Transformation of glutamate to delta-aminolevulinic acid by soluble extracts of Synechocystis sp. PCC 6803 and other oxygenic prokaryotes.

delta-Aminolevulinic acid is the first committed precursor in the biosynthesis of hemes, phycobilins, and chlorophylls. Plants and algae synthesize delta-aminolevulinic acid from glutamate via an RNA-dependent 5-carbon pathway. Previous reports demonstrated that cyanobacteria form delta-aminolevulinic acid from glutamate in vivo. We now report the direct measurement of this activity in vitro. Three oxygenic prokaryotes were examined, the unicellular cyanobacteria Synechocystis sp. PCC 6803 and Synechococcus sp. PCC 7002 (Agmenellum quadruplicatum PR-6) and the chlorophyll a- and b-containing filamentous prochlorophyte Prochlorothrix hollandica. delta-Aminolevulinic acid-forming activity was detected in soluble extracts of all three species. delta-Aminolevulinic acid formation by Synechocystis extracts was further characterized. Activity depended upon addition of reduced pyridine nucleotide, ATP, and Mg2+ to the incubation mixture. NADPH was a more effective pyridine nucleotide than NADH at low concentrations, but NADPH inhibited delta-amino-levulinic acid formation above 1 mM, whereas NADH did not. The pH optimum was about 7.6, and the ATP concentration optimum was 0.1 mM. Activity was stimulated by addition of RNA derived from Synechocystis or Chlorella, and abolished by preincubation with RNase A. After RNase inactivation, activity was restored by addition of RNasin to block further RNase action, followed by supplementation with Synechocystis RNA. Activity was inhibited by micromolar concentrations of hemin, as was previously found with plant and algal extracts. Complete dependence on added glutamate could not be achieved. Radioactivity was incorporated into delta-aminolevulinic acid when the incubation mixture contained 1-[14C]glutamate. Activity in the Synechocystis enzyme extract was stimulated by the addition of a partially purified enzyme fraction from Chlorella. It thus appears that prokaryotic oxygenic organisms share with chloroplasts the capacity for biosynthesis of photosynthetic pigments from glutamate via the RNA-dependent 5-carbon pathway.

Transformation of Glutamate to 6-Aminolevulinic Acid by Soluble Extracts of Synechocystis sp. PCC 6803 and Other Oxygenic Prokaryotes* (Received for publication, October 26, 1987) Siegfried RiebleS and Samuel I. Beale8 From the Division of Biology and Medicine, Brown Uniuersity, Providence, Rho& Island 02912 6-Aminolevulinic acid is the first committed precursor in the biosynthesis of hemes, phycobilins, and chlorophylls. Plants and algae synthesize 6-aminolevulinic acid from glutamate via an RNA-dependent 5-carbon pathway. Previous reports demonstrated that cyanobacteria form 6-aminolevulinic acid from glutamate in vivo. We now report the direct measurement of this activity in vitro. Three oxygenic prokaryotes were examined, the unicellular cyanobacteria Synechocystis sp. PCC 6803 and Synechococcus sp. PCC 7002 (Agmenellum quadruplicatum  and the chlorophyll aand b-containing filamentous prochlorophyte Prochlorothrix hollandica. 6-Aminolevulinic acidforming activity was detected in soluble extracts of all three species. 6-Aminolevulinic acid formation by Synechocystis extracts was further characterized. Activity depended upon addition of reduced pyridine nucleotide, ATP, and M 8 + to the incubation mixture. NADPH was a more effective pyridine nucleotide than NADH at low concentrations, but NADPH inhibited &aminolevulinic acid formation above 1 mM, whereas NADH did not. The pH optimum was about 7.6, and the ATP concentration optimum was 0.1 mM. Activity was stimulated by addition of RNA derived from Synechocystis or Chlorella, and abolished by preincubation with RNase A. After RNase inactivation, activity was restored by addition of RNasin to block further RNase action, followed by supplementation with Synechocystis RNA. Activity was inhibited by micromolar concentrations of hemin, as was previously found with plant and algal extracts. Complete dependence on added glutamate could not be achieved. Radioactivity was incorporated into 6-aminolevulinic acid when the incubation mixture contained l-['4C]glutamate. Activity in the Synechocystis enzyme extract was stimulated by the addition of a partially purified enzyme fraction from Chlorella. It thus appears that prokaryotic oxygenic organisms share with chloroplasts the capacity for biosynthesis of photosynthetic pigments from glutamate via the RNA-dependent 6-carbon pathway. 6-Aminolevulinic acid is the first committed precursor in the biosynthesis of tetrapyrroles including hemes, phycobilins, and chlorophylls (1). Two distinct pathways for 6-ami-* This work was supported by National Science Foundation Grant DMB85-18580. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C.
Section 1734 solely to indicate this fact.
$Recipient of a fellowship from the Deutscher Akademischer Austauschdienst (DAAD).
Active extracts have also been obtained from the methanogen Methanobacterium thermoautotrophicum (19). Soluble extracts of barley plastids (20) and Chlorella cells (21) have been partially purified and fractionated into three proteinaceous components, all of which are required, in addition to an RNA component, for the reconstitution of 6-aminolevulinic acidforming activity. These extracts require a reduced pyridine nucleotide, ATP, M$+, and a low molecular weight RNA as cofactors in vitro. A current working hypothesis for the route of 6-aminolevulinic acid formation from glutamate begins with activation of the C1 carboxyl group of glutamate by a reaction analogous or identical to the formation of glutamyl-tRNA for protein synthesis (Fig. 1). The activated glutamate is then reduced in a pyridine nucleotide-dependent step, and the reduced product, which may be free glutamate l-semialdehyde, is finally transaminated, perhaps internally, to form 6aminolevulinic acid. In the case of barley, the RNA component has been identified as tRNAG1"'UUC' by nucleotide sequence determination (22). It has recently been shown by anticodon-based affinity chromatography that the active RNA components from spinach, Euglena, Chlorella, and Cyanidium all carry the UUC glutamate anticodon (23). A UUC anticodon-containing RNA that was capable of supporting 6aminolevulinic acid formation by Chlorella preparations was also extracted from the cyanobacterium Synechocystis (23).
Cyanobacteria are potentially attractive experimental materials for isolation of the genes coding for the macromolecular components of the 6-aminolevulinic acid-forming system. Until now, the only direct evidence supporting the operation of the 5-carbon 6-aminolevulinic acid-forming pathway in cyanobacteria has been based on in vivo label incorporation (24)(25)(26)(27). We report here the formation of 6-aminolevulinic acid from glutamate in soluble extracts of two cyanobacteria, Synechocystis sp. PCC 6803 and Synechococcus sp. PCC 7002, and of another oxygenic prokaryote, the filamentous prochloro- phyte Prochlorothrix hollundica. A brief account of this work has appeared in abstract form (28).

MATERIALS AND METHODS
Growth of CeUs-Synechocystis sp. PCC 6803 was a gift from J. G.
K. Williams (DuPont-Central Research). Axenic cultures were maintained under photoheterotrophic conditions with continuous light (32 pE m-' s-I), supplied by equal numbers of cool-white and red fluorescent tubes. BG-11 growth medium (29) was supplemented with 50 mM dextrose. Growth temperature was 23 "C and cells were kept in suspension by rotary shaking. Synechococcus sp. PCC 7002 was a gift from R. Webb (Temple University, Philadelphia). Axenic cultures were grown in a water bath shaker at 32 "C. The growth medium consisted of 250 mM NaCl, 30 mM KN03, 10 mM glycerol, 2.5 mM MgSO,, 2.5 mM NaHzP04, 2.5 mM Na2HP04, 0.1 mM CaCh 20 p M NaEDTA, 20 p~ NaZEDTA, 20 p~ FeC13, 10 pM MnCh 10 pM ZnS04, 1 p M H3BOs, 1 pM Cus01,l pM Na3V04, 1 p M (NH&hfo&, 1 PM vitamin Biz, and 0.1 p~ CoClZ. The light conditions were the same as for Synechocystis. P. hollandica, a Chl a-and b-containing filamentous prochlorophyte (30) was obtained from K. R. Miller (Brown University) and grown photoautotrophically in BG-11 medium at 23 'C. Cultures were maintained in light conditions used for the other species described, and the culture flasks were continuously flooded with sterile air. E. gracilis Klebs strain Z Pringsheim was cultured at 23 "C in light in a glucose-based heterotrophic medium as described previously (31).
Preparation of Cell Extract-Cells of Synechocystis and Synechococcus were harvested by centrifugation at 4 "C for 10 min at 10,000 X g near the end of their exponential growth phase at a density between 0.9 X 10' and 1.1 X 10' cells ml-'. The cells were washed twice with ice-cold deionized water and then with 25 ml of ice-cold extraction buffer. Two different extraction buffers were used during the course of these experiments. Unless indicated otherwise, the standard extraction buffer was 150 mM Tricine' (pH 7.8), 500 mM glycerol, 40 mM MgClz, 1 mM DTT, 20 p M pyridoxal-P. six liters of cell culture usually yielded about 6 g wet wt of cells. The cells were suspended in ice-cold extraction buffer at a ratio of 1 ml buffer/g cells and passed twice through a French pressure cell at 138 kPa (20,000 p.s.i.). Unbroken cells and cell debris were removed by centrifugation at 10,000 X g for 10 min at 4 "C. The supernatant was adjusted to 0.5 M NaCl and stirred for 20 min on ice, and then centrifuged at 285,000 X g for 90 min at 2 "C. The supernatant was passed through a desalting column of Sephadex G-25 that was preequilibrated and eluted with assay buffer (50 mM Tricine (pH 7.8), 1 M glycerol, 40 mM MgClz, 1 mM DTT, 20 p~ pyridoxal-P). The fraction containing the soluble proteins was stored at -75°C until use.
Late in the course of these studies, a high-phosphate extraction buffer was employed. This buffer contained 0.75 M K-PO4 (pH 7.2), 150 mM Tricine, 500 mM glycerol, 1 mM DTT, and 20 p M pyridoxal-P. When cells were extracted into this buffer, the low-speed centrifugation and salt-stir steps were omitted, and the broken cell suspension was immediately centrifuged at high speed. The resulting highspeed supernatant was largely free of phycobiliproteins, which sedimented as membrane-bound intact phycobilisomes (32), and the supernatant contained the components of the 8-aminolevulinic acidforming system in a purer and more active form.
Assay for in Vitro 6-Aminoleuulinic Acid Formation-Incubations were carried out at 30°C in 1 ml of reaction volume. The reaction ' The abbreviations used are: Tricine, N-tridhydroxy-methy1)methylglycine; DTT, dithiothreitol; Hepes, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid. mixture had the composition of assay buffer but contained in addition (unless otherwise noted) 5 mM levulinic acid, 5 mM ATP, 5 mM NADH, and 1 mM glutamate added together from a concentrated cofactor-substrate mixture, and 1-3 mg of soluble protein. The reaction was started by adding the cofactor-substrate mixture, and stopped after 60 min of incubation by the addition of 100 pl of 1 M citric acid and 1 ml of 10% (w/v) sodium dodecyl sulfate.
Purification and Quantitation of 6-Aminoleuulinic Acid-&Aminolevulinic acid was purified by ion exchange chromatography as previously described (16). The purified 6-aminolevulinic acid was pyrrolized by reaction with ethylacetoacetate (33), the 1-methyl-2-carboxyethyl-3-propionic acid pyrrole product was reacted with Ehrlich-Hg reagent (34), and the As3 was measured. The Aam of unincubated control samples was subtracted from those of incubated samples to determine net Am values, and 6-aminolevulinic acid was calculated from a standard curve with samples containing known amounts of 8aminolevulinic acid.
Paper Chromatography of l-Methyl-2-carboryethyl-3-propwnic Acid Pyrrole-After reaction with ethylacetoacetate, the solution justed to pH 3.0 with HCl and then cooled to 0 "C. The solution was containing l-methyl-2-carboxyethyl-3-propionic acid pyrrole was adthen extracted with three 1-volume portions of diethyl ether. The combined ether extracts were cooled to -75 "C for 1 h. The ether was decanted from the ice that had formed from the water originally present in the ether solution, and the solution was then concentrated under a stream of N,. The l-methyl-2-carboxyethyl-3-propionic acid pyrrole was chromatographed on Whatman No. 3MM paper in a solvent of l-butanol:l-propanol:5% (w/v) aqueous NH4OH (2:l:l) (4). The l-methyl-2-carboxyethyl-3-propionic acid pyrrole was either visualized after chromatography by Ehrlich spray reagent (200 mg of pdimethylaminobenzaldehyde dissolved in 8 ml of ethanol and 2 ml of 12 N HCl) (4) or the lanes were cut into 1-cm segments, immersed in 5.5 ml of scintillation fluid (Econofluor), and radioactivity was determined by liquid scintillation spectroscopy. Synechocystis Small RNA Preparatwn-Low molecular weight RNA was isolated from the soluble cell extract after the high-speed centrifugation. The supernatant was diluted with RNA extraction buffer (10 mM Tris-HC1 (pH 7.5), 10 mM Mg(acetate)z, 100 mM NaCl, 10 mM 8-mercaptoethanol) to a final volume of 4 ml/g of cells (35). Sodium dodecyl sulfate was added from a 10% (w/v) stock solution to a final concentration of 1% (w/v). Proteins were removed by extraction with an equal volume phenol saturated with RNA extraction buffer. The aqueous phase was next extracted three or four times with equal volumes of ch1oroform:isoamyl alcohol (24:1, v/v). The nucleic acids in the aqueous phase were then precipitated by adding 2.5 volumes of ice-cold absolute ethanol and cooling overnight at -25'C. The precipitated nucleic acids were redissolved in RNA extraction buffer and chromatographed on DEAE-cellulose as previously described (35). The fractions containing tRNA were combined and precipitated by adding 2.5 volumes of ice-cold ethanol and cooling overnight at -25'C. Nucleic acids were deacylated by redissolving the precipitate in deacylation buffer (0.5 M Tris-HC1 (pH 8.0)), incubating at room temperature for 2 h, precipitating with ethanol, and washing twice with ethanol (35). The pellet was dried and dissolved in RNA storage buffer (10 mM Tris-HC1 (pH 7.5), 10 mM Mg(acetate),, 100 mM NaCl, 1 mM DTT) at a concentration of 200 AZW units ml-'. This RNA preparation was stored at -25°C for further use.
Extraction and Assay Procedures for 6-Aminolevulinic Acid Synthase-Cells of Synechocystis and Euglena were harvested as described above, but the buffer was changed to one of two d-aminolevulinic acid synthase incubation buffers. 6-Aminolevulinic acid synthase incubation buffer A was 75 mM Hepes (pH 7.8), 50 mM glycine, 25 mM succinic acid, 20 mM MgCl,, 5 mM EDTA, and 0.1 mM pyridoxal-P (31). 6-Aminolevulinic acid synthase incubation buffer B was 150 mM Tricine (pH 7.8), 500 mM glycerol, 50 mM glycine, 25 mM succinic acid, 20 mM MgC12, 1 mM DTT, and 20 p M pyridoxal-P. The supernatant derived from the first centrifugation step after cell disruption was used in the assays. Cell extracts were incubated for 30 min at 37 "C in 8-aminolevulinic acid synthase incubation buffer. The incubations were started by the addition of succinyl-CoA, which was generated by the method of Simon and Shemin (36). The total volume of the incubation mixture was 1 ml. Incubations were terminated by addition of 50 pl of 100% (w/v) aqueous trichloroacetic acid solution. Precipitate was removed by centrifugation for 1 min at 13,000 X g in the microcentrifuge. One-half milliliter of supernatant was adjusted to pH 6.8 with approximately 170 p1 of 0.5 M N&PO4, then 25 pl of ethylacetoacetate was added, and the solution was heated to 95 "C for b-Aminoleuulilccic Acid Formation in Extracts of Oxygenic Prokuryotes 15 min to form l-methyl-2-carboxyethyl-3-propionic acid pyrrole, which was quantitated as described above.
Other Procedures-RNase A (Sigma Type I-AS) was dissolved in water (1 mg rnl-'), heated to 100 "C for 5 min, cooled slowly, and stored at -25 "C. Cell population densities were determined with a Coulter Counter (Model ZBI, Coulter Electronics). Protein concentration was determined by the method of Bradford (37), using bovine serum albumin as a standard.
Chemicals-l-[14C]Glutamate and Econofluor scintillation fluid were obtained from DuPont-New England Nuclear. Cellulose DE-23 was from Whatman. RNasin and agarose-linked bakers' yeast hexokinase were from Sigma. All other reagents were from Sigma, Fisher, or Research Organics.

RESULTS
Characterization of the 6-Aminoleuulinic Acid-forming Reaction in Extracts of Synechocystis-Soluble extracts of Synechocystis formed 6-aminolevulinic acid upon incubation with substrate and cofactors appropriate for 6-aminolevulinic acid formation from glutamate. The incubation mixture contained 40 mM MgClz, 5 mM ATP, 5 mM NADH, 1 mM glutamate, and 5 mM levulinic acid. Levulinic acid, a competitive inhibitor of 6-aminolevulinic acid dehydratase, was included in the incubation medium to prevent further metabolism of the 6aminolevulinic acid. The absorption spectrum of the incubation product after reaction with ethylacetoacetate and Ehrlich reagent was identical to that of authentic 6-aminolevulinic acid. The product after reaction with ethylacetoacetate also had identical chromatographic behavior to that of the authentic 1-methyl-2-carboxyethyl-3-propionic acid pyrrole (see below). The soluble nature of all required cell components was indicated by the presence of activity in the supernatant fraction after high-speed centrifugation. This was true even when an alternate extraction procedure was used, in which cells were disrupted in high-phosphate buffer, which allowed the removal of a substantial fraction of cell protein by sedimenting intact membrane-bound phycobilisomes in the high-speed centrifugation step. The time course for 6-aminolevulinic acid formation during incubation indicated that under the conditions used, the reaction was completed after 30 min, and that no change in 6-aminolevulinic acid content occurred between 30 and 60 min. The amount of 6-aminolevulinic acid formed during the incubation was directly dependent on the amount of protein added to the incubation mixture over the range of 1.5 to 5 mg, the highest amount tested. Maximum &aminolevulinic acid formation occurred at approximately pH 7.6, and the temperature optimum was approximately 30 "C.
Synechocystis extract prepared in high-phosphate extraction buffer had a higher specific 6-aminolevulinic acid-forming activity than extract prepared in standard extraction buffer ( Table I). The high-phosphate concentration preserves the integrity of the phycobilisomes and facilitates their removal at the high-speed centrifugation step. Per mg of Synechocystis protein, the preparation was about 7.5-fold more active than earlier preparations containing extraneous phycobiliproteins. 6-Aminolevulinic acid formation in extracts prepared by either procedure had incomplete dependence on added RNA. This is attributable to endogenous RNA present in the incubation mixtures (see below).
Cofactor Requirements-Added pyridine nucleotide and Mg2+ were required as cofactors, and heat-denatured cell extract was inactive (Table I). Mg+ was not inhibitory at concentrations as high as 100 mM (data not sh wn). A comparison between the relative effectiveness of NADH and NADPH indicated that the NADPH is the pref 1 rred pyridine nucleotide (Fig. 2). However, at concentrations above 1 mM, NADPH inhibited 6-aminolevulinic acid formation, whereas NADH did not. At 5 mM or greater concentration, more 6-  aminolevulinic acid was formed with NADH as the pyridine nucleotide than with NADPH.
With cell extract prepared in standard extraction buffer, added ATP was required (Table I). However, with cell extract prepared in high-phosphate extraction buffer, sufficient residual ATP was carried through the gel filtration step to eliminate the need for added ATP. To demonstrate an ATP requirement in extract prepared in high-phosphate extraction buffer, the extract was preincubated for 10 min at 30 ' C with assay buffer plus agarose-linked hexokinase (10 units ml-' 1 and 25 mM dextrose. The agarose-linked hexokinase was then removed by filtration and the reaction was initiated by adding glutamate, NADPH, levulinic acid, and ATP. The optimal ATP concentration was 0.1 mM, and incubations without added ATP had about 20% of the maximal activity (Fig. 3).
Effects of RNA Supplementation and RNase Digestion-Added RNA was not absolutely required, but it stimulated 6aminolevulinic acid formation (Table I) tract, but the degree of stimulation, per Azm unit added, was less than for Synechocystis RNA. Increasing amounts of 6aminolevulinic acid were formed with increasing amounts of added Synechocystis RNA in the incubation mixture (Fig. 4).
The activity without added RNA is presumably due to the presence of endogenous RNA in the enzyme extract. Preincubation of the extract with RNase for 10 min prior to addition of substrates and cofactors resulted in nearly complete loss of the activity (Table 11). Preincubation of the extract alone without added RNase for 15 min before starting the reaction by the addition of substrate and cofactors had no effect on the resulting activity. The degree of inactivation was dependent on the amount of RNase added, with nearly complete inactivation occurring at 50 ng of added RNase (Fig. 5). After preincubation with RNase, activity could be restored to over 75% of the control value by adding RNasin to inhibit RNase, and then adding the low molecular weight RNA isolated from Synechocystis (Table 11). If RNasin was added before RNase, the amount of 6-aminolevulinic acid formed was 89% of the control value, indicating that neither reagent had other effects on the enzyme activity.
Identification of Glutamate as the 6-Aminoleuulinic Acid  Precursor-Absolute dependence of the reaction on added glutamate could not be demonstrated. In normally prepared cell extracts, omission of glutamate from the incubation mixture resulted in only slightly less 6-aminolevulinic acid accumulation than in complete incubation medium (Table 111). Even after passing the cell extract twice through Sephadex G-25, 6-aminolevulinic acid-forming activity in the absence of added glutamate was still about 80% of the value with complete incubation medium. Addition of 1 p1 of a saturated solution of phenylmethylsulfonyl fluoride in isopropyl alcohol (approximately 0.002% (w/v) final concentration) to the 1-ml incubation mixture had no effect on 6-aminolevulinic acid formation or glutamate dependence.

Effects of RNase digestion and RNA supplementation
An experiment was carried out to assess the possibility that the observed glutamate independence was due to the presence of glutamyl-tRNA in the extract and its co-elution with the protein-containing fraction during gel filtration. Cell extract was preincubated with RNase, then passed once or twice through Sephadex G-25, and finally incubated with RNasin plus deacylated tRNA. Under these conditions, activity in the absence of added glutamate was reduced to about 50% of the value with 1 mM glutamate added (Table 111).
Glutamate was identified as the substrate for the reaction by radioactive label incorporation. The specific activity of the 1-["Clglutamate added to the incubation mixture was 55.3 mCi mmol". After the incubation with l-['4C]glutamate, 6aminolevulinic acid that formed was reacted with ethylace- Effects of various treatments on glutamate dependence Synechocystis cells were grown and extracted, and 6-aminolevulinic acid-forming activity was determined as described in the text. Prior to incubation, cell extract was gel-filtered through Sephadex G-25 once or twice, and/or treated with RNase, followed by addition of RNasin to block further RNase action. The nature and order of treatments is indicated in the first column. Following the preincubation treatments, extract was incubated in medium ( RNasin added toacetate, and the identity and specific radioactivity of the 1methyl-2-carboxyethyl-3-propionic acid pyrrole were determined by paper chromatography. Only one Ehrlich-positive spot was detected on the paper chromatogram containing material derived from the incubation, and no Ehrlich-positive spots were detected in material derived from unincubated samples. The RF of the pyrrole derived from the incubation product was 0.50, and this value was identical to the RF of authentic l-methyl-2-carboxyethyl-3-propionic acid pyrrole derived from standard 6-aminolevulinic acid. Only one radioactive spot was detected in the incubated sample, and its position coincided with that of l-methyl-2-carboxyethyl-3propionic acid pyrrole (Fig. 6). Total radioactivity was determined in the region of an unstained paper chromatogram 4-6 cm from the origin, corresponding to the migration position of l-methyl-2-carboxyethyl-3-propionic acid pyrrole. By reference to the amount of extracted 6-aminolevulinic acid applied to the paper chromatogram, the specific radioactivity of the 6-aminolevulinic acid was calculated to be 16 mCi mmol".

6-Aminoleuulinic Acid Formation in Extracts of Oxygenic Prokaryotes
Inhibitors-Addition of increasing amounts of hemin to the incubation mixtures resulted in progressive inhibition of 6aminolevulinic acid formation (Fig. 7). Fifty percent inhibition occurred at approximately 5 ~L M hemin. a-Ketoglutarate (1 mM) inhibited 6-aminolevulinic acid formation by 20% when added to standard incubation mixture containing glutamate. Stimulation of 6-Aminolevulinic Acid Formation by Affinitypurified Chlorella Enzyme Fractions-Chlorella cell extract was previously fractionated by serial affinity chromatography into three enzyme components, all of which were required for in vitro 6-aminolevulinic acid formation, in addition to substrate, RNA, and other cofactors (21). The three Chlorella enzyme fractions were tested separately for the ability to stimulate 6-aminolevulinic acid formation when added to Synechocystis cell extract. One Chlorella fraction was highly stimulatory (Table IV). This fraction was the one which was Search for 6-Aminolevulinic Acid Synthase Activity in Synechocystis"Severa1 attempts were made to detect 6-aminole-  3 mCi mmol"). 6-Aminolevulinic acid was isolated from the incubation mixture, reacted to form 1-methyl-2-carboxyethyl-3-propionic acid pyrrole, extracted into diethyl ether, applied to Whatman No. 3MM paper, and chromatographed in 1-butanoE1-propanok5% (w/v) aqueous NH,OH (2:l:l). After solvent evaporation, the paper was cut into 1-cm strips centered on integral centimeter distances from the origin. The solvent front was at 10.5 cm from the origin, and the only Ehrlich-positive spot on the chromatogram was approximately 2-cm diameter and was centered at 5.2 cm from the origin. vulinic acid formation via 6-aminolevulinic acid synthasecatalyzed condensation of glycine and succinyl-CoA in extracts of Synechocystis. The low-speed supernatant after cell disruption was used to avoid possible loss of the enzyme activity. Extracts prepared from Euglena were used as positive controls (31). Incubations were carried out in two different 6aminolevulinic acid synthase incubation buffers, one (buffer A) corresponding to that used for measurement of &aminolevulinic acid synthase in Euglena extracts (31), and the other (buffer B) closer in composition to the buffer used for measuring 6-aminolevulinic acid formation from glutamate, but with glycine and succinyl-CoA added. 6-Aminolevulinic acid formation was not detected in Synechocystis extracts with either assay condition, whereas under the same conditions 6aminolevulinic acid formation (2.4-2.7 nmol of d-aminolevulinic acid) via the synthase reaction was readily detected in the Euglena extracts. The possibility was examined that the presence of 6-aminolevulinic acid synthase in the Synechocys-  tis extract was masked by an inhibitor. Enzyme fractions from Euglena and Synechocystis were mixed prior to incubation. The observed 6-aminolevulinic acid formation (2.6 nmol) could be accounted for entirely by the 6-aminolevulinic acid synthase activity in the Euglena extract, without any detectable stimulation or inhibition caused by the presence of the Synechocystis extract.

6-Aminoleuulinic Acid Formation in Extracts of Other
Oxygenic Prokaryotes-Cell extracts of Synechococczu sp. PCC 7002 and P. hollandica were also tested for 6-aminolevulinic acid-forming activity via the 5-carbon pathway. In the case of Synechococcus, the activity could be detected in the highspeed supernatant and was stimulated by adding RNA isolated from Synechocystis (Table V). 6-Aminolevulinic acidforming activity in Prochlorothrix extracts was in the lowspeed supernatant after disruption of the cells, but was located in both the pellet and the supernatant after high-speed centrifugation.

DISCUSSION
Cell-free extracts of three oxygenic prokaryotic species catalyze transformation of glutamate to 6-aminolevulinic acid.
The activity in extracts of all three species was sufficiently high to allow measurement of the 6-aminolevulinic acid formed in the reaction by spectrophotometric means. &Aminolevulinic acid was identified as the incubation product by comparison of the absorption spectrum after reaction with ethylacetoacetate and Ehrlich reagent to that of authentic 6aminolevulinic acid reaction product. Additional evidence was identical chromatographic behavior of the ethylacetoacetate reaction product with that of authentic 1-methyl-2-carboxyethyl-3-propionic acid pyrrole (see below).
6-Aminolevulinic acid-forming activity was further characterized in extracts of the unicellular cyanobacterium Synechocystis sp. PCC 6803. In most respects, the 6-aminolevulinic acid biosynthetic activity in Synechocystis extracts is very similar to those reported from eukaryotic cell or plastid extracts (13)(14)(15)(16)(17)(18). Activity required ATP, reduced pyridine nucleotide, M P , and a small tRNA-like molecule. The addition of an amino donor, other than glutamate, to the reaction mixture was not required. Heat-denatured extract was inactive.
The requirements for ATP and M P are consistent with the proposed mechanism of glutamate activation by aminoacylation of a tRNA (21,22,38,39). As with the eukaryotic cell extracts thus far examined, Synechocystis extracts had greater 6-aminolevulinic acid-forming activity with NADPH as the pyridine nucleotide cofactor than with NADH. The optimal concentration of NADPH was approximately 0.5 mM, and higher concentrations inhibited 6-aminolevulinic acid formation. With NADH, inhibition at higher concentrations was not observed. The physiological reason for the preference of the Synechocystis extracts for NADPH is unknown, but the preference of the plastid-derived extracts for NADPH is consistent with the generally greater availability of this reduced pyridine nucleotide, compared to NADH, in plastids (40).
Absolute dependence of 6-aminolevulinic acid formation on added glutamate could not be achieved in Synechocystis extracts. Cell extracts were passed through a Sephadex G-25 column to remove low molecular weight components before 6-aminolevulinic acid-forming activity was assayed. This procedure would be expected to deplete the extracts of endogenous glutamate and cause the reaction to become dependent on the addition of exogenous glutamate. However, the Synechocystis extracts still formed 6-aminolevulinic acid without added glutamate, and addition of glutamate stimulated the activity only by about 25%. The possibility was considered that the true carbon substrate was not glutamate, but aketoglutarate, which might be generated during the incubation. However, a-ketoglutarate inhibited 6-aminolevulinic acid formation by about 20%. Another possible explanation for the lack of glutamate dependence is that substrate quantities of glutamyl-tRNA might be present in the cell extract and co-elute with the protein fraction during gel filtration. This possibility was tested by digesting the cell extract with RNase and then gel filtering to remove the RNA fragments. After this treatment, 6-aminolevulinic acid-forming activity of the final extract was absolutely dependent on added RNA. The added RNA had been pretreated to ensure complete deacylation. Nevertheless, the dependence on glutamate was still not complete, and 6-aminolevulinic acid formation was stimulated only two fold by added glutamate.
One possible cause for the independence on added glutamate is that small, but sufficient, amounts of glutamate might be generated during the incubation by the action of proteolytic enzymes. Dependence on added glutamate was not increased by addition of the protease inhibitor phenylmethylsulfonyl 6-Aminolevulinic Acid Formation in Extracts of Oxygenic Prokaryotes fluoride. However, the responsible protease might not be inhibited by this reagent. Another possibility is that some glutamate might remain tightly bound to proteins through the gel filtration steps, and be released only during the incubation.
Although absolute dependence on added glutamate could not be achieved, glutamate was determined to be the carbon substrate for 6-aminolevulinic acid formation by measuring specific transfer of label from 1-["Clglutamate to 6-aminole-.
vulinic acid during the incubation. Glutamate labeled at C1 can transfer label to 6-aminolevulinic acid only via the 5carbon pathway, and this carbon atom is lost upon indirect incorporation of glutamate into 6-aminolevulinic acid via the 6-aminolevulinic acid synthase-catalyzed reaction (4,41). Therefore, the high degree of label transfer from 1-["C] glutamate to 6-aminolevulinic acid measured in the Synechocystis cell extract is interpreted to indicate a direct precursor role. Co-migration of the only radioactive spot with the 1methyl-2-carboxyethyl-3-propionic acid pyrrole upon paper chromatography indicates the high radiopurity of the isolated l-methyl-2-carboxyethyl-3-propionic acid pyrrole that was quantitated for specific radioactivity determination. The decrease in the specific radioactivity of the 6-aminolevulinic acid product to about one-third of the specific radioactivity of the glutamate added to the incubation mixture is consistent with the existence, or generation, of endogenous glutamate in the incubation mixture. The calculated %fold dilution of glutamate specific radioactivity is also consistent with the measured 50-100% stimulation of 6-aminolevulinic acid formation by added glutamate, with the unstimulated rate attributed to endogenous glutamate.
Because of earlier doubts raised by the apparent glutamate independence, an attempt was made to detect 6-aminolevulinic acid formation from glycine and succinyl-CoA in the Synechocystis extracts. As a positive control, 6-aminolevulinic acid synthase activity was measured in Euglena extracts, which were previously demonstrated to contain the enzyme (31). No 6-aminolevulinic acid synthase activity was detected in the Synechocystis extracts. Moreover, when Euglena extract was added to Synechocystis extract, all of the measured 6aminolevulinic acid synthase activity could be attributed to the added Euglena extract, with neither inhibition nor stimulation caused by the Synechocystis extract.
The 6-aminolevulinic acid formation in the Synechocystis extract is dependent on a small RNA. Activity with endogenous RNA was destroyed by RNase digestion of the extract prior to incubation. The extract was largely reactivated by addition of a fraction of Synechocystis RNA that was isolated by phenol-chloroform extraction and DEAE-cellulose chromatography. Addition of this RNA fraction to extracts that were not RNase-predigested stimulated 6-aminolevulinic acidforming activity. The degree of stimulation was proportional to the amount of added RNA. RNA isolated from Chlorella by the same procedure also stimulated 6-aminolevulinic acid formation, but not as effectively as Synechocystis RNA. Synechocystis RNA was previously found to stimulate &aminolevulinic acid formation in extracts of Euglena (17) and Chlorella (23). The cross-species acceptability of the RNA fractions from Synechocystis, Euglena, and Chlorella indicates that the RNAs are similar to each other, but different in some way from glutamyl-tRNAs from other species, which did not support 6-aminolevulinic acid formation in these cell extracts, even though these RNAs could be acylated with glutamate (17,42).
Hemin inhibited 6-aminolevulinic acid formation in Synechocystis extracts at micromolar concentrations. The sensitivity to hemin was somewhat lower than in Chlorella (16) or Chlamydomonas (43) extracts, but about the same as reported for barley plastid extracts (441, and suggests that heme may be an important physiological regulator of 6-aminolevulinic acid formation, acting by feedback inhibition at the enzyme level, in Synechocystis cells. 6-Aminolevulinic acid formation in Synechocystis extracts was highly stimulated by the addition of one of three affinitypurified enzyme fractions of the 6-aminolevulinic acid-forming system isolated from Chlorella (21). Preliminary results suggest that this fraction is also rate limiting in Chlorella extracts.* Based on the known affinity properties of the resins used in the separation, it was concluded that this component corresponds to the dehydrogenase enzyme (21). The stimulation caused by the Chlorella protein fraction was more than additive, since the Chlorella fractions, either singly or in any combination of two, formed only small amounts of &aminolevulinic acid in the absence of Synechocystis protein (21). The ability of a ChloreUa enzyme fraction to stimulate 6aminolevulinic acid formation in Synechocystis extracts suggests that the reaction intermediates are the same in both species, and that species-specific interaction of enzyme components is not required for 6-aminolevulinic acid formation.
The procedure for detection of 6-aminolevulinic acid formation in Synechocystis extracts was also used to measure 6aminolevulinic acid formation in extracts of another cyanobacterial species, Synechococcus sp. PCC 7002. Synechococcus extracts had somewhat lower activity than the Synechocystis extracts, and activity was stimulated by the addition of Synechocystis RNA. The lower activity of Synechococcus can be attributed to the facts that neither the growth conditions nor the extraction and incubation conditions were optimized for this species.
6-Aminolevulinic acid-forming activity was also detected in extracts of the filamentous prochlorophyte P. hollandica.
Prochlorothrix is the first member of the Prochlorophyta to be cultured in the laboratory (301, and the first in which the route of 6-aminolevulinic acid formation has been determined. We previously found that RNA isolated from Prochlorothrix supports 6-aminolevulinic acid formation in Euglena enzyme extracts (17). The observation that Prochlorothrix forms 6aminolevulinic acid via the 5-carbon pathway is consistent with the proposed candidacy of the chlorophyll a-and bcontaining Prochlorophyta as precursors to green algal and plant chloroplasts. Substantial progress has recently been made in several laboratories toward understanding the biochemistry of 6-aminolevulinic acid formation and its pivotal role in the regulation of tetrapyrrole biosynthesis. Progress with cell-free systems to date has been made almost exclusively with material derived from plants and eukaryotic algae. Oxygenic prokaryotes are especially attractive experimental materials for the next logical step in the investigation of mechanism of 6aminolevulinic acid formation and its regulation, isolation of the genes coding for the macromolecular components of the 6-aminolevulinic acid-forming system. The detection and partial characterization of 6-aminolevulinic acid formation in extracts of two genetically transformable cyanobacteria that have been used for molecular genetic studies (45, 46) is an important first step in the preparation for genetic studies.