Regulating apoptosis in mammalian cell cultures.

Cell culture technology has become a widely accepted method used to derive therapeutic and diagnostic protein products. Mammalian cells adapted to grow in bioreactors now play an integral role in the development of these biologicals. A major limiting factor determining the output efficiency of mammalian cell cultures however, is apoptosis or programmed cell death. Methods to delay apoptosis and increase the longevity of cell cultures can lead to more economical processes. Researchers have shown that both genetic and chemical strategies to block apoptotic signals can increase cell culture productivity. Here, we discuss various strategies which have been implemented to improve cellular viabilities and productivities in batch cultures.

Cloning and expression of human sialic acid pathway genes to generate CMP-sialic acids in insect cells.

The addition of sialic acid residues to glycoproteins can affect important protein properties including biological activity and in vivo circulatory half-life. For sialylation to occur, the donor sugar nucleotide cytidine monophospho-sialic acid (CMP-SA) must be generated and enzymatically transferred to an acceptor oligosaccharide. However, examination of insect cells grown in serum-free medium revealed negligible native levels of the most common sialic acid nucleotide, CMP-N-acetylneuraminic acid (CMP-Neu5Ac). To increase substrate levels, the enzymes of the metabolic pathway for CMP-SA synthesis have been engineered into insect cells using the baculovirus expression system. In this study, a human CMP-sialic acid synthase cDNA was identified and found to encode a protein with 94% identity to the murine homologue. The human CMP-sialic acid synthase (Cmp-Sas) is ubiquitously expressed in human cells from multiple tissues. When expressed in insect cells using the baculovirus vector, the encoded protein is functional and localizes to the nucleus as in mammalian cells. In addition, co-expression of Cmp-Sas with the recently cloned sialic acid phosphate synthase with N-acetylmannosamine feeding yields intracellular CMP-Neu5Ac levels 30 times higher than those observed in unsupplemented CHO cells. The absence of any one of these three components abolishes CMP-Neu5Ac production in vivo. However, when N-acetylmannosamine feeding is omitted, the sugar nucleotide form of deaminated Neu5Ac, CMP-2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (CMP-KDN), is produced instead, indicating that alternative sialic acid glycoforms may eventually be possible in insect cells. The human CMP-SAS enzyme is also capable of CMP-N-glycolylneuraminic acid (CMP-Neu5Gc) synthesis when provided with the proper substrate. Engineering the CMP-SA metabolic pathway may be beneficial in various cell lines in which CMP-Neu5Ac production limits sialylation of glycoproteins or other glycans.

Determination of nucleotides and sugar nucleotides involved in protein glycosylation by high-performance anion-exchange chromatography: sugar nucleotide contents in cultured insect cells and mammalian cells.

We have developed a simple and highly sensitive HPLC method for determination of cellular levels of sugar nucleotides and related nucleotides in cultured cells. Separation of 9 sugar nucleotides (CMP-Neu5Ac, CMP-Neu5Gc, CMP-KDN, UDP-Gal, UDP-Glc, UDP-GalNAc, UDP-GlcNAc, GDP-Fuc, GDP-Man) and 12 nucleotides (AMP, ADP, ATP, CMP, CDP, CTP, GMP, GDP, GTP, UMP, UDP, and UTP) was examined by reversed-phase HPLC and high-performance anion-exchange chromatography (HPAEC). Although the reversed-phase HPLC, using an ion-pairing reagent, gave a good separation of the 12 nucleotides, it did not separate sufficiently the sugar nucleotides for quantification. On the other hand, the HPAEC method gave an excellent and reproducible separation of all nucleotides and sugar nucleotides with high sensitivity and reproducibility. We applied the HPAEC method to determine the intracellular sugar nucleotide levels of cultured Spodoptera frugiperda (Sf9) and Trichoplusia ni (High Five, BTN-TN-5B1-4) insect cells, and compared them with those in Chinese hamster ovary (CHO-K1) cells. Sf9 and High Five cells showed concentrations of UDP-GlcNAc, UDP-Gal, UDP-Glc, GDP-Fuc, and GDP-Man equal to or higher than those in CHO cells. CMP-Neu5Ac was detected in CHO cells, but it was not detected in Sf9 and High Five cells. In conclusion, the newly developed HPAEC method could provide valuable information necessary for generating sialylated complex-type N-glycans in insect or other cells, either native or genetically manipulated.

Comparison of Bcl-2 to a Bcl-2 deletion mutant for mammalian cells exposed to culture insults.

Apoptosis has been found to occur in bioreactors as a result of environmental stresses. The overexpression of bcl-2 is a widely used strategy to limit the induction of apoptosis in mammalian cell cultures. In this study, the effectiveness of wild-type Bcl-2 was compared to a Bcl-2 mutant lacking the nonstructured loop domain in two commercially prominent cell lines, Chinese hamster ovary (CHO) and baby hamster kidney (BHK) cells. The generation of a DNA "ladder" and condensation of chromatin indicated that apoptosis occurred in these cell lines following Sindbis virus infection and serum deprivation. When cells were engineered to overexpress the bcl-2 mutant, cell death due to Sindbis virus was inhibited in a concentration-dependent manner. Furthermore, the Bcl-2 mutant provided increased protection as compared to wild-type Bcl-2 following two model insults, Sindbis virus infection and serum deprivation. Total production for a heterologous protein encoded on the Sindbis virus was increased in cell lines expressing the Bcl-2 variants compared to the parental cell line. In order to understand the reasons for the improved anti-apoptosis properties of the mutant, wild-type Bcl-2 and mutant Bcl-2 were examined by Western blot following each model insult. Wild-type Bcl-2 was observed to degrade into a 23 kDa fragment following both Sindbis virus infection and serum withdrawal in both cell lines, while the mutant Bcl-2 protein was not degraded during the same period. The processing of Bcl-2 was found to correlate with reduced cell viabilities following the two external insults to suggest that Bcl-2 degradation may limit its ability to inhibit apoptosis. These studies indicate that the cells regulate anti-apoptosis protein levels and these processing events can limit the effectiveness of cell death inhibition strategies in mammalian cell culture systems.

Investigating the secretory pathway of the baculovirus-insect cell system using a secretory green fluorescent protein.

The secretory pathway is important in actively transporting proteins into the extracellular environment of eucaryotic cells. In this study a green fluorescent protein (GFP) mutant engineered to contain a secretion signal was used as a model protein in order to visualize the secretion process inside insect cells. Fluorescent microscopy indicated that significant amounts of secreted green fluorescent protein (sGFP) accumulated in High-Five, Trichoplusia ni, cells following infection with a baculovirus vector containing the gene under the polyhedrin promoter. Laser scanning confocal microscopy was used to reconstruct whole cell images of the infected High-Five cells at multiple days postinfection. While the protein was widely distributed at 2 days postinfection, certain intracellular regions appeared to contain higher or lower concentrations of the sGFP. A layer by layer examination indicated pockets in which sGFP was absent, and these appear to be vesicles that have recently released the sGFP or are not yet accumulating sGFP. By 3 days postinfection, the sGFP in some cells was concentrated in a number of widely dispersed globules, which may represent the vesicle remnants of a deteriorating secretory pathway. In contrast, nonsecreted GFP was more uniformly distributed in the cells than sGFP and did not accumulate in vesicles. In addition to GFP, the lectins wheat germ agglutinin (WGA) and concanavalin A (ConA), which have affinities for sugar residues, were used to examine the secretory pathway. The WGA, which is a Golgi marker, was distributed around the nucleus prior to infection but then was found to be polarized in one region of the cell following the baculovirus infection. The expansion of other cellular compartments following the baculovirus infection may have caused a change in intracellular distribution of the Golgi. While some of the sGFP was found to colocalize with the WGA label, much of the sGFP was outside this Golgi region. In contrast, ConA labeling, which was not as specific as WGA, was found throughout the cell both before and after infection similar to the sGFP distribution. These studies demonstrate that confocal visualization of fluorescent proteins can be used as an in vivo tool for examining secretory processing in insect cells.

N-glycan patterns of human transferrin produced in Trichoplusia ni insect cells: effects of mammalian galactosyltransferase.

The N-glycans of human serum transferrin produced in Trichopulsia ni cells were analyzed to examine N-linked oligosaccharide processing in insect cells. Metabolic radiolabeling of the intra- and extracellular protein fractions revealed the presence of multiple transferrin glycoforms with molecular weights lower than that observed for native human transferrin. Consequently, the N-glycan structures of transferrin in the culture medium were determined using three-dimensional high performance liquid chromatography. The attached oligosaccharides included high mannose, paucimannosidic, and hybrid structures with over 50% of these structures containing one fucose, alpha(1,6)-, or two fucoses, alpha(1,6)- and alpha(1,3)-, linked to the Asn-linked N-acetylglucosamine. Neither sialic acid nor galactose was detected on any of the N-glycans. However, when transferrin was coexpressed with beta(1,4)-galactosyltransferase three additional galactose-containing hybrid oligosaccharides were obtained. The galactose attachments were exclusive to the alpha(1, 3)-mannose branch and the structures varied by the presence of zero, one, or two attached fucose residues. Furthermore, the presence of the galactosyltransferase appeared to reduce the number of paucimannosidic structures, which suggests that galactose attachment inhibits the ability of hexosaminidase activity to remove the terminal N-acetylglucosamine. The ability to promote galactosylation and reduce paucimannosidic N-glycans suggests that the oligosaccharide processing pathway in insect cells may be manipulated to mimic more closely that of mammalian cells.

Cloning and expression of the human N-acetylneuraminic acid phosphate synthase gene with 2-keto-3-deoxy-D-glycero- D-galacto-nononic acid biosynthetic ability.

Sialic acids participate in many important biological recognition events, yet eukaryotic sialic acid biosynthetic genes are not well characterized. In this study, we have identified a novel human gene based on homology to the Escherichia coli sialic acid synthase gene (neuB). The human gene is ubiquitously expressed and encodes a 40-kDa enzyme. The gene partially restores sialic acid synthase activity in a neuB-negative mutant of E. coli and results in N-acetylneuraminic acid (Neu5Ac) and 2-keto-3-deoxy-D-glycero-D-galacto-nononic acid (KDN) production in insect cells upon recombinant baculovirus infection. In vitro the human enzyme uses N-acetylmannosamine 6-phosphate and mannose 6-phosphate as substrates to generate phosphorylated forms of Neu5Ac and KDN, respectively, but exhibits much higher activity toward the Neu5Ac phosphate product.

Part I. Bcl-2 and Bcl-x(L) limit apoptosis upon infection with alphavirus vectors.

Viral expression systems offer the ability to generate high levels of a particular protein within a relatively short period of time. In particular, alphavirus constructs based on Sindbis virus (SV) and Semliki Forest virus (SFV) are promising vehicles as they are cytoplasmic vectors with the potential for high expression levels. Two such alphavirus vectors were utilized during the current study to infect two commercially relevant cell lines, baby hamster kidney (BHK) and Chinese hamster ovary (CHO); the first was a fully competent SV derivative carrying the gene for chloramphenicol acetyltransferase (dsSV-CAT), while the second was a replication deficient SFV construct containing the human interleukin-12 (IL-12) p35 and p40 genes (SFV-IL-12). Since infection with these vectors induced apoptosis in both cell lines, the present effort was dedicated to determining the ability of anti-apoptosis genes to limit the cell death associated with these virus constructs. Infection with the dsSV-CAT vector resulted in the rapid death of BHK and CHO cells within 4 days, a phenomenon which was considerably delayed by stably overexpressing bcl-2 or bcl-x(L). In fact, cellular lifespans were doubled in both BHK-bcl2 and CHO-bclx(L) cells relative to the parental cell lines. Furthermore, the presence of these gene products provided increases of up to 2-fold in recombinant CAT production. Overexpression of bcl-2 and bcl-x(L) also altered the response of these cells upon infection with SFV-IL-12. While the parental cell lines were completely nonviable within 1 week, the BHK-bcl2, BHK-bclx(L), and CHO-bclx(L) cells each recovered from the infection, resuming exponential growth and regaining viabilities of over 90% by 9 days post-infection. Total IL-12 productivities were nearly doubled by Bcl-2 and Bcl-x(L) in the CHO cells, although this effect was apparently cell-line specific, as the native BHK cells were able to secrete more IL-12 than either of its transfected derivatives. Regardless, the presence of the anti-apoptosis genes allowed the production of IL-12 to be maintained, albeit at low levels, from each of the cell lines for the duration of the culture process. Therefore, overexpression of bcl-2 family members can have a significant impact on culture viabilities and recombinant protein production during alphavirus infections of mammalian cells.

Part II. Overexpression of bcl-2 family members enhances survival of mammalian cells in response to various culture insults.

A number of bioreactor configurations have been developed for the manufacture of products from mammalian cell hosts. Even in the most efficient of these, however, problems such as nutrient exhaustion, growth factor deprivation, and toxin accumulations may arise. Consequently, the current effort focused on the feasibility of overexpressing anti-apoptosis genes in baby hamster kidney (BHK) and Chinese hamster ovary (CHO) cells as a means of limiting cell death upon exposure to three such insults. Extended periods of glucose deprivation, serum withdrawal, and treatment with ammonium chloride each caused significant damage, often apoptotic in nature, to BHK and CHO cells, typically rendering cultures completely nonviable. The overexpression of bcl-2 and bcl-x(L), however, was able to abrogate the cell death in BHK cultures, though to varying degrees. For instance, the presence of Bcl-2, which did little to suppress apoptosis upon glucose deprivation, significantly improved the viabilities of these cells during serum withdrawal. In contrast, bcl-x(L) overexpression provided BHK cells with enhanced protection in the absence of glucose, allowing cultures to remain viable throughout the entire three week study. CHO cultures, on the other hand, displayed similar trends in survival in response to both glucose and serum deprivation. During these studies, Bcl-x(L) was consistently able to afford cells the highest degree of protection, though Bcl-2 also enhanced culture viabilities and viable numbers. Death suppression following exposure to 50 mM ammonium chloride was observed to a limited extent in both BHK and CHO cells overexpressing bcl-2 and bcl-x(L). However, even during such harsh treatment, Bcl-x(L) was able to enhance the survival of both cultures, providing CHO cells with viable numbers that were nearly 20-fold that of the controls after five days of exposure. Furthermore, the extensions in cell survival provided by the anti-apoptosis gene products enabled the recovery of many of the cultures during rescue attempts in which the death-inducing stimulus was removed. Clearly, engineering cells to better withstand and recover from the insults common during the large scale cultivation of mammalian cells has a number of potential applications in the biopharmaceutical industries where cell death can limit culture productivities.

Antiapoptosis chemicals prolong productive lifetimes of mammalian cells upon Sindbis virus vector infection.

Viral expression systems allow for the rapid production of large amounts of recombinant protein in cell culture. In particular, Sindbis virus vectors now exist that make possible the expression of a variety of heterologous proteins in mammalian culture systems. Unfortunately, infection of cultured cells with Sindbis virus vectors typically results in apoptotic cell death, as demonstrated in the current study by DNA laddering and fluorescence microscopy. Fortunately, it has recently been demonstrated that apoptosis can be inhibited in vitro by certain chemical reagents that are capable of blocking specific steps during the cell death cascade. In this study, a rat prostate carcinomal cell line, AT3-neo, was infected with a Sindbis virus vector containing the gene for chloramphenicol acetyltransferase (dsSV-CAT) in the presence of several representative antiapoptotic chemicals and analyzed for cell viability as well as recombinant protein production. N-acetylcysteine (NAC), pyrrolidine dithiocarbamate (PDTC), bongkrekic acid (BA), and N-benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (Z-VAD.fmk) all exhibited the capacity to limit apoptosis in the infected cells. In fact, after just 1 day, percentage viabilities of the cells exposed to chemical reagents were between 72% and 91%, compared with 44% for the untreated controls. Furthermore, cells maintained on these agents were able to survive the infection from 1 to 3 days longer than the control samples. In addition to providing gains in cell viability, chemical treatment allowed for higher levels of recombinant protein production in most cases. Maximum chloramphenicol acetyltransferase (CAT) productivities in cells maintained on BA, NAC, and Z-VAD.fmk were 1.7-, 2.2-, and 3.9-fold higher than those obtained from the untreated cultures. Consequently, the addition of chemical reagents to culture media as a means of inhibiting apoptosis may be a valuable tool in the cell culture industry, where cell death severely limits productivity levels and adds significantly to production costs.

Molecular chaperones stimulate the functional expression of the cocaine-sensitive serotonin transporter.

The serotonin transporter (SERT) is an N-glycosylated integral membrane protein that is predicted to contain 12 transmembrane regions. SERT is the major binding site in the brain for antidepressant drugs, and it also binds amphetamines and cocaine. The ability of various molecular chaperones to interact with a tagged version of SERT (Myc-SERT) was investigated using the baculovirus expression system. Overexpression of Myc-SERT using the baculovirus system led to substantial quantities of inactive transporter, together with small amounts of fully active and, therefore, correctly folded molecules. The high levels of inactive Myc-SERT probably arose because folding was rate-limiting due, perhaps, to insufficient molecular chaperones. Therefore, Myc-SERT was co-expressed with the endoplasmic reticulum (ER) molecular chaperones calnexin, calreticulin and immunoglobulin heavy chain binding protein (BiP), and the foldase, ERp57. The expression of functional Myc-SERT, as determined by an inhibitor binding assay, was enhanced nearly 3-fold by co-expressing calnexin, and to a lesser degree on co-expression of calreticulin and BiP. Co-expression of ERp57 did not increase the functional expression of Myc-SERT. A physical interaction between Myc-SERT-calnexin and Myc-SERT-calreticulin was demonstrated by co-immunoprecipitation. These associations were inhibited in vivo by deoxynojirimycin, an inhibitor of N-glycan precusor trimming that is known to prevent the calnexin/calreticulin-N-glycan interaction. Functional expression of the unglycosylated SERT mutant, SERT-QQ, was also increased on co-expression of calnexin, suggesting that the interaction between calnexin and SERT is not entirely dictated by the N-glycan. SERT is the first member of the neurotransmitter transporter family whose folding has been shown to be assisted by the molecular chaperones calnexin, calreticulin, and BiP.

Modifying secretion and post-translational processing in insect cells.

The baculovirus-insect cell system is a valuable tool for the expression of heterologous proteins. Due to limitations in the intracellular processing environment, however, heterologous secreted and membrane proteins are often insoluble, poorly processed, or contain 'non-human' modifications. Recent attempts to modify the insect cell secretory pathway by overexpressing processing factors have demonstrated the potential to overcome these limitations.

A bacterial signal peptidase enhances processing of a recombinant single chain antibody fragment in insect cells.

The production of an antibody single chain fragment (scFv) in insect cells was accompanied by the formation of an insoluble intracellular precursor even with the inclusion of the bee melittin signal peptide. The presence of the precursor polypeptide suggests a limitation in the processing of the signal peptide so a baculovirus containing a signal peptidase from Bacillus subtilis (SipS) was constructed for expression studies. When the wild type SipS was coexpressed with scFv, preprocessed scFv fragments were no longer detected in insect cell lysates. Conversely, coexpression of scFv alone or with an inactive mutant SipS resulted in at least 30% of the intracellular polypeptide in an unprocessed form at 3 days post infection. Production of scFv in the medium was also enhanced in the presence of SipS; however, low secretion levels indicate the presence of a post-processing bottleneck.

Overcoming apoptosis: new methods for improving protein-expression systems.

Cellular suicide, or apoptosis, is responsible for a significant proportion of cell death in many bioprocesses. With the progressive elucidation of the biochemical and genetic events leading to this form of cell death, it is now possible to implement strategies for extending the productive lifetimes of cells in culture. These strategies may include nutritional, genetic and chemical methods that enhance cell survival and performance during the critical stages of a culture process, leading to improvements in the production capacity for valuable biotechnological products.

Overexpression of a cytosolic chaperone to improve solubility and secretion of a recombinant IgG protein in insect cells.

The secretion of heterologous IgG proteins in the baculovirus-insect cell expression system is accompanied by substantial insoluble immunoglobulin in the infected cells. The accumulation of these insoluble forms suggests a limitation in the processing and secretory pathway of the infected cells. As a result, cytosolic hsp70 chaperones, which are known to associate and prevent aggregation of polypeptides in vitro, have been coexpressed in the infected cells. The hsp70 protein coprecipitated with the immunoglobulin to indicate the formation of a specific hsp70-immunoglobulin complex in vivo. Immunoblot and pulse chase studies indicated that coexpression of hsp70 increased intracellular immunoglobulin solubility. Metabolic labeling experiments revealed that hsp70 increased secreted immunoglobulin levels after several days infection as compared to infection with control baculoviruses. Pulse chase studies indicated that hsp70 increases the solubility of immunoglobulin precursors that are then processed and assembled into the complete antibody oligomer. A comparison of the action of cytosolic hsp70 chaperone to the endoplasmic reticulum chaperone BiP suggests sequential action in which hsp70 increases the solubility of preprocessed immunoglobulin, while BiP enhances the solubility of processed immunoglobulin chains.

I. Study of protein aggregation due to heat denaturation: A structural approach using circular dichroism spectroscopy, nuclear magnetic resonance, and static light scattering.

The objective of this study was to investigate the relationship between oxidized RNase A protein structure and the occurrence of protein aggregation using several spectroscopic techniques. Circular dichroism spectroscopy (CD) measurements taken at small temperature intervals were used to determine the protein's melting temperature, Tm, of approximately 65 degrees C in deionized water. A more detailed examination of the protein structure was undertaken at several temperatures around Tm using near- and far-UV CD and one-dimensional nuclear magnetic resonance (NMR) measurements. These measurements revealed the presence of folded structures at 55 degrees C and below, while denatured structures appeared at 65 degrees C and above. Concurrent static light scattering (SLS) measurements, employed to detect the presence of RNase A aggregates, showed that RNase A aggregation was observed at 65 degrees C and above, when much of the protein was denatured. Subsequent NMR time-course data demonstrated that aggregates forming at 75 degrees C and pH 7.8 were indeed derived from heat-denatured protein. However, aggregation was also detected at 55 degrees C when the spectroscopic data suggested the protein was present predominantly in the folded configuration. In contrast, heat denaturation did not lead to RNase A aggregation in a very acidic environment. We attribute this phenomenon to the effect of charge-charge repulsion between the highly protonated RNase A molecules in very acidic pH.

II. Electrostatic effect in the aggregation of heat-denatured RNase A and implications for protein additive design.

In the previous study (part I), heat-denatured RNase A aggregation was shown to depend on the solution pH. Interestingly, at pH 3.0, the protein did not aggregate even when exposed to 75 degrees C for 24 h. In this study, electrostatic repulsion was shown to be responsible for the absence of aggregates at that pH. While RNase A aggregation was prevented at the extremely acidic pH, this is not an environment conducive to maintaining protein function in general. Therefore, attempts were made to confer electrostatic repulsion near neutral pH. In this study, heat-denatured RNase A was mixed with charged polymers at pH 7.8 in an attempt to provide the protein with excess surface cations or anions. At 75 degrees C, SDS and dextran sulfate were successful in preventing RNase A aggregation, whereas their cationic, nonionic, and zwitterionic analogs did not do so. We believe that the SO3- groups present in both additives transformed the protein into polyanionic species, and this may have provided a sufficient level of electrostatic repulsion at pH 7.8 and 75 degrees C to prevent aggregation from proceeding.

Thioredoxin domain non-equivalence and anti-chaperone activity of protein disulfide isomerase mutants in vivo.

Coexpression of the enzyme, protein disulfide isomerase (PDI), has been shown to increase soluble and secreted IgG levels from baculovirus-infected insect cells (Hsu, T.-A., Watson, S., Eiden, J. J., and Betenbaugh, M. J. (1996) Protein Expression Purif. 7, 281-288). PDI is known to include catalytic active sites in two separate thioredoxin-like domains, one near the amino terminus and another near the carboxyl terminus. To examine the role of these catalytic active sites in enhancing immunoglobulin solubility, baculovirus constructs were utilized with cysteine to serine mutations at the first cysteine of one or both of the CGHC active site sequences. Trichoplusia ni insect cells were coinfected with a baculovirus vector coding for IgG in concert with either the wild-type human PDI virus, amino-terminal mutant (PDI-N), carboxyl-terminal mutant (PDI-C), or mutant with both active sites altered (PDI-NC). Western blot analysis revealed that both immunoglobulins and PDI protein were expressed in the coinfected cells. To evaluate the effect of the PDI variants on immunoglobulin solubility and secretion, the infected cells were labeled with 35S-amino-acids for different periods, and the soluble immunoglobulins were immunoprecipitated from clarified cell lysates and culture medium using anti-IgG antibodies. Only coinfections with the wild-type PDI and PDI-N mutant led to increased immunoglobulin solubility and higher IgG secretion. In contrast, infection with the PDI-C and PDI-NC variants actually lowered immunoglobulin solubility levels below those achieved with a negative control virus. Immunoprecipitation with anti-PDI antibody revealed that heterologous PDI-C and PDI-NC were insoluble, even though PDI-N and wild-type PDI protein were detected in soluble form. The capacity for PDI-N to increase immunoglobulin solubility whereas the PDI-C mutant lowered solubility indicates that the amino- and carboxyl-terminal thioredoxin domains of PDI are functionally distinct in vivo following mutations to the active site. Furthermore, mutations at the active site of the carboxyl-terminal thioredoxin domain result in PDI variants that can act as anti-chaperones of immunoglobulin solubility in vivo as has been observed in vitro for lysozyme aggregation by wild-type PDI and PDI mutants (Puig, A., and Gilbert, H. F. (1994) J. Biol. Chem. 269, 7764-7771).

Differential N-glycan patterns of secreted and intracellular IgG produced in Trichoplusia ni cells.

Structures of the N-linked oligosaccharide attached to the heavy chain of a heterologous murine IgG2a produced from Trichoplusia ni (TN-5B1-4, High Five) insect cells were characterized. Coexpression of the chaperone immunoglobulin heavy chain-binding protein (BiP) in the baculovirus-infected insect cells increased the soluble intracellular and secreted IgG level. This facilitated the detailed analysis of N-glycans from both intracellular and secreted IgG. Following purification of the immunoglobulins using Protein A-Sepharose, glycopeptides, prepared by trypsin-chymotrypsin digestion, were further digested with glycoamidase from sweet almond emulsin to obtain the oligosaccharide moieties. The resulting oligosaccharides were then reductively aminated with 2-aminopyridine and the structures identified by two-dimensional high performance liquid chromatography mapping (Tomiya, N., Awaya, J., Kurono, M., Endo, S., Arata, Y., and Takahashi, N. (1988) Anal. Biochem. 171, 73-90). The N-glycans obtained from the secreted IgG contain 35% complex type, some with terminal galactose residues at either alpha1, 3-Man or alpha1,6-Man branches of the Man3GlcNAc2 core. The remaining oligosaccharides detected in the secreted IgG were principally hybrid (30%) and paucimannosidic (35%) type N-glycans. Most (84%) of these secreted glycoforms contained fucose alpha1, 6-linked to the innermost GlcNAc residue and the presence of a potentially allergenic fucose alpha1,3-linked to the innermost GlcNAc residue was also detected. In contrast, the intracellular immunoglobulins included 50% high mannose-type N-glycans with lower levels of complex, hybrid, and paucimannosidic-type structures. Reverse phase one-dimensional high performance liquid chromatography analysis of the IgG N-glycans in the absence of heterologous BiP exhibited a similar distribution of intracellular and secreted glycoforms. These studies indicate that Trichoplusia ni TN-5B1-4 cells are capable of terminal galactosylation. However, the processing pathways in these cell lines appear to diverge from mammalian cells in the formation of paucimannosidic structures, in the presence of alpha1,3-fucose linkages, and in the absence of sialylation.