Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, June 1996, p. 1913–1921 0099-2240/96/$04.0010 Copyright q 1996, American Society for Microbiology

Vol. 62, No. 6

Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase Gene Expression and Diversity of Lake Erie Planktonic Microorganisms H. HOWARD XU

AND

F. ROBERT TABITA*

Department of Microbiology, The Ohio State University, Columbus, Ohio 43210-1292 Received 2 January 1996/Accepted 22 March 1996

The Calvin-Benson-Bassham reductive pentose phosphate pathway is the major pathway for organisms to fix CO2 into organic carbon. This pathway is shared by diverse organisms, from bacteria to algae to green plants, with the crucial reaction of this pathway catalyzed by the enzyme ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO). RubisCO is the most abundant protein on Earth (12) and is among the most intensively studied of enzymes, presumably because this bifunctional enzyme controls both the reduction of CO2 and the oxygenolysis of ribulose-1,5-bisphosphate. Thus, this protein and its catalytic activity have vast agricultural and environmental significance. Since RubisCO is responsible for nearly all the carbon fixed on this planet, nearly all primary production is linked to the function of this enzyme. RubisCO also plays a major role in modulating the levels of atmospheric CO2, which is thought to be a major component of global warming. RubisCO from prokaryotic and eukaryotic organisms is composed primarily of two types of subunits, large (L) and small (S) subunits in a basic hexadecameric L8S8 structure (45). The nucleotide and amino acid sequences of small subunits vary extensively among diverse organisms. Although significant dif-

ferences amongst various organisms are evident and are diagnostic of the species (46), the primary structure of the large subunit of RubisCO is relatively well conserved. Regulation of RubisCO gene expression and control of enzyme activity are exerted at different levels and by multiple mechanisms (46). Regulation of rbcL gene expression at the transcriptional level is an important and sensitive way to adapt to changes in the environment. However, merely determining the abundance of RubisCO genes in a particular ecosystem only indicates the potential capacity for transcription/translation and subsequent carbon fixation by representatives of the population. By contrast, quantification of specific rbcL transcripts (mRNA) is reflective of active cell metabolism and gene expression. Recently, Pichard and Paul (39) showed that determination of both rbcL mRNA and rbcL gene levels could be used as a specific measure of gene expression in environmental samples. This parameter, or the gene expression per gene dose, obtained by dividing the mRNA abundance by the target DNA concentration, normalizes gene expression for different target population sizes (39). Lake Erie is one of the largest lakes in the world. Like all other lakes, Lake Erie is undergoing eutrophication. This process is accelerated by human activities through inputs of nutritive materials that enrich the aquatic system. In recent years, the effect of eutrophication has been compounded by the introduction of an exotic species: the zebra mussel (8, 17). Phy-

* Corresponding author. Mailing address: Department of Microbiology, The Ohio State University, 484 W. 12th Ave., Columbus, OH 43210-1292. Phone: (614) 292-4297. Fax: (614) 292-6337. Electronic mail address: [email protected]. 1913

Downloaded from http://aem.asm.org/ on January 2, 2017 by guest

Carbon dioxide fixation is carried out primarily through the Calvin-Benson-Bassham reductive pentose phosphate cycle, in which ribulose-1,5-bisphosphate carboxylase/oxygenase (RubisCO) is the key enzyme. The primary structure of the large subunit of form I RubisCO is well conserved; however, four distinct types, A, B, C, and D, may be distinguished, with types A and B and types C and D more closely related to one another. To better understand the environmental regulation of RubisCO in Lake Erie phytoplanktonic microorganisms, we have isolated total RNA and DNA from four Lake Erie sampling sites. Probes prepared from RubisCO large-subunit genes (rbcL) of the freshwater cyanobacterium Synechococcus sp. strain PCC6301 (representative of type IB) and the diatom Cylindrotheca sp. strain N1 (representative of type ID) were hybridized to the isolated RNA and DNA. To quantitate rbcL gene expression for each sample, the amount of gene expression per gene dose (i.e., the amount of mRNA divided by the amount of target DNA) was determined. With a limited number of sampling sites, it appeared that type ID (diatom) rbcL gene expression per gene dose decreased as the sampling sites shifted toward open water. By contrast, a similar trend was not observed for cyanobacterial (type IB) rbcL gene expression per gene dose. Complementary DNA specific for rbcL was synthesized from Lake Erie RNA samples and used as a template for PCR amplification of portions of various rbcL genes. Thus far, a total of 21 clones of rbcL genes derived from mRNA have been obtained and completely sequenced from the Ballast Island site. For surface water samples, deduced amino acid sequences of five of six clones appeared to be representative of green algae. In contrast, six of nine sequenced rbcL clones from 10-m-deep samples were of chromophytic and rhodophytic lineages. At 5 m deep, the active CO2-fixing planktonic organisms represented a diverse group, including organisms related to Chlorella ellipsoidea, Cylindrotheca sp. strain N1, and Olisthodiscus luteus. Although many more samplings at diverse sites must be accomplished, the discovery of distinctly different sequences of rbcL mRNA at different water depths suggests that there is a stratification of active CO2-fixing organisms in western Lake Erie.

1914

XU AND TABITA

APPL. ENVIRON. MICROBIOL.

toplanktonic microorganisms constitute one of the main groups of primary producers in Lake Erie (5); however, in recent years, the number of such organisms has been shown to be decreasing significantly in the Western Basin of Lake Erie, because of a reduction of phosphorus loading and the introduction of the zebra mussel (29, 33). Although there have been thorough surveys of the abundance and diversity of the phytoplankton in Lake Erie (29), little is known of how the CO2 fixation machinery operates and responds to environmental changes in this and other aquatic ecosystems. Reverse transcription-PCR (RT-PCR) has been used extensively to identify viral particles in vivo (7, 23, 31) and in various natural environments (27, 31, 48–50). However, RT-PCR has not been widely used to study transcriptional regulation in natural populations. Recently, RT-PCR was used to quantitate fungal mRNAs in Phanerochaete chrysosporium-colonized soil (26), and Song and Osborn combined RT-PCR with restriction fragment length polymorphism to examine the expression of homologous genes in polyploid plants (44). Finally, most relevant to the current study, Pichard et al. (36) used RT-PCR and sequence analysis to examine taxonomically diverse phytoplanktonic communities in natural oceanic environments. Thus, the objective of this study was to determine rbcL gene expression per gene dose for rbcL genes from two groups of CO2-fixing organisms, representing organisms containing types IB and ID RubisCO, at various depths at different sites in Lake Erie. On the basis of partial rbcL gene sequences derived from RT-PCR amplification of rbcL mRNA from natural samples, we also attempted to identify and categorize individual organisms actively involved in CO2 fixation.

MATERIALS AND METHODS Sampling procedures. Samples were collected at four stations in the western basin of Lake Erie (Fig. 1). Station G was near Gibraltar Island (41839.509N, 82849.109W), station B was north of Ballast Island (41841.349N, 82847.329W), station M was outside Marblehead (41833.749N, 82845.099W), and station S was inside Sandusky Bay (41828.919N, 82845.649W). These stations were selected because they represent different lake environments. Sandusky Bay acts as a reservoir for large amounts of agricultural runoff and municipal discharges. It is also a largely enclosed water body, where little mixing with water of other parts of the lake occurs; in the absence of seiche activity (standing waves), water flows unidirectionally into the main body of Lake Erie. Marblehead station is a relatively shallow nearshore environment. Ballast Island station is one of the deepest sites in the western basin of Lake Erie. The Gibraltar Island site is near the boat dock facing Put-in-Bay; extensive water plant growth was evident at this site. Surface and subsurface water samples (0 to 10 m deep) were collected with a 4-gal (15-liter) Wildco water sampler (Wildlife Supply Co., Saginaw, Mich.). The water temperature was recorded immediately for each sample collected. All sites were isothermal between different depths, indicating lack of physical stratification. RNA extraction. Reagents were prepared in glassware that had been baked at 2608C for 5 h. All solutions were made with distilled water treated with diethylpyrocarbonate (Sigma Chemical Co., St. Louis, Mo.). Lake water samples were filtered onto 47-mm Millipore Durapore membranes (pore size, 0.22 mm) with a Nalgene filtering apparatus. Four or five membranes were used per sample to collect enough biomass for RNA extraction and gene expression analysis. The membranes were then cut into small pieces with a pair of baked scissors; threaded membranes were placed into a baked 25-ml Corex heavy-duty centrifuge tube (with a screw cap). A 5-ml volume of GIPS extraction reagent (4 M guanidine isothiocyanate [GIBCO-BRL, Gaithersburg, Md.], 0.5% Sarkosyl, 25 mM sodium acetate [pH 7.0], 0.1 M 2-mercaptoethanol) (39) was added to the centrifuge tube, and the tube was immediately capped and frozen in a dry-ice– ethanol bath for 15 min. All frozen samples were kept on dry ice throughout the sampling trip until they were returned to the laboratory for further processing. They were then thawed at 508C for 10 min and then refrozen in a dry-ice–ethanol bath. The freeze-thaw cycle was repeated two more times to help break the cells of microorganisms. The material was extracted with an equal volume of acidic buffer-saturated phenol (pH 4.3; AMRESCO, Solon, Ohio) at 658C, and the extraction mixture was incubated in a 658C bath for 5 min. A half volume


FIG. 1. Sampling stations of the western basin of Lake Erie.

VOL. 62, 1996

1915

428C for 60 min. The reaction was stopped by adding 80 ml of Tris-EDTA (TE) buffer. For PCR amplification of rbcL cDNA, 2 ml of RT mixture was combined with 10 ml of 103 PCR buffer (U.S. Biochemicals), 0.01% gelatin, 200 mM each dNTP, 1 mM primer K, 1 mM primer V, and 0.5 ml (2.5 U) of Taq polymerase (U.S. Biochemicals) in a final volume of 100 ml. The contents of the reaction vial were thoroughly mixed, centrifuged, and overlaid with 100 ml of mineral oil. Amplification was performed in a DNA thermal cycler (Perkin-Elmer Corp., Norwalk, Conn.) with 39 cycles of 1 min at 948C, 2 min at 508C, and 1 min at 728C. These cycles were followed by a 20-min incubation at 728C. PCR products were electrophoresed on 0.9% agarose gels and stained with ethidium bromide solution for visualization. The expected size of the PCR-amplified rbcL gene is 0.5 kb. The amplified products were transferred to a GeneScreen Plus membrane and probed with an rbcL gene probe to confirm the identity of the PCR products. Cloning of PCR amplified rbcL gene fragments. PCR products were extracted with phenol and chloroform once each and precipitated with ethanol. One-fifth of the PCR products were digested with EcoRI and BamHI and ligated with EcoRI-BamHI-digested pK18 (40). E. coli JM109 competent cells were transformed with the ligation mixtures. Recombinant colonies were picked, and plasmid DNA was isolated from each clone. Restriction endonuclease digestion was performed to confirm the existence of a 0.5-kb insert in each recombinant plasmid. DNA sequencing and phylogenetic analysis. Plasmids containing PCR-amplified inserts were used directly for sequencing. Plasmids were purified with either CsCl2 gradients (43) or QIAGEN-tip 20 (QIAGEN Inc., Chatsworth, Calif.) as specified by the manufacturer. The inserts were completely sequenced from both DNA strands by the Taq Dye-Deoxy primer and terminator cycle sequencing kits (Applied Biosystems, Inc., Foster City, Calif.). Sequence reaction mixtures were electrophoresed with an Applied Biosystems model 373 DNA sequencer. Nucleotide sequence data were analyzed by the sequence analysis programs of the University of Wisconsin Genetics Computer Group, as well as by MacVector sequence analysis software (International Biotechnologies, Inc., New Haven, Conn.). Amino acid sequence data were aligned by using the CLUSTAL W package (47). A phylogenetic tree was constructed by the neighbor-joining program of the PHYLIP package (version 3.5c) (13). Bootstrap analyses for 100 replicates (13) were performed to provide confidence estimates for tree topology. Nucleotide sequence accession numbers. Lake Erie Ballast Island rbcL partial cDNA clones and corresponding GenBank accession numbers are as follows: 0m1, U42639; 0m10, U43056; 0m2, U43057; 0m3, U43058; 0m4, U43059; 0m6, U43060; 10m1, U43061; 10m10, U43062; 10m11, U43063; 10m12, U43064; 10m14, U43065; 10m2, U43066; 10m3, U43067; 10m7, U43068; 10m9, U43069; 5m1, U43070; 5m2, U43071; 5m26, U43072; 5m4, U43073; 5m6, U43074; 5m7, U43075.

RESULTS rbcL gene expression per gene dose. All sampling sites were isothermal, indicating that western Lake Erie was not physically stratified. To estimate rbcL gene expression per gene dose among two broad groups of CO2 fixation organisms, blotted total RNA and DNA were separately hybridized with antisense RNA probes prepared from rbcL genes of Cylindrotheca sp. strain N1 (representative of type ID RubisCO) and Synechococcus sp. strain PCC 6301 (representative of type IB RubisCO), respectively (Fig. 2). The hybridization conditions were such that the rbcL DNA from Cylindrotheca sp. strain N1 did not hybridize with the rbcL antisense RNA probe of Synechococcus sp. strain PCC 6301, and vice versa (Fig. 3). Similarly, rbcL mRNA from Cylindrotheca sp. strain N1 did not cross-react with an rbcL antisense RNA probe of Synechococcus sp. strain PCC 6301, and vice versa (data not shown). Quantitative analysis (Fig. 4A) indicated that the Marblehead 0- and 4-m water samples appeared to contain the highest levels of cyanobacterium-type and diatom-type rbcL, respectively, based on the level of rbcL DNA contents. In all samples, cyanobacteriumtype rbcL DNA and mRNA were generally more abundant than those of diatom-type rbcL (Fig. 4A and B). Diatom-type rbcL gene expression per gene dose was highest in Sandusky Bay, and this parameter decreased as the sampling sites moved toward open waters (Fig. 4C). Cyanobacterium-type rbcL gene expression per gene dose was similar at the Sandusky Bay (S) and Marblehead (M) sites, regardless of depth, but increased significantly with depth at the Ballast Island (B) site (Fig. 4C). Both diatom-type and cyanobacterium-type rbcL gene expres-


(compared with the volume of phenol used) of chloroform-isoamyl alcohol (24:1) was added to the extraction mixture to help form two phases. The tube was centrifuged for 15 min at 10,000 rpm (12,000 3 g). The aqueous phase was carefully saved and similarly extracted one more time. The resulting aqueous phase was then extracted once with chloroform-isoamyl alcohol (24:1). Onetenth volume of 3 M potassium acetate (pH 5.5) and 100 mg of glycogen (Boehringer Mannheim, Indianapolis, Ind.) were added to the final aqueous phase, and nucleic acids were precipitated overnight with 2.5 volumes of cold ethanol. The nucleic acids were collected by centrifugation at 15,000 rpm (27,000 3 g) for 20 min. The pellet was rinsed once with cold 70% ethanol, and the dried pellet was dissolved in 1 ml of 50 mM Tris-HCl (pH 7.8)–5 mM MgCl2. RNase-free DNase I (5 U) (Ambion, Inc., Austin, Tex.) was added to the nucleic acid solution, and the mixture was incubated at 378C for 25 min. The contents were then extracted once with acidic phenol and once with chloroform-isoamyl alcohol (24:1). RNA was precipitated similarly overnight, centrifuged, and then dried. The precipitate was dissolved in diethylpyrocarbonate-treated water (in 1/10,000 of the original water volume). DNA extraction. To prepare a total-DNA sample, equal amounts of water from the same sample used for the RNA extraction were filtered on four or five Durapore membranes, as described above. Threaded membranes were immediately immersed in 4.5 ml of DNA extraction buffer (20 mM Tris-HCl [pH 8.0], 100 mM NaCl, 50 mM EDTA) (22) and frozen in a dry-ice–ethanol bath for 15 min. These samples were kept on dry ice until they were returned to the laboratory for DNA extraction, using a protocol modified from that of Jacobs et al. (22). Each tube was thawed at 508C for 10 min. Then, 250 ml of 20% sodium dodecyl sulfate (SDS) was added, the solution was mixed, and 250 ml of proteinase K (10 mg/ml) was added. The mixture was incubated for a further 10 min at 508C and frozen again for 15 min in a dry-ice–ethanol bath. The freeze-thaw procedure was repeated two more times, and the thawed mixture was extracted twice with buffered phenol (pH 8.0) and finally once with chloroform-isoamyl alcohol (24:1). Nucleic acids were precipitated by the addition of 1/10 volume of 3 M potassium acetate (pH 5.5) and 2 volumes of ethanol and were then dried and dissolved in distilled water. All DNA solutions were treated with RNase and extracted with phenol-chloroform. The DNA was precipitated and dissolved in distilled water as described above. Preparation of radiolabeled antisense RNA probes and rbcL mRNA and DNA standards. 35S-radiolabeled antisense RNA probes were synthesized from plasmids pRLD1 (containing the Cylindrotheca sp. strain N1 rbcL gene) and pLC1 (containing the Synechococcus sp. strain PCC 6301 rbcL gene) (38) with a MAXIscript in vitro transcription kit (Ambion Inc.). Plasmids pRLD1 and pLC1 contain rbcL genes on riboprobe vector pGEM-3Z (Promega Corp., Madison, Wis.) from which single-stranded RNA probes or RNA transcripts may be conveniently prepared. By using the same in vitro transcription kit, unlabeled rbcL mRNAs from both Cylindrotheca sp. strain N1 and Synechococcus sp. strain PCC 6301 were also synthesized to construct mRNA standard curves. To obtain rbcL DNA, plasmids pRLD1 and pLC1 were digested with appropriate restriction endonucleases and rbcL DNA fragments were isolated and purified from an agarose gel by Geneclean (Bio101, Inc., Vista, Calif.) protocols following electrophoretic separation of the DNA fragments. Both mRNA and DNA were quantified with a DNA fluorometer (TK0100; Hoefer Scientific Instruments, San Francisco, Calif.). Nucleic acid blotting and hybridization. Denatured RNA and DNA samples, as well as mRNA and DNA standards, were vacuum blotted (in duplicate) onto GeneScreen Plus membranes (NEN/Du Pont) with Bio-Dot slot format blotting apparatus (Bio-Rad, Hercules, Calif.) as specified by the manufacturer. Blotted membranes were baked for 30 min at 808C. Cylindrotheca sp. strain N1 and Synechococcus sp. strain PCC 6301 rbcL antisense probes (about 106 cpm) were used to probe the abundance of rbcL mRNA in RNA samples and the amount of rbcL DNA in DNA samples blotted to the membranes. Hybridizations were carried out overnight at 428C in 50% formamide–63 SSPE (13 SSPE is 150 mM NaCl, 10 mM NaH2PO4 z H2O, plus 1 mM EDTA)–53 Denhardt’s solution– 0.5% SDS. The blots were washed twice in 23 SSPE–0.5% SDS at 588C for 15 min. The final wash was carried out in 0.13 SSPE–0.5% SDS at 588C for 30 min. Radioactive rbcL mRNA and DNA signals were quantified with an Instant Imager 2024 (Packard Instrument Co., Meriden, Conn.). rbcL mRNA and DNA levels from samples were obtained from rbcL mRNA and DNA standard curves. RT-PCR amplification of rbcL mRNA of Lake Erie samples. rbcL mRNA was amplified from total RNA samples from Lake Erie by the method of BeckerAndre and Hahlbrock (2) with degenerate oligonucleotide primers designed on the basis of two highly conserved domains (amino acid stretches KPKLGLS and VVGKLEG) of RubisCO. The primer corresponding to amino acids KPKLGLS (primer K) is 59-GCGAATTCAA(AG)CC(TA)AA(AG)(TC)TAGG(TG)(CT) T(AT)TC-39, and the primer corresponding to amino acids VVGKLEG (primer V), is 59-AGGGATCC(TC)TC(TC)A(AG)(TC)TTACC(AT)AC(GAT)AC-39. EcoRI and BamHI restriction sites were added to the 59 ends of primers K and V, respectively, to aid the cloning process. Conversion of rbcL mRNA to cDNA was carried out in a final volume of 20 ml containing 2 mg or less of total RNA, 4 ml of 53 RT buffer (Boehringer Mannheim), 100 ng of primer V, 0.5 mM each deoxynucleoside triphosphate (dNTP), 12.5 U of RNasin (Ambion), and 24 U of avian myeloblastosis virus reverse transcriptase (Boehringer Mannheim). Before addition of RNasin and reverse transcriptase, the reaction mixture was heated at 658C for 5 min and cooled slowly to room temperature. RT was carried out at

ENVIRONMENTAL rbcL GENE EXPRESSION

1916

XU AND TABITA


sions per gene dose were lowest at the Gibraltar Island (G) site (Fig. 4C). RT-PCR of rbcL mRNA. The previous experiments, which indicated the amount of mRNA expressed per gene dose, broadly indicated which group(s) of organisms was actively engaged in CO2 fixation and at what level of activity. Not surprisingly, it was very difficult to determine which species were dominant in a particular environment solely on the basis of hybridization experiments. Therefore, RT-PCR was used in an attempt to obtain double-stranded DNA fragments derived from actively transcribed rbcL genes in the natural environments examined in this study. Theoretically, complete rbcL PCR products defined by the primers would yield fragments of either 502 or 499 bp (chromophytic and rhodophytic RubisCO large subunits contain one less amino acid than do large subunits of cyanobacteria, green algae, and higher plants within

FIG. 3. Hybridization of Cylindrotheca sp. strain N1 rbcL DNA (C) and Synechococcus sp. strain PCC 6301 rbcL DNA (S) with the Cylindrotheca sp. strain N1 rbcL antisense RNA probe and Synechococcus sp. strain PCC 6301 rbcL antisense RNA probe. The amount of rbcL DNA blotted is indicated on the left.

the amplified region [see below]). As expected, agarose gel electrophoresis of RT-PCR products (Fig. 5A) indicated that the major DNA fragments were about 500 bp. Southern hybridization analysis with the diatom Cylindrotheca sp. strain N1 rbcL antisense RNA probe confirmed that these 500-bp bands were indeed rbcL-specific DNA fragments (Fig. 5B). Sequence analysis of PCR-amplified rbcL genes. To determine the identities of individual species performing CO2 fixation, RT-PCR-generated fragments from each sample were separately cloned. Thus far, a total of 21 RT-PCR-amplified rbcL gene fragments from the Ballast Island station have been completely sequenced. Six clones were each derived from 0and 5-m depths, and nine clones were derived from a depth of 10 m at this site. All these sequences may be classified into two groups: those of 450 bp and those of 447 bp, as defined by primer locations. Interestingly, the first group contains all green algal rbcL sequences, as indicated (see Fig. 6 and 7). Some sequences were nearly identical: clones 10m11 and 0m2 differed by only one nucleotide; clones 0m6 and 0m4 differ by two nucleotides; clones 10m11 and 10m7 differ by three nucleotides. The second group includes sequences that align well with the large subunit of the marine diatom Cylindrotheca sp. strain N1 or the multicellular red alga, Antithamnion sp. Two clones (10m1 and 10m2) were identical; clone 10m9 was nearly identical to 10m1 and 10m2, with only two nucleotides different; and clones 10m3 and 5m4 differ by one nucleotide. Deduced amino acid sequences and phylogenetic analysis. Deduced amino acid sequences of 21 partial rbcL genes from Lake Erie and of several representative known rbcL genes were aligned (Fig. 6). It is apparent that, just like the nucleotide sequences, 21 deduced amino acid sequences from Lake Erie rbcL clones and 13 other representative form I rbcL deduced amino acid sequences can be initially divided into two groups based on the length of the amino acid sequence defined by the two primers (Fig. 6). The top 18 sequences (from Spinacia oleracea to Thiobacillus ferrooxidans) all contain 150 amino acids within this stretch, while the bottom 16 sequences


FIG. 2. Nucleic acid hybridization analysis with rbcL probes. (A) Hybridization with the Cylindrotheca sp. strain N1 rbcL antisense RNA probe. Total RNA (2 ml) from each sample was vacuum blotted onto a membrane together with Cylindrotheca sp. strain N1 rbcL mRNA standards. Total DNA (20 ml) from each sample was also blotted, along with Cylindrotheca sp. strain N1 rbcL DNA standards. (B) Hybridization with the Synechococcus sp. strain PCC 6301 rbcL antisense RNA probe. Total RNA (2 ml) from each sample was vacuum blotted onto a membrane together with Synechococcus sp. strain PCC 6301 rbcL mRNA standards. Total DNA (20 ml) from each sample was also blotted along with Synechococcus sp. strain PCC 6301 rbcL DNA standards. Quantities of all standard rbcL mRNA and DNA are as follows (top to bottom): 10, 5, 2.5, 1.25, 0.625, 0.3125, 0.15, and 0.075 ng. Sample designations: S0 and S3, Sandusky Bay station 0- and 3-m-deep samples; M0 and M4, Marblehead station 0- and 4-m-deep samples; B0, B5, and B10, Ballast Island station 0-, 5-, and 10-m-deep samples; G0, Gibraltar Island station 0-m-deep sample.

VOL. 62, 1996


1917

diverse group of microorganisms, namely, organisms related (although sometimes distantly) to Chlorella ellipsoidea, Cylindrotheca sp. strain N1, Olisthodiscus luteus, and Antithamnion sp. DISCUSSION

(from Rhodobacter sphaeroides I to Antithamnion sp.) all contain 149 amino acids. This division of two groups of largesubunit sequences can also be seen on the dendrogram (Fig. 7). Clones 10m14 and 5m26 are closest to the position of diatom (Cylindrotheca sp. strain N1) RbcL. Five clones from 10 m deep (10m1, 10m2, 10m9, 10m10, and 10m3) and one clone from 5 m deep (5m4) might belong to other chromophytic and rhodophytic organisms (Fig. 7). Clones 0m1 and 5m6 appeared to be distantly related to RbcLs of rhodophytes, prymnesiophytes, and cryptophytes. On the other hand, three clones (0m2, 10m11, and 10m7) clustered closely with RbcL of the green alga Chlamydomonas reinhardtii. Two 5-m-deep clones (5m7 and 5m1) branched out from Chlorella ellipsoidea RbcL (Fig. 7). Five clones (10m12, 0m3, 0m6, 0m4, and 0m10) might thus be related to other chlorophytes. From an ecological standpoint, it may be summarized that five of the six clones from the 0-m-deep sample were chlorophytes while six of nine clones from the 10-m-deep sample were from organisms of chromophytic and rhodophytic lineage. At 5 m deep, the transcriptionally active planktonic community consisted of a rather


FIG. 4. Quantitation of rbcL DNA and mRNA and rbcL gene expression per gene dose for each sample. (A) rbcL DNA concentrations for both diatom-type and cyanobacterium-type organisms; (B) rbcL mRNA concentrations for both diatom-type and cyanobacterium-type organisms; (C) rbcL gene expression per gene dose (ratio of mRNA to DNA). Symbols: f, diatom type; o, cyanobacterium type.

Pichard and Paul (39) were the first to report the detection of gene expression in an indigenous community by mRNA analysis. They combined guanidine isothiocyanate RNA extraction with the use of antisense and sense RNA probes to analyze gene expression of the neomycin phosphotransferase gene (nptII) in a simulated microcosm and the RubisCO largesubunit gene (rbcL) in a natural phytoplankton community. Light-responsive rbcL expression was observed in this phytoplankton community. Pichard et al. developed the concept of gene expression per gene dose as a specific measurement of gene expression in the environment (37, 39). These authors used the target gene dose as a normalizing factor for mRNA quantification from environmental samples, and this concept was used to monitor the regulated expression of the catechol2,3-dioxygenase gene (xylE) contained on a thermoregulated plasmid in a marine Vibrio strain in a simulated environment (39). When rbcL gene expression of oceanic phytoplankton was examined, a trend was observed such that the levels of rbcL gene expression (mRNA/DNA ratio) decreased between three- and eightfold from estuarine to offshore environments. In an offshore vertical profile, the subsurface maximum in the rbcL mRNA/DNA ratio coincided with the 60-m maximum in photosynthetic assimilation rates. In a simulated incubator, rbcL gene expression in the light was five times higher than in the dark (37). On the basis of sequence homologies, form I RubisCO proteins from various sources may be grouped into four main evolutionary lineages, two of which, types A and B and types C and D, are more closely related to each other than to members of the opposite pair (46). However, practical use of knowledge of the evolutionary lineage of natural populations of CO2fixing organisms has not been fruitful. In the present study, we have attempted to use the gene-expression-per-gene-dose concept to analyze rbcL expression in a large freshwater ecosystem (Lake Erie), a system known to undergo fluctuations in phytoplankton appearance (29, 32). We have shown experimentally that purified rbcL genes and mRNA from type ID did not hybridize to rbcL gene probes from a representative of type IB (Fig. 3) (RNA data not shown). Moreover, the two rbcL probes from representative species of these two groups (e.g., Cylindrotheca sp. strain N1 and Synechococcus sp. strain PCC 6301) could definitely be used to estimate rbcL gene expression per gene dose in Lake Erie environments. Phosphate has been implicated as a limiting nutrient for freshwater phytoplankton production (6, 42). There is evidence that Lake Erie has reversed its course of eutrophication and has become mesotrophic in recent years. The total chlorophyll concentrations in the western basin of Lake Erie in 1991 were typical of oligotrophic areas of the upper Great Lakes (33). In addition, the phosphate concentration has decreased (3). The hybridization results indicated that the diatom (type ID) rbcL gene expression per gene dose decreased as sampling sites moved from Sandusky Bay toward open-water environments. It is thus possible that phosphate is a limiting factor in offshore areas of Lake Erie for certain groups of organisms, such as diatoms. Our results, which indicated that the Gibraltar Island site showed the lowest rbcL gene expression for both types of organisms, may be partially explained by the fact that the sample was collected near a boat dock, where

1918

XU AND TABITA


excessive water plant growth was visible. Perhaps the heavy vegetation produced a shading effect or competed for nutrients with the planktonic microorganisms present in this region. Nucleotide sequence analysis and phylogenetic analysis of rbcL gene clones from the rbcL mRNA of the Ballast Island site suggests that within the detection limits of the current assay, CO2 fixation in the surface water was catalyzed predominantly by green algae whereas CO2 fixation in the deeperwater zone (10 m deep) was performed mainly by diatoms. Intermediate water zones (5 m deep) appeared to consist of diverse groups of microorganisms; present in this zone were organisms which possessed rbcL sequences that were very similar to those of Cylindrotheca sp. strain N1, the green alga Chlorella ellipsoidea, the chromophyte Olisthodiscus luteus, and, to a lesser extent, rhodophytes, prymnesiophytes, and cryptophytes (Fig. 7). These results suggest that the 5-m-deep zone may be a transition zone between surface water and deep water. These results are extremely interesting, because a similar “partition” of cyanobacterium-type and diatom-type rbcL gene expression was also observed between shallow- and deepwater zones in marine environments (35). Unlike marine offshore environments, the western basin of Lake Erie is generally not stratified year-round (1, 5) and the water column was isothermal from 0- to 10-m depths at the Ballast Island station. Theoretically, planktonic community compositions at a particular site should be very similar because of the mixing of the water column. Here we present evidence of a functional stratification, i.e., CO2 fixation conducted by different groups of organisms at different depths, in western Lake Erie. The finding that different groups of microorganisms were involved in CO2 fixation with respect to water depth in western Lake Erie (as well as in marine offshore environments) may be attributable to differences in light-harvesting pigments, which preferentially absorb photons of different wavelength ranges, among these groups of organisms. Further research is needed to test this hypothesis, and, certainly, many more samplings at different sites are necessary before the aforementioned results may be considered a trend.

Also significant in this investigation was the identification of partial rbcL genes from several apparently different species of chromophytic or rhodophytic algae from the Ballast Island station. Unlike land plants, green algae, and photosynthetic bacteria, the complete RubisCO gene sequences from chromophytes and rhodophytes are less abundant (4, 10, 14, 15, 20, 24). It appeared that in Lake Erie, a diverse group of chromophytic and rhodophytic algae were actively involved in primary production and that several different strains of the same species (or several closely related species), represented by clones 10m1, 10m2, 10m9, 10m10, 10m3, 10m14, 5m4, and 5m26 (Fig. 7), were the dominant CO2 fixers at the subsurface water zone. The availability of large numbers of chromophytic or rhodophytic rbcL sequences, albeit partial, enabled us to analyze homologies of functional peptide domains for this group of organisms. Two stretches of RubisCO amino acid sequence regions that were implicated as parts of the active site (18, 21, 28) were well conserved among both representative species and the newly discovered Lake Erie clones (Fig. 6). Particularly interesting is that 9 of the 11 clones of the chromophytic or rhodophytic lineage contained Asn (N) instead of His (H) at position 118 (Fig. 6), the same position that was previously found to be an exception (20). The presence of histidine at this position in large subunits of 2 of the 11 clones supports the suggestion that this residue is not important in 3-phosphoglycerate binding (20), at least in chromophytes and rhodophytes. In conclusion, the present study illustrates the application of two different rbcL gene probes for the determination of RubisCO gene expression in Lake Erie and demonstrates the successful use of RT-PCR to obtain partial nucleotide sequences of actively transcribed genes (rbcL) of diverse evolutionary lineages in natural freshwater environments. The eventual cultivation of the dominant CO2-fixing species and the cloning and sequencing of the complete rbcL genes of these species will be important to further understand the environmental regulation of RubisCO gene expression and should contribute to an understanding of the population dynamics important for CO2 fixation in a given environment.


FIG. 5. RT-PCR of Lake Erie rbcL mRNA. (A) Agarose gel electrophoresis of PCR amplification products. Lanes: 1, PCR amplification of B5 total-RNA sample (2 ml) as a negative control; 2 through 9, RT-PCR amplification products of total RNA from G0, B0, B5, B10, M0, M4, S0, and S3 samples, respectively. (B) Autoradiograph of Southern hybridization of PCR products transferred to GeneScreen Plus membranes. The blot was hybridized with the Cylindrotheca sp. strain N1 rbcL antisense RNA probe prepared as described in Materials and Methods. Lad represents a 1-kb ladder (GIBCO-BRL, Gaithersburg, Md.).

VOL. 62, 1996


1919


FIG. 6. Amino acid sequence alignment of transcriptionally active rbcL (cbbL) genes from the Ballast Island site in Lake Erie with corresponding regions of amino acid sequences of representative rbcL genes. 0mN, 5mN, and 10mN represent clones obtained from 0-, 5-, and 10-m depths, respectively. R. sphaeroides I, Rhodobacter sphaeroides form I RbcL (16); X. flavus, Xanthobacter flavus RbcL (30); Cylindrotheca N1, Cylindrotheca sp. strain N1 RbcL (19); O. luteus, Olisthodiscus luteus RbcL (4); Antithamnion, Antithamnion sp. strain RbcL (24); S. oleracea, Spinacia oleracea RbcL (54); T. aestivum, Triticum aestivum RbcL (34); C. ellipsoidea, Chlorella ellipsoidea RbcL (53); C. reinhardtii, Chlamydomonas reinhardtii RbcL (11); Anabaena PCC7120, Anabaena sp. strain PCC 7120 RbcL (9); Synechococcus 6301, Synechococcus sp. strain PCC 6301 RbcL (41); C. paradoxa, Cyanophora paradoxa RbcL (51); T. ferrooxidans, Thiobacillus ferrooxidans RbcL (25). Two stretches of amino acids implicated in catalysis are indicated by Ç. Residue 118 is marked by {.

1920

XU AND TABITA

ACKNOWLEDGMENTS We are grateful to S. L. Pichard and J. H. Paul for providing plasmids pRLD1 and pLC1, for designing the oligonucleotide primers used for PCR amplification of rbcL mRNA, and for instructing us in the sampling and preparative procedures of environmental samples. We thank D. A. Culver for valuable discussions, the Biochemical Instrument Center for use of the Applied Biosystems model 373 DNA sequencer, and C. Daniels for use of the radioactivity imaging system (Instant Imager 2024). G. Watson’s help with ClustalW and PHYLIP programs was invaluable and is greatly appreciated. This work was supported by NSF and DOE grants OCE-9218517 and DE-FG02-93ER61700 (F.R.T.), the Ohio Sea Grant Development Fund (F.R.T. and H.H.X.), and an appointment to the Global Change Distinguished Postdoctoral Fellowships (H.H.X.) sponsored by the U.S. Department of Energy, Office of Health and Environmental Research, and administered by the Oak Ridge Institute for Science and Education. REFERENCES 1. Bartish, T. M. 1984. Thermal stratification in the Western Basin of Lake Erie: its characteristics, mechanisms of formation, and chemical and biological consequences. M.S. thesis. The Ohio State University, Columbus. 2. Becker-Andre, M., and K. Hahlbrock. 1989. Absolute mRNA quantification using the polymerase chain reaction (PCR). A novel approach by a PCR

aided transcript titration assay (PATTY). Nucleic Acids Res. 17:9437–9446. 3. Bertram, P. E. 1993. Total phosphorus and dissolved oxygen trends in the central basin of Lake Erie, 1970–1991. J. Gt. Lakes Res. 19:224–236. 4. Boczar, B. A., T. P. Delaney, and R. A. Cattolico. 1989. Gene for the ribulose1,5-bisphosphate carboxylase small subunit protein of the marine chromophyte Olisthodiscus luteus is similar to that of a chemoautotrophic bacterium. Proc. Natl. Acad. Sci. USA 86:4996–4999. 5. Bolsenga, S., and C. E. Herdendorf. 1993. Lake Erie and Lake St. Clair handbook. Wayne State University, Detroit. 6. Chang, C. C. Y., J. S. Kuwabara, and S. P. Pasilis. 1992. Phosphate and iron limitation of phytoplankton biomass in Lake Tahoe. Can. J. Fish. Aquat. Sci. 49:1206–1215. 7. Chen, J., S. A. Macfarlane, and T. M. Wilson. 1995. An analysis of spontaneous deletion sites in soil-borne wheat mosaic virus RNA2. Virology 209: 213–217. 8. Cooley, J. M. 1991. Zebra mussels. J. Gt. Lakes Res. 17:1–2. 9. Curtis, S. E., and R. Haselkorn. 1983. Isolation and sequence of the gene for the large subunit of ribulose-1,5-bisphosphate carboxylase from the cyanobacterium Anabaena 7120. Proc. Natl. Acad. Sci. USA 80:1835–1839. 10. Douglas, S. E., D. G. Durnford, and C. W. Morden. 1990. Nucleotide sequence of the gene for the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase from Cryptomonas f: evidence supporting the polyphyletic origin of plastids. J. Phycol. 26:500–508. 11. Dron, M., M. Rahire, and J.-D. Rochaix. 1982. Sequence of the chloroplast DNA region of Chlamydomonas reinhardii containing the gene of the large subunit of ribulose bisphosphate carboxylase and parts of its flanking genes. J. Mol. Biol. 162:775–793. 12. Ellis, R. J. 1979. The most abundant protein in the world. Trends Biochem. Sci. 4:241–244. 13. Felsenstein, J. 1993. PHYLIP (Phylogeny Inference Package), version 3.5c. Department of Genetics, University of Washington, Seattle. 14. Freshwater, D. W., S. Fredericq, B. S. Butler, M. H. Hommersand, and M. W. Chase. 1994. A gene phylogeny of the red algae (Rhodophyta) based on plastic rbcL. Proc. Natl. Acad. Sci. USA 91:7281–7285. 15. Fujiwara, S., H. Iwahashi, J. Someya, S. Nishikawa, and N. Minaka. 1993. Structure and cotranscription of the plastid-encoded rbcL and rbcS genes of Pleurochrysis carterae (Preymnesiophyta). J. Phycol. 29:347–355. 16. Gibson, J. L., D. L. Falcone, and F. R. Tabita. 1991. Nucleotide sequence, transcriptional analysis, and expression of genes encoded within the form I CO2 fixation operon of Rhodobacter sphaeroides. J. Biol. Chem. 266:14646– 14653. 17. Griffiths, R. W., D. W. Schloesser, J. H. Leach and W. P. Kovalak. 1991. Distribution and dispersal of the zebra mussel (Dreissena polymorpha) in the Great Lakes region. Can. J. Fish. Aquat. Sci. 48:1381–1388. 18. Hartman, F. C., and E. H. Lee. 1989. Examination of the function of active site lysine 329 of ribulose-bisphosphate carboxylase/oxygenase as revealed by the proton exchange reaction. J. Biol. Chem. 264:11784–11789. 19. Hwang, S.-R., and F. R. Tabita. 1989. Cloning and expression of the chloroplast-encoded rbcL and rbcS genes from the marine diatom Cylindrotheca sp. strain N1. Plant Mol. Biol. 13:69–79. 20. Hwang, S.-R., and F. R. Tabita. 1991. Cotranscription, deduced primary structure, and expression of the chloroplast-encoded rbcL and rbcS genes of the marine diatom Cylindrotheca sp. strain N1. J. Biol. Chem. 266:6271–6279. 21. Igarashi, Y., B. A. McFadden, and T. El-Gul. 1985. Active site histidine in spinach ribulosebisphosphate carboxylase/oxygenase modified by diethyl pyrocarbonate. Biochemistry 24:3957–3962. 22. Jacobs, J. D., J. R. Ludwig, M. Hildebrand, A. Kukel, T.-Y. Feng, R. W. Ord, and B. E. Volcani. 1992. Characterization of two circular plasmids from the marine diatom Cylindrotheca fusiformis: plasmids hybridize to chloroplast and nuclear DNA. Mol. Gen. Genet. 233:302–310. 23. Kariwa, H., M. Kamimura, J. Arikawa, K. Yoshimatsu, I. Takashima, and N. Hashimoto. 1995. Characterization of the mode of Hantaan virus infection in adult mice using a nested reverse transcriptase polymerase chain reaction: transient virus replication in adult mice. Microbiol. Immunol. 39:35–41. 24. Kostrzewa, M., K. Valentin, U. Maid, R. Radetzky, and K. Zetsche. 1990. Structure of the rubisco operon from the multicellular red alga Antithamnion spec. Curr. Genet. 18:465–469. 25. Kusano, T., T. Takeshima, C. Inoue, and K. Sugawara. 1991. Evidence for two sets of structural genes for ribulose bisphosphate carboxylase in Thiobacillus ferrooxidans. J. Bacteriol. 173:7313–7323. 26. Lamar, R. T., B. Schoenike, A. Vanden Wymelenberg, P. Stewart, D. M. Dietrich, and D. Cullen. 1995. Quantitation of fungal mRNAs in complex substrates by reverse transcription PCR and its application to Phanerochaete chrysosporium-colonized soil. Appl. Environ. Microbiol. 61:2122–2126. 27. Le Guyader, F., E. Dubois, D. Menard, and M. Pommepuy. 1994. Detection of hepatitis A virus, rotavirus, and enterovirus in naturally contaminated shellfish and sediment by reverse transcription-seminested PCR. Appl. Environ. Microbiol. 60:3665–3671. 28. Lorimer, G. H. 1981. Ribulosebisphosphate carboxylase: amino acid sequence of a peptide bearing the activator carbon dioxide. Biochemistry 20:1236–1240. 29. Makarewicz, J. C. 1993. Phytoplankton biomass and species composition in


FIG. 7. Dendrogram derived from deduced amino acid sequences (partial) of 21 Lake Erie transcriptionally active RubisCO genes and of rbcL (cbbL) or cbbM genes from representative species. The form II RubisCO sequence (CbbM) of Rhodobacter sphaeroides was used as the outgroup. The scale bar represents 0.1 substitution per site. Bootstrap values above 50 (percentages) are indicated. Source of sequence: Rhodobacter sphaeroides form II (52), Cryptomonas f (10), Pleurochrysis carterae (15), Mastocarpus stellatus (14), Ceramium diaphnum (14), and Ahnfeltia fastigiata (14). The rest were the same as in Fig. 6.


VOL. 62, 1996

1921

Loosdrecht lakes (The Netherlands). Limnol. Oceanogr. 31:983–988. 43. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 44. Song, K., and T. C. Osborn. 1994. A method for examining expression of homologous genes in plant polyploids. Plant Mol. Biol. 26:1065–1071. 45. Tabita, F. R. 1988. Molecular and cellular regulation of autotrophic carbon dioxide fixation in microorganisms. Microbiol. Rev. 52:155–189. 46. Tabita, F. R. 1995. The biochemistry and metabolic regulation of carbon metabolism and CO2 fixation in purple bacteria, p. 885–914. In R. E. Blankenship, M. T. Madigan, and C. E. Bauer (ed.), Anoxygenic photosynthetic bacteria. Kluwer Academic Publishers, Dordrecht, The Netherlands. 47. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 9:3251–3270. 48. Tsai, Y.-L., M. D. Sobsey, L. R. Sangermano, and C. J. Palmer. 1993. Simple method of concentrating enteroviruses and hepatitis A virus from sewage and ocean water for rapid detection by reverse transcriptase-polymerase chain reaction. Appl. Environ. Microbiol. 59:3488–3491. 49. Tsai, Y.-L., B. Tran, and C. J. Palmer. 1995. Analysis of viral RNA persistence in seawater by reverse transcriptase-PCR. Appl. Environ. Microbiol. 61:363–366. 50. Tsai, Y.-L., B. Tran, L. R. Sangermano, and C. J. Palmer. 1994. Detection of poliovirus, hepatitis A virus, and rotavirus from sewage and ocean water by triplex reverse transcriptase PCR. Appl. Environ. Microbiol. 60:2400–2407. 51. Valentin, K., and K. Zetsche. 1990. Nucleotide sequence of the gene for the large subunit of rubisco from Cyanophora paradoxa—phylogenetic implications. Curr. Genet. 18:199–202. 52. Wagner, S. J., S. E. Stevens Jr., B. T. Nixon, D. H. Lambert, R. G. Quivey, Jr., and F. R. Tabita. 1988. Nucleotide and deduced amino acid sequence of the Rhodobacter sphaeroides gene encoding form II ribulose-1,5-bisphosphate carboxylase/oxygenase and comparison with other deduced form I and II sequences. FEMS Microbiol. Lett. 55:217–222. 53. Yoshinaga, K., T. Ohta, Y. Suzuki, and M. Sugiura. 1988. Chlorella chloroplast DNA sequence containing a gene for the large subunit of ribulose-1,5bisphosphate carboxylase/oxygenase and a part of a possible gene for the b9 subunit of RNA polymerase. Plant Mol. Biol. 10:245–250. 54. Zurawski, G., B. Perrot, W. Bottomley, and P. R. Whitfeld. 1981. The structure of the gene for the large subunit of ribulose 1,5-bisphosphate carboxylase from spinach chloroplast DNA. Nucleic Acids Res. 9:3251–3270.


Lake Erie, 1970 to 1987. J. Gt. Lakes Res. 19:258–274. 30. Meijer, W. G., A. C. Arnberg, H. G. Enequist, P. Terpstra, M. E. Lidstrom, and L. Dijkhuizen. 1991. Identification and organization of carbon dioxide fixation genes in Xanthobacter flavus H4-14. Mol. Gen. Genet. 225:320–330. 31. Monceyron, C., and B. Grinde. 1994. Detection of hepatitis A virus in clinical and environmental samples by immunomagnetic separation and PCR. J. Virol. Methods 46:157–166. 32. Munawar, M., and I. F. Munawar. 1976. A Lakewide survey of phytoplankton biomass and its species composition in Lake Erie, April–December 1970. J. Fish. Res. Board Can. 33:581–600. 33. Nicholls, K. H., and G. J. Hopkins. 1993. Recent changes in Lake Erie (North Shore) phytoplankton: cumulative impacts of phosphorus loading reductions and the zebra mussel introduction. J. Gt. Lakes Res. 19:637–647. 34. Ogihara, Y., T. Terachi, and T. Sasakuma. 1991. Molecular analysis of the hot spot region related to length mutations in wheat chloroplast DNAs. I. Nucleotide divergence of genes and intergenic spacer regions located in the hot spot region. Genetics 129:873–884. 35. Paul, J. H., and S. L. Pichard. 1996. Molecular approaches to study natural communities of autotrophs, p. 301–309. In M. E. Lidstrom and F. R. Tabita (ed.), Microbial growth on C1 compounds. Kluwer Academic Publishers, Dordrecht, The Netherlands. 36. Pichard, S. L., J. Brown, and J. H. Paul. 1994. RubisCO gene expression in marine phytoplankton communities: a taxonomic perspective, abstr. N-125, p. 354. In Abstracts of the 95th General Meeting of the American Society for Microbiology 1995. American Society for Microbiology, Washington, D.C. 37. Pichard, S. L., M. E. Frischer, and J. H. Paul. 1993. Ribulose bisphosphate carboxylase gene expression in subtropical marine phytoplankton populations. Mar. Ecol. Prog. Ser. 101:55–65. 38. Pichard, S. L., and J. H. Paul. 1991. Detection of gene expression in genetically engineered microorganisms and natural phytoplankton populations in the marine environment by mRNA analysis. Appl. Environ. Microbiol. 57: 1721–1727. 39. Pichard, S. L., and J. H. Paul. 1993. Gene expression per gene dose, a specific measure of gene expression in aquatic microorganisms. Appl. Environ. Microbiol. 59:451–457. 40. Pridmore, R. D. 1987. New and versatile cloning vectors with kanamycinresistance marker. Gene 56:309–312. 41. Reichelt, B., and S. F. Delaney. 1983. The nucleotide sequence for the large subunit of ribulose 1,5-bisphosphate carboxylase from a unicellular cyanobacterium, Synechococcus PCC 6301. DNA 2:121–129. 42. Riegman, R., and L. R. Mur. 1986. Phytoplankton growth and phosphate uptake (for P limitation) by natural phytoplankton populations from the
