- Email: [email protected]
Importance of disulfide bridge formation on folding of phospholipase D from Streptomyces antibioticus
BIOENGWEERING Vol. 89, No. 5, 506-508. 2000
JOURNAL OF BIOSCIENCE AND
Importance of Disulfide Bridge Formation on Folding of Phospholipase D from Streptomyces antibioticus YUGO IWASAKI,
TATSUAKI
NISHIYAMA, YASUAKI KAWARASAKI, HIDE0 NAKANO, AND TSUNEO YAMANE* Laboratory of Molecular Biotechnology, Department of Biological Mechanisms and Functions, Graduate School of Bio- and Agro-Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan Received 7 January 2OOWAccepted21 February 2000
The effects of redox conditions on the folding of phospholipase D (PLD) of Streptomyces antibioticus were investigated. Although the enzyme was very stable even in the presence of l.OM guanidinehydrocbrolide (Gdn-HCl), the coexistence of dithiothreitol (DTT) and Gdn-HCl inactivated the enzyme completely. The inactivated enzyme recovered its activity by dialysis in which DTT was removed prior to Gdn-HCl, whereas its activity was not recovered when Gdn-HCl was removed prior to DTT. In vitro protein synthesis was used for further analyses of the folding process. Active PLD was synthesized in the absence of DTT. The activity increased as the protein synthesis proceeded. In contrast, inactive PLD was synthesized in the presence of DTT. The inactive PLD could not be effectively activated by simple removal of the reductant, while incubation with Gdn-HCl and subsequent removal of DTT followed by that of Gdn-HCl was a much more effective method for the synthesis of active enzymes. From these results, it is suggested that: (i) PLD contains disulfide bridge(s), which is (are) necessary for maintaining its tertiary structure, (ii) correct formation of the disulfide bridge(s) is a critical step in the early stage of the (re)folding process, and (ii) the disulfide bridge(s) further facilitate the folding process, resulting in the synthesis of the active enzymes with the correct structure. [Key words: phospholipase D, disulfide bridge, protein folding, in vitro protein synthesis] Disulfide bridges are important type of intra- or intermolecular interactions for maintaining the tertiary structure of proteins. Incorrect or incomplete formation of disulfide bridge(s) often cause misfolding or incomplete folding of proteins, resulting in formation of proteins with biologically impaired functions. Phospholipase D (PLD) catalyzes the hydrolysis of phospholipids and transphosphatidylation, a kind of transesterification by which useful phospholipids are synthesized (1). Our previous attempt to express the pld gene of Streptomyces antibioticus through fusion with a portion of lac Z resulted in the formation of insoluble and inactive enzymes in the cytoplasm of a recombinant strain of Escherichia coli, whereas the use of a signal peptide-assisted expression resulted in accumulation of soluble and active enzymes in the periplasm (2-4). The difference in folding between these two cases might depend on the formation of a disulfide bridge(s), because the cytoplasm is under a more reduced atmosphere than the periplasm, and because PLD has eight cysteine residues in its primary structure (2). The present study is concerned with the effects of redox conditions on the folding of PLD of S. antibioticus. Using the in vitro protein synthesis system as well as conventional denaturation and renaturation experiments of the authentic PLD, the importance of disulfide bond formation on the correct folding of the enzyme has been demonstrated. The authentic PLD was prepared from a recombinant strain of E. coli BL21 (DE3) containing pPELB-PLD3 (3, 4). The enzyme (55 pg/ml, 54 units/ml) was incubated in a basal buffer (30mM Tris-HCl pH 7.5, 30mM NaCl) containing various amounts of dithiothreitol (DTT) and guanidine hydrochloride (Gdn-HCl) at 37°C
for 2 h. Residual PLD activity was measured using soy lecithin as the substrate (2). Figure 1 shows the effects of the denaturants on the activity of the authentic PLD. The activity of enzyme was approximately 70% of the initial activity even after incubation with 1 .O M Gdn-HCl. In the presence of even a low concentration of DTT (0.5 mM) together with Gdn-HCl, however, the enzyme lost its activity almost completely. These results suggest that the PLD contains some disulfide bond(s), which are necessary for maintaining active enzyme structure. PLD, completely inactivated with 10 mM DTT and 1 .O M Gdn-HCl, was used for refolding experiments by dialyses. Two different methods for dialysis were compared: method 1, the inactivated enzyme solution (0.5 ml) was dialyzed against 100 ml of basal buffer containing 1 .O
g ; 75 .C .t H m 5o .c .5 E 25 I? 0
0
0.25 Gdn-HCI
05
075
concentration
10 (ht)
FIG. 1. Effects of DTT and Gdn-HCl on inactivation of PLD. The authentic PLD was incubated with DTT and Gdn-HCl at 37°C for 2 h. The remaining activity relative to that after incubation without the denaturants is shown. In fact, PLD is so stable that the incubation without the denaturants did not affect its activity. DTT concentrationtestedwereO(O),0.5(~),1.0(~)and10mM(~).
* Corresponding author.
506
NOTES
VOL. 89, 2000
0
20 40 60 Reaction time (min)
DlT concentration
507
(mM)
FIG. 2. PLD synthesis by the in vitro system. (A) Time course of the synthesis. The reactions were performed in the presence(2 mM, circles), and absence (0 mM, squares) of DTT. Open and closed symbols denote the amount of the protein and PLD activity, respectively. (B) Effect of DTT concentration on the PLD expression. The reactions were performed for 1 h at various concentrations of DTT. Open and closed circles indicate the amount of synthesized protein and the enzymatic activity, respectively.
M Gdn-HCl, then against 1OOml of basal buffer (to remove DTT prior to Gdn-HCl); and method 2, the inactivated enzyme solution (0.5 ml) was dialyzed against IOOml of basal buffer containing 10mM DTT, and then against lOOm1 of basal buffer (so as to remove Gdn-HCI prior to DTT). Each step of the dialysis was carried out for 8 h at room temperature twice. Consequently, 5.7% (3.1 units/ml) of the initial enzymatic activity was recovered by method 1, whereas the activity was not recovered by method 2. The results imply that the disulfide bond(s) must be formed during the early stage of the refolding process for the correct folding of the enzyme. For further analyses of the folding process, we chose an in vitro protein synthesis system (5). The advantages of the in vitro system are that the redox conditions can be easily changed, and that the progress of folding can be monitored with that of protein synthesis (6, 7). A plasmid, pPLD12, with the pld gene downstream of the T7 promoter, was used as the template DNA. An in vitro coupled transcription/translation reaction was performed as described previously (7) using E. coli S30 extract. Figure 2A shows the time courses of the in vitro protein synthesis in the presence (2mM) or absence (OmM) of DTT. Between the reduced and the nonreduced conditions, the amounts of the synthesized protein (monitored by scintillation counting of 14C-labelled leucine incorporated into the hot trichroloacetic acidinsoluble fraction) were similar. However, PLD activity was found only under the nonreduced condition. The enzymatic activity increased in accordance with the increase in the synthesized protein, suggesting that PLD folded rapidly after or during protein synthesis in the absence of DTT. Figure 2B shows the effect of the concentration of DTT on the PLD expression. The amount of the synthesized protein was nearly constant regardless of the DTT concentration, while the enzymatic activity was detected only at very low DTT concentrations. This result indicates that even a very low concentration (0.3 mM) of DTT is sufficient to inhibit the synthesized PLD from transforming into the active form. Takeshita et al. successfully obtained functional hu-
man nerve growth factor (hNGF) and brain-derived neurotrophic factor (BDNF) by in vitro translation in the presence of DTT followed by simple dialysis of the reaction mixture to remove the reductant (8). Unlike in the case of hNGF and BDNF, however, our attempts to activate the in vitro-synthesized inactive PLD by simple removal of DTT was unsuccessful, producing only 0.1 units/ml of active PLD. In contrast, the inactive PLD could be activated much more effectively (1.4 units/ml) by incubation with 1.0 M Gdn-HCl for 2 h at 37°C and sequential removal of DTT followed by that of GdnHCI. These results imply that the inactive PLD synthesized under the reduced conditions might be in an incorrectly folded form, from which it cannot be correctly folded by simple removal of the reductant (i.e., it requires to be denatured prior to refolding). The incorrect folding is possibly triggered by incomplete disulfide bridge formation. Hence, it is reiterated that the disulfide bridge(s) must be formed during the early stage of the folding process for correct folding. A similar result is reported for a bacterial lipase which has a disulfide bond in its structure (7). Based on the above consideration, we speculate that the folding process proceeds as follows. Immediately after or during peptide synthesis, some of the eight cysteine residues in the primary structure form a disulfide bridge(s). The formation of the bridge(s) is a critical step so that its failure would generate a misfolded protein which cannot be corrected unless the incorrect structure is denatured. The correct formation of disulfide bridge(s) may further facilitate the folding process, resulting in the synthesis of active enzymes with the correct structure. REFERENCES 1. Juneja, L.R., Kazuoka, T., Goto, N., Yamane, T., and Shimizu, S.: Conversion of phosphatidylcholine to phosphatidylserine by various phospholipase D in the presence of Lor D-serine. Biochim. Biophys. Acta, 1003, 277-283 (1989). 2. Iwasaki, Y., Nakano, H., and Yamane, T.: Phospholipase D from Streptomyces antibioticus: cloning, sequencing, expression, and relationship to other phospholipases. Appl. Micro-
508
IWASAKI ET AL.
biol. Biotechnol., 42, 290-299 (1994). 3. Iwasaki, Y., Mishima, N., Mizamoto, K., Nakano, H., and Yamane, T.: Extracellular production of phospholipase D of Streptomyces antibioticus using recombinant Escherichia coli. J. Ferment. Bioeng., 79, 417-421 (1995). 4. Mishima. N.. Mlzamoto. K.. Iwasaki. Y.. Nakano. H.. and Yamane,‘T.: ‘Insertion of stabilizing loci in ‘vectors of T7 ‘RNA polymerase-mediated Escherichia coli expression system: a case study on the plasmids involving foreign phospholipase D gene. Biotechnol. Prog., 13, 864-868 (1997). 5. Nakano, H. and Yamane, T.: Cell-free protein synthesis system. Biotechnol. Adv., 16, 367-384 (1998).
J. BIOSCI. BIOENG., 6. Kolb, V. A., Makeyv, E. V., and Spirin, A. S.: Folding of firefly luciferase during translation in a cell free system. EMBO J., 13, 3631-3637 (1994). 7. Yang, J., Kobayashi, K., Iwasaki, Y., Nakano, H., and Yamane, T.: In vitro analysis of roles of a disulfide bridge and a calcium binding site in activation of Pseudomonas sp. strain KWI-56 lipase. J. Bacterial., 182, 295-302 (2000). 8. Takeshita, T., Fujimori, K., and Shim&u, N.: Synthesis of biologically active human nerve growth factor and brain-derived neurotrophic factor by a cell-free system. J. Ferment. Bioeng., 81, 13-17 (1996).
JOURNAL OF BIOSCIENCE AND
Importance of Disulfide Bridge Formation on Folding of Phospholipase D from Streptomyces antibioticus YUGO IWASAKI,
TATSUAKI
NISHIYAMA, YASUAKI KAWARASAKI, HIDE0 NAKANO, AND TSUNEO YAMANE* Laboratory of Molecular Biotechnology, Department of Biological Mechanisms and Functions, Graduate School of Bio- and Agro-Sciences, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan Received 7 January 2OOWAccepted21 February 2000
The effects of redox conditions on the folding of phospholipase D (PLD) of Streptomyces antibioticus were investigated. Although the enzyme was very stable even in the presence of l.OM guanidinehydrocbrolide (Gdn-HCl), the coexistence of dithiothreitol (DTT) and Gdn-HCl inactivated the enzyme completely. The inactivated enzyme recovered its activity by dialysis in which DTT was removed prior to Gdn-HCl, whereas its activity was not recovered when Gdn-HCl was removed prior to DTT. In vitro protein synthesis was used for further analyses of the folding process. Active PLD was synthesized in the absence of DTT. The activity increased as the protein synthesis proceeded. In contrast, inactive PLD was synthesized in the presence of DTT. The inactive PLD could not be effectively activated by simple removal of the reductant, while incubation with Gdn-HCl and subsequent removal of DTT followed by that of Gdn-HCl was a much more effective method for the synthesis of active enzymes. From these results, it is suggested that: (i) PLD contains disulfide bridge(s), which is (are) necessary for maintaining its tertiary structure, (ii) correct formation of the disulfide bridge(s) is a critical step in the early stage of the (re)folding process, and (ii) the disulfide bridge(s) further facilitate the folding process, resulting in the synthesis of the active enzymes with the correct structure. [Key words: phospholipase D, disulfide bridge, protein folding, in vitro protein synthesis] Disulfide bridges are important type of intra- or intermolecular interactions for maintaining the tertiary structure of proteins. Incorrect or incomplete formation of disulfide bridge(s) often cause misfolding or incomplete folding of proteins, resulting in formation of proteins with biologically impaired functions. Phospholipase D (PLD) catalyzes the hydrolysis of phospholipids and transphosphatidylation, a kind of transesterification by which useful phospholipids are synthesized (1). Our previous attempt to express the pld gene of Streptomyces antibioticus through fusion with a portion of lac Z resulted in the formation of insoluble and inactive enzymes in the cytoplasm of a recombinant strain of Escherichia coli, whereas the use of a signal peptide-assisted expression resulted in accumulation of soluble and active enzymes in the periplasm (2-4). The difference in folding between these two cases might depend on the formation of a disulfide bridge(s), because the cytoplasm is under a more reduced atmosphere than the periplasm, and because PLD has eight cysteine residues in its primary structure (2). The present study is concerned with the effects of redox conditions on the folding of PLD of S. antibioticus. Using the in vitro protein synthesis system as well as conventional denaturation and renaturation experiments of the authentic PLD, the importance of disulfide bond formation on the correct folding of the enzyme has been demonstrated. The authentic PLD was prepared from a recombinant strain of E. coli BL21 (DE3) containing pPELB-PLD3 (3, 4). The enzyme (55 pg/ml, 54 units/ml) was incubated in a basal buffer (30mM Tris-HCl pH 7.5, 30mM NaCl) containing various amounts of dithiothreitol (DTT) and guanidine hydrochloride (Gdn-HCl) at 37°C
for 2 h. Residual PLD activity was measured using soy lecithin as the substrate (2). Figure 1 shows the effects of the denaturants on the activity of the authentic PLD. The activity of enzyme was approximately 70% of the initial activity even after incubation with 1 .O M Gdn-HCl. In the presence of even a low concentration of DTT (0.5 mM) together with Gdn-HCl, however, the enzyme lost its activity almost completely. These results suggest that the PLD contains some disulfide bond(s), which are necessary for maintaining active enzyme structure. PLD, completely inactivated with 10 mM DTT and 1 .O M Gdn-HCl, was used for refolding experiments by dialyses. Two different methods for dialysis were compared: method 1, the inactivated enzyme solution (0.5 ml) was dialyzed against 100 ml of basal buffer containing 1 .O
g ; 75 .C .t H m 5o .c .5 E 25 I? 0
0
0.25 Gdn-HCI
05
075
concentration
10 (ht)
FIG. 1. Effects of DTT and Gdn-HCl on inactivation of PLD. The authentic PLD was incubated with DTT and Gdn-HCl at 37°C for 2 h. The remaining activity relative to that after incubation without the denaturants is shown. In fact, PLD is so stable that the incubation without the denaturants did not affect its activity. DTT concentrationtestedwereO(O),0.5(~),1.0(~)and10mM(~).
* Corresponding author.
506
NOTES
VOL. 89, 2000
0
20 40 60 Reaction time (min)
DlT concentration
507
(mM)
FIG. 2. PLD synthesis by the in vitro system. (A) Time course of the synthesis. The reactions were performed in the presence(2 mM, circles), and absence (0 mM, squares) of DTT. Open and closed symbols denote the amount of the protein and PLD activity, respectively. (B) Effect of DTT concentration on the PLD expression. The reactions were performed for 1 h at various concentrations of DTT. Open and closed circles indicate the amount of synthesized protein and the enzymatic activity, respectively.
M Gdn-HCl, then against 1OOml of basal buffer (to remove DTT prior to Gdn-HCl); and method 2, the inactivated enzyme solution (0.5 ml) was dialyzed against IOOml of basal buffer containing 10mM DTT, and then against lOOm1 of basal buffer (so as to remove Gdn-HCI prior to DTT). Each step of the dialysis was carried out for 8 h at room temperature twice. Consequently, 5.7% (3.1 units/ml) of the initial enzymatic activity was recovered by method 1, whereas the activity was not recovered by method 2. The results imply that the disulfide bond(s) must be formed during the early stage of the refolding process for the correct folding of the enzyme. For further analyses of the folding process, we chose an in vitro protein synthesis system (5). The advantages of the in vitro system are that the redox conditions can be easily changed, and that the progress of folding can be monitored with that of protein synthesis (6, 7). A plasmid, pPLD12, with the pld gene downstream of the T7 promoter, was used as the template DNA. An in vitro coupled transcription/translation reaction was performed as described previously (7) using E. coli S30 extract. Figure 2A shows the time courses of the in vitro protein synthesis in the presence (2mM) or absence (OmM) of DTT. Between the reduced and the nonreduced conditions, the amounts of the synthesized protein (monitored by scintillation counting of 14C-labelled leucine incorporated into the hot trichroloacetic acidinsoluble fraction) were similar. However, PLD activity was found only under the nonreduced condition. The enzymatic activity increased in accordance with the increase in the synthesized protein, suggesting that PLD folded rapidly after or during protein synthesis in the absence of DTT. Figure 2B shows the effect of the concentration of DTT on the PLD expression. The amount of the synthesized protein was nearly constant regardless of the DTT concentration, while the enzymatic activity was detected only at very low DTT concentrations. This result indicates that even a very low concentration (0.3 mM) of DTT is sufficient to inhibit the synthesized PLD from transforming into the active form. Takeshita et al. successfully obtained functional hu-
man nerve growth factor (hNGF) and brain-derived neurotrophic factor (BDNF) by in vitro translation in the presence of DTT followed by simple dialysis of the reaction mixture to remove the reductant (8). Unlike in the case of hNGF and BDNF, however, our attempts to activate the in vitro-synthesized inactive PLD by simple removal of DTT was unsuccessful, producing only 0.1 units/ml of active PLD. In contrast, the inactive PLD could be activated much more effectively (1.4 units/ml) by incubation with 1.0 M Gdn-HCl for 2 h at 37°C and sequential removal of DTT followed by that of GdnHCI. These results imply that the inactive PLD synthesized under the reduced conditions might be in an incorrectly folded form, from which it cannot be correctly folded by simple removal of the reductant (i.e., it requires to be denatured prior to refolding). The incorrect folding is possibly triggered by incomplete disulfide bridge formation. Hence, it is reiterated that the disulfide bridge(s) must be formed during the early stage of the folding process for correct folding. A similar result is reported for a bacterial lipase which has a disulfide bond in its structure (7). Based on the above consideration, we speculate that the folding process proceeds as follows. Immediately after or during peptide synthesis, some of the eight cysteine residues in the primary structure form a disulfide bridge(s). The formation of the bridge(s) is a critical step so that its failure would generate a misfolded protein which cannot be corrected unless the incorrect structure is denatured. The correct formation of disulfide bridge(s) may further facilitate the folding process, resulting in the synthesis of active enzymes with the correct structure. REFERENCES 1. Juneja, L.R., Kazuoka, T., Goto, N., Yamane, T., and Shimizu, S.: Conversion of phosphatidylcholine to phosphatidylserine by various phospholipase D in the presence of Lor D-serine. Biochim. Biophys. Acta, 1003, 277-283 (1989). 2. Iwasaki, Y., Nakano, H., and Yamane, T.: Phospholipase D from Streptomyces antibioticus: cloning, sequencing, expression, and relationship to other phospholipases. Appl. Micro-
508
IWASAKI ET AL.
biol. Biotechnol., 42, 290-299 (1994). 3. Iwasaki, Y., Mishima, N., Mizamoto, K., Nakano, H., and Yamane, T.: Extracellular production of phospholipase D of Streptomyces antibioticus using recombinant Escherichia coli. J. Ferment. Bioeng., 79, 417-421 (1995). 4. Mishima. N.. Mlzamoto. K.. Iwasaki. Y.. Nakano. H.. and Yamane,‘T.: ‘Insertion of stabilizing loci in ‘vectors of T7 ‘RNA polymerase-mediated Escherichia coli expression system: a case study on the plasmids involving foreign phospholipase D gene. Biotechnol. Prog., 13, 864-868 (1997). 5. Nakano, H. and Yamane, T.: Cell-free protein synthesis system. Biotechnol. Adv., 16, 367-384 (1998).
J. BIOSCI. BIOENG., 6. Kolb, V. A., Makeyv, E. V., and Spirin, A. S.: Folding of firefly luciferase during translation in a cell free system. EMBO J., 13, 3631-3637 (1994). 7. Yang, J., Kobayashi, K., Iwasaki, Y., Nakano, H., and Yamane, T.: In vitro analysis of roles of a disulfide bridge and a calcium binding site in activation of Pseudomonas sp. strain KWI-56 lipase. J. Bacterial., 182, 295-302 (2000). 8. Takeshita, T., Fujimori, K., and Shim&u, N.: Synthesis of biologically active human nerve growth factor and brain-derived neurotrophic factor by a cell-free system. J. Ferment. Bioeng., 81, 13-17 (1996).