Mar. Biotechnol. 2, 577–586, 2000 DOI: 10.1007/s101260000044

© 2000 Springer-Verlag New York Inc.

Molecular Cloning and Antiserum Development of Cyclin Box in the Brown Tide Alga Aureococcus anophagefferens Senjie Lin,1,* Erika Magaletti,2,† and Edward J. Carpenter2 1 2

Department of Marine Sciences, University of Connecticut, Groton, CT 06340, U.S.A. Marine Sciences Research Center, State University of New York, Stony Brook, NY 11794, U.S.A.

Abstract: Cyclins can be useful cell cycle markers for growth rate studies on harmful algal blooms. In this study, a gene fragment corresponding to cyclin box was cloned for the brown tide alga Aureococcus anophagefferens. This algal gene fragment, designated as Btcycl1, was most similar to cyclin B. Oligopeptides based on the deduced amino acid sequence were synthesized and used to raise an antiserum that reacted on Western blots with a protein of about 63 kDa, the same size as cyclin B in other organisms. The cyclin B–like protein recognized by this antiserum, and the messenger RNA amplified using the primers, were more abundant in exponential cultures and decreased markedly in stationary cultures. This protein also appeared to be cell cycle dependent. Immunofluorescence labeling showed that this antiserum specifically stained a protein in Aureococcus cells and had no cross-reaction with bacteria that were present in the algal culture. The Btcycl1 sequence and the antiserum will provide a useful tool for studies on regulation of in situ growth rate for this brown tide alga. Key words: antibody, Aureococcus, brown tide, cell cycle, cyclin box, phytoplankton.

I NTRODUCTION Brown tide, the intense blooms of the pelagophyceae microalga Aureococcus anophagefferens (Sieburth et al., 1988; DeYoe et al., 1995), has had a serious impact on the environment and local economy in Long Island and some other areas along the northeast coast of the United States. Since its first outbreak in 1985 in several coastal embayments of

Received May 22, 2000; accepted July 13, 2000. †Present address: ICRAM Istituto Centrale per la Ricerca Scientifica e Tecnologica, Applicata al Mare Via di Casalotti, 300 00166 Roma, Italy *Corresponding author: telephone 860-405-9168; fax 860-405-9153; e-mail [email protected]

Long Island, as well as New Jersey and Rhode Island (Cosper et al., 1987; Olsen, 1989; Smayda and Villareal, 1989), brown tide blooms have recurred intermittently in several Long Island embayments with different intensity and duration (Anderson et al., 1993, and data from Dr. Robert Nuzzi, Suffolk County Health Department). Similar blooms occurred in other areas as far away as South Africa (Pitcher et al., 2000). The rapid growth and accumulation of this microalga can be so high (>109 cells/L) that it changes the impacted water into a brownish color. Major effects of this nuisance bloom include losses in the area of eelgrass beds (Zostera marina) owing to the severe reduction in light penetration (Dennison et al., 1989) and recruitment failure and mortality of commercially valuable bay scallops (Agropecten irradians irradians) and mussels (Mytilus edulis), which are

578

Senjie Lin et al.

unable to graze efficiently on A. anophagefferens cells (Tracey, 1988; Bricelj and Kuenstner, 1989). The causes of brown tide are still elusive (Bricelj and Lonsdale, 1997), largely because no information on its in situ growth rate is available. Earlier studies have attempted to attribute formation and demise of the blooms to metereological and hydrographical events (Cosper et al., 1989; Smayda and Villareal, 1989; Vieria and Chant, 1993), fluctuation in groundwater flow (La Roche et al., 1997), organic and micro nutrients (Cosper et al., 1989; Dzurica et al., 1989; Gobler and Cosper, 1996; Berg et al., 1997), grazing (Caron et al., 1989), and virus infection (Milligan and Cosper, 1994). These factors are compounded in nature, and information on the in situ growth rate is necessary to assess the relative importance of these factors. The poor knowledge of in situ growth rates is due to the lack of a proper method. One promising method relies on cell cycle analysis, which requires a cell cycle marker. Cyclins are a group of cell-cycle-dependent proteins (Evans et al., 1983) that potentially can be useful for growth rate studies (Carpenter et al., 1998). Cyclin B, in particular, can be very useful because it is highly accumulated during late G2 and M phases of the cell cycle (Murray and Hunt, 1993), and thus can be a marker for a terminal event when using the cell cycle approach (Carpenter and Chang, 1988; Lin et al., 1996). This protein is a regulatory component in the CDC2 kinase complex that regulates the cell cycle transition from G2 to M phase (Nigg, 1995). In a typical eukaryotic cell cycle, cyclin B reaches a threshold level before cells in G2 can progress into mitosis (King et al., 1994). Essentially, it binds to the p34cdc2 protein (CDC2) and triggers a cascade of phosphorylation and desphosphorylation and activates CDC2, leading to mitosis. Upon completion of mitosis, cyclin B is bound with ubiquitin and degraded (Glotzer et al., 1991), and the CDC2 kinase returns to its inactive state. In cyclin B, there is a domain, named the cyclin box, which is conserved among other cyclins and has important functions in binding to and activating CDC2 and other cell-cycledependent kinases (Lees and Harlow, 1993; Horton and Templeton, 1997). Unfortunately, no cyclin genes have been cloned and characterized for marine phytoplankton including harmful algal bloom species Aureococcus. In the present study, a cyclin box (Btcycl1) was identified for this brown tide alga that was homologous to cyclin B, and an antiserum was developed that can be used to detect the cyclin B–like protein with Western blotting and immunofluorescence. This antiserum appears to be potentially useful for growth rate studies for the brown tide alga.

M ATERIALS

AND

M ETHODS

Batch Cultures Cultures of A. anophagefferens (kindly provided by E. Cosper) were grown in 2 L of modified f/2 medium (Guillard and Ryther, 1962) at 18°C, with a 12:12 light-to-dark photocycle and a photon flux of 112 µE m−2 s−1. Modifications of the original f/2 medium included the addition of selenium (Na2SeO3, 10 nM) and replacement of EDTA with citric acid (50 µM) as a chelator of trace metals and inorganic phosphate with glycerophosphate (36 µM). The cell concentration was followed daily by use of a hemocytometer.

Reverse Transcriptase Polymerase Chain Reaction When a culture was growing exponentially, about 1 L was harvested by centrifugation at 890 g at 4°C for 50 minutes. The cell pellet was resuspended in 1 ml of Trizol Reagent (Gibco BRL, MD) and vortex mixed, followed by incubation at 50°C for about 15 minutes under continuous vigorous shaking. Previous tests showed that this incubation step in addition to the vortex mixing recommended by Gibco BRL enhanced lysis of cells and maximized the yield of RNA. Subsequently, RNA was isolated essentially following the manufacturer’s instructions and quantified spectrophotometrically at 260 nm. Four micrograms of the total RNA was used to synthesize complementary DNA with oligo(dT)17–20 as a primer using a preamplication kit (Superscript, Gibco BRL). A pair of primers were derived from known cyclin B sequences from other organisms, which flanked the highly conserved cyclin box region (see Figure 2). The forward primer (cyclin 1) was 5⬘-ATGCGIGGIATHYTIRYIGAYTGG-3⬘ whereas the reverse primer (cyclin 2) was 5⬘-GGRTAIATYTCYTCRTAYTT-3⬘ (I stands for inosine), which corresponded to amino acid residues MRGILIDW and KYEEIYP, respectively. Cyclin 1 was biased toward cyclin B of the fission yeast (Schizosaccharomyces pombe) and the higher plant Brassica. Cyclin 2 was biased toward cyclin B of higher plants and animals. PCR was performed with these primers on a Hybaid Omni-E thermocycler, under conditions as previously described (Lin and Carpenter, 1998), with an annealing temperature of 55°C. The PCR product was resolved by 1% agarose gel electrophoresis. The specific DNA band was excised and DNA was extracted using Spin-X (Costa), following the manufacturer’s instructions.

Cyclin Box in Brown Tide Alga 579

Cloning and Sequencing The purified PCR product was ligated into a TA cloning vector (Invitrogen). Cloning and sequencing were performed essentially as previously described (Lin and Carpenter, 1998). Plasmids were isolated using the Miniplasmid Spin Purification Kit (Qiagen). Both strands of the plasmid DNA were sequenced using an ABI Prizm automated sequencing system, based on the dideoxynucleotide reaction termination technique. The gene sequence was analyzed using GCG (Genetic Computing Group, Wis.) and PHYLIP software packages (Felsenstein, 1989).

Production of Antiserum Based on the deduced amino acid sequence, 2 oligopeptides were synthesized and the mixture of the 2 peptides was used to immunize 2 rabbits (Biosynthesis, Tex.). Their sequences were as follows: peptide 1, NH2-CHLKFKMLQPTIYLTVQICOOH; peptide 2, NH2-CTVQIIDRYLSAKQIDRNQLQ-COOH. The sequence selection was based on preference for uniqueness and maximum hydrophilicity. Unique regions were targeted to obtain Aureococcus cyclin-specific antibodies. High hydrophilicity was selected to maximize the probability that the epitopes in native protein would be exposed and accessible for antibody binding. The peptides were conjugated with the carrier keyhole limpet hemocyanin (KLH) before injection. Test bleeds were collected at week 6, 8, and 10, and the final bleed of about 40 ml for each rabbit was obtained (week 11).

Western Blotting When cultures entered early exponential and stationary stages, 200 ml was harvested by centrifugation and the pellet was stored at −80°C for protein extraction later. Cell lysates were prepared by sonicating the cells in a RIPA buffer (TrisHCl 50 mM, pH 7.4; 1% Nonidet P-40, 1 mM EGTA) supplemented with the protease inhibitors PMSF (0.5 mM), aprotinin (0.15 µM), leupeptin (1 µM), and pepstatin A (0.85 µM). Sonication was performed (Branson Sonifier 250, VWR Scientific) at a setting of 1.5 output, 15% duty, and 2 cycles of 10 seconds. Protein concentration was measured (DC Protein Assay, Bio-Rad). Proteins in the crude extract (50 µg of total proteins per lane) were then separated by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) and proteins containing the cyclin box were detected with Western blotting, as previously described (Lin et al., 1994). The antiserum against the brown

tide cyclin box produced in this study was used with a dilution of 1:500. After a final rigorous washing, the detected bands were visualized using an enhanced chemiluminescence system (Amersham Pharmacia, N.J.). The bands were digitalized using LogiTech ScanMan and band intensity measured with SigmaGel (SPSS, Calif.).

Determination of Gene Transcription To investigate transcription of cyclin box in relation to growth stages, 200-ml samples were collected from exponential and stationary growth phases, using the same procedure as described above. Two microliters of total RNA from each sample was used to prepare cDNA, using random hexamer primers provided by Gibco BRL. Subsequently, cyclin 1 and cyclin 2 primers were used to amplify 10% of the cDNA (2 µl) for 20 cycles. A primer set for 28S ribosomal RNA was also used to concurrently amplify this gene, as a control to calibrate variation in initial RNA quantity used for cDNA synthesis and in PCR efficiency. The sequences of the 28S primers were as follows: 28S1, 5⬘GCATATCAATAAGCGGAGGAAAAGAAAC -3⬘; and 28S2, 5⬘GGTCCGTGTTTCAAGACGG-3⬘. The low cycle number was used to ensure exponential growth of PCR product. Preliminary tests with 10, 20, and 30 cycles showed that within 20 cycles, both the cyclin box and the 28S rRNA cDNAs increased exponentially. Next, 5 µl of each PCR product was resolved in 1% agarose gel electrophoresis. The bands of 28S rDNA were bright enough to quantify without further treatment whereas bands of Btcycl1 mRNA needed to be enhanced by Southern hybridization. After proper treatment of the gel, DNA transfer and hybridization were carried out according to the manufacturer’s instructions of the ECL Nucleic Acid detection kit (Amersham Pharmacia). The cloned cyclin box fragment was labeled with labeling reagents in the kit and used as the probe. Ten nanograms of the probe DNA was dissolved in 10 µl of sterile ddH2O and incubated for 1 hour at 37°C with 10 µl of labeling reagent and the same volume of glutaraldehyde. After prehybridization for 1 hour at 49°C in the prehybridization buffer, the labeled probe was added and hybridization was undertaken overnight at 42°C with continuous rotation in a hybridization oven (Hybaid). The membrane was then washed at 42°C twice, each time for 15 minutes, in the primary wash solution (0.5× SSC, 0.4% SDS) with gentle agitation, followed by 2 secondary washes at room temperature with 2× SSC. Following ECL detection, the membrane was exposed to an x-ray film for 5 minutes, which was then developed using Kodak

580 Senjie Lin et al.

Figure 1. Cloning of cyclin box in Aureococcus anophagefferense. A: Agarose gel electrophoresis (1%) of PCR product stained with ethidium bromide. Lane M, 1-kb marker; the size of bands from top to bottom is 1, 0.51, 0.40, 0.34, 0.30, and 0.22 kb. Lane 1, the PCR product, whose size is indicated by the number on the right. B: Nucleotide and deduced amino acid sequence of the PCR product (cyclin box). Underlined are sequences used for antibody production. Gene sequence has been deposited in GenBank under accession number AF036313.

D-19 developer and Kodak fixative. The band intensity was quantified as described above.

Whole-Cell Immunostaining Immunofluorescence was conducted for two purposes: to optimize an immunostaining protocol for Aureococcus; and to examine whether the antiserum cross-reacted with bacteria that were associated with Aureococcus cultures. Some modifications were made to a previously established protocol (Lin and Carpenter, 1996) to suit the small size of the alga and to facilitate working in the field. Exponentially growing cells were fixed in 2% PFA by adding an equal volume of 4% PFA prepared in filtered artificial seawater (pH 8.1, salinity 28%). This fixation procedure was designed to facilitate field sampling in future studies, when a centrifuge is not available to obtain a cell pellet. Fixed samples were maintained at 4°C for 6 hours and then transferred into chilled methanol (−20°C) to improve cell membrane permeabilization and to extract chlorophyll, as described in Lin and Carpenter (1996). The fixation procedure was tested by maintaining the samples in methanol overnight, and for 2 weeks, 5 weeks, and 2 months. Cells were filtered onto a 0.6-µm Millipore filter with a vacuum of less than 1 atm to avoid cell damage. Five milliliters of 0.2 M phosphate buffer was added and filtered through to wash the cells. Then the filters were placed on poly-L-lysinecoated slides with sample-side down and were briefly centrifuged (1 minute) to allow the transfer of cells from the filters onto the slides. Next, the cells adhered to the slides were incubated for 15 minutes at 4°C in dimethylsulfoxide

(DMSO, 0.5%, vol/vol) to permeablize the cell membrane for antibody penetration, followed by 3 washes with PBS (2 minutes each). Prior to antibody incubation, cell samples were treated with normal goat serum for cell surface antigen staining or bovine serum albumin for intracellular protein staining to reduce background fluorescence. The cells were then incubated for 4 hours at room temperature in the primary antibodies diluted in PBS. Antiserum against cell surface antigen (kindly provided by E. Cosper) was used at 1:5000 dilution; anti-Rubisco (provided by P.G. Falkowski) was used at 1:500; anti-BtCYCL1 produced in this study was used at a dilution of 1:250. After the incubation, three 2-minute washes with PBS were carried out and followed by a 1-hour incubation with the secondary antibody conjugated with fluorescein isothiocyanate (FITC) (Sigma) at a dilution of 1:100. Finally, the cells were rinsed 3 times and stained with DAPI. The slides were mounted with a coverslip by adding 1 drop of Gel-Mount (Biomeda, Calif.). Immunofluorescence and DAPI fluorescence of cells were examined under an epifluorescence microscope by using blue and UV light excitation, respectively.

R ESULTS Gene Sequence PCR with primers cyclin 1 and cyclin 2 generated a product that appeared in a 1% agarose gel as a band of around 180 bp (Figure 1, A). Cloning of this amplicon yielded a nucleotide sequence that is most closely related to cyclin B as

Cyclin Box in Brown Tide Alga 581

Figure 2. Multi-alignment and phylogenetic tree of cyclin box. A: Btcycl1 aligned with counterparts in other organisms using CLUSTALW 1.8 and BOXSHADE 3.2. Black background highlights identical amino acid residues; gray indicates similar residues; nonshaded residues are dissimilar. The lines under the alignment indicate regions from which primers were designed. B: Maximum parsimony tree constructed with PHYLIP. Bootstrapping was run with 100 replicates and values over 50% are shown at the nodes. C: Neighbor-joining analysis conducted with PHYLIP. Scale bar = 0.1

substitution on each site. Species names and GenBank accession numbers (in parentheses) are as follows, Medsa, Medicago sativa (X68741); Arath, Arabidopsis thaliana (M80190); Brana, Brassica napus (S53003); Bt, brown tide = Aureococcus (AF036313); clam (P13952); Dun, Dunaliella tertiolecta (AF036312); human (P14635); Isochrysis (AF036315); Schpo, Schizosaccharomyces pombe (P10815); Skeletonema (AF036318); soy, soybean (Z26331); tobacco (T03021); Xenla, Xenopus laevis (J03166); Zea (U10077).

shown by FASTA analysis. Sequencing of 4 clones yielded identical sequences. Further comparative analyses on this gene, tentatively designated as Btcycl1, revealed an identity of about 50% to 75% to cyclin box in other phytoplankton species and higher organisms. This stretch of nucleotide sequence codes for 62 amino acid residues (Figure 1, B); the amino acid sequence was about 60% identical and 80% similar to counterparts in other organisms (Figure 2, A). Maximum parsimony and neighbor-joining analyses generated trees that indicated clustering of A. anophagefferens

with the fission yeast (S. pombe) and two other phytoplanktons (I. galbana and Skeletonema costatum; S. Lin, unpublished data), while green algae and higher plants fell in another clade (Figure 2, B). As expected, animal cyclin B sequences were more distant from Btcycl1.

Antibody Specificity On Western blots, the antiserum developed in this study reacted with a major protein band of 63 kDa and 3 minor

582 Senjie Lin et al.

Figure 4. Variation in the expression of Btcycl1 gene with growth stage. A: Western blot with proteins extracted from exponential (lane 1) and stationary (lane 2) growth stages; 50 µg of total protein was loaded to each lane. B: RT-PCR-Southern hybridization of Btcycl1 (upper arrow at right) and RT-PCR of the 28S rRNA (lower arrow at right). Lane 1, sample from exponential growth stage; lane 2, sample from stationary stage. Figure 3. Western blot of Aureococcus. Fifty micrograms of total protein was loaded to each lane. Lane 1, preimmune serum; lane 2, 6-week test serum; lane 3, final serum. On the left is the protein molecular weight marker.

bands of higher molecular weights (Figure 3). This major protein band was absent in the negative control in which the preimmune serum was used in place of the antiserum. The antiserum titer appeared to increase from the first test bleed (6 weeks after first immunization) to the final bleed (11 weeks after first immunization).

Gene Expression The abundance of the major protein band varied with growth stage, from a high level in the exponential growth stage (µ = 0.45/d) to a markedly lower level in the stationary stage (µ = 0.12/d). The same trend was observed for mRNA, which was determined using RT-PCR coupled with Southern blot hybridization (Figure 4, B). With the same total protein loaded to each lane, the cyclin B–like protein in the exponential growth stage was over 5 times as abundant as that in the stationary stage. Because the same amount of total RNA was used for cDNA synthesis and then in the PCR reaction, the band intensity of the 28S rRNA amplified simultaneously with Btcycl1 mRNA was similar between the sample from the exponential culture and that from the stationary culture. When normalized to 28S rRNA band intensity, the Btcycl1 mRNA level was about 3.5 times as high in the exponential stage as in the stationary stage. In the cell cycle that was partially synchronized by the light-dark cycle,

the cyclin B homologue detected on the Western blot was more abundant late in the light period, when the majority of cells were about to divide (indicated by major increase in cell concentration; Figure 5). The abundance of this protein decreased considerably late in the dark period when the cell division rate decreased, probably suggesting degradation of this cyclin B–like protein in G1 cells.

Immunofluorescence Good immunostaining was obtained for cell surface antigen and Rubisco, on samples preserved for up to 2 months (Figure 6, A–C; Magaletti, 1998). Cell surface staining is not shown here, because it has been shown elsewhere (Lin and Carpenter, 1996; Magaletti, 1998). Rather, intracellular staining is the focus of this article. No difference was noticed in Rubisco staining (or cell surface staining) for samples preserved overnight and for 2, 5, and 8 weeks (Figure 6, A–C). With the antibody developed in this study, positive staining was also obtained for Aureococcus, and no cross-reaction was observed for bacteria that were present in the sample (Figure 6, D and E).

D ISCUSSION This is the first attempt to clone a cyclin gene fragment and use it as a tool to study the growth rate of phytoplankton. This study used synthetic oligopeptides to develop an antibody against the protein for which the genetic codes were

Cyclin Box in Brown Tide Alga 583

Figure 5. Variation in abundance of the cyclin B–like protein with the cell cycle. A: Diel increase of the cell concentration. B: Western blot using anticyclin box. Shown are both the blot autoradiograph and the band intensity values (arbitrary unit). The same amount of total protein was loaded to each lane. Lane numbers correspond to sample numbers shown on the growth curve in A. Black bar on the top depicts dark period of the photocycle.

obtained. This study provides a framework for future studies on other cell cycle proteins that are of ecological significance (Lin and Carpenter, 1999). In addition, this study reveals some important characteristics of the cell cycle control machinery in brown tide alga.

Conservation of Cyclin-Based Cell Cycle Regulation The eukaryotic cell cycle engine is essentially composed of CDK kinases and cyclins, which are highly conserved in general (Murray and Hunt, 1993). Cyclins are a complex multigene family with different members present in different organisms, among which cyclin B appears to be the most common member. Cyclin box is a domain conserved among these different cyclins and appears to be functional in binding to and activating CDKs (e.g., Lees and Harlow, 1993) as well as controlling specificity for them (Horton and Templeton, 1997). Multiple cyclin box sequences would thus be expected when highly conserved sequences

Figure 6. Micrograph of immunofluorescence and DAPI staining. Rubisco immunostaining on samples preserved for 2 weeks (A), 5 weeks (B), and 2 months (C). Cyclin box FITC immunofluorescence visualized under blue light excitation (D) and DAPI staining of the same samples as visualized under UV light (E). Scale bars = 10 µm; the one in A applies for A–C, while that in D applies for D and E. Thick arrows indicate Aureococcus cells that were stained by both anti-BtCYCL1 (D) and DAPI (E), whereas thin arrows point to bacteria that were stained only by DAPI.

were used as primers in cloning attempts. However, only one gene sequence was isolated from Aureococcus and some other algae (S. Lin, unpublished data). This is probably because the primers used in this study strongly selected cyclin B gene and suggests that the cyclin box domain in this alga may be variable among different cyclins. In accordance with this, the antiserum produced in this study strongly reacted with only one protein band after thorough washing of the blots following antibody incubation. The major protein band appears to be a homologue of cyclin B by molecular size (63 kDa, the same as mammalian cyclin B), in agreement with the results of gene sequence analysis. Therefore, identifying other cyclins may require a different cloning strategy or using more universal primers in PCR reactions. Nevertheless, results shown here demonstrate that the cyclin B–based cell cycle regulation is conserved in this brown tide alga. With the short sequence obtained in this study, it is not feasible to derive a strongly reliable phylogenetic tree with fine-scale resolution. However, two interesting points can be made with coarse-scale analyses. First, the trees built with two different algorithms, maximum parsimony and neigh-

584 Senjie Lin et al.

bor joining, showed that the cyclin box isolated from the brown tide alga was more closely related to mitotic cyclins of higher plants and the fission yeast than animal counterparts. Second, Aureococcus appears to be more closely related to chromophyte algae than green algae and higher plants. This result agrees well with the phylogenetic relation established by 18S rRNA sequence (DeYoe et al., 1995), where Aureococcus was located more closely with diatoms than Chlorella.

Association of the Cyclin B–like Protein with Growth Stage and the Cell Cycle The cyclin B homologue in Aureococcus was associated with active growth phase on both protein and mRNA levels. This is congruent with the fact that cyclin B is a mitotic cyclin associated with actively dividing cell populations (Murray and Hunt, 1993). Fluctuation in abundance modulates the activity of this cyclin in the cell division cycle (Evans et al., 1983). Therefore, in a nondividing cell population (e.g., differentiated tissues in multicellular organisms), cyclin B may diminish or disappear. Association of cyclin B homologues with actively proliferative cell population has also been found for D. tertiolecta (Lin et al., 1996) and Prorocentrum minimum (S. Lin, unpublished data). In accordance with the nature of cyclin B observed in other organisms, the cyclin B–like protein identified in this study also appeared to be dependent on the predivision (S-G2-M) phase of the cell division cycle (as seen from the growth curve). The expression pattern in relation to growth stage and the cell cycle renders this protein a promising cell cycle marker for growth rate studies.

Potential for In Situ Studies For a cell cycle marker to be useful for in situ growth rate studies, it needs to be reactive, specific, and amenable to whole cell (i.e., in situ) detection. First, there must be an effective immunostaining protocol. Immunofluorescence results from the present study demonstrated that the protocol modified from a general immunofluorescence protocol was effective for preserving and immunostaining the samples. Homogeneous staining was obtained for cell surface antigens and Rubisco, on samples preserved for an extended period of time. Good staining of Rubisco and BtCYCL1 also indicate that use of DMSO was effective to permeablize the brown tide cells that were fixed with the method reported here. The modification of the protocol

simplifies sample fixation and will facilitate field collection and in situ preservation of the samples. Use of filtration rather than centrifugation in the first several steps of immunostaining was intended to reduce cell loss normally encountered with centrifugation and expedite the process. Second, it is also evident that the anti-BtCYCL1 produced in this study had reasonably high affinity to the cyclin B–like protein in Aureococcus, as shown by both the Western blot and immunostaining. Third, the anti-BtCYCL1 did not cross-react with bacteria that are commonly associated with cultures of this alga. The gene sequence will provide a basis for designing a specific gene probe for RNA hybridization or primers for RT-PCR for field-collected samples, and the antiserum may prove particularly useful for field studies on this brown tide alga.

A CKNOWLEDGMENTS We thank Ms. Pilar Heredia for technical support. This research is funded by New York Sea Grant and National Science Foundation grant OCE9529970.

R EFERENCES Anderson, D.M., Keafer, B.A., Kulis, D.M., Water, R.M., and Nuzzi, R. (1993). An immunofluorescence survey of the brown tide chrysophyte Aureococcus anophagefferens along the northeast coast of the United States. J Plankton Res 15:563–580. Berg, G.M., Glibert, P.M., Lomas, M.W., and Burford, M.A. (1997). Organic nitrogen uptake and growth by the chrosophyte Aureococcus anophagefferens during a brown tide event. Mar Biol 129:377–387. Bricelj, V.M., and Kuenstner, S.H. (1989). Effects of the “brown tide” on the feeding, physiology and growth of bay scallops and mussels. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag. Bricelj, V.M., and Lonsdale, D.J. (1997). Aureococcus anophagefferens: causes and ecological consequences of brown tides in U.S. mid-Atlantic coastal waters. Limnol Oceanogr 42:1023–1038. Caron, D.A., Lim, E., Kunze, H., Cosper, E.M., and Anderson, D.M. (1989). Trophic interactions between nanno and microzooplankton and the “brown tide.” In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag. Carpenter, E.J., and Chang, J. (1988). Species-specific phytoplank-

Cyclin Box in Brown Tide Alga 585

ton growth rates via diel DNA synthesis cycles, I: concept of the method. Mar Ecol Prog Ser 43:105–111.

C-terminus of cyclins A and E specify CDK activation and substrate specificity. Oncogene 14:491–498.

Carpenter, E.J., Lin, S., and Chang, J. (1998). Phytoplankton growth studies by cell cycle analysis. In: Molecular Approaches to the Study of the Ocean, Cooksey, K.E. (ed.). New York: Chapman & Hall.

King, R.W., Jackson, P.K., Kirschner, M.W. (1994). Mitosis in transition. Cell 79:563–571.

Cosper, E.M., Dennison, W.C., Carpenter, E.J., Bricelj, V.M., Mitchell, J.G., Kuenstner, S.H., Colfelsh, D.C., and Dewey, M. (1987). Recurrent and persistent “brown tide” blooms perturb coastal marine ecosystem. Estuaries 10:284–290. Cosper, E.M., Dennison, W., Milligan, A.J., Carpenter, E.J., Lee, C., Holzapfel, J., and Milanese, L. (1989). An examination of the environmental factors important to initiating and sustaining “brown tide” blooms. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag. Dennison, W.C., Marshall, G.J., and Wigand, C. (1989). Effect of “brown tide” shading on eelgrass (Zostera marina L.) distributions. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag. DeYoe, H.R., Chan, A.M., and Suttle, C.A. (1995). Phylogeny of Aureococcus anophagefferens and a morphologically similar bloomforming alga from Texas as determined by 18S ribosomal RNA sequence analysis. J Phycol 31:413–418. Dzurica, S., Lee, C., Cosper, E.M., and Carpenter, E.J. (1989). Role of environmental variables, specifically organic compounds and micronutrients, in the growth of the chrysophyte Aureococcus anophageffeens. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: SpringerVerlag. Evans, T., Rosenthal, E.T., Youngblom, J., Distel, D., and Hunt, T. (1983). Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell 33:389– 396. Felsenstein, J. (1989). PHYLIP—Phylogeny Inference Package (Version 3.2). Cladistics 5:164–146. Glotzer, M., Murray, A.W., and Kischner, M.W. (1991). Cyclin is degraded by the ubiquitin pathway. Nature 349:132–137. Gobler, C., and Cosper, E.M. (1996). Stimulation of “brown tide” blooms by iron. In: Harmful and Toxic Algal Blooms. Proc. 7th Int. Conf. on Toxic Phytoplankton. Intergov. Oceanogr. Comm. UNESCO. Guillard, R.R., and Ryther, J.H. (1962). Studies of marine planktonic diatoms, I: Cyclotella nana Hustedt and Detonula confervacea (Cleve) Gran. Can J Microbiol 18:229–239. Horton, L.E., and Templeton, D.J. (1997). The cyclin box and

La Roche, J., Nuzzi, R., Waters, R., Wyman, K., Falkowski, P.G., and Wallace, M.W.R. (1997). Brown tide blooms in Long Island’s coastal waters linked to interannual variability in groundwater flow. Global Change Biol 3:397–410. Lees, E.M, and Harlow, E. (1993). Sequences within the conserved cyclin box of human cyclin A are sufficient for binding to and activation of cdc2 kinase. Mol Cell Biol 13:1194–1201. Lin, S., and Carpenter, E.J. (1998). Identification and preliminary characterization of PCNA gene in the marine phytoplankton Dunaliella tertiolecta and Isochrysis galbana. Mol Mar Biol Biotechnol 71:62–71. Lin, S., and Carpenter, E.J. (1999). Molecular ecology of marine Phytoplankton: a practical procedure. In: Proceedings of International Symposium on Progress and Prospect of Marine Biotechnology (ISPPMB’98), Xu, H-S., and Colwell, R.R. (eds.). Beijing: China Ocean Press. Lin, S., Carpenter, E.J., and Chang, J. (1996). Detection of p34cdc2- and cyclin B-like proteins in Dunaliella tertiolecta (Chlorophyceae). Mar Biol 125:603–610. Lin, S., Chang, J., and Carpenter, E.J. (1994). Detection of proliferating cell nuclear antigen analog in four species of marine phytoplankton. J Phycol 30:449–456. Magaletti, E. (1998). Detection and characterization of cell cycle– related proteins in the brown tide alga Aureococcus anophagefferens: a potential tool for growth rate estimations. M.S. thesis, State University of New York at Stony Brook. Milligan, K.L.D., and Cosper, E.M. (1994). Isolation of virus capable of lysing the brown tide microalga, Aureococcus anophagefferens. Science 266:805–807. Murray, A., and Hunt, T. (1993). The Cell Cycle: An Introduction. New York: Oxford University Press. Nigg, E.A. (1995). Cyclin-dependent protein kinases: key regulators of the eukaryotic cell cycle. BioEssays 17:471–480. Olsen, P.S. (1989). Development and distribution of a brownwater algal bloom in Barnegat Bay, New Jersey with perspective on resources and other red tides in the region. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag. Pitcher, G.C., Probyn, T.A., Pienaar, R.N., and Bernard, S. (2000). Brown tides in an embayment on the South African coast. Pre-

586 Senjie Lin et al.

sented at 9th International Conference on Harmful Algal Blooms. Hobart, Tasmania, Australia, February 2000.

summer. In: Novel Phytoplankton Blooms, Cosper, E.M., Bricelj, V.M., and Carpenter, E.J. (eds.). New York: Springer-Verlag.

Sieburth, J.McN., Johnson, P.W., and Hargraves, P.E. (1988). Ultrastructure and ecology of Aureococcus anophagefferens gen. et sp. nov. (Chrysophyceae); the dominant picoplankter during a bloom in Narragansett Bay, Rhode Island, Summer 1985. J Phycol 24:416–425.

Tracey, G.A. (1988). Feeding reduction, reproductive failure, and mortality in Mytilus edulis during the 1985 “brown tide” in Narragansett Bay, Rhode Island. Mar Ecol Prog Ser 50:73–81.

Smayda, T.J., and Villareal, T.A. (1989). The 1985 ‘brown-tide’ and the open phytoplankton niche in Narragansett Bay during

Vieira, M.E.C., and Chant, R. (1993). On the contribution of subtidal volume fluxes to algal blooms in Long Island estuaries. Estuarine Coastal Shelf Sci 36:15–29.