Amino acid sequence homology between N- and C-terminal halves of a carbonic anhydrase in Porphyridium purpureum, as deduced from the cloned cDNA.

Carbonic anhydrase (CA) from Porphyridium purpureum, a unicellular red alga, was purified >209-fold to a specific activity of 1,147 units/mg protein. cDNA clones for this CA were isolated. The longest clone, comprising 1,960 base pairs, contained an open reading frame which encoded a 571-amino acid polypeptide with a calculated molecular mass of 62,094 Da. The N- and C-terminal halves of the putative mature Porphyridium CA have amino acid sequence homology to each other (>70%) and to other prokaryotic-type CAs. Both regions contain, at equivalent positions, one set of three possible zinc-liganding amino acid residues conserved among prokaryotic-type CAs. CA purified from Porphyridium contained two atoms of zinc per molecule. We propose that the Porphyridium CA has evolved by duplication of an ancestral CA gene followed by the fusion of the duplicated CA gene. The CA truncated into the putative mature form was overexpressed in Escherichia coli, and the expressed protein was active. Clones expressing separately the N- and C-terminal halves of the CA were constructed. CA activity was present in extracts of E. coli cells expressing the N-terminal half, while no detectable activity was found in cells expressing the C-terminal half.

evolutionary origin of these two types of CAs (6). Although the crystal structure of the prokaryotic-type CAs has not yet been resolved, investigation of spinach CA by site-directed mutagenesis and extended x-ray absorption fine structure analysis showed that this enzyme binds an active-site zinc through two cysteine and one histidine residues (7).
Various microalgae exhibit much lower apparent K 0.5 (CO 2 ) for photosynthesis when they were grown under a low-CO 2 (air level, 0.04% CO 2 ) condition (low-CO 2 cells), than when grown under a high-CO 2 (4 -5% CO 2 ) condition (high-CO 2 cells) (8). Correspondingly, low-CO 2 cells have much higher carbonic anhydrase activity than high-CO 2 cells. Some microalgal species have the CA activity both inside cells and on the surface of the cells, while others only have intracellular activity (9). It has been thought that both the intracellular and extracellular CAs are effective in increasing the supply of CO 2 to ribulose-1,5bisphosphate carboxylase/oxygenase under CO 2 -limiting conditions (10,11).
Extracellular CA of Chlamydomonas reinhardtii has been extensively investigated. This enzyme is a heterotetramer with a native size of 76 kDa (12), consisting of two large subunits (35 and 36.5 kDa) and two small subunits (4 kDa). Isolation and sequencing of its cDNA revealed the existence of possible functional amino acid residues conserved in eukaryotic-type of CAs (13). On the other hand, although the presence and physiological importance of intracellular CA were reported in C. reinhardtii (14) and other eukaryotic microalgae (8,9), their molecular characteristics have not been elucidated. In cyanobacteria, the genomic DNA region, which complements the Synechococcus PCC7942 mutant requiring a higher concentration of inorganic carbon for growth, contains an open reading frame homologous to CAs of higher plants (6).
CA from Porphyridium purpureum R-1, a unicellular red alga, was partially purified, characterized, and shown to be a monomer with a native molecular mass of 55 kDa (15). This CA was detected only inside the cells. Immunogold electron microscopy using an antiserum raised against the partially purified enzyme suggested that it was located mainly in the chloroplast stroma (16).
We have isolated and characterized cDNA clones encoding CA from P. purpureum, and describe here the primary structure of this enzyme, which is composed of two almost identical domains, each of which corresponds to a monomer of prokaryotic-type CAs. named P. cruentum R-1, Culture Collection of the Institute of Molecular Cell Bioscience, University of Tokyo) were grown as described by Yagawa et al. (15) except that the culture medium was buffered with 10 mM TES-NaOH (pH 7.5). To obtain low-CO 2 cells, the culture medium was constantly bubbled with air containing 5% CO 2 for several days and then with air (0.04% CO 2 ) for 2 days.
Purification of Porphyridium CA-Low-CO 2 cells were harvested by centrifugation and disrupted in extraction buffer containing 100 mM Tris-HCl (pH 9.0), 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride, by using a chilled French pressure cell at 147 megapascals. The homogenate was clarified by centrifugation (20,000 ϫ g, 60 min), and the supernatant was subjected to ammonium sulfate fractionation. Precipitate obtained at 30 -65% saturation was resuspended and dialyzed against 5 mM Tris-HCl (pH 9.5). The dialysate was charged onto an affinity column (p-methylaminobenzene sulfonamide-conjugated Sepharose 6B) which was prepared according to Yang et al. (18) and pre-equilibrated with 5 mM Tris-HCl (pH 9.5). After extensive washing with 5 mM Tris-HCl (pH 8.5), the fraction with CA activity was eluted with 50 mM Tris-HCl (pH 8.5) containing 80 mM NaClO 4 . The fraction was concentrated and dialyzed against 100 mM Tris-HCl (pH 9.0) by ultrafiltration (Molcut-L, LGC, Millipore). The dialysate was applied to TSK-gel DEAE-5PW (7.5 mm ϫ 7.5 cm, Tosoh, Tokyo, Japan) equilibrated with 100 mM Tris-HCl (pH 9.0). The proteins were eluted with 60 ml of linear 0 -0.5 M NaCl gradient in 100 mM Tris-HCl (pH 9.0) at a flow rate of 1.0 ml/min, and fractions containing CA activity were collected and kept at 4°C. All purification procedures were carried out at 4°C, with the exception of anion-exchange chromatography, which was run at ambient temperature.
The CA activity was measured as the time needed for a pH change from 8.3 to 7.3 (t) after addition of 2 ml of CO 2 saturated water to 12 mM sodium 5,5-diethylbarbiturate-HCl buffer (pH 8.3) containing an enzyme solution and 5 mM NaCl (final concentration in a total volume of 5 ml). The reaction was carried out at 2°C. An enzyme activity unit was calculated using the equation, unit ϭ (t 0 Ϫ t)/t, where t 0 is the time required for the pH change using buffer without enzyme. SDS-PAGE was carried out according to the method of Laemmli (19) using 12.5% gel. Protein concentrations were determined by the method of Bradford (20) using Bio-Rad dye reagents and bovine serum albumin as a standard.
Peptide Sequencing-Pyridylethylated CA was digested with Achromobacter protease I (EC 3.4.21.50) purchased from Wako Chemicals (Tokyo, Japan). The resultant peptide fragments were separated by reversed-phase HPLC with a TSK-gel ODS-120T column (4.6 mm ϫ 15 cm, Tosoh). The column was equilibrated with solution A (0.1% trifluoroacetic acid) and developed with a linear gradient from 0 to 60% solution B (0.085% trifluoroacetic acid in 90% acetonitrile) in solution A at a flow rate of 0.8 ml/min. The sequences of fractionated peptide fragments were analyzed by a peptide sequencer (PSQ 2; Shimadzu, Kyoto, Japan).
Isolation of RNA-Because P. purpureum cells are surrounded by high amounts of polysaccharides, total RNA isolation was performed as described for marine macroalgae (21) with the exception of using a lysis buffer (4 M guanidine thiocyanate, 100 mM 2-mercaptoethanol, 0.5% sodium N-lauroyl sarcosine, and 25 mM sodium citrate, pH 7.0). Purification of poly(A) ϩ RNA was carried out using oligotex 30 (Takara Shuzo Co., Ltd.).
Amplification of a Partial cDNA Fragment-Four oligonucleotide primers, PrS1, PrA1, PrS2, and PrA2, were synthesized based on the sequence of two peptide fragments. These primers were mixtures of all possible nucleotide sequences determining the sequences of a proteasecleaved peptide fragments, Asn-Ile-Phe-Ala-Asn-Asn-Glu-Ala (PrS1 and PrA1, 23-mer) or Phe-Val-Asn-Asn-Glu-Asn-Trp-Arg (PrS2 and PrA2, 23-mer). PrS1 and PrS2 correspond to sense strand of cDNA, while PrA1 and PrA2 to antisense strand. cDNA was synthesized by the method of Gubler and Hoffman (22) and used as a template. PCR was performed in a thermal cycler with a cycle of 1 min at 94°C, 2 min at 54°C, and 3 min at 72°C using a primer-combination of either PrS1-PrA2 or PrS2-PrA1. The amplified PCR product was cloned into pCR vector (Stratagene) and sequenced.
Screening and Nucleotide Sequence Analysis of cDNA Clones-A cDNA library was constructed in gt10 (cDNA cloning kit, Amersham) according to the manufacture's protocol. The PCR-amplified cDNA fragment described above was labeled with [ 32 P]dCTP by a random-primed method. Plaque hybridization was carried out using nylon membrane (Hybond-Nϩ, Amersham) in a solution containing 6 ϫ SSC (1 ϫ SSC ϭ 0.15 M NaCl, 15 mM sodium citrate), 5 ϫ Denhardt's solution (0.02% bovine serum albumin, 0.02% Ficoll, and 0.02% polyvinylpyrrolidone), 0.1% SDS, and 100 g of salmon sperm DNA/ml at 68°C for 12 h and then the membrane was sequentially washed by 2 ϫ SSC, 0.1% SDS at 20°C for 30 min; 1 ϫ SSC, 0.1% SDS at 65°C for 1 h; and 0.1 ϫ SSC, 0.1% SDS at 65°C for 1 h. The cDNA inserts isolated from purified positive plaques were subcloned into pUC119 and the nucleotide sequences were determined on both strands by the dideoxy chain termination method using universal primers or synthetic oligonucleotide primers with a DNA sequencer (Applied Biosystems, model 373A). Computer analysis and comparison of DNA sequences were performed using GENETYX genetic information processing software (Software Development, Tokyo, Japan).
Zinc Content Determination-For the metal analysis, the CA from P. purpureum was purified as described above and loosely bound zinc was removed using Chelex 100 (Bio-Rad). Zinc concentration was analyzed by an inductively coupled plasma atomic emission spectrometer (ICPS-1000III, Simadzu). A zinc standard solution for metal content measurement was purchased from Wako Chemicals (Tokyo, Japan). As a negative control, buffer alone was treated similarly. Protein concentration of the sample was determined by amino acid content analysis using an amino acid analysis system (Tosoh).
Plasmid Construction for Expression of CA in E. coli-In order to construct a plasmid expressing a truncated CA (encoded in gtPCA1, the clone with the longest insert) without a putative hydrophobic transit peptide, cDNA encoding the N-terminal portion of the CA was amplified by PCR using two primers as follows: M1, 5Ј-GGGGAAGCTT-GAAGCTCGCGGCAGGCATGGG-3Ј; M2, 5Ј-GTCACGAAGCTTGC-CGTCACC-3Ј. M1 partially corresponds to the sense strand bases from 229 to 248, determining the N-terminal end of the putative mature CA. M2 corresponds to the antisense strand bases from 874 to 894, including a HindIII site inside the cDNA. M1 was tagged by the HindIII cleavage site, therefore the amplified product was cleaved by HindIII into a segment of 0.8-kb. The 1.0-kb HindIII-HindIII fragment of plasmid pUC119 containing the entire cDNA insert was replaced by the 0.8-kb HindIII fragment and the absence of base substitution in the amplified region was confirmed by a DNA sequence analysis. The resultant plasmid containing the 0.8-kb HindIII fragment in correct orientation, encodes 7 amino acid residues of LacZ fused to the CA lacking 76 amino acid residues at the N-terminal. The plasmid called pPCA was used to transform Escherichia coli JM109 cells.
Plasmid Constructions for the Expression of the N-and C-terminal Halves of the CA-As described under "Results," the Porphyridium CA consists of homologous N-and C-terminal halves, each of which exhibits a sequence similarity to other prokaryotic-type CAs. To construct plasmids for the expression of the N-and C-terminal halves, EcoRI and HindIII sites were introduced between the two halves by using PCRs. The primers used are as follows: N1, 5Ј-CTGTGCTGCAGTACGCGG-3Ј; N2, 5Ј-GTCCGACAAGCTTGAATTCTTGGTGCGGTAGAACTTTGAA-3Ј; C1, 5Ј-ACCAAGAATTCAAGCTTGTCGGACAGTGGTGCGCTC-3Ј; C2, 5Ј-CTTAACCTTGAGGTATTGCACGGC-3Ј. N1 and N2 correspond to the sense and antisense strands, bases from 557 to 574 and partial bases from 927 to 951, respectively, while C1 and C2 correspond to the sense and antisense strands, partial bases from 943 to 966 and bases from 1333 to 1356, respectively. N2 and C1 contain nucleotide mismatches to create EcoRI and HindIII site. First, two PCRs were performed either with N1 and N2 or with C1 and C2. Next, using a mixture of the products of these two PCRs, the second PCR was carried out with N1 and C2, and the product was cloned into a pCR vector (pCR-EH). The introduction of EcoRI and HindIII restriction sites and the absence of mutation in the amplified region were confirmed by a DNA sequence analysis. The plasmid pCR-EH was digested with PstI and the 0.8-kb fragment was gel-purified and ligated into the PstI-cleaved pPCA plasmid. The resultant plasmid containing the pCR-EH-derived 0.8-kb PstI fragment in correct orientation was designated as pPCA-EH.
For expression of the N-terminal half (molecular mass approximately 28.1 kDa) of the CA, the 0.9-kb EcoRI-EcoRI fragment of pPCA-EH encoding the C-terminal half was replaced by the omega fragment (23). The omega fragment terminates transcription and translation coming into the region at the both ends (23). The resultant plasmid called pNCA was used to transform E. coli strain JM109.
For the expression of the C-terminal half (molecular mass approximately 28.9 kDa) of the CA, the pPCA-EH plasmid was digested with HindIII and the 4.1-kb fragment was self-ligated, and the resultant plasmid called pCCA was used to transform E. coli strain JM109.
Expression and Measurement of the Activity of CA and Polypeptides of N-and C-terminal Halves-E. coli clones containing a control plasmid lacking the insert, pPCA, pNCA, or pCCA were grown overnight, diluted into fresh Luria-Bertani broth containing 100 g of ampicillin/ ml, and grown at 37°C until the cultures reached an A 600 of 0.2. Subsequently, the cultures were incubated in the presence or absence of 0.1 mM isopropyl-␤-D-thio-galactopyranoside for 5 h at 28°C. Cells were harvested by centrifugation and resuspended in a lysis buffer containing 1 mM phenylmethylsulfonyl fluoride, 0.1 mM EDTA, and 50 mM Tris-HCl (pH 9.0). The cells were disrupted by sonication and subjected to centrifugation (12,000 ϫ g, 15 min), and the CA activity of the supernatant fractions was determined as described above.

RESULTS
Purification of CA-CA from P. purpureum cells was purified by 209-fold (27% recovery) with a specific activity of 1,147 units/mg protein (Table I). It was thought that the sulfanilamide-conjugated affinity chromatography used in the purification of Chlamydomonas CA would not be suitable for that of Porphyridium CA, since this compound showed little inhibitory effect on the CA activity of the latter alga (15). However, under weak ionic strength and alkaline conditions, namely in 5 mM Tris-HCl (pH 9.5), Porphyridium CA was bound to the affinity column and efficiently purified. Using almost the same conditions for binding and elution, the chloroplastic CA had been effectively purified from pea (24). This observation suggests that both Porphyridium CA and pea chloroplastic CA exhibit similar interaction with sulfanilamide.
After the step of an anion-exchange HPLC, the active protein fraction migrated as a doublet band with an apparent molecular mass of around 59 kDa on SDS-PAGE (Fig. 1). These bands showed similar amino acid compositions (data not shown). In addition, these bands reacted with the antisera raised against the CA expressed in E. coli (data not shown). It is therefore possible that the smaller band represent either proteolytic product generated during the purification or CA isomer present in P. purpureum cells. The latter possibility was supported by the isolation of cDNA clones encoding two CAs different in 4 amino acid residues (see below).
cDNA Cloning of CA-In order to obtain a partial cDNA fragment for screening of the clones with full-length cDNA insert, two PCRs were carried out using a combination of degenerated primers, either PrS1-PrA2 or PrS2-PrA1. Amplification occurred only when the set of PrS1-PrA2 was used. The PCR product (0.8-kb) was cloned and sequenced. It contained two primer sequences at both ends and five amino acid sequences identical with those obtained by sequencing the proteolytic fragments. In addition, the amino acid sequences deduced from the nucleotide sequence of the PCR product contained the motifs conserved in the CAs of prokaryotes and higher plants (see below). It was, therefore, concluded that the PCR product was amplified from the cDNA of Porphyridium CA, and is suitable for use as a probe to screen a cDNA library for the clone encoding the CA.
A gt10 cDNA library was constructed using poly(A) ϩ RNA of P. purpureum cells which were adapted to air for 3 h to induce CA activity (data not shown). The cDNA library, containing 1.5 ϫ 10 5 recombinant phages, was screened by the 32 P-labeled PCR-amplified partial cDNA fragment, and more than 100 independent plaques were positive. Twenty clones having the strongest signals were purified. The EcoRI fragment inserts of positive clones were from 1.4 to 2.0 kb in length. Using these phage DNAs as a template and two primers mentioned above, 0.8-kb DNAs were PCR-amplified. Several clones with the longest inserts were analyzed by restriction mapping and classified into two groups based on the number of HindIII sites. A clone having the longest insert with one HindIII site was designated as gtPCA1 and another clone with the longest insert of the second group having 2 HindIII sites as gtPCA2.
Nucleotide and Deduced Amino Acid Sequence of CA cDNA-The cDNA insert of gtPCA1 was sequenced on both strands. Analysis of the nucleotide sequence of the gtPCA1 insert (GenBank accession number D86050) revealed that it consisted of a 90-bp 5Ј-untranslated region, a 1,713-bp open reading frame encoding 571-amino acid polypeptide with a calculated molecular mass of 62,094 Da, a 141-bp 3Ј-untranslated region, and a 16-bp poly(A) segment as shown in Fig. 2. The positions of the synthetic PCR primers were indicated by arrows. The first ATG codon at the 91st nucleotide was determined as the initiation codon based on the following observations: 1) no other ATG codon in the frame was found upstream from Leu-78 whose presence was confirmed by peptide sequencing. 2) The nucleotide sequence TCACC just upstream from the first ATG codon resembles the consensus sequence CC(A/G)CC for the eukaryotic initiation site proposed by Kozak (25).
The amino acid sequence of Porphyridium CA deduced from the cDNA nucleotide sequence is also shown in Fig. 2. Amino acid sequences of peptide fragments obtained by the Achromobacter Protease I digestion of the purified CA were identical with the amino acid sequences underlined in Fig. 2 except for Met-108 and Asn-389. The differences were explained by assuming the existence of two or more CA isozymes in P. purpureum. In fact, the deduced amino acid sequence from the second clone, gtPCA2, showed complete agreement with the amino acid sequences of these peptide fragments. Briefly, the insert of gtPCA2 (1,905-bp in length) encodes a polypeptide of 571 amino acid residues (62,078 Da) in which 4 residues were substituted when compared with those encoded by gtPCA1: Thr-30, Lys-50, Val-108, and Lys-389. The nucleotide sequence of gtPCA2 has been submitted to GenBank (accession number D86051). The gtPCA1 having the longest insert was used for further analyses.
Although the N-terminal amino acid residue could not be determined by direct protein sequencing, the sequence of highly hydrophobic amino acid residues, such as alanine and proline, in the N-terminal region of the protein encoded in the open reading frame would be regarded as a transit peptide which functions in transport of the nascent enzyme into the chloroplast. The size of the possible mature protein estimated from deduced amino acid sequence showed good agreement with the molecular mass calculated by SDS-PAGE. The presence of a transit peptide is consistent with the chloroplastic location of the enzyme suggested by immunogold electron microscopy (16).
Of particular interest is the finding that in the amino acid sequence of the putative mature CA, extensive sequence similarity was observed between the N-and C-terminal halves (Fig.  3). The alignment was carried out to maximize the similarities between these two halves. When they were compared, 72% of the amino acid residues are identical and 23% are conservative substitutions in the alignment shown in Fig. 3.
Furthermore, each half contains a set of amino acid residues conserved in CAs from higher plants and two prokaryotes (Fig.  4). The homologies of the N-terminal half with CAs from Synechococcus (6), E. coli (26), spinach (27), and barley (28) were 29, 27, 28, and 32%, respectively, and those of the C-terminal half were 28, 25, 24, and 28%, respectively. Although no information is available for crystal structure of these CAs, Bracey et al. (7) investigated the role of conserved amino acid residues of spinach CA for the zinc binding by site-directed mutagenesis analysis. Mutation at Cys-150, His-210, or Cys-213 of the spinach CA (numbers used in Ref. 7) caused inactivation of the CA and reduction of zinc binding, suggesting that the active-site zinc of spinach CA is coordinated with these two cysteine and one histidine residues. These three amino acid residues were The nucleotide numbered as ϩ1 corresponds to the translation initiation site. Amino acid sequences that are identical to those determined by peptide sequencing of fragments obtained by the Achromobacter protease I digestion are underlined. The position of primers used for the amplification of a partial cDNA fragment to make a CA-specific probe are indicated by arrows. Translation termination codon (TAA) are indicated by an asterisk. The underlined TATAAA at the 3Ј end of the cDNA represents a putative polyadenylation signal sequence.
found at equivalent positions in the sequence of both N-terminal half (Cys-149, His-205, and Cys-208) and C-terminal half (Cys-403, His-459, and Cys-462) of the Porphyridium CA (numbers used in this report). In pea CA, Glu-204 (numbers used in Ref. 29) is additionally shown to be essential to the activity (29), however, the glutamic acid residues were replaced by glutamine residues in both halves (Gln-189 and Gln-443). Aside from these three amino acid residues, 18 (N-terminal half) or 17 (C-terminal half) additional amino acid residues were also conserved in all other prokaryotic-type CAs so far reported. It is possible that these conserved residues are important in catalytic activity or in forming the functional conformation of these enzymes.
Zinc Content of CA-Deduced amino acid sequence of the Porphyridium CA suggested that this enzyme possesses two sets of possible ligands to the active-site zinc on a single polypeptide, while the binding of only one zinc atom to a monomer has been shown in the higher plant enzymes (2). In the native Porphyridium CA, the ratio of zinc atoms to the enzyme molecule was 1.92, suggesting that this CA bound zinc in a ratio of two zinc atoms per molecule.
Functional Expression of CA-In order to demonstrate that gtPCA1 encodes an active CA, the protein was subjected to expression in E. coli. The plasmid expressing CA was designed so as to produce possible mature enzyme excluding 76 amino acid residues at the N-terminal end. The significant CA activity was detected in extracts of the cells transformed with pPCA after isopropyl-␤-D-thio-galactopyranoside induction but not in those of uninduced cells or induced cells harboring control plasmid, pUC119 (Table II). The expressed CA was purified to homogeneity by the same procedures as used in the purification from Porphyridium cells (i.e. ammonium sulfate fractionation, affinity chromatography, and anion-exchange HPLC) and the purified material showed properties identical to those of the native enzyme: (i) 50% inhibition of activity of the CA purified from the E. coli cells was observed by ethoxyzolamide and acetazolamide at a final concentration of 2 ϫ 10 Ϫ7 and 6 ϫ 10 Ϫ8 M, respectively. The I 50 values were almost identical to those determined for purified native Porphyridium CA (the values for ethoxyzolamide and acetazolamide were 1 ϫ 10 Ϫ7 and 6 ϫ 10 Ϫ8 M, respectively.). (ii) The activity of CA purified from E. coli was stimulated by chloride ion at millimolar concentrations as was the native Porphyridium CA (15). (iii) Both the CA purified from the E. coli and P. purpureum showed higher activities when it had been kept at an alkaline (pH 9 -10) condition (15).
Expression of N-terminal and C-terminal Halves of CA-In order to evaluate the catalytic activity of N-and C-terminal halves of the CA, both polypeptides were separately expressed in E. coli. The expression of the truncated CA polypeptides in E. coli cells containing pNCA or pCCA was confirmed by immunoblotting with anti-Porphyridium CA antisera (Fig. 5). Each clone synthesized a soluble polypeptide of the expected molecular weight. The CA activity was detected in the crude extract of E. coli cells containing pNCA, although the specific activity was much lower than that expressed with pPCA. How- ever, no detectable CA activity was expressed from E. coli cells containing pCCA (Table II). DISCUSSION CA purified from low-CO 2 cells of P. purpureum with affinity column and DEAE-HPLC column chromatography contained multiple polypeptide species showing similar but distinct electrophoretic mobility on SDS-PAGE. These polypeptides would represent isozymes of the CA. This proposal was supported by the isolation of two independent cDNA clones, gtPCA1 and gtPCA2, each encoding CA with a 4-amino acid residue difference from the other. Each clone was isolated from a cDNA library derived from the mRNA of the low-CO 2 cells and encoded a polypeptide containing putative transit peptide with high hydrophobicity. Therefore, these cDNAs may encode two chloroplastic CAs which are induced under low-CO 2 condition. This conclusion is in agreement with the previous observation by immunogold electron microscopy suggesting that the CA is located within the chloroplast of low-CO 2 cells in P. purpureum (16). In addition, these cDNA inserts contain poly(A) tails, indicating that these CAs are encoded in the nuclear genome of P. purpureum, similarly to higher plants whose CA genes are also encoded in the nuclear genome (30).
Fukuzawa et al. (6) proposed that the following two distinct groups of CAs have evolved independently. One is the prokaryotic-CA group which includes CAs from prokaryotes and chloroplast of higher plants, and the other, the eukaryotic-CA group which includes CAs from various higher vertebrates and two extracellular isozymes of C. reinhardtii (31). Recently, the third group of CAs was discovered in the archaeon Methanosarcina thermophila (4). Among the three groups, prokaryotictype CAs are characterized by three conserved amino acid residues possibly involved in the zinc binding and some other conserved amino acid residues. We revealed that the Porphyridium CA consists of homologous N-and C-terminal halves and these conserved amino acid residues are present in both halves of the enzyme at equivalent positions, indicating that this CA is a member of prokaryotic-type CAs (Fig. 4). It has been reported that this CA does not cross-react with the spinach CA antibody (16). This might result from relatively broad regions with less similarity, except those containing conserved amino acid residues.
Striking sequence similarity (Ͼ70%) between the two halves of putative mature Porphyridium CA (Fig. 3) and the presence of amino acid residues possibly essential to the catalytic activity of prokaryotic-type CAs, within each of them (Fig. 4), suggest that the gene encoding the CA monomer of a putative ancestral organism would be duplicated and then fused to form Porphyridium CA or its ancestral enzyme through the process of evolution. In fact, the Porphyridium CA has a molecular mass of about 60 kDa, while those of monomer of higher plant CA, cyanobacterial CA, and an intracellular CA of Coccomyxa, a unicellular green alga, are around 24 (24,27,28), 30 (6), and 26 kDa (32), respectively. Using a CA-directed photoaffinity label, intracellular CA of Chlamydomonas was identified as a polypeptide with an apparent molecular mass of 30 kDa (33). However, its partial amino acid sequence exhibited homology to eukaryotic-type CA (34). It will be interesting to examine whether or not the duplicated CA structure is found in other red algae and cyanobacteria. In other words, distribution of this CA structure would reflect phylogenetic relationships of the algal species. In this connection, we recently found that in Porphyra (a multicellular red alga having a stellate chloroplast as in Porphyridium), polypeptide of a molecular mass of about 50 kDa cross-reacted with the antisera raised against the Porphyridium CA. 2 Three mammalian hexokinase isozymes (Types I, II, and III) are well characterized examples of the enzymes consisting of highly homologous N-terminal and C-terminal halves (e.g. Ref. 35). It is proposed that these isozymes (Types I, II, and III) of Ϸ100 kDa have evolved by a process of duplication and fusion of a gene encoding an ancestral hexokinase similar to the mammalian glucokinase (Type IV hexokinase) and the yeast enzyme of Ϸ50 kDa. Site-directed mutagenesis studies of the rat Type I isozyme have shown that catalytic activity is associated solely with the C-terminal half while the N-terminal half is catalytically inactive and thought to be involved in a regulatory function (36). On the other hand, more recently, sitedirected mutagenesis of rat Type II isozyme showed that both halves of the enzyme retain comparable catalytic activities (37).
In the present study, the N-and C-terminal halves of Porphyridium CA were expressed separately, and it was demonstrated that the N-terminal half was catalytically active, while no catalytic activity was found in the C-terminal half, although both were expressed as soluble proteins. One of the simplest interpretations of this observation is that the N-terminal half expresses the CA activity, while the C-terminal half is not functional due to substitutions in some essential amino acid residues. However, we showed that one molecule of Porphyridium CA binds two zinc atoms, in contrast to other prokaryotic-type CAs which have been shown to bind one zinc atom per 2 S. Mitsuhashi et al., manuscript in preparation.  molecule (2). Thus we rather expected that each half of Porphyridium CA binds one zinc atom and should exhibit CA activity. We wonder whether the folding of the C-terminal half into a catalytically active form is possible only when the Nterminal half is present. To test this hypothesis, we are in the process of constructing various mutants of Porphyridium CA. In any mutant forms thus far obtained, the N-terminal half can take a catalytically active configuration. A functional expression system developed in this study may be useful to further clarify the structure/function relationship and the biochemical and physiological significance of this enzyme structure.