This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.
Agricola
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, February 2003, p. 1108-1113, Vol. 69, No. 2
0099-2240/03/$08.00+0     DOI: 10.1128/AEM.69.2.1108-1113.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.

A Single Mutation in the Activation Site of Bovine Trypsinogen Enhances Its Accumulation in the Fermentation Broth of the Yeast Pichia pastoris

José Hanquier,1 Yannick Sorlet,2 Dominique Desplancq,2 Laurence Baroche,2 Marc Ebtinger,3 Jean-François Lefèvre,2 Franc Pattus,2 Charles L. Hershberger,1 and Alain A. Vertès3*

Lilly Research Laboratories, Lilly Corporate Center, Eli Lilly & Co., Indianapolis, Indiana 46285 ,1 GIB-ESBS, F-67400 Illkirch,2 Lilly France, 67642 Fegersheim Cedex, France3

Received 4 September 2002/ Accepted 8 November 2002


arrow
ABSTRACT
 
We produced bovine trypsinogen in the yeast Pichia pastoris. Little or no trypsinogen was detected when the gene with its native leader sequence was expressed under the control of the strong aox1 promoter, suggesting that expression of the wild-type bovine trypsinogen was toxic to the cells. We altered the trypsinogen native propeptide sequence by replacing the lysine at position 6 with an aspartic acid, thus destroying the site in the propeptide cleaved by enterokinase and by trypsin. This mutant accumulated up to 10 mg of trypsinogen per liter in shake flask cultures and about 40 mg/liter in 6-liter fermentors. Trypsinogen could be activated in vitro with a dipeptidyl-aminopeptidase, which selectively removed the modified trypsinogen propeptide; the resulting trypsin was fully active and showed evidence of glycosylation. Thus, we have developed a novel protein production scheme that can be used for the expression of proteins, such as proteases, that are deleterious to the producing organism. This system relies on the expression of a zymogen that cannot be activated in vivo coupled with its in vitro purification and activation.


arrow
INTRODUCTION
 
Trypsin is a serine protease that plays a key role in the activation cascade of pancreatic enzymes involved in digestion. The trypsin precursor secreted by the pancreas is an inactive zymogen, trypsinogen, which is activated in the duodenum by enterokinase. Enterokinase recognizes the (Asp)4-Lys sequence in the propeptide and cleaves after the lysine residue to release active trypsin (36). Removal of the propeptide results in conformational changes that lead to the structural formation of an active site (22). Six disulfide bridges are present in trypsin (25). Trypsin cleaves peptide bonds at the carboxyl-terminal end of lysine and arginine residues and can activate its own zymogen. This self-activation is the in vivo basis of the activation amplification cascade.

Bovine trypsin is widely used in mammalian cell culture applications (trypsinization) and is an important component of several industrial processes in the pharmaceutical industry (5, 17). At present, bovine trypsin is produced by extraction from cow pancreas. However, the emergence of diseases such as bovine spongiform encephalopathies has raised concerns about the use of products of animal origin in industrial processes, especially in the pharmaceutical industry. Thus, alternative industrial processes are now sought to circumvent this risk. For example, Casolari (4) suggested changes in food processing technology to inactivate the prion. Furthermore, the European Union has proposed regulations to minimize the risk of transmitting animal spongiform encephalopathy agents via medicinal products (Annex to the Council Directive 75/318/EEC, Commission of the European Communities, Directorate General for Industry). Consequently, there is increased interest in manufacturing bovine trypsin by fermentation of a microbial host to provide industrial quantities of this enzyme.

Escherichia coli has been the host of choice for the production of rat trypsin, and altered forms of rat trypsin have been reported to be produced at levels of as high as 1 mg/liter (15). More recently, rat trypsinogen was produced in Saccharomyces cerevisiae with yields of up to 10 mg/liter (20, 38). However, human trypsinogen is insoluble when expressed in E. coli (28), and expression of bovine trypsinogen in E. coli results in the formation of inclusion bodies (C. L. Hershberger, unpublished observation). These differences in solubility of trypsinogens originating from various organisms in E. coli may be attributed to differences at the amino acid sequence level, as was demonstrated for antibody fragments (24).

The methylotrophic yeast Pichia pastoris is an attractive alternative to E. coli. Extensive development of P. pastoris biotechnology over the past 10 years now permits the construction of efficient host-vector systems for heterologous protein expression. A number of proteins that are insoluble in E. coli have been expressed in an active and soluble form in P. pastoris (23), as have various eukaryotic proteins (7, 9, 13, 26, 29, 30, 32, 39). P. pastoris possesses the complex eukaryotic secretion machinery and can secrete proteins containing a high number of disulfide bridges. Moreover, high P. pastoris cell densities can be attained in inexpensive media (11), and scale-up to industrial processes is relatively straightforward (43).

The methanol-inducible aox1 promoter, which controls the expression of the alcohol oxidase AOX1 in P. pastoris (14), and the strong glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter (12) have been preferentially used for the expression of heterologous proteins. In the presence of methanol, the AOX1 protein represents up to 30% of total cellular protein (14). The GAPDH promoter has been reported to be repressed in the presence of glucose and induced under conditions of glucose limitations. This expression pattern confers an important industrial advantage over expression regulated by the aox promoter, since methanol is a toxic, flammable compound (8, 10, 12, 31, 35, 41).

Our objectives in this study were to express and secrete bovine trypsinogen by using P. pastoris. The main hypothesis tested was that disabling in vivo activation of the zymogen would suffice to enable production of a deleterious protein and that in vitro activation was an efficient method to produce active trypsin. In addition to providing a novel heterologous production scheme as well as a source of bovine trypsin not derived from animal raw materials, this work results in the availability of trypsin totally devoid of contaminating pancreatic enzymes, such as bovine chymotrypsin.


arrow
MATERIALS AND METHODS
 
Host cells, vectors, and chemicals.
Plasmid pPIC9 and P. pastoris GS115 (his4) were purchased from Invitrogen (De Shelp, The Netherlands). Vectors pRMG5 and pLDG43 (16) and the P. pastoris protease mutant SMD1163 (his4 pep4) were from the Eli Lilly (Indianapolis, Ind.) laboratory collection. The gene coding for the bovine trypsinogen used in this study was derived from published sequences (SwissProt accession number Q29463). The portion of pHKY603 containing the promoter-operator sequence and the bovine trypsinogen gene was sequenced after cloning to rule out any sequence alteration. E. coli XL1-Blue was used to propagate pPIC9 expression vectors. Medium components were purchased from Difco (Detroit, Mich.). Restriction enzymes were purchased from New England Biolabs (Beverley, Mass.). Oligonucleotides were synthesized by Eurogentec (Liège, Belgium). The dipeptidyl-diaminopeptidase (dDAP) originating from Dictyostelium discoideum was obtained from Lilly France. Trypsin (T8918) and trypsinogen (T1143) standards were obtained from Sigma (St. Quentin Fallavier, France).

Mutation of the trypsinogen leader sequence.
The bovine trypsinogen gene is available as an NdeI-BamHI cassette in vector pRMG5, a pSK derivative plasmid from the Eli Lilly laboratory collection. The {alpha}-mating factor signal sequence (3) was inserted into this plasmid upstream of the trypsinogen-coding sequence by cloning an oligonucleotide cassette containing an XhoI site upstream of the signal sequence. A sequence corresponding to the amino-terminal region of the bovine trypsinogen gene (SwissProt accession number Q29463) up to the NarI restriction site was subsequently inserted. The oligonucleotides were phosphorylated, annealed, and inserted into pRMG5 following digestion with XhoI and NarI and dephosphorylation (Fig. 1). The nucleotide sequence of the junction of the {alpha}-mating factor signal sequence with the trypsinogen gene was confirmed by DNA sequencing. The trypsinogen gene together with the leader sequence thus inserted was then subcloned in plasmid pPIC9 as an XhoI-XbaI fragment ligated to XhoI- and AvrII-digested dephosphorylated vector DNA (Fig. 1). The oligonucleotides TCGAG AAA AGA GTC GAC GAT GAT GAC AAG ATC GTC GGA GGT TAT ACA TGT GG and CGCC ACA TGT ATA ACC TCC AAC GAT CTT GTC ATC ATC GTC GAC TCT TTT were used to construct the control plasmid with the native trypsinogen sequence. The half XhoI and NarI sites, respectively, are indicated in boldface, and the nucleotides represented by boldface underlined characters are at the sites that were mutated in the sequence of plasmid pPIC9VD5NADP. A similar scheme was used to construct the mutated signal sequence, by using the following oligonucleotides: TCGAG AAA AGA GTC GAC GAT GAT GAC GAT ATC GTC GGA GGT TAT ACA TGT GG and CGCC ACA TGT ATA ACC TCC AAC GAT ATC GTC ATC ATC GTC GAC TCT TTT (for plasmid pPIC9VD5NADP).



View larger version (21K):
[in this window]
[in a new window]
  		FIG. 1. Constructs for trypsin production. (A) Expression vectors pPIC9 and pLGD43. Plasmid pPIC9 (Invitrogen) includes the following elements: the P. pastoris alcohol oxidase 1 promoter (AOX1), the {alpha}-mating factor signal sequence ({alpha}MF), the transcription terminator and polyadenylation signal from the alcohol oxidase 1 gene (3'AOX1TT), the histidinol dehydrogenase gene (HIS4), and sequences of the alcohol oxidase gene 3' from the transcription terminator (3'AOX1). Plasmid pLGD43 includes the following elements: the GAPDH promoter, the human serum albumin signal sequence (HSASS), the transcription terminator and polyadenylation signal from the alcohol oxidase 1 gene (3'AOX1TT), the histidinol dehydrogenase gene (HIS4), and sequences of the alcohol oxidase gene 3' from the transcription terminator (3'AOX1). Tpn, site of insertion of the trypsinogen gene. (B) Sequences of the trypsinogen variants. Boxes indicate the trypsinogen propeptide, and arrows show the cleavage sites of enterokinase, trypsin, and dDAP. The mutation Lys (AAG) to Asp (GAT) is indicated in boldface.

Transformation of P. pastoris.
Expression vectors encoding wild-type and VD5-trypsinogen were linearized by digestion with BglII to favor integration by homologous recombination at the aoxI locus. P. pastoris strains were transformed by electroporation (21). VD5 refers to the amino terminus of the mutated trypsinogen [Val-(Asp)5 versus Val-(Asp)4-Lys]. Integrants (his+) were selected at 30°C on MD agar without histidine (13.4 g of Bacto yeast nitrogen base per liter, 4 µg of biotin per liter, 10 g of dextrose per liter, 15 g of agar per liter). Determination of the MutS (methanol utilization slow) phenotype was performed on MD agar containing 0.5% methanol as the sole carbon source. The wild-type strain was used as a control for determination of the MutS phenotype. Approximately 14% of the his+ clones grow slowly on MD plus methanol, indicating gene replacement at the aox1 locus.

Quantitative PCR.
The presence of the trypsinogen-encoding gene on the chromosome was verified by quantitative PCR with recombinant P. pastoris chromosomal DNA as a template. This experiment also suggested that the trypsinogen gene is present at no more than one copy per chromosome. The forward primer used in this experiment was 5'-CGT CTG GGC GAG GAT AAC AC-3', the reverse primer was 5'-GAC TTG GAT GCG GAG ATG AAC-3', and the probe was 5'-6-FAM-GCT CAT TGC CCT CCA CGA CGT TG-TAMRA-3', which was labeled with the fluorescent dyes 6-carboxyfluorescein (6-FAM) and 6-carboxytetramethylrhodamine (TAMRA). These primers were synthesized by Synthegen (Houston, Tex.). All quantitative PCRs were based on the TaqMan Universal PCR Master Mix protocol (Applied Biosystems). Reactions were run in duplicate in adjacent wells. Final concentrations of all primers were 200 nM, and those of all probes were 100 nM. The following thermal cycle was used: 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Data were collected by the ABI Prism 7700 sequence detection system.

Shake flask fermentation.
Clones were stored at -80°C as 1-ml aliquots in YES-glycerol medium (20 g of Bacto Soytone per liter, 5 g of yeast extract per liter, 0.5 g of NaCl per liter, 5 g of glucose per liter, 10% glycerol) at a cell density corresponding to an optical density at 600 nm of approximately 20. One aliquot was thawed to inoculate 75 ml of BMGY medium (100 mM phosphate buffer [pH 6], 4 µg of biotin per liter, 13.4 g of Bacto yeast nitrogen base per liter, 10 g of yeast extract per liter, 20 g of Bacto Peptone per liter, 1% glycerol). The culture was grown for 24 h at 30°C and 250 rpm. Cells in which the trypsinogen gene is under control of the aox1 promoter were harvested by centrifugation and resuspended to an optical density at 600 nm of approximately 60 in 25 ml of BMMY medium (BMGY medium plus 0.5% methanol). To enable methanol induction of the aox1 promoter, 125 µl of methanol was then added and incubation was continued for an additional 48 h. The culture supernatant was subsequently harvested by centrifugation, and VD5-trypsin was assayed.

Recombinant clones in which the trypsinogen gene is under control of the GAPDH promoter were first grown in glucose to generate cell mass and subsequently transferred to a medium containing glycerol as a sole carbon source, since the GAPDH promoter is induced only when glucose is limited in the culture medium. For these recombinant strains, culture conditions and protocols were identical to those used for the aox1 promoter-controlled clones, except that cells were first grown in BMGlcY (same as BMGY but glycerol is replaced by 5% glucose). Cells were subsequently transferred to BMGY medium to allow for induction of the GAPDH promoter-controlled trypsinogen gene and grown for 48 h. The culture supernatant was subsequently harvested by centrifugation, and VD5-trypsin was assayed.

Pilot scale fermentation.
A 6-liter bench top fermentor (LSL Biolafitte, St-Germain-en-Laye, France) was autoclaved with 2.5 to 3.2 liters of BSM medium (CaSO4 · 2H2O, 0.93 g/liter; K2SO4, 18.2 g/liter; MgSO4 · 7H2O, 14.9 g/liter; KOH, 4.1 g/liter; NH4Cl, 4.3 g/liter; trisodium-citrate · 2H2O, 1.5 g/liter). In order to favor optimal cell growth, the pH was adjusted to 5.4 with 25% NH4OH or 45% H3PO4. The substrate mix was aseptically added to reach final concentrations of 5% glycerol, 4 ml of PTM solution [CuSO4 · 5H2O, 6.0 g/liter; Fe2(SO4)3 · 6H2O, 65 g/liter; ZnCl2, 20 g/liter; MnSO4 · H2O, 3 g/liter; CoCl2, 0.5 g/liter; NaMoO4 · 2H2O, 0.24 g/liter; NaI, 80 mg/liter; H3BO3 · 20 mg/liter; 97% H2SO4, 5 ml/liter), and 1 mg of biotin per liter. The fermentor was inoculated with a 24-h P. pastoris culture at approximately 10% of the initial fermentation volume. The temperature was set at 29.5°C. After exhaustion of the initial glucose in approximately 24 h, as judged from the drop in O2 consumption, a continuous 50% (vol/vol) glycerol feed containing 12 ml of PTM solution per liter and 3 mg of biotin per liter was started at approximately 15 ml/h per liter of culture. This feed strategy was designed mainly as a means to accumulate biomass, while GAPDH promoter-controlled trypsinogen production was allowed to occur. To decrease biomass growth yield and thereby channel more of the carbon into recombinant protein production, the pH was allowed to vary from 5.4 to 3 as a result of cellular metabolism. Following 4 to 6 h of high-glycerol feeding, the trypsinogen production rate was increased by decreasing the carbon source feed to 3.0 to 3.5 ml/h per liter of culture. This phase of production was continued for 48 to 120 h.

Quantitative VD5-trypsin assay.
VD5-trypsin production was measured by purifying VD5-trypsin in batch on SP Toyopearl 550C resin (TosoHaas, Stuttgart, Germany). Five milliliters of supernatant was diluted 10-fold in 10 mM acetate buffer, pH 3.5. The pH of the resulting solution was adjusted to 3.0 with 10 M acetic acid. The VD5-trypsin solution was incubated overnight at 4°C under gentle rotatory stirring in the presence of 125 µl of SP Toyopearl 550C resin equilibrated in 10 mM acetate buffer, pH 3.0. Beads were washed twice with 50 ml of 10 mM acetate buffer, pH 3.0. Protein was eluted by incubating the resin twice with 500 µl of 10 mM acetate buffer (pH 3.0), supplemented with a final concentration of 0.5 M NaCl, for 10 min with gentle stirring. The two eluates were pooled. The yield of the purification step was measured by using standard solutions containing known amounts of purified trypsinogen. A total of 50 µl of eluate was diluted twofold in 10 mM acetate buffer (pH 3.0) and activated overnight at 25°C with 3 µl of dDAP at 40 U/liter.

Trypsin-specific activity assay.
Trypsin-specific activity was measured with a TAME assay (42) as follows: 33 µl of trypsin solution was added to 867 µl of TAME buffer (46 mM Tris [pH 8.1], 11.5 mM CaCl2, 0.44 mg of p-toluene-sulfonyl-L-arginine methyl ester per ml) and incubated at 25°C, and the kinetics were monitored for 2 min at 247 nm.

Preparative VD5-trypsin purification.
The VD5-trypsin secreted in the culture supernatant was concentrated at 4°C by ammonium sulfate precipitation at 70% saturation. The precipitate was resuspended in 20 ml of 10 mM sodium acetate (pH 5) and dialyzed overnight against the same buffer at 4°C. The dialyzed solution was subsequently loaded onto a CM-Sepharose CL6B column (Pharmacia, Uppsala, Sweden) equilibrated in 10 mM acetate buffer, pH 5. The VD5-trypsin was eluted with an NaCl gradient from 10 to 200 mM. Fractions containing VD5-trypsin were pooled and concentrated on Macrosep MF centrifugal filters (Pall Filtron Corp., Northborough, Mass.).

Preparative VD5-trypsin activation and trypsin purification.
Trypsin was obtained by digesting purified VD5-trypsin in acetate buffer with 500 U of dDAP per g of trypsinogen overnight at 25°C. The pH of the resulting digest was adjusted to 8.1 with 10% NaOH, and its NaCl concentration was brought to 1 M. The resulting solution was loaded on a benzamidine-Sepharose CL6B column (Pharmacia) equilibrated with 50 mM Tris (pH 8.1)-100 mM NaCl-50 mM CaCl2. The column was washed with 100 mM acetate buffer (pH 3)-100 mM NaCl-50 mM CaCl2, and trypsin was eluted with 100 mM acetate buffer (pH 3)-1 M NaCl-50 mM CaCl2.

PNGase F treatment.
N-Endoglycopeptidase F (PNGase F) was purchased from New England Biolabs (Beverly, Mass.) and used according to the manufacturer's instructions, as follows. Ten milligrams of purified recombinant trypsin was ethanol precipitated and heated at 100°C for 10 min in 10 ml of denaturation buffer (0.5% sodium dodecyl sulfate [SDS], 1% ß-mercaptoethanol). Denatured protein samples were subsequently incubated for 1 h at 37°C in 50 mM sodium phosphate buffer (pH 7.5) supplemented with 1% NP-40 and 2 U of PNGase F in a total reaction volume of 20 ml. Deglycosylated samples were analyzed by SDS-12% polyacrylamide gel electrophoresis (SDS-12% PAGE).

Electrophoresis and Western blotting.
SDS-PAGE was carried out under reducing conditions. Proteins were stained with Coomassie blue or silver stain according to standard procedures (33). For Western blotting, proteins were separated by SDS-12% PAGE and transferred to nylon membranes (Millipore Corp., Bedford, Mass.). The membranes were probed with a rabbit polyclonal antitrypsin antibody (Research Diagnostic Inc., Flanders, N.J.) at a 1:5,000 dilution in TBS buffer (50 mM Tris [pH 8.0], 150 mM NaCl). The VD5-trypsin was identified with a polyclonal goat anti-rabbit immunoglobulin G-horseradish peroxidase conjugate by using the ECL detection system (Amersham, Little Chalfont, England). Immunoelectrofocusing (pH 3 to 10) was performed with the Bio-Rad isoelectric focusing system according to the protocol recommended by the manufacturer.


arrow
RESULTS
 
Expression of a trypsinogen analog not activatable by trypsin.
Initially we attempted to secrete wild-type trypsinogen from P. pastoris. We recovered very low levels of tryptic activity and low levels of trypsin instead of trypsinogen, as demonstrated by Western blot analysis (data not shown). We also detected numerous proteins that were not observed when the untransformed parental strain was grown, suggesting that tryptic activity was causing cellular damage and the release of intracellular proteins.

To prevent the activation of trypsinogen to trypsin, we altered the DNA sequence corresponding to the amino terminus of the trypsinogen gene to destroy the trypsinogen site cleaved by enterokinase and by trypsin. Quantitative PCR suggested that the modified trypsinogen gene was present at no more than one copy per chromosome. The engineered trypsinogen could not be activated in vivo but could be activated in vitro, for example, by using a diamino-exopeptidase. The lysine in position 6 was replaced by an aspartic acid, resulting in the change of the wild-type sequence from VD4K to VD5. We observed two discrete bands in Western blots, corresponding to proteins with apparent molecular masses of 28 and 24 kDa, when the supernatant of cells carrying plasmid pPIC9VD5NADP grown under induced conditions was analyzed. No bands were observed in the supernatant of cells carrying pPIC9NADP, which encodes the wild-type bovine trypsinogen. The film was overexposed to enable the detection of even very faint signals, and the two bands appear blurred for this reason (Fig. 2). The 24-kDa VD5-trypsin band, which is the approximate size of the expected trypsinogen analog, was relatively faint at moderate exposures (not shown), while the 28-kDa band was much more intense, and they are clearly distinct in silver-stained gels (see Fig. 4, lane 1). Yields of 5 to 10 mg/liter could be routinely achieved in shake flasks when cultures reached 10 g of biomass (wet weight) per 100 ml. VD5-trypsin did not accumulate intracellularly, as demonstrated by Western blots of whole-cell protein extracts of P. pastoris strain GS115 carrying pPIC9VD5NADP.



View larger version (104K):
[in this window]
[in a new window]
  		FIG. 2. Western blotting to detect VD5-trypsin in P. pastoris supernatant. Cultures of P. pastoris GS115 (lanes 1 and 2) or GS115 transformed with pPIC9NADP (lanes 3 and 4) or pPIC9VD5NADP (lanes 5 and 6) were induced with methanol. Ten microliters of culture supernatant harvested prior to induction and following 48 h of induction was resolved by SDS-12% PAGE. Lanes 1, 3, and 5, noninduced conditions; lanes 2, 4, and 6, induced conditions. The rabbit immunoglobulin G recognizes both trypsin and VD5-trypsin. The Western blot was overexposed to detect faint signals, and the two discrete bands in lane 6 appear blurred for this reason.



View larger version (52K):
[in this window]
[in a new window]
  		FIG. 4. The VD5-trypsin produced in P. pastoris is glycosylated. Bovine VD5-trypsin produced by high-cell-density fermentation of P. pastoris and activated by dDAP digestion was subjected to PNGase F digestion (1 U/10 µg of recombinant trypsin) as described in Material and Methods. Treated and untreated samples were analyzed by reducing SDS-12% PAGE, and the proteins (2 µg of protein was loaded per lane) were visualized by silver staining. The N-glycosylated trypsin in the untreated sample migrates at approximately 28 kDa (lane 1), while unglycosylated and deglycosylated trypsin migrate at approximately 24 kDa (lane 2). The numbers on the left of the gel represent the positions of size standards present in the 10-kDa molecular mass marker (Gibco BRL). The band originating from glycosylated recombinant trypsin is the major band in lane 1 and the minor band indicated by an arrow in lane 2; the major band in lane 2 originates mainly from deglycosylated recombinant trypsin.

Fermentation to produce recombinant VD5-trypsin.
Since our objective was to enable production of trypsin on an industrial scale, VD5-trypsin was placed under the control of the GAPDH promoter by cloning the VD5-trypsin-coding sequence in plasmid pLGD43 (Fig. 1) downstream of the human serum albumin signal sequence. Shake flask fermentation results from 45 pPIC9-derived clones expressing VD5-trypsin from the aox1 promoter were compared to those from shake flask fermentations of 15 pLGD43-derived clones expressing VD5-trypsin from the GAPDH promoter. The best producers of each pool were significantly different in expression yields. While the comparison between the two promoters is not unambiguous, since different secretion signals were used, both the {alpha}-mating factor and the human serum albumin signal sequences were correctly recognized and processed by the P. pastoris signal peptidases. One of the GAPDH promoter-controlled clones, P. pastoris EL336, was selected for the fermentation study and characterized further.

Fermentation was performed in a three-step process including an initial batch phase in 5% glucose. This initial phase was followed by a glycerol-based high-feed batch fermentation to serve as a biomass growth phase, but which also was required for trypsin production driven by the GAPDH promoter. Finally, there was a low-feed production phase characterized by a very low growth rate in which glycerol also was the main carbon source (Fig. 3). The trypsin production rate was significantly increased during the third phase compared to that during the second phase. Typically, yields of 33 mg of VD5-trypsin per liter were attained in a 6-liter fermentor at a cell density of 60 g (dry weight)/liter.



View larger version (24K):
[in this window]
[in a new window]
  		FIG. 3. High-cell-density production of bovine VD5-trypsin by P. pastoris. The dry cell weight (dcw) and bovine VD5-trypsin content were determined throughout the fermentation. The three stages of the fermentation process are indicated beneath the curve. An initial fed-batch fermentation was started with 5% glucose and conducted at pH 5.4 for 24 h. Glycerol feeding was initiated at a high rate (>5 g/h per liter) to further increase the biomass, allowing the pH of the medium to decrease to 3.0 as a result of cellular metabolism. Trypsinogen production driven by the GAPDH promoter was nevertheless concomitant with the biomass growth phase, as expected. Once the culture reached a cell density of 55 g (dry cell weight)/liter, the glycerol feeding rate was reduced to 1 g/h per liter in order to further divert the carbon flow from cell growth to trypsinogen production, which lasted for 48 h. Error bars are omitted, as the error is less than 10% of the value of the points. Broken line, dry cell weight; solid line, trypsinogen. High, initiation of the high glycerol feed rate; low, initiation of the low glycerol feed rate.

Activation of the non-trypsin-activatable trypsinogen.
Recombinant VD5-trypsin was activated to trypsin by removal of the propeptide Val-Asp5 (Fig. 1) with dDAP. In the absence of dDAP, no trypsin activity was detected. By using 0.4 U of dDAP per mg of VD5-trypsin, the yield of active trypsin approached 100% after 24 h. The enzymatic activity of recombinant trypsin remained constant for 24 h at 25°C following dDAP activation (dDAP not removed), demonstrating that trypsin produced in P. pastoris was stable under these conditions. The amino terminus of the dDAP-activated trypsin was identical to that of native trypsin extracted from cow pancreas. The specific activity of recombinant trypsin was determined following affinity purification on a benzamidine-Sepharose column. The recombinant enzyme and extracted enzyme (bovine trypsinogen from Sigma, activated with the same protocol) had comparable activities (data not shown), confirming that the properties of the recombinant enzyme are equivalent to the properties of the natural enzyme.

Characterization of the nonactivatable trypsin produced in P. pastoris.
We purified VD5-trypsinogen from the clear supernatant of cultures carrying the trypsinogen-encoding derivatives of plasmid pLGD43 or pPIC9. We sequenced the first 10 N-terminal amino acids from both the 28- and 24-kDa VD5-trypsin variants. In both cases, these trypsin species had the same amino terminus as the native protein, demonstrating the correct processing of the {alpha}-mating factor and the human serum albumin signal sequences (Fig. 1). Activated 24- and 28-kDa VD5-trypsin species were treated with PNGase F. Following treatment, the 28-kDa species comigrated with the untreated 24-kDa species (Fig. 4), whereas the migration of the 24-kDa species was unaffected by the PNGase F treatment.


arrow
DISCUSSION
 
We demonstrated that inhibiting the in vivo activation pathway of a zymogen form of a deleterious protein can enable its production in a heterologous host. The zymogen is subsequently activated in vitro. Inhibition of the natural activation of trypsinogen to trypsin by P. pastoris-derived proteases enabled the production of milligram amounts of VD5-trypsin by this methylotrophic yeast. We abolished in vivo activation by modifying the amino acid sequence of the zymogen without affecting the amino acid sequence of the mature trypsin. A single mutation changing the lysine residue at position 6 to an aspartic acid was sufficient to destroy the activation site cleaved by enterokinase and by trypsin. Two forms of the VD5-trypsin were observed by electrophoresis, which corresponded to proteins with apparent molecular masses of 24 and 28 kDa.

The 24- and 28-kDa proteins could differ in the amount of glycosylation. P. pastoris often glycosylates recombinant proteins (1, 19, 34, 37, 40). Montesino et al. (27) found that carbohydrates on heterologous proteins from various microorganisms expressed in P. pastoris vary in size, with Man8GlcNAc2 and Man9GlcNAc2 structures the most frequent. Nevertheless, short oligomannosides and much longer structures up to Man18GlcNAc2 also were observed. The observation that migration of the 24-kDa species was unaffected by PNGase treatment but that the 28-kDa species comigrated with the 24-kDa species following this treatment (Fig. 4) suggests that the 28- and 24-kDa species differ in their glycosylation pattern and that this glycosylation pattern involves N-linked carbohydrates. This result also is consistent with the observation that human {alpha}-galactosidase A produced in P. pastoris exhibits a glycosylation pattern with a predominant presence of N-linked high-mannose structures rather than complex carbohydrates (6).

Purified trypsinogen was activated in vitro with dDAP. The enzyme dDAP is naturally secreted by D. discoideum, its pH optimum is 3.5, and it has no activity at neutral pH (2). Working at acidic pH is an advantage in this process, since activation by dDAP can occur but autodigestion of the activated trypsin by itself cannot, because trypsin is fully and reversibly inactive in the range of pH 5 to 3.0. With minimal internal cleavage, the resulting enzyme preparation is enriched in highly active ß-chain trypsin (one chain) (18, 44). The absence of further proteolysis of trypsin by dDAP may be explained by steric hindrance, as the cysteine at position 7 is involved in a disulfide bridge constraining the three-dimensional structure of the protein and could prevent further digestion by dDAP (Fig. 1). Alternatively, dDAP may not use isoleucine-valine dipeptides as a substrate. For industrial purposes, the activated trypsin could be further purified by affinity column chromatography to remove dDAP.

Conclusions.
The main objective of this study was to develop a recombinant process for trypsin production on an industrial scale. This work paves the way to such a system in the methylotrophic yeast P. pastoris. In addition, we provide a "proof of concept" for the prevention of the in vivo activation of deleterious proteins, such as proteases, in order to enable their production in a heterologous host. To our knowledge, this is the first report of a scheme involving production of an engineered zymogen, the native form of which is toxic to the host cell but activatable in vitro, with the protein maturation being effected by a suitable exopeptidase. This heterologous trypsin production process also ensures that the resulting enzyme preparation is totally devoid of contaminating pancreatic proteases, while the observed glycosylation does not seem to dramatically affect enzymatic activity. As such, production of trypsin in a foreign host is an optimal method for the preparation of trypsin devoid of contaminating pancreatic enzymes, e.g., as reference material. Moreover, use of recombinantly produced trypsin, as opposed to trypsin extracted from bovine pancreas, circumvents the problems associated with infectious agents potentially associated with animal-derived raw materials.

In this study, altering the zymogen activation site was the key to the efficient production of trypsin. Moreover, our observations are consistent with the conclusion that P. pastoris is an important alternative microbial host for the expression of proteins when production in E. coli generates inclusion bodies.

The process described in this report requires further work to be cost-effective. In particular, in addition to optimization of the fermentation conditions and downstream processing, optimization of the dDAP activation step is needed to reduce the amount of dDAP required. Thus, this process, while scientifically promising, is not yet in commercial use.


arrow
FOOTNOTES
 
* Corresponding author. Present address: Pfizer Global Research & Development, 3-9 Rue de la Loge, BP 100, F-94265, Fresnes Cedex, France. Phone: 01 40 96 74 00. Fax: 01 46 68 16 44. E-mail: alain.vertes{at}pfizer.com. Back


arrow
REFERENCES
 
    1
  1. Abdulaev, N. G., M. P. Popp, W. C. Smith, and K. D. Ridge. 1997. Functional expression of bovine opsin in the methylotrophic yeast Pichia pastoris. Protein Expr. Purif. 10:61-69.[CrossRef][Medline]
    2. 2
  2. Atkinson, P. R., M. D. Hilton, and P. K. Lambooy. 1995. Purification and preliminary characterization of dDAP, a novel dipeptidylaminopeptidase from Dictyostelium discoideum. Biochemistry 34:10827-10834.[CrossRef][Medline]
    3. 3
  3. Brake, A. J., J. P. Merryweather, D. G. Coit, U. A. Heberlein, F. R. Masiarz, et al. 1984. Alpha-factor-directed synthesis and secretion of mature foreign proteins in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 81:4642-4646.[Abstract/Free Full Text]
    4. 4
  4. Casolari, A. 1998. Heat resistance of prions and food processing. Food Microbiol. 15:59-63.
    5. 5
  5. Chance, R., R. Di Marchi, B. Frank, and J. E. Shields. December 1997. Process for preparing insulin analogs. U.S. patent 5,700,662.
    6. 6
  6. Chen, Y., M. Jin, T. Egborge, G. Coppola, J. Andre, and D. H. Calhoun. 2000. Expression and characterization of glycosylated and catalytically active recombinant human {alpha}-galactosidase A produced in Pichia pastoris. Protein Exp. Purif. 20:472-484.[CrossRef][Medline]
    7. 7
  7. Clare, J. J., F. B. Rayment, S. P. Ballantine, K. Sreekrishna, and M. A. Romanos. 1991. High-level expression of tetanus toxin fragment in Pichia pastoris strains containing multiple tandem integrations of the gene. Bio/Technology 9:455-460.[CrossRef][Medline]
    8. 8
  8. Clare, J. J., and M. A. Romanos. 1995. Expression of cloned genes in the yeasts Saccharomyces cerevisiae and Pichia pastoris. Methods Mol. Cell. Biol. 5:319-329.
    9. 9
  9. Cregg, J. M., J. F. Tschopp, C. Stillman, R. Siegel, M. Akong, W. S. Craig, R. G. Buckohltz, K. R. Madden, P. A. Kellaris, G. R. Davis, B. L. Smiley, J. Cruze, R. Torregossa, G. Velicelebi, and G. P. Thill. 1987. High level expression and efficient assembly of hepatitis B surface antigen in the methylotrophic yeast Pichia pastoris. Bio/Technology 5:479-485.[CrossRef]
    10. 10
  10. Cregg, J. M., and D. R. Higgins. 1995. Production of foreign proteins in the yeast Pichia pastoris. Can. J. Bot. 73(Suppl. 1):S891-S897.
    11. 11
  11. Cregg, J. M., T. S. Vedvick, and W. C. Raschke. 1993. Recent advances in the expression of foreign genes in Pichia pastoris. Bio/Technology 11:905-910.[CrossRef][Medline]
    12. 12
  12. Döring, F., M. Klapper, S. Theis, and H. Daniel. 1998. Use of the glyceraldehyde-3-phosphate dehydrogenase promoter for production of functional mammalian membrane transport proteins in the yeast Pichia pastoris. Biochem. Biophys. Res. Commun. 250:531-535.[CrossRef][Medline]
    13. 13
  13. Eldin, P., M. E. Pauza, Y. Hieda, G. Lin, M. P. Murtaugh, P. R. Pentel, and C. A. Pannel. 1997. High level secretion of two antibody single chain Fv fragments by Pichia pastoris. J. Immunol. Methods 201:67-75.[CrossRef][Medline]
    14. 14
  14. Ellis, S. B., P. F. Brust, P. J. Koutz, A. F. Waters, M. M. Harpold, and T. R. Gingeras. 1985. Isolation of alcohol oxidase and two other methanol regulatable genes from the yeast Pichia pastoris. Mol. Cell. Biol. 5:1111-1121.[Abstract/Free Full Text]
    15. 15
  15. Evnin, L. B., and C. S. Craik. 1988. Development of an efficient method for generating and screening active trypsin and trypsin variants. Ann. N.Y. Acad. Sci. 542:61-74.[Medline]
    16. 16
  16. Fayerman, J. T., D. P. Greenen, C. L. Hershberger, J. L. Larson, J. L. Sterner, and Z. Haichao. September 1997. DNA sequences encoding porcine pancreatic carboxypeptidase B. U.S. patent 5,672,496.
    17. 17
  17. Frank, B., W. Prouty, H. Richard, and M. Walden. October 1995. Process for transforming a human insulin precursor to human insulin. U.S. patent 5,457,066.
    18. 18
  18. Gabel, D., and V. Kasche. 1973. Autolysis of ß-trypsin. Acta Chem. Scand. 27:1971-1981.[Medline]
    19. 19
  19. Gemmill, T. R., and R. B. Trimble. 1999. Overview of N- and O-linked oligosaccharide structures found in various yeast species. Biochim. Biophys. Acta 1426:227-237.[Medline]
    20. 20
  20. Hedstrom, L., T. Y. Lin, and W. Fast. 1996. Hydrophobic interactions control zymogen activation in the trypsin family of serine proteases. Biochemistry 35:4515-4523.[CrossRef][Medline]
    21. 21
  21. Hood, M. T., and C. Stachow. 1990. Transformation of Schizosaccharomyces pombe by electroporation. Nucleic Acids Res. 18:688.[Free Full Text]
    22. 22
  22. Huber, H., and W. Bode. 1978. Structural basis of the activation and action of trypsin. Acc. Chem. Res. 11:114-122.[CrossRef]
    23. 23
  23. Kim, T. R., Y. Goto, N. Hirota, K. Kuwata, H. Denton, S. Y. Wu, K. Sawyer, and C. A. Batt. 1997. High level expression of the bovine ß-lactoglobulin in Pichia pastoris and characterization of its chemical properties. Prot. Eng. 10:1339-1345.[Abstract/Free Full Text]
    24. 24
  24. Knappik, A., and A. Pluckthun. 1995. Engineered turns of a recombinant antibody improves its in vivo folding. Protein Eng. 8:81-90.[Abstract/Free Full Text]
    25. 25
  25. Kossiakoff, A. A., J. L. Chambers, L. M. Kay, and R. M. Stroud. 1977. Structure of bovine trypsinogen at 1.9 A resolution. Biochemistry 16:654-664.
    26. 26
  26. Mason, A. B., R. C. Woodworth, R. W. A. Oliver, B. N. Green, L. N. Lin, J. F. Brandts, B. M. Tam, A. Maxwell, and R. T. A. MacGillway. 1996. Production and isolation of the recombinant N-lobe of human serum transferrin from the yeast Pichia pastoris. Protein Exp. Purif. 8:119-125.[CrossRef][Medline]
    27. 27
  27. Montesino, R., R. Garcia, O. Quintero, and J. A. Cremata. 1998. Variation in N-linked oligosaccharide structures on heterologous proteins secreted by the methylotrophic yeast Pichia pastoris. Protein Exp. Purif. 14:197-207.[CrossRef][Medline]
    28. 28
  28. Nyaruhucha, C. N. M., M. Kito, and S. I. Fukuoka. 1997. Identification and expression of the cDNA-encoding human mesotrypsin(ogen), an isoform of trypsin with inhibitor resistance. J. Biol. Chem. 272:10573-10578.[Abstract/Free Full Text]
    29. 29
  29. Paifer, E., E. Margolles, J. Cremata, R. Montesino, L. Herrera, and J. M. Delgado. 1994. Efficient expression and secretion of recombinant alpha amylase in Pichia pastoris using two different signal sequences. Yeast 10:1415-1419.[CrossRef][Medline]
    30. 30
  30. Ridder, R., R. Schmitz, F. Legay, and H. Gram. 1995. Generation of rabbit monoclonal antibody fragments from a combinatorial phage display library and their production in the yeast Pichia pastoris. Bio/Technology 13:255-260.[CrossRef][Medline]
    31. 31
  31. Rosenberg, S., D. Coit, and P. Tekamp-Olson. 1990. Glyceraldehyde-3-phosphate dehydrogenase derived expression cassettes for constitutive synthesis of heterologous proteins. Methods Enzymol. 185:341-351.[Medline]
    32. 32
  32. Rosenfeld, S. A., D. Nadeau, J. Tirado, G. F. Hollis, R. M. Knabb, and S. Jiab. 1996. Production and purification of recombinant hirudin expressed in the methylotrophic yeast Pichia pastoris. Protein Exp. Purif. 7:423-430.[CrossRef][Medline]
    33. 33
  33. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
    34. 34
  34. Scorer, C. A., R. G. Buckholtz, J. J. Clare, and M. A. Romanos. 1993. The intracellular production and secretion of HIV-1 envelope protein in the methylotrophic yeast Pichia pastoris. Gene 136:111-119.[CrossRef][Medline]
    35. 35
  35. Sears, I. B., J. O'Connor, O. W. Rossanese, and B. S. Glick. 1998. A versatile set of vectors for constitutive and regulated gene expression in Pichia pastoris. Yeast 14:783-790.[CrossRef][Medline]
    36. 36
  36. Stroud, R. M., A. A. Kossiakoff, and J. L. Chambers. 1977. Mechanisms of zymogen activation. Annu. Rev. Biophysics Bioeng. 6:177-193.
    37. 37
  37. Tanner, W., and L. Lehle. 1987. Protein glycosylation in yeast. Biochim. Biophys. Acta 906:81-99.[Medline]
    38. 38
  38. Venekei, I., L. Graf, and W. J. Rutter. 1996. Expression of rat chymotrypsin in yeast: a study on the structural and functional significance of the chymotrypsinogen propeptide. FEBS Lett. 379:139-142.[CrossRef][Medline]
    39. 39
  39. Ventura, S., V. Villegas, J. Sterner, J. Larson, J. Vendrell, C. L. Hershberger, and F. X. Aviles. 1999. Mapping of the pro-region of carboxypeptidase by protein engineering. J. Biol. Chem. 247:19925-19933.
    40. 40
  40. Vozza, L. A., L. Wittwer, D. R. Higgins, T. J. Purcell, M. Bergsfeld, L. A. Collins-Racie, E. R. Lavallie, and J. P. Hoefler. 1996. Production of a recombinant bovine enterokinase catalytic subunit in the methylotrophic yeast Pichia pastoris. Bio/Technology 14:77-81.[CrossRef][Medline]
    41. 41
  41. Waterham, H. R., M. E. Digan, P. J. Koutz, S. V. Lair, and J. M. Cregg. 1997. Isolation of the Pichia pastoris glyceraldehyde-3-phosphate dehydrogenase gene and regulation and use of its promoter. Gene 186:37-44.[CrossRef][Medline]
    42. 42
  42. Webster, M. E., and J. V. Pierce. 1961. Action of kallikrein on synthetic substrates. Proc. Soc. Exp. Biol. Med. 107:1564-1583.
    43. 43
  43. Wegner, E. H. November 1983. Biochemical conversions by yeast fermentation at high cell density. U.S. patent 4,414,329.
    44. 44
  44. Wu, H. L., C. Kundrot, and M. L. Bender. 1982. The denaturation of trypsin. Biochem. Biophys. Res. Commun. 107:742-745.[Medline]

Applied and Environmental Microbiology, February 2003, p. 1108-1113, Vol. 69, No. 2
0099-2240/03/$08.00+0     DOI: 10.1128/AEM.69.2.1108-1113.2003
Copyright © 2003, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Beckham, S. A., Boyd, S. E., Reynolds, S., Willis, C., Johnstone, M., Mika, A., Simerska, P., Wijeyewickrema, L. C., Smith, A. I., Kemp, D. J., Pike, R. N., Fischer, K. (2009). Characterization of a Serine Protease Homologous to House Dust Mite Group 3 Allergens from the Scabies Mite Sarcoptes scabiei. J. Biol. Chem. 284: 34413-34422 [Abstract] [Full Text]  
  • Vertes, A. A., Inui, M., Yukawa, H. (2005). Manipulating Corynebacteria, from Individual Genes to Chromosomes. Appl. Environ. Microbiol. 71: 7633-7642 [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.
Agricola
Right arrow Articles by Hanquier, J.
Right arrow Articles by Vertès, A. A.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»