Secretion of human glucocerebrosidase from stable transformed ...

148

Secretion of human glucocerebrosidase from stable transformed insect cells using native signal sequences Graham Sinclair, Tom A. Pfeifer, Thomas A. Grigliatti, and Francis Y.M. Choy

Abstract: The lysosomal hydrolase, glucocerebrosidase (GBA), catalyses the penultimate step in the breakdown of membrane glycosphingolipids. An inherited deficiency of this enzyme activity leads to the onset of Gaucher disease, the most common lysosomal storage disorder. Affected individuals range from adults with hepatosplenomegaly, haematological complications, and bone pain (type 1 disease) to children and neonates with severe neuronopathy leading to neurological degradation and premature death (type 2 and type 3 disease). Enzyme replacement therapy has become the standard of treatment for type I Gaucher disease but remains an expensive option, in part because of the cost of recombinant enzyme production using mammalian cell culture. Using a nonlytic integrative plasmid expression system, we have successfully produced active human GBA in stable transformed Sf 9 (Spodoptera frugiperda) cells. Both the 39 and 19 amino acid native GBA signal sequences were capable of endoplasmic reticulum targeting, which led to secretion of the recombinant protein, although approximately 30% more enzyme was produced using the longer signal sequence. The secreted product was purified to apparent electrophoretic homogeneity using hydrophobic interaction chromatography and found to be produced in a fully glycosylated and a hypoglycosylated form, both of which cross-reacted with a human GBA-specific monoclonal antibody. The pH optimum (at pH 5.5) for activity of the recombinant enzyme was as expected for human GBA using the artificial substrate 4-methyl-umbelliferyl-β-D-glycopyranoside. With initial nonoptimized expression levels estimated at 10–15 mg/L using small-scale batch cultures, stable transformed insect cells could provide a viable alternative system for the heterologous production of human GBA when grown under optimized perfusion culture conditions. Key words: Gaucher disease, glucocerebrosidase, protein expression, enzyme purification, Sf 9 cells. Résumé : La glucocérébrosidase (GBA), une hydrolase lysosomale, catalyse l’avant-dernière étape de la dégradation des glycosphingolipides membranaires. Une déficience héréditaire dans l’activité de cette enzyme mène à l’apparition de la maladie de Gaucher, la maladie affectant l’entreposage lysosomal la plus commune. Le type d’individus affectés varie, allant d’adultes présentant une hépatosplénomégalie, des complications hématologiques et des douleurs osseuses (maladie de type 1), aux enfants et aux nouveaux-nés présentant une neuropathie sévère conduisant à la dégradation neurologique et le décès prématuré (maladie de types 2 et 3). La thérapie de remplacement enzymatique s’est révélée un traitement standard de la maladie de Gaucher de type 1, mais elle demeure une option onéreuse, dû entre autres au coût de production de l’enzyme recombinante en système de culture de cellules mammifères. Grâce à un système d’expression de plasmide recombinant intégré non-lytique, nous avons réussi à produire une GBA humaine active dans des cellules Sf 9 transformées stablement. Les deux séquences signal de 39 et 19 acides aminés de la GBA native ont été capables de cibler le protéine au RE, résultant en sécrétion de protéine recombinante, quoique approximativement 30 % plus d’enzyme était produite par l’utilisation de la séquence la plus longue. Le produit sécrété a été purifié à homogénéité apparente selon des critères électrophorétiques, par chromatographie d’hydrophobicité et s’est révélé être sous forme totalement glycosylée ou hypo-glycosylée, les deux formes réagissant avec un anticorps monoclonal spécifique à la GBA humaine. Le pH optimal (pH 5,5) à l’activité de l’enzyme recombinante correspond à ce qui était attendu de la GBA humaine pour le substrat artificiel 4-méthyl-umbelliferyl-β-D-glycopyranoside. Avec des niveaux d’expression non optimisés estimés à 10–15 mg/L en culture non renouvelée à petite échelle, les cellules d’insectes transformées stablement pourraient fournir une alternative viable à la production hétérologue de GBA lorsque cultivées sous des conditions de cultures optimales en milieu perfusé. Mots clés : maladie de Gaucher, glucocérébrosidase, expression des protéines, purification enzymatique, cellules Sf 9. [Traduit par la Rédaction]

Sinclair et al.

156

Received 20 July 2005. Revision received 17 October 2005. Accepted 20 October 2005. Published on the NRC Research Press Web site at http://bcb.nrc.ca on 3 March 2006. G. Sinclair and F.Y.M. Choy.1 Biomedical Research Centre, Department of Biology, University of Victoria, PO Box 3020, Station CSC, Victoria, BC V8W 3N5, Canada. T.A. Pfeifer and T.A. Grigliatti. Department of Zoology, University of British Columbia, 6270 University Blvd., Vancouver, BC V6T 1Z4, Canada. 1

Corresponding author (e-mail: [email protected]).

Biochem. Cell Biol. 84: 148–156 (2006)

doi:10.1139/O05-165

© 2006 NRC Canada

Sinclair et al.

Introduction Gaucher disease is a lysosomal storage disorder resulting from an inherited deficiency in glucocerebrosidase (GBA) activity (Brady et al. 1965). Individuals affected with this disease range from adults with hepatosplenomegaly, haematological complications, and bone pain (type 1 disease) to children and neonates with neuronopathy leading to severe disability and premature death (type 2 and type 3 disease) (Beutler and Grabowski 2001). Although there is no cure for Gaucher disease, enzyme replacement is now a viable, though expensive, long-term therapy for many affected individuals (predominantly type 1). Alglucerase, macrophage-targeted GBA isolated from human placenta, was approved for clinical use in 1991 and released under the trade name Ceredase™ (Barton et al. 1990, 1991; Beutler 1993). It has been found highly effective in decreasing hepatosplenomegaly and overall sphingolipid storage levels in a large number of type 1 Gaucher cases (Altarescu et al. 2000; Brady and Barton 1994; Grabowski et al. 1998; Mistry et al. 1996; Moscicki and Taunton-Rigby 1993). A recombinant form of the enzyme, imiglucerase, has more recently been developed in a Chinese hamster ovary (CHO) cell expression system and marketed for clinical use under the trade name Cerezyme™ (Hoppe 2000). Unfortunately, the cost of this therapy remains a burden ($382 000/yr), resulting in highly selective inclusion criteria for treatment and a limitation of therapy for those with neuronopathy to controlled clinical trials. (Figueroa et al. 1992) Heterologous expression of human GBA has been attempted in a number of expression systems with varying results. GBA is a highly hydrophobic glycoprotein requiring extensive post-translational modification to ensure proper folding and activity. The post-translational modifications are crucial for enzyme function and cannot be provided by prokaryotic cell expression systems (Grace and Grabowski 1990). Higher eukaryotic systems, such as the aforementioned CHO cells, can provide the necessary post-translation modifications and trafficking but relatively low protein production levels, low achievable cell density, and labour-intensive culturing requirements. Downstream purification and processing also drive up production costs and can limit the availability of the therapeutic protein (Geisse et al. 1996). Insect cell expression systems using Autographa californica multicapsid nucleopolyhedrosis (AcMNPV) baculovirus vectors have been used extensively for the expression of foreign eukaryotic proteins (Blissard and Rohrmann 1990; Bonning and Hammock 1996; Hasnain et al. 1997; Jarvis 2003). Although high levels of transcription have been associated with AcMNPV expression systems, the quantity and quality of the secreted heterologous protein produced using baculovirus systems have been inconsistent because of the nature of viral infection (Geisse et al. 1996; Jarvis and Finn 1995). As part of the infection process, the baculovirus tends to restrict the production of native insect cell proteins in favour of those encoded by the viral genome (Kretzchmar et al. 1994). While this leads to increased production levels of simple proteins, the reduced availability of chaperones, glycosidases, and other cellular factors required for the maturation of complex glycoproteins has been shown to affect both the quantity and quality of heterologous proteins, particularly late in infection (Chazenbalk and Rapoport 1995).

149

However, using an AcMNPV-based expression system, insect cells have been shown to produce active GBA at a diagnostic level, and some investigations of the utility of this system for large-scale production have been undertaken (Martin et al. 1988). In contrast to the lytic, baculovirusbased systems, a plasmid-mediated, nonlytic expression system would allow for the stable genomic integration of the GBA transgene and should allow high-level expression. Stable cell line expression systems using a number of baculovirus immediate-early promoters to drive expression of a heterologous gene have been created to produce high levels of human proteins in a variety of dipteran and lepidopteran cell lines (Jarvis et al. 1996; Johansen et al. 1989; Hegedus et al.1998; Pfeifer 1998). This avoids the transient problems of baculovirus expression systems and allows both the scale-up to large bioreactors and the continuous production in perfusion bioreactors in serum-free, protein-free media (Gorenflo et al. 2004). Hence we chose to examine the production of human GBA in a stable insect cell-based integrative plasmid expression system and to test the function of the resulting human GBA produced in Sf 9 (Spodoptera frugiperda) cells.

Materials and methods Sf 9 vector construction All Sf 9 GBA constructs were made using the vector p2ZOp2F, with transcription of the transgene driven from the Orgyia pseudotsugata nucleopolyhedrosis virus (OpNPV) immediate-early promoter (Hegedus et al. 1998). This vector also contained a Zeocin™ resistance gene for the selection of recombinant plasmids in Escherichia coli and selection of stable genomic integrants in transformed Sf 9 cells. GBA constructs were cloned into the p2ZOp2F vector at the EcoRI site, downstream of the OpNPV immediate-early promoter to produce the pOpfGBA and pOpsGBA vectors (Fig. 1). The 2 constructs were designed to direct transcription of the entire GBA coding region, including both the full 39-amino acid (fGBA) and shorter 19-amino acid (sGBA) native leader sequences. The primers used for Pfu PCR amplification of the cDNAs were as follows: both constructs (reverse 5′GCTGAATTCTTTAATGCCCAGGCTG-3′), fGBA (forward 5′-ACTCGAATTCTCTTCATCTAAGGACCCTGAGG-3′), and sGBA (forward 5′-TACCGAATTCATGGCTGGCAGCCTCACAGG-3′). All primers included linker regions to introduce EcoRI cloning sites (noted in bold), and the fGBA forward primer included a single T to A mismatch to eliminate a possible cryptic ATG start site (noted by underline). Sf 9 cell transfection and stable cell line selection Prior to transfection, Sf 9 cells were grown at 26 °C to mid-log phase in ESF921 serum-free, protein-free medium (Expression Systems, Woodland, Calif.). For transient transfections, 1–2 × 106 cells were seeded in a 6-well tissue culture plate with 1 mL of Grace’s minimal medium (Life Technologies Canada, Burlington, Ont.) and allowed to attach. CellFectin™ (Life Technologies) lipofection reagent (10 µL) was mixed with 1.0 µg of plasmid DNA in 1 mL of Grace’s medium, incubated for 30 min, and added to the seeded Sf 9 cells. Following a 4 h incubation, the transfection mix was replaced with 2 mL of ESF921 medium, © 2006 NRC Canada

150

Biochem. Cell Biol. Vol. 84, 2006

Fig. 1. Sf 9 cell expression constructs. pOpfGBA is the Sf 9 expression vector directing expression of the GBA cDNA with its full-length leader (fGBA) from the OpMNPV immediate-early promoter (OpIE2). pOpsGBA contains the GBA cDNA with a shorter, 19-amino acid leader (sGBA). OpIE2 polyadenylation signal (pA), E. coli origin of replication (Col E1), and Zeocin™ resistance gene (Zeo) are noted.

and the cultures were incubated at 26 °C for 48 h. For the selection of stable cell lines, the cells were transferred to T25 tissue culture flasks and allowed to grow to confluence in the presence of 1 mg/mL Zeocin™. Selection on Zeocin™ was maintained for a total of 3 passages to obtain a stable polyclonal cell line. These stable lines were scaled up to 250 mL Erlenmyer flasks and incubated at 26 °C and 100 rpm. When required, the cultures were further scaled up to 2 cultures of 500 mL using 2 L Erlenmyer flasks and grown as noted earlier. GBA activity assay GBA activity in the crude cell lysates and culture medium was assayed using the fluorescent substrate 4-methylumbelliferyl-β-D-glucopyranoside (4MUGP) (Sigma–Aldrich Canada, Oakville, Ont.) as described previously (Choy 1984). Sf 9 cell lysates were prepared by means of 5 cycles of freeze–thaw, and culture medium was either used directly or the soluble proteins were concentrated by ammonium sulfate precipitation (50% w/v) prior to activity assay. All protein concentrations were calculated using the Biorad Reagent (BioRad Laboratories, Hercules, Calif.) as adapted from Bradford (1976). Protein analysis Sf 9 cell lysates and culture medium were prepared as described earlier for the analysis of protein expression by Tris–glycine SDS–PAGE as described by Spector et al. (1998). Proteins were visualized on SDS-PAGE gels using Gel Code stain (Pierce Biotechnology Inc., Rockford, Ill.) or silver staining as follows. Gels were microwaved at maximum power for 90 s in fixative (50% methanol, 12% acetic acid, 0.1% formaldehyde) and then for 90 s in 50% ethanol. The gels were then pretreated in 0.02% sodium thiosulfate pentahydrate for 90 s in the microwave, washed in deionized water for 90 s at room temperature, and stained with 2 mg/mL silver nitrate in 0.075% formaldehyde by microwaving twice for 40 s. Bands were resolved in developer (60 mg/mL sodium carbonate, 0.05% formaldehyde, 0.002% sodium thiosulfate

pentahydrate) and stopped in 50% methanol following a 90 s water wash. For Western blotting, proteins were electroblotted from SDS–PAGE gels onto Hybond-P polyvinylidene difluoride (PVDF) membrane (Amersham Pharmacia Biotech, Piscataway, N.J.) for 1 h at 100 V in 10% methanol transfer buffer (25 mmol Tris–HCl/L, 0.2 mol glycine/L) using a Miniprotean II Electroblot Apparatus (Bio–Rad Laboratories). PVDF membranes were washed in TTBS (20 mmol Tris– HCl/L, pH 7.5, 0.05% Tween-20, 500 mmol sodium chloride/L) for 5 min at room temperature, and this was followed by 1 h of blocking in TTBS with 7.5% w/v dry skimmed-milk powder. The membrane was washed twice for 5 min in TTBS before a 1 h incubation with the GBA specific primary antibody diluted 1:200–1:400 in blocking solution (7.5% dry milk in TTBS). The antibody used was a mouse monoclonal antibody, AA16B3, raised against the native human placental enzyme purified to homogeneity and donated by Dr. Ernest Beutler. Following incubation with the primary antibody, membranes were washed 4× for 5 min in TTBS and incubated in a goat anti-mouse secondary antibody conjugated to horseradish peroxidase (Clontech, Palo Alto, Calif.) diluted 1:4000–1:6000 in blocking solution for 1 h. Membranes were washed 4× in TTBS and incubated in ECL Plus chemiluminescent reagent (Amersham Pharmacia Biotech, Piscataway, N.J.) for luminescent detection on Biomax autoradiography film (Eastman Kodak Co., Rochester, N.Y.) or fluorescent detection using a Molecular Dynamics Storm 860 phosphorimager (Molecular Dynamics, Sunnyvale, Calif.). Peptide:N-glycosidase F (PNGase F) digestion was performed under denaturing conditions as described in the product literature (Sigma–Aldrich Canada). Fast protein liquid chromatography (FPLC) purification Proteins were separated from Sf 9 stable cell line medium by hydrophobic interaction chromatography (HIC) with a Biogel TSK phenyl-sepharose column (150 mm × 21.5 mm) (Bio–Rad Laboratories) using a Pharmacia FPLC system (Amersham Pharmacia Biotech). Culture medium was prepared © 2006 NRC Canada

Sinclair et al.

for FPLC by the addition of ammonium sulfate to 1.7 mol/L, centrifugation at 14 000g at 4 °C for 30 min, and filtration– degassing through a 0.45 µm membrane (Pall Gelman Laboratories, Ann Arbor, Mich.). Samples were loaded on to the column previously equilibrated with 1.7 mol ammonium sulfate/L at a flow rate of 5 mL per min and held at 100% buffer A (1.7 mol ammonium sulfate/L) until the void had passed and a stable baseline monitor reading (λ = 230 nm) was established. Proteins were eluted from the column with a linear gradient from 100% buffer A to 100% buffer B (0.1 mol NaCl/L) over 4 column volumes followed by a linear gradient over 12 column volumes to 100% buffer C (5% w/v cholic acid). Fractions were collected at 2 min intervals and screened by 4MUGP assay and protein dot blot for the presence of GBA enzyme activity or cross-reactivity. 4MUGP assays were performed as previously described (Choy 1984) with the following alterations. Total volumes were reduced to allow for a 96-well plate format, and active fractions were determined visually under ultraviolet light rather than through quantification by a fluorimeter to speed the screening process. Protein dot blotting was performed by soaking PVDF membranes in methanol and washing in TTBS before manually dotting 2.5 µL aliquots of each FPLC fraction onto the membrane. The membranes were dried to fix the protein dots, and crossreactive material was determined following the immunodetection protocol already described in the Western blotting section (Protein analysis). Volume reports were generated for each dot using the spotfinder function of the Storm 860 phosphorimager (Molecular Dynamics) to estimate relative amounts of cross-reactive material in each FPLC fraction as compared with the pre-column sample. Using these values an overall purification yield could be calculated.

Results Human GBA has been produced in a CHO cell expression system for enzyme replacement therapy for Gaucher disease, but it remains an expensive treatment option. In this study, a plasmid-based nonlytic insect cell expression system was investigated as a potential expression host for the production of human GBA. In human cells, the enzyme is translated with either a 39-amino acid or 19-amino acid signal sequence that directs the nascent polypeptide to transit the endoplasmic reticulum (ER) membrane before being cleaved in the ER lumen. The enzyme is then glycosylated and shuttled through the transgolgi network into maturing lysosomes through an unknown targeting mechanism. Because of the presence of these native signal sequences on each GBA construct used in this study, it was hypothesized that measurable quantities of the nascent GBA would be directed to the insect cell lysosome or remain unlocalized intracellularly as is seen with baculovirus expression (Grabowski et al. 1989; Xu and Grabowski 1998). Sf 9 cells were transfected with the pOpfGBA and pOpsGBA vectors, and both cell lysates and culture media were harvested after 72 h. Crude cell lysate activity was assayed using the fluorogenic substrate (4MUGP) (Choy et al. 1996) but failed to reveal any functional GBA protein from the transfected Sf 9 cells (data not shown). However, analysis of the culture medium revealed considerable GBA activity. To

151

confirm this, proteins from the culture medium of both fGBAand sGBA-transformed Sf 9 cells were analyzed by Western blotting using a mouse anti-human GBA-specific monoclonal antibody (AA16B3). These Western blots revealed a distinctive doublet of cross-reactive material at approximately 60– 63 kDa (Fig. 2A) in media derived from transiently transfected cells and zero cross-reactivity in the cell lysates (Fig. 2B). Thus, expression of human GBA gene construct with either native signal sequence, in Sf 9 cells, leads to secretion of the human GBA protein into the medium with no visible degradation of the product. Stable transformed polyclonal cell lines were then established using a Zeocin™ selection scheme. With the selected stable polyclonal lines, there was a 30% difference in GBA expression between the fGBA and sGBA lines with 4MUGP specific activities of 1942 (±15.2) nmol·h–1·mg protein–1 for the fGBA clone as compared with 1616 (±109) nmol·h–1·mg protein–1 for the sGBA clone (Table 1). HIC was performed to purify the GBA protein present in the culture medium of Sf 9 cells selected for stable integration of the transgene. For the fGBA clone, a single broad peak of 4MUGP activity was eluted from the column at approximately 2.5% cholate (Fig. 3). A protein dot blot was performed to check all eluted fractions from the HIC run for the presence of cross-reactive material. The major peak of cross-reactive material correlated directly with the peak in 4MUGP activity, confirming the identity of the GBA protein. Western blot analysis of the active fractions separated by SDS–PAGE confirmed the presence of a 63 kDa cross-reactive protein (Fig. 4A). However, direct protein staining revealed the presence of an additional minor band of 60 Da in 2 fractions (fractions 1 and 2, Fig. 4B), suggesting the presence of either a minor species that lacks the epitope recognized by the anti-GBA monoclonal antibody or contaminant in these fractions. Identical results were obtained for the purification of GBA from the sGBA clone (data not shown). PNGase F deglycosylation was performed on the active fraction 3 (Fig. 4B) from the HIC run, and upon silver staining of the protein the 63 kDa band decreased to a species of 58 kDa, confirming the presence of N-linked glycans and the appropriate size of the deglycosylated protein (Fig. 5). The HIC-purified GBA was also assayed using the artificial 4MUGP substrate under varying pH conditions, from pH 4.0 to 8.0 at 0.5 pH increments, to confirm its identity as an acid β-glucosidase. The activity curve of the purified enzyme was noted to peak at pH of 5.5 (data not shown), which is identical to the activity maximum of the native human enzyme (Peters et al. 1976). This confirmed that active human GBA had been secreted from transformed Sf 9 cells and purified to electrophoretic homogeneity. Scanning volume reports were generated on protein dot blots from the HIC purification using the Storm 860 phosphorimager to estimate the yield of the purification. As all the material eluted from the column was collected and sampled (including the flowthrough), volume reports from all collected fractions were summed and compared with dots of the sample prior to loading (data not shown). From this comparison it was determined that 67% of the GBA crossreactivity applied to the column was contained within the pooled fractions representing the single activity peak. Although this analysis was an estimate of the GBA yield, a similar © 2006 NRC Canada

152

Biochem. Cell Biol. Vol. 84, 2006

Fig. 2. Western blot of culture medium (A) and cell lysates (B) from transient transfected Sf 9 cells expressing GBA (GBA). Transfections with GBA constructs with the full-length (fGBA) or shorter (sGBA) native leader sequence are presented along with a nontransfected control Sf 9 cell line. A positive control lane of baculovirus-expressed human GBA (rHuman GBA) (Choy et al. 1996) is included in panel B. Sixty microlitres of culture medium samples (3 d post-transfection) or 5 µg of total protein from cell lysates were loaded for 10% Tris–glycine SDS–PAGE separation. Immunodetection was done using a mouse anti-human GBA-specific monoclonal antibody (AA16B3).

densitometric analysis of a Western Blot including pooled cross-reactive fractions and the pre-column sample estimated a yield of 44% (data not shown). On the basis of a protein concentration of 6.6 µg/mL in the pooled cross-reactive GBA fractions and a purification yield of 44%–67%, this suggests an initial protein production of 10–15 mg/L of culture medium.

Discussion Functional human GBA was expressed in stable transformed Sf 9 cells using either full-length or shortened signal sequence GBA constructs. In both cases, the product was secreted into the culture medium. The secretion of recombinant GBA appears to be a host-cell-dependent phenomenon with some cell types secreting significant amounts of protein and other cell types releasing no measurable GBA to the culture medium. For example, Xu and Grabowski (1998) found that CHO and mouse myoblast (C2C12) cell lines expressing GBA through a retroviral vector secreted 50%– 75% of the recombinant GBA into the culture medium, whereas transformed fibroblasts secreted no enzyme. GBA secretion was also seen with baculovirus expression in Sf 9 cells from the AcMNPV polyhedron promoter with 50%– 75% of the enzyme activity present in the medium at 48 h after infection (Xu and Grabowski 1998). In contrast to these findings, Berg-Fussman et al. (1993) reported that little enzyme was secreted from Sf 9 or COS-1 cells expressing GBA using the same promoter and harvest times. Grabowski et al. (1989) were also unable to find GBA activity in the culture medium of Sf 9 cells until 5–7 d after infection using a baculovirus-based GBA expression system, suggesting a lytic release rather than active secretion. Unfortunately, most other studies expressing GBA with baculovirus-infected Sf 9 cells have focused on cellular activities and included no

Table 1. Glucocerebrosidase activities (assayed using 4MUGP a) from the culture medium of stable Sf 9 transformants. Sf 9 clone

4MUGP activityb (nmol·h–1·mg protein–1)

Controlc fGBA sGBA

249.1±5.59 1942±15.2 1616±109

a

4MUGP, 4-methyl-umbelliferyl-β-D-glucopyranoside. Activities represent mean values ± the standard error from 3 independent assays 72 h post-seeding. c Untransformed Sf 9 cells. b

investigation of GBA secretion to the culture medium (Grace and Grabowski 1990; Grace et al. 1994). Despite the conflicting data in the literature, it appears that stable transformed Sf 9 cell lines expressing the human GBA gene from templates integrated in the insect genome are capable of secreting the majority of the human protein into the culture medium. Although the native signal sequence from GBA is required for localization to the ER, subsequent lysosomal localization in human cells is independent of this sequence, and the lysosomal targeting mechanism for GBA remains to be elucidated (Glickman and Kornfeld 1993; Zimmer et al. 1999). The required processing pathways, binding partners, or targeting signals may not be efficiently recognized in Sf 9 cells, and the default pathway is secretion of the recombinant product out of the cell. The near absolute secretion of GBA by stable transformed Sf 9 cells as compared with the partial (or absent) secretion of baculovirus-based systems could result from the maintenance of regular cellular metabolism in stable transformed cells. Baculovirus infection inhibits the transcriptional activity of many of the insect host genes, and © 2006 NRC Canada

Sinclair et al.

153

Fig. 3. Elution profile from hydrophobic interaction chromatography (HIC-FPLC) of stable transfected Sf 9 cell medium expressing GBA construct fGBA (full-length native leader). The grey trace in the upper portion of the figure represents, from left to right, the desalting gradient from 100% buffer A (1.7 mol ammonium sulfate/L) to 0% buffer A (100% B, 0.1 mol NaCl/L) and the final cholate gradient to 100% buffer C (5% w/v cholic acid). Those fractions displaying activity on the artificial 4MUGP substrate are highlighted below the absorbance trace (λ = 230 nm). A single broad peak of active and cross-reactive protein eluted at ~2% to 2.5% cholate.

preferential transcription of the viral genome occurs. In addition, baculovirus infection causes degradation of cellular enzymes and cofactors involved in protein processing, which alters the post-translational maturation of host proteins and may alter the trafficking of proteins (Ailor and Betenbaugh 1999; Hegedus et al. 1998). This could lead to the retention of misfolded or improperly targeted recombinant proteins in the endomembrane system or cytoplasm of the cell in baculovirus-based expression systems, in contrast to the secretion of properly processed proteins using a stable transformed cell system. The GBA construct containing both native initiator ATGs (fGBA) produced approximately 30% more enzymatic activity than the construct with only the second ATG and a shorter 19 amino acid leader (sGBA). This result is in agreement with the expression of GBA in retrovirus-transformed human cells (HeLa), in which transcription from the upstream initiator ATG alone, or in concert with the second ATG, produced 25%–50% more protein activity than constructs that contained only the second ATG (Pasmanik-chor et al. 1996). Although the fGBA construct could initiate translation from either in-frame ATG, these results suggest that the upstream initiator is preferred in Sf 9 cells. Indeed, translation of the human GBA gene in mammalian systems, both in vivo and in vitro, appears to occur more efficiently for the upstream ATG (Pasmanik-chor et al. 1996; Sorge et al. 1987). Hence, translation preferences in the insect system appear to parallel

mammalian systems. Western blots suggest that alteration in leader sequence length had little impact on GBA secretion in Sf 9 cells, as both constructs yielded minimal intracellular expression. Heterologous expression in murine cells (NIH/3T3) supports this finding, as no alterations in intracellular trafficking were observed between GBA constructs translated from either initiation codon (Sorge et al. 1987). The presence of a 63 kDa species in the culture medium and partially purified active HIC pools corresponds to the published size of expressed GBA (Grabowski et al. 1989; Grace and Grabowski 1990; Grace et al. 1990, 1994). The presence of a second cross-reactive species of 60 kDa (Fig. 2A) suggests that there is some heterogeneity in the glycosylation of the secreted product under transient transfection conditions. The simplest interpretation is that these 2 isoforms correspond to different glycoforms of GBA either because of variable occupancy of the various possible N-linked glycosylation sites or difference in processing of the oligosaccharide structures. Enzymatic deglycosylation of the partially purified product resolved a single 58 kDa species, as presented in Fig. 5. A reduction of the 63 kDa GBA species by approximately 2 kDa was observed by Grace and Grabowski (1990) following PNGase digestion of GBA under nondenaturing conditions, and this reduction in estimated molecular mass correlated with the removal of a single glycan from the mature protein. In this present study, GBA was denatured prior to PNGase digestion, suggesting that the 58 kDa band observed © 2006 NRC Canada

154 Fig. 4. GBA purified from the culture medium of a stable Sf 9 cell line expressing the GBA construct containing the full-length native leader sequence (fGBA). (A) Western blotting analysis of pre-column media and the major cross-reactive fractions (1, 2, and 3) eluted from the hydrophobic interaction column during the cholate gradient using mouse anti-human GBA-specific monoclonal antibody AA16B3. (B) Gelcode Blue™ (Pierce Chemical Co., Rockford, Ill.) protein staining of the highly cross-reactive fractions (1, 2, and 3) of the Western blot. Broad Range Protein Standards (New England Biolabs Inc., Beverley, Mass.) were used for both (A) and (B).

represents a fully deglycosylated form, although altered size due to incorrect processing or proteolytic degradation cannot be unequivocally ruled out. Protein glycosylation in Sf 9 cells differs from that in mammalian cells by the extent to which trimmed glycan structures are expanded. While mammalian glycosylation pathways extend core glycans into bi- and tri-antennary sialylated complex glycans, insect systems tend not to expand glycans beyond the paucimannose (Man3GlcNAc2-based structure) core (Kost et al. 2005). Although proper glycosylation is required for the function of the GBA protein, it is sequon occupancy rather than glycan structure that is central to that function. In fact, appropriate in vivo targeting of CHO-cellderived human GBA for enzyme replacement therapy has been achieved by trimming the complex mammalian glycans to produce mannose-terminated structures. These glycans are recognized by mannose receptors on the surface of reticuloendothelial cells leading to the internalization of the therapeutic protein. This targeting mechanism has been successful in the treatment of Gaucher disease and has subsequently been exploited in the development of therapeutics for a number

Biochem. Cell Biol. Vol. 84, 2006 Fig. 5. PNGase F enzymatic deglycosylation of partially purified GBA produced in stable transfected Sf 9 cells. The original GBA sample (GBA), and GBA incubated in digestion buffer under denaturing conditions in the presence (PNGase+) and absence (PNGase–) of the N-glycanase (PNGase F, Sigma–Aldrich Canada, Oakville, Ont.) are presented. Protein size markers correspond to the Broad Range Protein Standards (New England Biolabs Inc.). Proteins were separated on 10% Tris–glycine gels and silver stained.

of other lysosomal storage disorders. Accordingly, the production of mannose-terminated glycans on Sf 9-produced human GBA increases its direct applicability as a potential biotherapeutic product. While these experiments prove that stable transformed insect cells are capable of producing functional human GBA, the production yield can likely be improved by using alternative secretion vectors and through more detailed analysis of clonal lines for GBA production. With an estimated production yield of 10–15 mg/L for catalytically active human GBA using a small-scale batch culture, a great deal of improvement should be possible with optimized culture conditions. For example, Gorenflo et al. (2004) were able to show a 10-fold increase in volumetric productivity by switching to an Sf 9 perfusion culture system with the expression of Factor X fusion protein. Reported optimized heterologous protein expression levels in stable transformed Sf 9 cells have ranged from 4 to 25 mg/L (Pfeifer et al. 2001). It appears that a stable transfected Sf 9 cell expression system could be suitable for larger-scale expression of this valuable protein for basic research and potential biotherapeutics following the appropriate optimization and scale-up. © 2006 NRC Canada

Sinclair et al.

Acknowledgements This work was supported by a Natural Sciences and Engineering Research Council (NSERC) operating grant, No. 138216-01 (F.C.), an NSERC Strategic Project Grant, No. 234718 (T.G.), an NSERC postgraduate fellowship (G.S.), and a student research grant from the Scottish Rite Charitable foundation (G.S.). The authors thank Dr. Ernest Beutler for providing the mouse monoclonal antibody directed against homogenous GBA, and Jamie Haddon and D’Arcy Deacon for their technical assistance.

References Ailor, E., and Betenbaugh, M.J. 1999. Modifying secretion and post-translational processing in insect cells. Curr. Opin. Biotechnol. 10: 142–145. Altarescu, G., Schiffmann, R., and Parker, C.C. 2000. Comparative efficacy of dose regimes in enzyme replacement therapy of type I Gaucher disease. Blood Cells Mol. Dis. 26: 285–290. Barton, N.W., Furbish, F.S., Murray, G.J., Garfield, M., and Brady, R.O. 1990. Therapeutic response to intravenous infusions of glucocerebrosidase in a patient with Gaucher disease. Proc. Natl. Acad. Sci. U.S.A. 87: 1913–1916. Barton, N.W., Brady, R.O., and Dambrosia, J.M., Di Bisceglie, A.M., Doppelt, S.H., Hill, S.C. et al. 1991. Replacement therapy for inherited enzyme deficiency — macrophage targeted glucocerebrosidase for Gaucher’s Disease. N. Engl. J. Med. 324: 1464–1470. Berg-Fussman, A., Grace, M.E., Ioannou, Y., and Grabowski, G.A. 1993.Human acid β-glucosidase. N-glycosylation site occupancy and the effect of glycosylation on enzymatic activity. J. Biol. Chem. 268: 14861–14866. Beutler, E. 1993. Gaucher disease as a paradigm of current issues regarding single gene mutations of humans. Proc. Natl. Acad. Sci. U.S.A. 90: 5384–5390. Beutler, E., and Grabowski, G.A. 2001. Gaucher disease. In The metabolic and molecular bases of inherited disease. Edited by C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle. McGraw-Hill, New York, pp. 3645–3668. Blissard, G.W., and Rohrmann, G.F. 1990. Baculovirus diversity and molecular biology. Annu. Rev. Entomol. 35: 127–155. Bonning, B.C., and Hammock, B.D. 1996. Development of recombinant baculoviruses for insect control. Annu. Rev. Entomol. 41: 191–210. Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Ann. Biochem. 72: 248–254. Brady, R.O., and Barton, N.W. 1994. Enzyme replacement therapy for Gaucher disease: critical investigations beyond demonstration of clinical efficacy. Biochem. Med. Metab. Biol. 52: 1–9. Brady, R.O., Kanfer, J.N., and Shapiro, D. 1965. Metabolism of glucocerebrosidase II. Evidence of enzymatic deficiency in Gaucher’s disease. Biochem. Biophys. Res. Commun. 18: 221–225. Chazenbalk, G.D., and Rapoport, B. 1995. Expression of the extracellular domain of the thyrotropin receptor in the baculovirus system using a promoter active earlier than the polyhedrin promoter. J. Biol. Chem. 270: 1543–1549. Choy, F.Y.M. 1984. Gaucher disease: the effects of phosphatidylserine on glucocerebrosidase from normal and Gaucher fibroblasts. Hum. Genet. 67: 432–436. Choy, F.Y.M. 1986. Purification of human placental glucocerebrosidase

155 using a two-step high-performance hydrophobic and gel permeation column chromatography method. Anal. Biochem. 156: 515–520. Choy, F.Y.M., Wei, C., and Levin, D.1996. Gaucher disease: functional expression of the normal glucocerebrosidase and Gaucher T1366G and G1604A alleles in Baculovirus-transfected Spodoptera frugiperda cells. Am. J. Med. Genet. 65: 184–189. Figueroa, M.L., Rosenbloom, B.E., Kay, A.C., Garver, P., Thurston, D.W., Koziol, J.A. et al. 1992. A less costly regimen of alglucerase to treat Gaucher’s disease. N. Engl. J. Med. 327: 1632–1636. Geisse, S., Gram, H., Kleuser, B., and Kocher, H.P. 1996. Eukaryotic expression systems: a comparison. Protein Expr. Purif. 8: 271–282. Glickman, J.N., and Kornfeld, S. 1993. Mannose 6-phosphateindependent targeting of lysosomal enzymes in I-cell disease B lymphoblasts, J. Cell Biol. 123: 99–108. Gorenflo, V.M., Pfeifer, T.A., Grigliatti, T.A., Lesnicki, G., Kwan, E.M., Kilburn, D.G., and Piret, J.M. 2004. Production of a selfactivating CBM-factor X fusion protein in a stable transformed Sf 9 insect cell line using high cell density perfusion culture. Cytotechnology, 44: 93–102. Grabowski, G.A., White, W.R., and Grace, M.E. 1989. Expression of functional human acid β-glucosidase in COS-1 and Spodoptera frugiperda cells. Enzyme, 41: 131–142. Grabowski, G.A., Leslie, N., and Wenstrup, R. 1998. Enzyme therapy for Gaucher disease: the first 5 years. Blood Rev. 12: 115–133. Grace, M.E., and Grabowski, G.A. 1990. Human acid β-glucosidase: glycosylation is required for catalytic activity. Biochem. Biophys. Res. Commun. 168: 771–777. Grace, M.E., Graves, P.N., Smith, F.I., and Grabowski, G.A. 1990. Analyses of catalytic activity and inhibitor binding of human acid β-glucosidase by site-directed mutagenesis. J. Biol. Chem. 265: 6827–6835. Grace, M.E., Newman, K.M., Scheinker, V., Berg-Fussman, A., and Grabowski, G.A. 1994. Analysis of human acid β-glucosidase by site-directed mutagenesis and heterologous expression, J. Biol. Chem. 269: 2283–2291. Hasnain, S.A., Jain, A., Habib, S., Ghosh, S., Chatterji, U., Ramachandran, A. et al. 1997. Involvement of host factors in transcription from baculovirus very late promoters — a review. Gene, 190: 113–118. Hegedus, D.D., Pfeifer, T.A., Hendry, J., Theilmann, D.A., and Grigliatti, T.A. 1998. A series of broad host range shuttle vectors for constitutive inducible expression of heterologous proteins in insect cell lines. Gene, 207: 241–249. Hoppe, H. 2000. Cerezyme — recombinant protein treatment for Gaucher’s disease. J. Biotechnol. 76: 259–261. Jarvis, D.L. 2003. Developing baculovirus-insect cell expression systems for humanized recombinant glycoprotein production. Virology, 310: 1–7. Jarvis, D.L., and Finn, E.E. 1995. Biochemical analysis of the N-glycosylation pathway in baculovirus-infected lepidopteran insect cells. Virology, 212: 500–511. Jarvis, D.L., Weinkauf, C., and Guarino, L.A. 1996. Immediateearly baculovirus vectors for foreign gene expression in transformed or infected insect cells. Protein Expr. Purif. 8: 191–203. Johansen, H., Van der Straten, A., Sweet, R., Otto, E., Maroni, G., and Rosenberg, M. 1989. Regulated expression at high copy number allows production of a growth-inhibitory oncogene product in Drosophila Schneider cells. Genes Dev. 3: 882–889. Kost, A.T., Condreay, J.A., and Jarvis, D.L. 2005. Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nat. Biotechnol. 23: 567–575. Kretzchmar, E. Geyer, R., and Klenk, H. 1994. Baculovirus infection © 2006 NRC Canada

156 does not alter N-glycosylation in Spodoptera frugiperda cells. J. Biol. Chem. 375: 323–327. Martin, B.M., Tsuji, S., LaMarca, M.E., Maysak, K., Eliason, W., and Ginns, E.I. 1988. Glycosylation and processing of high levels of active human glucocerebrosidase in invertebrate cells using a baculovirus expression vector. DNA, 7: 99–106. Mistry, P.K., Wraight, E.P., and Cox, T.M. 1996. Therapeutic delivery of proteins to macrophages: implications for treatment of Gaucher’s disease. Lancet, 348: 1555–1559. Moscicki, R.A., and Taunton-Rigby, A. 1993. Treatment of Gaucher’s Disease. N. Engl. J. Med. 328: 1564–1568. Pasmanik-chor, M., Elroy-Stein, O., Aerts, H., Agmon, V., Gatt, S., and Horowitz, M. 1996. Overexpression of human glucocerebrosidase containing different-sized leaders. Biochem. J. 317: 81–88. Peters, S.P., Coyle, P., and Glew, R.H. 1976. Differentiation of βglucocerebrosidase from β-glucosidase in human tissues using sodium taurocholate. Arch. Biochem. Biophys. 175: 569–582. Pfeifer, T.A. 1998. Expression of heterologous proteins in stable insect cell culture. Curr. Opin. Biotechnol. 9: 518–521.

Biochem. Cell Biol. Vol. 84, 2006 Pfeifer, T.A., Guarna, M.M., Kwan, E.M., Lesnicki, G., Theilmann, D.A., Grigliatti, T.A., and Kilburn, D.G. 2001. Expression analysis of a modified factor X in stably transformed insect cell lines. Protein Expr. Purif. 23: 233–241. Sorge, J.A., West, C., Kuhl, W., Treger, L., and Beutler, E. 1987. The human glucocerebrosidase gene has two functional ATG initiator codons. Am. J. Hum. Genet. 41: 1016–1024. Spector, D.L., Goldman, R.D., and Leinwand, L.A. 1998. Cells, a laboratory manual. Cold Spring Harbor Laboratory Press, Plainview, N.Y. Xu, Y.H., and Grabowski, G.A. 1998. Translational inefficiency of acid beta-glucosidase mRNA in transgenic mammalian cells. Mol. Genet. Metab. 64: 87–98. Zimmer, K., le Coutre, P., Aerts, J.M.F.G., Harzer, K., Fukuda, M., O’Brien, J.S., and Naim, H.Y. 1999. Intracellular transport of acid β-glucosidase and lysosome-associated membrane proteins is affected in Gaucher disease (G202R mutation). J. Pathol. 188: 407–414.

© 2006 NRC Canada