J. Biochem. 112, 828-833 (1992)
Inactivation of Bacillus subtilis Glutamine Synthetase by Metal-Catalyzed Oxidation Kinuko Kimura and Shigeyoshi Sugano Laboratory of Biochemistry, College of Science, Rikkyo (SL Paul's) University, Toshima-ku, Tokyo 171 Received for publication, May 26, 1992
Mixed-function oxidation has recently been identified as a post-translational covalent modification of proteins (1). Many enzymes are susceptible to mixed-function oxidation, and are converted to catalytically inactive forms which are readily degraded by a number of proteases (1-3). The mixed-function oxidation of glutamine synthetase from Escherichia coli has been extensively studied, with respect to protein turnover (4). The enzyme is oxidized and inactivated by several metal-catalyzed oxidation systems (5, 6), and the oxidized enzyme is preferentially degraded by a variety of purified proteinases (7, 8). Glutamine synthetase plays a central role in the nitrogen metabolism of microorganisms. Consequently, the enzyme is subject to intensive regulatory control. Regulation of the activity and synthesis of the glutamine synthetase in E. coli and several other Gram-negative bacteria has been extensively studied (9-12). In contrast to E. coli, Bacillus species do not appear to have a global system of regulation of nitrogen metabolism. Furthermore, their glutamine synthetases are not regulated by an adenylylation system (13, 14). In earlier papers, we reported the isolation and characterization of the glutamine synthetase from B. subtilis and B. cereus (15, 16). The activity of B. subtilis glutamine synthetase was highly sensitive to limited proteolysis by proteases (16, 17) and the protease-sensitive regions in the B. subtilis enzyme were quite different from the segment in the E. coli enzyme (18). From these results, we suggested that inactivation by limited proteolysis of B. subtilis glutamine synthetase may serve as an important regulatory 828
mechanism and may replace adenylylation of the E. coli enzyme. On the other hand, B. subtilis glutamine synthetase was quite unstable in cell-free extracts without addition of EDTA (16). We show here that the stabilizing effects of EDTA in crude extracts are caused by protection from oxidative inactivation and not by inhibition of proteolysis by Ca2+-dependent intracellular proteases (ISP) in this organism. To understand the structural basis of the inactivation, oxidative modification of glutamine synthetase was studied by utilizing a nonenzymic model oxidation system consisting of ascorbate, oxygen, and iron. MATERIALS AND METHODS Glutamine Synthetase—B. subtilis glutamine synthetase was purified from E. coli YMC11 harboring pSGS2 which contains the glnA gene of B. subtilis PCI219 (19). The enzyme was purified as described previously (20) and assayed by the y-glutamyltransferase method at pH 7.8 (22). The specific activity of the purified enzyme was constant for each isolation: 130 units/mg at 30"C. E. coli glutamine synthetase was purified from E. coli YMCll harboring pEGSl (a plasmid containing the glnA gene of E. coU). Preparation of Cell-Free Extracts—B. subtilis strains PCI219, KN2 (deficient in three proteases) (22) and RM141/pYKl (carrying a plasmid containing an intracellular serine protease gene) (23) were grown in L broth at 37*C to the stationary phase. The harvested cells were washed by the method of Nakayama et al. (24) to remove J. Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/112/6/828/789969 by University Of Otago user on 18 January 2019
Instability of Bacillus subtilis glutamine synthetase in crude extracts was attributed to site-specific oxidation by a mixed-function oxidation, and not to limited proteolysis by intracellular serine proteases (ISP). The crude extract from B. subtilis KN2, which is deficient in three intracellular proteases, inactivated glutamine synthetase similarly to the wild-type strain extract. To understand the structural basis of the functional change, oxidative modification of B. subtilis glutamine synthetase was studied utilizing a model system consisting of ascorbate, oxygen, and iron salts. The inactivation reaction appeared to be first order with respect to the concentration of unmodified enzyme. The loss of catalytic activity was proportional to the weakening of subunit interactions. B. subtilis glutamine synthetase was protected from oxidative modification by either 5 mM Mn2+ or 5 mM Mn2+ plus 5 mM ATP, but not by Mg2+. The CD-spectra and electron microscopic data showed that oxidative modification induced relatively subtle changes in the dodecameric enzyme molecules, but did not denature the protein. These limited changes are consistent with a site-specific free radical mechanism occurring at the metal binding site of the enzyme. Analytical data of the inactivated enzyme showed that loss of catalytic activity occurred faster than the appearance of carbonyl groups in amino acid side chains of the protein. In B. subtilis glutamine synthetase, the catalytic activity was highly sensitive to minute deviations of conformation in the dodecameric molecules and these subtle changes in the molecules could be regarded as markers for susceptibility to proteolysis.
Oxidative Inactivation of B. subtilis Glutamine Synthetase
829
extracellular proteases. Cell-free extracts from each strain were prepared by centrifugation after rupturing the cells by sonication (25). Superoxide dismutase (SOD) was purified from Mycobacterium smegmatis according to the method of Kusunose et al. (25). Protease Assay—One unit of protease was equivalent to the amount required to hydrolyze 1 //g of azocoll (Sigma Chemical) per min at 37*C (26). Oxidative Modification—Glutamine synthetase was oxidized by an ascorbate system (4, 5, 7). Unless otherwise specified, the oxidative cleavage reaction was performed by incubating 20^M/subunit of protein at 30'C in 50 mM Tris-HCl buffer, pH 7.2, containing 24 mM ascorbate and 20 //M FeCl3 or CuCl2. The oxidation reaction was stopped by adding EDTA (40 /iM) and suitable amounts of samples were analyzed on SDS-PAGE (28). The remaining glutamine synthetase activity was measured by means of the y-glutamyltransferase assay (15). Circular Dichroism Measurements—CD spectra were recorded on a JASCO model J-500 instrument at room temperature. Measurements were performed in a quartz cell with an optical path length of 10 mm for the near-UV region, and 1 mm for the far-UV region. The concentration of native or oxidized enzyme was 1.0 mg/ml (near-UV) or 0.1 mg/ml (far-UV) in 50 mM Tris-HCl buffer, pH7.2, containing 0.1 mM MnCl2. Electron Microscopy and Image Analysis—Electron microscopic examination was conducted as described in the previous paper (18) using a JEOL model JEM-1200EXII at 120 kV, at a magnification of 150,000. Images were processed with a JEOL model JIM-5000. Amino Acid Composition and Sequence Analysis—Following hydrolysis of glutamine synthetase with 6 N HC1 containing 2% phenol at 110'C for 24 h in an evacuated tube, the amino acid composition was determined on a Hitachi automatic amino acid analyzer. The amino acid sequences of the peptides were determined by manual Edman degradation as described previously (18, 27), or with a Shimadzu PSQ protein sequencer. Further, the sequence and mass of the peptide fragments were confirmed by mass spectroscopy with a JEOL JMS-SX102 mass spectrometer equipped with a JMA-DA6000 data system. Determination of Protein Carbonyl Content with Fluoresceinamine (FlNHt)—Fluoresceinamine derivatization of
the oxidativfe modified glutamine synthetase was performed as described by Climent et al. (28). Carbonyl content was calculated from the maximum absorbance at 490 nm using £„ for F1NH2 in 0.1 M NaOH of 86,800 NT1 • cm"1. Protein concentration was calculated from the absorbance at 290 nm in 0.1 M NaOH (eM =44,900), corrected for FlNHj absorbance.
Additions (mM)
Remaining activity* (%)
None 43.9 Phenylmethylsulfonyl fluoride (PMSF) (2) 71.2 Diisopropyl fluorophosphate (2) 57.3 Dithiothreitol (2) 35.8 2-Mercaptoethanol (2) 19.3 100 EDTA(l) 100 EGTA(l) 100 EDTA(1) + Ca1+ (2) Mn1+ (7.5) 80.3 ATP(7.5)+Mn 2+ (7.5) 90.4 1+ 100 ATP (7.5)+Mn (7.5)+PMSF (2) "The endogenous GS activity in freshly prepared cell-free extract (100%) was about 3.0 //mol y-glutamylhydroxamate/min/mg protein, and is expressed as a percentage of the activity remaining after incubation at 37*C for 1 h. Vol. 112, No. 6, 1992
Stability of Endogenous Glutamine Synthetase in Crude Extracts—Glutamine synthetase in cell-free extracts was quite unstable; about 60% of the activity was lost within 1 h at 37*C. An attempt was made to stabilize the enzyme by addition of various agents as shown in Table I. Although many divalent cation chelators, such as EDTA and EGTA at 1 mM, stabilized the enzyme, mercaptoethanol and dithiothreitol inactivated the enzyme. Serine protease inhibitors such as PMSF and DFP at 2 mM slightly stabilized the enzyme; however, the combination of substrates, Mn2+ATP (7.5 mM) and PMSF (2 mM) completely stabilized the enzyme. These results suggested that the glutamine synthetase in cell-free extracts may be inactivated by an intracellular serine protease (ISP) which is characterized by its sensitivity to EDTA due to an absolute requirement for Ca2+ for stability and activity (23). ISP Activity in Cell-Free Extracts and Inactivation of Exogenous Purified Glutamine Synthetase—B. subtilis produces a variety of extracellular and intracellular proteases. Cell-free extracts were prepared from the B. subtilis strains PCI219, KN2, and RM141/pYKl as described in "MATERIALS AND METHODS." The protease activity of these extracts was determined by using azocoll as a substrate, and isolated glutamine synthetase was incubated with each cell-free extract at 37*C. As shown in Table II, the extract from B. subtilis KN2 (a strain deficient in three proteases) had no detectable protease activity as expected; however, glutamine synthetase activity was inactivated. The extract from RMl41/pYKl (carrying a plasmid containing an ISP gene) had high protease activity, but glutamine synthetase inactivation was the weakest of the three extracts. The extracts from strain KN2 inactivated endogenous glutamine synthetase as well as exogenous glutamine synthetase isolated from B. subtilis PCI219. These results suggested that the ISP may not be the major factor in inactivating glutamine synthetase, and some other EDTA-sensitive factors may exist in the cell-free extracts. The inactivation reaction by each of the three cell-free preparations was inhibited completely by 2 mM EDTA and stimulated by either dithiothreitol, mercaptoethanol, re-
TABLE II. The effect of some intracellular protease activities on the inactivation of glutamine synthetase (GS). Cell-free extract from B. BubtUis
Protease activity* Remaining GS activity* 0/g/min/mg protein) (%) 0.0 58.0 9.7 73.9 0.46 18.6
KN2 RM141/pYKl PCI219 •One unit of protease was equivalent to the amount hydrolyzing 1 n% of azocoll per min at 37*C (29). "Isolated GS (0.17 mg) was incubated with 80 //I of each cell-free extract (~34 mg/ml) at 37*C for 60 min, and the remaining y-glutamyltransferase activity was expressed as a percentage of the activity without any addition.
Downloaded from https://academic.oup.com/jb/article-abstract/112/6/828/789969 by University Of Otago user on 18 January 2019
TABLE I. Stability of glutamine synthetase (GS) in cell-free extracts under different conditions.
RESULTS
830
K. Kimura and S. Sugano 00
1
=^t—-1—_
s 7 6
C\
1
—
x
-
100
—
5 4
\
3
\
2
\ \
g 7
:
e 5
•
4
2
B 1
1 -
-
.8 7 i .
30
40
t
10
Time (min)
12
3 4 5
3 4 5
i
i
20
30
40
Time (min)
Fig. 1. Inactivation of glutamine synthetase by ascorbate/FeI+. A: Effect of FeCU concentration. The reaction mixtures contained 30 mM Tris-HCl buffer (pH7.2), 60^M MnCl,, 20^M glutamine synthetase (subunit equivalent), 24 mM ascorbate, and 0 to 100 ^M Fed, in a final volume of 0.5 ml. 0, • ; 10, O; 20, A; 50, D; and 100 fiM, A. The mixtures were held at 30"C, and residual enzymatic activity was determined by means of the y-glutamyltransferase assay. B: Effect of MnCl, concentration. The reaction mixtures contained 20 /iM glutamine synthetase, 20 //M FeClj, 24 mM ascorbate, and 50 to
12
I
is!
W*
— -66
B
Inactivation of Bacillus subtilis Glutamine Synthetase by Metal-Catalyzed Oxidation Kinuko Kimura and Shigeyoshi Sugano Laboratory of Biochemistry, College of Science, Rikkyo (SL Paul's) University, Toshima-ku, Tokyo 171 Received for publication, May 26, 1992
Mixed-function oxidation has recently been identified as a post-translational covalent modification of proteins (1). Many enzymes are susceptible to mixed-function oxidation, and are converted to catalytically inactive forms which are readily degraded by a number of proteases (1-3). The mixed-function oxidation of glutamine synthetase from Escherichia coli has been extensively studied, with respect to protein turnover (4). The enzyme is oxidized and inactivated by several metal-catalyzed oxidation systems (5, 6), and the oxidized enzyme is preferentially degraded by a variety of purified proteinases (7, 8). Glutamine synthetase plays a central role in the nitrogen metabolism of microorganisms. Consequently, the enzyme is subject to intensive regulatory control. Regulation of the activity and synthesis of the glutamine synthetase in E. coli and several other Gram-negative bacteria has been extensively studied (9-12). In contrast to E. coli, Bacillus species do not appear to have a global system of regulation of nitrogen metabolism. Furthermore, their glutamine synthetases are not regulated by an adenylylation system (13, 14). In earlier papers, we reported the isolation and characterization of the glutamine synthetase from B. subtilis and B. cereus (15, 16). The activity of B. subtilis glutamine synthetase was highly sensitive to limited proteolysis by proteases (16, 17) and the protease-sensitive regions in the B. subtilis enzyme were quite different from the segment in the E. coli enzyme (18). From these results, we suggested that inactivation by limited proteolysis of B. subtilis glutamine synthetase may serve as an important regulatory 828
mechanism and may replace adenylylation of the E. coli enzyme. On the other hand, B. subtilis glutamine synthetase was quite unstable in cell-free extracts without addition of EDTA (16). We show here that the stabilizing effects of EDTA in crude extracts are caused by protection from oxidative inactivation and not by inhibition of proteolysis by Ca2+-dependent intracellular proteases (ISP) in this organism. To understand the structural basis of the inactivation, oxidative modification of glutamine synthetase was studied by utilizing a nonenzymic model oxidation system consisting of ascorbate, oxygen, and iron. MATERIALS AND METHODS Glutamine Synthetase—B. subtilis glutamine synthetase was purified from E. coli YMC11 harboring pSGS2 which contains the glnA gene of B. subtilis PCI219 (19). The enzyme was purified as described previously (20) and assayed by the y-glutamyltransferase method at pH 7.8 (22). The specific activity of the purified enzyme was constant for each isolation: 130 units/mg at 30"C. E. coli glutamine synthetase was purified from E. coli YMCll harboring pEGSl (a plasmid containing the glnA gene of E. coU). Preparation of Cell-Free Extracts—B. subtilis strains PCI219, KN2 (deficient in three proteases) (22) and RM141/pYKl (carrying a plasmid containing an intracellular serine protease gene) (23) were grown in L broth at 37*C to the stationary phase. The harvested cells were washed by the method of Nakayama et al. (24) to remove J. Biochem.
Downloaded from https://academic.oup.com/jb/article-abstract/112/6/828/789969 by University Of Otago user on 18 January 2019
Instability of Bacillus subtilis glutamine synthetase in crude extracts was attributed to site-specific oxidation by a mixed-function oxidation, and not to limited proteolysis by intracellular serine proteases (ISP). The crude extract from B. subtilis KN2, which is deficient in three intracellular proteases, inactivated glutamine synthetase similarly to the wild-type strain extract. To understand the structural basis of the functional change, oxidative modification of B. subtilis glutamine synthetase was studied utilizing a model system consisting of ascorbate, oxygen, and iron salts. The inactivation reaction appeared to be first order with respect to the concentration of unmodified enzyme. The loss of catalytic activity was proportional to the weakening of subunit interactions. B. subtilis glutamine synthetase was protected from oxidative modification by either 5 mM Mn2+ or 5 mM Mn2+ plus 5 mM ATP, but not by Mg2+. The CD-spectra and electron microscopic data showed that oxidative modification induced relatively subtle changes in the dodecameric enzyme molecules, but did not denature the protein. These limited changes are consistent with a site-specific free radical mechanism occurring at the metal binding site of the enzyme. Analytical data of the inactivated enzyme showed that loss of catalytic activity occurred faster than the appearance of carbonyl groups in amino acid side chains of the protein. In B. subtilis glutamine synthetase, the catalytic activity was highly sensitive to minute deviations of conformation in the dodecameric molecules and these subtle changes in the molecules could be regarded as markers for susceptibility to proteolysis.
Oxidative Inactivation of B. subtilis Glutamine Synthetase
829
extracellular proteases. Cell-free extracts from each strain were prepared by centrifugation after rupturing the cells by sonication (25). Superoxide dismutase (SOD) was purified from Mycobacterium smegmatis according to the method of Kusunose et al. (25). Protease Assay—One unit of protease was equivalent to the amount required to hydrolyze 1 //g of azocoll (Sigma Chemical) per min at 37*C (26). Oxidative Modification—Glutamine synthetase was oxidized by an ascorbate system (4, 5, 7). Unless otherwise specified, the oxidative cleavage reaction was performed by incubating 20^M/subunit of protein at 30'C in 50 mM Tris-HCl buffer, pH 7.2, containing 24 mM ascorbate and 20 //M FeCl3 or CuCl2. The oxidation reaction was stopped by adding EDTA (40 /iM) and suitable amounts of samples were analyzed on SDS-PAGE (28). The remaining glutamine synthetase activity was measured by means of the y-glutamyltransferase assay (15). Circular Dichroism Measurements—CD spectra were recorded on a JASCO model J-500 instrument at room temperature. Measurements were performed in a quartz cell with an optical path length of 10 mm for the near-UV region, and 1 mm for the far-UV region. The concentration of native or oxidized enzyme was 1.0 mg/ml (near-UV) or 0.1 mg/ml (far-UV) in 50 mM Tris-HCl buffer, pH7.2, containing 0.1 mM MnCl2. Electron Microscopy and Image Analysis—Electron microscopic examination was conducted as described in the previous paper (18) using a JEOL model JEM-1200EXII at 120 kV, at a magnification of 150,000. Images were processed with a JEOL model JIM-5000. Amino Acid Composition and Sequence Analysis—Following hydrolysis of glutamine synthetase with 6 N HC1 containing 2% phenol at 110'C for 24 h in an evacuated tube, the amino acid composition was determined on a Hitachi automatic amino acid analyzer. The amino acid sequences of the peptides were determined by manual Edman degradation as described previously (18, 27), or with a Shimadzu PSQ protein sequencer. Further, the sequence and mass of the peptide fragments were confirmed by mass spectroscopy with a JEOL JMS-SX102 mass spectrometer equipped with a JMA-DA6000 data system. Determination of Protein Carbonyl Content with Fluoresceinamine (FlNHt)—Fluoresceinamine derivatization of
the oxidativfe modified glutamine synthetase was performed as described by Climent et al. (28). Carbonyl content was calculated from the maximum absorbance at 490 nm using £„ for F1NH2 in 0.1 M NaOH of 86,800 NT1 • cm"1. Protein concentration was calculated from the absorbance at 290 nm in 0.1 M NaOH (eM =44,900), corrected for FlNHj absorbance.
Additions (mM)
Remaining activity* (%)
None 43.9 Phenylmethylsulfonyl fluoride (PMSF) (2) 71.2 Diisopropyl fluorophosphate (2) 57.3 Dithiothreitol (2) 35.8 2-Mercaptoethanol (2) 19.3 100 EDTA(l) 100 EGTA(l) 100 EDTA(1) + Ca1+ (2) Mn1+ (7.5) 80.3 ATP(7.5)+Mn 2+ (7.5) 90.4 1+ 100 ATP (7.5)+Mn (7.5)+PMSF (2) "The endogenous GS activity in freshly prepared cell-free extract (100%) was about 3.0 //mol y-glutamylhydroxamate/min/mg protein, and is expressed as a percentage of the activity remaining after incubation at 37*C for 1 h. Vol. 112, No. 6, 1992
Stability of Endogenous Glutamine Synthetase in Crude Extracts—Glutamine synthetase in cell-free extracts was quite unstable; about 60% of the activity was lost within 1 h at 37*C. An attempt was made to stabilize the enzyme by addition of various agents as shown in Table I. Although many divalent cation chelators, such as EDTA and EGTA at 1 mM, stabilized the enzyme, mercaptoethanol and dithiothreitol inactivated the enzyme. Serine protease inhibitors such as PMSF and DFP at 2 mM slightly stabilized the enzyme; however, the combination of substrates, Mn2+ATP (7.5 mM) and PMSF (2 mM) completely stabilized the enzyme. These results suggested that the glutamine synthetase in cell-free extracts may be inactivated by an intracellular serine protease (ISP) which is characterized by its sensitivity to EDTA due to an absolute requirement for Ca2+ for stability and activity (23). ISP Activity in Cell-Free Extracts and Inactivation of Exogenous Purified Glutamine Synthetase—B. subtilis produces a variety of extracellular and intracellular proteases. Cell-free extracts were prepared from the B. subtilis strains PCI219, KN2, and RM141/pYKl as described in "MATERIALS AND METHODS." The protease activity of these extracts was determined by using azocoll as a substrate, and isolated glutamine synthetase was incubated with each cell-free extract at 37*C. As shown in Table II, the extract from B. subtilis KN2 (a strain deficient in three proteases) had no detectable protease activity as expected; however, glutamine synthetase activity was inactivated. The extract from RMl41/pYKl (carrying a plasmid containing an ISP gene) had high protease activity, but glutamine synthetase inactivation was the weakest of the three extracts. The extracts from strain KN2 inactivated endogenous glutamine synthetase as well as exogenous glutamine synthetase isolated from B. subtilis PCI219. These results suggested that the ISP may not be the major factor in inactivating glutamine synthetase, and some other EDTA-sensitive factors may exist in the cell-free extracts. The inactivation reaction by each of the three cell-free preparations was inhibited completely by 2 mM EDTA and stimulated by either dithiothreitol, mercaptoethanol, re-
TABLE II. The effect of some intracellular protease activities on the inactivation of glutamine synthetase (GS). Cell-free extract from B. BubtUis
Protease activity* Remaining GS activity* 0/g/min/mg protein) (%) 0.0 58.0 9.7 73.9 0.46 18.6
KN2 RM141/pYKl PCI219 •One unit of protease was equivalent to the amount hydrolyzing 1 n% of azocoll per min at 37*C (29). "Isolated GS (0.17 mg) was incubated with 80 //I of each cell-free extract (~34 mg/ml) at 37*C for 60 min, and the remaining y-glutamyltransferase activity was expressed as a percentage of the activity without any addition.
Downloaded from https://academic.oup.com/jb/article-abstract/112/6/828/789969 by University Of Otago user on 18 January 2019
TABLE I. Stability of glutamine synthetase (GS) in cell-free extracts under different conditions.
RESULTS
830
K. Kimura and S. Sugano 00
1
=^t—-1—_
s 7 6
C\
1
—
x
-
100
—
5 4
\
3
\
2
\ \
g 7
:
e 5
•
4
2
B 1
1 -
-
.8 7 i .
30
40
t
10
Time (min)
12
3 4 5
3 4 5
i
i
20
30
40
Time (min)
Fig. 1. Inactivation of glutamine synthetase by ascorbate/FeI+. A: Effect of FeCU concentration. The reaction mixtures contained 30 mM Tris-HCl buffer (pH7.2), 60^M MnCl,, 20^M glutamine synthetase (subunit equivalent), 24 mM ascorbate, and 0 to 100 ^M Fed, in a final volume of 0.5 ml. 0, • ; 10, O; 20, A; 50, D; and 100 fiM, A. The mixtures were held at 30"C, and residual enzymatic activity was determined by means of the y-glutamyltransferase assay. B: Effect of MnCl, concentration. The reaction mixtures contained 20 /iM glutamine synthetase, 20 //M FeClj, 24 mM ascorbate, and 50 to
12
I
is!
W*
— -66
B
