to p lan k to n . P ro c N a d A c a d Sei U S A 107: 1 4 6 7 9 -1 4 6 8 4 .

25. W o rd e n A Z , Ja n o u sk o v e c J , M cR o se D , E n g m a n A , W elsh R M , e t al. (2012) G lo b a l d istr ib u tio n o f a w ild a lga rev ea led b y ta rg e te d m etag en o m ics. C u r r Biol 22: R 6 7 5 -R 6 7 7 .

26. M assan a R , T e r ra d o R , F o rn I, Lovejoy G , Pedrós-A lió G (2006) D is tr ib u tio n a n d a b u n d a n c e o f u n c u ltu red h e te ro tro p h ic flagellates in th e w o rld oceans. E n v iro n M icro b io l 8: 1 5 1 5 -1 5 2 2 .

27. K im D Y , C o u n tw a y P D , G a s t R J, C a ro n D A (2011) R a p id shifts in th e stru c tu re a n d co m p o sitio n o f a p ro tis ta n assem blage d u r in g b o td e in cu b a d o n s affect e stim ates o f to ta l p ro tis ta n species richness. M ic ro b E co l 62: 383—398.

28. W e b er F , D e l C a m p o J , W ylez ich G, M assan a R , J ü rg e n s K (2012) U nveilling tro p h ic fu n cd o n s o f u n c u ltu red p ro tis t tax a b y in cu b a d o n ex p e rim en ts in the b rack ish B alde Sea. P loS O N E 7 (7): e41970.

29. S ensen C W , H e im a n n K , M elk o n ian M (1993) T h e p ro d u c tio n o f c lo n a l a n d a x en ic cu ltu res o f m ic ro alg ae using fluo rescen ce -ac tiv a ted cell so rting (FACS). E u r J P hyco l 28: 9 3 -9 7 .

30. V e ld h u is M JW , K ra a y G W (2000) A p p lica tio n o f flow cy to m etry in m arin e p h y to p la n k to n research : c u rre n t app lica tions a n d fu tu re p erspectives. Sei M a r 64: 12 1 -1 3 4 .

31. H ey w o o d J L , S ieracki M E , Bellows W , P o u lto n N J, S tep an au sk as R (2011) C a p tu r in g d iversity o f m arin e h e te ro tro p h ic pro tists: o n e cell a t a tim e. IS M E J5: 6 7 4 -6 8 4 .

32. Y ilm az S, S ingh A K (2012) Single cell g en o m e sequencing . C u r r O p in B io tech n o l 23: 437—443.

33. S tep an au sk as R (2012) Single cell genom ics: a n in d iv id u a l look a t m icrobes. C u r r O p in M icro b io l 15: 6 1 3 -6 2 0 .

34. N o t F, V a le n tin K , R o m a ri K , Lovejoy G , M a ssa n a R , e t al. (2007) Picobiliphytes: a m arin e p ico p lan k to n ic a lgal g ro u p w ith u n k n o w n affinities to o th e r eukaryo tes. Science 315: 253—255.

35. N o t F , G au slin g R , A z a m F, H e id e lb e rg J F , W o rd e n A Z (2007) V ertica l d istr ib u tio n o f p ico eu k a ry o tic d iversity in th e Sargasso Sea. E n v iro n M ic ro b io l 9: 1 2 3 3 -1 2 5 2 .

36. G uvelier M L , O r t iz A , K im E , M o eh lig H , R ic h a rd so n D E (2008) W id e sp rea d d istr ib u tio n o f a u n iq u e m arin e p ro tis ta n lineage. E n v iro n M ic ro b io l 10: 1621— 1634.

37. L e G all F , R ig a u t-Ja la b e rt F, M a rie D , G a rczarek L , V ip rey M , e t al. (2008) P ic o p la n k to n d iv ers ity in th e S o u th -E a s t P ac ific O c e a n f ro m cu ltu res . B iogeosciences 5: 203—214.

38. L i L , H u a n g Q W u S, L in D , C h e n J , e t al. (2008) T h e sp a tia l a n d tem p o ra l d istr ib u tio n o f m ic ro alg ae in th e S o u th C h in a Sea: ev idence fro m G IS -b ased analysis o f 18S rD N A sequences. Sei C h in a S er G: Life sei 51: 1 1 2 1 -1 1 2 8 .

39. M assan a R , P e rn ice M , B u n g e J A , D e l C a m p o J (2011) S eq u en ce diversity a n d novelty o f n a tu ra l assem blages o f p ico eu k ary o te s f ro m th e In d ia n O c e a n . IS M E J 5: 1 8 4 -1 9 5 .




40. K im E , H a rr iso n J W . S udek S. J o n e s M D M , W ilcox H M . e t al. (2011) N ew ly id en tified a n d d iverse p las tid -b ea rin g b ra n c h o n th e eu k ary o tic tre e o f life. P ro c N a d A cad Sei U S A 108: 1 4 9 6 -1 5 0 0 .

41. Y o o n H S , P rice D C , S tep an au sk as R , R a ja h V D , S ieracki M E , e t al. (2011) Single-cell genom ics reveals o rg an ism a l in te rac tio n s in u n cu ltiv a ted m arin e p ro tists . Science 332: 7 1 4 -7 1 7 .

42. M c F a d d en G I, M e lk o n ian M (1986) U se o f H e p e s b u ffe r fo r m icro alg a l cu ltu re m ed ia a n d fix a tio n fo r e lec tro n m icroscopy . Phyco log ia 25: 551—557.

43. G e im e r S, M e lk o n ian M (2004) T h e u ltra s tru c tu re o f th e Chlamydomonas reinhardtii b a sa l ap p a ra tu s: id en tificatio n o f a n early m ark e r o f rad ia l asy m m etry in h e re n t in th e b asa l body . J C ell Sei 117: 2663—2674.

44. K re m e r J R , M a stro n a rd e D N , M c In to sh J R (1996) C o m p u te r v isua liza tion o f th ree -d im en s io n a l im ag e d a ta using IM O D . J S tru c t B iol 116: 71—76. h t t p : / / b io 3 d .c o lo ra d o .e d u /im o d /.

45. M a rtin -C e re ce d a M , R o b e r ts E C , W o o tto n E C , B onaccorso E , D y al P , e t al. (2009) M o rp h o lo g y , u ltra s tru c tu re , a n d sm all su b u n it rD N A ph y lo g en y o f the m arin e h e te ro tro p h ic flagella te Goniomonas aff. amphinema. J E u k a ry o t M icrob io l 57: 1 5 9 -1 7 0 .

46. M ed lin L K , M etfies K , M eh l H , W iltsh ire K , V a le n tin K (2006) P icoeukaryo tic p lan k to n diversity a t th e H e lg o la n d tim e series site as assessed b y th re e m o lecu lar m ethods. M ic ro b E co l 52: 5 3 -7 1 .

47. P e e r Y V D , W u y ts J , P e rrie G , B e rn a rd A C (2004) T h e E u ro p e a n rib o so m al R N A d a tab ase . N u cle ic A cids R es 32: D 1 0 1 -D 1 0 3 .

48. Z u k e r M (2003) M fo ld w eb server fo r nuc le ic ac id fo ld ing a n d h y b rid iza tio n p red ic tio n . N u c le ic A cids R es 31: 3 4 0 6 -3 4 1 5 .

49. S tam atak is A (2006) R A x M L -V I-H P C : m ax im u m lik e lihood-based phy lo g en e tic analyses w ith th o u san d s o f tax a a n d m ix ed m odels. B io in fo rm atics 22: 2 688— 2690.

50. Sw offord D L (2002) PA U P*. P hy logenetic A nalysis U sin g P a rsim o n y V e rs io n 4. {*and O th e r M ethods). S in au er A ssociates, S u n d erlan d .

51. P o sad a D , C ra n d a ll K A (1998) M O D E L T E S T : tes tin g th e m o d el o f D N A su bstitu tion . B io in fo rm atics 14: 8 1 7 —818.

52. R o n q u is t F , H u e lsen b eck J P (2003) M rB ayes 3: B ayesian p h y lo g en e tic in ference u n d e r m ix ed m odels. B io in fo rm atics 19: 1572—1574.

53. A ltek ar G , D w ark ad as S, H u e lsen b eck J P (2004) P a ralle l M etro p o lis co u p led M ark o v c h a in M o n te C a rlo fo r B ayesian p h y lo g en e tic in ference. B io inform atics 20: 4 0 7 M 1 5 .

54. O k a m o to N , C h a n ta n g si C , H o rá k A , L e a n d e r BS, K ee lin g I J (2009) M o lecu lar ph y lo g en y a n d d escrip tio n o f th e novel k a ta b le p h a r id Roombia truncata gen. e t sp. nov ., a n d e s tab lish m en t o f th e H a c ro b ia ta x o n nov. P L oS O N E 4: 11.

55. A d i S M , S im p so n A G B , L an e C E , Lukes J , Bass D , e t al. (2012) T h e revised classification o f eukaryo tes. J E u k a ry o t M icro b io l 59: 4 2 9 M 9 3 .

56. B urki F , O k a m o to N , P o m b e r tJ -F , K ee lin g F f (2012) T h e e v o lu tio n ary h istory o f h ap to p h y tes a n d c ryp tophy tes: p h y lo g en o m ic ev id en ce for s e p ara te origins. P ro c R o y Soc B 279: 2 2 4 6 -2 2 5 4 .

57. M o estru p 0 (2000) T h e flagella te cy toskeleton: In tro d u c tio n o f a g en e ra l term in o lo g y for m ic ro tu b u la r flagellar ro o ts in pro tists. In L ea d b e a te r B S C , G re e n J C (eds) T h e F lagellates. Syst Assoc S pecial V ol. Ser 59 , T a y lo r & F rancis, L o n d o n , N ew Y ork , 6 9 -9 4 .

58. O k a m o to N , In o u y e I (2005) T h e k a ta b le p h arid s a re a d istan t sister g ro u p o f the C ry p to p h y ta : a p ro p o sa l fo r K a ta b le p h a rid o p h y ta disisio n o v a /K a th a b le p h a r id a p h y lu m n o v u m b a sed o n S S U rD N A a n d b e ta - tu b u lin phy logeny . P ro tis t 156: 16 3 -1 7 9 .

59. S h a lch ia n -T ab riz i K , E ik rem W , K laveness D , V a u lo t D , M in g e M A , e t al. (2006) T e lo n em ia , a n ew p ro tis t p h y lu m w ith affinity to ch ro m ist lineages. P ro c R o y Soc B 273: 1 8 3 3 -1 8 4 2 .

60. Y ab u k i A , In ag ak i Y , Ish id a K (2010) Palpitomonas bilix gen. e t sp. N ov.: a novel d e e p -b ra n c h in g h e te ro tro p h possib ly re la te d to A rch aep las tid a o r H a c ro b ia . P ro d s t 161: 5 2 3 -5 3 8 .

61. C ra ig S R (1986) P ic o p lan k to n size d istrib u tio n s in m arin e a n d fresh w aters: P ro b lem s w ith filter fra c d o n a tio n studies. F E M S M icro b io l E co l 38: 171—177.

62. W o rd e n A Z (2006) P ico eu k ary o te d iversity in coasta l w aters o f the Pacific O c e a n . A q u a t M ic ro b E co l 43: 16 5 -1 7 5 .

63. L i W K W (1990) P ard cles in “p a rtic le -free” seaw ater: g ro w th o f u ltrap h y to - p lan k to n a n d im p licad o n s fo r d ilu tion exp erim en ts . C a n T F ish A q u a t Sei 47: 1 2 5 8 -1 2 6 8 .

64. M a tz C , B oenigk J , A rn d t H , J ü rg e n s K (2002) R o le o f b a c te r ia l p h e n o ty p ic tra its in selective feed in g o f th e h e te ro tro p h ic n anflage lla te Spumella sp. A q u a t M ic ro b E co l 27: 1 37-148 .

65. E rk e n M , F a rre n sc h o n N , S p e c k m a n n S, A rn d t H , W e ite re M (2012) Q u a n d fic a d o n o f in d iv id u a l f lag e lla te -b acteria in te rac tio n s w ith in sem i-n a tu ra l b iofilm s. P ro tis t 163: 6 3 2 -6 4 2 .

66. W ells M L , G o ld b e rg E D (1991) O c c u rre n c e o f sm all colloids in sea w ater. N a tu re 353: 342—344.

67. W a k eh a m S G , P ease T K , B en n er R (2003) H y d ro x y fatty acids in m arin e d isso lved o rg an ic m a tte r as in d ica to rs o f b a c te r ia l m e m b ra n e m ateria l. O rg G eo ch em 34: 8 5 7 -8 6 8 .

68. S u td e C A (2005) V iruses in th e sea. N a tu re 437: 3 5 6 -3 6 1 .69 . M a rie D , B ru ssaa rd C P D , T h y rh a u g R , B ra tb a k G , V a u lo t D (1999)

E n u m e ra d o n o f m a rin e viruses in c u ltu re a n d n a tu ra l sam ples b y flow cy tom etry . A p p l E n v iro n M ic ro b io l 65: 45—52.

70. R o sa rio K , D uffy S, B re itb a rt M (2012) A field gu ide to eu k a ry o d c c ircu la r s in g le -s tran d ed D N A viruses: insights g a in ed fro m m etagenom ics. A rc h V iro l 157: 1 8 5 1 -1 8 7 1 .

71. S ieracki M , P o u lto n N , C ro sb ie N (2005) A u to m a te d iso lad o n tech n iq u es for m icro alg ae . In A n d ersen R A (Ed) A lgal C u ltu r in g T ech n iq u es . Elsevier, B u rlin g to n M A , 1 0 1 -1 1 6 .

72. M assan a R , Pedrós-A lió C (2008) U n v eilin g n ew m ic ro b ia l euk ary o tes in the surface o cean . C u r r O p in M ic ro b io l 11: 213—218.



open 3 access freely available online © plosi o

Documents