Profiles of pyrimidine biosynthesis,salvage and degradation in disks of potato (Solanum tuberosum L.) tubers

In order to obtain general metabolic profiles of pyrimidine ribo- and deoxyribonucleotides in potato (Solanum tuberosum L.) plants, the in situ metabolic fate of various (14)C-labelled precursors in disks from growing potato tubers was investigated. The activities of key enzymes in potato tuber extracts were also studied. The following results were obtained. Of the intermediates in de novo pyrimidine biosynthesis, [(14)C]carbamoylaspartate was converted to orotic acid and [2-(14)C]orotic acid was metabolized to nucleotides and RNA. UMP synthase, a bifunctional enzyme with activities of orotate phosphoribosyltransferase (EC 2.4.2.10) and orotidine 5'-monophosphate decarboxylase (EC 4.1.1.23), exhibited high activity. The rates of uptake of pyrimidine ribo- and deoxyribonucleosides by the disks were high, in the range 2.0-2.8 nmol (g FW)(-1) h(-1). The pyrimidine ribonucleosides, uridine and cytidine, were salvaged exclusively to nucleotides, by uridine/cytidine kinase (EC 2.7.1.48) and non-specific nucleoside phosphotransferase (EC 2.7.1.77). Cytidine was also salvaged after conversion to uridine by cytidine deaminase (EC 3.5.4.5) and the presence of this enzyme was demonstrated in cell-free tuber extracts. Deoxycytidine, a deoxyribonucleoside, was efficiently salvaged. Since deoxycytidine kinase (EC 2.7.1.74) activity was extremely low, non-specific nucleoside phosphotransferase (EC 2.7.1.77) probably participates in deoxycytidine salvage. Thymidine, which is another pyrimidine deoxyribonucleoside, was degraded and was not a good precursor for nucleotide synthesis. Virtually all the thymidine 5'-monophosphate synthesis from thymidine appeared to be catalyzed by phosphotransferase activity, since little thymidine kinase (EC 2.7.1.21) activity was detected. Of the pyrimidine bases, uracil, but not cytosine, was salvaged for nucleotide synthesis. Since uridine phosphorylase (EC 2.4.2.3) activity was not detected, uracil phosphoribosyltransferase (EC 2.4.2.9) seems to play the major role in uracil salvage. Uracil was degraded by the reductive pathway via beta-ureidopropionate, but cytosine was not degraded. The activities of the cytosine-metabolizing enzymes observed in other organisms, pyrimidine nucleoside phosphorylase (EC 2.4.2.2) and cytosine deaminase (EC 3.5.4.1), were not detected in potato tuber extracts. Operation of the de novo synthesis of deoxyribonucleotides via ribonucleotide reductase and of the salvage pathway of deoxycytidine was demonstrated via the incorporation of radioactivity from both [2-(14)C]cytidine and [2-(14)C]deoxycytidine into DNA. A novel pathway converting deoxycytidine to uracil nucleotides was found and deoxycytidine deaminase (EC 3.5.4.14), an enzyme that may participate in this pathway, was detected in the tuber extracts.     

Hexose metabolism in discs excised from developing potato (Solanum tuberosum L.) tubers

Metabolism of radiolabelled hexoses by discs excised from developing potato (Solanum tuberosum L.) tubers was been investigated in the presence of acid invertase to prevent accumulation of labelled sucrose in the bathing medium (Viola, 1996, Planta 198: 179–185). When the discs were incubated with either [U-¹⁴C]glucose or [U-¹⁴C]fructose without unlabelled hexoses, the unidirectional rate of sucrose synthesis was insignificant compared with that of sucrose breakdown. The inclusion of unlabelled fructose in the medium induced a dramatic increase in the unidirectional rate of sucroses synthesis in the tuber discs. Indeed, the decline in the sucrose content observed when discs were incubated without exogenous sugars could be completely prevented by including 300 mM fructose in the bathing medium. On the other hand, the inclusion of unlabelled glucose in the medium did not significantly affect the relative incorporation of [U-¹⁴C]glucose to starch, sucrose or glycolytic products. Substantial differences in the intramolecular distribution of ¹³C enrichment in the hexosyl moieties of sucrose were observed when the discs were incubated with either [2-¹³C]fructose or [2-¹³C]glucose. The pattern of ¹³C enrichment distribution in sucrose suggested that incoming glucose was converted into sucrose via the sucrose-phosphate synthase pathway whilst fructose was incorporated directly into sucrose via sucrose synthase. Quantitative estimations of metabolic fluxes in vivo in the discs were also provided. The apparent maximal rate of glucose phosphorylation was close to the extractable maximum catalytic activity of glucokinase. On the other hand, the apparent maximal rate of fructose phosphorylation was much lower than the maximum catalytic activity of fructokinase, suggesting that the activity of the enzyme (unlike that of glucokinase) was regulated in vivo. Although in the discs incubated with or without fructose the rates of starch synthesis or glycolysis were similar, the relative partitioning of metabolic intermediates into sucrose was much higher in discs incubated with fructose (0.6% and 32.6%, respectively). It is hypothesised that the equilibrium of the reaction catalysed by sucrose synthase in vivo is affected in discs incubated with fructose as a result of the accumulation of the sugar in the tissue. This results in the onset of sucrose cycling. Incubation with glucose enhanced all metabolic fluxes. In particular, the net rate of starch synthesis increased from 2.0 mol · hexose · g FW^–1 · h^–1 in the absence of exogenous glucose to 3.7 mol · hexose · g FW^–1 · h^–1 in the presence of 300 mM glucose. These data are taken as an indication that the regulation of fructokinase in vivo may represent a limiting factor in the utilisation of sucrose for biosynthetic processes in developing potato tubers.Abbreviations ADPGlc adenosine 5-diphosphoglucose - Glc6P glucose-6-phosphate - hexose-P hexose phosphate - NMR nuclear magnetic resonance - UDPGlc uridine 5-diphosphoglucose Many thanks to L. Sommerville for skillfull assistance and to J. Crawford and J. Liu for useful discussions on flux analysis. The research was funded by the Scottish Office Agriculture and Fisheries Department.     

Pathways of starch and sucrose biosynthesis in developing tubers of potato (Solanum tuberosum L.) and seeds of faba bean (Vicia faba L.)

Tissue slices from developing potato tubers (Solanum tuberosum L.) and developing cotyledons of faba bean (Vicia faba L.) were incubated with specifically labelled [¹³C]glucose and [¹³C]ribose. Enriched[¹³C]glucose released from starch granules was analysed by nuclear magnetic resonance (NMR). Spectral analyses were also performed on sucrose purified by high-performance liquid chromatography. In both tissues a low degree of randomisation (< 11 % in potato and < 14% in Vicia) was observed between carbon positions 1 and 6 in glucose released from starch when material was incubated with [¹³C]glucose labelled in positions 6 and 1, respectively. Similarly, with [2-¹³C]glucose a low degree of randomisation was observed in position 5. These findings indicate that extensive transport of three-carbon compounds across the amyloplast membrane does not occur in storage organs of either species. This is in agreement with previously published data which indicates that sixcarbon compounds are transported into the plastids during active starch synthesis. When [1-¹³C]ribose was used as a substrate, ¹³C-NMR spectra of starch indicated the operation of a classical pentose-phosphate pathway. However, with [2-¹³C]glucose there was no preferential enrichment in either carbon positions 1 or 3 relative to 4 or 6 of sucrose and starch (glucose). This provides evidence that entry of glucose in this pathway may be restricted in vivo. In both faba bean and potato the distribution of isotope between glucosyl and fructosyl moieties of sucrose approximated 50%. The degree of randomisation within glucosyl and fructosyl moieties ranged between 11 and 19.5%, indicating extensive recycling of triose phosphates.Abbreviation NMR nuclear magnetic resonance We are grateful to Dr. George Ratcliffe for his critical reading of the text and Dr. Bernard Goodman for helpful suggestions on the NMR measurements. The research was funded by a European Economic Community research grant, which the authors duly acknowledge.     

Immunocytochemical localization of patatin,the major glycoprotein in potato (Solanum tuberosum L.) tubers

Patatin is a family of glycoproteins with an apparent molecular weight of 40 kDa. The protein is synthesized as a pre-protein with a hydrophobic signal sequence of 23 amino acids. Using different immunocytochemical methods we determined the tissue-specific as well as subcellular localization of the patatin protein. Since antibodies raised against patatin showed crossreactivity with glycans of other glycoproteins, antibodies specific for the protein portion of the glycoprotein were purified. Using these antibodies for electron-microscopical immunocytochemistry, the protein was found to be localized mainly in the vacuoles of both tubers and leaves of potatoes (Solanum tuberosum L.) induced for patatin expression. Neither cell walls nor the intercellular space contained detectable levels of patatin protein. Concerning the tissue specificity, patatin was mainly found in parenchyma cells of potato tubers. The same distribution was observed for the esterase activity in potato tubers.Abbreviations PHA phytohemagglutinin - TFMS trifluoromethanesulfonic acid     

Subcellular pyrophosphate metabolism in developing tubers of potato (Solanum tuberosum)

PPi has previously been implicated specifically in the co-ordination of the sucrose–starch transition and in the broader context of its role as co-factor in heterotrophic plant metabolism. In order to assess the compartmentation of pyrophosphate (PPi) metabolism in the potato tuber we analysed the effect of expressing a bacterial pyrophosphatase in the amyloplast of wild type tubers or in the cytosol or amyloplast of invertase-expressing tubers. The second and third approaches were adopted since we have previously characterized the invertase expressing lines to both exhibit highly altered sucrose metabolism and to contain elevated levels of PPi (Farré et al. (2000a) Plant Physiol 123:681) and therefore this background rendered questions concerning the level of communication between the plastidic and cytosolic pyrophosphate pools relatively facile. In this study we observed that the increase in PPi in the invertase expressing lines was mainly confined to the cytosol. Accordingly, the expression of a bacterial pyrophosphatase in the plastid of either wild type or invertase-expressing tubers did not lead to a decrease in total PPi content. However, the expression of the heterologous pyrophosphatase in␣the cytosol of cytosolic invertase-expressing tubers led to strong metabolic changes. These results are discussed both with respect to our previous hypotheses and to current models of the compartmentation of potato tuber metabolism.     

Immunolocalisation of two tropinone reductases in potato (Solanum tuberosum L.) root, stolon, and tuber sprouts

Tropinone reductases (TRs) are essential enzymes in the tropane alkaloid biosynthesis, providing either tropine for hyoscyamine and scopolamine formation or providing pseudotropine for calystegines. Two cDNAs coding for TRs were isolated from potato (Solanum tuberosum L.) tuber sprouts and expressed in E. coli. One reductase formed pseudotropine, the other formed tropine and showed kinetic properties typical for tropine-forming tropinone reductases (TRI) involved in hyoscyamine formation. Hyoscyamine and tropine are not found in S. tuberosum plants. Potatoes contain calystegines as the only products of the tropane alkaloid pathway. Polyclonal antibodies raised against both enzymes were purified to exclude cross reactions and were used for Western-blot analysis and immunolocalisation. The TRI (EC 1.1.1.206) was detected in protein extracts of tuber tissues, but mostly in levels too low to be localised in individual cells. The function of this enzyme in potato that does not form hyoscyamine is not clear. The pseudotropine-forming tropinone reductase (EC 1.1.1.236) was detected in potato roots, stolons, and tuber sprouts. Cortex cells of root and stolon contained the protein; additional strong immuno-labelling was located in phloem parenchyma. In tuber spouts, however, the protein was detected in companion cells.     

Patatin,a major soluble protein of the potato (Solanum tuberosum L.) tuber is synthesized as a larger precursor

Antibodies were raised against the highly purified glycoprotein patatin. They were used to characterize the product synthesized in a wheatgerm cell-free translation system, programmed with polyadenylated RNA from potato tubers. Sodium dodecylsulfate-polyacrylamide gel electrophoresis revealed that the immunoprecipitated protein had a molecular mass of 43 kDa compared to 40 for the authentic patatin. It is assumed that patatin is synthesized in vivo as a larger precursor which is processed to the mature protein by cleavage of a signal peptide. Our results are in agreement with sequence-analysis data of patatin complementary DNA which indicate a signal peptide of about 23 amino acids (Mignery et al., 1984; Nucleic Acids Res. 12, 7987–8000).Abbreviation Poly(A)⁺ RNA polyadenylated RNA - SDS-PAGE sodium dodecyl sulfate-polyacrylamide gel electrophoresis Preliminary results were published in Mitteilungsband, Botaniker Tagung in Wien, p. 180, Wien, September 1984     

Consequences of the salvage of purine compounds on the proliferation of rat T-lymphocytes with normal or inhibited purine de novo synthesis

We studied the ability of purine compounds to restore the proliferation of concanavalin-A-stimulated rat T-lymphocytes under conditions of purine de novo synthesis inhibition and, on the other hand, the inhibition by purine nucleosides of the response of these cells to a mitogenic stimulation under conditions of normal purine de novo synthesis. The use of 50 μM azaserine, a potent inhibitor of purine de novo synthesis, allowed us to define the physiologically active salvage pathways of purine bases, ribo- and deoxyribonucleosides in concanavalin-A-stimulated rat T-lymphocytes. Except for guanylic compounds, all purines completely restored cell proliferation at a concentration of 50 μM. Guanine, guanosine and 2′-deoxyguanosine at concentrations up to 500 μM did not allow us to restore more than 50% of the cell proliferation. In conditions of normal purine de novo synthesis, the addition of 1000 μM adenine, adenosine, 2′-deoxyadenosine or 100 μM 2′-deoxyguanosine inhibited rat T-lymphocyte proliferation. The differences between the degree of inhibition of cell proliferation could be explained only in part by the differences between the capacities of salvage of these compounds. Furthermore, the fact that 2′-deoxyguanosine toxicity was dependent and 2′-deoxyadenosine toxicity independent on the activation state of the cells provided more evidence that the biochemical mechanisms of inhibition of cell proliferation should be different for these two nucleosides.     

Chloroplast DNA evolution in potato (Solanum tuberosum L.)

Summary A deletion specific to chloroplast (ct) DNA of potato (Solanum tuberosum ssp. tuberosum) was determined by comparative sequence analysis. The deletion was 241 bp in size, and was not flanked by direct repeats. Five small, open reading frames were found in the corresponding regions of ctDNAs from wild potato (S. tuberosum ssp. andigena) and tomato (Lycopersicon esculentum). Comparison of the sequences of 1.35-kbp HaeIII ctDNA fragments from potato, tomato, and tobacco (Nicotiana tabacum) revealed the following: the locations of the 5 ends of both rubisco large subunit (rbcL) and ATPase beta subunit (atp) mRNAs were probably the same as those of spinach (Spinacia oleracea); the promoter regions of the two genes were highly conserved among the four species; and the 5 untranslated regions diverged at high rates. A phylogenetic tree for the three potato cultivars, one tomato cultivar, and one tobacco cultivar has been constructed by the maximum parsimony method from DNA sequence data, demonstrating that the rate of nucleotide substitution in potato ctDNA is much slower than that in tomato ctDNA. This fact might be due to the differences in the method of propagation between the two crops.     

Accumulation of vitamin E in potato (Solanum tuberosum) tubers

Vitamin E (tocopherol) is a powerful antioxidant essential for human health and synthesized only by photosynthetic organisms. The effects of over-expression of tocopherol biosynthetic enzymes have been studied in leaves and seeds, but not in a non-photosynthetic, below-ground plant organ. Genetic and molecular approaches were used to determine if increased levels of tocopherols can be accumulated in potato (Solanum tuberosum L.) tubers through metabolic engineering. Two transgenes were constitutively over-expressed in potato: Arabidopsis thaliana p-hydroxyphenylpyruvate dioxygenase (At-HPPD) and A. thaliana homogentisate phytyltransferase (At-HPT). α-Tocopherol levels in the transgenic plants were determined by high-performance liquid chromatography. In potato tubers, over-expression of At-HPPD resulted in a maximum 266% increase in α-tocopherol, and over-expression of At-HPT yielded a 106% increase. However, tubers from transgenic plants still accumulated approximately 10- and 100-fold less α-tocopherol than leaves or seeds, respectively. The results indicate that physiological and regulatory constraints may be the most limiting factors for tocopherol accumulation in potato tubers. Studying regulation and induction of tocopherol biosynthesis should reveal approaches to more effectively engineer crops with enhanced tocopherol content.     

Protein kinase activity in different stages of potato (Solanum tuberosum L.) microtuberization

In vitro culture was used to study morphogenetic aspects of the tuberization process under controlled conditions in potato (Solanum tuberosum L.) plants. This paper accurately defines four stages of tuber development and their correlation to external morphological characteristics and histological structures. Protein kinase activity, assayed in each stage using Historic HAS as substrate, was differentially expressed during the tuberization process. Phosphorylation was maximum in the first stages of tuber formation. The incorporation of [³²PO₄ ^–1] to endogenous peptides containing serine/threonine amino acidic residues followed the same pattern that the protein kinase activity did.Abbreviations EDTA Ethylenediaminetetraacetic acid - EGTA ethylenebis (oxyethylenenitrilo) tetraacetic acid - MOPS 4-morpholine-propanesulfonic acid     

Effect of photoperiod on in vitro tuberization of potato (Solanum tuberosum L.)

Single-node cuttings of potato cultivars Jemseg, Katahdin, Russet Burbank and Superior were cultured on a multiplication medium containing MS salts and no growth regulators. Cultures were exposed to 8 h (SD) and 16 h (LD) photoperiodic regimes. The subsequent plantlets were excised and single node cuttings from each photoperiodic regime were placed under SD or LD on a second medium containing growth regulators which promoted tuberization. Production of microtubers was strongly influenced by genotype and by photoperiodic treatments. Superior produced stunted plantlets and some microtubers under SD conditions in the multiplication medium. The number of microtubers formed by Jemseg was not influenced by photoperiod. However, Katahdin and Russet Burbank formed fewer microtubers under LD-LD conditions compared to LD-SD, SD-SD and SD-LD regimes. Compared with the other regimes, LD-SD photoperiod generally promoted microtuber formation with larger diameters and significantly (p<0.05) greater fresh weight. The intensity of the tuberization stimulus was affected by daylength, and this was characterized by microtubers with secondary tubers, the growth of more than one axillary microtuber, and microtubers subtended by stolons. The maturity group of the potato cultivars and photoperiodic regime in vitro strongly influenced the production of microtubers. These results can be employed to adapt light regimes for multiplication and tuberization to the specific requirements for cultivars from different maturity groups, and thus increase the efficiency of potato multiplication protocols.     

Allotriploid somatic hybrids of diploid tomato (Lycopersicon esculentum Mill.) and monoploid potato (Solanum tuberosum L.)

Allotriploid somatic hybrids were obtained from fusions between protoplasts of diploid tomato and monohaploid potato. The selection of fusion products was carried out in two different ways: (1) The fusion of nitrate reductase-deficient tomato with potato gave rise only to hybrid calli if selection was performed on media lacking ammonium. Parental microcalli were rarely obtained and did not regenerate. (2) The fusion of cytoplasmic albino tomato with potato gave rise to albino and green hybrid calli and plants. Allotriploids were identified from the two somatic hybrid populations by counting chloroplast numbers in leaf guard cells and by flow cytometry of leaf tissue. Although some pollen fertility of allotriploids and pollen-tube growth of tomato, potato andLycopersicon pennellii into the allotriploid style were observed, no progeny could be obtained. The relevance of allotriploid somatic hybrids in facilitating limited gene transfer from potato to tomato is discussed.     

Characterization of chromosome instability in interspecific somatic hybrids obtained by X-ray fusion between potato (Solanum tuberosum L.) and S. brevidens Phil

Summary Asymmetric somatic hybrids between Solanum tuberosum L. and S. brevidens Phil. have been obtained via the fusion of protoplasts from potato leaves and from cell suspension culture of S. brevidens. The wild Solanum species served as donor after irradiation of its protoplasts with a lethal X-ray dose (200 Gy). Selection of the putative hybrids was based on the kanamycin-resistance marker gene previously introduced into the genome of Solanum brevidens by Agrobacterium-mediated gene transfer. Thirteen out of the 45 selected clones exhibited reduced morphogenic potential. The morphological abnormalities of the regenerated plantlets were gradually eliminated during the extended in vitro culture period. Cytological investigations revealed that the number of chromosomes in the cultured S. brevidens cells used as protoplast source ranged between 28–40 instead of the basic 2n=24 value. There was a high degree of aneuploidy in all of the investigated hybrid clones, and at least 12 extra chromosomes were observed in addition to the potato chromosomes (2n=48). Interand intraclonal variation and segregation during vegetative propagation indicated the genetic instability of the hybrids, which can be ascribed to the pre-existing and X-ray irradiation-induced chromosomal abnormalities in the donor S. brevidens cells. The detection of centromeric chromosome fragments and long, poly-constrictional chromosomes in cytological preparations as well as non-parental bands in Southern hybridizations with restriction fragment length polymorphism (RFLP) markers revealed extensive chromosome rearrangements in most of the regenerated clones. On the basis of the limited number of RFLP probes used, preferential loss of S. brevidens specific markers with a non-random elimination pattern could be detected in hybrid regenerants.     

Cotransformation and differential expression of introduced genes into potato (Solanum tuberosum L.) cv Bintje

Summary The Dutch potato cultivar Bintje has been transformed by Agrobacterium strain LBA1060KG, which contains two plasmids carrying three different DNAs (TL- and TR-DNA on the Agrobacterium rhizogenes plasmid and TKG-DNA on the pBI121 plasmid). Several transformed root clones were obtained after transformation of leaf, stem, and tuber segments, and plants were then regenerated from these root clones. The expression of the various marker genes [rol, opine, -glucuronidase (GUS), and neomycin phosphotransferase (NPTII)] was determined in several root clones and in regenerated plants. The selection of vigorously growing root clones was as efficient as selection for kanamycin resistance. In spite of the location of NPTII and GUS genes on the same T-DNA, 17% of the root clones did not show GUS activity. Nevertheless, Southern blot analysis showed that these root clones contained at least three copies of the GUS gene. Sixty-four per cent of the root clones contained opines. The expression of these genes, however, was negatively correlated with plant regeneration capacity and normal plant development. The differential expression of the marker genes in the transgenic potato tissues is discussed.     

Inheritance of three electrophoretically determined protein bands in potato (Solanum tuberosum L.)

Summary A modified polyacrylamide gel electrophoresis technique is employed to resolve proteins for use as biochemical gene markers in potato. Dominant, duplicate dominant and complementary gene action are three modes of inheritance that adequately explain the segregation of three respective protein bands in two generations of crossing within diploid Phureja X haploid Tuberosum families.Scientific Journal Seires Article 10,171 of the Minnesota Agricultural Experiment Station     

Isolation of an amylose-free starch mutant of the potato (Solanum tuberosum L.)

Summary An amylose-free potato mutant was isolated after screening 12,000 minitubers. These minitubers had been induced on stem segments of adventitious shoots, which had been regenerated on leaf explants of a monoploid potato clone after Röntgen-irradiation. The mutant character is also expressed in subterranean tubers and in microspores. Starch granules from the mutant showed a strongly reduced activity of the granule bound starch synthase and loss of the major 60 kd protein from the starch granules.     

Efficient transformation of potato (Solanum tuberosum L.) using a binary vector in Agrobacterium rhizogenes

Summary We transformed three potato (Solanum tuberosum L.) genotypes by using A. rhizogenes or a mixture of A. rhizogenes and A. tumefaciens. Inoculations of potato stem segments were performed with Agrobacterium rhizogenes AM8703 containing two independent plasmids: the wild-type Ri-plasmid, pRI1855, and the binary vector plasmid, pBI121. In mixed inoculation experiments, Agrobacterium rhizogenes LBA1334 (pRI1855) and Agrobacterium tumefaciens AM8706 containing the disarmed Ti-plasmid (pAL4404) and the binary vector plasmid (pBI121) were mixed in a 11 ratio. The T-DNA of the binary vector plasmid pBI121 contained two marker genes encoding neomycin phosphotransferase, which confers resistance to kanamycin, and -glucuronidase. Both transformation procedures gave rise to hairy roots on potato stem segments within 2 weeks. With both procedures it was possible to obtain transformed hairy roots, able to grow on kanamycin and possessing -glucuronidase activity, without selection pressure. The efficiency of the A. rhizogenes AM8703 transformation, however, was much higher than that of the mixed transformation. Up to 60% of the hairy roots resulting from the former transformation method were kanamycin resistant and possessed -glucuronidase activity. There was no correlation between the height of the kanamycin resistance and that of the -glucuronidase activity in a root clone. Hairy roots obtained from a diploid potato genotype turned out to be diploid in 80% of the cases. Transformed potato plants were recovered from Agrobacterium rhizogenes AM8703-induced hairy roots.     

Subcellular distribution of carbonic anhydrase in Solanum tuberosum L. leaves

The intracellular compartmentation of carbonic anhydrase (CA; EC 4.2.1.1), an enzyme that catalyses the reversible hydration of CO₂ to bicarbonate, has been investigated in potato (Solanum tuberosum L.) leaves. Although enzyme activity was mainly located in chloroplasts (87% of total cellular activity), significant activity (13%) was also found in the cytosol. The corresponding CA isoforms were purified either from chloroplasts or crude leaf extracts, respectively. The cytosolic isoenzyme has a molecular mass of 255 000 and is composed of eight identical subunits with an estimated M _r of 30000. The chloroplastic isoenzyme (M _r 220000) is also an octamer composed of two different subunits with M _r estimated at 27 000 and 27 500, respectively. The N-terminal amino acid sequences of both chloroplastic CA subunits demonstrated that they were identical except that the M _r-27 000 subunit was three amino acids shorter than that of the M _r-27 500 subunit. Cytosolic and chloroplastic CA isoenzymes were found to be similarly inhibited by monovalent anions (Cl^–, I^–, N ₃ ^- and NO ₃ ^- ) and by sulfonamides (ethoxyzolamide and acetozolamide). Both CA isoforms were found to be dependent on a reducing agent such as cysteine or dithiothreitol in order to retain the catalytic activity, but 2-mercaptoethanol was found to be a potent inhibitor. A polyclonal antibody directed against a synthetic peptide corresponding to the N-terminal amino acid sequence of the chloroplastic CA monomers also recognized the cytosolic CA isoform. This antibody was used for immunocytolocalization experiments which confirmed the intracellular compartmentation of CA: within chloroplasts, CA is restricted to the stroma and appears randomly distributed in the cytosol.Abbreviations BSA bovine serum albumin - CA carbonic anhydrase - PMSF phenylmethylsulphonyl fluoride - BAM benzamidine - DTT dithiothreitol - 2-ME 2-mercaptoethanol - PVDF polyvinylidene difluoride The authors thanks P. Carrier and Dr. B. Dimon for technical assistance with the mass-spectrometry measurements.