Studying non-model organisms is crucial in the context of the current development of genomics and transcriptomics for both physiological experimentation and environmental characterization. We investigated the transcriptomes of two marine planktonic ciliates, the mixotrophic oligotrich Strombidium rassoulzadegani and the heterotrophic choreotrich Strombidinopsis sp., and their respective algal food using Illumina RNAseq. Our aim was to characterize the transcriptomes of these contrasting ciliates and to identify genes potentially involved in mixotrophy. We detected approximately 10,000 and 7,600 amino acid sequences for S. rassoulzadegani and Strombidinopsis sp., respectively. About half of these transcripts had significant BLASTP hits (E-value <10−6) against previously-characterized sequences, mostly from the model ciliate Oxytricha trifallax. Transcriptomes from both the mixotroph and the heterotroph species provided similar annotations for GO terms and KEGG pathways. Most of the identified genes were related to housekeeping activity and pathways such as the metabolism of carbohydrates, lipids, amino acids, nucleotides, and vitamins. Although S. rassoulzadegani can keep and use chloroplasts from its prey, we did not find genes clearly linked to chloroplast maintenance and functioning in the transcriptome of this ciliate. While chloroplasts are known sources of reactive oxygen species (ROS), we found the same complement of antioxidant pathways in both ciliates, except for one enzyme possibly linked to ascorbic acid recycling found exclusively in the mixotroph. Contrary to our expectations, we did not find qualitative differences in genes potentially related to mixotrophy. However, these transcriptomes will help to establish a basis for the evaluation of differential gene expression in oligotrichs and choreotrichs and experimental investigation of the costs and benefits of mixotrophy.
Citation: Santoferrara LF, Guida S, Zhang H, McManus GB (2014) De Novo Transcriptomes of a Mixotrophic and a Heterotrophic Ciliate from Marine Plankton. PLoS ONE 9(7): e101418. https://doi.org/10.1371/journal.pone.0101418
Editor: Ross Frederick Waller, University of Cambridge, United Kingdom
Received: May 5, 2014; Accepted: June 6, 2014; Published: July 1, 2014
Copyright: © 2014 Santoferrara et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files.
Funding: This research was funded by the Gordon and Betty Moore Foundation through Grant GBMF2637 to NCGR and by the US National Science Foundation through Grant OCE1130033 to GBM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The most diverse and abundant ciliates in euphotic marine waters correspond to two sister subclasses, Oligotrichia and Choreotrichia (class Spirotrichea) . Oligotrich and choreotrich ciliates are globally distributed  and episodically dominate microzooplankton –. They are major consumers of small algae, thus channeling energy through the microbial loop and higher levels in the planktonic food web . One of the most prominent physiological differences between these two ciliate groups is that many oligotrich species practice mixotrophy, while this nutrition mode has not been confirmed for any choreotrich –.
Mixotrophs obtain nutrients and energy by combining heterotrophy and autotrophy  and play key roles as both primary and secondary producers . The mechanism for mixotrophy in oligotrichs is chloroplast sequestration, or kleptoplasty, in which a primarily herbivorous organism retains functional chloroplasts from its algal food and uses them for photosynthesis. For example, the oligotrich Strombidium rassoulzadegani captures chloroplasts from algal prey and uses them to grow rapidly in the light, although chloroplasts are not able to divide in the ciliate and eventually need to be replaced –. Apart from ciliates, kleptoplasty has been widely reported to occur within dinoflagellates, foraminiferans, and even in some molluscs –.
It is unclear how a kleptoplastic organism can keep functional chloroplasts. Most genes needed to regulate these organelles are nuclear-encoded, but the algal nucleus is usually not retained in the host cell . An exception is the ciliate Mesodinium rubrum, in which the nuclei of ingested algae remain transcriptionally active . The most popular hypothesis on the genetic basis of kleptoplasty is related to the horizontal transfer of genes involved in chloroplast functioning and maintenance from algae to the host nucleus , . For example, five plastid-targeting proteins that function in photosystem stabilization and metabolite transport have been found encoded in the nucleus of the kleptoplastic dinoflagellate Dinophysis acuminata and have probably been acquired through horizontal gene transfer from multiple algal sources . In contrast, no support for horizontal gene transfer has been found in another kleptoplastic protist, the foraminiferan Elphidium margaritaceum, and thus other hypotheses related to chloroplast stability have been suggested , .
Kleptoplasty provides the advantage of a photosynthetic energy subsidy, but it is unclear if this strategy provides other benefits or costs to the cell . One hypothetical cost of kleptoplasty is the necessity for mitigation of reactive oxygen species (ROS) produced during photosynthesis. ROS are produced during respiration and normal metabolism both in heterotrophic and autotrophic organisms, which have multiple mechanisms of detoxification , . Additional ROS are produced and detoxified in the chloroplasts of autotrophs –. It is unknown how a kleptoplastic ciliate mitigates the extra ROS produced by the sequestered chloroplasts. Maintaining a different or more active detoxification mechanism and the risk of additional oxidative stress may represent costs of mixotrophy for ciliates. This is also interesting from the evolutionary point of view, as an enhanced ability to deal with ROS would partially explain why only some ciliates can harbor photosynthetic symbionts , . On the other hand, some accumulation of ROS may provide defense against predation , thus helping to explain why mixotrophic ciliates appear to be less vulnerable than heterotrophic ones to copepod grazing .
The Marine Microbial Eukaryote Transcriptome Sequencing Project (MMETSP; http://marinemicroeukaryotes.org) has provided an unprecedented amount of genetic information on ciliates and other non-model marine protists that have been strongly underrepresented in previous genomics and transcriptomics efforts . As part of this initiative, we performed RNAseq on two ciliates, the mixotrophic oligotrich S. rassoulzadegani and the heterotrophic choreotrich Strombidinopsis sp., and provide here the first transcriptome analyses for these groups. To eliminate potential food contamination, and given the lack of whole-genome information for any algae we could use as food, we also obtained the transcriptomes of the prey used for ciliate culturing. Our aim was to characterize the transcriptomes of these contrasting ciliates and, in particular, to explore the hypothesis that additional genes involved in kleptoplasty and ROS mitigation are expressed by the mixotroph compared to the heterotroph. Given the complete novelty of this kind of data, our study advances the understanding of the physiology of mixotrophic and heterotrophic ciliates and lays the molecular groundwork necessary for further experimentation.
Materials and Methods
Strombidium rassoulzadegani and Strombidinopsis sp. were sampled from a tide pool and a dock, respectively, at the UConn Avery Point campus, Connecticut, USA (41.32° N, 72.06° W). No special permits were needed and field collection did not involve endangered or protected species.
Cultures and RNA extraction
Both ciliates were isolated and maintained in autoclaved, filtered seawater supplemented with mineral nutrients (salinity c. 30 Practical Salinity Units and nutrients added as f/2 or f/20 for the mixotroph and the heterotroph, respectively ). S. rassoulzadegani has been kept in our laboratory for almost 10 years using the prasinophyte Tetraselmis chuii (strain PLY429), which is the prey that provides the most efficient growth of this ciliate . Strombidinopsis sp. has been periodically isolated and cultured in our laboratory using the cryptophyte Rhodomonas lens (strain RHODO) as food. For RNA isolation, new cultures of S. rassoulzadegani and Strombidinopsis sp. were started using T. chui PLY429 and R. lens RHODO, respectively, as prey.
When the ciliate cultures were in the exponential growth stage and the food algae in the culture were largely consumed, ciliate cells were harvested. To minimize food contamination in the ciliate RNA extracts, individual cells were picked with a micropipette under a stereo microscope and pooled into a 15-ml tube containing 5 ml of Tri-Reagent (MRC Inc., Cincinnati, OH, USA). A total of 22,000 cells for S. rassoulzadegani and 10,000 for Strombidinopsis sp. were isolated. Also, the two food algae (∼107 cells) were harvested from axenic cultures by centrifugation at 3,000 x g and the cell pellets were fixed in Tri-Reagent. RNA was extracted from all four samples following the modified Zymo column purification method using Direct-zol RNA MiniPrep Kit (Zymo Research, Irvine, CA, USA) as reported previously .
Library preparation and RNAseq
RNA samples were quantified using Qubit Q32855 (Invitrogen, Carlsbad, CA, USA) and their quality was assessed using the Agilent 2100 Bioanalyzer. Libraries for each of the four species were made from 2 µg RNA using the TruSeq RNA Sample Preparation Kit (Illumina, San Diego, CA, USA). Libraries were sequenced on the Illumina HiSeq 2000 to obtain paired-end, 50-bp-long reads. Approximately 2 Gbp of sequence data was generated per library.
Transcriptome assembly was carried out using the internal pipeline BPA1.0 (Batch Parallel Assembly version 1.0) of the National Center for Genome Resources. Sequence reads were preprocessed using SGA  for quality trimming (swinging average) at Q15. Reads shorter than 25 nucleotides (nt) after trimming were discarded. Preprocessed sequence reads were assembled into contigs with ABySS v. 1.3.0 , using 20 unique kmers between k = 26 and k = 50. ABySS was run requiring a minimum kmer coverage of 5, and bubble popping at >0.9 branch identity with the scaffolding flag enabled to maintain contiguity for divergent branching. Paired-end scaffolding was performed on each kmer. Sequence read pairing information was used in GapCloser v. 1.10 as part of the SOAP de novo package  to walk in on gaps created during scaffolding in each individual kmer assembly. Contigs from all gap-closed kmer assemblies were combined. The OLC (overlap layout consensus) assembler miraEST  was used to identify minimum 100 base pair overlaps between the contigs and assemble larger contigs, while collapsing redundancies. BWA  was used to align sequence reads back to the contigs. Alignments were processed by SAMtools mpileup (http://samtools.sourceforge.net) to generate consensus nucleotide calls at positions where IUPAC bases were introduced by miraEST , and read composition showed a predominance of a single base. In an attempt to remove incomplete sequences, the consensus contigs were filtered at a minimum length of 150 nt to produce the final set of contigs. Sequences are available in the CAMERA Portal (http://camera.calit2.net/mmetsp/list.php ) under the unique MMETSP identifiers included in Table 1.
Prediction of coding regions and elimination of food transcripts
DNA coding sequences (CDS) and the corresponding amino acid sequences (AAS) were predicted using ESTScan , . Resulting CDS and AAS numbers were slightly different given the different length cut-offs used for each dataset (150 nt vs. 30 aa, respectively). A Bacillariophyta scoring matrix was used based on availability of well-annotated mRNA entries in NCBI RefSeq. Illumina sequence reads were aligned back to the nucleotide motifs of the assembled contigs and predicted CDS using BWA  to assess assembly quality.
Putative algal sequences in the ciliate data were identified with BLASTN  using the food algae transcripts as reference database and an E-value of 10−6 as cut-off. Sequences with a significant hit were removed from the ciliate datasets using custom scripts.
Ciliate AAS datasets were contrasted with OrthoMCL using default settings . First, all-against-all BLASTP searches (E-value <10−6) were done to identify reciprocal best hits between species. Then, homologous AAS were grouped and each group was putatively classified as orthologous (gene families separated by speciation) or paralogous (gene duplications subsequent to speciation).
Predicted AAS were annotated using Blast2GO . The NCBI non-redundant NR database and an E-value cut-off of 10−6 were used for BLASTP . Annotated Gene Ontology (GO) terms were complemented with results from InterProScan and associated KEGG pathways were retrieved. In addition to Blast2GO, AAS characterization was done also with the more sensitive method HMMER3  against the Pfam-A , TIGRFAM  and SUPERFAMILY  databases. Information on proteins of particular interest (e.g. related to photosynthesis or response to ROS) was retrieved manually from both Blast2GO and HMMER3 results. This strategy was used also to confirm absence of certain AAS in Strombidinopsis sp. or algae datasets.
For one protein of interest (Nec3, see below), ciliate transcripts and other amino acid sequences downloaded from NCBI GenBank were combined and aligned with MUSCLE . Overlapping regions were trimmed resulting in a final alignment of 225 sites. For phylogenetic inferences, both Neighbor Joining (as implemented in MEGA ) and Maximum Likelihood (RAxML ) analyses were carried on with 1,000 bootstrap replicates. The evolution model LG with a Г model of rate heterogeneity and a proportion of invariable sites was used as selected by ProtTest under the Akaike Information Criterion .
Results and Discussion
Transcriptome assemblies, filtering of food transcripts, and ciliate AAS datasets
We sequenced the transcriptomes of two marine planktonic ciliates, Strombidium rassoulzadegani and Strombidinopsis sp., as well as their two respective algal foods. The number of Illumina reads and assembled contigs obtained in this study ranged from ca. 14 to 43 million and 12 to 33 thousand per species, respectively (Table 1). Half of the total assembled nucleotides were contained in sequences of 1,300 nt or larger as indicated by N50 values (minimum size cut-off = 150 nt). The fact that over 80% of Illumina reads were realigned to these contigs confirms the adequate quality of the assemblies. However, there are no reference genomes for any of the ciliates and algae sequenced and thus we cannot make any conclusions about the completeness of the transcriptomes. For ciliates, the closest species with a known genome is Oxytricha trifallax, which belongs to a different subclass (Stichotrichia) of the Spirotrichea. The genome of this species, which is fragmented into thousands of nanochromosomes, is about 50 Mb long and is estimated to encode ca. 18,400 genes . This gene content is within the range of transcripts assembled for S. rassoulzadegani and Strombidinopsis sp. (Table 1), although it is unclear how comparable the O. trifallax genome is to those of Oligotrichia and Choreotrichia species.
Given that oligotrich and choreotrich ciliates cannot be cultured independently of their food, we had to include a step to eliminate putative algal transcripts from the ciliate data. Although ciliate cells were picked individually to avoid contamination, algal 18S rRNA was detected in the ciliate samples, possibly due to prey being digested within the ciliates. A total of 7.7% and 24.0% of the CDS (equivalent to 5% and 6% of Illumina reads) obtained from S. rassoulzadegani and Strombidinopsis sp. cultures, respectively, were removed as food transcripts. Analysis of GC content in CDS indicated that the filtering procedure was successful (Fig. 1A-B). The frequency distribution of GC content per CDS was identical between algal transcripts and transcripts eliminated from ciliate data. In contrast, the distribution of filtered ciliate transcripts showed distinctive peaks, with maximum frequency of CDS with 60% and 40–50% GC content in S. rassoulzadegani and Strombidinopsis sp., respectively.
A-B Frequency distribution of GC content in ciliate transcripts, algal food transcripts, and putative food transcripts filtered from ciliate data. C- Venn diagram of all-against-all BLASTP hits between ciliates. D- Venn diagram of orthologous and paralogous amino acid sequence groups in ciliates.
We detected 10,015 AAS for S. rassoulzadegani and 7,608 AAS for Strombidinopsis sp. (Table 1). Using OrthoMCL, 2,279 out of the total 17,623 AAS were identified as reciprocal best hits between the two species (Fig. 1C). In addition, 1,310 out of 3,150 total AAS groups were shared between species (orthologous), while the remaining groups were identified as paralogs within S. rassoulzadegani or Strombidinopsis sp. (Fig. 1D). Thus, our transcriptome data indicated only 13% reciprocal best hits pairs and 42% orthologous groups between the two ciliates.
Strombidium rassoulzadegani and Strombidinopsis sp. transcriptome annotation
About half of Strombidium rassoulzadegani and Strombidinopsis sp. transcripts matched previously known sequences. A total of 44% and 55% predicted AAS (equivalent to 70% and 85% Illumina reads) had significant BLASTP hits (NCBI non redundant NR database, E-value <10−6) for S. rassoulzadegani and Strombidinopsis sp., respectively (Fig. 2). In both cases, the maximum proportion of hits corresponded to Oxytricha trifallax. Only eight ciliate genomes have been sequenced so far , thus explaining the proportion of unknown sequences. However, this proportion is relatively low in comparison to that found for other non-model protist transcriptomes (e.g. 72% unknown sequences for a marine euglenoid ). From the low proportion of our AAS that matched to sequences from groups other than ciliates, most of them corresponded to lineages such as amoebozoa and opisthokonts. Less than 0.5% of hits corresponded to the same lineages as the food algae (prasinophytes or cryptophytes).
A- Proportion of amino acid sequences. B- Proportion of Illumina reads realigned to the corresponding coding regions.
From the AAS with significant BLASTP hits, 40% (S. rassoulzadegani) and 47% (Strombidinopsis sp.) had a confident assignment of Gene Ontology (GO) terms, which were retrieved mostly from the UniProt database. Complementation with InterProScan results increased annotations by 19% (S. rassoulzadegani) and 22% (Strombidinopsis sp.). GO terms distribution was similar between species (Fig. 3), with binding and catalytic activity as the main molecular functions, cellular and metabolic process as the main biological processes and nuclear-related structures as the main cellular components represented in both transcriptomes.
For each ciliate dataset, transcripts were included in 92 KEGG pathways, 81 of which were present in both species (Table S1). Most of these pathways corresponded to the metabolism of carbohydrates, lipids, amino acids, nucleotides, glycan, terpenoids, and vitamins and cofactors. In addition, some sequences were related to biosynthesis of secondary metabolites such as some antibiotics (e.g. streptomycin, neomycin) and degradation of xenobiotics such as some toxic aromatic hydrocarbons (e.g. xylene, toluene). A few of the results obtained by automatic annotations with Blast2GO were unexpected for protists. Some GO terms (e.g. ‘multicellular organismal process’ or ‘immune system process’; Fig. 3) may actually represent ancient eukaryotic genes with broader functions . Similar conclusions may apply for some KEGG pathways linked to both ciliate datasets (e.g. ‘peptidoglycan biosynthesis’; Table S1).
Transcripts potentially related to kleptoplasty in Strombidium rassoulzadegani
A kleptoplastic organism engulfs photosynthetic prey and digests all but their chloroplasts, which remain temporarily functional despite lacking control from the algal nucleus. Some kleptoplastic organisms express algal genes involved in chloroplast functioning and photosynthesis, likely integrated in the host nucleus by horizontal gene transfer , . We found transcripts linked to the GO term ‘plastid’ and the KEGG pathway ‘carbon fixation in photosynthetic organisms’ in the kleptoplastic Strombidium rassoulzadegani, but this was detected in the heterotrophic Strombidinopsis sp. as well (Fig. 3; Table S1). Specific searches of transcripts assigned to GO terms ‘plastid’ and ‘photosynthesis’ in both ciliates (Table S2) indicated that most of those transcripts 1) had a significant BLASTP hit with O. trifallax or other non-photosynthetic organisms and/or 2) are not clearly specific to plastids or photosynthesis according to their GO terms descriptions and associated KEGG pathways. Thus, most of these transcripts probably have more general functions. Apart from not providing data on chloroplast functioning in S. rassoulzadegani, some of these sequences may correspond to the <0.5% potential food transcripts not filtered from the ciliate data (see above), especially the few of them that had significant BLASTP hits with the same lineages as the algal prey (Table S2).
Alternative explanations for kleptoplasty include that retained chloroplasts are simply stable and thus remain functional for some time ,  or that they are transcriptionally active and can regulate themselves inside the host. The methods used in this study prevent us from discriminating if chloroplast genes from the food alga are expressed in S. rassoulzadegani. If this is the case, chloroplast genes expressed within the ciliate may have been removed by poly-A selection during library preparation and/or by filtering sequences that matched with algal transcripts in BLASTN searches. Similar to our results, transcriptome data on a kleptoplastic foraminiferan were also unable to provide information on genes potentially related to chloroplast functioning in the host . Thus, this approach may be insufficient to explain the mechanics of kleptoplasty in some organisms.
ROS detoxification in Strombidium rassoulzadegani and Strombidinopsis sp
Both under normal physiological conditions and as a response to oxidative stress, autotrophic and heterotrophic cells have enzymatic and non-enzymatic mechanisms to control ROS concentrations , , . We found evidence for these pathways in the transcriptomes of Strombidium rassoulzadegani and Strombidinopsis sp. (Fig. 4, Tables S3 and S4). In this case, transcripts were clearly linked to known antioxidant enzymes and most of them had highly significant BLASTP hits against ciliates or other non-photosynthetic organisms, thus minimizing the possibility that these genes belong to the food algae (Tables S3 and S4). Although we show simplified antioxidant pathways, these mechanisms are usually interrelated by reciprocal control and each of them has a higher activity in a certain cell compartment or against a certain oxidant, including ROS, lipid peroxides and reactive nitrogen species .
Pathways evidenced in both ciliates (white background) or only in Strombidium rassoulzadegani (yellow background) are shown. In red, ROS (O2- = superoxide radical; H2O2 = hydrogen peroxide); in dark blue, antioxidant enzymes (SOD = superoxide dismutase; Cat = catalase; APX = ascorbate peroxidase; GPX = glutathione peroxidase; TPX = thioredoxin peroxidase); in light blue, enzymes involved in metabolism of non-enzymatic antioxidants (Nec3 = bifunctional monodehydroascorbate reductase/carbonic anhydrase nectarin-3-like; TRed = thioredoxin reductase; GRed = glutathione reductase; GSyn = GSH synthetase; GCLig = Glutamate-cysteine ligase); in green, non-enzymatic antioxidants (AsA = ascorbic acid; GSH = glutathione; Trx = thioredoxin). MDAs = monodehydroascorbate; GSSG = glutathione disulfide; TrxS2 = thioredoxin disulfide; NAD(P)+ = nicotinamide adenine dinucleotide (phosphate). Based on literature , , .
Superoxide dismutase, catalase and peroxidases, the major enzymes that directly modulate ROS, were detected in the transcriptomes of both S. rassoulzadegani and Strombidinopsis sp. (Fig. 4, Table S3). Superoxide dismutase (SOD) reduces superoxide radicals to hydrogen peroxide and it exists in two forms in both ciliates: Cu/Zn SOD and Fe/Mn SOD (cytosolic and mitochondrial forms, respectively, in higher eukaryotes ). Catalase and peroxidases reduce hydrogen peroxide to water, the latter using non-enzymatic antioxidants as electron donors . Ascorbate peroxidase (APX), glutathione peroxidase (GPX) and thioredoxin peroxidase (TPX) oxidize ascorbic acid (AsA), glutathione (GSH) and thioredoxin (Trx), respectively, in order to reduce hydrogen peroxide , . Catalase, APX, GPX and TPX sequences were found in both ciliates. The multiple sequences detected for each enzyme may correspond to isoforms with different cell localizations, as known for example for APX and GPX in algae and plants . They may also correspond to mRNA precursors, which are usually difficult to distinguish in RNAseq data . An additional cause for these multiple sequences may be inability to condense some similar transcripts during the assembly.
Among non-enzymatic antioxidants, GSH is a tri-peptide that can be synthetized and recycled in both ciliates, according to the transcriptome data (Fig. 4, Table S4). Similarly, transcripts for the protein Trx and its recycling enzyme were found in both species. In contrast, transcripts related to the synthesis of AsA or other antioxidants found in plants and algae, such as carotenoids and tocopherols , were not detected in the ciliates.
Interestingly, we found evidence for a group of enzymes that can recycle AsA (nectarin-3-like enzymes, Nec3) in S. rassoulzadegani but not in Strombidinopsis sp. (Fig. 4, Table S4). Nec3 has monodehydroascorbate reductase (MDAR) activity, i.e. it transforms the oxidized form of AsA back to its reduced form, thus providing the advantage of keeping constant levels of AsA without the necessity of a constant supply –. Although AsA can also be recycled spontaneously or through a cycle that involves GSH (the AsA-GSH cycle), it is more rapidly regenerated by MDAR activity . Therefore, Nec3 may help S. rassoulzadegani to maintain high pools of AsA, which is important both as anti-oxidant and as cofactor and regulator during photosynthesis . Phylogenetic inferences showed that Nec3 from S. rassoulzadegani clustered with sequences from other non-photosynthetic organisms (the ciliate Oxytricha trifallax, one fungus and two animals) and formed a group apart from those of plants, although there are no sequences available for algae (Fig. S1). These preliminary results suggest that 1) Nec3 belongs to the ciliate and not the food alga, and 2) there is no evolutionary link between S. rassoulzadegani and photosynthetic organisms regarding Nec3. In this context, clarifying the origin and role of Nec3 in S. rassoulzadegani and its potential role in mixotrophy deserves further experimentation.
ROS detoxification occurs in several parts of the eukaryotic cell. In both heterotrophic and autotrophic cells these mechanisms act in the cytosol, in mitochondria and, in many eukaryotes, also in peroxisomes. Peroxisome presence has a patchy distribution among ciliates and other protist taxa  and they have not been observed in oligotrichs or choreotrichs to our knowledge, but we detected sequences related to this organelle in our data (Fig. 3). In autotrophs, enzymes such as SOD, APX and GPX scavenge ROS also in the chloroplast and they are key in order to keep this organelle active , , . However, these enzymes are nuclear encoded ,  and, even if they initially exist in kleptochloroplasts of S. rassoulzadegani, their activity is very likely lost after some time (Fig. 4). Inactivation of chloroplast antioxidant enzymes is known to limit photosynthetic efficiency . Thus, oxidative damage may contribute to the lack of kleptochloroplast functionality and the fact that a continuous supply of fresh chloroplasts is needed for the growth of S. rassoulzadegani .
We used RNAseq to characterize the transcriptomes of two non-model microbial eukaryotes. This approach provided information about genes with known functions as well as multiple potentially novel genes. We experienced the typical challenges of studying non-model organisms and using automatic annotation tools that do not detect the whole spectrum of protist physiological features. Limitations such as unknown levels of genome coverage, high proportion of sequences not similar to those available in databases, and annotations not compatible with protist biology have been common in this kind of study so far. Additional effort was required for sequencing and filtering food transcripts, given that the ciliates under study cannot be cultured independent of their prey.
The transcriptomes of Strombidium rassoulzadegani and Strombidinopsis sp. provide baselines for analyzing ciliate metabolism, ecological roles in the planktonic food web and relationships with the environment. Our observations are noteworthy in two ways. First, we analyzed the first transcriptomes from oligotrichs and choreotrichs, which are the most diverse and abundant ciliates in marine plankton. Second, the species we chose practice two contrasting nutritional modes, heterotrophy and mixotrophy, and hence have somewhat different ecological roles. Although the transcriptomes differed in general features such as GC content distribution and had a homology lower than 50%, they provided similar annotations for GO terms and KEGG pathways, which were related mostly to housekeeping activity. As more ciliate reference genomes become available, we expect that more pathways, including novel ones, will be revealed in the data.
Transcriptome information alone provided limited insights on genes related to mixotrophy. We did not find transcripts clearly related to the maintenance and functioning of retained chloroplasts in S. rassoulzadegani and we identified very similar antioxidant mechanisms in both mixotrophic and heterotrophic ciliates. The relevance of one enzyme potentially related to ascorbate recycling in the mixotroph as well as the potential differences in regulation and expression levels of all the identified genes require future experimentation in order to understand the implications of antioxidant pathways for physiology and evolution of mixotrophs.
Maximum Likelihood tree inferred from Nec3 sequences. Nodes with support higher than 50% are indicated with a black (Maximum Likelihood) and/or a grey circle (Neighbor Joining). The scale bar represents ten substitutions per 100 nucleotides.
KEGG pathways identified in Strombidium rassoulzadegani and Strombidinopsis sp.
Results of specific searches for GO terms related to plastids and photosynthesis in Strombidium rassoulzadegani and Strombidinopsis sp. transcriptomes.
Transcripts of antioxidant enzymes detected in Strombidium rassoulzadegani and Strombidinopsis sp.
Transcripts related to metabolism of non-enzymatic antioxidants in Strombidium rassoulzadegani and Strombidinopsis sp.
Samples MMETSP0449, MMETSP0463, MMETSP0484 and MMETSP0491 were sequenced and assembled at the National Center for Genome Resources (NCGR). We are grateful to NCGR staff, including project manager Callum J. Bell, former manager Arvind K. Bharti, Connor T. Cameron for BPA pipeline design, Kelly B. Schilling for bioinformatic assistance, and Peter B. Ngam, Jennifer L. Jacobi, and Pooja E. Umale for laboratory work and sequencing. Gary Wikfors of the National Marine Fisheries Service laboratory in Milford CT provided axenic cultures of algae. We acknowledge the editor and three anonymous reviewers for useful comments.
Conceived and designed the experiments: GBM HZ. Performed the experiments: GBM HZ. Analyzed the data: LFS SG. Contributed reagents/materials/analysis tools: GBM HZ. Contributed to the writing of the manuscript: LFS SG HZ GBM.
- 1. Lynn DH (2008) The Ciliated Protozoa. Characterization, classification, and guide to the literature. Dordrecht: Springer Verlag. 606 p.
- 2. Agatha S (2011) Global Diversity of Aloricate Oligotrichea (Protista, Ciliophora, Spirotricha) in Marine and Brackish Sea Water. PLoS ONE 6: e22466.
- 3. Sherr E, Sherr B (2002) Significance of predation by protists in aquatic microbial food webs. Antonie van Leeuwenhoek 81: 293–308.
- 4. Strom SL, Macri EL, Olson MB (2007) Microzooplankton grazing in the coastal Gulf of Alaska:Variations in top-down control of phytoplankton. Limnol Oceanogr 52: 1480–1494.
- 5. Santoferrara L, Gómez MI, Alder V (2011) Bathymetric, latitudinal and vertical distribution of protozooplankton in a cold-temperate shelf (southern Patagonian waters) during winter. J Plankton Res 33: 457–468.
- 6. Pierce RW, Turner JT (1992) Ecology of planktonic ciliates in marine food webs. Reviews in Aquatic Sciences 6: 139–181.
- 7. Laval-Peuto M, Rassoulzadegan F (1988) Autofluorescence of marine planktonic Oligotrichina and other ciliates. Hydrobiologia 159: 99–110.
- 8. Stoecker DK, Johnson MD, De Vargas C, Not F (2009) Acquired phototrophy in aquatic protists. Aquat Microb Ecol 57: 279–310.
- 9. Esteban GF, Fenchel T, Finlay BJ (2010) Mixotrophy in ciliates. Protist 161: 621–641.
- 10. Johnson MD (2011) Acquired phototrophy in ciliates: A review of cellular interactions and structural adaptation. J Eukar Microbiol 58: 185–195.
- 11. Stoecker DK (1998) Conceptual models of mixotrophy in planktonic protists and some ecological and evolutionary implications. Europ J Protistol 34: 281–290.
- 12. Flynn KJ, Stoecker DK, Mitra A, Raven JA, Glibert PM, et al. (2013) Misuse of the phytoplankton–zooplankton dichotomy: the need to assign organisms as mixotrophs within plankton functional types. J Plankton Res 35: 3–11.
- 13. McManus GB, Zhang H, Lin S (2004) Marine planktonic ciliates that prey on macroalgae and enslave their chloroplasts. Limnol Oceanogr 49: 308–313.
- 14. McManus GB, Xu D, Costas BA, Katz LA (2010) Genetic identities of cryptic species in the Strombidium stylifer/apolatum/oculatum cluster, including a description of Strombidium rassoulzadegani n. sp. J Eukar Microbiol 57: 369–378.
- 15. Mcmanus GB, Schoener D, Haberlandt K (2012) Chloroplast symbiosis in a marine ciliate: ecophysiology and the risks and rewards of hosting foreign organelles. Front Microbiol 3: 321
- 16. Schoener DM, McManus GB (2012) Plastid retention, use, and replacement in a kleptoplastidic ciliate. Aquat Microb Ecol 67: 177–187.
- 17. Johnson MD (2011) The acquisition of phototrophy: adaptive strategies of hosting endosymbionts and organelles. Photosynth Res 107: 117–32.
- 18. Minnhagen S, Carvalho WF, Salomon PS, Janson S (2008) Chloroplast DNA content in Dinophysis (Dinophyceae) from different cell cycle stages is consistent with kleptoplasty. Environ Microbiol 10: 2411–7.
- 19. Clark KB, Jensen KR, Strits HM (1990) Survey of functional kleptoplasty among West Atlantic Ascoglossa ( = Sacoglossa) (Mollusca: Opistobranchia). The Veliger 33: 339–345.
- 20. Johnson MD, Oldach D, Delwiche CF, Stoecker DK (2007) Retention of transcriptionally active cryptophyte nuclei by the ciliate Myrionecta rubra. Nature 445: 426–428.
- 21. Wisecaver J, Hackett J (2010) Transcriptome analysis reveals nuclear-encoded proteins for the maintenance of temporary plastids in the dinoflagellate Dinophysis acuminata. BMC Genomics 11: 366.
- 22. Pierce SK, Fang X, Schwartz JA, Jiang X, Zhao W, et al. (2012) Transcriptomic evidence for the expression of horizontally transferred algal nuclear genes in the photosynthetic sea slug, Elysia chlorotica. Mol Biol Evol 29: 1545–56.
- 23. Pillet L (2013) The role of horizontal gene transfer in kleptoplastidy and the establishment of photosynthesis in the eukaryotes. Mobile Genetic Elements 3: e24773.
- 24. Pillet L, Pawlowski J (2013) Transcriptome analysis of foraminiferan Elphidium margaritaceum questions the role of gene transfer in kleptoplastidy. Mol Biol Evol 30: 66–69.
- 25. Lesser MP (2006) Oxidative stress in marine environments: Biochemistry and physiological ecology. Annu Rev Physiol 68: 253–278.
- 26. Vonlaufen N, Kanzok SM, Wek RC, Sullivan Jr WJ (2008) Stress response pathways in protozoan parasites. Cell Microbiol 10: 2387–2399.
- 27. Apel K, Hirt H (2004) Reactive oxygen species: Metabolism, oxidative stress, and signal transduction. Annu Rev Plant Biol 55: 373–399.
- 28. Asada K (2006) Production and scavenging of reactive oxygen species in chloroplasts and their functions. Plant Physiol 141: 391–396.
- 29. Foyer CH, Shigeoka S (2011) Understanding oxidative stress and antioxidant functions to enhance photosynthesis. Plant Physiol 155: 93–100.
- 30. Kawano T, Kadono T, Kosaka T, Hosoya H (2004) Green paramecia as an evolutionary winner of oxidative symbiosis: a hypothesis and supportive data. Z Naturforsch C 59: 538–542.
- 31. Ohkawa H, Hashimoto N, Furukawa S, Kadono T, Kawano T (2011) Forced symbiosis between synechocystis spp. PCC 6803 and apo-symbiotic Paramecium bursaria as an experimental model for evolutionary emergence of primitive photosynthetic eukaryotes. Plant Signal Behav 6: 773–776.
- 32. Flores HS, Wikfors GH, Dam HG (2012) Reactive oxygen species are linked to the toxicity of the dinoflagellate Alexandrium spp. to protists. Aquat Microb Ecol 66: 199–209.
- 33. Pérez MT, Dolan JR, Fukai E (1997) Planktonic oligotrich ciliates in the NW Mediterranean: Growth rates and consuption by copepods. Mar Ecol Prog Ser 155: 89–101.
- 34. Keeling PJ, Burki F, Allam B, Allen E, Armbrust G, et al. (2014) MMETSP: Illuminating the functional diversity of life in the oceans through transcriptome sequencing. PLoS Biology, in press.
- 35. Guillard RR, Ryther JH (1962) Studies of marine planktonic diatoms. I. Cyclotella nana Hustedt, and Detonula confervacea (Cleve) Gran. Can J Microbiol 8: 229–239.
- 36. Zhang H, Finiguerra M, Dam HG, Huang Y, Xu D, et al. (2013) An improved method for achieving high-quality RNA for copepod gene transcriptomic studies. J Exp Mar Biol Ecol 446: 57–66.
- 37. Simpson JT, Durbin R (2012) Efficient de novo assembly of large genomes using compressed data structures. Genome Res 22: 549–556.
- 38. Simpson JT, Wong K, Jackman SD, Schein JE, Jones SJ, et al. (2009) ABySS: a parallel assembler for short read sequence data. Genome Res 19: 1117–1123.
- 39. Li R, Li Y, Kristiansen K, Wang J (2008) SOAP: Short oligonucleotide alignment program. Bioinformatics 25: 713–714.
- 40. Chevreux B, Pfisterer T, Drescher B, Driesel AJ, Müller WE, et al. (2004) Using the miraEST assembler for reliable and automated mRNA transcript assembly and SNP detection in sequenced ESTs. Genome Res 14: 1147–1159.
- 41. Li H, Durbin R (2009) Fast and accurate short read alignment with Burrows-Wheeler Transform. Bioinformatics 25: 1754–1760.
- 42. Sun S, Chen J, Li W, Altintas I, Lin A, et al. (2011) Community cyberinfrastructure for Advanced Microbial Ecology Research and Analysis: the CAMERA resource. Nucleic Acids Res 39: D546–551.
- 43. Iseli C, Jongeneel CV, Bucher P (1999) ESTScan: a program for detecting, evaluating, and reconstructing potential coding regions in EST sequences. Proc Int Conf Intell Syst Mol Biol 138–148.
- 44. Lottaz C, Iseli C, Jongeneel CV, Bucher P (2003) Modeling sequencing errors by combining Hidden Markov models. Bioinformatics 19: ii103–ii112.
- 45. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tool. J Molec Biol 215: 403–410.
- 46. Li L, Stoeckert CJ, Roos DS (2003) OrthoMCL: Identification of ortholog groups for eukaryotic genomes. Genome Res 13: 2178–2189.
- 47. Conesa A, Götz S, Garcia-Gomez JM, Terol J, Talon M, et al. (2005) Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics 21: 3674–3676.
- 48. Zhang Z, Wood WI (2003) A profile hidden Markov model for signal peptides generated by HMMER. Bioinformatics 19: 307–308.
- 49. Finn RD, Tate J, Mistry J, Coggill PC, Sammut SJ, et al. (2010) The Pfam protein families database. Nucleic Acids Res 38: D211–222.
- 50. Haft DH, Loftus BJ, Richardson DL, Yang F, Eisen JA, et al. (2001) TIGRFAMs: a protein family resource for the functional identification of proteins. Nucl Acids Res 29: 41–43.
- 51. Gough J, Karplus K, Hughey R, Chothia C (2001) Assignment of homology to genome sequences using a library of hidden Markov models that represent all proteins of known structure. J Molec Biol 313: 903–919.
- 52. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32: 1792–1797.
- 53. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, et al. (2011) MEGA5: Molecular Evolutionary Genetics Analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 2731–2739.
- 54. Stamatakis A (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22: 2688–2690.
- 55. Abascal F, Zardoya R, Posada D (2005) ProtTest: selection of best-fit models of protein evolution. Bioinformatics 21: 2104–2105.
- 56. Swart EC, Bracht JR, Magrini V, Minx P, Chen X, et al. (2013) The Oxytricha trifallax macronuclear genome: A complex eukaryotic genome with 16,000 tiny chromosomes. PLoS Biology 11: e1001473.
- 57. Ellegren H (2014) Genome sequencing and population genomics in non-model organisms. Trends Ecol Evol 29: 51–63.
- 58. Kuo RC, Zhang H, Zhuang Y, Hannick L, Lin S (2013) Transcriptomic study reveals widespread spliced leader trans-splicing, short 5′-UTRs and potential complex carbon fixation mechanisms in the euglenoid alga Eutreptiella sp. PLoS ONE 8: e60826.
- 59. Grant JR, Lahr DJG, Rey FE, Burleigh JG, Gordon JI, et al. (2012) Gene discovery from a pilot study of the transcriptomes from three diverse microbial eukaryotes: Corallomyxa tenera, Chilodonella uncinata, and Subulatomonas tetraspora. Protist Genomics 1: 3–18.
- 60. Bowler C, Montagu MV, Inze D (1992) Superoxide dismutase and stress tolerance. Annu Rev Plant Physiol Plant Mol Biol 43: 83–116.
- 61. Müller S, Liebau E, Walter RD, Krauth-Siegel RL (2003) Thiol-based redox metabolism of protozoan parasites. Trends Parasitol 19: 320–328.
- 62. Krauth-Siegel RL, Leroux AE (2012) Low-molecular-mass antioxidants in parasites. Antioxid Redox Signal 17: 583–607.
- 63. Shigeoka S, Ishikawa T, Tamoi M, Miyagawa Y, Takeda T, et al. (2002) Regulation and function of ascorbate peroxidase isoenzymes. J Exp Bot 53: 1305–1319.
- 64. Garber M, Grabherr MG, Guttman M, Trapnell C (2011) Computational methods for transcriptome annotation and quantification using RNA-seq. Nat Methods 8: 469–477.
- 65. Di Mascio P, Murphy ME, Sies H (1991) Antioxidant defense systems: the role of carotenoids, tocopherols, and thiols. Am J Clin Nutr 53: 194S–200S.
- 66. Carter C, Thornburg R (2004) Tobacco Nectarin III is a bifunctional enzyme with monodehydroascorbate reductase and carbonic anhydrase activities. Plant Mol Biol 54: 415–425.
- 67. Murthy SS, Zilinskas BA (1994) Molecular cloning and characterization of a cDNA encoding pea monodehydroascorbate reductase. J Biol Chem 269: 31129–31133.
- 68. Leterrier M, Corpas FJ, Barroso JB, Sandalio LM, del Río LA (2005) Peroxisomal monodehydroascorbate reductase. Genomic clone characterization and functional analysis under environmental stress C\conditions. Plant Physiol 138: 2111–2123.
- 69. Lunde C, Baumann U, Shirley N, Drew D, Fincher G (2006) Gene structure and expression pattern analysis of three monodehydroascorbate reductase (Mdhar) genes in Physcomitrella patens: Implications for the evolution of the MDHAR family in plants. Plant Mol Biol 60: 259–275.
- 70. Hossain MA, Teixeira da Silva JA, Fujita M (2011) Glyoxalase system and reactive oxygen species detoxification system in plant abiotic stress response and tolerance: An intimate relationship. In: Arun Shanke, editor. Abiotic stress in plants - Mechanisms and Adaptations. InTech. Available: http://wwwintechopencom/books/abiotic-stress-in-plants-mechanismsand-adaptations/glyoxalase-system-and-reactive-oxygen-species-detoxification-system-in-plant-abiotic-stressresponse. Accessed 10 January 2014.
- 71. Gabaldón T (2010) Peroxisome diversity and evolution. Philos Trans R Soc Lond B Biol Sci 365: 765–773.
- 72. Takeda T, Ishikawa T, Shigeoka S (1997) Metabolism of hydrogen peroxide by the scavenging system in Chlamydomonas reinhardtii. Physiol Plant 99: 49–55.
- 73. Yoon H-S, Lee H, Lee I-A, Kim K-Y, Jo J (2004) Molecular cloning of the monodehydroascorbate reductase gene from Brassica campestris and analysis of its mRNA level in response to oxidative stress. Biochim Biophys Acta - Bioenergetics 1658: 181–186.
- 74. Ishikawa T, Shigeoka S (2008) Recent advances in ascorbate biosynthesis and the physiological significance of ascorbate peroxidase in photosynthesizing organisms. Biosci, Biotechnol Biochem 72: 1143–1154.
- 75. Rhee SG, Kang SW, Chang T-S, Jeong W, Kim K (2001) Peroxiredoxin, a novel family of peroxidases. IUBMB Life 52: 35–41.