natW' ' n“ - o 4 o E, I E; r A Michigan Scene University ., ILI!lllllllflzllflllllllflllflljllflllllllljlljfllfll This is to certify that the thesis entitled EFFECTS OF IRON-DEFICIENCY ON HEME BIOSYNTHESIS IN RHIZOBIUM JAPONICUM presented by Paul G. Roessler has been accepted towards fulfillment of the requirements for M. S . degree in Botany Major professor 93W 0-7639 OVERDUE FINES: 25¢ per day per item RETURNIMS LIBRARY MATERIALS: Place in book return to remove charge from circulation records EFFECTS OF IRON-DEFICIENCY ON HEME BIOSYNTHESIS IN R’IZOBIUM JAPONICUM By Paul G. Roessler A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1980 Cflfl/Jfivg ABSTRACT EFFECTS OF IRON-DEFICIENCY ON HEME BIOSYNTHESIS IN REIIZOBIUM JAPONICUM By Paul G. Roessler The heme moiety of leghemoglobin, a protein required for nitrogen fixation in the legume root nodule, is produced by Rhizobium bacteroids. This investigation involves a study on the role of iron in the control of heme biosynthesis in Rhizobium japonicum. Several effects of iron-deficient growth of B. japonicum were noted. Iron-deficient cells had increased rates of iron uptake and decreased cytochrome levels as compared to iron-replete cells. Protoporphyrin (PROTO) excretion occurred in iron-limited cultures, but not in iron-sufficient cultures. The activities of the heme biosynthetic enzymes G-aminolevulinic acid synthase (ALAS) and d-aminolevulinic acid dehydrase (ALAD) were decreased as cells entered an iron-limited stage of growth, but succinic thiokinase Paul G. Roessler activity did not decrease. The addition of 1 mM FeCl3. FeSOA, or certain iron chelators to assay mixtures were without effect on ALAS and ALAD activities, thus suggesting that iron does not directly affect these enzymes. PROTO and hemin (200 uM) also did not have a direct effect on ALAS or ALAD activity. PROTO may feedback regulate the activities of ALAS and ALAD since the addition of 10 mM levulinic acid prevents PROTO accumulation and also blocks the loss of ALAS and ALAD activities in iron-deficient cultures. In addition, growth of cells in iron-sufficient conditions in the presence of PROTO and dimethyl sulfoxide results in decreased ALAS activity. These results suggest a mechanism by which heme synthesis may be controlled in the legume root nodule. to my parents ii ACKNOWLEDGMENTS I would like to thank the members of my guidance committee, Drs. Philip Filner and Harold Sadoff, for helpful discussions and suggestions concerning my research. Their suggestions dealing with the preparation of this thesis are also appreciated. I would also like to thank Drs. Norman Good and Clifford Pollard for their help in the writing of this thesis. The help of Dr. Kenneth Poff with regard to spectro- photometric determinations of cytochromes and of Dr. Lee Jacobs concerning atomic absorption spectroscopy is also greatly appreciated. The companionship and help of all my friends here at Michigan State have made my stay here an enjoyable one, and I thank them all from the bottom of my heart. Finally, I would especially like to thank my major professor, Dr. Kenneth Nadler, for all of his help through- out my graduate career. I thank him not only for his suggestions and ideas concerning my research, but also for the constant encouragement and inspiration that he gave to me. He has made my stay here most rewarding, and I thank him for it. iii TABLE OF CONTENTS LIST OF FIGURES LIST OF TABLES. LIST OF ABBREVIATIONS INTRODUCTION. Structure, Function, and Biosynthesis of Leghemoglobin. . The Biosynthesis of Heme Effects of Iron on Enzymatic Activity. MATERIALS AND METHODS Media. Growth Conditions. . Measurement of Iron Uptake Cytochrome Difference Spectra. Measurement of Culture Fluorescence. Porphyrin Absorption Spectra . . Porphyrin Methyl Ester Chromatography. Preparation of Cell- Free Extracts. Enzyme Assays. . Measurement of Heme .Biosynthetic Rates Solubilization of Protoporphyrin and Hemin : Protein Determination. RESULTS Characteristics of Iron-Limited Growth . Effects of Iron- Limitation on Heme Biosynthesis Possible Mechanisms of Iron Effects. DISCUSSION. BIBLIOGRAPHY. iv . vi . viii 36 36 48 61 77 89 LIST OF FIGURES Figure Page 1 The biosynthesis of heme . . . . . . . . . . 6 2 Uptake of iron by R. japonicum . . . . . . . 38 3 Reduced minus oxidized difference spectra of cell-free extracts of B. japonicum . . . 4O 4 Reduced minus oxidized difference spectra of B. japonicum whole cells . . . . . . . . 43 5 Fluorescence of B. japonicum cultures as a function of culture age . . . . . . . . . 46 6 Acid spectra of protoporphyrin IX and the porphyrin excreted by R. japonicum grown in LI medium . . . . . . . . . . . . . . . . 49 7 Neutral spectra of the methyl esters of protoporphyrin IX and the porphyrin excreted by B. japonicum grown in LI medium . . . . . 51 8 ALAS activity in B. japonicum as a function of culture age . . . . . . . . . . . . . . . 54 9 Half-life of ALAS in cell-free extracts of R. japonicum . . . . . . . . . . . . . . . . 65 10 Apparent feedback inhibition of ALAD by protoporphyrin IX . . . . . . . . . . . . . 70 11 Effect of dimethyl sulfoxide on the growth of R. japonicum . . . . . . . . . . . . . . 74 LIST OF TABLES Table Page 1 Chromatographic properties of the methyl ester of the porphyrin excreted by R. japonicum grown in LI medium . . . . . . . 53 2 ALAS activity in cell-free extracts of R. japonicum grown in the presence of iron and other metal salts . . . . . . . . . . . . 57 3 Activities of several enzymes of heme biosynthesis in cell-free extracts of R. japonicum grown in HI medium and LI medium 58 4 Rates of incorporation of 2-14C-glycine into heme in R. japonicum grown in HI medium and LI medium . . . . . . . . . . . . . . . . . . 6O 5 Effects of ferric and ferrous iron on ALAS and ALAD activities in cell-free extracts of R. japonicum grown in LI medium . . . . . . . 62 6 Effects of iron chelating agents on ALAS and ALAD activities in cell-free extracts of R. japonicum grown in HI medium . . . . . . . 63 7 ALAS and ALAD activities in cell-free extracts of R. japonicum containing hemin and proto- porphyrin IX . . . . . . . . . . . . . . . . . 67 8 ALAS and ALAD activities in cell-free extracts of R. japonicum grown in the presence of hemin and protoporphyrin IX . . . . . . . . . 68 vi Table Page 9 Effect of levulinic acid on ALAS and ALAD activities in cell-free extracts of R. japonicum grown in LI medium . . . . . . . 72 lO ALAS activity in cell-free extracts of R. japonicum grown in HI medium in the presence of dimethyl sulfoxide and certain tetrapyrroles . . . . . . . . . . . . . . . . 75 vii ALA ALAD ALAS COPRO'gen DMSO HI LI PBG PROTO STK TCA URO'gen LIST OF ABBREVIATIONS é-aminolevulinic acid S-aminolevulinic acid dehydrase é-aminolevulinic acid synthase coproporphyrinogen dimethyl sulfoxide high iron low iron porphobilinogen protoporphyrin succinic thiokinase trichloroacetic acid uroporphyrinogen yeast extract-mannitol viii INTRODUCTION Structure, Function, and Biosynthesis of Leghemoglobin Leghemoglobins, monomeric hemoproteins having molecu- lar weights of 15,000 to 17,000 daltons, are found within the root nodules of many plants in the legume family when these plants are infected with bacteria of the genus Rhizobium. .They are among the most abundant proteins in the root nodule, constituting 20 - 40% of the total solu- ble protein (1). Most legumes have several chromato- graphically distinguishable leghemoglobins in their nod- ules. An example of this phenomenon is found in soybean, where five different leghemoglobins have been isolated. These appear structurally similar (as determined by circu- lar dichroism spectral studies) but differ slightly in antigenic properties and in amino acid sequence (2). All leghemoglobin molecules are similar enough to one another, however, that the general term "leghemoglobin" is used to refer to the entire group of such hemoproteins found in legume root nodules. Leghemoglobin has been shown to resemble vertebrate myoglobin in molecular weight, in binding characteristics with ligands such as 02 and carbon monoxide (3), in tertiary structure (4,5), and even in amino acid sequence (6). The prosthetic group common to all leghemoglobins is protoheme IX. Electron paramagnetic and optical spectral studies indicate that the fifth ligand-binding position of the coordinated iron atom of the heme molecule is occupied by the imidazole ring of a histidine residue in the protein, while the sixth ligand-binding position remains open for binding of free molecules, including molecular oxygen (7). It is now a widely accepted hypo- thesis that this 02-binding property of the hemoprotein molecule provides leghemoglobin with its functional role. It has been known for a long time that leghemoglobin is always associated with nitrogen fixation within the legume nodule (8). This observation led early investi- gators to believe that leghemoglobin had a direct role in nitrogen fixation as an electron transferring agent, via oxidation state changes in the iron of the heme molecule. Subsequent investigations showed, however, that isolated bacteroids were able to reduce molecular nitrogen in the absence of leghemoglobin (9,10,11) and that root nodules of non-legumes (including a non-legume nodulated by Rhizobium) do not contain leghemoglobin, and yet still maintain rates of nitrogen fixation comparable to those of legumes (12). These studies suggest that leghemoglobin does not have a direct role in nitrogen fixation, but rather some type of supportive role. Nitrogenase, the enzyme that catalyzes the reduction“ of molecular nitrogen to ammonia in Rhizobium, is irrever- sibly inhibited by molecular oxygen (13). At the same time, however, this enzyme requires large amounts of energy in the form of ATP before metabolically produced reducing equivalents can convert molecular nitrogen to ammonia. The large amounts of ATP required are generated by respiration utilizing 02 as a terminal electron accep- tor. Thus, oxygen has a beneficial effect and a deleter- ious effect on nitrogen fixation in the root nodule. It is this situation that led to the development of the hypothesis that leghemoglobin contributes to nitrogen fixation by acting as an oxygen "buffer”. It can provide a large resevoir of readily available oxygen for the rapidly respiring Rhizobium cells while at the same time maintaining the oxygen tension in the nodule at such a low level that nitrogenase is not inhibited. This hypo- thesis has gained considerable experimental support. It has been known for a long time that leghemoglobin has a great affinity for O2 (14). Leghemoglobin is half-satur- ated with 02 at about 0.05 mm mercury (15,16), a value much lower than that of other hemoglobins. The O2 combi- nation rate constant for leghemoglobin and the 02 dis- sociation rate constant for oxyleghemoglobin, as measured by Wittenberg et al. (3 ), are particularly well suited for facilitated oxygen diffusion at a very low mean oxygen pressure. The biosynthesis of leghemoglobin occurs only as a result of the symbiosis between the legume host plant and the rhizobia. The heme molecule is produced by the sym- biotic bacteria (bacteroids) while the apoprotein is produced by the host plant. It was shown that the heme moiety is produced by the bacteroids in radioisotope tracer studies (17,18) and by the demostration that the activity of certain enzymes of the heme biosynthetic route increases in soybean bacteroids as nodule leghemoglobin content increases, whereas this increase in heme biosyn- thetic ability is not seen in the host plant (19). It is also clear that the apoprotein is produced by the host plant since the electrophoretic mobilities of the various leghemoglobins are determined by the strain of plant used rather than by the strain of Rhizobium infecting the roots (20,21). It has also been shown by in vitro protein synthesis experiments that the apoprotein is encoded by a 93 poly(A)-containing mRNA that is associated with 803- type plant ribosomes in vivo (22). In addition, it has been recently shown that the genes for leghemoglobin are contained within the host plant DNA by hybridization studies involving complimentary DNA prepared from the above-mentioned mRNA (23). The fact that certain parts of the leghemoglobin molecule are made separately by each of the two symbionts presents the problem of the regulation of heme and globin synthesis in such a way that neither of the moieties are overproduced. The absence of such regulation would be an extremely wasteful situation which could lead to deleter- ious effects on the symbiosis. It is the purpose of this project to examine possible modes of regulation in the biosynthesis of the heme moiety of leghemoglobin. The Biosynthesis of Heme The biosynthesis of heme in animals and microorgan- isms occurs via the pathway shown in Figure l. The first step in this pathway is the reaction of succinyl-CoA with glycine to yield é-aminolevulinic acid (ALA). The glycine is activated by pyridoxal phosphate in a way that a stable carbanion is produced from the a-carbon. The electro- phylic acyl-carbon atom of succinyl-CoA then condenses_ with the glycine to form a-amino-B-ketoadipic acid, Which immediately decarboxylates to form ALA. The reaction is catalyzed by G-aminolevulinic acid synthase (ALAS). The second step in the reaction sequence is the condensation of two molecules of ALA to form the dicarboxylic mono- pyrrole porphobilinogen (PBG). This reaction is catalyzed by G-aminolevulinic acid dehydrase (ALAD). The first Figure 1. The biosynthesis of heme. Abbreviations: ALA PBG URO' G-aminolevulinic acid porphobilinogen gen = urOporphyrinogen COPRO'gen = coproporphyrinogen PROTO = protoporphyrin M = A p = V = Pyr- “CH3 -CH3COOH -CH CH COOH 3 3 P04 = pyridoxal phosphate 5523 a ._ _ ~x ole I v :3»; . I: < Nxc Al 6.. .- + «a «37:53.; tetrapyrrole intermediate in heme biosynthesis, uropor- phyrinogen III, results from the condensation of four molecules of PBG. This step is catalyzed by two enzymes, PBG deaminase and uroporphyrinogen cosynthetase. Four molecules of ammonia are eliminated in this reaction. Coproporphyrinogen III, the next intermediate, arises from the decarboxylation of the acetyl side chains of uropor- phyrinogen III, catalyzed by uroporphyrinogen III decar- boxylase. Subsequent oxidative decarboxylation of two of the propionyl side chains of coproporphyrinogen III yields protoporphyrin IX. This step is catalyzed by copropor- phyrinogen oxidase. The final step in heme biosynthesis involves the insertion of ferrous iron into protopor- phyrin IX, a step catalyzed by ferrochelatase. The regulation of heme biosynthesis has been exten- sively studied in both animals and microorganisms. Feed— back inhibition and repression of enzyme synthesis have both been shown to be control mechanisms in heme biosyn- thesis. Work by Lascelles on Rhodopseudomonas spheroides showed that heme (10 uM) was able to repress the formation of ALAS while other porphyrins were without effect (24). Work in the same laboratory revealed that partially puri- fied ALAS was extremely sensitive to feedback inhibition by heme. The enzyme was 50% inhibited by 1 uM heme. Other metalloporphyrins and metal-free porphyrins were also inhibitory, but concentrations required for 50% inhibition were ten times higher (25). ALAS in SpiriZZum itersonii (26) and Micrococcus denitrificans (27) is also inhibited by low concentrations of heme. Heme also appears to play a role in the regulation of its own synthesis in animal systems. Granick (28) demonstrated in studies on embryonic chick liver cultures that the rate of formation of ALAS in this system appears to be con- trolled by feedback repression by heme. Indirect evidence has also been provided for a direct inhibition of ALAS by heme in rabbit reticulocyte preparations (29). Recent evidence suggests, however, that mammalian hepatic ALAS is not controlled by direct feedback inhibition by heme (30). The second enzyme of the heme biosynthetic pathway, ALAD, appears to be a secondary control point in several organisms. In R. spheroides, feedback inhibition of ALAD by heme has been observed, with half-maximal inhibition occurring at a concentration of 40 uM heme (25). Inhibi- tion of ALAD by heme has also been noted in mouse liver (32) and in human erythrocytes (33). The synthesis of ALAD is repressed in R. spheroides by the same concentration of heme required to repress the synthesis of ALAS (24). Repression of ALAS and ALAD formation by ALA has also been reported for this organism (24). This suggests that coordinate repression occurs as in many other biosynthetic pathways subject to control by 10 repression. Indeed, a linear arrangement of the genes reponsible for heme biosynthesis has been shown to occur on the chromosome of Staphylococcus aureus (34). These genes are tightly linked and co-transducible. However, in Escherichia chi it appears that the genes coding for the enzymes of heme biosynthesis are found in different regions of the chromosome (35). It has been demonstrated that the synthesis of ALA in R. spheroidcs is rate-limiting in heme biosynthesis, thus raising the question of why there would exist a secondary control point at the step catalyzed by ALAD. One possible explanation involves a postulated succinate-glycine cycle for the production of ALA as proposed by Shemin (36). In this proposed cycle, ALA formed via ALAS would lose NH3 via transamination, yielding y-ketoglutaraldehyde, which would in turn lose the terminal aldehyde group to yield succinate and a ”Cl" compound. It can be seen why a secondary control point at ALAD would be desirable if this cycle is indeed in operation. It should be noted at this point that ALAD from Neurospora crassa appears to be rate-limiting in heme biosynthesis. The enzyme can be induced by iron and inhibited by coproporphyrinogen and by protoporphyrin (37). There is also evidence that ALAD from yeast may be the rate-limiting enzyme in heme biosynthesis in this organism (38,39). Several factors have been shown to 11 control the activity of this enzyme in this organism. There have been several reports of iron having an effect on heme biosynthesis in various organisms. These reports will be discussed in the following section. Effects of Iron on Enzymatic Activity Iron has been shown to be a controlling factor in the activity of many enzymes. Iron can affect enzymatic activity by altering the catalytic efficiency of the enzyme directly or indirectly, or by causing a change in the amount of the enzyme via induction or repression. The "active sites” of many enzymes contain iron. A widespread example of this is seen in the various non- heme iron-sulfur proteins. NADH dehydrogenase, succinate dehydrogenase, and the ferredoxins are common examples of enzymes containing iron-sulfur centers. These enzymes are typically involved in oxidoreduction reactions, a direct result of the ability of the iron molecule to undergo oxidation state changes. They have been shown to have decreased activities in various organisms when grown under iron-deficient conditions (40,41). Nitrogenase and hydrogenase, enzymes that are both found in Rhizobium, also contain iron within the polypeptide structure. Both of these enzymes are involved in electron transfer reac- tions. Some enzymes not involved in oxidoreduction 12 reactions have also been shown to require iron for acti- vity. One such enzyme containing iron is 3-deoxy-D-ara- bino-heptulosonate-7-phosphate synthase, the enzyme catalyzing the first committed step in the biosynthesis of aromatic compounds in bacteria and plants (42). Aconitase, an enzyme of the tricarboxylic acid cycle, also requires ferrous iron for activity. In addition to such enzymes containing iron directly within the polypeptide structure, several enzymes owe their catalytic abilities to the presence of the iron- containing prosthetic group, heme. Peroxidase, catalase, and the cytochromes are common examples of heme-containing proteins. Decreased peroxidase and catalase activities have been reported for several plants, fungi, and bacteria grown under iron-deficient conditions (43,44,45). Decreased cytochrome content is also commonly observed during iron-limited growth (46,41), but whether these findings are due simply to the lack of heme molecules, or to a concomitant decrease in apoprotein levels is not known in all cases. In addition to those enzymes already mentioned which require iron for catalytic activity, there are several enzymes whose catalytic activities are enhanced by, but not absolutely dependent upon the presence of iron. Several enzymes which contain pyridoxal phosphate as a cofactor are in this category. 13 The general mechanism of pyridoxal phosphate cata- lyzed reactions first involves Schiff base formation between an a-amino acid and the aldehyde group of the pyridoxal phosphate molecule. The electrophylic nature of the heterocyclic nitrogen atom in the pyridine ring then causes a displacement of electrons from one of the bonds surrounding the a-carbon of the amino acid. From this point, transamination, decarboxylation, racemization, B-elimination, or certain condensations can occur. Metzler ct aZ. (47) were able to show that various metal ions, including iron, enhance numerous non-enzymatic reactions involving amino acids which are catalyzed by pyridoxal. The metal ion appears to enhance the catalytic activity of pyridoxal (and pyridoxal phosphate) by three modes of action (48); (1) promotion of Schiff base form- ation between the amino acid and the aldehyde group of the pyridoxal molecule, (2) stabilization of the conjugated intermediate in a planar configuration essential for the postulated electron shifts, and (3) provision of an additional electrophylic group that intensifies displace- ment of the electrons from the bonds surrounding the d-carbon atom of the amino acid. Although the above reports deal with non-enzymatic reactions, several other reports suggest that a similar phenomenon occurs in pyridoxal phosphate-containing enzymes. Patwardhan (49) demonstrated that the activity 14 of aspartic - glutamic transaminase in Dolichcs Zablab was increased when ferrous iron was included in the reac- tion mixture. Likewise, an aminobenzoic acid decarbox- ylase from E.chi (50) and a histidine decarboxylase (51) have been shown to require ferric iron for activity. All of these enzymes require pyridoxal phosphate as a cofactor, as does ALAS. Besides having a direct effect on enzymatic activity, iron appears to be able to alter the level of several enzymes by repression, derepression, or induction. A well characterized example of this is the derepression of the synthesis of certain iron chelators which are produced and excreted by many microorganisms during iron-limited growth. Receptor sites for the chelators are often con- comitantly developed. The chelation of iron by these compounds and the subsequent uptake of the complex allows the organism to continue growth despite the low amount of iron present. A specific example of the phenomenon can be found in E. coli. This organism will excrete entero- chelin (tricyclic 2,3-dihydroxy-N-benzoyl-L-serine) into the medium when grown under iron-deficient conditions. There are eight enzymes responsible for the production, transport, and breakdown of enterochelin, and the genes corresponding to these proteins are all located around minute 14 of the E. coli chromosome. The synthesis of all but one of the gene products has been shown to be 15 repressed by iron, suggesting that the genes are clustered in one or more operons (52,53,54,55). Repression might occur by the combination of iron (or an iron-containing molecule) with an aporepressor protein to yield a functional repressor. The iron-containing heme molecule also appears to be able to control the rate of protein synthesis in several organisms. Some studies suggest that the formation of the protein moieties of the cytochromes requires the pres- ence of heme. Lascelles et al. (56) showed that in a heme-requiring mutant of SpiriZZum itersonii, reconsti- tution of the normal cytochrome spectrum did not occur upon the addition of exogenous hemin to the cell extracts. However, the addition of hemin to whole cells resulted in cytochrome formation which could be blocked by protein synthesis inhibitors. In another study on S. itersonii, Garrard (57) demonstrated that the synthesis of the apo- protein of cytochrome c stops when the endogenous heme pool is depleted. There is also substantial evidence suggesting that globin synthesis in mammalian reticulo- cytes is controlled by heme (58,59). It appears that heme inhibits the activity of a cyclic AMP-dependent protein kinase which is responsible for the phosphorylation (activation) of a cyclic AMP-independent protein kinase. The active cyclic AMP-independent protein kinase is then able to phosphorylate (inactivate) an initiating factor 16 for protein synthesis. It can be seen that in this cascade of reactions, a small amount of heme allows protein synthesis to occur at maximal rates. There is also evidence, however, which suggests that in some organisms the apoproteins of cytochromes and other hemoproteins are synthesized in the absence of heme. Heme-deficient Staphylococcus epidermidis cells exhibit a rapid increase in catalase activity and respiration upon the addition of exogenous heme, despite the presence of protein synthesis inhibitors (60). A similar situation has been reported for E. coli (61). There have been several reports of iron having an effect on heme biosynthesis in various organisms. These reports fall into two general categories; cases in which iron-deficiency leads to a decrease in heme biosynthesis merely due to the lack of iron for insertion into proto- porphyrin, and cases in which it has been demonstrated that the activities of certain enzymes of heme biosynthe- sis are altered under iron deficient conditions. One of the first reports of iron having an effect on the activity of an enzyme involved in heme biosynthesis was made by Brown in 1958. He noted that the addition of ferrous iron to chick erythrocyte preparations overcame the inhibition of ALAS by 5,6-dimethylbenzimidazole (62) and by a,a'-dipyridyl (an iron chelator) (63). It was known that 5,6-dimethylbenzimidazole inhibits the step l7 involving activation of glycine, and therefore Brown suggested that the catalytic activity of ALAS-bound pyridoxal phosphate was enhanced in the presence of iron. Vogel ct a2. (64) found that heme and porphyrin synthesis was decreased in iron-deficient duck blood preparations, and that the rates could be brought back to control levels by the addition of ferrous iron. It was determined by radioisotope tracer studies that the iron had its effect on ALA synthesis, but addition of ferrous iron to parti- culate fractions of the reticulocytes did not cause an increase in ALAS activity. Although there is no evidence to date which suggests that mammalian hepatic ALAS requires iron for activity, it has been reported that the induction of ALAS synthesis in rat liver by allylisopropylacetamide can be greatly“ enhanced by oral administration of ferric citrate (65).. Iron also appears to play a role in ALAS acti- vity in certain plants. There has been a report present— ing evidence that iron-deficiency results in decreased ALA production in cowpea leaves, although the remainder of the pathway up to protoporphyrin is unaffected (66). Iron has also been shown to affect the activity of heme biosynthetic enzymes in several microorganisms. ALAS activity is low in Micrococcus denitrificans grown under iron-deficient conditions (67). Addition of ferric citrate to cultures leads to an increase in the activity 18 of this enzyme, and this increase can be blocked by chloramphenicol. ALAD, the rate-limiting enzyme of heme biosynthesis in Neurospora crassa has also been shown to have decreased activity when this organism is grown under iron-deficient conditions (68). Activity can be maximally induced in thirty minutes by the addition of ferric citrate to the culture. This induction can be blocked at the level of transcription by 8-azaguanine and cordycepin. Many microorganisms excrete porphyrins into the growth medium when grown under iron-deficient conditions. In almost all cases, coproporphyrin , and not protopor- phyrin, has been shown to be the predominant porphyrin excreted (69). This suggests that iron may play a role in the conversion of coproporphyrinogen to protoporphyrin. Coproporphyrinogen oxidase activity in mammalian (70) and plant (71) mitochondria was decreased upon the addition of iron chelators, but no such effect could be observed in extracts of Chromatium (72), R. spheroides (73), E. coli (74), or Pseudomonas (74). It is of interest to note that iron chelators will decrease the rate of conversion of coproporphyrinogen to protoporphyrin in anaerobically- maintained R. spheroidcs extracts, however (73). Korstee (75) observed that iron enhanced the conversion of coproporphyrinogen to protoporphyrin in Arthrobacter, but determined that the iron increased the phosphatidyletha- nolamine content of the cells, and that this phospholipid 19 was actually responsible for the enhanced conversion. The metabolism of iron by Rhizobium and the effects that iron may have on various metabolic processes in this organism are especially important. The reduction of molecular nitrogen in the legume root nodule is dependent upon rhizobial nitrogenase, which is functional only when the oxygen concentration within the nodule is maintained at very low levels, a situation made possible by the pres- ence of leghemoglobin. In addition, nitrogenase requires large amounts of energy in the form of ATP, which is generated by cytochrome-mediated respiration. Nitrogenase, leghemoglobin, and the cytochromes are all iron-containing molecules, and this fact alone illustrates the importance of investigating iron metabolism in Rhizobium. As mentioned earlier, the biosynthesis of heme by Rhizobium is an important aspect of legume root nodule development because of the role of leghemoglobin in symbiotic nitrogen fixation. It would be beneficial to the symbiosis if rhizobial heme synthesis and/or apoprotein synthesis by the host plant was tightly regulated so that heme and apoprotein production rates would be approximately the same. Whether this situation indeed exists in the nodule is unknown, however. Since the bacteroids are dependent upon the host plant for all nutrients, control of delivery of a certain substrate by the host plant to the bacteroids could provide a convenient means of control 20 of rhizobial heme synthesis. Iron is undoubtedly one of the nutrients which must be transferred from the host plant to the bacteroids. In view of this fact, an investigation of the effects of iron on heme biosynthesis in Rhizobium japonicum was undertaken. MATERIALS AND METHODS Media Low-iron (LI) medium contained (in grams per liter of distilled, deionized water): mannitol (Sigma Chem. Co.), 2; NazHP04-7H20 (Fisher Scientific Co., Certified AC8), 1.3; NaH2P04°H20 (Fisher Scientific Co., Certified AC8), 0.1; Mg804o7H20, 0.05; NaCl, 0.5; (NH4)2804, l; Ca(N03)2-4H20, 0.02; biotin (Sigma Chem. Co.), 0.0025; and casamino acids (Difco), l. High-iron (HI) medium was obtained by adding 4 mg of ferric citrate (Baker Chem. Co.; 16% iron by assay) per liter of LI medium. All components were autoclaved together except for the biotin, casamino acids, and ferric citrate, which were added aseptically at the time of inoculation. Yeast extract-mannitol (YM) medium contained (in grams per liter of distilled water): mannitol (Sigma Chem. Co.), 10; yeast extract (Difco), l; KZHPO4 (Mallinckrodt), 0.5; MgSOA-7H20, 0.2; NaCl, 0.1; and CaSOA-ZHZO, 0.1. All components were autoclaved together. 21 22 The chemicals used for media preparation were chosen from readily available sources for having low amounts of iron contamination. In addition, all culture flasks were allowed to soak overnight in 6 N HCl in order to remove contaminating iron. The acid was removed by exhaustive rinsing with distilled, deionized water. The concentration of iron in LI medium was determined by flameless atomic absorption spectroscopy. A Varian model AA6 atomic absorption spectrophotometer with carbon rod attachment ORA—90 was used for these determinations. The sample was dried for 25 sec at 100°C, ashed for 15 sec at 800°C, and atomized at 23oo°c with a hold time of 1.5 sec. The ramp speed was set for 400°C per second. Known amounts of iron were added in increasing concentrations to vials containing LI medium. The absorption due to iron atoms was determined in these samples and in a sample to which no iron had been added. Extrapolation through these points to the baseline then allows for determination of iron in the medium. LI medium.was determined to contain 14 ng of iron per ml. This corresponds to a concentration of 0.25 umoles of iron per liter. HI medium contained 11.7 umoles of iron per liter. 23 Growth Conditions Rhizobium japonicum 3Ilb-110 was kindly supplied by D. C. Weber, USDA, Beltsville, MD. Bacterial cultures were maintained on YM-agar slants. Primary inoculum cul- tures were prepared every 10-15 days via inoculation from a slant into a flask containing 100 m1 of medium. Cultures for experimental assay were prepared by inoculation of 500 ml of medium in a one liter flask with 3-5 ml aliquots from primary inoculum cultures. Cultures were grown at 30°C on a reciprocal shaker. Cells were routinely har- vested at early stationary phase. The growth and purity of cultures was routinely checked by plating out serially- diluted cultures on YM-agar plates. Measurement of Iron Uptake 9Fe) was used to determine the Radioactive iron (5 rate of iron uptake from the medium by R. japonicum cells. Cells were harvested and then washed twice in ice-cold 0.1 M HEPES buffer (pH 7.0) and then resuspended in 50 ml of fresh buffer. 10 ml of this suspension was then placed in a 50 m1 Erlenmeyer flask and allowed to preincubate at 30°C for 30 min with shaking. 0.25 uCi of 59Fe (specific activity = 0.16 uCi per nmol iron) was then added to the cell suspension and a 0.5 ml sample was immediately 24 removed and filtered through a millipore filter (pore size = 0.45 pm). The cells remaining on the filter were then immediately washed with 7 ml of ice-cold 0.1 M HEPES buffer (pH 7.0) containing 10 mM EDTA. 0.5 m1 samples were subsequently taken at various time points and filtered in the same manner. The cell suspension was shaken at 30°C except for the brief periods during sam- pling. | The filters were then placed in 10 ml of scintil- lation cocktail (ACS, Amersham) and counted in a Packard Tri-Carb model 578 scintillation counter at an efficiency of 88%. The radioactivity of the filter from the initial time point was subtracted from the subsequent time points to account for non-specific adsorption of 59Fe. Non- specific adsorption accounted for less than 10% of the radioactivity that was detected after the final time point. Cytochrome Difference Spectra Cytochromes were detected in cell-free extracts and in whole cell suspensions by difference spectroscopy. Cell-free extracts were obtained by washing and resuspend- ing cells in 0.1 M sodium phosphate buffer (pH 7.0) and then sonicating for six 15 sec bursts at power setting 3 with a Branson model S-125 sonifier. Each burst was 25 followed by a 15 sec cooling period. The sonicates were then clarified by centrifugation at 30,000 x g for 15 min. The protein concentration of the clarified extracts were then adjusted to 3 mg/ml. One sample was reduced by the addition of a few grains of sodium dithionite, and another sample was oxidized by the addition of 50 ul of a 3 mM K3Fe(CN)6 solution. The reduced minus oxidized spectrum was then obtained with a Cary 15 recording spectrophoto- meter. Whole cell suspensions were obtained by washing and resuspending cells in 0.1 M phosphate buffer (pH 7.0) at a concentration of 15% (wet weight/volume). One sample was reduced by the addition of sodium dithionite and another sample was oxidized by passing oxygen gas through the suspension for 10 min. The samples were placed in cylindrical cuvettes which were subsequently placed in transparent Dewar vessels containing liquid nitrogen. The absorption spectrum was then obtained for each sample with a single-beam spectrophotometer (Cary 14, modified) with the data being stored in an on-line Hewlett—Packard 2108 mX minicomputer. The difference spectrum was then determined by the computer by subtraction of the absorp- tion spectrum of the oxidized sample from that of the reduced sample. The computer also provided the fourth derivative of the resulting difference spectrum. The fourth derivative analysis was used to detect minor peaks 26 and shoulders on major peaks. Measurement of Culture Fluorescence R. japonicum cells grown in LI medium excreted por- phyrins that were insoluble in the culture medium. These porphyrins adhered to the sides of the flask or formed insoluble aggregates which could be removed from the med- ium along with the cells by centrifugation at 11,000 x g for 20 min. Passing the supernatant through a 1 cm pad of neutral alumina (Fisher Scientific Co.; Brockman activity I, 80-200 mesh) equilibrated with 3% acetic acid did not reveal the presence of soluble porphyrins. The fluorescence of cultures due to the presence of porphyrins was measured by the use of a Turner fluoro- meter (model 111) equipped with a 405 nm excitation filter and a 23A secondary filter (wavelength cutoff = 570 nm). 5 ml samples were periodically removed from cultures and mixed with 1 m1 of concentrated HCl. The resulting solution was placed in the dark at 4°C until the fluoro- metric determination was to be made. Samples were centrifuged at 30,000 x g for 15 min immediately prior to fluorometric determinations. A 10 nM solution of protoporphyrin in HCl yields 25 relative fluorescence units under these conditions. 27 Porphyrin Absorption Spectra The acid spectrum of the excreted porphyrin was determined by repeatedly rinsing the sides of the culture flask with 5 ml of 1 N HCl. The resulting solution was clarified by centrifugation at 30,000 x g for 20 min. The absorption spectrum was then obtained directly with a Cary 15 recording spectrophotometer. The methyl ester of the excreted porphyrin was obtained by repeatedly rinsing the sides of the culture flask with a methanol/sulfuric acid solution (20:1, v/v). After allowing this solution to remain in the dark over- night, the porphyrin methyl ester was purified by the method of Doss and Ulshofer (76). First, the solution was neutralized by the addition of 5 volumes of a 5% NaHCO3 solution and the porphyrin methyl ester was immediately extracted into one volume of chloroform. The chloroform layer was washed five times with 0.5 volumes of a 7% NaCl solution and then the remaining water was removed from the organic layer by the addition of excess NaCl crystals. The neutral spectrum of the porphyrin ester could now be obtained with a Cary 15 recording spectrophotometer. Evaporation of most of the chloroform provided a sample suitable for thin layer chromatography of the porphyrin methyl ester. 28 Porphyrin Methyl Ester Chromatography The methyl ester of the porphyrin obtained as described above was analyzed by thin layer chromatography along with porphyrin methyl ester standards (obtained from Sigma Chem. Co.) in three different systems: 1. benzene-ethyl acetate-ethanol, l90:20:7.5 (v/v/v) on silica gel G plates (Brinkman) (76); 2. petroleum ether-paraffin oil-chloroform, 1 1:10 (v/v/v) on Eastman Chromagram sheets (77); 3. water-acetonitrile- p-dioxane, 2 7:1 (v/v/v) on Whatman KC18 reverse-phase plates. The solvent con- tained 0.5 M NaCl to stabilize the bonding of the hydrocarbon chains. The porphyrin esters were located by their fluorescence under ultraviolet light. Preparation of Cell-Free Extracts Cells were harvested by centrifugation in 250 ml bottles at 11,000 x g for 20 min. Cells were then washed in ice-cold sonication buffer and recentrifuged at 25,000 x g for 10 min. The washed cells were then resuspended in 5-10 ml of ice-cold sonication buffer and the chilled sample was sonicated for six 15 sec bursts at power setting 3 with a Branson model S-125 sonifier. Each burst was 29 followed by a 15 sec cooling period. The sonicates were then centrifuged at 30,000 x g for 10 min to obtain the cell-free extracts. The extracts were then diluted with sonication buffer so that the protein concentration was between 1.0 and 2.0 mg per ml. The sonication buffer for measurement of ALAS and succinic thiokinase activities contained (in mmoles per liter): HEPES buffer (pH 8.0), 100; MgClZ, 1; and 2-mer- captoethanol, 1. The sonication buffer for measurement of ALAD activity contained (in mmoles per liter): HEPES buffer (pH 7.5), 30; MgSOA, 10; and Z-mercaptoethanol, 2.5. Enzyme Assays §-aminolevulinic acid synthase (ALAS) ALA production in cell-free extracts was measured by a modification of the method of Mauzerall and Granick (78). The reaction mixture for assaying ALAS activity contained (in mmoles per liter): HEPES buffer (pH 8.0), 100; MgClz, 32; Naz-succinate, 200; glycine, 200; ATP, 14; Coenzyme A, 0.1; and pyridoxal phosphate, 0.6. The reaction was initi- ated by the addition of 0.5 m1 of the cell-free extract to 0.5 m1 of ALAS reaction mixture. The reaction was allowed to proceed for 2 hours in test tubes placed in a 30°C water bath. The reaction was stopped by placing the test tubes on ice and immediately adding 0.2 m1 of 33% trichloroacetic 3O acid (TCA). The tubes were then vortexed and allowed to stand on ice overnight. Precipitated protein was removed by centrifugation at 800 x g for 10 min. One ml of the supernatant from each tube was placed in another tube to which 0.25 ml of 0.75 m Na3PO4 and 0.1 ml of ethylaceto- acetate were added. The contents of the tubes were thor- oughly mixed and then placed on a boiling water bath for 15 min. After the tubes had cooled on ice, 1.35 ml of modified Ehrlich's reagent (79) was added to each tube and the resulting color complex was analyzed after 15 min in a spectrophotometer (Gilson model 240) at a wavelength of 553 nm. Modified Ehrlich's reagent contains 1.0 g of p-dimethylaminobenzaldehyde and 0.175 g of HgCl2 dissolved in 42 ml of glacial acetic acid to which has been added 10 m1 of 70% perchloric acid. The molar extinction coeffi- cient of the ALA-ethylacetoacetate-Ehrlich's complex at 553 nm was determined to be 6.2 x 104. A unit (U) of ALAS activity is defined as l nanomole of ALA formed per hour. é-aminolevulinic acid dehydrase (ALAD) Porphobilinogen (PBG) production in cell-free extracts was measured by a modification of the procedure used by Mauzerall and Granick (78). The reaction mixture for assaying ALAD activity contained (in mmoles per liter): HEPES buffer (pH 7.5), 30; MgSOA, 10; 2-mercaptoethanol, 25; and ALA, 5. The reaction was initiated by the addition 31 of 0.5 m1 of the cell-free extract to 0.5 ml of ALAD reaction mixture. The reaction was allowed to proceed for 2 hours in test tubes placed in a 30°C water bath. The reaction was stopped by placing the tubes on ice and adding 0.25 ml of 20% TCA which was saturated with HgClZ. The tubes were then vortexed and allowed to stand on ice over- night. Precipitated protein was removed by centrifugation at 800 x g for 10 min. One ml portions of the supernatants were placed in separate tubes to which one ml of modified Ehrlich's reagent was added. The absorption of the result- ing color complex at 555 nm was then determined in a spec- trophotometer. The molar extinction coefficient of the PBG-Ehrlich's complex was 5.0 x 104. A unit (U) of ALAD activity is defined as l nanomole of PBG formed per hour. Succinic thiokinase (STK) STK activity was estimated by the detection of succinic hydroxamate generated from succinyl-CoA. A modi- fication of the procedure used by Burnham and Lascelles (25) was used. The reaction mixture contained (in mmoles per liter): Naz-succinate, 125; Coenzyme A, 0.5; ATP, 6.25; 2-mercaptoethanol, 12.5; MgClZ, 6.25; and freshly prepared neutral NHZOH, 1000. The neutral NHZOH was produced by titrating 4 M NHZOH-HCl with 7 M NaOH. The reaction was initiated by the addition of 1.0 ml of the cell-free extract to 4.0 ml of the reaction mixture. 32 Incubation was allowed to proceed for 30 min in a 30°C water bath. The reaction was stopped by pipetting 1.0 ml portions into chilled test tubes containing 1.2 m1 of the FeCl3 reagent of Lipmann and Tuttle (80) which contains equal volumes of 5% FeCl3 in 0.1 N HCl, 3 N HCl, and 12% (w/v) TCA. The test tubes were then centrifuged for 5 min at 800 x g. Absorbance at 540 nm was then determined in a spectrophotometer. The molar extinction coefficient of the resulting color complex was determined to be 6.1 x 104. A unit (U) of STK activity is defined as 1 umole of succinyl CoA formed per hour. Results are presented as the average value of two or more separate experiments in which the assays were done in triplicate. Measurement of Heme Biosynthetic Rates The rate of heme biosynthesis in intact R. japonicum cells was determined by measuring the rate of incorporation of [2-14CJ-glycine into heme. Cells were harvested by centrifugation and washed once in ice-cold 0.1 M HEPES buffer (pH 7.0). The cells were then resuspended in fresh buffer at a concentration of approximately one mg protein per ml. 20 m1 of the cell suspension were placed in a 50 m1 Erlenmeyer flask and 20 ul of a ferric citrate solution (4 mg per ml) were added. The flask was then placed in a 33 30°C water bath for 20 min with shaking. After this pre- incubation period, [2-14CJ-glycine (Specific activity = 49.73 mCi per mmole) was added at a concentration of 0.5 uCi per ml. The cells were allowed to incubate for 30 min at 30°C with shaking. Heme synthesis was terminated by pipetting 5 ml samples of the cell suspension into centrifuge tubes containing 30 m1 of ice-cold acetone con- taining 1% HCl. After allowing the tubes to sit on ice for one hour, the precipitated material was removed by centrifugation at 15,000 x g for 20 min. Heme was then purified by the procedure of Cutting and Schulman (17). 500 ug of hemin in 0.1 ml of pyridine was added to each tube and the heme was then extracted into 22.5 ml of diethyl ether. The ether layer was then washed three times with 5 ml of 15% HCl, followed by three wash- ings with 5 ml of water. Any water remaining in the ether layer was removed by the addition of NaCl crystals. The ether layer was then evaporated to dryness and the heme was redissolved in 0.2 ml of pyridine, followed by the addition of 2 ml of water. The heme was precipitated by the addition of 2 ml of 2 N HCl with subsequent heating at 60°C for a few minutes. The precipitated heme was collected on milli- pore filters (pore size = 0.45 pm) which were then dried and glued to planchettes for counting in a gas flow detector (Chicago Nuclear model 1042). Counting effi— ciency was 20%. 34 Thin layer chromatography of the resulting product dissolved in pyridine was performed on silica gel plates (E-M Laboratories) developed with 2,6-lutidine/water, 5:3 (v/v). The vapor phase in this system was saturated with ammonium hydroxide. Solubilization of Protoporphyrin and Hemin In certain experiments, it was required to add hemin and protoporphyrin IX (PROTO) to preparations. Hemin and PROTO were solubilized by adding 1 m1 of 50 mM NaOH to the crystals and mixing. The solutions were then diluted to 15 ml with 0.1 M HEPES buffer (pH 7.0). The solutions were then centrifuged at top speed in a clinical centri- fuge to remove undissolved crystals. In certain cases it was necessary to sterilize the hemin and PROTO solutions. This was accomplished by filtration through millipore filters (pore size = 0.45 pm). The resulting concentra- tion of heme was determined by measuring the difference between the peak at 557 nm and the trough at 540 nm that is observed in the pyridine hemochromogen difference spectrum, assuming a millimolar extinction coefficient of 20.7. The concentration of the PROTO solution was determined by measuring the absorbance of the peak at 556 nm in the acid spectrum of PROTO, assuming a millimolar extinction coefficient of 13.5. 35 Protein Determination The concentration of protein in cell-free extracts was determined by the method of Bradford (81), using bovine y-globulin as a standard. This technique involves protein-dye binding with Coomassie blue. 0.05 ml volumes of the cell-free extracts were added to 5 ml of the dye solution and the resulting color complex was analyzed at 595 nm in a spectrophotometer (Gilson model 240). RESULTS Characteristics of Iron-Limited Growth In order to examine the effects of iron on heme biosynthesis in R. japonicum, it is necessary to limit the growth of the cultures by decreasing the amount of iron in the growth medium that is available to the cells. The growth rates of R. japonicum cultures in L1 medium and HI medium were identical for several generations. At a certain point, however, the growth rate of cultures in LI medium rapidly decreased,presumably due to the lack of available iron (refer to Figure 8). A short stationary phase was observed, followed by a fairly rapid decrease in cell viability. Addition of iron to stationary phase cultures in LI medium resulted in a resumption of growth. Assays were performed on cells in stationary phase. A well-characterized phenomenon associated with iron- limited growth of bacteria involves the excretion of compounds into the medium which have the ability to chelate iron molecules. In addition, there is often a concomitant synthesis of receptor sites for these compounds in the membranes of the iron-starved cells. The 36 37 resulting complexes formed from these chelators and iron are more readily taken up by the cells than naturally- occurring iron-containing compounds. It was of interest to know whether this phenomenon is also operative in R. japonicum. The uptake of iron from the medium by R. japonicum grown in HI medium and LI medium was measured by allowing the cells to incubate in a buffered solution 59Fe. As shown in Figure 2, cells that were 59 containing grown in L1 medium were able to take up Fe at an initial rate 2 to 3 times greater than cells grown in HI medium. Increased capacity for iron uptake under low-iron condi- tions was also suggested by the fact that R. japonicum cultures grown in L1 medium had initial growth rates identical to those of cultures grown in HI medium, despite the large difference in the amount of iron in each of the media. This was especially apparent in that there was not a long transition period into the iron-limited stationary phase. Another common phenomenon associated with iron- limited growth of aerobic organisms involves a reduction in the amount of cytochromes present in the cells. In the present investigation, cytochromes of R. japonicum grown in LI medium and HI medium were examined by means of reduced minus oxidized difference spectra. Difference spectra of both cell-free extracts and whole cell suspen- sions were obtained. Since the various cytochromes have 20 15 pmoles Fe /109 cells or 3 38 \ |——-o—-l |-—O—-l I/I .I. l 3 5 IO 15 Minutes Figure 2. Uptake of iron by R. japonicum. Cells were harvested, washed, and then resuspended in a buffered solution to which was added [59Fe]Cl3. At various time points, samples were removed and the cells were collected on filters and subsequently washed with ice-cold buffer containing 10 mM EDTA. An initial time point was taken, and this value was subtracted from the other time points to account for non-specific adsorption. Symbols: (I) cells grown in HI medium; (0) cells grown in L1 medium. 39 different difference spectra maxima, it is possible to both qualitatively and quantitatively estimate the pres- ence of cytochromes by this technique. The dithionite- reduced minus ferricyanide-oxidized difference spectra of cell-free extracts of R. japonicum are shown in Figure 3. In these spectra, peaks are present only at 552, 523, and 420 nm. The peaks at 552 and 523 nm correspond to the a-peak and B-peak, respectively, of cytochrome c. The level of this cytochrome in extracts of cells grown in H1 medium (Figure 3A) is approximately three times the level in extracts from iron-deficient cells (Figure 3B). The peak at 420 nm could in part be due to the y-peak of cyto- chrome c, but apparently another component contributes to this peak since the decrease in height of this peak is much more pronounced in iron-deficient cells than the decrease in height of the other peaks associated with cytochrome c. It is possible that the peak at 420 nm is associated with the soluble Rhizobium hemoglobin reported by Appleby (82) having a peak at 428 nm. Cytochromes of the a and b-type were not apparent in these spectra, either due to small quantities or because they were removed by centrifugation. In order to detect particle-bound cytochromes, dithionite-reduced minus 02-oxidized difference spectra were obtained of whole cell suspensions. Spectra were obtained with a single beam spectrophotometer, and the difference spectra were calculated by an on-line computer. Figure 3. 40 Reduced minus oxidized difference spectra of cell-free extracts of R. japonicum. Harvested cells were washed, resuspended in phosphate buffer, and sonicated. Soni- cates were clarified by centrifugation and diluted to a protein concentration of 3 mg/ml. One sample was reduced with sodium dithionite and another oxidized with K3Fe(CN)6 in order to obtain the difference spectrum. A, cells grown in HI medium; B, cells grown in L1 medium. The dashed line represents the baseline. 41 420 Absorbance 523 552 1A=.01 430 550 630 Wavelength, nm 42 The difference spectrum of cells grown in HI medium (Figure 4A) reveals the presence of a, b, and c-type cytochromes (a-peaks at 598, 559 and 556, and 548 nm, respectively). These peak locations are very similar to those reported for R. japonicum by Appleby (82). There is also a large peak in the 420-430 region. It is difficult to decipher the difference spectrum obtained from iron- deficient cells (Figure 4B) due to the presence of proto- porphyrin, which has broad difference spectrum maxima at 643, 595, 560, 531, and 485 nm. These peaks make quanti- tation of cytochromes impossible, but do not prevent detection of b-(a-peak at 559 nm) and c-(a-peak at 548 nm) type cytochromes. (The 4 nm shift in the a-peak of cyto- chrome c is a result of freezing the sample in liquid nitrogen.) In addition, the fourth derivatives of these difference spectra reveal that the B-peaks (505-530 nm) and the y-peaks (415-450 nm) for cytochromes a, b, and c are identical in iron-deficient and iron-sufficient cells. A striking difference in the spectrum of iron-deficient cells as compared to iron-sufficient cells is apparent, however, in that there is a drastic reduction in absorb- ance in the 420-430 nm region. Whether this absorbance can be attributed to the previously mentioned Rhizobium hemoglobin is unknown at the present time. The results of these experiments suggest that quantitative, but not qualitative differences exist in the cytochromes of Figure 4. 43 Reduced minus oxidized difference spectra of R. japonicum whole cells. Cells were harvested, washed, and resuspended in phosphate buffer at a con- centration of 15% (wet weight/volume). One sample was reduced with sodium dithionite and another oxidized by passing 02 through the suspension in order to obtain the difference spectra. A, cells grown in H1 medium; B, cells grown in L1 medium. 44 428 Absorbance 548 428 r-559 430 I 560 ‘ 630 Wavelength, nm 45 iron-deficient and iron-sufficient R. japonicum cells. Numerous microorganisms have been shown to excrete porphyrins into the medium when grown under iron-limiting conditions. When R. japonicum cells were grown in L1 medium, porphyrin excretion was also observed. The excreted porphyrins could be observed either adhering to the sides of the culture flask or as insoluble aggregates suspended in the medium. The excretion of porphyrins by R. japonicum grown in L1 medium could be detected fluoro- metrically. As shown in Figure 5, there was no fluores- cence above background levels in cultures grown in LI medium until the growth rate of the cells began to fall off due to iron-limitation. At this point, a large increase in fluorescence was observed, which subsided as culture viability began to decrease. The decrease in fluorescence can probably be attributed to porphyrin aggregation and adherence to the sides of the culture flask. Cultures grown in HI medium did not exhibit an increase in fluorescence. Levulinic acid competitively inhibits the second enzyme of the heme biosynthetic path- way, ALAD, and thus is able to block porphyrin synthesis. The addition of 10 mM levulinic acid to cultures grown in LI medium completely blocked the increase in fluorescence that was normally observed (see Figure 5). It is inter- esting to note that the addition of levulinic acid resulted in a slightly decreased rate of cell death. Figure 5. 46 Fluorescence of R. japonicum cultures as a function of culture age. Cultures grown in H1 medium and LI medium were periodically measured for growth and fluorescence due to porphyrin accumulation. One ml of concentrated HCl was added to 5 m1 of culture fluid and then clarified by centrifugation prior to determination of relative fluorescence in a fluorometer. Closed symbols represent cell counts; open symbols represent relative fluorescence units. Symbols: (0,0) cells grown in HI medium; (I,D) cells grown in L1 medium; (A,A) cells grown in LI medium to which 10 mM levulinic acid had been added at the point indicated by the arrow. 25 relative fluorescence units are equivalent to 10 pmoles of protoporphyrin per m1. 47 «:2: 0053203.”. .3: 5 0 5 0 5 n. w 7 5 2 5 q u q d O A n 0 MA to A\ n\ n\ 0A \ 1 1 ,o 8 T- 0/ A._o o A. o to O / / \ 16 IV.- Iv om /o-/ 8.“ on 0010 2 Pp. . . b - p - *pF .L . . D 9 8 7 O O _ W .5 \ 2.8 “32> Hours 48 By obtaining the absorption spectrum of the excreted porphyrin in l N HCl (Figure 6) and of the porphyrin methyl ester in chloroform (Figure 7), the porphyrin that is excreted by R. japonicum grown in L1 medium was ident- ified as protoporphyrin IX. In addition, the methyl ester of the excreted porphyrin co-chromatographed with authen- tic protoporphyrin IX dimethyl ester in three thin layer chromatography systems (Table 1), including a reverse- phase system. Since some porphyrins are more soluble in an aqueous medium than protoporphyrin, an attempt was made to remove any such porphyrins from centrifuge-clarified culture medium. Any soluble porphyrins would have been removed from the culture medium by filtration through a pad of neutral alumina. This technique did not reveal the presence of any soluble porphyrins, however. Effects of Iron-Limitation on Heme Biosynthesis The rate-limiting enzyme of heme biosynthesis in mammals, many bacteria, and R. japonicum has been shown to be the first enzyme of the pathway, 6-aminolevulinic acid synthase (ALAS). Since the rate-limiting enzyme of a particular biosynthetic pathway is often the regulatory step, the effect of iron-deficiency on ALAS activity in R. japonicum was examined. As can be seen in Figure 8, the activity of this enzyme rapidly decreased as cells 49 Figure 6. Acid spectra of protoporphyrin IX and the porphyrin excreted by R. joponicum grown in L1 medium. Porphyrins were dissolved in l N HCl prior to spectrophotometric determinations. A, porphyrin excreted by iron-deficient cells; B, authentic protoporphyrin IX. 50 ’ 408 555 I I p 408 555 Absorbance 10x .3. L /\—\ 1x 400 530 600 Wavelength, nm 51 Figure 7. Neutral spectra of the methyl esters of protoporphyrin IX and the porphyrin excreted by R. japonicum grown in LI medium. Porphyrins were methylated as described in MATERIALS AND METHODS and then dissolved in chloroform prior to spectrophotometric determinations. A, methyl ester of the porphyrin excreted by iron-deficient R. japonicum; B, authentic protoporphyrin IX dimethyl ester. Absorbance 52 400 560 63b Wavelength, nm 53 Table 1. Chromatographic properties of the methyl ester of the porphyrin excreted by R. japonicum grown in LI medium. Rf Values System URO I+III COPRO III PROTO IX Excreted Standard Standard Standard Porphyrin I 0.13 0.34 0.49 0.49 II ‘0.18 0.40 0.48 0.48 III 0.93 0.68 0.38 0.38 Systems: I = benzene/ethyl acetate/ethanol (l90/20/7.5) (v/v/v) on Brinkman Silica Gel G TLC plates II = petroleum ether/paraffin oil/chloroform (l/l/lO,v/v/v) on Eastman Chromagram Sheets III = water/acetonitrile/p-dioxane (2/7/1, v/v/v) on Whatman KC1 reverse phase TLC plates (Solvent contained 0.5 M NaCl to stabilize the hydocarbon chains.) Abbreviations: URO = uroporphyrin; COPRO = coproporphyrin; PROTO = protoporphyrin Figure 8. 54 ALAS activity in R. japonicum as a function of culture age. ALAS activity was measured as described in MATERIALS AND METHODS. Closed symbols refer to cell counts; open symbols refer to ALAS activity. Symbols: (0,0) cells grown in HI medium; (I,D) cells grown in L1 medium; (A,A) cells grown in LI medium to which ferric citrate (4 mg/l) had been added at the point indicated by the arrow. 55 A5226. 9S: 33:2 m<5< 0 1.8642 -bbbbb D P h -PPDD P b/b \ l L 120 so Hours 9 o 1 o .E\u__oo can; 7 100 56 grown in LI medium entered the iron-limited stage of growth. This decrease in activity was observed before the rapid decline in culture viability took place. On the other hand, ALAS activity of cells grown in HI medium remained constant throughout the growth cycle (see Figure 8). The increased ALAS activity of cells grown in ferric citrate-containing HI medium was shown to be truly a result of the iron, and not of the citrate ion, because both ferric citrate and ferric chloride additions to the medium resulted in increased ALAS activity, while growth in the presence of sodium citrate without added iron resulted in low ALAS activity (Table 2). Decreased ALAS activity in the absence of high levels of iron was observed in both a complex medium (yeast extract-mannitol) and in a more defined medium (LI medium). In addition, transition metals other than iron were not able to main- tain high ALAS activity when iron was withheld from the growth medium (Table 2). As can be seen in Table 3, the second enzyme of the heme biosynthetic pathway, ALAD, also appears to be affected by the supply of iron to R. japonicum cells. This enzyme decreased by approximately the same degree as ALAS when the cells were subjected to iron-limitation. This observation suggests that there may be some sort of coordinate regulation of the heme biosynthetic pathway 57 Table 2. ALAS activity in cell-free extracts of R. japonicum grown in the presence of iron and other metal salts. Additions to ALAS ACtiVitY (U/mg protein) growth medium YM Medium LI Medium none 3.3 i 0.7 3.0 i 1.3 Fe citrate (4 mg/l) 6.6 i 0.5 8.8 i 2.1 FeCl3 (11.7 uM)a 7.9 i 0.3 10.1 i 1.3 Na3 citrate (17.8 uM)a 3.8 i 0.3 3.3 i 1.3 CoCl2 (10 uM ) ----- 3.3 i 0.7 ZnSO4 (10 0M) ----- 2.8 i 1.1 CuSO4 (10 uM) ----- 2.3 i 1.2 Cells were grown in either complex medium (yeast extract- mannitol, YM) or defined medium (LI medium) containing additional salts as indicated. 3 Concentrations equivalent to the amount of iron or citrate in media containing 4 mg ferric citrate / 1. 58 Table 3. Activities of several enzymes of heme biosynthesis in cell-free extracts of R. japonicum grown in H1 medium and LI medium. Growth STK Activity ALAS Activity ALAD Activity Medium (U/mg protein) (U/mg protein) (U/mg protein) LI ---- 4.l-_l-0.9 2.1102 HI ---- 8.9 i 1.6 6.9 i 1.0 LI 35.2+7.8 2.8+0.1 ..... HI 28.6 + 7.8 9.8 + 1.3 ----- 59 imposed by iron-limited growth. Succinic thiokinase (STK) is an enzyme that is responsible for the production of succinyl-CoA, a small part of which is used as a substrate for the production of ALA by ALAS. As shown in Table 3, the activity of this enzyme did not decrease during iron- limited growth. In fact, there was a slight increase in this enzyme's activity. Since growth of R. japonicum under iron-deficient conditions led to decreased activities of ALAS and ALAD in vitro, it was of interest to know whether this effect was also manifest in vivo. This was determined by washing and resuspending the cells in a buffered solution to which [2-14CJ-glycine was added. The rate at which this labelled compound is incorporated into heme allows the investigator to measure the relative rates of heme biosyn- thesis in different batches of cells. It was necessary to purify the heme produced from the [2-14C]-glycine in order to eliminate any radioactivity not due to heme. Thin layer chromatography of the purified heme revealed that the majority of the radioactivity did, in fact, co- migrate with authentic heme. As shown in Table 4, cells grown in HI medium incorporated [Z-IACJ-glycine into heme at a rate slightly greater than twice that of cells grown in LI medium. Thus, the decreased activities of heme biosynthetic enzymes in iron-limited cells which were observed in vitro also appeared to be manifest in vivo. 60 14C -glycine into Table 4. Rates of incorporation of 2- heme in R. japonicum grown in HI medium and LI medium. figggfi; cpm/mg protein/hour LI 312 i 73 HI 720 + 135 Cells were harvested, washed, and resuspended in HEPES buffer (pH 7.0) at a concentration of approximately 1 mg/ml. [2-14CJ-glycine was added and the suspensions were allowed to incubate for 30 min at 300 C. Heme was purified as described in MATERIALS AND METHODS. Radio- active heme was measured with a gas flow detector at an efficiency of 20%. 61 Possible Mechanisms of Iron Effects It does not appear that iron directly affects the activity of ALAS or ALAD. The addition of 1 mM FeCl3 or 1 mM FeSOA to cell-free extracts of iron-limited cells did not cause a detectable increase in the activity of either of these enzymes (Table 5). Concentrations of FeCl3 and FeSO4 in the range of 10 uM to 200 uM were also without effect (data not shown), thus precluding the possibility that a concentration of 1 mM may have been inhibitory due to its being a supraoptimal concentration. The addition of sodium dithionite to the cell-free extracts containing FeSO4 did not lead to enhanced ALAS or ALAD activity, thus discounting the possibility that the ferrous iron was not having an effect as a result of oxidation to ferric iron. Additional evidence suggesting that iron does not play a direct role in the activity of ALAS and ALAD is the fact that iron chelating agents did not diminish the activity of ALAS and ALAD in cell-free extracts of cells grown in HI medium. Desferal (desferrioxamine mesylate) is a chelator with a high affinity for ferric iron, while u,a'-dipyridy1 is a chelator with a high affinity for ferrous iron. Neither of these compounds substantially decreased the activity of either ALAS or ALAD when added to cell-free extracts at a concentration of 1 mM (Table 6). 62 Table 5. Effects of ferric and ferrous iron on ALAS and ALAD activities in cell-free extracts of R. japonicum grown in LI medium. Additions to ALAS Activity ALAD Activity reaction mixture (U/mg protein) (U/mg protein) . none 2.7 i 0.5 2.1 t 0.2 FeCl3 (1 mM) 2.7 i 0.5 2.1 i 0.2 FeSO4 (1 mM) 2.7 i 0.6 2.1 i 0.2 FeCl3 and FeSO4 were added to the reaction mixtures at the onset of the 2 hour incubation. 63 Table 6. Effects of iron chelating agents on ALAS and ALAD activities in cell-free extracts of R. japonicum grown in HI medium. Additions to ALAS Activity ALAD Activity reaction mixture (U/mg protein) (U/mg protein) none 7.8 i 1.0 7.6 i 0.9 a,0'-dipyridyl (1 mM) 7.7 i 0.9 7.6 i 0.8 Desferal (1 mM) 7.8 + 1.0 7.6 + 0.9 Chelating agents were added to the reaction mixtures at the onset of the 2 hour incubation. 64 Another possible mechanism which could result in decreased enzymatic activity in iron-deficient cells could be that the enzyme was more stable in the presence of iron or some iron-containing compound. To investigate this possibility, cell-free extracts were obtained from iron-deficient and iron-sufficient cells. FeCl3 (50 uM) was then added to the extract from the iron-sufficient cells. The cell-free extracts were maintained at 00 C, and aliquots were periodically removed and subsequently assayed for ALAS activity. As shown in Figure 9, the in vitro half-life of ALAS from iron-deficient and iron- sufficient cells was not substantially different. These results suggest that iron does not appear to be able to directly stabilize ALAS. As described in the introduction, iron has been shown to enhance the activities of several pyridoxal phosphate- containing enzymes. This is attributed to increased catalytic activity of the pyridoxal phosphate molecule in the presence of iron. Several pyridoxal phosphate - mediated reactions can be carried out non-enzymatically (e.g. transamination), and the reaction is often enhanced by the addition of metal ions, including iron. An attempt was made to form ALA non-enzymatically by combining 100 pM concentrations of succinyl-CoA, glycine, and pyridoxal phosphate,and heating at 750 C. However, no detectable 65 100 70 tn CD °/., Activity on CD T i 6 12 18 24, 30 36 Hours Figure 9. Half-life of ALAS in cell-free extracts of R. japonicum. Cell-free extracts were prepared as previously described and diluted to a protein concen- tration of 0.5 mg/ml. FeCl3 (50 uM) was added to the extract from cells grown in H1 medium. The extracts were maintained at 0°C and samples were periodically removed and measured for ALAS activity as previously described. Symbols: (0) cells grown in HI medium; (I) cells grown in L1 medium. 66 ALA was formed under these conditions, even upon the addi- tion of 100 uM FeC13 or FeSO4 Since iron-deficient cells do not have available iron to convert protoporphyrin to heme, a possible explanation for decreased ALAS and ALAD activities might involve the need for activation or induction of these enzymes by heme. If such a situation existed, a decrease in the heme content of the cells would lead to decreased ALAS and ALAD activities. To test this possibility, hemin was added to cell-free extracts of iron-deficient cells. As can be seen in Table 7, however, the addition of hemin at concen- trations up to 200 uM did not increase the activities of ALAS or ALAD in these extracts. Growth of R. japonicum in LI medium with added hemin (50 uM) also did not result in increased ALAS or ALAD activity (Table 8), but it is not known whether the exogenously added hemin was able to enter the cells. Additions of hemin in the range of 50 - 200 uM to cell-free extracts of cells grown in HI medium also did not appear to affect ALAS or ALAD activity (Table 7). Thus, heme does not appear to be able to con- trol its own synthesis via feedback inhibition of these enzymes, either. h Since iron-limited R. japonicum cells excrete proto- porphyrin (PROTO) into the medium, an attempt was made to determine if this compound has a direct feedback inhibi- tory effect on ALAS and ALAD. .As shown in Table 7, 67 mHQB xH sawhsmhoaouOHa H 090mm "macaumfi>ounn< .COHumbnocH H50: N ecu mo uomco gnu um moHSONHE cOHuomoH mflu on women mma0huzamuuoe .poumowccw mm finance H4 H0 Hm Hmnuflo ca cBOHw mneB mHHoU N o H o N m o H H m Hz: ooNV :Hse: HH I I N .Hexm N o + o N m o + H m use: HH ..... m o H m N Hz: QONV oeome Hm H o H a m m o H a N Hz: OONV tHatm H: H .uexm H o H a m m.o H e.N one: Hm AcHououa wE\DV Aaflououm wE\Dv ousuxwfi Goauomou sawed: HHH>HHU< aHuo< mwuom Qoun£< .cowuflppm wasp umumm meson em vthompom mums mewmmm oE>Nam .nu30pw mo mmmsa HmHucmaomxo mama ozu aw mouduaso on women onB m0a0humamuum9 m o H H N m o H N m Hz: one cHsmm HH I I N .uexm e o + w H o H + N m econ HH w o H m e m o H e N Hz: ooHV oeome Hz 8 o H m e m H H e m Hz; omv cHsmm Hm H .Hexm w o H m e e o H m w etc: He AcHououm wE\Dv Acflmuoua wE\DV shaves £u3ouu Shana: NHH>HHU< oHHU< mwuom QoH C033 .aowuwppm menu momma mHDO£ ma pthomuma oho3 mzmmmm ofihwcm .owmum Hmwucmaoaxm mama cw mousuaso ou woven me cwom UHCHHS>0H .poumoHpGH cosz - -- . . H28 OHV - - N o + e N eHem UHcHHs>mH econ m.o H e.N 8.0 H H.m meet eHetHMficmww>eH N.o + m.H N.o + o.m ococ econ Acwmuoum wE\Dv Acwmuoua wENDv ousuxwa cowuomwn EDHpoE HA huH>Huo< Qfluo< mHuee omH Ho esteem .e mHHte 73 of cells harvested 18 hours after the addition of levulinic acid were nearly twice as high as those of extracts from iron-limited cells to which no levulinic acid had been added. It should be noted that these enhanced activities were not quite as high as cells grown in HI medium, however. The increased activity cannot be attributed to in vitro effects of levulinic acid, since there was no enhancement of enzymatic activity when 10 mM levulinic acid was added directly to the cell-free extracts. An attempt was made to gain further evidence for PROTO-mediated reduction of ALAS activity by adding PROTO to cultures growing in HI medium containing dimethyl sulfoxide (DMSO). DMSO is known to permeabilize cell membranes by interfering with the lipid-water interface. As shown in Figure 11, 5% DMSO does not have an appreci- able effect on growth of R. joponicum in HI medium. An experiment was performed in which DMSO and PROTO were added to late exponential stage cultures grown in HI medium, with subsequent assay for ALAS activity performed 24 hours after this addition. As shown in Table 10, ALAS activity in cell-free extracts from cells grown in the presence of 5% DMSO and 100 uM PROTO was 33% lower than in extracts from cells grown in the presence of 5% DMSO only. However, this decreased ALAS activity was not as low as that of iron-limited cells. The addition of DMSO 74 - o I ./ j;:- o 100 :- /' i t t +- 1/ P o ”a? 3: - C: a P “ 1; o 3.910 .__ :5 l— ; — C) b 5 - e r- ] l J l J 20 4O 60 80 Hours Figure 11. Effect of dimethyl sulfoxide on the growth of R. japonicum. Dimethyl sulfoxide (DMSO) was added to cultures growing in H1 medium at a concen- tration of 5% (v/v) at the point indicated by the arrow. Symbols: (I) growth in the absence of DMSO; (I) growth in the presence of DMSO. 75 Table 10. ALAS activity in cell-free extracts of R. japonicum grown in H1 medium in the presence of dimethyl sulfoxide and certain tetrapyrroles. Additions to ALAS Activity‘ 0 HI medium (U/mg protein) A decrease DMSO (5%) 8.8 i 0.3 ---- Expt. 1 DMSO (5%) + PROTO (100 uM) 5.9 i 0.5 33% DMSO (5%) 7.6 i 0.5 ---- Expt. 2 DMSO (5%) + Hemin (10 0M) 4.1 + 0.6 46% DMSO and tetrapyrroles were added to cultures in the late exponential phase of growth. ALAS assays were performed 24 hours after this addition. Abbreviations: DMSO = dimethyl sulfoxide PROTO = protoporphyrin IX 76 alone did not result in ALAS activity that was lower than what was normally observed in extracts from iron-sufficient cells. The addition of hemin (10 uM) to cultures growing in H1 medium in the presence of DMSO also resulted in decreased ALAS activity (Table 10). In fact, the inhibi- tion caused by hemin was substantially greater than that caused by PROTO. It should be mentioned at this point that growth of R. joponicum in DMSO and PROTO or hemin resulted in a great deal of slime production which made harvesting of cells very difficult. The nature of this extracellular material is not known. This slime production was not observed in cells grown in the presence of DMSO alone. DISCUSSION The growth of R. japonicum was limited by decreasing the amount of iron available to the cells. Several characteristics of this iron-limited state of growth have been examined, with special emphasis on the effects of iron-deficiency on heme biosynthesis. A common phenomenon associated with iron-limited growth of many microorganisms involves an increased capacity for the uptake of iron from the growth medium. This is generally attributed to the fact that the lack of iron derepresses the synthesis of various compounds that have the ability to chelate iron molecules. These compounds are of two general types; secondary hydroxamates and phenolic acid-containing molecules. Specific trans- port systems for these compounds are also often produced as a result of iron-deficiency, and thus the chelated iron molecules are more available to the cells than naturally- occurring forms of iron. R. japonicum cells grown in L1 medium were shown in this study to have an increased rate of iron uptake from the medium as compared to cells grown in HI medium (see 77 ’s 78 Figure 2). This finding might indicate that R. japonicum may also have a chelator-mediated system for enhanced iron uptake that is repressed in cells grown in media contain- ing sufficient iron. Further work must be done in order to characterize this phenomenon, however. The presence of an iron-uptake system in Rhizobium that has increased activity in low-iron conditions would be beneficial to the root nodule symbiosis since it would allow heme synthesis to occur at maximal rates. In addition, it would provide sufficient iron for the rhizobial iron-containing proteins that are important in the symbiosis, such as nitrogenase. Another finding suggesting that R. japonicum cells grown in LI medium are truly iron-deficient was the fact that the cytochrome content of these cells was markedly decreased from that of cells grown in medium containing high levels of iron (Figure 3). This phenomenon has commonly been observed in other organisms grown under iron-limiting conditions (69). As discussed in the intro- duction, there is evidence suggesting that the decreased cytochrome level observed in other cells may be a result of decreased apoprotein synthesis and not merely a result of insufficient amounts of iron to form the heme prosthetic group. In such cases, the addition of heme to iron- deficient cells is generally able to stimulate apoprotein synthesis, thus suggesting that the regulation of heme and apoprotein biosynthesis may be integrally related (56,85). 79 It appears that the heme, or some heme-containing compound, is able to control protein synthesis at the level of transcription or translation. The mechanism by which heme is able to control apoprotein synthesis is not known, however. The stimulating effect of heme on apoprotein synthesis is not universal, however, since in several organisms reconstitution of normal hemoprotein levels is obtained immediately upon the addition of heme to heme- deficient cells (60,83). Also, an ALA-requiring mutant of E. coZi forms the apoproteins of the cytochromes despite the lack of heme molecules (84). The regulation of apoprotein synthesis by heme is of interest to the present investigation because a similar phenomenon might be responsible for regulation of leghemoglobin apoprotein synthesis in the legume root nodule. A situation could exist that is similar to the well-characterized effect of heme on globin synthesis in mammalian reticulocytes. In this case, the heme stimu- lates globin synthesis by preventing the activation of an endogenous translational inhibitor. There is no specificity as to which proteins have heme-stimulated rates of synthesis in this type of regulation, but since globin synthesis accounts for the vast majority of the protein produced by reticulocytes, there may not be a need for specific regulation. An analogous situation could exist in the legume root nodule; since leghemoglobin 80 accounts for up to 40% of the total soluble protein in the nodule, it is quite possible that the plant cells are differentiated to the extent that leghemoglobin apoprotein is the major protein being produced. Whether this type of control indeed exists in the nodule is unknown, how- ever. An investigation of possible heme-regulated apo- leghemoglobin synthesis in legume root nodules should certainly be undertaken. There is evidence produced from experimentation with SpiPiZZum itcrsonii which suggests that in addition to decreased cytochrome apoprotein synthesis under heme- deficient conditions, heme synthesis may be decreased under conditions in which apoprotein synthesis is blocked (85). Thus, dual control mechanisms may be present to ensure equal rates of synthesis of the apoprotein and of the heme prosthetic group. In the same report, it was shown that decreasing the aeration of S. itcrsonii cultures led to a parallel increase in cytochrome content and ALAS activity. It is interesting to note that a similar phenomenon has been observed in R. japonicum (87). By decreasing the aeration of cultures, a drastic increase in both ALAS activity and the heme content of this organ- ism was observed. Whether all of the heme was associated with cytochromes is unknown, however. Another effect of iron-limited growth of R. japonicum involved the excretion of protoporphyrin (PROTO) into the 81 medium (Figure 5). This was apparently due simply to the lack of iron for insertion into the porphyrin ring to make heme. It should be noted, however, that the growth of many microorganisms under iron-deficient con- ditions normally leads to c6proporphyrin excretion rather than PROTO excretion. Coproporphyrin is not a true inter- mediate of the heme biosynthetic pathway, but arises via oxidation of accumulated coproporphyrinogen. This fact suggests that coproporphyrinogen oxidase in some way requires iron for activity in many organisms. As discussed in the introduction, however, attempts to inhibit the conversion of coproporphyrinogen to PROTO by the addition of iron chelating agents have only been successful in a few organisms. At any rate, the fact that PROTO was the only porphyrin excreted by R. japonicum grown under iron- deficient conditions may indicate that iron is not required for coproporphyrinogen oxidase activity in this organism. Another significant finding of this investigation was that ALAS activity rapidly decreased as R. japonicum cells entered an iron-limited stage of growth. Addition of ferric citrate to iron-limited cultures brought about a slow increase both in the growth rate and in ALAS activity (see Figure 8). The second enzyme of the heme biosynthetic pathway, ALAD, also was shown to decrease under iron-deficient conditions (Table 3). The relative decrease of ALAD activity was approximately the same 82 as that of ALAS, suggesting that there may be coordinated regulation of these enzymes. Decreased capacity for heme biosynthesis in vivo was suggested by the fact that [2-14CJ-glycine was incorporated into heme at a slower rate in iron-deficient cells as compared to iron-suffi- cient cells (Table 4). The activity of succinic thio- kinase did not exhibit a decrease in activity in iron- deficient cells, however (Table 3). The observed effects of iron-limitation on the acti- vities of ALAS and ALAD would result if these enzymes required the presence of iron for catalytic activity. However, no evidence was obtained for a direct activation of ALAS or ALAD by iron. The addition of ferric and ferrous iron to extracts of iron-deficient cells did not lead to enhanced activities of these enzymes (Table 5). It is possible that the iron had to be metabolized to some other form before it could have an enhancing effect, but this possibility could not be checked in cell-free extracts due to a rapid loss of enzymatic activity at room temperature. The fact that the addition of ferric citrate to iron-deficient cultures led to a rather slow increase in ALAS activity (see Figure 8) also suggests that there is not a direct effect of iron on ALAS acti- vity. In addition, mixing of extracts from iron-deficient cells and iron-sufficient cells led to ALAS and ALAD activities midway between the extremes, further casting 83 doubt on the presence of direct activators or inhibitors of these enzymes. The addition of iron chelating agents to cell-free extracts of iron-sufficient R. japonicum did not result in decreased ALAS and ALAD activities, thus further sup- porting the hypothesis that iron plays no direct role in ALAS or ALAD activity. These experiments also suggest that the enhanced ALAS activity in high-iron conditions is not the result of increased pyridoxal phosphate cata- lytic ability due to the presence of iron. It is uncer- tain, however, if the chelators used would have been able to remove iron that was tightly bound within the polypeptide framework of the protein, and therefore this possibility cannot be eliminated. Iron was not able to directly stabilize ALAS in cell- free extracts maintained at 0°C. This experiment does not preclude the possibility that the in vivo stability of ALAS may be affected by the iron status of the cell, however. It is unlikely, however, that compounds are produced during iron-limited growth which cause a specific degradation of ALAS (e.g. - an ALAS-specific protease), since the mixing of extracts from iron-deficient and iron- sufficient cells led to an ALAS activity that was the average of the two extremes. It is possible, however, that there could be a compound which is metabolically produced from iron that is able to stabilize ALAS. 84 Unfortunately, the rapid loss of ALAS activity at room temperature does not allow lengthly preincubation studies to be carried out to test this possibility. Since iron-limited growth of R. japonicum results in protoporphyrin excretion and a concomitant reduction in ALAS and ALAD activities, it is possible that PROTO feed- back regulates ALAS and ALAD. On the other hand, decreased ALAS and ALAD activities would be a result of iron-defi- ciency if the presence of normal levels of heme was required for maximal activities of these enzymes. However, no evidence could be obtained to support the latter possi- bility (Tables 7 and 8). An attempt was also made to detect feedback inhibition of ALAS and ALAD by heme. No inhibition of either of these enzyme activities could be detected when concentrations of heme up to 200 uM were added to cell-free extracts (Table 7). These findings are significant when one considers the fact that feedback inhibition of ALAS (and often ALAD) by heme is a rather common phenomenon. For example, ALAS activity in R. spheroides is 50% inhibited at a concentration of 1 uM heme (25). This same degree of inhibition of ALAS in S. itersonii occurs at a concentration of 10 uM heme (85). There is evidence that heme biosynthesis from 14C-ALA in R. japonicum appears to be 50% inhibited when 100 uM heme is added to the cell suspensions (86). This suggests that there may be another point of regulation of 85 the pathway that is affected by heme. The possibility also exists, however, that the exogenously added heme may have had an indirect effect, such as inhibiting the uptake of 14 C-ALA, but this possibility was not examined by the investigators. A Protoporphyrin (PROTO) also does not appear to directly feedback inhibit ALAS at concentrations up to 200 uM (Table 7). Experiments which attempted to show feedback inhibition of ALAD by PROTO were inconclusive, due to the interference of PROTO with the colorimetric reaction involving p-dimethylaminobenzaldehyde (Figure 10). The interference that was observed may be especially significant when one considers the fact that several investigators have reported apparent feedback inhibition of ALAD by PROTO when using the same assay procedure. Growth of R. japonicum in 100 uM PROTO also did not reveal a significant decrease in ALAS or ALAD activity (Table 8). However, it was quite possible that the PROTO was unable to enter the cells. Indirect evidence was obtained, however, which suggests that PROTO may be able to regulate ALAS and ALAD activities. The addition of 10 mM levulinic acid to cultures growing in LI medium was able to completely block PROTO accumulation. ALAS and ALAD activities in extracts of cells harvested 18 hours after the addition of 10mM levulinic acid to cultures growing in L1 medium were almost twice as high 86 as those of extracts from cells grown in L1 medium in the absence of levulinic acid (Table 9). Thus, the pre- vention of PROTO accumulation in iron-deficient cells permits relatively high ALAS and ALAD activities to persist. A point to consider, however, is that since levulinic acid inhibits heme biosynthesis, there is probably a decreased sink for iron, and therefore the cul- tures containing levulinic acid may not be truly iron- deficient. If this is the case, it cannot be said that the higher activities of ALAS and ALAD in levulinic acid- containing cells is due to decreased PROTO accumulation. In addition, growth of R. japonicum in HI medium to which dimethyl sulfoxide (DMSO) and 100 0M PROTO had been added resulted in ALAS activities that were approximately 1/3 lower than those of cells grown in the presence of DMSO without added PROTO (Table 10). However, hemin also decreased ALAS activity when added to cultures growing in H1 medium in the presence of DMSO. There is no reason to exclude the possibility that the exogenously added PROTO was metabolized to heme, and that a resulting high level of heme was actually causing the decreased ALAS activity. A mutant of R. japonicum with a lesion in the heme biosyn- thetic pathway at the step catalyzed by ferrochelatase would be helpful in clarifying this problem. Therefore, these experiments suggest, but do not conclusively prove, that decreased ALAS and ALAD 87 activities may be the result of some type of feedback regulation that is mediated by PROTO. An earlier report from this laboratory (87) indicated that R. japonicum grown under decreased aeration had increased ALAS and ALAD activities, and yet these cells excreted PROTO into the growth medium. These results are contradictory to a control mechanism involving any type of feedback regu- lation involving PROTO. It is possible that there are two different regulatory processes occurring simultaneously; one involving control by oxygen and the other involving control by feedback regulation. If this is indeed the case, it is apparent that the control mediated by oxygen tension has precedence over feedback regulation by PROTO. In conclusion, this work has shown that the biosyn- thesis of heme in R. japonicum is affected by the iron status of the cell. Evidence has been provided that indi- cates a decreased capacity for heme biosynthesis in iron- deficient cells. This decrease in heme biosynthetic ability may possibly be due to feedback regulation by protoporphyrin. In addition, it has been shown that the cytochrome levels of iron-deficient cells are markedly reduced from the levels present in iron-sufficient cells. In the legume root nodule, the host plant provides the Rhizobium bacteroids with all of the nutrients necessary for growth. Iron is undoubtedly one of the nutrients which must be obtained by the bacteroids from ht 88 the plant, and this fact allows for the possibility that the release of iron from the host plant to the bacteroids could provide a means for regulating heme biosynthesis for the production of leghemoglobin. Additionally, since heme has been shown to control the synthesis of the apo- protein moieties of various hemoproteins in many organisms, the investigation of heme biosynthesis by Rhizobium takes on even greater significance. BIBLIOGRAPHY BIBLIOGRAPHY Nash, D.T., and H.M. Schulman. 1976. Leghemoglobin and nitrogenase activity during soybean root 1 nodule development. Can. J. Bot. 54:2790-2797. Appleby, C.A., N.A. Nicola, J.G.R. Hurrell, and S.J. Leach. 1975. Characterization and improved separation of soybean leghemoglobins. Biochemistry: 14:4444-4450. Wittenberg, J.A., C.A. Appleby, and B.A. Wittenberg. 1972. The kinetics of the reactions of leghemo- globin with oxygen and carbon monoxide. J. Biol. Chem. 247:527-531. Vainshtein, B.K., E.H. Harutyunyan, I.P. Kuranov, V.V. Borisov, N.I. Sosfenov, A.G. Pavlovsky, A.I. Grebenko, and N.V. Konareva. 1975. Structure of leghemoglobin from lupin root nodules at 5 A resolution. Nature 254:163-164. Nicola, N. A. E. Minasian, C. A. Appleby, and S. J. Leach. 1975. Circular dichroism studies of myog lo- bin and leghemoglobin. Biochemistry 14: 5141- 5149. Hunt, J.A. 1972. in Atlas of Protein Sequence and Structure (M.O. Dayhoff, ed.) National Biomedical Research Foundation. Washington. pp. 369-370. Appleby, C.A., W.E. Blumberg, J. Peisach, B.A. Wittenberg. and J.A. Wittenberg. 1976. Leghemo- globin: An electron paramagnetic resonance and optical spectral study of the free protein and its complexes with nicotinate and acetate. J. Biol. Chem. 251:6090-6096. Virtanen, A.I., J. Erkama, and H. Linkola. 1947. On the relation between nitrogen fixation and leghemoglobin content of leguminous root nodules. Acta Chem. Scand. 1:861-870. 89 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. 90 Bergersen, F.J., and G.L. Turner. 1967. Nitrogen fixation by the bacteroid fraction of breis of soybean root nodules. Biochim. Biophys. Acta 141:507-515. Keister, D.L. 1975. Acetylene reduction by pure cultures of Rhizobium. J. Bact. 123:1265-1268. Laane, C., H. Haaken, and C. Veeger. 1978. Involve- ment of the cytoplasmic membrane in nitrogen fixation by Rhizobium Zeguminosarum bacteroids. Eur. J. Biochem. 87:147-153. Coventry, D.R., M.J. Trinick, and C.A. Appleby. 1976. A search for a leghemoglobin-like compound in root nodules of Trema cannabina Lour. Biochim. Biophys. Acta 420:105-111. Burris, R.H., and W.H. Orme-Johnson. 1974. Nitro- genase and its EPR properties, in Microbial Iron Metabolism (J.B. Neilands, ed.). Academic Press: New York. Keilin, D., and Y.L. Wang. 1945. Hemoglobin in the root nodules of leguminous plants. Nature 155: 227-229. Appleby, C.A. 1962. The oxygen equilibrium of leg— hemoglobin. Biochim. Biophys. Acta 60:226-235. Imamura, T., A. Riggs, and Q.H. Gibson. 1972. Equilibria and kinetics of ligand binding by leghemoglobin from soybean root nodules. J. Biol. Chem. 247:521-526. Cutting, J.A., and H.M. Schulman. 1969. The site of heme synthesis in soybean root nodules. Biochim. Biophys. Acta 192:486-493. Godfrey, C.A., and M.J. Dilworth. 1971. Haem biosyn- thesis from 14C-6-aminolevulinic acid in labora- tory-grown and root nodule Rhizobium Zupini. J. Gen. Micro. 69, 385-390. Nadler, K.D., and Y. Avissar. 1977. Heme biosynthesis in soybean root nodules. 1. On the role of bacter- oid ALAS and ALAD in the synthesis of the heme of leghemoglobin. Plant Physiol. 60:433-436. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 91 Cutting, J.A., and H.M. Schulman. 1971. The biogen- esis of leghemoglobin: The determinant in the Rhizobium-legume symbiosis for leghemoglobin specificity. Biochim. Biophys. Acta 229:58-62. Dilworth, M.J. 1969. The plant as the genetic deter- minant of leghemoglobin production in the legume root nodule. Biochim. Biophys. Acta 184 432-441. Verma, D.P.S., D.T. Nash, and H.M. Schulman. 1974. Isolation and in vitro translation of soybean leghemoglobin mRNA. Nature 251:74-77. Baulcombe, D., and D.P.S. Verma. 1978. Preparation of complimentary DNA for leghemoglobin and direct demonstration that leghemoglobin is encoded by the soybean genome. Nucleic Acids Res. 5:4141-4153. Lascelles, J. 1960. The synthesis of enzymes concerned in bacteriochlorophyll formation in growing cultures of Rhodopseudomonas spheroidcs. J. Gen. Micro. 23:487-498. Burnham, B.F., and J. Lascelles. 1963. Control of porphyrin biosynthesis through a negative-feedback mechanism. Studies with preparations of ALAS and ALAD from Rhodopscudomonas spheroides. Biochem. J. 87:462-472. Burnham, B.F. 1971. Metabolism of porphyrins and corrinoids, in Metabolic Pathways, 3rd ed. Vol III. Amino Acids and Tetrapyrroles. Academic Press: New York. Tait, G.H. 1972. 5-aminolevulinate synthetase of Micrococcus denitrificans. Biochem. J. 128:32P. Granick, S. 1966. The induction in vitro of the synthesis of ALAS in chemical porphyria: A response to certain drugs, sex hormones, and foreign chem- icals. J. Biol. Chem. 241:1359-1375. Karibian, D., and I.M. London. 1965. Control of heme synthesis by feedback inhibition. Biochem. Biophys. Res. Comm. 18:243-249. Wolfson, S.J., A. Bartszak, and J.R. Bloomer. 1979. Effect of endogenous heme generation on ALAS acti- vity in rat liver mitochondria. J. Biol. Chem. 254:3543-3546. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 92 Coleman, D.L. 1966. Purification and properties of 6-aminolevulinate dehydratase from tissues of two strains of mice. J. Biol. Chem. 241:5511-5517. Callisano, P., D. Bonsignore, and C. Cartasegna. 1966. Control of heme synthesis by feedback inhibition on human erythrocyte ALAD. Biochem. J. 101:550-552. Tien, W., and D.C. White. 1968. Linear sequential arrangement of genes for the biosynthetic pathway of protoheme in Staphylococcus aureus. Proc. Nat. Acad. Sci. 61:1392-1398. Sasarman, A., M. Surdeanu, G. Szegli, T. Horodniceanu, V. Greceanu, and A. Dumitrescu. 1968. Hemin-defi- cient mutants of E. coli K12. J. Bact. 96:570-572. Shemin, D., and C.S. Russell. 1953. ALA, its role in the biosynthesis of porphyrins and purines. J. Am. Chem. Soc. 75:4873-4874. Muthukrishnan, S., K. Malathi, and G. Padmanaban. 1972. ALAD, the regulatory enzyme of the haem biosynthetic pathway in Neurospora crassa. Biochem. J. 129:31-37. Jayaraman, J., G. Padmanaban, K. Malathi, and P.S. Sarma. 1971. Haem synthesis during mitochondro- genesis in yeast. Biochem. J. 121:531-535. Labbe-Bois, R., and C. Volland. 1977. Changes in the activities of the protoheme-synthesizing system during the growth of yeast under different con- ditions. Arch. Biochem. Biophys. 179:565-577. Knight, E., and R.W.F. Hardy. 1966. Isolation and characteristics of flavodoxin from nitrogen- fixing Clostridium pasteurianum. J. Biol. Chem. 241:2752-2756. Imai, K., A. Asano, and R. Sato. 1968. Oxidative phosphorylation in Micrococcus denitrificans. J. Biochem. 63 219-225. McCandliss, R.J., and K.M. Herrman. 1978. Iron, an essential element for biosynthesis of aromatic compounds. Proc. Nat. Acad. Sci. 75:4810-4813. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 93 Waring, W.S., and C.H. Werkman. 1943. Growth of bacteria in an iron-free medium. Arch. Biochem. 1:303-310. Webley, D.M., R.B. Duff, and G. Anderson. 1962. The metabolism of iron-, zinc-, and manganese- deficient Nocardia opaca. J. Gen. Micro. 29: 179-187. Healy, W.B., S.C. Cheng, and W.D. McElroy. 1955. Metal toxicity and iron deficiency effects on enzymes in Neurospora. Arch. Biochem. Biophys. 54:206-214. Light, P A., and P.B. Garland. 1971. A comparison of mitochondria from Torulopsis utilis grown in con- tinuous culture with glycerol, iron, ammonium, magnesium, or phosphate as the growth-limiting nutrient. Biochem. J. 124:123-134. Metzler, D.E., M. Ikawa, and E.E. Snell. 1954. A general mechanism for vitamin B5-catalyzed reactions. J. Am. Chem. Soc. 76:648-652. Snell, E.E. 1958. Chemical structure in relation to biological activities of vitamin Be. Vitamins and Hormones 16:77-125. Patwardhan, M.V. 1958. Role of iron in enzymatic transaminations. Nature 181:187. McCullough, W.G., J.T. Piligran, and I.J. Daniel. 1957. Enzymatic decarboxylation of the aminoben- zoates. J. Am. Chem. Soc. 79:628-630. Guirard, B.M. and E.E. Snell. 1954. Pyridoxal phos- phate and metal ions as cofactors for histidine decarboxylase. J. Am. Chem. Soc. 76:4745-4746. Young, I.G., and F. Gibson. 1969. Regulation of the enzymes involved in the biosynthesis of 2,3- dihydroxybenzoic acid in Acrobacter acrogenes and E. coli. Biochim. Biophys. Acta 177:401—411. Young, I.G., G.B. Cox, and F. Gibson. 1967. 2,3-dihy- droxybenzoate as a bacterial growth factor and its route of biosynthesis. Biochim. Biophys. Acta 141:319-331. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 94 O'Brien, I.G., G.B. Cox, and F. Gibson. 1971. Enterochelin hydrolysis and iron metabolism in E. coli. Biochim. Biophys. Acta 237:537-549. Brot, N., and J. Goodwin. 1968. Regulation of 2,3-dihydroxybenzoylserine synthetase by iron. J. Biol. Chem. 243:510-513. Lascelles, J., and T.P. Hatch. 1969. Bacteriochloro- phyll and heme biosynthesis in Rhodopseudomonas spheroidcs: possible role of heme in regulation of the branched biosynthetic pathway. J. Bact. 98:712-720. Garrard, W.T. 1972. Synthesis, assembly, and locali- zation of periplasmic cytochrome c. J. Biol. Chem. 247:5935-5943. Gross, M., and M. Rabinovitz. 1972. Control of globin synthesis by hemin: factors influencing formation of an inhibitor of globin chain initiation in rzticulocyte lysates. Biochim. Biophys. Acta 287: 3 0-352. Datta, A., C. deHaro, J.M. Sierra, and S. Ochoa. 1977. Mechanism of translational control by hemin in reticulocyte lysates. Proc. Nat. Acad. Sci. 74: 3326-3329. Jacobs, N.J., and S.F. Conti. 1965. Effect of hemin on the formation of the cytochrome system of anaerobically-grown Staphylococcus epidermidis. J. Bact. 89:675-679. McConville, M.L., and H.P. Charles. 1979. Mutants of E. coli K12 permeable to hemin. J. Gen. Micro. 113:165-168. Brown, E.G. 1958. Mode of action of 5,6-dimethyl- benzimidazole and certain keto acids as inhibitors of haem biosynthesis. Nature 182:1091-1092. Brown, E.G. 1958. Evidence for the involvement of ferrous iron in the biosynthesis of ALA by chicken erythrocyte preparations. Nature 182:313- 315. Vogel,W., D.A. Richert, B.Q. Pixley, and M.P. Schulman. 1960. Heme synthesis in iron-deficient duck blood. J. Biol. Chem. 235:1769-1775. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 95 Stein, J.A., D.P. Tschudy, P.L. Corcoran, and A. Collins. 1970. 6-aminolevulinic acid synthetase III. Synergistic effect of chelated iron on induction. J. Biol. Chem. 245:2213-2218. Marsh, H.V., H.J. Evans, and G. Matrone. 1960. Investigations of the role of iron in chlorophyll metabolism. II. Effect of iron deficiency on chlorophyll synthesis. Plant Physiol. 38:638-642. Tait, G.H. 1973. Aminolaevulinate synthetase in Micrococcus denitrificans. Purification and prop- erties of the enzyme, and the effect of growth conditions on the enzyme activity in cells. Biochem. J. 131:389-403. Chandrika, S.R., and G. Padmanaban. 1977. Site of action of iron in the induction of ALAD in Neurospora crassa. Biochem. Biophys. Res. Comm. 77:35-41. Lascelles, J. 1964. Tetrapyrrole Biosynthesis and its Regulation. W.A. Benjamin Co.; New York. Sano, S., and S. Granick. 1961. Mitochondrial copro- porphyrinogen oxidase and protoporphyrin formation. J. Biol. Chem. 236:1173-1180. Hsu, W.P. and G.W. Miller. 1970. Coproporphyrinogen- ase in tobacco (Nicotiana tabacum L.). Biochem. J. 117:215-220. Mori, M., and S. Sano. 1968. Protoporphyrin formation from coproporphyrinogen III by Chromatium cell extracts. Biochem. Biophys. Res. Comm. 32: 610-615. Tait, G.H. 1969. Coproporphyrinogenase activity in extracts from Rhodopseudomonas spheroides. Biochem. Biophys. Res. Comm. 37:116-122. Jacobs, N.J., J.M. Jacobs, and P. Brent. 1971. Char- acterization of the late steps of microbial heme synthesis: conversion of coproporphyrinogen to protoporphyrin. J. Bact. 107 203-209. Korstee, G.J.J. 1970. Iron deficiency and porphyrin formation by Arthrobactcr globiformis. Antonie van Leeuwenhoek J. Microbial Serol. 36:579-580. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 96 Doss, M., and B. Ulshofer. 1971. Porphyrin stability as a function of the number of carboxylic acid side chains. Biochim. Biophys. Acta 237:356-360. Chu, T.C., and E.J. Chu. 1966. Thin layer chromato- graphy of porphyrins, chlorins, and related com- pounds on Eastman "Chromagram" sheets. J. Chroma- tography 21:46-51. Mauzerall, D., and S. Granick. 1956. The occurrence and determination of ALA and porphobilinogen in urine. J. Biol. Chem. 219:435-446. Urata, C., and S. Granick. 1963. Biosynthesis of a-aminoketones and the metabolism of aminoacetone. J. Biol. Chem. 238:811-820. Lipmann, F., and L.C. Tuttle. 1945. A specific micro- method for the determination of acyl phosphates. J. Biol. Chem. 159:21-28. Bradford, M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principal of protein-dye binding. Anal. Biochem. 72:248-254. Appleby, C.A. 1969. Electron transport systems of Rhizobium japonicum. II. Rhizobium hemoglobin, cytochromes, and oxidases in free-living cells. Biochim. Biophys. Acta 172:88-105. Cheng, J.B., and J. Lascelles. 1963. Nitrate reductase in cell-free extracts of a hemin-requiring strain of Staphylococcus aureus. Biochem. J. 89:503-510. Haddock, B.A., and H.U. Schairer. 1973. Electron transport chains of E. coli. Eur. J. Biochem. 35:34-45. Clark-Walker, G.D., B. Rittenberg, and J. Lascelles. 1967. Cytochrome synthesis and its regulation in Spirillum itersonii. J. Bact. 94:1648-1655. Cutting, J.A., and H.M. Schulman. 1972. The control of heme synthesis in soybean root nodules. Biochim. Biophys. Acta 261 321-327. Avissar, Y., and K.D. Nadler. 1978. Stimulation of tetrapyrrole formation in Rhizobium japonicum by restricted aeration. J. Bact. 135:782-789.