Eur. J . Biochem. 59. 73-77 (1975)
5-Aminolevulinic-Acid Synthetases from Rhodopseudomonas spheroides Y Comparison of the Purification and Properties of Enzymes Extracted from Bacteria Grown in Different Iron Concentrations Jcnny D. CLEMENT-METRAL and Michele FANICA-GAIGNIER Labordtoire de Photosynthese, Centre National de la Recherche Scientifique, Gif-sur-Yvette (Received April 11 /August 1, 1975)
The two 5-aminolevulinic acid synthetases of Rhodopseudomonas spheroides Y. were extracted from cells grown in a ‘low-iron’ medium and purified. They have a specific activity 10-fold higher than the ‘high-iron’ enzymes described by us previously and have the same properties except that they do not contain any iron and have one free - SH group more per mole of enzyme (2 for E, ; 8 for E2). Their inhibition by adenosine triphosphate and iron and their oxidation-reduction sensitivity are discussed in terms of light, oxygen and heme feed-back regulation of bacteriochlorophyll synthesis.
5-Aminolevulinic acid synthetase has been detected in relatively few organisms and only three isolations of pure enzyme have so far been reported [l - 31. Some of the most active sources are the photosynthetic bacteria of the Athiorhodacae. Rhodopseudomonas spheroides (Van Niel collection [4] or strain Y [ 5 ] ) contains two different isofunctional enzymes which have been purified to homogeneity and characterized [6,7]. Although the pure 5-aminolevulinic acid synthetases of the mammalian or bacterial sources share in common most of their properties [8], the two enzymes isolated from Rps spheroides Y had a much higher specific activity compared to the others. Two explanations could be discarded: inhomogeneity of the proteins and difference in the assay conditions. A possible explanation for this low specific activity might be that we used a 17 pM iron growth medium whereas in earlier studies of this enzyme, animals (anemic rabbits) or bacteria (grown in an iron-deprived medium) were severely deprived of iron. This work was undertaken to find out if iron, a well-known inhibitor of 5-aminolevulinic acid synthetase, was the cause of the apparent inactivation and to compare the purification and properties of the enzymes grown in high-iron or low-iron medium. Previous papers in this series appeared in this journal [1,6]. Abbreviarion. Rps, Rhodopseudomonas. Enzyme. 5-Aminolevulinic acid synthetase (EC 2.3.1.37).
MATERIALS AND METHODS Rps spheroides Y was grown photosynthetically in either medium ‘021’ described by Reiss et al. [9] in which the iron concentration is 1 pM or in ‘021 Fe’ in which iron concentration is 17pM. Cells were harvested during late exponential phase, resuspended and washed twice in 0.05 M phosphate buffer pH 7.4 and stored at - 15 “Cuntil used. Purification and assay of 5-aminolevulinic acid synthetase activity were carried out as previously described [1,5]. - SH groups [ l ] and porphyrins [lo] were estimated as described previously. Hemin was prepared following the method of Karibian and London [ll]. Iron was assayed according to Doeg and Ziegler [12] using bathophenanthroline as a reagent.
+
RESULTS Comparison of the Purification of High-Iron and Low-Iron 5-Aminolevulinic Acid Synthetases
The results of the purification of 5-aminolevulinic acid synthetase extracted from Rps spheroides Y grown in low-iron medium are summarized in Table 3 and compared to the results of the purification obtained from cells grown in high-iron medium in Table 2. The two different isofunctional enzymes, in both cases, are separated only upon the DEAE-Sephadex
5-Aminolevulinic-Acid Synthetases from Rps spheroides Y
71
.I ,ibk 1. Purifi'cation of5-aminolevulinic acid synihetase E, und E,, ( I und 11) extracted from low-iron-grow cells The activity was assayed by determining the amount of 5-aminolevulinic acid formed at 22 "C after 15 min in a reaction mixture (a) containing 400 pmol Tris bufler pH 7.8, 600 nmol succinyl CoA, 200 pmol glycine, 0.2 pmol pyridoxal phosphate and enzyme extract in a final volume of 1 ml. The reaction was terminated by 1 ml of 10% trichloroacetic acid and 5-aminolevulinic acid formed determined by modified Ehrlich reagent [13]; alternatively the formation of CoASH was determined in the same reaction mixture (a) containing 0.2 pmol of 5,5'-dithiobis(2nitrobenzoic acid) measured at 412 nm ( E = 13600 M-' cm-') [14]. 1 unit = 1 pmol 5-aminolevulinic acid or CoASH formed per min ~~
~
Fraction
~~
Volume I
I1
ml
Protein concentration
Total proteins
I
I
I1
mdml
I1
mg
Specific activity
Total activity
Purification factor
I
I
I
I1
units/mg protein
11
units
11
-fold
.-
Crude extract Spinco 110 Dialysed 12 fraction Sephadex 46 c-200 DEAEScphadex 20 22 Electro6 5 focusing Sephadex9 22 G-200
2.7
297
0.0008
0.24
1
5.2
62.4
0.0036
0 22
5
1.2
55.2
0.03
1.66
40
0.2
0.1
0.2
0.03
0.01
0.005
2.2
0.15
0.2
0.60
0.44
200
270
1.2
0.15
0.36
0.57
0.43
0.08
500
740
0.09
0.11
2.6
1.9
0.23
0.20
3500
2500
4.0
Table 2. Comparison ofthe purification of 5-aminolevulinir acid synthetasrs I and 11 extracted from cells grown in high-iron (Fe') or lowiron ( F P - ) medium 1 unit = 1 pmol of 5-aminolevulinic acid formed per min -~
Fraction
Protein concentration
Specific activity
Fe'
Fe'
Fe-
I
11
I
I1
mdml Crude extract Spinco 2.3 Dialysis 4.1 Scphadex (3-200 1.5 DEAESephadex 0.06 Electrofocusing 0.075 Sephadex G-200 0.022
Fe-
I1
I
+
Fe'
I
I1
units/mg protein
Fe-
I1
I
I
I1
-fold
2.7 5.2
0.0001 0.0011
0.0008 0.0036
1 10
5
1.2
0.0057
0.03
57
40
1
0.15
0.2
0.1
0.027
0.02
0.15
0.2
270
200
200
270
0.15
0.2
0.03
0.05
0.047
0.36
0.57
500
470
500
470
0.03
0.01
0.005
0.37
0.17
2.6
1.9
3700
1700
3500
2500
chromatography; they are not differentiated before and the activities in the three first steps correspond to the sum E, E,,. Table 2 shows that the specific activity for the lowiron enzymes at the beginning of the purification is ten times that obtained with high-iron growth medium for El E,,. This enhancement is kept all along the purification and the pure enzymes (homogenous in disc electrophoresis, for details see [l]) are also ten
+
Purification factor
times more active. The purification is almost the same, purification factors are similar and the activation obtained by dialysis and Sephadex G-200 filtration is also observed. ofTheir Properties
The molecular weight of both species (I and 11) is 100000 20000 by filtration on Sephadex G-200. The
J. D. Clement-Metral and M.Fanica-Gaignier
Enzyme
A
loo r
Table 3. Tirrution of - S H groups in enzyme e.rtructed,frorn cells grown in high or /OM.iron Amount of protcin = 1 rng
- S H groups .~
high-iron
low-iron
rno1-l -.
~
A
El Ell
1.2 f 0.3 7.1 f 0.3
2.2 f 0.3 8.1 k 0.3
isoelectric points are 5.1 for El, 6.0 for Ell. The absorption spectra maxima corresponding to the prosthetic group: pyridoxal5'-phosphate (maxima at 330 and 415 nm at pH 6.5 [l]) are the same for enzymes I and 11. These properties are identical for cells grown in high or low iron. The -SH titration of native low-iron 5-aminolevulinic acid synthetases gives one - SH group more than in high-iron enzymes (Table 3). Dosage of Fe on pure enzymes extracted from low-iron medium shows that within the limits of experimental error they are free of Fe, whereas the same dosage made on highiron enzymes, which retain only 10% of the activity, gives about one atom Fe per mole of enzyme. The addition to the iron-free enzymes of increasing amounts of FeCl,, at room temperature and pH 7.5 (Fig. 1 A), shows that activity of El as well as E,, is inhibited to a constant level when one mole of FeCl, is bound per mole of enzyme. With an excess of FeCl, (0.1 mM) the inhibition becomes total. Addition of imidazole (0.2 mM) to the enzymes prior to FeCl, prevents the inhibition. As was shown by Porra et al. [15], 0.2 mM imidazole cannot reverse the inhibition when added after the enzyme preparation and iron had been mixed. A similar observation had already been made by Keilin and Hartee [16] in their work on succinate dehydrogenase. Porra et al. suggested that the inhibition of 5aminolevulinic acid synthetase by hemin in cell-free extracts of Rps spheroides was due to the formation, through its iron atom, of a coordination complex with the enzymes. As with FeCl,, iron-free 5-aminolevulinic acid synthetases are inhibited (Fig. 1 B) to a constant level when one mole of hemin is fixed to one mole of enzyme. Inhibition is partially prevented by previous addition of imidazole. The inhibition by hemin is allosteric and therefore differs from FeCl, inhibition. As was shown by Yubisui and Yoneyama [17] with a partially purified preparation, protoporphyrin IX is also an inhibitor of 5-aminolevulinic acid synthetase. 0.1 pM protoporphyrin IX inhibits totally the two pure enzymes. Hemin must then inhibit 5-aminolevulinic acid synthetases by binding on two different sites, one of them involving a complex with iron.
c
w
of
I
I
1
,
1
1
I
I
2 FeC13 (rnol/mol enzyme)
I
2
p........+o
[.......... j d o
Hemin (mol I mol enzyme)
Fig. 1. Inhibition of 5-aminolevulinic ucid syntlietuse uctivitj h j I A ) iron and ( B ) hemin. For clarity El only is shown, the same phenomenon is observed with Ell.(A) FeCI, added to 0.5 mg enzyme (molecular weight 100000) in the absence (I) or presence (11) of 0.2 mM irnidazole. (B) Hemin added to 0.5 rng enzyme in the absence (I) or presence (11) of 0.2 mM imidazole. 1 unit = 1 prnol/h
Control of Low-Iron Enzyme Activity by Adenosine Triphosphate
An important property of high-iron Rps spheroides Y 5-aminolevulinic acid synthetases is their inhibition by ATP or pyrophosphate in vivo [lo] as well as in vitro [5,6]. ADP, A M P or phosphate have no effect. We have previously shown [lo] that, in an ironcontaining medium, exogenous 1 mM adenosine triphosphate inhibits bacteriochlorophyll synthesis and porphyrin excretion, without changing the rate of growth. This is a direct effect of ATP as Gajdos et al. [18] using double-labelled ATP (32P and 14C) have shown that the intact ATP molecule enters Rps spheroides cells. The ATP inhibition ofbacteriochlorophyll synthesis is prevented by addition of 5-aminolevulinc acid in the medium, which suggests that ATP affects the first stage of bacteriochlorophyll synthesis, i . e. at the enzymic level of 5-aminolevulinic acid synthetdse. Identical results are obtained in low-iron medium. Table 4 shows that, in vitro, both enzymes I and I1 exhibit similar behaviour in the presence of adenosine triphosphate and that the extent of inhibition is the same as with high-iron enzymes. For that reason the whole kinetic study was not undertaken and the results obtained with high-iron enzymes [6] (inhibition con-
76
5-Aminolevulinic-Acid Synthetases from Rps spheroides Y
Table 4. Inhibition of 5-aminolevulinic acid synthetases activity by adenosine triphosphate These results were obtained at the electrofocusing step. 1 unit = 1 pmol formed per h Enzyme
Specific activity ~
+ 1mM ATP
alone units/mg protein
Inhibition by ATP
Table 5. Fixation of adenosine triphosphate after complete inhibition of - S H groups by potassium ietrafhionate or sodium arsenite binding on pure enzyme I Measurements were made on enzyme El after blocking of all - SH groups by the inhibitor. Inhibitor concentration was 1 mM Inhibitor
"/,
Inhibition of activity
ATP fixed
x
mol/mol enzyme
90 100
1.2 0.9
~~
El
22
4.4
80
EII
34
9.5
15
stants, type of inhibition, lack of inhibition with dinucleotides or mononucleotides) assumed to be valid in the case of low-iron enzymes. However the fixation of ATP to the protein was studied. This was done only with pure enzyme I as previously described [6]. Pure 5-aminolevulinic acid synthetase I was incubated with adenosine ["PI triphosphate (15 mol/ mol enzyme in 0.1 M Tris-HC1 pH 8 for 30 min at room temperature), then filtered on Sephadex G-200 using 0.1 M Tris pH 8 [39]. Fractions were assayed for protein by Lowry's procedure [20] and compared to the pure enzyme protein standard curve. "P was determined in 10ml Bray's solution in a Packard scintillation Tri-carb counter [21]. The isolated radioactive complex contained 1 mol of adenosine ["PI triphosphate per mol of enzyme; the complex was almost inactive. As the incubation with [a-32P]ATP or [y32P]ATPgives a radioactive 32P-labeledenzyme complex, it is probable that ATP is entirely bound to the enzyme. Experiments with [14C]ATP are necessary to confirm this hypothesis. It is well known that 5-aminolevulinic acid synthetase contains an -SH group at its active center [22,1] and that the active forms are those that are reduced [l].Fixation of ATP masks one - SH group: no free -SH groups are titratable in E, form highiron-grown cells [6], and only one free - SH per molecule of E, in low-iron-grown cells (as compared to two in the active enzyme, which does not contain iron, Table 3). Denaturation by 0.5 % sodium dodecylsulfate, 4 or 8 M urea does not unmask any more - SH groups in the active enzyme or in the ATP . enzyme complex. After incubation with p-hydroxymercuribenzoate, which blocks - SH groups, no ATP fixation was obtained. These results seemed to imply that an -SH group is involved in the ATP linkage with 5aminolevulinic acid synthetase. However p-hydroxymercuribenzoate is known to induce conformational changes in proteins [23]; it could also cause steric obstruction and prevent ATP from reaching the inhibitor site on the enzyme. The fixation of ATP on the enzyme blocked by other -SH inhibitors was thus investigated in the case of low-iron enzymes. Table 5 shows that a free - SH group is not implicated in the
Sodium arsenite Potassium tetrathionate
0.3 0.2
ATP fixation. The completely - SH-blocked inactivated enzyme still fixes one mole of ATP. Masking of - SH with ATP fixation is thus probably due to steric hindrance or to reaction between the ATP . enzyme and the dye used in the - SH titration: nitrothiophenol absorbing at 412 nm [14] which is liberated from 5,5'-dithiobis(2-nitrobenzoicacid) after reaction with a free -SH. One molecule of nitrothiophenol instead of being estimated could react with the ATP . enzyme, liberate ATP and form nitrothiophenol . enzyme complex. Kitagawa et al. [24] have indeed shown that this reaction occurs in myosin, resulting in the masking of an -SH group after fixation of ATP; masking was insensitive to urea.
DISCUSSION The non-sulphur bacterium Rps spheroides can grow anaerobically in the light, or aerobically in both light and dark. Light intensity, medium composition and oxygen tension all play important roles in determining the extent of synthesis of the photosynthetic pigment bacteriochlorophyll [25]. The importance of the first enzyme of the biosynthetic path, 5-aminolevulinic acid synthetase, as a main regulation point in the bacteriochlorophyll synthesis is well known from investigations by Lascelles, Neuberger, Tuboi and their coworkers. Besides repression and derepression, the production of 5-aminolevulinicacid synthetase is regulated by modification of the enzyme activity. We have been working on the purification of 5-aminolevulinic acid synthetases from Rps spheroides in the belief that a full understanding of the control of bacteriochlorophyll synthesis will only be possible when it is available in pure form. Most of the studies on regulation of bacteriochlorophyll synthesis were made using Rps spheroides grown in the so-called 'S' medium of Lascelles [26] deprived of iron. Our choice of an iron-containing medium which we thought more physiological, was a good one in that the enzymes extracted from cells grown in such conditions are more stable ; however their specific activity was low compared to that of the single 5-amino-
J. D. Clement-Mctral and M. Fanica-Gaignier
levulinic acid synthetase isolated by Warnick and Burnham [ 3 ] from Rps spheroides grown in S medium (2.1 units/mg protein). The specific activity of the 5aminolevulinic acid synthetases grown in low-iron medium obtained in this work is high and similar to Warnick’s value (2.2 units/mg protein). Their properties are the same as that of the previously described enzymes [l ] except that they donot contain iron, whereas the enzymes with low activity contain about 1 mole per mole of enzyme, and they have one free - S H group more per mole than the iron-containing enzymes (2 for E,; 8 for EJ. These results are in agreement with those of Yubisui and Yoneyama [I71 who found that 2-mercaptoethanol reversed to some extent the iron inhibition of a partially purified preparation. Hemin is well known as a feed-back regulator of 5-aminolevulinic acid synthetase [27], bound to the enzyme on two different sites, one of which is through its iron atom [15]. Although Rps spheroides 5-aminolevulinic acid synthetases do not seem to fit the general model of allosteric enzymes, binding of the small ligand iron could be an interesting way of regulation in the case where free heme does not accumulate, since free pigments do not accumulate to any obvious extent in nature. Besides this important possible role, iron has perhaps a stabilizing effect on these very oxidation/ reduction-sensitive enzymes, protecting the activecenter - SH group [l] from oxidation. The alterations in the rate of bacteriochlorophyll which occur in cells when the light intensity is changed are too rapid to be accounted for by induction or repression of enzyme synthesis. All the results of this and previous work [5,6,10,28] are consistent with the view that it has to be correlated with the fast changes in cellular adenosine triphosphate levels with light intensity [29,30]. The inhibition of bacteriochlorophyll synthesis in vivo by exogenous adenosine triphosphate in high [lo] and low-iron medium is probably due to the inhibition of 5-aminolevulinic acid synthetase as adenosine triphosphate inhibits the pure enzyme by forming a complex. The very specific inhibition of 5-aminolevulinic acid synthetases by adenosine triphosphate or pyrophosphate [6] is consistent with the interpretation that adenosine triphosphate is perhaps the primary agent through which light intensity controls bacteriochlorophyll production in Rps spheroides Y.
77 We are grateful to Dr B. Diner for critically reading our manuscript and to Mrs D. Clerot for skilled technical assistance.
REFERENCES 1 . Fanica-Gaignier, M. & Clement-Metral, J. D. (1973) Eur. J.
Biochem. 40, 13- 18. 2. Aoki, Y., Wada, 0.. Urata, G., Takaku, F. & Nakao, K. (1971) Biochem. Biophys. Res. Comrnun. 42, 568- 515. 3. Warnick, G . R. & Burnham, 9. F. (1971) J . Biol. Chem. 246, 6880-6885. 4. Tuboi, S., Kim, H. J. & Kikuchi, G. (1970) Arch. Biochem. Biophys. 138, 141- 154. 5. Fanica-Gaignier, M. & Clement-Metral, J. D. (1971) Biochc,m. Biophys. Res. Commun. 44, 192- 198. 6. Fanica-Gaignier, M. & Clement-Metral, J. D. (1973) Eur. J . Biochem. 40, 19-24. 7. Reference deleted. 8. Jordan, P. M. & Shemin, D. (1912) in The Enzymes, 3rd edn (Boyer,P. D., ed.) pp. 339-356, Academic Press, New York. 9. Reiss-Husson, F., De Klerk, G., Jolchine, G., Jauneau, E. & Karnen. M. D. (1971) Biochim. Biophys. Acta. 234, 73-82. 10. Fanica-Gaignier, M. (1 971) IInd lnternafional Congress on Photosynthesis (Forti, G., Avron, M. & Melandri, A,, eds) vol. 111, pp. 2721-2726, Dr Junk,W., N.V. Publishers, The Hague. 11. Karibian, D. & London, I. M. (1965) Biochem. Biophys. Res. Commun. 18,243-249. 12. Doeg, K. A. & Ziegler, D. M. (1962) Arch. Biochem. Biophys. 97, 37 -40. 13. Urata, G. & Granick, S . (1963) J. Biol. Chem. 238, 811-820. 14. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 15. Porra, R. J., Irving, E. A. & Tennick, A. M. (1972) Arch. Biochem. Biopli.vs. 148, 37-43. 16. Keilin, D. & Hartree, E. F. (1947) Biochem. J . 41, 503- 506. 17. Yubisui, T. & Yoneyama, U. (1972) Arch. Biochern. Biophys. 150, 17- 85. 18. Gajdos, A., Gajdos-Torok, M., Gorchein, A , , Neuberger, A. & Tait, G. H. (1968) Biochem. J . 106, 185. 19. Das, N., Cottam, G . L. & Srere. P. A. (1971) Arch. Biochez. Biophys. 143, 602- 608. 20. Lowry, 0. H., Rosebrough, N. J., Farr, A . L. & Randall, R. J. (1951) J . Biol. Chem. 193,265-215. 21. Bray, G. A. (1960) Anal. Biochem. 1, 279-285. 22. Scholnick, P. L., Hammaker, L. E. & Marver, H. S. (1972) J . Biol. Chem. 247, 4132. 23. Gibson, Q.H. (1973) J . Biol. Chem. 248, 1281- 1284. 24. Kitagawa, S . , Chiang, K. K. & Tonomura, Y . (1964) Biochim. Biopliys. Acta, 82, 83 -91. 25. Cohen-Bazire, G., Sistrom, W. R. & Stanier, R. Y . (1957) J. Cell Comp. Physiol. 49,25 - 68. 26. Lascelles, J. (1956) Biochem. J. 62, 78-93. 27. Laxelles, J. (1968) Biochem. SOC.Symp. 28, 49-59. 28. Tuboi, S.,Kim, H. J. & Kikuchi, G. (1970) Arch. Biochem. Biophys. 138, 155- 159. 29. Welsch, F . & Smith, L. (1969) Biochemistry, 8,3403-3408. 30. Fdnica-Gaignier, M., Clement-Metral, J. D. & Kamen, M. D. (1971) Biochim. Biophys. Acta, 226, 135- 143.
J. D. Clement-Metral, Laboratoire de PhotosynthZse du C.N.R.S., F-91190 Gif-sur-Yvette, France M. Fanica-Gdignier, Laboratoire d’Enzymologie du C.N.R.S. F-91190 Gif-sur-Yvette, France
5-Aminolevulinic-Acid Synthetases from Rhodopseudomonas spheroides Y Comparison of the Purification and Properties of Enzymes Extracted from Bacteria Grown in Different Iron Concentrations Jcnny D. CLEMENT-METRAL and Michele FANICA-GAIGNIER Labordtoire de Photosynthese, Centre National de la Recherche Scientifique, Gif-sur-Yvette (Received April 11 /August 1, 1975)
The two 5-aminolevulinic acid synthetases of Rhodopseudomonas spheroides Y. were extracted from cells grown in a ‘low-iron’ medium and purified. They have a specific activity 10-fold higher than the ‘high-iron’ enzymes described by us previously and have the same properties except that they do not contain any iron and have one free - SH group more per mole of enzyme (2 for E, ; 8 for E2). Their inhibition by adenosine triphosphate and iron and their oxidation-reduction sensitivity are discussed in terms of light, oxygen and heme feed-back regulation of bacteriochlorophyll synthesis.
5-Aminolevulinic acid synthetase has been detected in relatively few organisms and only three isolations of pure enzyme have so far been reported [l - 31. Some of the most active sources are the photosynthetic bacteria of the Athiorhodacae. Rhodopseudomonas spheroides (Van Niel collection [4] or strain Y [ 5 ] ) contains two different isofunctional enzymes which have been purified to homogeneity and characterized [6,7]. Although the pure 5-aminolevulinic acid synthetases of the mammalian or bacterial sources share in common most of their properties [8], the two enzymes isolated from Rps spheroides Y had a much higher specific activity compared to the others. Two explanations could be discarded: inhomogeneity of the proteins and difference in the assay conditions. A possible explanation for this low specific activity might be that we used a 17 pM iron growth medium whereas in earlier studies of this enzyme, animals (anemic rabbits) or bacteria (grown in an iron-deprived medium) were severely deprived of iron. This work was undertaken to find out if iron, a well-known inhibitor of 5-aminolevulinic acid synthetase, was the cause of the apparent inactivation and to compare the purification and properties of the enzymes grown in high-iron or low-iron medium. Previous papers in this series appeared in this journal [1,6]. Abbreviarion. Rps, Rhodopseudomonas. Enzyme. 5-Aminolevulinic acid synthetase (EC 2.3.1.37).
MATERIALS AND METHODS Rps spheroides Y was grown photosynthetically in either medium ‘021’ described by Reiss et al. [9] in which the iron concentration is 1 pM or in ‘021 Fe’ in which iron concentration is 17pM. Cells were harvested during late exponential phase, resuspended and washed twice in 0.05 M phosphate buffer pH 7.4 and stored at - 15 “Cuntil used. Purification and assay of 5-aminolevulinic acid synthetase activity were carried out as previously described [1,5]. - SH groups [ l ] and porphyrins [lo] were estimated as described previously. Hemin was prepared following the method of Karibian and London [ll]. Iron was assayed according to Doeg and Ziegler [12] using bathophenanthroline as a reagent.
+
RESULTS Comparison of the Purification of High-Iron and Low-Iron 5-Aminolevulinic Acid Synthetases
The results of the purification of 5-aminolevulinic acid synthetase extracted from Rps spheroides Y grown in low-iron medium are summarized in Table 3 and compared to the results of the purification obtained from cells grown in high-iron medium in Table 2. The two different isofunctional enzymes, in both cases, are separated only upon the DEAE-Sephadex
5-Aminolevulinic-Acid Synthetases from Rps spheroides Y
71
.I ,ibk 1. Purifi'cation of5-aminolevulinic acid synihetase E, und E,, ( I und 11) extracted from low-iron-grow cells The activity was assayed by determining the amount of 5-aminolevulinic acid formed at 22 "C after 15 min in a reaction mixture (a) containing 400 pmol Tris bufler pH 7.8, 600 nmol succinyl CoA, 200 pmol glycine, 0.2 pmol pyridoxal phosphate and enzyme extract in a final volume of 1 ml. The reaction was terminated by 1 ml of 10% trichloroacetic acid and 5-aminolevulinic acid formed determined by modified Ehrlich reagent [13]; alternatively the formation of CoASH was determined in the same reaction mixture (a) containing 0.2 pmol of 5,5'-dithiobis(2nitrobenzoic acid) measured at 412 nm ( E = 13600 M-' cm-') [14]. 1 unit = 1 pmol 5-aminolevulinic acid or CoASH formed per min ~~
~
Fraction
~~
Volume I
I1
ml
Protein concentration
Total proteins
I
I
I1
mdml
I1
mg
Specific activity
Total activity
Purification factor
I
I
I
I1
units/mg protein
11
units
11
-fold
.-
Crude extract Spinco 110 Dialysed 12 fraction Sephadex 46 c-200 DEAEScphadex 20 22 Electro6 5 focusing Sephadex9 22 G-200
2.7
297
0.0008
0.24
1
5.2
62.4
0.0036
0 22
5
1.2
55.2
0.03
1.66
40
0.2
0.1
0.2
0.03
0.01
0.005
2.2
0.15
0.2
0.60
0.44
200
270
1.2
0.15
0.36
0.57
0.43
0.08
500
740
0.09
0.11
2.6
1.9
0.23
0.20
3500
2500
4.0
Table 2. Comparison ofthe purification of 5-aminolevulinir acid synthetasrs I and 11 extracted from cells grown in high-iron (Fe') or lowiron ( F P - ) medium 1 unit = 1 pmol of 5-aminolevulinic acid formed per min -~
Fraction
Protein concentration
Specific activity
Fe'
Fe'
Fe-
I
11
I
I1
mdml Crude extract Spinco 2.3 Dialysis 4.1 Scphadex (3-200 1.5 DEAESephadex 0.06 Electrofocusing 0.075 Sephadex G-200 0.022
Fe-
I1
I
+
Fe'
I
I1
units/mg protein
Fe-
I1
I
I
I1
-fold
2.7 5.2
0.0001 0.0011
0.0008 0.0036
1 10
5
1.2
0.0057
0.03
57
40
1
0.15
0.2
0.1
0.027
0.02
0.15
0.2
270
200
200
270
0.15
0.2
0.03
0.05
0.047
0.36
0.57
500
470
500
470
0.03
0.01
0.005
0.37
0.17
2.6
1.9
3700
1700
3500
2500
chromatography; they are not differentiated before and the activities in the three first steps correspond to the sum E, E,,. Table 2 shows that the specific activity for the lowiron enzymes at the beginning of the purification is ten times that obtained with high-iron growth medium for El E,,. This enhancement is kept all along the purification and the pure enzymes (homogenous in disc electrophoresis, for details see [l]) are also ten
+
Purification factor
times more active. The purification is almost the same, purification factors are similar and the activation obtained by dialysis and Sephadex G-200 filtration is also observed. ofTheir Properties
The molecular weight of both species (I and 11) is 100000 20000 by filtration on Sephadex G-200. The
J. D. Clement-Metral and M.Fanica-Gaignier
Enzyme
A
loo r
Table 3. Tirrution of - S H groups in enzyme e.rtructed,frorn cells grown in high or /OM.iron Amount of protcin = 1 rng
- S H groups .~
high-iron
low-iron
rno1-l -.
~
A
El Ell
1.2 f 0.3 7.1 f 0.3
2.2 f 0.3 8.1 k 0.3
isoelectric points are 5.1 for El, 6.0 for Ell. The absorption spectra maxima corresponding to the prosthetic group: pyridoxal5'-phosphate (maxima at 330 and 415 nm at pH 6.5 [l]) are the same for enzymes I and 11. These properties are identical for cells grown in high or low iron. The -SH titration of native low-iron 5-aminolevulinic acid synthetases gives one - SH group more than in high-iron enzymes (Table 3). Dosage of Fe on pure enzymes extracted from low-iron medium shows that within the limits of experimental error they are free of Fe, whereas the same dosage made on highiron enzymes, which retain only 10% of the activity, gives about one atom Fe per mole of enzyme. The addition to the iron-free enzymes of increasing amounts of FeCl,, at room temperature and pH 7.5 (Fig. 1 A), shows that activity of El as well as E,, is inhibited to a constant level when one mole of FeCl, is bound per mole of enzyme. With an excess of FeCl, (0.1 mM) the inhibition becomes total. Addition of imidazole (0.2 mM) to the enzymes prior to FeCl, prevents the inhibition. As was shown by Porra et al. [15], 0.2 mM imidazole cannot reverse the inhibition when added after the enzyme preparation and iron had been mixed. A similar observation had already been made by Keilin and Hartee [16] in their work on succinate dehydrogenase. Porra et al. suggested that the inhibition of 5aminolevulinic acid synthetase by hemin in cell-free extracts of Rps spheroides was due to the formation, through its iron atom, of a coordination complex with the enzymes. As with FeCl,, iron-free 5-aminolevulinic acid synthetases are inhibited (Fig. 1 B) to a constant level when one mole of hemin is fixed to one mole of enzyme. Inhibition is partially prevented by previous addition of imidazole. The inhibition by hemin is allosteric and therefore differs from FeCl, inhibition. As was shown by Yubisui and Yoneyama [17] with a partially purified preparation, protoporphyrin IX is also an inhibitor of 5-aminolevulinic acid synthetase. 0.1 pM protoporphyrin IX inhibits totally the two pure enzymes. Hemin must then inhibit 5-aminolevulinic acid synthetases by binding on two different sites, one of them involving a complex with iron.
c
w
of
I
I
1
,
1
1
I
I
2 FeC13 (rnol/mol enzyme)
I
2
p........+o
[.......... j d o
Hemin (mol I mol enzyme)
Fig. 1. Inhibition of 5-aminolevulinic ucid syntlietuse uctivitj h j I A ) iron and ( B ) hemin. For clarity El only is shown, the same phenomenon is observed with Ell.(A) FeCI, added to 0.5 mg enzyme (molecular weight 100000) in the absence (I) or presence (11) of 0.2 mM irnidazole. (B) Hemin added to 0.5 rng enzyme in the absence (I) or presence (11) of 0.2 mM imidazole. 1 unit = 1 prnol/h
Control of Low-Iron Enzyme Activity by Adenosine Triphosphate
An important property of high-iron Rps spheroides Y 5-aminolevulinic acid synthetases is their inhibition by ATP or pyrophosphate in vivo [lo] as well as in vitro [5,6]. ADP, A M P or phosphate have no effect. We have previously shown [lo] that, in an ironcontaining medium, exogenous 1 mM adenosine triphosphate inhibits bacteriochlorophyll synthesis and porphyrin excretion, without changing the rate of growth. This is a direct effect of ATP as Gajdos et al. [18] using double-labelled ATP (32P and 14C) have shown that the intact ATP molecule enters Rps spheroides cells. The ATP inhibition ofbacteriochlorophyll synthesis is prevented by addition of 5-aminolevulinc acid in the medium, which suggests that ATP affects the first stage of bacteriochlorophyll synthesis, i . e. at the enzymic level of 5-aminolevulinic acid synthetdse. Identical results are obtained in low-iron medium. Table 4 shows that, in vitro, both enzymes I and I1 exhibit similar behaviour in the presence of adenosine triphosphate and that the extent of inhibition is the same as with high-iron enzymes. For that reason the whole kinetic study was not undertaken and the results obtained with high-iron enzymes [6] (inhibition con-
76
5-Aminolevulinic-Acid Synthetases from Rps spheroides Y
Table 4. Inhibition of 5-aminolevulinic acid synthetases activity by adenosine triphosphate These results were obtained at the electrofocusing step. 1 unit = 1 pmol formed per h Enzyme
Specific activity ~
+ 1mM ATP
alone units/mg protein
Inhibition by ATP
Table 5. Fixation of adenosine triphosphate after complete inhibition of - S H groups by potassium ietrafhionate or sodium arsenite binding on pure enzyme I Measurements were made on enzyme El after blocking of all - SH groups by the inhibitor. Inhibitor concentration was 1 mM Inhibitor
"/,
Inhibition of activity
ATP fixed
x
mol/mol enzyme
90 100
1.2 0.9
~~
El
22
4.4
80
EII
34
9.5
15
stants, type of inhibition, lack of inhibition with dinucleotides or mononucleotides) assumed to be valid in the case of low-iron enzymes. However the fixation of ATP to the protein was studied. This was done only with pure enzyme I as previously described [6]. Pure 5-aminolevulinic acid synthetase I was incubated with adenosine ["PI triphosphate (15 mol/ mol enzyme in 0.1 M Tris-HC1 pH 8 for 30 min at room temperature), then filtered on Sephadex G-200 using 0.1 M Tris pH 8 [39]. Fractions were assayed for protein by Lowry's procedure [20] and compared to the pure enzyme protein standard curve. "P was determined in 10ml Bray's solution in a Packard scintillation Tri-carb counter [21]. The isolated radioactive complex contained 1 mol of adenosine ["PI triphosphate per mol of enzyme; the complex was almost inactive. As the incubation with [a-32P]ATP or [y32P]ATPgives a radioactive 32P-labeledenzyme complex, it is probable that ATP is entirely bound to the enzyme. Experiments with [14C]ATP are necessary to confirm this hypothesis. It is well known that 5-aminolevulinic acid synthetase contains an -SH group at its active center [22,1] and that the active forms are those that are reduced [l].Fixation of ATP masks one - SH group: no free -SH groups are titratable in E, form highiron-grown cells [6], and only one free - SH per molecule of E, in low-iron-grown cells (as compared to two in the active enzyme, which does not contain iron, Table 3). Denaturation by 0.5 % sodium dodecylsulfate, 4 or 8 M urea does not unmask any more - SH groups in the active enzyme or in the ATP . enzyme complex. After incubation with p-hydroxymercuribenzoate, which blocks - SH groups, no ATP fixation was obtained. These results seemed to imply that an -SH group is involved in the ATP linkage with 5aminolevulinic acid synthetase. However p-hydroxymercuribenzoate is known to induce conformational changes in proteins [23]; it could also cause steric obstruction and prevent ATP from reaching the inhibitor site on the enzyme. The fixation of ATP on the enzyme blocked by other -SH inhibitors was thus investigated in the case of low-iron enzymes. Table 5 shows that a free - SH group is not implicated in the
Sodium arsenite Potassium tetrathionate
0.3 0.2
ATP fixation. The completely - SH-blocked inactivated enzyme still fixes one mole of ATP. Masking of - SH with ATP fixation is thus probably due to steric hindrance or to reaction between the ATP . enzyme and the dye used in the - SH titration: nitrothiophenol absorbing at 412 nm [14] which is liberated from 5,5'-dithiobis(2-nitrobenzoicacid) after reaction with a free -SH. One molecule of nitrothiophenol instead of being estimated could react with the ATP . enzyme, liberate ATP and form nitrothiophenol . enzyme complex. Kitagawa et al. [24] have indeed shown that this reaction occurs in myosin, resulting in the masking of an -SH group after fixation of ATP; masking was insensitive to urea.
DISCUSSION The non-sulphur bacterium Rps spheroides can grow anaerobically in the light, or aerobically in both light and dark. Light intensity, medium composition and oxygen tension all play important roles in determining the extent of synthesis of the photosynthetic pigment bacteriochlorophyll [25]. The importance of the first enzyme of the biosynthetic path, 5-aminolevulinic acid synthetase, as a main regulation point in the bacteriochlorophyll synthesis is well known from investigations by Lascelles, Neuberger, Tuboi and their coworkers. Besides repression and derepression, the production of 5-aminolevulinicacid synthetase is regulated by modification of the enzyme activity. We have been working on the purification of 5-aminolevulinic acid synthetases from Rps spheroides in the belief that a full understanding of the control of bacteriochlorophyll synthesis will only be possible when it is available in pure form. Most of the studies on regulation of bacteriochlorophyll synthesis were made using Rps spheroides grown in the so-called 'S' medium of Lascelles [26] deprived of iron. Our choice of an iron-containing medium which we thought more physiological, was a good one in that the enzymes extracted from cells grown in such conditions are more stable ; however their specific activity was low compared to that of the single 5-amino-
J. D. Clement-Mctral and M. Fanica-Gaignier
levulinic acid synthetase isolated by Warnick and Burnham [ 3 ] from Rps spheroides grown in S medium (2.1 units/mg protein). The specific activity of the 5aminolevulinic acid synthetases grown in low-iron medium obtained in this work is high and similar to Warnick’s value (2.2 units/mg protein). Their properties are the same as that of the previously described enzymes [l ] except that they donot contain iron, whereas the enzymes with low activity contain about 1 mole per mole of enzyme, and they have one free - S H group more per mole than the iron-containing enzymes (2 for E,; 8 for EJ. These results are in agreement with those of Yubisui and Yoneyama [I71 who found that 2-mercaptoethanol reversed to some extent the iron inhibition of a partially purified preparation. Hemin is well known as a feed-back regulator of 5-aminolevulinic acid synthetase [27], bound to the enzyme on two different sites, one of which is through its iron atom [15]. Although Rps spheroides 5-aminolevulinic acid synthetases do not seem to fit the general model of allosteric enzymes, binding of the small ligand iron could be an interesting way of regulation in the case where free heme does not accumulate, since free pigments do not accumulate to any obvious extent in nature. Besides this important possible role, iron has perhaps a stabilizing effect on these very oxidation/ reduction-sensitive enzymes, protecting the activecenter - SH group [l] from oxidation. The alterations in the rate of bacteriochlorophyll which occur in cells when the light intensity is changed are too rapid to be accounted for by induction or repression of enzyme synthesis. All the results of this and previous work [5,6,10,28] are consistent with the view that it has to be correlated with the fast changes in cellular adenosine triphosphate levels with light intensity [29,30]. The inhibition of bacteriochlorophyll synthesis in vivo by exogenous adenosine triphosphate in high [lo] and low-iron medium is probably due to the inhibition of 5-aminolevulinic acid synthetase as adenosine triphosphate inhibits the pure enzyme by forming a complex. The very specific inhibition of 5-aminolevulinic acid synthetases by adenosine triphosphate or pyrophosphate [6] is consistent with the interpretation that adenosine triphosphate is perhaps the primary agent through which light intensity controls bacteriochlorophyll production in Rps spheroides Y.
77 We are grateful to Dr B. Diner for critically reading our manuscript and to Mrs D. Clerot for skilled technical assistance.
REFERENCES 1 . Fanica-Gaignier, M. & Clement-Metral, J. D. (1973) Eur. J.
Biochem. 40, 13- 18. 2. Aoki, Y., Wada, 0.. Urata, G., Takaku, F. & Nakao, K. (1971) Biochem. Biophys. Res. Comrnun. 42, 568- 515. 3. Warnick, G . R. & Burnham, 9. F. (1971) J . Biol. Chem. 246, 6880-6885. 4. Tuboi, S., Kim, H. J. & Kikuchi, G. (1970) Arch. Biochem. Biophys. 138, 141- 154. 5. Fanica-Gaignier, M. & Clement-Metral, J. D. (1971) Biochc,m. Biophys. Res. Commun. 44, 192- 198. 6. Fanica-Gaignier, M. & Clement-Metral, J. D. (1973) Eur. J . Biochem. 40, 19-24. 7. Reference deleted. 8. Jordan, P. M. & Shemin, D. (1912) in The Enzymes, 3rd edn (Boyer,P. D., ed.) pp. 339-356, Academic Press, New York. 9. Reiss-Husson, F., De Klerk, G., Jolchine, G., Jauneau, E. & Karnen. M. D. (1971) Biochim. Biophys. Acta. 234, 73-82. 10. Fanica-Gaignier, M. (1 971) IInd lnternafional Congress on Photosynthesis (Forti, G., Avron, M. & Melandri, A,, eds) vol. 111, pp. 2721-2726, Dr Junk,W., N.V. Publishers, The Hague. 11. Karibian, D. & London, I. M. (1965) Biochem. Biophys. Res. Commun. 18,243-249. 12. Doeg, K. A. & Ziegler, D. M. (1962) Arch. Biochem. Biophys. 97, 37 -40. 13. Urata, G. & Granick, S . (1963) J. Biol. Chem. 238, 811-820. 14. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 15. Porra, R. J., Irving, E. A. & Tennick, A. M. (1972) Arch. Biochem. Biopli.vs. 148, 37-43. 16. Keilin, D. & Hartree, E. F. (1947) Biochem. J . 41, 503- 506. 17. Yubisui, T. & Yoneyama, U. (1972) Arch. Biochern. Biophys. 150, 17- 85. 18. Gajdos, A., Gajdos-Torok, M., Gorchein, A , , Neuberger, A. & Tait, G. H. (1968) Biochem. J . 106, 185. 19. Das, N., Cottam, G . L. & Srere. P. A. (1971) Arch. Biochez. Biophys. 143, 602- 608. 20. Lowry, 0. H., Rosebrough, N. J., Farr, A . L. & Randall, R. J. (1951) J . Biol. Chem. 193,265-215. 21. Bray, G. A. (1960) Anal. Biochem. 1, 279-285. 22. Scholnick, P. L., Hammaker, L. E. & Marver, H. S. (1972) J . Biol. Chem. 247, 4132. 23. Gibson, Q.H. (1973) J . Biol. Chem. 248, 1281- 1284. 24. Kitagawa, S . , Chiang, K. K. & Tonomura, Y . (1964) Biochim. Biopliys. Acta, 82, 83 -91. 25. Cohen-Bazire, G., Sistrom, W. R. & Stanier, R. Y . (1957) J. Cell Comp. Physiol. 49,25 - 68. 26. Lascelles, J. (1956) Biochem. J. 62, 78-93. 27. Laxelles, J. (1968) Biochem. SOC.Symp. 28, 49-59. 28. Tuboi, S.,Kim, H. J. & Kikuchi, G. (1970) Arch. Biochem. Biophys. 138, 155- 159. 29. Welsch, F . & Smith, L. (1969) Biochemistry, 8,3403-3408. 30. Fdnica-Gaignier, M., Clement-Metral, J. D. & Kamen, M. D. (1971) Biochim. Biophys. Acta, 226, 135- 143.
J. D. Clement-Metral, Laboratoire de PhotosynthZse du C.N.R.S., F-91190 Gif-sur-Yvette, France M. Fanica-Gdignier, Laboratoire d’Enzymologie du C.N.R.S. F-91190 Gif-sur-Yvette, France
