Inducers of glycinebetaine synthesis in barley.

Glycinebetaine is an osmoprotectant accumulated by barley (Hordeum vulgare) plants in response to high levels of NaCl, drought, and cold stress. Using barley seedlings in hydroponic culture, we characterized additional inducers of glycinebetaine accumulation. These included other inorganic salts (KCl, MgCl(2), LiCl, and Na(2)SO(4)), oxidants (H(2)O(2) and cumene hydroperoxide), and organic compounds (abscisic acid, polymixin B, n-butanol, salicylic acid, and aspirin). Stress symptoms brought on by high NaCl and other inducers, and not necessarily correlated with glycinebetaine accumulation, include wilting, loss of chlorophyll, and increase in thiobarbituric acid reacting substances. For NaCl, Ca(2+) ions at 10 to 20 mM decrease these stress symptoms without diminishing, or even increasing, glycinebetaine induction. Abscisic acid induces glycinebetaine accumulation without causing any of the stress symptoms. NaCl, KCl, and H(2)O(2) (but not other inducers) induce glycinebetaine at concentrations below those needed for the other stress symptoms. Mg(2+) at 10 to 20 mM induces both stress symptoms and glycinebetaine, but only at low (0.2 mM) Ca(2+). Although illumination is needed for optimal induction, a significant increase in the leaf glycinebetaine level is found in complete darkness, also.

An isozyme of betaine aldehyde dehydrogenase in barley.

Betaine aldehyde dehydrogenase (BADH) is an important enzyme for Gly betaine synthesis. We isolated two types of BADH cDNAs (BBD1 and BBD2) from barley. As BBD1 contained the signal sequence (SKL) targeting to microbodies, BBD2 was more similar to previously reported genes coding for BADH in dicotyledons (chloroplast type) than those in monocotyledons (microbody type). The two barley BADH genes showed different expression patterns. The BBD1 transcript was more abundant in roots than leaves and was induced to higher levels by salt, drought and abscisic acid (ABA) treatment. BBD2 transcript was more abundant in leaves and induced by salt, drought, PEG and ABA treatment. To understand the processing of these BADH proteins, we partially purified both enzymes and determined their N-terminal sequences. Based on comparisons of the N-terminal sequences to their deduced amino acid sequence, neither BBD1 nor BBD2 is processed at the N-terminus. These results suggest that BBD2 codes for a new type of BADH, which is not localized in either chloroplasts or mitochondria.

The chloroplast-located homolog of bacterial DNA recombinase.

The cDNA for the chloroplast-located homolog of bacterial RecA protein, designated recA-AT, was placed in a plasmid appropriate for in vitro transcription and translation. Translation with 35S-labeled Met permitted demonstration of uptake of the protein product into isolated pea chloroplasts, and processing to a mature size. Preliminary evidence for the first amino acid was estimated from results using both 35S-Met and 3H-Leu for in vitro transcription and translation, followed by uptake into chloroplasts and processing. The labeled protein was subject to sequential amino acid hydrolyses, and radioactivity was measured in each round. Induction of gene transcription in leaves infiltrated with the DNA-damaging agent, methyl methane-sulfonate was shown by Northern blot analysis. Further constructs were made for over-expression of the gene in E. coli; and one out of many tried permitted production of some soluble protein. Extracts from transformed bacteria were shown to have RecA activity using the "POM" assay [Bertrand et al. (1993) Nucl. Acids Res. 21:3653] for DNA strand transfer. The protein was purified to close to homogeneity using methods developed for E. coli RecA isolation.

Some unique characteristics of thylakoid unisite ATPase.

Under unisite conditions (ratio of ATP to chloroplast coupling factor (CF0CF1), approximately 1:2.8), spinach thylakoid ATPase depends on prior reductive activation of CF1, just as multisite ATPase does, and is sensitive to removal of CF1 by EDTA. Faster rates in room light than in semidarkness and up to 80% inhibition by uncouplers only in room light indicate a strong effect of protonmotive force, which can be provided by room light. In addition, unisite ATPase is inhibited by azide as long as some ADP is bound to the CF1. Several differences were found between unisite and multisite ATPase. 1) The unisite activities of both membrane-bound and free enzyme were stimulated up to 3-fold by 4 mM free MgCl2 (a strong inhibitor of multisite ATPase). 2) Thylakoid unisite ATPase was inhibited by sulfite (50% inhibition at 5 mM), a powerful activator of multisite ATPase. This inhibition is attributed to a nonspecific ionic strength effect. 3) Unisite ATPase was inhibited by trypsin treatment, which increases multisite ATPase severalfold. 4) The pH profile of thylakoid unisite ATPase is somewhat different from that of multisite. 5) Alkylation of Cys-89 of the gamma subunit by N-ethylmaleimide did not affect the unisite activity, but inhibited multisite activity more than 90%.

Chloroplast molecular chaperone-assisted refolding and reconstitution of an active multisubunit coupling factor CF1 core.

The chloroplast coupling factor 1 (CF1) is composed of five kinds of subunits with a stoichiometry of alpha 3 beta 3 gamma delta epsilon. Reconstitution of a catalytically active alpha 3 beta 3 gamma core from urea-denatured subunits at a physiological pH is reported here. A restoration of approximately 90% of the CF1 ATPase activity has been observed. The reconstitution was achieved by using subunits overexpressed in Escherichia coli, purified, and combined in the presence of MgATP, K+, and a mixture of several chloroplast molecular chaperones at pH 7.5. The combination of chaperonin 60 and chaperonin 24 failed to reconstitute the active CF1 core, as did the GroEL/GroES pair (E. coli chaperonin 60/10 homologues). Characteristics of the reconstituted ATPase were very close to those of the native complex, including methanol-reversible inhibition by the purified epsilon subunit of CF1 and sensitivity to inhibition by azide and by tentoxin. In reconstitution with a mixture of tentoxin-resistant and -sensitive beta subunits, the extent of inhibition by tentoxin depended on the proportion of sensitive subunits in the reconstitution mixture. Finally, a model for the assembly of the CF1 core alpha 3 beta 3 gamma structure is proposed.

A Purified Zinc Protease of Pea Chloroplasts, EP1, Degrades the Large Subunit of Ribulose-1,5-Bisphosphate Carboxylase/Oxygenase.

A previously reported endopeptidase (EP1) from pea chloroplasts was purified over 11,000-fold using a four-step protocol involving ultrafiltration, sucrose gradient centrifugation, isoelectric focusing, and high performance liquid chromatography gel filtration. The enzyme was determined to be a metalloprotease requiring bound Zn2+ and added Mg2+ or Ca2+ for proper activity. Its localization in the stroma of pea chloroplasts was confirmed by demonstrating its insensitivity to thermolysin when the envelope was intact. A contaminating serine protease that attacks EP1 was found. The contaminating protease was inhibited by 4-(2-aminoethyl)-benzenesulfonyl fluoride, but not by o-phenanthroline, whereas EP1 sensitivities were the reverse. EP1 is able to hydrolyze the large subunit of native ribulose-1,5-bisphosphate carboxylase/oxygenase under physiological conditions.

DNA Strand-Transfer Activity in Pea (Pisum sativum L.) Chloroplasts.

The occurrence of DNA recombination in plastids of higher plants is well documented. However, little is known at the enzymic level. To begin dissecting the biochemical mechanism(s) involved we focused on a key step: strand transfer between homologous parental DNAs. We detected a RecA-like strand transfer activity in stromal extracts from pea (Pisum sativum L.) chloroplasts. Formation of joint molecules requires Mg2+, ATP, and homologous substrates. This activity is inhibited by excess single-stranded DNA (ssDNA), suggesting a necessary stoichiometric relation between enzyme and ssDNA. In a novel assay with Triton X-100-permeabilized chloroplasts, we also detected strand invasion of the endogenous chloroplast DNA by 32P-labeled ssDNA complementary to the 16S rRNA gene. Joint molecules, analyzed by electron microscopy, contained the expected displacement loops.

Treatment of pea (Pisum sativum L.) protoplasts with DNA-damaging agents induces a 39-kilodalton chloroplast protein immunologically related to Escherichia coli RecA.

Organisms must have efficient mechanisms of DNA repair and recombination to prevent alterations in their genetic information due to DNA damage. There is evidence for DNA repair and recombination in plastids of higher plants, although very little is known at the biochemical level. Many chloroplast proteins are of eubacterial ancestry, suggesting that the same could be true for the components of a DNA repair and recombination system. A 39-kD protein, immunologically related to Escherichia coli RecA, is present in chloroplasts of pea (Pisum sativum L.). Bandshift gel assays suggest that it binds single-stranded DNA. Its steady-state level is increased by several DNA-damaging agents. These results are consistent with it being a plastid homolog of E. coli RecA protein, presumably involved in DNA repair and recombination, and with the existence of an SOS-like response in pea leaf cells. Experiments with protein synthesis inhibitors suggest that the 39-kD chloroplast protein is encoded in the nucleus.

Import and assembly of the beta-subunit of chloroplast coupling factor 1 (CF1) into isolated intact chloroplasts.

The transit peptide gene of the small subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase from Nicotiana plumbaginifolia was fused to the tentoxin-resistant beta-subunit gene of chloroplast coupling factor 1 (CF1) from Nicotiana tabacum via a linker sequence. The consequent fusion gene encodes the entire gene sequences of both the transit peptide and the beta-subunit of CF1 without a single change of amino acid residues. The fusion gene was in vitro expressed in a coupled transcription-translation system as a 62-kDa precursor and was imported into isolated intact chloroplasts of Nicotiana longiflora. The imported precursor was found to be processed to the expected mature beta-subunit size. Evidence is presented that radioactively labeled beta-subunit was incorporated into mature CF1 and not just nonspecifically associated with the thylakoid membranes. Since the fusion protein containing only the transit peptide sequence was imported and apparently correctly processed, it may not be necessary to include N-terminal amino acids of the mature small subunit for correct proteolytic cleavage in the chloroplast stroma.

Inhibition of Thylakoid ATPase by Venturicidin as an Indicator of CF1-CF0 Interaction.

Venturicidin inhibits the F0 portion of membrane-located, H+-pumping ATPases. We find it meets the criteria for an energy transfer inhibitor for spinach (Spinacia oleracea) thylakoids: complete inhibition of photophosphorylation and of photophosphorylation-stimulated and basal electron flow rates, but not of electron flow under uncoupled conditions. The extent of H+ uptake in the light is stimulated by venturicidin (vtcd), as expected for a compound blocking H+ efflux through CF0. Vtcd had no effect on the nonproton pumping, methanol-stimulated ATPase of thylakoids or on soluble CF1 ATPase. Under totally uncoupled conditions (saturating NH4Cl + gramicidin), vtcd can still inhibit sulfite-stimulated thylakoid ATPase completely. The concentration of vtcd needed for inhibition of ATPase was proportional to the concentration of thylakoids present in the assay, with an apparent stoichiometry of about 10 vtcd molecules per CF1/CF0 for 50% inhibition. Vtcd raised the Km for ATP somewhat, but had a stronger effect on the Vmax with respect to ATP. Inhibition by saturating vtcd ranged from 50 to 100%, depending on the condition of the thylakoids. Grinding leaves in buffer containing 0.2 M choline chloride (known to provide superior photophosphorylation rates) helped bring on maximum vtcd inhibition; trypsin treatment or aging of thylakoids brought on vtcd-resistant ATPase. We conclude that the extent of inhibition by vtcd can be used as an indicator of the tightness of coupling between CF1 and CF0.

The Isolation of Actin from Pea Roots by DNase I Affinity Chromatography.

Native actin can be isolated from pea (Pisum sativum L.) roots by DNase I affinity chromatography, but the resulting yields and quality of actin are variable. By use of two assays for actin, a DNase I inhibition assay and a gel scanning assay, we identified several factors that increased actin yield. ATP is required for the actin in crude pea root extracts to bind to immobilized DNase I. Low amounts of ATP are hydrolyzed rapidly by an endogenous ATPase in the extract, and the actin then irreversibly loses the ability to bind to DNase I. High ATP concentrations (5-10 mm) or inhibition of the ATPase (with 10 mm pyrophosphate) are required for pea actin to retain DNase I binding ability. When adequate amounts of ATP are present, actin binding from the extract is further enhanced by basic pH, formamide, and soluble polyvinyl-pyrrolidone. Once actin is bound to the DNase I-agarose and washed free of extract, high ATP concentrations are not required to keep actin bound. Actin eluted from the DNase I-agarose with formamide retained its ability to polymerize into filaments with the addition of KCl and Mg(2+). The advantages and disadvantages of this procedure and its application to other plant materials are discussed.

A homolog of Escherichia coli RecA protein in plastids of higher plants.

Studies of chloroplast DNA variations, and several direct experimental observations, indicate the existence of recombination ability in algal and higher plant plastids. However, no studies have been done of the biochemical pathways involved. Using a part of a cyanobacterial recA gene as a probe in Southern blots, we have found homologous sequences in total DNA from Pisum sativum and Arabidopsis thaliana and in a cDNA library from Arabidopsis. A cDNA was cloned and sequenced, and its predicted amino acid sequence is 60.7% identical to that of the cyanobacterial RecA protein. This finding is consistent with our other results showing both DNA strand transfer activity and the existence of a protein of the predicted molecular mass crossreactive with antibodies to Escherichia coli RecA in the stroma of pea chloroplasts.

Further characterization of ribosome binding to thylakoid membranes.

Previous work indicated more polysomes bound to pea (Pisum sativum cv Progress No. 9) thylakoids in light than in the dark, in vivo (LE Fish, AT Jagendorf 1982 Plant Physiol 69: 814-825). With isolated intact chloroplasts incubated in darkness, addition of MgATP had no effect but 24 to 74% more RNA was thylakoid-bound at pH 8.3 than at pH 7. Thus, the major effect of light on ribosome-binding in vivo may be due to higher stroma pH. In isolated pea chloroplasts, initiation inhibitors (pactamycin and kanamycin) decreased the extent of RNA binding, and elongation inhibitors (lincomycin and streptomycin) increased it. Thus, cycling of ribosomes is controlled by translation, initiation, and termination. Bound RNA accounted for 19 to 24% of the total chloroplast RNA and the incorporation of [(3)H]leucine into thylakoids was proportional to the amount of this bound RNA. These data support the concept that stroma ribosomes are recruited into thylakoid polysomes, which are active in synthesizing thylakoid proteins.

Site of Synthesis of the Enzymes of the Pyrimidine Biosynthetic Pathway in Oat (Avena sativa L.) Leaves.

Heat-bleached oat (Avena sativa L. cv Porter) leaves lacking 70S chloroplast ribosomes have been used to demonstrate that four chloroplast-localized enzymes of pyrimidine nucleotide biosynthesis: aspartate carbamoyl-transferase, dihydroorotase, orotidine phosphoribosyl-transferase, and orotidine-5'-phosphate decarboxylase, are synthesized on cytoplasmic ribosomes. Two other chloroplast enzymes, carbamoyl phosphate synthetase, involved in both pyrimidine and arginine biosynthesis, and ornithine carbamoyltransferase, an enzyme of arginine biosynthesis, were also shown to be made on 80S ribosomes.

Extended X-ray absorption fine structure of Mn2+ and Mn2+ X ATP complex bound to coupling factor 1 of the H+-ATPase from chloroplasts.

The spinach chloroplast ATPase, coupling factor 1, contains three tight Mn2+-binding sites which interact cooperatively. The bound manganese coordinations were studied by x-ray absorption fine structure analysis. Mn2+ was found to be bound to the enzyme with an average Mn-O bond length of 2.15 +/- 0.15 A, significantly shorter than the 2.15 +/- 0.15 A of the Mn-O bond of the average first hydration shell for Mn2+ in aqueous solution. On adding ATP to the manganese-enzyme mixture, a tertiary complex of Mn2+ X ATP X enzyme was formed as indicated by the appearance of a second shell. Mn-P bond distances were estimated at 4.95 +/- 0.15 A in the tertiary Mn2+ X ATP X enzyme complex, which was considerably longer than the Mn-P bond distance of 3.36 +/- 0.15 A for the Mn2+ X ATP complex in aqueous solution. The Mn-P bond distance in the tertiary Mn2+ X ATP X enzyme complex decreased to 4.32 +/- 0.15 A when selenite, a potent effector of ATPase activity, was added. Based on these results, it is suggested that the tertiary complex is required for catalysis. The stimulation of ATP hydrolysis by anions such as selenite may be the result of shortening the distance between Mn2+ and the ATP phosphates in the enzyme active site.

Neutral peptidases in the stroma of pea chloroplasts.

One endopeptidase (EP1) and at least three aminopeptidases (AP1, AP2, and AP3) were discovered in the stroma of chloroplasts isolated from pea seedlings (Pisum sativum L.), and purified over 100-fold. EP1 requires added Mg(2+) or Ca(2+) for activity, may have an additional tightly bound metal atom, and is inhibited by sulfhydryl reagents but not by serine residue-directed inhibitors. It is reversibly inhibited by dithiothreitol. Its specificity is for the bond between two adjacent Ala or Gly residues. Its molecular mass is 93 kilodaltons, estimated on a gel filtration column. Aminopeptidase activities were detected with the aid of different amino acyl-beta-naphthylamides as substrates. They were resolved into at least three individual proteins by gel filtration and DEAE-cellulose chromatography, having apparent molecular masses of 269,000 (AP1), 84,000 (AP2), and 42,000 (AP3) daltons, respectively. Each has a unique specificity for substrates, with AP1 hydrolyzing only the Prolyl-beta-naphthylamide. None of the APs require added divalent cations for activity, but the possibility of a tightly bound metal function was suggested in AP2 and AP3 (not AP1) from effects of inhibitors. A probable sulfhydryl residue function was indicated for all three, from inhibition by p-hydroxymercuribenzoate and Zn(2+). All these peptidases had pH optima at 7.7.

Evidence for multiple effects in the methanol activation of chloroplast coupling factor 1.

Activation of the latent ATPase of soluble CF1 by methanol is shown to involve several distinct effects. CaATPase activity of whole, but not epsilon-deficient or heat-activated CF1, is stimulated by methanol. This suggests that one effect of methanol is to overcome inhibition by the epsilon subunit. In contrast, the MgATPase activities of both whole and epsilon-deficient CF1 are further stimulated by methanol. This second activating effect can be traced in part to a greatly increased affinity of CF1, due to methanol, for those anions which reverse the inhibitory effect of Mg2+. Since the inhibition by free Ca2+ is much less severe than that caused by Mg2+, anions have relatively little effect on CaATPase. Thus methanol has little or no effect when Ca2+ is the divalent cation, but stimulates the reaction when Mg2+ is used. Methanol also stimulates the MgATPase activity of epsilon-deficient CF1 in the complete absence of activating anions. This additional effect is shown to arise from an increase in the Vmax rather than from changes in either the Km for MgATP or the Ki for free Mg2+. Since this change in Vmax occurs with the MgATPase but not the CaATPase, it can be inferred that different steps are rate-limiting in the two activities.