- Email: [email protected]
Human Pancreatic Triglyceride Lipase Expressed in Yeast Cells: Purification and Characterization
PROTEIN EXPRESSION AND PURIFICATION ARTICLE NO.
13, 36–40 (1998)
PT980874
Human Pancreatic Triglyceride Lipase Expressed in Yeast Cells: Purification and Characterization Yanqing Yang* and Mark E. Lowe*,†,1 *Department of Pediatrics and †Department of Molecular Biology and Pharmacology, Washington University School of Medicine, One Children’s Place, St. Louis, Missouri 63110
A cDNA clone encoding human pancreatic triglyceride lipase was cloned into a yeast expression vector so that the yeast PHO1 signal peptide replaced the native signal peptide. Pichia pastoris cells were transfected with the vector, and clones expressing human pancreatic triglyceride lipase were isolated. Recombinant human pancreatic lipase was expressed in broth cultures and was purified from the medium by DEAE blue Sepharose and hydroxyapatite chromatography. The highly purified lipase had specific activities for various triglyceride substrates identical to those of tissuepurified human pancreatic triglyceride lipase; it was inhibited by bile salts, required colipase for activity, and demonstrated interfacial activation. This expression system is suitable for the rapid, efficient production of human pancreatic triglyceride lipase in amounts adequate for biophysical studies. q 1998 Academic Press
Human pancreatic triglyceride lipase is an enzyme of interest to physiologists, to clinicians, and to biochemists. It is central to the digestion of dietary fats. It is a serum marker for acute pancreatitis and it is the archetype of lipases, hydrophilic enzymes that act on water-insoluble substrates at an oil–water interface. The molecular details of lipase action have implications for the understanding of the many biological processes that occur at oil–water interfaces. Over the past 5 years, multiple laboratories have contributed to our understanding of the molecular mechanisms of lipolysis (1). Many of these studies have concentrated on pancreatic triglyceride lipase. The recently solved three-dimensional structures of human pancreatic triglyceride lipase in two different conformations provided a plausible model to explain the preference of this enzyme for water-insoluble substrates 1 To whom correspondence should be addressed. Fax: (314) 4544218. E-mail: [email protected].
over water-soluble substrates (2,3). These studies also identified candidates for the active site residues, the acyl chain-binding sites, and regions that interact with colipase, a protein cofactor of pancreatic triglyceride lipase. Site-specific mutagenesis of human pancreatic triglyceride lipase has been an important complement to the biophysical studies of the enzyme structure. The active site residues and the role of a surface lid in catalysis were detailed by characterizing mutant human pancreatic lipases (4,5). Mutagenesis studies provide important information about the relationships of structure to lipase function, but are hampered by the inability to quickly produce mutant proteins in available expression systems, particularly in quantities that are suitable for biophysical measurements on the mutants. Structure determinations on the mutants are critical components of any interpretation of kinetic results with the mutants and are necessary before the contribution of a particular amino acid or region can be fully understood. The requirements for a system that is easily and quickly manipulated, that processes and secretes mammalian proteins correctly, and that can produce large amounts of protein can potentially be met by expression in Pichia pastoris. We tested the ability of these yeast to express human pancreatic triglyceride lipase and found them suitable for producing milligram quantities of protein that is easily purified and that possesses enzymatic properties that are indistinguishable from human pancreatic lipase expressed in baculovirus-infected insect cells or the protein isolated from human pancreas. MATERIALS AND METHODS
Materials. The accompanying paper lists the sources for reagents (6). Transformation of yeast cells. The cDNA encoding human pancreatic lipase was subcloned into the pHILS1 expression vector. To accomplish the subcloning, an
36
AID
1046-5928/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
PEP 0874
/
6q1e$$$$61
05-11-98 10:36:28
pepa
AP: PEP
HUMAN PANCREATIC LIPASE EXPRESSION
EcoRI site was added to the 3*-end and an XhoI site was added to the 5*-end by polymerase chain reaction. The 5*-primer overlapped the sequence encoding the mature amino-terminus of the protein and the sequence encoding the carboxy-terminus of the yeast PHO1 signal peptide. This product, after being subcloned into the vector, replaced the native pancreatic triglyceride lipase signal peptide with the yeast PHO1 signal peptide and replaced the amino-terminal lysine with arginine to recreate the yeast signal peptidase cleavage site. The sequence of the cDNA was determined by dideoxynucleotide sequence analysis. The vector was linearized with Bgl II and transformed into GS115 by electroporation, and positive clones were selected as described (6). The medium from the methanol-induced cultures was analyzed by immunoblot and tributyrin assay as previously described (4,7). Expression of human pancreatic triglyceride lipase in yeast. Expression was accomplished as described in the accompanying paper with several minor exceptions (6). One liter of buffered minimal glycerol-complex medium divided into two 2-liter baffled flasks was inoculated with a 25-ml overnight culture and grown until the OD600 was between 6.0 and 10.0. The harvested cell pellet was resuspended in 200 ml of buffered minimal methanol-complex medium. Expression was continued for 4 to 5 days after methanol induction. Purification of human pancreatic triglyceride lipase. The medium was concentrated to 50 ml over a Filtron (Northborough, MA) Ultrasette tangential flow filter with a pore size of 30,000. The buffer was changed by dilution with 10 mM Tris–Cl, pH 8.0, and concentration with the Ultrasette. The sample was passed over a DEAE blue Sepharose column (1.5 1 10 cm) equilibrated in the same buffer (7). Human pancreatic lipase was collected in the pass-through from this column and dialyzed against 0.05 M NaPO4 , pH 7.0. The sample was applied to a CHT-I ceramic hydroxyapatite column from Bio-Rad (Richmond, CA) that was connected to a Pharmacia Akta Explorer chromatography system. The column was equilibrated in the dialysis buffer and lipase was eluted with a 10 column volume linear gradient from 0.05 to 0.35 M NaPO4 , pH 7.0. Elution was monitored at OD280 and activity was measured on each fraction. Lipase eluted as a single peak in the middle of the gradient. The fractions containing human pancreatic lipase were dialyzed against 10 mM Tris–Cl, pH 8.0, and 100 mM NaCl. After dialysis the sample was brought to 2 mM CaCl2 by adding the appropriate volume of 1 M CaCl2 . The recombinant lipase could be stored at 0807C for at least a year without loss of activity. Human pancreatic triglyceride lipase was purified from human pancreas and from baculovirus-infected
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
37
FIG. 1. Time course of lipase expression after methanol induction. Aliquots of the culture were removed every day after the switch to buffered minimal methanol-complex medium. The cells were removed by centrifugation and 20 ml of medium was assayed with tributyrin. 2 mg of colipase was added to each assay. The values are plotted as activity (mmol fatty acid released/min/20 ml of medium) versus day after induction.
Sf9 cells as previously described (7). Colipase was purified from human pancreatic tissue or recombinant colipase was produced in yeast and purified as described (6,8). Protein methods and lipase assay. Amino-terminus sequencing was performed at the Protein and Nucleic Acid Chemistry laboratory (Washington University School of Medicine, St. Louis) using an Applied Biosystems gas-phase sequencer. The presence of carbohydrate was detected with the Immun-Blot Kit for Glycoprotein Detection from Bio-Rad. The manufacturer’s instructions for total carbohydrate labeling in solution, protocol 2A, were followed. The labeled proteins were separated on SDS–PAGE and transferred to an Immobilon-P membrane (Millipore, Bedford, MA) as previously described (4). The blot was stained with a streptavidin–alkaline phosphatase conjugate following the Bio-Rad instructions. The concentration of purified human pancreatic triglyceride lipase and colipase were determined using an extinction coefficient at OD280 of E1% Å 1.2 and 0.3, respectively. Lipase activity was determined in the pH-STAT as previously described (7). Interfacial activation was determined against varying concentrations of tripropionin stabilized in 2% gum arabic (9). RESULTS AND DISCUSSION
Expression of human pancreatic triglyceride lipase. To determine the time course of expression, aliquots were removed before and each day following methanol induction. Lipase activity was not present prior to induction but was detected within 24 h after induction (Fig. 1). The activity rose slowly over the next 3 days.
pepa
AP: PEP
38
YANG AND LOWE
FIG. 2. Time course of lipase expression analyzed by SDS–polyacrylamide gel electrophoresis and immunoblot. 20-ml aliquots of medium from the samples described in the legend to Fig. 1 were mixed with 20 ml of 21 SDS sample buffer and heated at 1007C for 5 min. The samples were prepared in duplicate. Both were separated by SDS–polyacrylamide gel electrophoresis on two separate gels. One gel was stained with Coomassie blue (A) and the other was transferred to a membrane and stained with anti-human pancreatic triglyceride lipase antibody (B). The numbers below each lane are the days after induction. M, markers. The molecular weights of the markers are given to the left of Panel A. The arrows mark the position of lipase.
No lipase activity was detected in the medium from yeast transformed with a colipase vector, indicating that the yeast did not produce an endogenous, secreted lipase (6). Lipase was also detected by Coomassie blue staining of SDS–polyacrylamide gels (Fig. 2). A band migrating at the proper position for human pancreatic triglyceride lipase was the predominant band in the medium from 24 h after induction through at least 4 days. This band was not present in the medium prior to induction, indicating that protein expression is tightly regulated by methanol. The protein band stained with a polyclonal antibody against human pancreatic triglyceride lipase. Multiple yeast proteins also stained in the immunoblot, but most were not induced by methanol and were detected with preimmune serum (data not shown). The presence of lipase activity and the appearance of a protein of the molecular size of human pancreatic lipase that binds an antibody against hu-
man pancreatic triglyceride lipase demonstrate the expression of this protein in P. pastoris. Lipase was purified from the medium as described under Materials and Methods. The yield at each step of the isolation procedure is given in Table 1. SDS– polyacrylamide gel analysis of the medium (Fig. 2) suggested that the large amount of protein detected in the medium by the BCA assay was from the peptone added to the medium rather than from proteins secreted from the yeast. A 10-fold increase in purity was achieved by chromatography over DEAE blue Sepharose. This fraction still had a light brown color and SDS–PAGE analysis revealed that the majority of the protein was lipase but several minor contaminants were still present (data not shown). Both the color and the contaminating proteins were removed by hydroxylapatite chromatography. The fractions from this column contained a single, major protein band that had a specific activity for tributyrin comparable to that previously reported for purified human pancreatic triglyceride lipase (Table 1 and Fig. 3) (10). The amino-terminal sequence of the recombinant lipase was REVCYERLG, confirming that the recombinant protein is human pancreatic lipase and that the signal peptide was properly cleaved by the yeast signal peptidase. The tissue-purified lipase migrated as two major bands with several minor, faster migrating bands also present (Fig. 3). All of these bands reacted with antibody to human pancreatic triglyceride lipase and the smaller fragments probably represent degradation products (data not shown). Although it was possible that the difference in size between the two major bands was also caused by proteolysis, it seemed likely that the size difference resulted from posttranslational processing. Only one amino-terminal sequence was present, KEVCYERLG, demonstrating that there was no internal cleavage of the peptide chain. The preparation had a specific activity for several different substrates that was comparable to those reported by others, suggesting that the two species had identical activities (see below). Secreted human pancreatic triglyceride lipase is glycosylated (11). Thus, we determined whether glycosyla-
TABLE 1
Purification of Recombinant Human Pancreatic Triglyceride Lipase from Yeast Cell Medium Purification step
Volume (ml)
Total protein (mg)
Total activity (Units)
Specific activity (Units/mg)
% Yield
Medium DEAE Hydroxylapatite
200 75 23
205 32.8 15.2
1.1 1 108 1.8 1 108 1.2 1 108
537 5488 7895
100 164 109
Note. Activity was determined as described under Materials and Methods and is expressed in units of mmol fatty acid released/min. A 5 M excess of colipase was present in the assay.
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
pepa
AP: PEP
HUMAN PANCREATIC LIPASE EXPRESSION
FIG. 3. SDS–polyacrylamide gel electrophoresis of purified recombinant and tissue-purified human pancreatic triglyceride lipase. The proteins were purified as described under Materials and Methods. 10 mg of yeast expressed lipase and 20 mg of tissue-purified lipase were dissolved in SDS sample buffer and heated at 1007C for 5 min. The gel was stained with Coomassie blue. M, markers; 1, yeast expressed lipase; 2, tissue-purified lipase.
tion accounted for the size difference between the two bands. We oxidized any carbohydrate residues on tissue-purified and recombinant human pancreatic triglyceride lipase and reacted with biotin. The samples were run on SDS–polyacrylamide gels and were transferred to a membrane. The presence of carbohydrate was detected with streptavidin–alkaline phosphatase and color development reagents. The upper bands in tissue-purified and recombinant lipase were detected by this method, indicating that they were glycosylated (Fig. 4). The lower band in the tissue-purified lipase was not detected. These results suggest that the two forms of lipase present in the tissue preparation differ by the absence or presence of an oligosaccharide chain rather than proteolytic cleavage.
39
FIG. 4. Glycoprotein staining of purified recombinant and tissuepurified human pancreatic triglyceride lipase. The proteins were processed as described under Materials and Methods. An equal volume of 21 SDS sample buffer was added to each sample and the samples were heated at 1007C for 5 min. The samples were separated by SDS– polyacrylamide gel electrophoresis and transferred to an Immobilon membrane. The membranes were stained as described under Materials and Methods. Lanes 1, 10 mg ovalbumin; 2, 10 mg tissue-purified lipase; 3, 5 mg yeast expressed lipase; 4, 20 mg albumin; 5, 20 mg yeast expressed lipase; 6, 40 mg tissue-purified lipase. M, markers. The arrows mark the position of lipase.
Activity of recombinant human pancreatic triglyceride lipase. The activity of the yeast expressed lipase was compared to that of tissue-purified lipase. Substrates with different acyl chain lengths were tested in the presence of 4 mM taurodeoxycholate (Table 2). No activity was detected in the absence of colipase, indicating that the recombinant lipase was inhibited by bile salts. When colipase was added, recombinant lipase had specific activities for all three substrates
TABLE 2
Specific Activity of Human Pancreatic Triglyceride Lipase from Various Sources Specific activity (mmol fatty acid/min/mg) Lipase
Tributyrin
Trioctanoin
Triolein
Tissue-purifieda Baculovirusa P. pastoris
8200 { 630 7720 { 550 8028 { 780
2579 { 180 2480 { 230 2451 { 175
1650 { 85 1500 { 90 1716 { 60
Note. Activity was determined as described under Materials and Methods. a Values taken from Ref. (7).
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
FIG. 5. Interfacial activation of yeast expressed pancreatic triglyceride lipase. 2.5 mg of yeast expressed lipase and 1 mg of colipase were incubated with various amounts of tripropionin as described under Materials and Methods. Assays were done in the pH STAT. Activity is mmol fatty acid released/min. The arrow indicates the solubility limit of tripropionin as determined by light scattering.
pepa
AP: PEP
40
YANG AND LOWE
that were indistinguishable from the values for tissue-purified and baculovirus expressed lipase. These data demonstrate that the recombinant human pancreatic triglyceride lipase faithfully reproduces the important properties of bile salt inhibition, colipase dependence, and substrate specificity found in tissuepurified lipase. Another important kinetic property of pancreatic triglyceride lipase is the preference for water-insoluble substrates over water-soluble substrates. The large increase in activity seen when lipase encounters an interface, a property termed interfacial activation, is characteristic of this enzyme. We determined whether the recombinant lipase showed interfacial activation against tripropionin (Fig. 5). Recombinant human pancreatic triglyceride lipase had low levels of activity until the solubility of tripropionin was exceeded. When water-insoluble substrate particles formed, the activity increased markedly. The yeast-produced lipase retained the property of interfacial activation. Conclusion. Human pancreatic triglyceride lipase was successfully produced in a yeast expression system. The best expression was obtained by replacing the native signal peptide with a yeast signal peptide. This required changing the mature amino-terminus from lysine to arginine. There was no detectable effect of this change on the enzymatic properties of the recombinant lipase. The ability to easily manipulate and grow the yeast and to produce milligram quantities of protein in small culture volumes makes this system useful for expressing native and mutant human pancreatic triglyceride lipase.
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
ACKNOWLEDGMENT This work was supported by NIH Grant HD/DK 33060.
REFERENCES 1. Lowe, M. E. (1997) Structure and function of pancreatic lipase and colipase. Annu. Rev. Nutr. 17, 141–158. 2. van Tilbeurgh, H., Sarda, L., Verger, R., and Cambillau, C. (1992) Structure of the pancreatic lipase–procolipase complex. Nature 359, 159–162. 3. van Tilbeurgh, H., Egloff, M. P., Martinez, C., Rugani, N., Verger, R., and Cambillau, C. (1993) Interfacial activation of the lipase–procolipase complex by mixed micelles revealed by X-ray crystallography. Nature 362, 814–820. 4. Lowe, M. E. (1992) The catalytic site residues and interfacial binding of human pancreatic lipase. J. Biol. Chem. 267, 17069– 17073. 5. Jennens, M. L., and Lowe, M. E. (1994) A surface loop covering the active site of human pancreatic lipase influences interfacial activation and lipid binding. J. Biol. Chem. 269, 25470–25474. 6. Cordle, R. A., and Lowe, M. E. (1997) Purification and characterization of human procolipase expressed in yeast cells. Prot. Exp. Purif., 13, 30–35. 7. Lowe, M. E. (1996) Mutation of the catalytic site Asp177 to Glu177 in human pancreatic lipase produces an active lipase with increased sensitivity to proteases. Biochim. Biophys. Acta 1302, 177–183. 8. Lowe, M. E. (1994) Human pancreatic procolipase expressed in insect cells: Purification and characterization. Prot. Exp. Purif. 5, 583–586. 9. Carriere, F., Thirstrup, K., Hjorth, S., Ferrato, F., Nielsen, P. F., Withers-Martinez, C., Cambillau, C., Boel, E., Thim, L., and Verger, R. (1997) Pancreatic lipase structure–function relationships by domain exchange. Biochemistry 36, 239–248. 10. Rogalska, E., Cudrey, C., Ferrato, F., and Verger, R. (1993) Stereoselective hydrolysis of triglycerides by animal and microbial lipases. Chirality 5, 24–30. 11. DeCaro, A., Figarella, C., Amic, J., Michel, R., and Guy, O. (1977) Human pancreatic lipase: A glycoprotein. Biochim. Biophys. Acta 490, 411–419.
pepa
AP: PEP
13, 36–40 (1998)
PT980874
Human Pancreatic Triglyceride Lipase Expressed in Yeast Cells: Purification and Characterization Yanqing Yang* and Mark E. Lowe*,†,1 *Department of Pediatrics and †Department of Molecular Biology and Pharmacology, Washington University School of Medicine, One Children’s Place, St. Louis, Missouri 63110
A cDNA clone encoding human pancreatic triglyceride lipase was cloned into a yeast expression vector so that the yeast PHO1 signal peptide replaced the native signal peptide. Pichia pastoris cells were transfected with the vector, and clones expressing human pancreatic triglyceride lipase were isolated. Recombinant human pancreatic lipase was expressed in broth cultures and was purified from the medium by DEAE blue Sepharose and hydroxyapatite chromatography. The highly purified lipase had specific activities for various triglyceride substrates identical to those of tissuepurified human pancreatic triglyceride lipase; it was inhibited by bile salts, required colipase for activity, and demonstrated interfacial activation. This expression system is suitable for the rapid, efficient production of human pancreatic triglyceride lipase in amounts adequate for biophysical studies. q 1998 Academic Press
Human pancreatic triglyceride lipase is an enzyme of interest to physiologists, to clinicians, and to biochemists. It is central to the digestion of dietary fats. It is a serum marker for acute pancreatitis and it is the archetype of lipases, hydrophilic enzymes that act on water-insoluble substrates at an oil–water interface. The molecular details of lipase action have implications for the understanding of the many biological processes that occur at oil–water interfaces. Over the past 5 years, multiple laboratories have contributed to our understanding of the molecular mechanisms of lipolysis (1). Many of these studies have concentrated on pancreatic triglyceride lipase. The recently solved three-dimensional structures of human pancreatic triglyceride lipase in two different conformations provided a plausible model to explain the preference of this enzyme for water-insoluble substrates 1 To whom correspondence should be addressed. Fax: (314) 4544218. E-mail: [email protected].
over water-soluble substrates (2,3). These studies also identified candidates for the active site residues, the acyl chain-binding sites, and regions that interact with colipase, a protein cofactor of pancreatic triglyceride lipase. Site-specific mutagenesis of human pancreatic triglyceride lipase has been an important complement to the biophysical studies of the enzyme structure. The active site residues and the role of a surface lid in catalysis were detailed by characterizing mutant human pancreatic lipases (4,5). Mutagenesis studies provide important information about the relationships of structure to lipase function, but are hampered by the inability to quickly produce mutant proteins in available expression systems, particularly in quantities that are suitable for biophysical measurements on the mutants. Structure determinations on the mutants are critical components of any interpretation of kinetic results with the mutants and are necessary before the contribution of a particular amino acid or region can be fully understood. The requirements for a system that is easily and quickly manipulated, that processes and secretes mammalian proteins correctly, and that can produce large amounts of protein can potentially be met by expression in Pichia pastoris. We tested the ability of these yeast to express human pancreatic triglyceride lipase and found them suitable for producing milligram quantities of protein that is easily purified and that possesses enzymatic properties that are indistinguishable from human pancreatic lipase expressed in baculovirus-infected insect cells or the protein isolated from human pancreas. MATERIALS AND METHODS
Materials. The accompanying paper lists the sources for reagents (6). Transformation of yeast cells. The cDNA encoding human pancreatic lipase was subcloned into the pHILS1 expression vector. To accomplish the subcloning, an
36
AID
1046-5928/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
PEP 0874
/
6q1e$$$$61
05-11-98 10:36:28
pepa
AP: PEP
HUMAN PANCREATIC LIPASE EXPRESSION
EcoRI site was added to the 3*-end and an XhoI site was added to the 5*-end by polymerase chain reaction. The 5*-primer overlapped the sequence encoding the mature amino-terminus of the protein and the sequence encoding the carboxy-terminus of the yeast PHO1 signal peptide. This product, after being subcloned into the vector, replaced the native pancreatic triglyceride lipase signal peptide with the yeast PHO1 signal peptide and replaced the amino-terminal lysine with arginine to recreate the yeast signal peptidase cleavage site. The sequence of the cDNA was determined by dideoxynucleotide sequence analysis. The vector was linearized with Bgl II and transformed into GS115 by electroporation, and positive clones were selected as described (6). The medium from the methanol-induced cultures was analyzed by immunoblot and tributyrin assay as previously described (4,7). Expression of human pancreatic triglyceride lipase in yeast. Expression was accomplished as described in the accompanying paper with several minor exceptions (6). One liter of buffered minimal glycerol-complex medium divided into two 2-liter baffled flasks was inoculated with a 25-ml overnight culture and grown until the OD600 was between 6.0 and 10.0. The harvested cell pellet was resuspended in 200 ml of buffered minimal methanol-complex medium. Expression was continued for 4 to 5 days after methanol induction. Purification of human pancreatic triglyceride lipase. The medium was concentrated to 50 ml over a Filtron (Northborough, MA) Ultrasette tangential flow filter with a pore size of 30,000. The buffer was changed by dilution with 10 mM Tris–Cl, pH 8.0, and concentration with the Ultrasette. The sample was passed over a DEAE blue Sepharose column (1.5 1 10 cm) equilibrated in the same buffer (7). Human pancreatic lipase was collected in the pass-through from this column and dialyzed against 0.05 M NaPO4 , pH 7.0. The sample was applied to a CHT-I ceramic hydroxyapatite column from Bio-Rad (Richmond, CA) that was connected to a Pharmacia Akta Explorer chromatography system. The column was equilibrated in the dialysis buffer and lipase was eluted with a 10 column volume linear gradient from 0.05 to 0.35 M NaPO4 , pH 7.0. Elution was monitored at OD280 and activity was measured on each fraction. Lipase eluted as a single peak in the middle of the gradient. The fractions containing human pancreatic lipase were dialyzed against 10 mM Tris–Cl, pH 8.0, and 100 mM NaCl. After dialysis the sample was brought to 2 mM CaCl2 by adding the appropriate volume of 1 M CaCl2 . The recombinant lipase could be stored at 0807C for at least a year without loss of activity. Human pancreatic triglyceride lipase was purified from human pancreas and from baculovirus-infected
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
37
FIG. 1. Time course of lipase expression after methanol induction. Aliquots of the culture were removed every day after the switch to buffered minimal methanol-complex medium. The cells were removed by centrifugation and 20 ml of medium was assayed with tributyrin. 2 mg of colipase was added to each assay. The values are plotted as activity (mmol fatty acid released/min/20 ml of medium) versus day after induction.
Sf9 cells as previously described (7). Colipase was purified from human pancreatic tissue or recombinant colipase was produced in yeast and purified as described (6,8). Protein methods and lipase assay. Amino-terminus sequencing was performed at the Protein and Nucleic Acid Chemistry laboratory (Washington University School of Medicine, St. Louis) using an Applied Biosystems gas-phase sequencer. The presence of carbohydrate was detected with the Immun-Blot Kit for Glycoprotein Detection from Bio-Rad. The manufacturer’s instructions for total carbohydrate labeling in solution, protocol 2A, were followed. The labeled proteins were separated on SDS–PAGE and transferred to an Immobilon-P membrane (Millipore, Bedford, MA) as previously described (4). The blot was stained with a streptavidin–alkaline phosphatase conjugate following the Bio-Rad instructions. The concentration of purified human pancreatic triglyceride lipase and colipase were determined using an extinction coefficient at OD280 of E1% Å 1.2 and 0.3, respectively. Lipase activity was determined in the pH-STAT as previously described (7). Interfacial activation was determined against varying concentrations of tripropionin stabilized in 2% gum arabic (9). RESULTS AND DISCUSSION
Expression of human pancreatic triglyceride lipase. To determine the time course of expression, aliquots were removed before and each day following methanol induction. Lipase activity was not present prior to induction but was detected within 24 h after induction (Fig. 1). The activity rose slowly over the next 3 days.
pepa
AP: PEP
38
YANG AND LOWE
FIG. 2. Time course of lipase expression analyzed by SDS–polyacrylamide gel electrophoresis and immunoblot. 20-ml aliquots of medium from the samples described in the legend to Fig. 1 were mixed with 20 ml of 21 SDS sample buffer and heated at 1007C for 5 min. The samples were prepared in duplicate. Both were separated by SDS–polyacrylamide gel electrophoresis on two separate gels. One gel was stained with Coomassie blue (A) and the other was transferred to a membrane and stained with anti-human pancreatic triglyceride lipase antibody (B). The numbers below each lane are the days after induction. M, markers. The molecular weights of the markers are given to the left of Panel A. The arrows mark the position of lipase.
No lipase activity was detected in the medium from yeast transformed with a colipase vector, indicating that the yeast did not produce an endogenous, secreted lipase (6). Lipase was also detected by Coomassie blue staining of SDS–polyacrylamide gels (Fig. 2). A band migrating at the proper position for human pancreatic triglyceride lipase was the predominant band in the medium from 24 h after induction through at least 4 days. This band was not present in the medium prior to induction, indicating that protein expression is tightly regulated by methanol. The protein band stained with a polyclonal antibody against human pancreatic triglyceride lipase. Multiple yeast proteins also stained in the immunoblot, but most were not induced by methanol and were detected with preimmune serum (data not shown). The presence of lipase activity and the appearance of a protein of the molecular size of human pancreatic lipase that binds an antibody against hu-
man pancreatic triglyceride lipase demonstrate the expression of this protein in P. pastoris. Lipase was purified from the medium as described under Materials and Methods. The yield at each step of the isolation procedure is given in Table 1. SDS– polyacrylamide gel analysis of the medium (Fig. 2) suggested that the large amount of protein detected in the medium by the BCA assay was from the peptone added to the medium rather than from proteins secreted from the yeast. A 10-fold increase in purity was achieved by chromatography over DEAE blue Sepharose. This fraction still had a light brown color and SDS–PAGE analysis revealed that the majority of the protein was lipase but several minor contaminants were still present (data not shown). Both the color and the contaminating proteins were removed by hydroxylapatite chromatography. The fractions from this column contained a single, major protein band that had a specific activity for tributyrin comparable to that previously reported for purified human pancreatic triglyceride lipase (Table 1 and Fig. 3) (10). The amino-terminal sequence of the recombinant lipase was REVCYERLG, confirming that the recombinant protein is human pancreatic lipase and that the signal peptide was properly cleaved by the yeast signal peptidase. The tissue-purified lipase migrated as two major bands with several minor, faster migrating bands also present (Fig. 3). All of these bands reacted with antibody to human pancreatic triglyceride lipase and the smaller fragments probably represent degradation products (data not shown). Although it was possible that the difference in size between the two major bands was also caused by proteolysis, it seemed likely that the size difference resulted from posttranslational processing. Only one amino-terminal sequence was present, KEVCYERLG, demonstrating that there was no internal cleavage of the peptide chain. The preparation had a specific activity for several different substrates that was comparable to those reported by others, suggesting that the two species had identical activities (see below). Secreted human pancreatic triglyceride lipase is glycosylated (11). Thus, we determined whether glycosyla-
TABLE 1
Purification of Recombinant Human Pancreatic Triglyceride Lipase from Yeast Cell Medium Purification step
Volume (ml)
Total protein (mg)
Total activity (Units)
Specific activity (Units/mg)
% Yield
Medium DEAE Hydroxylapatite
200 75 23
205 32.8 15.2
1.1 1 108 1.8 1 108 1.2 1 108
537 5488 7895
100 164 109
Note. Activity was determined as described under Materials and Methods and is expressed in units of mmol fatty acid released/min. A 5 M excess of colipase was present in the assay.
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
pepa
AP: PEP
HUMAN PANCREATIC LIPASE EXPRESSION
FIG. 3. SDS–polyacrylamide gel electrophoresis of purified recombinant and tissue-purified human pancreatic triglyceride lipase. The proteins were purified as described under Materials and Methods. 10 mg of yeast expressed lipase and 20 mg of tissue-purified lipase were dissolved in SDS sample buffer and heated at 1007C for 5 min. The gel was stained with Coomassie blue. M, markers; 1, yeast expressed lipase; 2, tissue-purified lipase.
tion accounted for the size difference between the two bands. We oxidized any carbohydrate residues on tissue-purified and recombinant human pancreatic triglyceride lipase and reacted with biotin. The samples were run on SDS–polyacrylamide gels and were transferred to a membrane. The presence of carbohydrate was detected with streptavidin–alkaline phosphatase and color development reagents. The upper bands in tissue-purified and recombinant lipase were detected by this method, indicating that they were glycosylated (Fig. 4). The lower band in the tissue-purified lipase was not detected. These results suggest that the two forms of lipase present in the tissue preparation differ by the absence or presence of an oligosaccharide chain rather than proteolytic cleavage.
39
FIG. 4. Glycoprotein staining of purified recombinant and tissuepurified human pancreatic triglyceride lipase. The proteins were processed as described under Materials and Methods. An equal volume of 21 SDS sample buffer was added to each sample and the samples were heated at 1007C for 5 min. The samples were separated by SDS– polyacrylamide gel electrophoresis and transferred to an Immobilon membrane. The membranes were stained as described under Materials and Methods. Lanes 1, 10 mg ovalbumin; 2, 10 mg tissue-purified lipase; 3, 5 mg yeast expressed lipase; 4, 20 mg albumin; 5, 20 mg yeast expressed lipase; 6, 40 mg tissue-purified lipase. M, markers. The arrows mark the position of lipase.
Activity of recombinant human pancreatic triglyceride lipase. The activity of the yeast expressed lipase was compared to that of tissue-purified lipase. Substrates with different acyl chain lengths were tested in the presence of 4 mM taurodeoxycholate (Table 2). No activity was detected in the absence of colipase, indicating that the recombinant lipase was inhibited by bile salts. When colipase was added, recombinant lipase had specific activities for all three substrates
TABLE 2
Specific Activity of Human Pancreatic Triglyceride Lipase from Various Sources Specific activity (mmol fatty acid/min/mg) Lipase
Tributyrin
Trioctanoin
Triolein
Tissue-purifieda Baculovirusa P. pastoris
8200 { 630 7720 { 550 8028 { 780
2579 { 180 2480 { 230 2451 { 175
1650 { 85 1500 { 90 1716 { 60
Note. Activity was determined as described under Materials and Methods. a Values taken from Ref. (7).
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
FIG. 5. Interfacial activation of yeast expressed pancreatic triglyceride lipase. 2.5 mg of yeast expressed lipase and 1 mg of colipase were incubated with various amounts of tripropionin as described under Materials and Methods. Assays were done in the pH STAT. Activity is mmol fatty acid released/min. The arrow indicates the solubility limit of tripropionin as determined by light scattering.
pepa
AP: PEP
40
YANG AND LOWE
that were indistinguishable from the values for tissue-purified and baculovirus expressed lipase. These data demonstrate that the recombinant human pancreatic triglyceride lipase faithfully reproduces the important properties of bile salt inhibition, colipase dependence, and substrate specificity found in tissuepurified lipase. Another important kinetic property of pancreatic triglyceride lipase is the preference for water-insoluble substrates over water-soluble substrates. The large increase in activity seen when lipase encounters an interface, a property termed interfacial activation, is characteristic of this enzyme. We determined whether the recombinant lipase showed interfacial activation against tripropionin (Fig. 5). Recombinant human pancreatic triglyceride lipase had low levels of activity until the solubility of tripropionin was exceeded. When water-insoluble substrate particles formed, the activity increased markedly. The yeast-produced lipase retained the property of interfacial activation. Conclusion. Human pancreatic triglyceride lipase was successfully produced in a yeast expression system. The best expression was obtained by replacing the native signal peptide with a yeast signal peptide. This required changing the mature amino-terminus from lysine to arginine. There was no detectable effect of this change on the enzymatic properties of the recombinant lipase. The ability to easily manipulate and grow the yeast and to produce milligram quantities of protein in small culture volumes makes this system useful for expressing native and mutant human pancreatic triglyceride lipase.
AID
PEP 0874
/
6q1e$$$$62
05-11-98 10:36:28
ACKNOWLEDGMENT This work was supported by NIH Grant HD/DK 33060.
REFERENCES 1. Lowe, M. E. (1997) Structure and function of pancreatic lipase and colipase. Annu. Rev. Nutr. 17, 141–158. 2. van Tilbeurgh, H., Sarda, L., Verger, R., and Cambillau, C. (1992) Structure of the pancreatic lipase–procolipase complex. Nature 359, 159–162. 3. van Tilbeurgh, H., Egloff, M. P., Martinez, C., Rugani, N., Verger, R., and Cambillau, C. (1993) Interfacial activation of the lipase–procolipase complex by mixed micelles revealed by X-ray crystallography. Nature 362, 814–820. 4. Lowe, M. E. (1992) The catalytic site residues and interfacial binding of human pancreatic lipase. J. Biol. Chem. 267, 17069– 17073. 5. Jennens, M. L., and Lowe, M. E. (1994) A surface loop covering the active site of human pancreatic lipase influences interfacial activation and lipid binding. J. Biol. Chem. 269, 25470–25474. 6. Cordle, R. A., and Lowe, M. E. (1997) Purification and characterization of human procolipase expressed in yeast cells. Prot. Exp. Purif., 13, 30–35. 7. Lowe, M. E. (1996) Mutation of the catalytic site Asp177 to Glu177 in human pancreatic lipase produces an active lipase with increased sensitivity to proteases. Biochim. Biophys. Acta 1302, 177–183. 8. Lowe, M. E. (1994) Human pancreatic procolipase expressed in insect cells: Purification and characterization. Prot. Exp. Purif. 5, 583–586. 9. Carriere, F., Thirstrup, K., Hjorth, S., Ferrato, F., Nielsen, P. F., Withers-Martinez, C., Cambillau, C., Boel, E., Thim, L., and Verger, R. (1997) Pancreatic lipase structure–function relationships by domain exchange. Biochemistry 36, 239–248. 10. Rogalska, E., Cudrey, C., Ferrato, F., and Verger, R. (1993) Stereoselective hydrolysis of triglycerides by animal and microbial lipases. Chirality 5, 24–30. 11. DeCaro, A., Figarella, C., Amic, J., Michel, R., and Guy, O. (1977) Human pancreatic lipase: A glycoprotein. Biochim. Biophys. Acta 490, 411–419.
pepa
AP: PEP