| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Unité de Physiologie de la Reproduction et des Comportements, Institut National de la Recherche Agronomique (INRA), Centre National de la Recherche Scientifique (CNRS) et Université François Rabelais de Tours, 37 380 Nouzilly, France
1 Unité de Pathologie Comparée, UMR INRA-CNRS, 30380 Saint-Christol-lez-Alès, France
(Requests for offprints should be addressed to S Legardinier; Email: legardin{at}tours.inra.fr)
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
and ß polypeptide chains, but eCG subunits are much more heavily glycosylated and sialylated. Consequently, eCG exhibits a much longer half-life than eLH in blood. Recombinant eLH/CG, expressed in Sf9 and Mimic insect cells, were compared with one another and to the natural hormones eCG and eLH. Mimic cells are stably-transformed Sf9 cells, expressing five mammalian genes encoding glycosyltransferases involved in the synthesis of complex N-carbohydrate chains. Recombinant eLH/CG expressed in Mimic cells exhibited a higher apparent molecular weight (MW) than that expressed in Sf9 cells, suggesting that its N-glycosylation was, as expected, more complete. Nevertheless, the two recombinant eLH/CG exhibited lower MW than natural eCG from pregnant mare plasma. The two eLH/CG produced in Sf9 and Mimic cells were found to be active in in vitro LH and FSH bioassays, with potencies similar to those of eCG. By contrast, they exhibited no significant in vivo bioactivity, neither in the specific follicle-stimulating hormone (FSH) assay nor in the specific eCG assay. Although recombinant eLH/CG produced in Mimic cells bears more elaborate carbohydrate chains than recombinant eLH/CG from Sf9 cells, it exhibits no significant in vivo bioactivity, probably because of insufficient terminal sialylation of its carbohydrate chains, leading to its rapid removal from blood.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
- and ß-subunits. Within the same species, the -subunit is encoded by a single gene and is common to all glycoprotein hormones, whereas different genes encode ß-subunits that confer specificity to the glycoprotein hormone heterodimers (Goverman & Pierce 1981, Combarnous 1992). In equids, in contrast to primates, both LH and CG ß-subunits are encoded by the same gene (Sherman et al. 1992) and consequently the recombinant hormone is called eLH/CG. Both placental eCG and pituitary eLH exhibit dual LH and FSH in non equine species with identical FSH/LH activity ratios (Stewart & Allen 1979, Combarnous et al. 1984). Although eCG and eLH exhibit identical and ß polypeptide chains, their carbohydrate contents are different. Indeed, eCG is the most heavily glycosylated of all glycoprotein hormones, with 45% carbohydrate by weight (Harbon-Chabbat et al. 1961, Gospodarowicz 1972) versus 30% for eLH (Roser et al. 1984, Butnev et al. 1996). The -subunit, composed of 96 amino acids, bears two complex-type N-linked oligosaccharide chains located at asparagines (Asn) 56 and 82 whereas the ß-subunit composed of 149 amino acids has only one at Asn 13. In addition to N-glycans, both eLH and eCG ß-subunits possess a carboxy-terminal peptide (CTP) of 28 amino acids (ß122–149), which is O-glycosylated at the same twelve serine or threonine residues (Bousfield & Butnev 2001). Placental and pituitary hormones also strongly differ by their N-glycan termination with sialic acids (Sia2,3 Gal) on eCG and sulphated N-acetylgalactosamines (SO4–4-GalNAc) on eLH (Smith et al. 1993, Matsui et al. 1994). The remarkable difference in their molecular weight is essentially due to the presence of longer disialylated poly-N-acetyllactosaminyl O-glycans on eCG, with a greater percentage of O-glycans attached at serine and threonine sites (Hokke et al. 1994, Bousfield & Butnev 2001) (Fig. 1
). These structural differences explain why eCG has such an exceptional half-life compared with eLH. Owing to its biological properties, eCG has been used for a long time in fertilization programs. Nevertheless, commercial preparations of partially purified eCG from PMSG could contain contaminants with potential sanitary risks. It is of great interest to produce, in large quantities, a bioactive substitute for eCG and other gonadotropins used as therapeutic agents. Only recombinant gonadotropins produced in mammalian cells have been shown so far to exhibit in vivo biological activity. Such hormones, like hFSH expressed in chinese hamster ovary (CHO) cells (Keene et al. 1989), have even been put on the market as Gonal-F (Serono, Geneva, Switzerland) and Puregon (NV Organon, Oss, The Netherlands). However, the great advantage of non-mammalian systems like yeasts, insect cell lines or plants would be to produce recombinant proteins in larger amounts, at a lower cost, and in the absence of fetal calf serum. Some gonadotropins of zootechnical interest have already been expressed in baculovirus systems (Huang et al. 1991, Inaba et al. 1998, Kato et al. 1998, van de Wiel et al. 1998), in the methylotropic yeast Pichia pastoris (Fidler et al. 1998, 2003, Richard et al. 1998), and in plants (Dirnberger et al. 2001). However, only in vitro biological activities have been reported in these studies, with no information concerning the in vivo potencies of the recombinant hormones produced. Recently, single-chain eLH/CG (Galet et al. 2001) and bFSH (Coulibaly et al. 2002) have been produced in the milk of transgenic rabbits. In vivo biological activity was determined only for eLH/CG, and it was found inactive because its half-life was too short, most likely because the carbohydrate moieties of recombinant glycoproteins in milk (Koles et al. 2004) are unusual compared with those found in serum glycoproteins. The present study deals with the analysis of the biological activities of recombinant eLH/CG produced in two insect cell lines, Sf9 and Mimic cells, the latter derived from Sf9 cells and expressing several mammalian glycosyltransferase genes (Hollister et al. 2002). Mimic cells have been claimed to produce glycoproteins with complex biantennary N-oligosaccharide chains terminating with sialic acids on one branch, instead of truncated trimannose N-glycans as usually found in insect cells (Fig. 1). The present work was undertaken to determine whether recombinant eLH/CG expressed in Mimic cells is more completely glycosylated than that expressed in Sf9 cells and whether, consequently, it exhibits in vivo bioactivities.
|
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Alignments of conserved amino acid sequences of equine (e), donkey (dk), human (h) and chimeric human/equine (he) gonadotropin -subunits are shown in Figure 2, with corresponding nucleotide sequences. The few human nucleotides of chimeric he cDNA (Chopineau et al. 1997) were replaced with those encoding the equine -subunit (e) (Ward et al. 1982, Chopineau & Stewart 1996) using common Bsm I restriction sites. This construct was used as a template to introduce XbaI restriction sites at each end and to remove the 3'untranslated region by PCR using primers A and B (Table 1). Then, the amplified fragment (430 bp) was digested and inserted downstream of the polyhedrin promoter (PH) at a unique XbaI site in a modified pGmAc116T baculovirus transfer vector (Poul et al. 1995) (Fig. 3A).
|
|
|
). The digested PCR product (535 bp) was inserted downstream of the protein p10 promoter (P10) into baculovirus transfer vector p119 (Chaabihi et al. 1993) using the Bgl II and Hind III cloning sites (Fig. 3B).
Truncated cDNA encoding equine LH/CG ß-subunit missing its 122–149 sequence corresponding to the CTP (ße
CTP) was obtained from the full-length cDNA as a template by PCR using primers E and F, introducing a TGA stop codon at position 122 (Table 1). Then, the amplified product (446 bp) was subcloned into pGEM-T and reinserted into baculovirus transfer vector p119 using the indicated restriction sites and sequenced (Fig. 3B).
The e cDNA construct was also cloned in the mammalian expression vector pCDM8 and co-transfected with pCDM8-ße (Chopineau et al. 1997) into COS-7 cells to produce the eße eLH/CG heterodimer in order to compare its in vitro bioactivity with that of the previously studied chimeric heße heterodimer.
PCR was carried out using Taq polymerase (Promega, Charbonnières, France) and amplified PCR products were subcloned into pGEM-T (Promega) and digested using the restriction enzymes indicated (Fig. 3). Resulting fragments were low-melting gel-purified using NuSieve GTG agarose (BMA, Tébu, Le Perray-en-Yvelines, France) and directly ligated into specific baculovirus transfer vectors as described above. The sequences of all the amplified products were checked by the dideoxy chain termination method (Genome Express, Meylan, France). Restriction enzymes were purchased from Roche (Meylan, France) and oligonucleotides from Eurogentec (Seraing, Belgium).
Transient transfections of COS-7 cells
COS-7 monkey kidney cells (ATCC-CRL 1651) were maintained at 37 °C in Dulbecco’s modified Eagle’s medium (DMEM) without sodium pyruvate; supplemented with 10% heat-inactivated fetal bovine serum (FBS), 10 mM HEPES buffer, 2 mM L-glutamine (Gln), 100 units/ml penicillin and 100 µg/ml streptomycin. COS-7 cells were transiently transfected using the calcium phosphate precipitation procedure. Briefly, cells were cotransfected at 75% of confluency in duplicate in 6 cm diameter Petri dishes with 1 µg of each construct in serum-free media. pCDM8-ße was co-transfected with pCDM8 containing he, or e cDNAs. After 48 h incubation at 37 °C, media were recovered by centrifugation at 100 g for 5 min and stored at –20 °C until assayed. Unless otherwise stated, all other reagents used in cell culture were purchased from Invitrogen (Cergy Pontoise, France).
Cells and virus infection
The Sf9 subclone of Spodoptera frugiperda Sf21 cells (Vaughn et al. 1977) was maintained at 28 °C in supplemented TC-100 growth medium containing 5% heat-inactivated FBS, 50 units/ml penicillin G and 50 µg/ml streptomycin. Mimic cells (Invitrogen) were cultured in complete Grace’s Insect Medium, supplemented with 10% FBS and antibiotics as above. Both insect cell lines were maintained as adherent cells at a density of 2•0 x 106 to 2•5 x 106 cells per ml in 25 cm2 flasks (Falcon) and passed over 80% confluency three times a week. To reach the same cell density as that obtained with Sf9 cells, more Mimic cells had to be passed because of their slower growth (but never more than 30 times, as recommended by the manufacturer). The viruses were propagated in both cell lines and recovered as described previously (Summers & Smith 1987). Cells were infected or co-infected with recombinant baculoviruses at a multiplicity of infection (MOI) ranging from 5 to 10 plaque forming units (Pfu) per cell. After 45 min incubation with viral suspensions, fresh culture media were added and cells were incubated at 28 °C until day 5 post-infection. The viral titers were determined by plaque assay (Summers & Smith 1987).
Construction of the recombinant baculoviruses
Sf9 cells were cotransfected with 5 µg transfer vector and 500 ng purified viral DNA using a lipofection method (DOTAP, Roche, Meylan, France). Two distinct viral DNA preparations were used: (i) viral DNA extracted from the wild-type Autographa californica multiple nucleocapsid polyhedrosis virus (AcMNPV), when a polyhedrin-specific transfer vector such as pGmAc116T-e was used or (ii) viral DNA extracted from the baculovirus AcSLP10 (Chaabihi et al. 1993) which possesses only one very late promoter, the P10 promoter, driving the expression of the polyhedrin gene when the P10 specific transfer vectors such as p119-ße and p119-ße Delta;CTP were used. In each case, recombinant baculoviruses were selected by plaque assay and distinguished from the wild-type progeny by their occlusion body-negative phenotype. The screening and purification of the recombinant baculoviruses NPV-e (PH), SLP10-ße (P10) and SLP10-ße Delta;CTP (P10) were carried out as previously described (Summers & Smith 1987).
Quantitation of expressed recombinant hormones in enzyme-linked immunosorbent assay (ELISA)
The concentrations of the recombinant eLH/CG dimeric hormones produced in COS-7 cells or insect cells culture media were determined in a sandwich ELISA as previously described (Galet et al. 2000). The microtitration plates from Nunc (Maxisorp, VWR, Strasbourg, France) were coated overnight at 4 °C with 100µl per well 89A2 mouse monoclonal antibody (1µg/ml) (Intervet, Boxmeer, Holland) recognizing the alpha chain as single subunit and as heterodimer in eCG (Chopineau et al. 1993). After five washes in 0•1% Tween 20–PBS (1•15 g/l Na2HPO4, 0•2 g/l KH2Po4, 8 g/l NaCl, 0•2 g/l KCl, Dulbecco) and saturation of non-specific sites with 0•2% bovine serum albumin (BSA) (Sigma, Saint-Quentin Fallavier, France) in 0•1% Tween 20–PBS, 100 µl per well of standard eCG NZY-01 (1500 IU/mg) (Lecompte et al. 1998) or media containing recombinant eLH/CG were incubated at different concentrations for 1 h at 4 °C. After five washes with 0•1% Tween20-PBS, a rabbit polyclonal antibody raised against highly-purified eCG (Cahoreau & Combarnous 1987) was added at a 1:100 000 dilution for 1 h at 4 °C. A peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, Interchim, Montluçon, France) added at 1:10 000 dilution for 1 h at 4 °C and, after five washes, 100 µl/well SureBlue TMB peroxidase substrate was added (KPL, Eurobio, Les Ulis, France) and the mixture was incubated for 20 min at room temperature in the dark. The reaction was stopped with 50 µl 2N H2SO4 and the absorbance was measured at 450 nm using a SpectraCount (Packard, Downers Grove, IL, USA) spectrophotometer and data were analyzed by the I-SMART software (Packard, Downers Grove, IL, USA).
Western blot analysis of the recombinant dimeric hormones and free - and ß-subunits
Sf9 or transgenic Mimic insect cells were seeded in 25 cm2 flasks and inoculated separately or simultaneously with recombinant NPV-e (PH), SLP10-ße (P10) and SLP10-ße Delta;CTP (P10). Supernatants were recovered by centrifugation at 100 g for 5 min and aliquots (30 µl) were diluted in Laemmli’s 4 x buffer (Laemmli 1970) under non reducing or reducing conditions (5% ß-mercaptoethanol). Unheated or heated samples were electrophoresed in 10 or 14% sodium dodecyl sulphate polyacrylamide gels (SDS-PAGE). After separation, proteins were transferred overnight at 4 °C onto reinforced nitrocellulose membranes (Schleicher and Shuell, Ecquevilly, France) within a mini-transblot apparatus in 20% v/v propanol–2/Tris–glycine (25 mM Tris, 0•192 M glycine) buffer from Amresco (Interchim, Montluçon, France). Membranes were satured for 30 min at room temperature in Tris buffer saline (TBS; 25 mM Tris, 140 mM NaCl, 3 mM KCl pH 7•4) Amresco containing 0•05% Tween 20 and 5% non-fat milk and then incubated for 2 h with specific antibodies. Dimeric hormones were specifically revealed using rabbit polyclonal anti-eCG antibody (Cahoreau & Combarnous 1987) at 1: 50 000 dilution. Free and ßsubunits with or without CTP were revealed using 10 ml mouse monoclonal antibody 89A2 (1µg/ml) (Intervet) (Chopineau et al. 1993) and a rabbit polyclonal antibody (dilution 1:5000) raised against a synthetic LH/CG ß1–9 peptide representing the nine N-terminal residues of LH and CG ß-subunits from numerous species (M Chopineau & C Galet, unpublished). This peptide with highly-conserved sequence among all LH and CG ß-subunits was conjugated to Keyhole Limpet hemocyanin using carbodiimide as the coupling agent. The membranes were washed three times for 10 min in TBS–0•05% Tween 20 and incubated for 45 min with peroxidase-conjugated goat anti-rabbit or anti-mouse IgG (Jackson ImmunoResearch; Interchim, Montluçon, France) diluted 1:5000. After three washes, the membranes were analysed using SuperSignal, West Pico (Pierce).
In vitro LH and FSH bioassays
LH and FSH bioactivities of recombinant hormones were estimated by the in vitro stimulation of progesterone production respectively in mouse Leydig tumoral cells MLTC-1 (ATCCCRL 2065) and in a mouse adrenal cortex tumor stably expressing the human FSH receptor, referred to as Y1 cells (kindly given by Ares Serono, Geneva, Switzerland) (Kelton et al. 1992). For the in vitro LH bioassay, about 1•5 x 105 MLTC-1 cells per well were incubated at 37 °C in 0•5 ml supplemented RPMI growth medium (10% FBS, 50 µg/ml gentamicin, 10 units/ml penicillin and 10 µg/ml streptomycin) in 48-well plates until 80% confluency. Just before stimulation, cells were incubated for 2 h in 0•5 ml serum-free RPMI media and then stimulated for 2 h with 0•5 ml serum-free RMPI media containing samples of recombinant hormones at different concentrations. The supernatants were recovered and stored at –20 °C until assayed. To estimate the in vitro FSH bioactivity, 1•5 x 105 Y1 cells were seeded in 0•5 ml Ham’s F10 growth medium (supplemented with 15% horse serum (HS), 2•5% FBS, 2 mM L-Gln and 80µg/ml Geneticin (G418) in 48-well plates and incubated at 37 °C for 3 or 4 days. Twenty-four hours before the stimulation, cells were cultured in 0•5 ml F10 media that was poorer in serum (5% HS, 0•8% FBS, 2 mM Gln and 80µg/ml G418). Finally, Y1 cells were stimulated with 0•5 ml of the same samples as above in serum-free F10 medium containing 1 mg/ml BSA and 2 mM L-Gln. After an incubation of 4 h at 37 °C, media were stored at –20 °C until assayed.
Secreted progesterone recovered in media from MLTC-1 or Y1 cells was measured by a specific radioimmunoassay (Saumande & Batra 1985). Relative LH and FSH potencies of recombinant hormones in COS-7 cells and in insect cells were determined respectively with heße and standard eCG NZY-01 as references. Relative biopotencies were calculated on the basis of ED50 corresponding to the dose necessary to obtain 50% of the maximal production.
Animals
Immature 25-day old female Wistar rats were obtained from the animal breeding facility of the laboratory. All the procedures used in this paper were in compliance with the European Community Council Directive of 24 November 1986 (86/609/EEC).
In vivo eCG and FSH bioassays
Supernatants containing recombinant eLH/CG from Sf9 or Mimic cells were harvested 5 days postinfection and were centrifuged for 10 min at 1000 g. After dialysis, the samples were frozen at –80 °C, lyophilised and diluted in 0•15 M saline on the basis of their immunoactivity in ELISA. The in vivo eCG bioactivity of recombinant hormones expressed in insect cells was determined in the eCG test of the Pharmacopoeia (Cole & Erway 1941). The in vivo FSH bioactivity was evaluated using a specific FSH bioassay of the Pharmacopoeia (Steelman & Pohley 1953). Five animals were injected s.c. for each dose of hormone and one to four doses were used per hormone to be tested.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The quantities of recombinant hormones secreted by COS-7 and insect cells were estimated in a sandwich ELISA using standard eCG NZY-01 (1500 IU/mg) as reference and also compared with pituitary eLH (Fig. 4). Briefly, the pCDM8-ße was cotransfected in COS-7 cells in serum free medium, either with pCDM8-he or with the pCDM8-e plasmid. After 48 h incubation, the media were recovered and assayed by ELISA. The productions of recombinant heße and e ße were 0•123 ± 0•003 µg/ml and 0•094 ± 0•006 µg/ml respectively. Mimic cells and Sf9 cells were both coinfected with NPV-e(PH) and SLP10-ße(P10) recombinant baculoviruses. Supernatants from infected cells were recovered by centrifugation at 100 g for 5 min, 5 days post-infection and assayed by ELISA. Maximal productions of eLH/CG in Sf9 and Mimic cells were found to be 4•15 ± 0•49 µg/ml and 5•33 ± 1•32 µg/ml, respectively, 5 days post infection (n=5) and thus 30–50 fold more than in COS-7 cells (not shown).
|
To express free and ß subunits with or without CTP in the baculovirus insect cell system, Sf9 or Mimic cells were seeded in 25 cm2 flasks and infected with recombinant NPV-e(PH), SLP10-ße(P10) and SLP10-ßeCTP(P10) respectively. Complete eLH/CG was produced by coinfection of insect cells with NPV-e(PH) and SLP10-ße(P10). After five days incubation at 28 ° C,supernatants were collected and analysed by Western blotting using specific antibodies.
Standard and recombinant dimeric hormones were detected using rabbit antisera directed against native eCG (Cahoreau & Combarnous 1987). Standard eCG exhibited a complex range of bands between 80 and 100 kDa while the recombinant dimeric hormones expressed in Sf9 cells and Mimic cells were more homogenous appearing as a doublet at about 45 and 50 kDa respectively (Fig. 5A). A strong specific signal was revealed at about 26 kDa for the -subunit, with native heated eCG using 89A2 monoclonal antibody recognizing conformational -subunit under non reducing conditions (Fig. 5B) (Chopineau et al. 1993). Recombinant -subunit appeared as a major band at about 22 kDa in Sf9 cells versus 26 kDa in Mimic cells.
|
). A single band was detected for the ßeCTP at about 16 kDa in Sf9 cells and a doublet at the same MW in Mimic cells (Fig. 5D). In vitro bioactivities
In vitro LH and FSH bioactivities of recombinant hormones were determined by their ability to stimulate progesterone production by MLTC-1 and Y1 cells, respectively. The dimeric eße hormone expressed in COS-7 cells exhibited LH and FSH potencies, respectively 22 and 27% lower than those of heße (Figs 6A and B, respectively) leading to a slight non-significant decrease in its FSH/LH activity ratio (Table 2).
|
|
). The in vitro LH biopotency of eLH/CG expressed in Mimic cells was only 20% lower than that of natural eCG and its in vitro FSH biopotency was unaffected (Fig. 7B). Equine LH/CG produced in Mimic cells possessed a FSH/LH ratio of 1•2, versus 2•0 when expressed in Sf9 cells (Table 2).
|
The eCG in vivo bioassay (Cole & Erway 1941) was used to compare the biopotency of eLH/CG produced in Sf9 cells with that produced in Mimic cells. The unique injection of standard eCG leads to a dose-dependent increase in ovarian weight. Equine LH/CG expressed by Sf9 or Mimic cells had no effect on ovary growth (Fig. 8). Pituitary eLH was also compared with eCG in this in vivo bioassay, on the basis of their respective potencies in the sandwich ELISA. Equine LH isoforms were found to be 10–30 fold less active than natural eCG in the in vivo eCG assay (not shown). The FSH in vivo bioassay (Steelman & Pohley 1953) was carried out to specifically test FSH bioactivity of Mimic eLH/CG. The repeated injection of standard hFSH or eCG NZY-01 over 4 days induced a dose-dependent increase in ovarian weight under hCG stimulation. However, both recombinant eLH/CG had no effect on ovary growth (Fig. 9).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
) but exhibit sulfated terminal N-acetylgalactosamine residues in addition to, or in place, of sialic acid residues (Smith et al. 1993, Matsui et al. 1994). The sites of O-glycosylation of eLH are also the same as is eCG but, on average, they represent a smaller proportion and the chains are shorter (Bousfield & Butnev 2001). Equine CG presents a much longer half-life (6 days) than eLH (1 h) in mares (Cole et al. 1967, Ginther et al. 1974) in relation to its carbohydrate content. Indeed, enzymatic removal of terminal sialic acid residues of eCG leads to a dramatic drop in its in vivo bioactivity in various species, including sheep (Martinuk et al. 1991) and rats (Rafelson et al. 1961). Consequently, the structure of carbohydrate side-chains in recombinant eLH/CG is expected to be an important issue for its clearance from blood, and consequently for its in vivo bioactivity. The present paper reports the production and biological characterization of recombinant eLH/CG produced in Sf9 and Mimic insect cell lines using the baculovirus expression system. The correction of the original chimeric he sequence into fully e sequence permitted the production of eLH/CG with its genuine sequence in COS-7 cells, as well as in Sf9 and Mimic cells. The replacement of the human N-terminal APDV sequence in he by the equine N-terminal FPDGEFTT sequence in e led to a slight decrease in both LH and FSH in vitro activities of the recombinant heterodimer formed with ße. Since the quantity of hormones tested in bioassays are determined by ELISA, this decrease actually relates to the bioactivity to immunoactivity ratio. Therefore, it cannot be ruled out that the observed decrease could in fact be due to a slightly better recognition of eße than heße in the ELISA. This would not be unexpected, as the -specific 89A2 monoclonal antibody used in the sandwich ELISA was raised against authentic eCG, thus having an N-terminal FPDGEFTT sequence in its -subunit. Moreover, the -subunit N-termini of hCG and hFSH are known to be on the opposite side relative to their hormone binding sites to their cognate receptors (Lapthorn et al. 1994, Wu et al. 1994, Fox et al. 2001). The in vivo biological activity of eCG strongly depends on the presence of sialylated complex saccharide side chains leading to its slow rate of elimination from blood (Klett et al. 2003). The baculovirus expression system allows high levels of production of recombinant proteins, but N-glycans processed in insect cells such as Sf9 cells are not complex and sialylated N-glycans such as those found in placental eCG. To overcome this problem, we expressed recombinant eLH/CG in Mimic cells that have been designed to produce biantennary, terminally sialylated N-glycans (Hollister et al. 2002). Recombinant eLH/CG with both subunits possessing the native amino acid equine sequences was successfully produced in two insect cell lines using the baculovirus expression system, with yields of 4–5 mg/l that are similar to those previously reported for the production of porcine FSH in insect cells (1 mg/l) (Kato et al. 1998) and of bovine FSH (1–5 mg/l) (van de Wiel et al. 1998).
Recombinant eLH/CG and its subunits expressed in Sf9 and Mimic cells exhibited lower apparent molecular weights compared with natural eCG and its subunits, highlighting the presence of shorter glycans in these cells. Nevertheless, the higher apparent molecular weights found for eLH/CG and its subunits expressed in Mimic cells compared with Sf9 cells strongly suggest that glycosylation was indeed improved in Mimic cells as expected (Hollister et al. 2002). The major differences in MW between natural and recombinant hormones seem to be essentially due to very short O-glycans in the CTP of eLH/CG produced in insect cells. Indeed, the -subunit, which is only N-glycosylated, exhibited the same apparent MW when expressed in Mimic cells as that of -subunit from natural eCG. On the contrary, the ß-subunit and eLH/CG exhibited significantly lower MW than their natural counterparts, suggesting that the major effect on MW arises from weaker O-glycosylation of the CTP. Indeed, the CTP is O-glycosylated in insect cells since its elimination in ßeDelta;CTP leads to a single band, while full-length ß-subunit exhibits higher and heterogeneous MW. In Sf9 cells, O-glycans are constituted by single N-acetylgalactosamine (GalNAc) residues or by GalNAc-Gal disaccharides (Thomsen et al. 1990) but for the time being no information on this matter is available for Mimic cells. Recombinant glycoproteins produced in Mimic cells referred to as SfSWT-1 were claimed to bear sialylated biantennary N-glycans (Hollister et al. 2002, 2003). Sialylation in Mimic cells is unexpected since Sf9 cells do not synthesize the CMP-sialic acid nucleotidic donor (CMP-NeuAc) as a substrate for 2,6-sialyltransferase. Our observation that recombinant eLH/CG from Mimic cells do not exhibit any in vivo activity in spite of full in vitro activity strongly suggests that the hormone is probably not sialylated. Indeed, desialylation of natural eCG leads to the complete loss of its in vivo activity without affecting its in vitro activity (Morell et al. 1971).
Both eLH/CGs expressed in Sf9 and Mimic cells exhibited full in vitro LH and FSH bioactivities, confirming that recombinant hormones possess a conformation similar to that of natural eCG. Equine LH/CG expressed in Mimic cells and Sf9 cells exhibit in vitro FSH/LH ratios that are not significantly different. On the basis of their concentrations measured by ELISA, their in vivo bioactivity was not detectable at the doses used and they are less than 1% of that of natural eCG. Recombinant eLH/CG produced in Mimic cells was also found to be inactive in the FSH in vivo bioassay (Steelman & Pohley 1953) that does not require a hormone with a very long half-life, in contrast to the eCG in vivo bioassay (Cole & Erway 1941). The absence of in vivo bioactivity of recombinant eLH/CG may be the result of a rapid clearance of the hormone in the plasma, as is the case for single-chain eLH/CG expressed in the milk of transgenic rabbits (Galet et al. 2001). Sialic acids are mainly responsible for prolonged half-life in plasma, suggesting they are partially or totally lacking on eLH/CG produced in Sf9 cells and in Mimic cells. Native eLH is known to bear sulphated N-glycans and sialylated O-glycans. Although displaying a similar in vitro FSH/LH ratio (Guillou & Combarnous 1983), native eLH that bears fewer sialic acids compared with eCG exhibits a much shorter half life. The presence of sialylated O-glycans is also involved in the half-life, since the elimination of the O-glycosylated CTP from hCG led to a drastic decrease of its half-life (Matzuk et al. 1990). The absence of in vivo bioactivity of eLH/CG expressed in Sf9 cells can be explained by the presence of trimannose side-chains that can be recognized by the macrophage mannose receptor (Childs et al. 1990), leading to removal of the hormone from blood. Although eLH/CGs expressed in Mimic cells present more elaborate carbohydrate side chains, they might still be recognized by the MMR, since its carbohydrate recognition domains bind fucose and N-acetyl-glucosamine in addition to mannose, and its cystine-rich domain binds sulfated GalNAc and various sulfated carbohydrates (Leteux et al. 2000). The rapid elimination of porcine LH from blood in rats has been shown to occur by rapid trapping by kidneys (Klett et al. 2003) but it does not seem to involve MMR, since the half-life of pLH is not increased in MMR–/– mice compared with wild-type mice (Lee et al. 2002). Nevertheless, we have no clue at the moment whether recombinant hormones expressed in insect cells are eliminated from the circulation in the same way as pLH. We have shown in the present paper that eLH exhibits low in vivo bioactivity (3–10% that of eCG) but this activity is easily detectable, in contrast to that of recombinant eLH/CG. This indicates that recombinant hormones are, indeed probably eliminated from blood more rapidly than pLH and eLH. Therefore, the involvement of MMR present at the surface of macrophages cannot be ruled out in the elimination of recombinant eLH/CG.
In brief, we succeeded in producing two recombinant eLH/CG in the baculovirus expression system using Sf9 cells and their transgenic progeny Mimic cells. Although having full in vitro LH and FSH bioactivities, eLH/CG expressed in both cell lines did not exhibit significant in vivo bioactivity. Nevertheless, our results clearly show that the glycosyltransferase activities in Mimic cells lead to larger carbohydrate moieties on baculovirus-expressed proteins. However, other modifications such as the addition of O-glycans and sialic acid residues are still very incomplete and will need further attention. At the least, Mimic cells are expected to synthesize glycoproteins with N-glycans up to terminal galactose residues and so are good substrates for a more complete in vitro sialylation in contrast to those of eLH/CG from Sf9 cells. Two strategies can be proposed to engineer N-and O-glycosylation: one by modifying the cell genome of insect cells and the other by introducing all necessary glycosyltransferase genes in the baculovirus genome. A recent study showed the successful production of sialylated human 1-antitrypsin in Estigmene acrea Ea4 cells transfected by a single recombinant baculo-virus co-expressing N-acetylglucosaminyl transferase II, ß1,4-galactosyl transferase and 2,6-sialyltransferase in addition to human 1-antitrypsin (Chang et al. 2003).
|
Acknowledgements |
|---|
The authors are grateful to D Aguer and MA Driancourt (Intervet Pharma R&D, Beaucouzé, France) and P Sondermeijer (Intervet, Boxmeer, The Netherlands) for valuable discussions. We are thankful to C Cahier, J C Braguer and J C Poirier (INRA Nouzilly France) for skilfull technical assistance. We acknowledge M Chopineau and C Galet (INRA Nouzilly, France) for the gift of donkey cDNA constructs and antipeptide serum. We also thank A Ythier (Serono Ares, Geneva) for the gift of Y1 cells expressing the human FSH receptor.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Bousfield GR & Butnev VY 2001 Identification of twelve O-glycosylation sites in equine chorionic gonadotropin beta and equine luteinizing hormone ss by solid-phase Edman degradation. Biology of Reproduction 64 136–147.
Butnev VY, Gotschall RR, Baker VL, Moore WT & Bousfield GR 1996 Negative influence of O-linked oligosaccharides of high molecular weight equine chorionic gonadotropin on its luteinizing hormone and follicle-stimulating hormone receptor-binding activities. Endocrinology 137 2530–2542.[Abstract]
Cahoreau C & Combarnous Y 1987 Comparison of two reference preparations for horse chorionic gonadotrophin in four in-vivo and in-vitro assays. Journal of Reproduction and Fertility 79 281–287.[Abstract]
Chaabihi H, Ogliastro MH, Martin M, Giraud C, Devauchelle G & Cerutti M 1993 Competition between baculovirus polyhedrin and p10 gene expression during infection of insect cells. Journal of Virology 67 2664–2671.
Chang GD, Chen CJ, Lin CY, Chen HC & Chen H 2003 Improvement of glycosylation in insect cells with mammalian glycosyltransferases. Journal of Biotechnology 102 61–71.[CrossRef][ISI][Medline]
Childs RA, Feizi T, Yuen CT, Drickamer K & Quesenberry MS 1990 Differential recognition of core and terminal portions of oligosaccharide ligands by carbohydrate-recognition domains of two mannose-binding proteins. Journal of Biological Chemistry 265 20770–20777.
Chopineau M & Stewart F 1996 Cloning and analysis of the cDNA for the common alpha-subunit of the donkey pituitary glycoprotein hormones. Journal of Molecular Endocrinology 16 9–13.[Abstract]
Chopineau M, Maurel MC, Combarnous Y & Durand P 1993 Topography of equine chorionic gonadotropin epitopes relative to the luteinizing hormone and follicle-stimulating hormone receptor interaction sites. Molecular and Cellular Endocrinology 92 229–239.[CrossRef][ISI][Medline]
Chopineau M, Stewart F & Allen WR 1995 Cloning and analysis of the cDNA encoding the horse and donkey luteinizing hormone beta-subunits. Gene 160 253–256.[CrossRef][ISI][Medline]
Chopineau M, Martinat N, Troispoux C, Marichatou H, Combarnous Y, Stewart F & Guillou F 1997 Expression of horse and donkey LH in COS-7 cells: evidence for low FSH activity in donkey LH compared with horse LH. Journal of Endocrinology 152 371–377.[Abstract]
Cole HH & Erway J 1941 48-Hour assay test for equine gonadotropin with results expressed in international units. Endocrinology 29 514–519.
Cole HH, Bigelow M, Finkel J & Rupp GR 1967 Biological half-life of endogenous PMS following hysterectomy and studies on losses in urine and milk. Endocrinology 81 927–930.[ISI][Medline]
Combarnous Y 1992 Molecular basis of the specificity of binding of glycoprotein hormones to their receptors. Endocrine Reviews 13 670–691.[CrossRef][ISI][Medline]
Combarnous Y, Guillou F & Martinat N 1984 Comparison of in vitro follicle-stimulating hormone (FSH) activity of equine gonadotropins (luteinizing hormone, FSH, and chorionic gonadotropin) in male and female rats. Endocrinology 115 1821–1827.[Abstract]
Coulibaly S, Besenfelder U, Miller I, Zinovieva N, Lassnig C, Kotler T, Jameson JL, Gemeiner M, Muller M & Brem G 2002 Expression and characterization of functional recombinant bovine follicle-stimulating hormone (boFSHalpha/beta) produced in the milk of transgenic rabbits. Molecular Reproduction and Development 63 300–308.[CrossRef][ISI][Medline]
Dirnberger D, Steinkellner H, Abdennebi L, Remy JJ & van de Wiel D 2001 Secretion of biologically active glycoforms of bovine follicle stimulating hormone in plants. European Journal of Biochemistry 268 4570–4579.[ISI][Medline]
Fidler AE, Lun S, Young W & McNatty KP 1998 Expression and secretion of a biologically active glycoprotein hormone, ovine follicle stimulating hormone, by Pichia pastoris. Journal of Molecular Endocrinology 21 327–336.[Abstract]
Fidler AE, Lin JS, Lun S, Ng Chie W, Western A, Stent V & McNatty KP 2003 Production of biologically active tethered ovine FSHbetaalpha by the methylotrophic yeast Pichia pastoris. Journal of Molecular Endocrinology 30 213–225.[Abstract]
Fox KM, Dias JA & Van Roey P 2001 Three-dimensional structure of human follicle-stimulating hormone. Molecular Endocrinology 15 378–389.
Galet C, Chopineau M, Martinat N, Combarnous Y & Guillou F 2000 Expression of an in vitro biologically active equine LH/CG without C-terminal peptide (CTP) and/or beta26–110 disulphide bridge. Journal of Endocrinology 167 117–124.[Abstract]
Galet C, Le Bourhis CM, Chopineau M, Le Griec G, Perrin A, Magallon T, Attal J, Viglietta C, Houdebine LM & Guillou F 2001 Expression of a single betaalpha chain protein of equine LH/CG in milk of transgenic rabbits and its biological activity. Molecular and Cellular Endocrinology 174 31–40.[CrossRef][ISI][Medline]
Ginther OJ, Pineda MH, Wentworth BC & Nuti L 1974 Rate of disappearance of exogenous LH from the blood in mares. Journal of Animal Science 39 397–403.
Gospodarowicz D 1972 Purification and physicochemical properties of the pregnant mare serum gonadotropin (PMSG). Endocrinology 91 101–106.[ISI][Medline]
Goverman JM & Pierce JG 1981 Differential effects of alkylation of methionine residues on the activities of pituitary thyrotropin and lutropin. Journal of Biological Chemistry 256 9431–9435.
Guillou F & Combarnous Y 1983 Purification of equine gonadotropins and comparative study of their acid-dissociation and receptor-binding specificity. Biochimica et Biophysica Acta 755 229–236.[Medline]
Harbon-Chabbat S, Legault-Demare J, Jolles P & Clauser H 1961 Chemical study of a pure serum gonadotropic hormone preparation (PMSG). Bulletin de la Société de Chimie Biologique (Paris) 43 13391345.
Hokke CH, Roosenboom MJ, Thomas-Oates JE, Kamerling JP & Vliegenthart JF 1994 Structure determination of the disialylated poly-(N-acetyllactosamine)-containing O-linked carbohydrate chains of equine chorionic gonadotropin. Glycoconjugate Journal 11 35–41.[CrossRef][ISI][Medline]
Hollister J, Grabenhorst E, Nimtz M, Conradt H & Jarvis DL 2002 Engineering the protein N-glycosylation pathway in insect cells for production of biantennary, complex N-glycans. Biochemistry 41 15093–15104.[CrossRef][Medline]
Hollister J, Conradt H & Jarvis DL 2003 Evidence for a sialic acid salvaging pathway in lepidopteran insect cells. Glycobiology 13 487–495.
Huang CJ, Huang FL, Chang GD, Chang YS, Lo CF, Fraser MJ & Lo TB 1991 Expression of two forms of carp gonadotropin alpha subunit in insect cells by recombinant baculovirus. PNAS 88 7486–7490.
Inaba T, Mori J, Ohmura M, Tani H, Kato Y, Tomizawa K, Kato T, Ihara T, Sato I & Ueda S 1998 Recombinant porcine follicle stimulating hormone produced in baculovirus-insect cells induces rat ovulation in vivo and gene expression of tissue plasminogen activator in vitro. Research in Veterinary Science 64 25–29.[CrossRef][ISI][Medline]
Kato Y, Sato I, Ihara T, Tomizawa K, Mori J, Geshi M, Nagai T, Okuda K, Kato T & Ueda S 1998 Expression and purification of biologically active porcine follicle-stimulating hormone in insect cells bearing a baculovirus vector. Journal of Molecular Endocrinology 20 55–65.[Abstract]
Keene JL, Matzuk MM, Otani T, Fauser BC, Galway AB, Hsueh AJ & Boime I 1989 Expression of biologically active human follitropin in Chinese hamster ovary cells. Journal of Biological Chemistry 264 4769–4775.
Kelton CA, Cheng SV, Nugent NP, Schweickhardt RL, Rosenthal JL, Overton SA, Wands GD, Kuzeja JB, Luchette CA & Chappel SC 1992 The cloning of the human follicle stimulating hormone receptor and its expression in COS-7, CHO, and Y-1 cells. Molecular and Cellular Endocrinology 89 141–151.[CrossRef][ISI][Medline]
Klett D, Bernard S, Lecompte F, Leroux H, Magallon T, Locatelli A, Lepape A & Combarnous Y 2003 Fast renal trapping of porcine luteinizing hormone (pLH) shown by 123I-scintigraphic imaging in rats explains its short circulatory half-life. Reproductive Biology and Endocrinology 1 64.
Koles K, van Berkel PH, Pieper FR, Nuijens JH, Mannesse ML, Vliegenthart JF & Kamerling JP 2004 N- and O-glycans of recombinant human C1 inhibitor expressed in the milk of transgenic rabbits. Glycobiology 14 51–64.
Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227 680–685.[CrossRef][Medline]
Lapthorn AJ, Harris DC, Littlejohn A, Lustbader JW, Canfield RE, Machin KJ, Morgan FJ & Isaacs NW 1994 Crystal structure of human chorionic gonadotropin. Nature 369 455–461.[CrossRef][Medline]
Lecompte F, Roy F & Combarnous Y 1998 International collaborative calibration of a preparation of equine chorionic gonadotrophin (eCG NZY-01) proposed as a new standard. Journal of Reproduction and Fertility 113 145–150.[Abstract]
Lee SJ, Evers S, Roeder D, Parlow AF, Risteli J, Risteli L, Lee YC, Feizi T, Langen H & Nussenzweig MC 2002 Mannose receptor-mediated regulation of serum glycoprotein homeostasis. Science 295 1898–1901.
Leteux C, Chai W, Loveless RW, Yuen CT, Uhlin-Hansen L, Combarnous Y, Jankovic M, Maric SC, Misulovin Z, Nussenzweig MC et al. 2000 The cysteine-rich domain of the macrophage mannose receptor is a multispecific lectin that recognizes chondroitin sulfates A and B and sulfated oligosaccharides of blood group Lewis(a) and Lewis(x) types in addition to the sulfated N-glycans of lutropin. Journal of Experimental Medicine 191 1117–1126.
Martinuk SD, Manning AW, Black WD & Murphy BD 1991 Effects of carbohydrates on the pharmacokinetics and biological activity of equine chorionic gonadotropin in vivo. Biology of Reproduction 45 598–604.[Abstract]
Matsui T, Mizuochi T, Titani K, Okinaga T, Hoshi M, Bousfield GR, Sugino H & Ward DN 1994 Structural analysis of N-linked oligosaccharides of equine chorionic gonadotropin and lutropin beta-subunits. Biochemistry 33 14039–14048.[CrossRef][Medline]
Matzuk MM, Hsueh AJ, Lapolt P, Tsafriri A, Keene JL & Boime I 1990 The biological role of the carboxyl-terminal extension of human chorionic gonadotropin [corrected] beta-subunit. Endocrinology 126 376–383.[Abstract]
Morell AG, Gregoriadis G, Scheinberg IH, Hickman J & Ashwell G 1971 The role of sialic acid in determining the survival of glycoproteins in the circulation. Journal of Biological Chemistry 246 1461–1467.
Poul MA, Cerutti M, Chaabihi H, Devauchelle G, Kaczorek M & Lefranc MP 1995 Design of cassette baculovirus vectors for the production of therapeutic antibodies in insect cells. Immunotechnology 1 189–196.[CrossRef][ISI][Medline]
Rafelson MEJr., Clauser H & Legault-Demare J 1961 Removal of sialic acid from serum gonadotropin by acidic and enzymic hydrolysis. Biochimica et Biophysica Acta 47 406–407.[Medline]
Richard F, Robert P, Remy JJ, Martinat N, Bidart JM, Salesse R & Combarnous Y 1998 High-level secretion of biologically active recombinant porcine follicle-stimulating hormone by the methylotrophic yeast Pichia pastoris. Biochemical and Biophysical Research Communications 245 847–852.[CrossRef][ISI][Medline]
Roser JF, Papkoff H, Murthy HM, Chang YS, Chloupek RC & Potes JA 1984 Chemical, biological and immunological properties of pituitary gonadotropins from the donkey (Equus asinus): comparison with the horse (Equus caballus). Biology of Reproduction 30 1253–1262.[Abstract]
Saumande J & Batra SK 1985 Superovulation in the cow: comparison of oestradiol-17 beta and progesterone patterns in plasma and milk of cows induced to superovulate; relationships with ovarian responses. Journal of Endocrinology 107 259–264.[Abstract]
Sherman GB, Wolfe MW, Farmerie TA, Clay CM, Threadgill DS, Sharp DC & Nilson JH 1992 A single gene encodes the beta-subunits of equine luteinizing hormone and chorionic gonadotropin. Molecular Endocrinology 6 951–959.[Abstract]
Smith PL, Bousfield GR, Kumar S, Fiete D & Baenziger JU 1993 Equine lutropin and chorionic gonadotropin bear oligosaccharides terminating with SO4–4-GalNAc and Sia alpha 2,3 Gal, respectively. Journal of Biological Chemistry 268 795–802.
Steelman SL & Pohley FM 1953 Assay of the follicle stimulating hormone based on the augmentation with human chorionic gonadotropin. Endocrinology 53 604–616.
Stewart F & Allen WR 1979 The binding of FSH, LH and PMSG to equine gonadal tissues. Journal of Reproduction and Fertility. Suppl 431–440.
Summers MD & Smith GE 1987 A manual of methods for baculovirus vectors and insect cell culture procedures. Texas Agricultural Experimental Station Bulletin No. 1555.
Thomsen DR, Post LE & Elhammer AP 1990 Structure of O-glycosidically linked oligosaccharides synthesized by the insect cell line Sf9. Journal of Cellular Biochemistry 43 67–79.[CrossRef][ISI][Medline]
van de Wiel DF, van Rijn PA, Meloen RH & Moormann RJ 1998 High-level expression of biologically active recombinant bovine follicle stimulating hormone in a baculovirus system. Journal of Molecular Endocrinology 20 83–98.[Abstract]
Vaughn JL, Goodwin RH, Tompkins GJ & McCawley P 1977 The establishment of two cell lines from the insect Spodoptera frugiperda (Lepidoptera; Noctuidae). In Vitro 13 213–217.[ISI][Medline]
Ward DN, Moore WT & Burleigh BD 1982 Structural studies on equine chorionic gonadotropin. Journal of Protein Chemistry 1 263–280.[CrossRef]
Wu H, Lustbader JW, Liu Y, Canfield RE & Hendrickson WA 1994 Structure of human chorionic gonadotropin at 2•6 Å resolution from MAD analysis of the selenomethionyl protein. Structure 2
) and culture media of COS-7 cells producing 
) were tested in a specific sandwich ELISA. The eCG NZY-01 (1500 IU/mg) was used as a standard (
). Pituitary eLH was also tested (
). Supernatant from Mimic cells infected with wild-type baculovirus AcMNPV was used as a negative control (
). All other hormones (from Sf9 media and 
). The concentrations were estimated by ELISA and each point corresponds to the mean±S.E.M. of three experiments.
