Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Durkin, C. A. Articles by Armbrust, E. V. Search for Related Content PubMed PubMed Citation Articles by Durkin, C. A. Articles by Armbrust, E. V.

 Previous Article  |  Next Article 

Eukaryotic Cell, July 2009, p. 1038-1050, Vol. 8, No. 7
1535-9778/09/$08.00+0     doi:10.1128/EC.00079-09
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Chitin in Diatoms and Its Association with the Cell Wall ^,

Colleen A. Durkin,¹ Thomas Mock,² and E. Virginia Armbrust^1*

School of Oceanography, University of Washington, Box 357940, Seattle, Washington 98195,¹ School of Environmental Sciences, University of East Anglia, Norwich NR4 7TJ, United Kingdom²

Received 8 March 2009/ Accepted 24 April 2009

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Chitin is a globally abundant polymer widely distributed throughout eukaryotes that has been well characterized in only a few lineages. Diatoms are members of the eukaryotic lineage of stramenopiles. Of the hundreds of diatom genera, two produce long fibers of chitin that extrude through their cell walls of silica. We identify and describe here genes encoding putative chitin synthases in a variety of additional diatom genera, indicating that the ability to produce chitin is more widespread and likely plays a more central role in diatom biology than previously considered. Diatom chitin synthases fall into four phylogenetic clades. Protein domain predictions and differential gene expression patterns provide evidence that chitin synthases have multiple functions within a diatom cell. Thalassiosira pseudonana possesses six genes encoding three types of chitin synthases. Transcript abundance of the gene encoding one of these chitin synthase types increases when cells resume division after short-term silicic acid starvation and during short-term limitation by silicic acid or iron, two nutrient conditions connected in the environment and known to affect the cell wall. During long-term silicic acid starvation transcript abundance of this gene and one additional chitin synthase gene increased at the same time a chitin-binding lectin localized to the girdle band region of the cell wall. Together, these results suggest that the ability to produce chitin is more widespread in diatoms than previously thought and that a subset of the chitin produced by diatoms is associated with the cell wall.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Chitin is the most abundant polymer in the ocean and the second most abundant polymer on earth, exceeded only by cellulose (see, for example, references 1, 16, and 27). The pervasiveness of chitin is attributed to its use by diverse eukaryotic organisms, including fungi, insects, mollusks, crustaceans, algae, and protists. These diverged lineages commonly use chitin to strengthen their cell walls or skeletons. Chitin synthase is used by all chitin-producing organisms to generate the β-(14)-linked N-acetylglucosamine polymers. The universality of this enzyme suggests an ancient eukaryotic origin (54). Chitin synthases are identified by the conserved amino acid motifs QXXEY, EDR, QXRRW, and additional lineage specific motifs that together are responsible for substrate specificity and reaction catalysis (38, 54). Chitin synthases are embedded across the cell membrane, and it is at the membrane junction that N-acetylglucosamine monomers are added to a growing polymer while it is transported across the membrane (55). While maintaining these general features, chitin synthases have also diverged and diversified within and between eukaryotic lineages. These enzymes have been most extensively studied within fungi where sequence phylogenies identify two major divisions of chitin synthases, with additional classes within these divisions. Phylogenetically distinct chitin synthases produce chitin at different cellular locations and for different biological functions (52). Thus, an individual organism often contains multiple types of chitin synthases.

Diatoms are known chitin producers that are of significant ecological importance. Globally, these unicellular algae are responsible for ca. 20% of annual primary production (13, 45) and support the most biologically productive regions of the ocean. Evolutionarily distinct from plants and metazoans, diatoms are members of the stramenopiles, a major eukaryotic lineage that includes plant pathogens, flagellates, and brown macroalgae. The most distinctive characteristic of diatoms is the nanopatterned silica cell wall, which is composed of two halves (thecae) connected by hooplike silica structures called girdle bands. Diatom cell wall formation and eventual dissolution controls the biogeochemical cycling of silicon in the ocean, and because it is required for growth, silicon availability affects carbon and other nutrient cycles (8). The formation of these intricate inorganic nanostructures has been studied intensively due to potential applications in nanotechnology that require controlled synthesis of microscopic structures (4, 47). These topics have motivated more detailed analysis of the biological mechanisms controlling cell wall formation (24) and molecules associated with the silica (14, 30, 31, 62). The centric diatom Thalassiosira pseudonana has emerged as a powerful model for understanding silica biomineralization because of the availability of the whole genome sequence (2).

Thalassiosira and Cyclotella are the two diatom genera known to produce chitin (34). Both are members of the order Thalassiosirales, which falls within the multipolar diatoms (class Mediophyceae). Interestingly, this class is evolutionarily intermediate to the more ancient centric diatoms (class Coscinodiscophyceae) and the more derived pennate diatoms (class Bacillariophyceae) (36) (see inset in Fig. 1). Thalassiosira and Cyclotella both produce long thin chitin fibers that extend from the theca through specialized pores within the silica known as fultoportulae (21, 28, 53). The chitin fibers are arranged in an unusual highly crystalline beta configuration, in which all polymers lie parallel to one another and are not associated with other molecules (9, 35, 61). Specialized vesicles targeted to an invaginated membrane just below the fultoportulae have been identified and were proposed to carry the molecular and chemical machinery necessary for chitin synthesis, which likely includes the chitin synthases (20). In T. weissflogii, chitin synthases are estimated to generate 700,000 N-acetylglucosamine linkages s^–1 µm of specialized membrane^–2 (20).

View larger version (32K): [in this window] [in a new window]

FIG. 1. Maximum-likelihood phylogenetic tree of translated diatom chitin synthase gene sequences. Sequences amplified by CODEHOP primers are followed by numerical identifiers 1 through 10, while sequences identified from genomes are followed by their protein ID number. The outgroup S. cerevisiae chitin synthase 3 (NCBI accession no. P29465 [UniProtKB/Swiss-Prot] ) was used to root the tree. Bootstrap values are indicated at nodes. A scale bar indicates the number of amino acid changes. The inset shows a simplified phylogeny of diatom evolution adapted from data from Medlin and Kaczmarska (36), and genera listed in the inset correspond to those present in the gene tree.

Analysis of the whole-genome sequences for T. pseudonana (2) and the pennate diatom Phaeodactylum tricornutum (3) suggests a previously unsuspected and complex role for chitin-related processes in diatoms. All genes necessary for chitin synthesis were found in T. pseudonana, as well as P. tricornutum, despite the fact that P. tricornutum does not produce chitin fibers. Both diatoms appear to encode multiple chitin synthases; moreover, T. pseudonana possesses more than 20 genes that encode putative chitin-degrading chitinases and more than 20 genes that encode putative proteins with chitin-binding domains. Two of the putative chitin-binding proteins in T. pseudonana localize to the girdle band region of the cell wall, a region not previously suspected to contain chitin (6). The genes encoding these chitin-binding proteins, named p150 and p150-like, are highly expressed when cells are grown in the presence of high concentrations of copper or when cells are starved of silicic acid or iron, conditions that all result in abnormally elongated theca (6, 40). We recently identified 84 genes upregulated when cells were deprived of iron or silicic acid; many of these coregulated genes encode proteins known to be involved in cell wall processes (40). Surprisingly, a gene encoding a chitin synthase was among this subset of upregulated genes, suggesting that chitin may play a role in diatom cell wall processes. Additional support for this possibility comes from a recent analysis of the organic composition of T. pseudonana cell walls using solid-state nuclear magnetic resonance (62). That study identified chitin as the main carbohydrate component embedded within the silica cell walls, although the authors attributed this presence to contaminating chitin fibers.

The genetic complexity of chitin-related genes in T. pseudonana, the identification of potential chitin synthases in P. tricornutum, and the unexpected expression pattern of one chitin synthase gene in T. pseudonana suggests a more complex role of chitin synthesis in diatoms than was previously thought. We describe here our characterization of diatom chitin synthase genes through comparative genetics and physiological experimentation in order to better understand the evolution, function, and ecological consequences of chitin synthesis in diatoms.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In silico analyses. Putative chitin synthase genes were identified in the T. pseudonana and P. tricornutum genomes by combining automated gene predictions (www.doe.jgi.gov) and homology of predicted proteins to known chitin synthases in the NCBI nonredundant database based on nBLAST and pBLAST tools (ncbi.nlm.nih.gov/blast/blast.cgi). BLAST identification of chitin synthases was considered significant when the e values were 10^–15 and when a majority of identified homologs were annotated as chitin synthases. Domain structures of chitin synthases were analyzed by Interpro (ebi.ac.uk/InterProScan/), Prodom (prodom.prabi.fr/prodom/current/html/form.php), the NCBI Conserved Domain Database (ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi), and TMHMM (cbs.dtu.dk/services/TMHMM-2.0/). Domain predictions were considered significant when the e values were 10^–5.

Phytoplankton cultures and experimental setup. An isolate of T. punctigera and T. rotula was provided by P. von Dassow (64) and an isolate of Ditylum brightwellii was provided by J. Koester; both isolates were collected from Puget Sound, WA. An isolate of Pseudo-nitzschia multiseries from the North Atlantic was provided by S. Bates and C. Leger. All other phytoplankton isolates (Table 1) were purchased from the Provasoli-Guillard National Center for Culture of Marine Phytoplankton (CCMP [ccmp.bigelow.org]). Cultures were maintained exponentially at 100 to 200 µmol of photons m^–2 s^–1 in 250 to 500 ml of f/2-amended seawater (17, 18) that had been filtered through a 0.45-µm-pore-size polycarbonate filter (Millipore).

View this table: [in this window] [in a new window]

TABLE 1. Organisms tested by PCR for the presence of chitin synthase genes and identification of those in which chitin synthase gene fragments were successfully amplified

Silicic acid starvation and recovery were monitored on both short and long time scales. Short-term recovery from silicic acid starvation was analyzed by first growing triplicate cultures in Aquil medium (50) under continuous light at 100 µmol of photons m^–2 s^–1 and constant bubbling with sterile air until the middle of the exponential growth phase. In a manner similar to those described by Hildebrand et al. (22), cells were harvested by centrifugation at 4,500 x g for 15 min, washed once with, and inoculated into Aquil medium without silicic acid at a final concentration of 0.6 x 10⁶ cells ml^–1. At 48 h, silicic acid was added to the cultures at a final concentration of 200 µM, and the cultures were monitored for an additional 8 h as they recovered from silicic acid starvation and resumed cell division. Long-term responses to silicic acid starvation were monitored in triplicate cultures transferred from exponential growth in fully amended f medium (17) into f medium with low silicic acid (0.067 µM) at 175 to 275 µmol of photons m^–2 s^–1. Changes in chlorophyll a fluorescence over time were measured with a fluorometer (Turner), and cell concentrations were determined with an InFlux cell sorter flow cytometer (Cytopeia) equipped with a 488-nm laser. An aliquot of 2-µm fluorescent bead stock (Polysciences) of known concentration was added to 1 ml of culture, and the particles were counted together. Photosynthetic capacity (Fv/Fm) was determined with a PhytoPAM fluorometer (Walz). To confirm that cell division had ceased due to silicic acid starvation, a final concentration of 214 µM silicic acid was added to a 50-ml aliquot of each triplicate culture, and the relative fluorescence was monitored for a day. Aliquots of the culture media on day 4 were also filtered through a 0.2-µm-pore-size filter and analyzed for dissolved silicic acid content by using the molybdate method (5). A pulse of silicic acid similar to the short-term experiment (214 µM) was added to the silicic acid-starved cultures after 6 days in stationary phase, and the response of the cells was monitored 24 and 48 h later. Samples were collected from both types of experiments for quantitative reverse transcriptase PCR (qRT-PCR) by filtering between 200 and 1,000 ml of culture onto four 2-µm-pore-size polycarbonate filters (Millipore) (see below). Filtered cells were also collected from previously described experiments (40) with nutrient-replete cultures and short-term (<1 day) silicic acid-, iron-, and nitrate-limited cultures.

DNA and RNA extraction and cDNA synthesis. DNA was extracted with a plant DNA extraction kit (Qiagen) from exponentially maintained cultures (Table 1) that had been concentrated and frozen onto 0.45-µm-pore-size polycarbonate filters (Millipore). RNA was extracted with the plant RNA isolation reagent (Invitrogen) from frozen cells on 2-µm-pore-size polycarbonate filters. Contaminating DNA was eliminated by incubating the isolated RNA with DNase I (Ambion) at 37°C for 1 to 2 h; DNA-free RNA was purified with an RNeasy MiniElute cleanup kit (Qiagen). cDNA was synthesized from 2 µg of purified RNA with a Superscript III first-strand synthesis system for RT-PCR (Invitrogen). The 20 µl of cDNA was subsequently diluted with water to 100 µl. One microliter of RNA was used in PCRs (see below) to ensure that all contaminating DNA had been eliminated.

Determination of full-length gene sequences. Full-length sequence of the T. pseudonana chitin synthase cDNAs was determined by DNA sequencing of PCR-generated fragments using a total of 40 primers (not listed) designed across the entire length of the modeled gene sequences (www.doe.jgi.gov), and open reading frames that appeared to extend up or downstream of gene models. PCRs consisted of 1 µl of cDNA, 1 µl of 10x Taq buffer, 3.125 mM MgCl₂, 0.7 U of Taq (Promega and Gene Choice), 0.4 mM deoxynucleoside triphosphates, and 0.5 µM concentrations of forward and reverse primers. Amplifications consisted of a denaturation at 94°C for 2 min, followed by 30 cycles of 94°C for 10 s, treatment at between 55 and 62°C for 30 s, depending on annealing temperature of the primer set, and then 72°C for 90 s. Amplified fragments were separated on an agarose gel, and fragments of the appropriate size were cut out of the gel and extracted with the QiaQuick gel extraction kit (Qiagen). Between 50 and 100 ng of purified PCR product was cycle sequenced with DYEnamic ET dye terminator cycle sequencing kit (GE Healthcare Biosciences Corp.) and analyzed on a MegaBACE 1000 (GE Healthcare Biosciences Corp.). Rapid amplification of cDNA ends (RACE) was performed with First Choice RLM-RACE kit (Ambion) to identify the 5' and 3' ends of the gene encoding protein 7305. The RACE primers are listed in Table 2. All sequences were analyzed and aligned in Sequencher version 4.6 and compared to the T. pseudonana genome with BLAST.

View this table: [in this window] [in a new window]

TABLE 2. Oligonucleotide primers used for amplifying chitin synthase gene fragments in phytoplankton and cDNA fragments from various genes in T. pseudonana

Gene expression. Transcript abundances in T. pseudonana were determined by qRT-PCR. Primers (Table 2) were designed to amplify a 100- to 250-bp fragment for four reference genes encoding the actin-like protein (protein identification [protein ID] 269504), actin (protein ID 25772), 40S rRNA (protein ID 31084), and beta tubulin (protein ID 31569). Primers were also designed to amplify a 100-bp fragment from genes encoding three chitin synthases (protein IDs 7305, 6575, and 4368) and two genes encoding girdle band-associated chitin-binding proteins p150 and p150-like (protein IDs 12594 and 26041). Triplicate qRT-PCRs included 2 µl of cDNA; 0.8 µM concentrations of forward and reverse primers; and 10, 20, or 25 µl of iQ Supermix (Bio-Rad, Richmond, CA) in a 20-, 40-, or 50-µl reaction volume. Amplifications were carried out in an iCycler (Bio-Rad) and consisted of an initial denaturation at 95°C for 3 min; followed by 45 cycles of 95°C for 10 s, 60°C for 30 s, and 72°C for 50 s; followed by a stepwise increase in melting temperature to verify the presence of a single melt peak fluorescent signal. Threshold cycle (C_T) values were calculated with iCycler software, and the PCR amplification curves were imported into LinReg (51) to calculate the amplification efficiency. Efficiency calculations were based on at least four points, and only values with an R of 0.998 and an efficiency between 1.7 and 2.2 were considered. Efficiencies calculated from all qRT-PCRs for a single primer set were averaged to yield a single efficiency for use in expression calculations. These averages were between 1.92 and 1.98 determined from between 34 and 187 separate PCRs, depending on the gene. Relative transcript abundance was calculated by the equation E_ref^CTref/E_target^CTtarget, where E_ref is the reference gene efficiency, E_target is the target gene efficiency, C_Tref is the C_T of the reference gene, and C_Ttarget is the C_T of the target gene (49). Relative transcript abundances measured in triplicate cultures entering silicic acid, iron, or nitrate limitation were compared to the control condition by using a Student t test. Relative transcript abundances in triplicate cultures in the hours after silicic acid replenishment or the days of growth in silicic acid-depleted media were log transformed, and time points were compared by using an analysis of variance test ( < 0.05) in SPSS version 16. Time points with significant differences were identified by using a Tukey test (66).

Phylogenetic analysis. CLUSTAL W (63) was used to align translated T. pseudonana chitin synthase genes with fungal chitin synthase protein sequences (GenBank numbers O13353 [UniProtKB/Swiss-Prot] , P29465 [UniProtKB/Swiss-Prot] , P30573 [GenBank] , Q01285 [UniProtKB/Swiss-Prot] , P78611 [UniProtKB/Swiss-Prot] , and O13394 [UniProtKB/Swiss-Prot] ) and PCR primers (Table 2) were designed by using CODEHOP (52a). Primers were designed to target the region encoding conserved chitin synthase motifs G(X)₄(Y/F)R and SWG and spanned the sequence encoding EDR and QRRRW residues, which define the enzyme (38, 43, 54). Reactions included 1 to 25 ng of DNA, 0.5 µM concentrations of forward and reverse primers, 1 µl of 10x buffer, 3.125 mM MgCl₂, 0.4 mM deoxynucleoside triphosphates, and 0.7 U of Taq polymerase (Promega and Gene Choice). Amplifications consisted of an initial denaturation of 95°C for 2 min, followed in turn by 30 cycles of 94°C for 10 s, 55 or 60°C for 30 s, and 72°C for 2 min. Amplified DNA fragments were separated and purified from an agarose gel as described above, ligated into the TOPO vector (Invitrogen), and used to transform Escherichia coli TOP 10 cells. Plasmids from 3 to 10 positive transformants were amplified with TempliPhi (GE Healthcare Biosciences Corp.) and sequenced with M13 forward and reverse primers as described above. Resulting DNA sequences were analyzed in Sequencher and aligned with CLUSTAL W. Positive identification of fragments corresponding to chitin synthase genes were based on the presence of the sequence that encoded EDR and QRRRW motifs. Sequences that differed by three or more nucleotides were considered distinct. A maximum-likelihood tree was calculated in proml in the PHYLIP software package (12) based on the alignment of all translated sequence fragments amplified with CODEHOP primers, T. pseudonana and P. tricornutum sequences, with Saccharomyces cerevisiae chitin synthase 3 (P29465 [UniProtKB/Swiss-Prot] ) as an outgroup. Sequence alignment was anchored by the EDR, QRRRW, and SWG motifs. Regions corresponding to amino acids 1007 to 1073 in T. pseudonana 7305, amino acids 573 to 770 in P. tricornutum 37908, and amino acids 1012 to 1078 in S. cerevisiae P29465 [UniProtKB/Swiss-Prot] were eliminated due to the uncertainty of alignment in this region. The resulting alignment was 111 amino acids long, and bootstrap values were calculated from the construction of 1,000 trees.

Localizing and quantifying chitin and silica deposition. Silica deposition was determined by incubating 50-ml aliquots of cultures with 0.5 µM 2-(4-pyridyl)-5{[4-(2-dimethylaminoethyl-aminocarbamoyl)-methoxy]phenyl}oxazole (PDMPO; Invitrogen) for 12 to 14 h in the same light and temperature conditions as the experimental conditions (56). Cells were analyzed with an Influx cell sorter flow cytometer (Cytopeia) equipped with a 355-nm UV laser. Fluorescent emission was detected at 460 nm (50-nm band-pass). Internal standards of 3-µm UV beads (Spherotech) were used for calibration. Epifluorescence was also visualized with an i80 microscope (Nikon) after excitation at 300 to 400 nm and emission detected with a 420-nm long-pass filter (Chroma). Potential chitin localization was determined by incubating cells with 4 µg of fluorescein isothiocyanate-labeled wheat germ agglutinin (FITC-WGA; Sigma) for 15 to 30 min. Cells were centrifuged at 15,000 x g for 5 min and resuspended in 1 ml of f/2 media or phosphate-buffered saline buffer (137 mM NaCl, 2.7 mM KCl, 10.4 mM Na₂HPO₄, 1.8 mM KH₂PO₄) to remove unbound FITC-WGA. The resulting fluorescent signal was quantified by flow cytometry after excitation with a 488-nm laser, and 2-µm yellow-green beads (Polysciences) were used as internal standards. Emission was detected at 530 nm (40 nm band-pass). Localization of FITC-WGA binding was visualized by fluorescence microscopy with a 500- to 570-nm wavelength filter (Chroma) after excitation at 455 to 500 nm. Competitive binding experiments with chitotriose and N-acetylglucosamine (41, 48) were used to determine whether the FITC-WGA bound specifically to chitin in T. pseudonana. First, 4 µg of FITC-WGA was incubated with 1 µg to 2.4 mg of chitotriose in 100 µl of water for 3 h, followed by a 30-min incubation with 100 µl of cell culture at >10⁶ cells ml^–1. Excess FITC-WGA was washed from the cells as described above, and the amount of FITC-WGA bound to cells was quantified by flow cytometry and microscopy. Afterward, similar incubations were conducted with 1 µg to 9.6 mg of N-acetylglucosamine as the competitive binding substrate. Flow cytometry signal comparisons were calculated in MatLab by using a two-sided Kolmogorov-Smirnov test with a 95% confidence interval. Modes were determined by creating a histogram of the 65,000 possible fluorescent intensities detected by the flow cytometer with a bin size of 200. If a distribution had two modes the average of the two numbers was used. Distributions were visualized by using FlowJo (Tree Star, Inc.).

Exponentially growing and silicic acid starved T. pseudonana cells were also incubated with FITC-labeled chitin-binding probe (FITC-chb) purified from Bacillus circulans (New England Biolabs) by pelleting the cells in a centrifuge at 15,000 x g for 10 min, resuspending them in 500 µl of TBS (0.05 M Tris base, 0.15 M NaCl), and then incubating them with 1 µl of FITC-chb for at least 3 h. The cells were washed, and FITC-chb localization was visualized as described above. Cells were also stained with 10 µg of calcofluor white (Sigma) for 30 min and visualized after excitation at 300 to 400 nm and emission detected with a 420-nm long-pass filter (Chroma). Additional diatom species T. punctigera, T. rotula, and S. costatum were also silicic acid starved and stained with FITC-WGA as described above.

Genetic sequences. Chitin synthase sequence fragments amplified from CODEHOP primers were deposited in the NCBI database under the following accession numbers: C. socialis (FJ544943 to FJ544945), L. undulatum (FJ544946 and FJ544947), S. costatum (FJ544948 to FJ544952), T. guillardii (FJ544953 to FJ544960), T. minuscula (FJ544961 to FJ544965), T. oceanica (FJ544966 to FJ544974), T. punctigera (FJ544975 to FJ544977), and T. wiessfloggii (FJ550076 to FJ550085).

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In silico and phylogenetic analyses of putative chitin synthases in diatoms. Automated annotation of the T. pseudonana whole genome sequence identified six genes that encoded putative chitin synthases (Table 3) based on detection of the chitin synthase active site domain with the conserved amino acid motifs EDR, QRRRW, and SWG. Full-length sequences and intron locations of the identified genes were confirmed through a combination of cDNA sequencing and RACE. The six T. pseudonana genes are located on three chromosomes. Chitin synthase genes occur as inverted repeats once on chromosome 4 and again on chromosome 7. Because the genes in both inverted repeats are identical to each other, it was not possible to determine whether retrieved cDNA sequences corresponded to both members of a repeat. Automated gene model predictions from version 3 of the T. pseudonana genome were consistent with transcribed regions confirmed by cDNA sequencing and RACE, except modeled protein ID 4368 incorrectly predicted a short exon at the 3' end.

View this table: [in this window] [in a new window]

TABLE 3. Genomic characteristics of chitin synthase genes discovered in the T. pseudonana and P. tricornutum genomes

Automated annotation of P. tricornutum whole genome sequence identified two genes that encoded putative chitin synthases (Table 3) that also contained the chitin synthase active-site domain with conserved amino acid motifs, despite the fact that this diatom is not known to produce chitin. The gene model for one chitin synthase (ID 44759) displayed expressed sequence tag (EST) support without additional open reading frames detected either upstream or downstream of the modeled gene and was therefore assumed to be accurate. The second gene (ID 37908) lacked EST support but displayed 61% nucleotide sequence identity and 57% translated nucleotide identity to the gene encoding 44759.

PCR products from the active site region of chitin synthase genes were successfully amplified from a variety of multipolar species including the known chitin fiber producers T. oceanica, T. punctigura, T. minuscula, T. guillardii, and T. weissflogii, as well as species not known to produce fibers, including Skeletonema costatum, Chaetoceros socialis, and Lithodesmium undulatum. PCR products were not successfully amplified from the other examined phytoplankton (Table 1). Phylogenetic analyses grouped the diatom chitin synthases into three clades with bootstrap support of 96, 71, and 100; a fourth potential clade has weaker bootstrap support of 45 (Fig. 1). The two P. tricornutum sequences formed a clade, identified here as clade D. The multipolar diatom sequences were distributed among three clades, identified here as clades A, B, and C. Each of these three clades contained sequences from multiple species. Moreover, most examined multipolar species encoded proteins that fell into more than one clade. Although clone libraries were not sequenced to saturation, it is of interest to note that only species possessing fultoportula encoded proteins associated with clade A sequences.

The large sequence divergence of clade D from clades A, B, and C is similar to the divergence between division 1 and division 2 fungal chitin synthases. The P. tricornutum chitin synthases do not have transmembrane domains N terminal to the QRRRW motif. This motif is also located relatively near the N-terminal end, which is typical of division 1 fungal and oomycete chitin synthases. In contrast, the T. pseudonana chitin synthases possess additional transmembrane domains and a cytochrome b₅ domain toward the N-terminal end and their QRRRW motif is located closer toward the C-terminal end, features that characterize division 2 fungal chitin synthases (54) (Fig. 2). The chitin synthases from clades A and C contain a myosin motor head domain and have identical protein domain predictions overall. The clade B chitin synthase lacks the myosin motor head domain and has an additional transmembrane domain. The three different protein domain structures further support the identification of phylogenetically distinct clades A, B, and D.

View larger version (16K): [in this window] [in a new window]

FIG. 2. Schematic of predicted protein domains for representatives from the four diatom chitin synthase clades and two classes of fungal chitin synthases. The species and protein or NCBI ID are indicated to the right; the clade or class designation is indicated to the left. Amino acid length is indicated by the top scale and represented in each protein by black lines. Functional domain symbols defined in the key at the bottom.

Differential expression of chitin synthase genes in T. pseudonana. Relative transcript abundance for the chitin synthase genes from the three clades was not significantly different (P > 0.01) under nutrient-replete or nitrate-depleted conditions (Fig. 3A). Relative transcript abundance for the genes from clades A and C was also not significantly different when growth was slowed by iron or silicic acid depletion relative to nutrient-replete conditions (P > 0.01). In contrast, the gene from clade B was upregulated by 7.6- ± 0.5-fold and 4.9- ± 0.5-fold (P < 0.01) relative to the control after experiencing less than 24 h of silicic acid or iron depletion, respectively. This treatment is expected to stop growth of a portion of cells in the same phase of the cell cycle (22).

View larger version (23K): [in this window] [in a new window]

FIG. 3. Transcript abundance of chitin synthase genes from clades A (protein ID 7305), B (protein ID 6575), and C (protein ID 4368) relative to the actin-like gene (protein ID 269504) in T. pseudonana cells under different nutrient conditions. (A) Cells maintained under nutrient-replete conditions or starved of nitrate, silicic acid, or iron for less than 1 day. (B) Time course of relative transcript abundance after replenishment of silicic acid to cells starved for silicic acid for 2 days. In panel A, asterisks indicate a significant difference from the nutrient replete control condition. Error bars represent the standard deviations of biological triplicates.

To assess transcript abundance in synchronized cells, cells were first starved of silicic acid for 48 h to block the majority of cells in the same phase of the cell cycle, and then replenished with sufficient silicic acid to resume cell division. This is similar to a treatment shown to synchronize T. pseudonana (22), modified in the present study with an additional day of silicic acid starvation. Transcript abundance of genes from clades A and C did not differ significantly (P > 0.01) among the 11 time points examined during the 8 h after replenishment of silicic acid. In contrast, transcript abundance of the clade B gene remained significantly lower (P < 0.01) during the early time points (5 min and 0.5, 1, and 2 h) after the addition of silicic acid than during the later time points (4, 6, and 8 h); transcript abundance also differed significantly between the 5-min and 7-h time points (Fig. 3B). This treatment appears to influence transcription of the clade B chitin synthase gene once cells reach a particular stage in cell division.

The effects of long-term silicic acid starvation and recovery were also examined. Asynchronous cells maintained in media with growth limiting concentrations of silicic acid increased exponentially (µ = 1.8 day^–1) for 3 days, after which point silicic acid was depleted to 0 µM in all bottles. Photosynthetic capacity decreased in these silicic acid starved cells, as measured by a drop in the Fv/Fm from 0.64 ± 0.02 to 0.5 ± 0.05 (Fig. 4A), the cell division was blocked, and cells began to display aberrant morphology and aggregation. The chlorophyll a relative fluorescence units (RFUs) doubled in 50-ml aliquots from each day-4 culture 1 day after the addition of 214 µM silicic acid. Together, these data indicate that growth of cells was stopped by silicic acid depletion. The tendency for aggregation persisted until the starved culture was replenished with silicic acid on day 9. By day 10, the Fv/Fm had increased to 0.62 ± 0.03 and cultures resumed asynchronous exponential growth (µ = 1.1 day^–1) (Fig. 4A).

View larger version (23K): [in this window] [in a new window]

FIG. 4. Growth and gene expression characteristics of cells grown in silicic acid deplete media and after replenishment of silicic acid on day 9, as indicated by the x-axis arrow. (A) Time course of cell concentration and Fv/Fm. Relative transcript abundance of chitin synthase genes (B) from clades A (protein ID 7305), B (protein ID 6575), and C (protein ID 4368) and the two girdle band-associated chitin-binding genes p150 and p150-like (protein IDs 12594 and 26041) (C) relative to the actin-like gene (protein ID 269504) (note the change in scales). Error bars represent the standard deviations of biological triplicates.

During the 3 days of exponential growth, the relative transcript abundance was low for the chitin synthase genes from clades A, B, and C and the two genes encoding chitin-binding proteins, p150 and p150-like, that localize to the girdle bands (days 2 and 3, Fig. 4B and C). A slight yet significant increase in clade B and a decrease in clade A chitin synthases occurred on day 3, suggesting that cells were beginning to experience silicic acid depletion. On day 4, when silicic acid was depleted and cell division stopped, the relative transcript abundance of both the clade B and C chitin synthases increased by 4.9- ± 0.6-fold and 2.3- ± 0.4-fold, respectively, relative to the previous day. This increased transcript abundance was significantly maintained through day 10 for the clade C chitin synthase and through day 11 for the clade B chitin synthase, although the levels did begin to drop as cells resumed unsynchronized division. Transcript abundance of the chitin binding gene p150 also increased 3.8- ± 0.8-fold on day 4. In contrast, transcript abundance of the clade A chitin synthase gene and p150-like gene did not increase during the period of silicic acid starvation. When related to different reference genes, the absolute value of transcript abundance, but not the direction of the daily changes, was affected by the choice of reference gene (see Fig. S1 in the supplemental material). Transcript abundance of the actin-like gene (ID 269504) appeared to maintain constant expression or increase during silicic acid starvation and decrease during exponential growth, depending on the reference gene used (see Fig. S1A in the supplemental material), and was thus identified as the most conservative reference gene for target gene expression. This gene was previously shown to have similar expression levels in different growth conditions (40).

Chitin localization in diatoms. Competitive binding experiments were conducted to determine whether FITC-WGA binds specifically to T. pseudonana chitin and thus can be used to determine chitin localization. As concentrations of chitotriose preincubated with FITC-WGA increased, reduced amounts bound to T. pseudonana cells, as detected with flow cytometry (Fig. 5A). Distributions of the FITC-WGA fluorescence of cells incubated with or without chitotriose were significantly different (P < 0.05) at every concentration, and the mode consistently decreased as chitotriose concentrations increased, to a final mode 70% less than noncompetitively bound cells. In contrast, the fluorescence distributions of FITC-WGA-bound cells with or without N-acetylglucosamine were not significantly different (P > 0.05) until concentrations of 22 mM were reached and the mode values varied by no more than ±5% (Fig. 5B). These results suggested that the FITC-WGA bound specifically to T. pseudonana chitin.

View larger version (28K): [in this window] [in a new window]

FIG. 5. Distributions of fluorescence intensities from T. pseudonana cells bound with FITC-WGA in the presence of a competitive binding substrate. (A) Cells bound with FITC-WGA preincubated with increasing concentrations of chitiotriose; (B) cells bound with FITC-WGA preincubated with increasing concentrations of N-acetylglucosamine. The concentrations of the competitive binding substrate are indicated in each panel. Thick black lines represent the fluorescence of cells in the presence of the competitive substrate, and thin gray lines represent the fluorescence of cells bound with FITC-WGA without a competitive substrate. Asterisks indicate a significant difference between distributions.

The impact of silicic acid starvation and recovery on silica deposition and chitin localization were examined by staining cells with FITC-WGA and PDMPO, a stain that incorporates into newly deposited silica (56). During the transition of T. pseudonana from exponential growth to silicic acid depletion, silica deposition and chitin exposure on the cell wall were inversely correlated. Exponentially growing cells had a relatively narrow distribution of high intensity (mode = 135 ± 5 RFUs) PDMPO staining as detected with flow cytometry (Fig. 6A). Epifluorescence microscopy confirmed that these exponentially growing cells were depositing silica at valves and girdle bands (Fig. 6B and C). In contrast, the FITC-WGA fluorescence of exponentially growing cells was slightly higher (mode = 66 ± 4 RFUs) than unstained cells (mode = 46 ± 2); FITC-WGA staining of cells was not detectable with epifluorescence microscopy. As cells entered silicic acid depletion on day 4 (Fig. 4A), the distribution of PDMPO fluorescence broadened, with more cells displaying lower intensity fluorescence, and the average mode decreased by 35% (Fig. 6A). Microscopic examination of cells indicated that many still deposited silica at the girdle bands but not the valves. Mode fluorescence of FITC-WGA per cell increased on day 4 by 35%. Microscopy revealed lectin localization around the girdle bands (Fig. 6C). The PDMPO fluorescent intensity per cell continued to decrease until the average mode was 1% of the original intensity on day 5, and no staining of cells was observed (Fig. 6C). The distribution of FITC-WGA fluorescence shifted to a higher intensity, with the highest average mode on day 9 being about twice the original intensity. Most cells had FITC-WGA bound to the girdle-band region. These cells also had an elongated and bent phenotype. This was particularly evident in the highest-fluorescing cells, where this exposed bent joint between the two thecae appeared to be larger (Fig. 6C). No binding of FITC-WGA was seen at the chitin fibers. After silicic acid replenishment on day 9, the mode of PDMPO fluorescence increased significantly (119 ± 5 RFUs) and the mode of FITC-WGA fluorescence decreased significantly (72 ± 1 RFUs) (P < 0.01).

View larger version (65K): [in this window] [in a new window]

FIG. 6. (A) Fluorescent signal of T. pseudonana cells bound with FITC-WGA (left column) and stained with PDMPO (right column) as measured by flow cytometry during transitions between silicic acid-replete and -limited conditions. The distribution of fluorescent signals from stained cells is indicated by colored histograms and from untreated cells by a black line. Silicic acid depletion was observed on day 4, and silicic acid was replenished on day 9. (B and C) Diagram of T. pseudonana frustules (B) and differential interference contrast and fluorescence micrographs of live cells during exponential growth (above) and stationary phase (below) (C). The scale bars in the light micrographs indicate 2 µm. FITC-WGA is localized by the intense green fluorescence, while chlorophyll autofluorescence appears as a diffuse green. PDMPO is localized by the blue fluorescence, while chlorophyll autofluorescence appears red.

To confirm the localization of chitin at the girdle bands, additional fluorescent chitin binding probes were used to label both exponentially growing and silicic acid-starved T. pseudonana cultures. Cells were incubated with FITC-WGA, calcofluor white, and the commercially available chitin-binding protein derived from bacteria, FITC-chb. All three stains localized to the girdle band regions of these cells and were especially visible on silicic acid-starved cells with elongated morphologies (Fig. 7). FITC-chb and calcofluor white also labeled the girdle bands of cells that were in the process of separating (Fig. 7). FITC-chb and calcofluor white also localized to the chitin fibers of T. pseudonana, whereas FITC-WGA did not (data not shown), illustrating differences in their chitin binding mechanisms and access to the substrate.

View larger version (105K): [in this window] [in a new window]

FIG. 7. Light and fluorescent micrographs of exponentially growing (above) and silicic acid-starved (below) T. pseudonana cells incubated with FITC-WGA, FITC-chb protein (FITC-chitin binding), or calcofluor white. The bar in the light micrograph indicates 2 µm. Multiple cells are shown from each treatment to illustrate cells in different stages of division. For the FITC-WGA and FITC-chb images, the brighter green fluorescence at the girdle bands (indicated by arrows) is the location of FITC-labeling, while the diffuse green is chlorophyll a autofluorescence. For the calcofluor white images, the calcofluor white appears blue, and chlorophyll a autofluorescence appears red.

The localization of chitin at the girdle band region in additional species of diatoms was confirmed by labeling silicic acid starved T. punctigera, T. rotula, and S. costatum with FITC-WGA (Fig. 8). All three species displayed an abnormal phenotype elongated at the theca. As in T. pseudonana, FITC-WGA localized to these elongated regions.

View larger version (77K): [in this window] [in a new window]

FIG. 8. Light and fluorescence micrographs of silicic acid-starved diatoms incubated with FITC-WGA. Arrows in the light micrographs indicate the location of FITC-WGA binding in the fluorescence micrographs. Scale bars indicate 2 µm for S. costatum and T. rotula and 10 µm for T. punctigera.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The discovery of chitin synthase genes in a variety of diatoms emphasizes the widespread distribution of chitin throughout eukaryotic lineages. Stramenopile chitin synthases have been sparsely characterized, and the present study provides a more complete picture of the evolutionary history of this polymer. Chitin is often widely distributed within lineages due to its fundamental structural function in the organisms; however, it can also evolve divergent functions and structures. For example, in fungi and insects chitin is a primitive, indispensable character that plays a fundamental structural role in forming the cell wall and exoskeleton and thus is present throughout these two groups of organisms. However, within the fungi, chitin has different functions during cell wall formation, bud scar formation, and cell division (52). In mollusks, chitin is also widely distributed and is used for diverse structural functions, such as the formation of the shells in bivalves (33), the structural pens in squid (26), and the radula of snails (60). Identification of chitin synthase genes in diatom genera (Phaeodactylum, Lithodesmium, Chaetoceros, and Skeletonema) that do not produce chitin fibers and the apparent localization of chitin to the cell wall suggests that a similar pattern of evolution has occurred within the diatoms. Our data suggest a broad distribution of chitin in diatoms due to a fundamental structural function in the cell wall with cases of additional functional diversification, such as fiber production. Consequently, the discovery of chitin as the major carbohydrate component in silica frustules by Tesson et al. (62) is possibly a true measurement rather than solely due to contamination from residual fibers.

The chitin synthase gene sequences from the representative pennate and multipolar centric species examined here are widely diverged, as expected for diatoms from evolutionarily distant lineages (3, 36). The predicted chitin synthases from P. tricornutum and T. pseudonana have different numbers and positions of transmembrane domains and thus are expected to assume different folded structures across the membrane. In addition, only the T. pseudonana chitin synthase genes encode a cytochrome b₅ domain, which can anchor proteins into lipid bilayers (15) and may have served as a template for the evolution of a lipid binding site (39). This difference in the presence of cytochrome b₅ domains is also observed between the two divisions of fungal chitin synthases. The acquisition or loss of this N terminus from the ancestral chitin synthase is hypothesized to have driven the divergence of the two fungal chitin synthase divisions (54). Fungi often contain chitin synthases from both divisions. In contrast, the few stramenopile lineages that have been examined do not contain this level of diversity within a single organism. The other chitin synthases identified in stramenopiles are from oomycetes, whose sequences are phylogenetically similar to division 1 fungal chitin synthases (54). In diatoms, T. pseudonana chitin synthases appear to be similar to division 2 fungal chitin synthases, while P. tricornutum chitin synthases are more similar to those of division 1. It is unclear whether P. tricornutum lost the cytochrome b₅ domain or T. pseudonana gained it. Clearly, there are interesting parallels to fungal chitin synthases within the stramenopiles, and the mechanism of this evolutionary connection may be clarified as more stramenopile chitin synthases are examined.

Three clades of chitin synthase genes were detected in multipolar diatoms; two were supported by high bootstrap values, and the distinctiveness of the third was supported by additional gene expression data. This suggests different evolutionary histories and possibly different functions for chitin within the cells. Each clade contains sequences from multiple species, and a single species commonly has chitin synthase genes in multiple clades. The lone exception to this is T. weissfloggi, with multiple sequences all assigned only to clade A; this could be due to insufficient sampling of clone libraries, and the large number of copies may be related to the large genome size of this strain (64). Clades B and C contain gene fragments from all successfully tested genera, including the chitin-producing Thalassiosira species and the non-chitin-fiber-producing C. socialis, L. undulatum, and S. costatum. Representatives from these two gene clades were likely present in the common ancestor prior to the evolution of the chitin-fiber-producing species. Interestingly, the encoded proteins from these two clades in T. pseudonana appear to utilize different modes of intracellular transport: the clade C protein possesses a myosin motor head domain N-terminal to the cytochrome b₅ domain, whereas the clade B protein lacks this domain. Clade A sequences are dominated by members of the genus Thalassiosira. The single exception is a gene fragment from Skeletonema costatum. This genus is derived from Thalassiosira (28) and also has fultoportula, the specialized pores in the silica wall through which chitin fibers emerge. However, rather than chitin fibers extending from the fultoportula, the S. costatum silica structures instead remain fused between divided cells and create distinctive chains of cells. Now, only species with fultoportula contain clade A chitin synthases. Therefore, clade A gene products are likely associated with the fultoportula structures, including the synthesis of the long chitin fibers. The clade A chitin synthase also possesses a myosin motor head domain in a similar location as the protein from clade C, suggesting a similar evolutionary history for these two gene families.

Support for functional differences between chitin synthases also comes from the differential transcription of genes in T. pseudonana cells exposed to different nutrient conditions. Relative transcript abundance of the clade B gene is particularly sensitive to short-term depletion of either silicic acid or iron. These conditions resulted in an increased transcript abundance of this clade B gene but not of the clade A or C genes. Coregulation of transcript abundance by silicic acid and iron depletion has been linked to cell wall processes (40). Transcript abundance of only the clade B gene varied depending on the amount of time elapsed since synchronized cells resumed division after short-term silicic acid starvation (22), suggesting that the clade B gene may have a role during a particular part of the cell cycle and particularly under silicic acid and iron stresses. When cells were submitted to longer-term silicic acid starvation, a response from both the clade B and clade C chitin synthases was detected, but the clade C response was of a much smaller magnitude. This gene may also be related to cell wall processes, although it appears to have less sensitivity to changing silicic acid conditions. Interestingly, similar differences in transcription level was found for the two genes encoding girdle-band associated chitin-binding proteins both in the present study and others (7). None of the tested conditions affected the transcript abundance of the clade A chitin synthase, suggesting that it is not directly related to these cell wall processes.

The chitin-specific lectin WGA was used to correlate clade B and C transcript abundance with chitin synthesis. The lectin bound to the girdle band region of cells. WGA can bind to multiple substrates, so its specificity for chitin on the cell was confirmed both by competitive binding experiments and by similar localization patterns using two additional chitin-binding probes with different binding mechanisms and access to the binding substrate. Interestingly, the lectin was not visibly bound to the chitin fibers extruded from the fultoportula. This might be due to the pure beta configuration of the fiber polymers, whose highly crystalline parallel arrangement may not be accessible to the lectin-binding sites. In contrast, both the chitin-binding probe and calcofluor white were visible on both the girdle-bands and the fultoportula chitin fibers, indicating a more general binding mechanism that does not exclude specific forms of chitin. When cells experienced silicic acid depletion and longer-term starvation, transcript abundance increased for the clade B and C chitin synthases and for a gene encoding a chitin-binding protein, p150, previously localized to cell wall girdle bands (6). The increase in transcripts corresponded to the increase in WGA binding at the girdle bands, which suggests that the clade B and C proteins are related to synthesis of chitin at the girdle bands in times of cell wall stress. The clade B gene also appears to be transcribed during normal progression through the cell cycle. The relative amount of transcript abundance of the clade B gene depended on length of exposure to silicic acid starvation and/or amount of time since recovery from different degrees of silicic acid starvation, events known to influence cell cycle progression (22). Further experiments are needed to determine the exact relation between chitin synthesis and cell cycle progression. However, both the chitin binding probe and calcofluor white localized to the girdle band region of exponentially growing cells preparing to separate after completing division. This supports a connection between chitin synthesis with the cell cycle and suggests that chitin synthesis is enhanced in stressed cells unable to precipitate silica and divide.

The identification of chitin synthase genes in a wide variety of diatoms and the localization of chitin at the girdle bands of T. pseudonana suggests the possibility that this use for chitin may be more widespread in diatoms. Silicic acid-starved T. punctigera, T. rotula, and S. costatum were labeled with WGA, and similar localization patterns were identified. The association of chitin with the cell wall should be considered in other diatom species; however, more detailed analysis within each lineage is needed.

Chitin appears to be intimately connected with silica processes because of its relationship to the cell wall, and this connection is illustrated in several ways. The distinctive fultoportula that serve as pores through the silica cell wall appear to have formed specifically to extrude chitin fibers (24), implying that this structure likely coevolved with the ability to produce chitin fibers at this cellular location, perhaps with the evolution of the clade A chitin synthase genes. Interestingly, chitin and silica have also evolved to form structures of similar functions. Diatoms can be found as solitary cells or in chains depending on the species. Either chitin fibers or silica spines can extend out from the cell to increase drag in the water and also form the connections between chains of cells. The evolution of either chitin or silica connections between cells may result from the different costs associated with production of rigid silica spines versus flexible chitin fibers; for example, rigid chains experience larger shear flow compared to flexible chains, which has the potential to increase their encounter rate with other particles and nutrients (29). The discovery of chitin at the girdle band region reveals another connection between chitin and silica. Our results suggest that chitin is localized to the girdle band region as a normal function of cell division and also under conditions when the cells are unable to precipitate additional silica. This stress-induced localization of chitin around the girdle band region is associated with cells that are elongated and bent and that tend to aggregate and sink. The potential interactions between silica and chitin likely reflect the necessary coordination of the organic and inorganic components of a functional cell wall.

Chitin synthase is proposed to have been present in a primitive eukaryote because of its widespread distribution throughout the eukaryotic tree of life (54). Thus, it seems probable that the ability to synthesize chitin was present in diatom ancestors as they evolved the ability to generate cell walls of silica. However, further work is needed to investigate whether chitin synthases can be identified in the more primitive centric lineage or a sister group to diatoms. Perhaps chitin has been maintained as an essential component of the cell wall, and it is only during times of cell wall stress or certain stages of the cell cycle that it can be detected through lectin binding at specialized locations. The use of chitin as a matrix for the precipitation of inorganic structures is a common strategy in eukaryotes (10, 11, 65), and the potential for chitin-silica structures specifically has been demonstrated in vitro (46). It is intriguing to speculate whether direct interactions between silica and chitin occur in diatoms; however, the current understanding of diatom silica precipitation is that it occurs in an acidic vesicle and does not rely on chitin (30-32). Interestingly, a recent microscopy-based study has shown that an uncharacterized organic matrix of fibers forms the core of the girdle bands onto which the silica structure is formed (23). Our data indicate a role for chitin in the cell wall, and future microscopy-based studies are needed to clarify that role. Chitin may be a component of the cell wall that is linked to silica but functionally separate. For example, chitin, rather than silica, may be used for more flexible components of the cell wall during the cell division cycle. Similarly, the apparently larger quantities of chitin detected at the girdle bands during silicic acid starvation may serve as a replacement cell wall material during conditions in which silica cannot be deposited. Notably, the increased synthesis of chitin at girdle bands in stressful conditions is reminiscent of chitinous cyst formation, a survival strategy used by other protists (42).

The relationship between silica and chitin has ecological implications for both diatoms and the global ocean. Chitin is the most abundant polymer in the ocean and serves as an enormous reservoir of organic carbon and nitrogen. Diatoms have traditionally been excluded from considerations of chitin production (27, 59). Our work suggests that diatoms are likely larger contributors to this organic reservoir than previously thought. Chitin is an attractive source of nutrients for microbes (37, 44), and the deposition of chitin at diatom girdle bands has the potential to facilitate diatom-microbe interactions (19, 25). The morphology of cell wall-stressed, chitin-producing cells is also of ecological importance. These cells are elongated and bent with a tendency to aggregate, which causes them to sink. If these processes observed in a laboratory translate to postbloom environmental conditions, it suggests that production of chitin on the cell wall is enhanced when growth is no longer possible, followed by sinking out of the surface layer. Sinking is proposed to be a survival strategy for cells in hostile surface environments and enables them to persist at depth until upwelling brings them back into favorable conditions (58). Thus, the deposition of chitin around the cell as it sinks to less hostile conditions may be a component of a survival strategy to persist through fluctuating ocean conditions. In addition, if cells sink below the mixing depth, sinking becomes a mechanism to pump carbon and other nutrients to the deep ocean (57).

 
This study was supported by a Gordon and Betty Moore Foundation Marine Microbiology Investigator Award to E.V.A.

We thank Rhonda Morales, Julie Koester, and Thomas Connolly for help with flow cytometry analysis; Ellen O. Lin and Audrey Djunaedi for help with sequencing; and Tiffany Truong for help with the RNA extraction and qRT-PCR. We also thank Julie Koester, Micaela Parker, and Nils Kröger for helpful comments and recommendations.

 
* Corresponding author. Mailing address: School of Oceanography, University of Washington, P.O. Box 357940, Seattle, WA 98195. Phone: (206) 616-1783. Fax: (206) 685-6651. E-mail: armbrust{at}u.washington.edu

Published ahead of print on 8 May 2009.

Supplemental material for this article may be found at http://ec.asm.org/.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Aluwihare, L. I., D. J. Repeta, S. Pantoja, and C. G. Johnson. 2005. Two chemically distinct pools of organic nitrogen accumulate in the ocean. Science 308:1007-1010.[Abstract/Free Full Text]
2
2. Armbrust, E. V., J. A. Berges, C. Bowler, B. R. Green, D. Martinez, N. H. Putnam, S. G. Zhou, A. E. Allen, K. E. Apt, M. Bechner, M. A. Brzezinski, B. K. Chaal, A. Chiovitti, A. K. Davis, M. S. Demarest, J. C. Detter, T. Glavina, D. Goodstein, M. Z. Hadi, U. Hellsten, M. Hildebrand, B. D. Jenkins, J. Jurka, V. V. Kapitonov, N. Kroger, W. W. Y. Lau, T. W. Lane, F. W. Larimer, J. C. Lippmeier, S. Lucas, M. Medina, A. Montsant, M. Obornik, M. S. Parker, B. Palenik, G. J. Pazour, P. M. Richardson, T. A. Rynearson, M. A. Saito, D. C. Schwartz, K. Thamatrakoln, K. Valentin, A. Vardi, F. P. Wilkerson, and D. S. Rokhsar. 2004. The genome of the diatom Thalassiosira pseudonana: ecology, evolution, and metabolism. Science 306:79-86.[Abstract/Free Full Text]
3
3. Bowler, C., A. E. Allen, J. H. Badger, J. Grimwood, K. Jabbari, A. Kuo, U. Maheswari, C. Martens, F. Maumus, R. P. Otillar, E. Rayko, A. Salamov, K. Vandepoele, B. Beszteri, A. Gruber, M. Heijde, M. Katinka, T. Mock, K. Valentin, F. Verret, J. A. Berges, C. Brownlee, J.-P. Cadoret, A. Chiovitti, C. J. Choi, S. Coesel, A. De Martino, J. C. Detter, C. Durkin, A. Falciatore, J. Fournet, M. Haruta, M. J. J. Huysman, B. D. Jenkins, K. Jiroutova, R. E. Jorgensen, Y. Joubert, A. Kaplan, N. Kroger, P. G. Kroth, J. La Roche, E. Lindquist, M. Lommer, V. Martin-Jezequel, P. J. Lopez, S. Lucas, M. Mangogna, K. McGinnis, L. K. Medlin, A. Montsant, M.-P. O.-L. Secq, C. Napoli, M. Obornik, M. S. Parker, J.-L. Petit, B. M. Porcel, N. Poulsen, M. Robison, L. Rychlewski, T. A. Rynearson, J. Schmutz, H. Shapiro, M. Siaut, M. Stanley, M. R. Sussman, A. R. Taylor, A. Vardi, P. von Dassow, W. Vyverman, A. Willis, L. S. Wyrwicz, D. S. Rokhsar, J. Weissenbach, E. V. Armbrust, B. R. Green, Y. Van de Peer, and I. V. Grigoriev. 2008. The Phaeodactylum genome reveals the evolutionary history of diatom genomes. Nature 456:239-244.[CrossRef][Medline]
4
4. Brott, L. L., R. R. Naik, D. J. Pikas, S. M. Kirkpatrick, D. W. Tomlin, P. W. Whitlock, S. J. Clarson, and M. O. Stone. 2001. Ultrafast holographic nanopatterning of biocatalytically formed silica. Nature 413:291-293.[CrossRef][Medline]
5
5. Brzezinski, M. A., and D. M. Nelson. 1986. A solvent-extraction method for the colorimetric determination of nanomolar concentrations of silicic acid in seawater. Mar. Chem. 19:139-151.[CrossRef]
6
6. Davis, A. K., M. Hildebrand, and B. Palenik. 2005. A stress-induced protein associated with the girdle band region of the diatom Thalassiosira pseudonana (Bacillariophyta). J. Phycol. 41:577-589.[CrossRef]
7
7. Davis, A. K., M. Hildebrand, and B. Palenik. 2006. Gene expression induced by copper stress in the diatom Thalassiosira pseudonana. Eukaryot. Cell 5:1157-1168.[Abstract/Free Full Text]
8
8. Dugdale, R. C., F. P. Wilkerson, and H. J. Minas. 1995. The role of a silicate pump in driving new production. Deep-Sea Res. I Oceanogr. Res. Papers 42:697-719.[CrossRef]
9
9. Dweltz, N. E., J. R. Colvin, and A. G. McInnes. 1968. Studies on chitan (β-(1-4)-linked 2-acetamido-2-deoxy-D-glucan) fibers of the diatom Thalassiosira fluviatilis Hustedt. 3. Structure of chitan from X-ray diffraction and electron microscope observations. Can. J. Chem. 46:1513-1521.
10
10. Ehrlich, H., M. Krautter, T. Hanke, P. Simon, C. Knieb, S. Heinemann, and H. Worch. 2007. First evidence of the presence of chitin in skeletons of marine sponges. II. Glass sponges (Hexactinellida: Porifera). J. Exp. Zool. B Mol. Dev. Evol. 308B:473-483.[Medline]
11
11. Falini, G., S. Albeck, S. Weiner, and L. Addadi. 1996. Control of aragonite or calcite polymorphism by mollusk shell macromolecules. Science 271:67-69.[Abstract]
12
12. Felsenstein, J. 1989. PHYLIP: phylogeny inference package (version 3.2). Cladistics 5:164-166.
13
13. Field, C. B., M. J. Behrenfeld, J. T. Randerson, and P. Falkowski. 1998. Primary production of the biosphere: integrating terrestrial and oceanic components. Science 281:237-240.[Abstract/Free Full Text]
14
14. Frigeri, L. G., T. R. Radabaugh, P. A. Haynes, and M. Hildebrand. 2006. Identification of proteins from a cell wall fraction of the diatom Thalassiosira pseudonana: insights into silica structure formation. Mol. Cell. Proteomics 5:182-193.[Abstract/Free Full Text]
15
15. George, S. K., L. Najera, R. P. Sandoval, C. Countryman, R. W. Davis, and G. M. Ihler. 1989. The hydrophobic domain of cytochrome B5 is capable of anchoring β-galactosidase in Escherichia coli membranes. J. Bacteriol. 171:4569-4576.[Abstract/Free Full Text]
16
16. Gooday, G. W. 1990. The ecology of chitin degradation, p. 387-429. In K. C. Marshall (ed.), Advances in microbial ecology, vol. 11. Plenum Press, Inc., New York, NY.
17
17. Guillard, R. R., and J. H. Ryther. 1962. Studies of marine planktonic diatoms. 1. Cyclotella nana Hustedt, and Detonula confervacea (Cleve) Gran. Can. J. Microbiol. 8:229.[Medline]
18
18. Guillard, R. R. L. 1975. Culture of phytoplankton for feeding marine invertebrates, p. 26-60. In W. L. Smith and M. H. Chanley (ed.), Culture of marine invertebrate animals. Plenum Press, Inc., New York, NY.
19
19. Haines, K. C., and R. R. L. Guillard. 1974. Growth of vitamin B₁₂-requiring marine diatoms in mixed laboratory cultures with vitamin B₁₂-producing marine bacteria. J. Phycol. 10:245-252.[CrossRef]
20
20. Herth, W. 1979. Site of beta-chitin fibril formation in centric diatoms. 2. Chitin-forming cytoplasmic structures. J. Ultrastruct. Res. 68:16-27.[CrossRef][Medline]
21
21. Herth, W., and E. Schnepf. 1982. Chitin-fibril formation in algae, p. 184-206. In R. M. Brown (ed.), Cellulose and other natural polymer systems. Plenum Press, Inc., New York, NY.
22
22. Hildebrand, M., L. G. Frigeri, and A. K. Davis. 2007. Synchronized growth of Thalassiosira pseudonana (Bacillariophyceae) provides novel insights into cell wall synthesis processes in relation to the cell cycle. J. Phycol. 43:730-740.[CrossRef]
23
23. Hildebrand, M., S. Kim, D. Shi, K. Scott, and S. Subramaniam. 2009. 3D imaging of diatoms with ion-abrasion scanning electron microscopy. J. Struct. Biol. 166:316-328.[CrossRef][Medline]
24
24. Hildebrand, M., E. York, J. I. Kelz, A. K. Davis, L. G. Frigeri, D. P. Allison, and M. J. Doktycz. 2006. Nanoscale control of silica morphology and three-dimensional structure during diatom cell wall formation. J. Materials Res. 21:2689-2698.[CrossRef]
25
25. Hünken, M., J. Harder, and G. O. Kirst. 2008. Epiphytic bacteria on the Antarctic ice diatom Amphiprora kufferathii Manguin cleave hydrogen peroxide produced during algal photosynthesis. Plant Biol. 10:519-526.[CrossRef][Medline]
26
26. Hunt, S., and A. E. Sherief. 1990. A periodic structure in the "pen" chitin of the squid Loligo vulgaris. Tissue Cell 22:191-197.[CrossRef][Medline]
27
27. Jeuniaux, C., and M. F. Vossfoucart. 1991. Chitin biomass and production in the marine environment. Biochem. Syst. Ecol. 19:347-356.[CrossRef]
28
28. Kaczmarska, I., M. Beaton, A. C. Benoit, and L. K. Medlin. 2005. Molecular phylogeny of selected members of the order Thalassiosirales (Bacillariophyta) and evolution of the fultoportula. J. Phycol. 42:121-138.[CrossRef]
29
29. Karp-Boss, L., and P. A. Jumars. 1998. Motion of diatom chains in steady shear flow. Limnol. Oceanogr. 43:1767-1773.
30
30. Kroger, N., R. Deutzmann, C. Bergsdorf, and M. Sumper. 2000. Species-specific polyamines from diatoms control silica morphology. Proc. Natl. Acad. Sci. USA 97:14133-14138.[Abstract/Free Full Text]
31
31. Kroger, N., R. Deutzmann, and M. Sumper. 1999. Polycationic peptides from diatom biosilica that direct silica nanosphere formation. Science 286:1129-1132.[Abstract/Free Full Text]
32
32. Kroger, N., S. Lorenz, E. Brunner, and M. Sumper. 2002. Self-assembly of highly phosphorylated silaffins and their function in biosilica morphogenesis. Science 298:584-586.[Abstract/Free Full Text]
33
33. Levi-Kalisman, Y., G. Falini, L. Addadi, and S. Weiner. 2001. Structure of the nacreous organic matrix of a bivalve mollusk shell examined in the hydrated state using cryo-TEM. J. Struct. Biol. 135:8-17.[CrossRef][Medline]
34
34. McLachlan, J., and J. S. Craigie. 1966. Chitan fibres in Cyclotella cryptica and growth of C. cryptica and Thalassiosira fluviatilis, p. 511-517. In H. Barnes (ed.), Some contemporary studies in marine science. George Allen and Unwin Ltd., London, England.
35
35. McLachlan, J., A. G. McInnes, and M. Falk. 1965. Studies on chitan (chitin-poly-n-acetylglucosamine) fibers of diatom Thalassiosira fluviatilis Hustedt. 1. Production and isolation of chitan fibers. Can. J. Bot. 43:707.[CrossRef]
36
36. Medlin, L. K., and I. Kaczmarska. 2004. Evolution of the diatoms. V. Morphological and cytological support for the major clades and a taxonomic revision. Phycologia 43:245-270.
37
37. Meibom, K. L., X. B. B. Li, A. T. Nielsen, C. Y. Wu, S. Roseman, and G. K. Schoolnik. 2004. The Vibrio cholerae chitin utilization program. Proc. Natl. Acad. Sci. USA 101:2524-2529.[Abstract/Free Full Text]
38
38. Merzendorfer, H. 2006. Insect chitin synthases: a review. J. Comp. Physiol. B Biochem. Syst. Environ. Physiol. 176:1-15.[Medline]
39
39. Mifsud, W., and A. Bateman. 2002. Membrane-bound progesterone receptors contain a cytochrome b₅-like ligand-binding domain. Genome Biol. 3:RESEARCH0068.[Medline]
40
40. Mock, T., M. P. Samanta, V. Iverson, C. Berthiaume, M. Robison, K. Holtermann, C. Durkin, S. S. BonDurant, K. Richmond, M. Rodesch, T. Kallas, E. L. Huttlin, F. Cerrina, M. R. Sussmann, and E. V. Armbrust. 2008. Whole-genome expression profiling of the marine diatom Thalassiosira pseudonana identifies genes involved in silicon bioprocesses. Proc. Natl. Acad. Sci. USA 105:1579-1584.[Abstract/Free Full Text]
41
41. Montgomery, M. T., N. A. Welschmeyer, and D. L. Kirchman. 1990. A simple assay for chitin: application to sediment trap samples from the sub-Arctic Pacific. Mar. Ecol. Prog. Ser. 64:301-308.[CrossRef]
42
42. Mulisch, M. 1993. Chitin in protistan organisms: distribution, synthesis, and deposition. Eur. J. Protistol. 29:1-18.
43
43. Nagahashi, S., M. Sudoh, N. Ono, R. Sawada, E. Yamaguchi, Y. Uchida, T. Mio, M. Takagi, M. Arisawa, and H. Yamadaokabe. 1995. Characterization of chitin synthase 2 of Saccharomyces cerevisiae: implication of two highly conserved domains as possible catalytic sites. J. Biol. Chem. 270:13961-13967.[Abstract/Free Full Text]
44
44. Nalin, D. R., V. Daya, A. Reid, M. M. Levine, and L. Cisneros. 1979. Adsorption and growth of Vibrio cholerae on chitin. Infect. Immun. 25:768-770.[Abstract/Free Full Text]
45
45. Nelson, D. M., P. Treguer, M. A. Brzezinski, A. Leynaert, and B. Queguiner. 1995. Production and dissolution of biogenic silica in the ocean: revised global estimates, comparison with regional data and relationship to biogenic sedimentation. Global Biogeochem. Cycles 9:359-372.[CrossRef]
46
46. Ogasawara, W., W. Shenton, S. A. Davis, and S. Mann. 2000. Template mineralization of ordered macroporous chitin-silica composites using a cuttlebone-derived organic matrix. Chem. Materials 12:2835.[CrossRef]
47
47. Parkinson, J., and R. Gordon. 1999. Beyond micromachining: the potential of diatoms. Trends Biotechnol. 17:190-196.[CrossRef][Medline]
48
48. Peters, W., and I. Latka. 1997. Wheat germ agglutinin-gold labeling, p. 33-40. In R. A. A. Muzzarelli and M. G. Peters (ed.), Chitin handbook. European Chitin Society, Grottammare, Italy.
49
49. Pfaffl, M. W. 2001. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 29:e45.[Abstract/Free Full Text]
50
50. Price, N. M., G. I. Harrison, J. G. Herring, R. J. Hudson, P. M. V. Nirel, B. Palenik, and A. F. M. Moorman. 1988/89. Preparation and chemistry of the artificial algal culture medium Aquil. Biol. Oceanogr. 6:443-461.
51
51. Ramakers, C., J. M. Ruijter, R. H. L. Deprez, and A. F. M. Moorman. 2003. Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci. Lett. 339:62-66.[CrossRef][Medline]
52
52. Roncero, C. 2002. The genetic complexity of chitin synthesis in fungi. Curr. Genet. 41:367-378.[CrossRef][Medline]
52
53. Rose, T. M., E. R. Schultz, J. G. Henikoff, S. Pietrokovski, C. M. McCallum, and S. Henikoff. 1998. Consensus-degenerate hybrid oligonucleotide primers for amplification of distantly related sequences. Nucleic Acids Res. 26:1628-1635.[Abstract/Free Full Text]
53
54. Round, F. E., R. M. Crawford, and D. G. Mann. 1990. The diatoms: biology and morphology of the genera. Cambridge University Press, Cambridge, England.
54
55. Ruiz-Herrera, J., J. M. Gonzalez-Prieto, and R. Ruiz-Medrano. 2002. Evolution and phylogenetic relationships of chitin synthases from yeasts and fungi. FEMS Yeast Res. 1:247-256.[Medline]
55
56. Saxena, I. M., R. M. Brown, M. Fevre, R. A. Geremia, and B. Henrissat. 1995. Multidomain architecture of β-glycosyltransferases: implications for mechanism. J. Bacteriol. 177:1419-1424.[Free Full Text]
56
57. Shimizu, K., Y. Del Amo, M. A. Brzezinski, G. D. Stucky, and D. E. Morse. 2001. A novel fluorescent silica tracer for biological silicification studies. Chem. Biol. 8:1051-1060.[CrossRef][Medline]
57
58. Smetacek, V. 1999. Diatoms and the ocean carbon cycle. Protist 150:25-32.[CrossRef][Medline]
58
59. Smetacek, V. S. 1985. Role of sinking in diatom life-history cycles: ecological, evolutionary, and geological significance. Mar. Biol. 84:239-251.[CrossRef]
59
60. Smucker, R. A. 1991. Chitin primary production. Biochem. Syst. Ecol. 19:357-369.[CrossRef]
60
61. Sollas, I. B. J. 1907. Memoirs: the molluscan radula: its chemical composition, and some points in its development. Q. J. Microsc. Sci. s2-51:115-136.
61
62. Sugiyama, J., C. Boisset, M. Hashimoto, and T. Watanabe. 1999. Molecular directionality of beta-chitin biosynthesis. J. Mol. Biol. 286:247-255.[CrossRef][Medline]
62
63. Tesson, B., S. Masse, G. Laurent, J. Maquet, J. Livage, V. Martin-Jezequel, and T. Coradin. 2008. Contribution of multi-nuclear solid state NMR to the characterization of the Thalassiosira pseudonana diatom cell wall. Anal. Bioanal. Chem. 390:1889-1898.[CrossRef][Medline]
63
64. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.[Abstract/Free Full Text]
64
65. von Dassow, P., T. W. Petersen, V. A. Chepurnov, and E. V. Armbrust. 2008. Inter- and intraspecific relationships between nuclear DNA content and cell size in selected members of the centric diatom genus Thalassiosira (Bacillariophyceae). J. Phycol. 44:335-349.[CrossRef]
65
66. Wainwright, S. A. 1963. Skeletal organization in coral, Pocillopora damicornis. Q. J. Microsc. Sci. 104:169-171.
66
67. Zar, J. H. 1996. Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ.

---

Eukaryotic Cell, July 2009, p. 1038-1050, Vol. 8, No. 7
1535-9778/09/$08.00+0     doi:10.1128/EC.00079-09
Copyright © 2009, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Supplemental material Alert me when this article is cited Alert me if a correction is posted Citation Map Services E-mail this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Durkin, C. A. Articles by Armbrust, E. V. Search for Related Content PubMed PubMed Citation Articles by Durkin, C. A. Articles by Armbrust, E. V.