High productivity of human recombinant beta-interferon from a low-temperature perfusion culture.

Recombinant human interferon-beta (ß-IFN), used in the therapeutic treatment of multiple sclerosis (MS), can be produced on a large-scale from genetically engineered Chinese hamster ovary (CHO) cells. However, its hydrophobicity causes non-reversible, molecular aggregation in culture. The parameters affecting aggregation were determined to be concentration, culture residence time, temperature and glycosylation. Although the protein can be produced in Escherichia coli in a non-glycosylated form, the addition of glycans confers a reduced rate of aggregation as well as a 10-fold higher bioactivity. We report on the application of a low temperature perfusion culture designed to control the parameters that cause aggregation. In this three-phase culture system there is a transition to a low temperature (32°C) in a batch mode prior to implementing perfusion at 1 volume/day using an acoustic cell separator. Perfusion at the low temperature resulted in a 3.5-fold increase in specific productivity and a 7-fold increase in volumetric productivity compared to the batch culture at 37°C. The percentage aggregation of ß-IFN was reduced from a maximum of 43% in batch culture to a minimum of 5% toward the end of the perfusion phase. The glycosylation profile of all samples showed predominantly sialylated biantennary fucosylated structures. The extent of sialylation, which is important for bioactivity, was enhanced significantly in the perfusion culture, compared to the batch culture.

Enhanced production of monomeric interferon-beta by CHO cells through the control of culture conditions.

The enhancement of recombinant protein expression of a transfected cell line is essential for the development of an efficient large-scale bioprocess. The effect of various media additives and temperature conditions were studied in an attempt to optimize protein production, stability, and protein glycosylation from a Chinese hamster ovary (CHO) cell line producing human beta-interferon (Hu-beta-IFN). We observed a decrease in the ELISA response of the glycoprotein in the later stages of batch cultures, which was attributed to molecular aggregation. Cells were subjected to various concentrations of glycerol, dimethyl sulfoxide (DMSO), and sodium butyrate (NaBu) in a variety of culture systems and conditions. The addition of both NaBu and DMSO resulted in higher specific productivities but reduced growth rates that resulted in a net reduction of interferon produced. Glycerol appeared to stabilize the secreted beta-IFN, resulting in reduced aggregation, despite a decrease in cell growth rate. Glycosylation analysis of isolated beta-IFN showed a time-dependent decrease in sialylation in batch culture that was ameliorated by the presence of glycerol. Low-temperature conditions (30 degrees C) had the greatest effect on productivity with a significant increase in beta-IFN titer as well as a reduction in the degree of molecular aggregation.

Erythropoietin production from CHO cells grown by continuous culture in a fluidized-bed bioreactor.

A Chinese hamster ovary (CHO) cell line that expresses human erythropoietin (huEPO) was in a 2-L Cytopilot fluidized-bed bioreactor with 400 mL macroporous Cytoline-1 microcarriers and a variable perfusion rate of serum-free and protein-free medium for 48 days. The cell density increased to a maximum of 23 x 10(6) cells/mL, beads on day 27. The EPO concentration increased to 600 U/mL during the early part of the culture period (on day 24) and increased further to 980 U/mL following the addition of a higher concentration of glucose and the addition of sodium butyrate. The EPO concentration was significantly higher (at least 2x than that in a controlled stirred-tank bioreactor, in a spinner flask, or in a stationary T-flask culture. The EPO accumulated to a total production of 28,000 kUnits over the whole culture period. The molecular characteristics of EPO with respect to size and pattern of glycosylation did not change with scale up. The pattern of utilization and production of 18 amino acids was similar in the Cytopilot culture to that in a stationary batch culture in a T-flask. The concentration of ammonia was maintained at a low level (< 2 mM) over the entire culture period. The specific rate of consumption of glucose, as well as the specific rates of production of lactate and ammonia, were constant throughout the culture period indicating a consistent metabolic behavior of the cells in the bioreactor. These results indicate the potential of the Cytopilot bioreactor culture system for the continuous production of a recombinant protein over several weeks.

Application of a serum-free medium for the growth of Vero cells and the production of reovirus.

Two strains of reovirus (serotype 1 Lang/TIL and serotype 3 Dearing/T3D) were propagated in Vero cells grown in stationary or agitated cultures in a serum-free medium, M-VSFM. Solid microcarriers (Cytodex-1) were used to support cell growth in agitated cultures with a normal doubling time of 25 h. Cell yields of 1 x 10(6) cells/mL were obtained from an inoculum of 2 x 10(5) cells/mL in 4 days in microcarrier cultures. The growth profile and cell yield was not significantly different from serum-supplemented cultures. The virus titer increased by 3-4 orders of magnitude over a culture period of 150 h. The maximum virus titer in stationary cultures reached >1 x 10(9) pfu/mL for both strains of reovirus in M-VSFM. M-VSFM also supported high viral yields in microcarrier cultures. Both the specific productivity and final viral yield was higher in M-VSFM than serum-supplemented cultures. The high viral productivity suggests that this is a suitable system for the production of reovirus as an oncolytic agent for human therapeutic use.

Linoleic acid improves the robustness of cells in agitated cultures.

The murine hybridoma (CC9C10) was subjected to high shear rates in a spinner flask to determine the effect of various culture additives on cell survival. At 500 rpm, the half-life of the viable cell concentration in a low protein serum-free medium was 50 min. Both bovine serum albumin and Pluronic F-68 had a significant effect in protecting cells under these conditions. The effects of the two supplements were additive, so that in the presence of both supplements there was minimal cell damage at 500 rpm. The survival rate of cells grown in media supplemented with linoleic acid improved significantly under high stirring rates. Cells grown for one passage in 50 muM linoleic acid and stirred at 500 rpm had a significantly higher survival rate than control cells. For cells grown over 5 passages in 25 muM linoleic acid, the survival rate at 470 rpm was x3 greater than that determined for control cells. This difference gradually decreased at higher stirring rates up to 610 rpm when the half-life of the viable cell population was reduced to approximately 10 min. Supplementation of cultures with linoleic acid has previously been shown to result in incorporation into all three cellular lipid fractions - polar, non-polar and free fatty acid (Butler et al., 1997). Our explanation for the increased survivability of the cells at high agitation rates in the presence of linoleic acid is that the structural lipid components of the cell including the outer membrane attained a higher unsaturated/saturated ratio which was more robust than that of control cells.

The effect of dissolved oxygen on the metabolic profile of a murine hybridoma grown in serum-free medium in continuous culture.

The murine B-lymphocyte hybridoma, CC9C10 was grown at steady state under serum-free conditions in continuous culture at dissolved oxygen (DO) concentrations in the range of 10% to 150% of air saturation. Cells could be maintained with this range at high viability in a steady state at a dilution rate of 1 d(-1), although with lower cell concentrations at higher DO. A higher specific antibody production measured at higher DO was matched by a decrease in the viable cell concentration at steady state, so that the volumetric antibody titre was not changed significantly. An attempt to grow cells at 250% of air saturation was unsuccessful but the cells recovered to normal growth once the DO was decreased.There was a requirement for cellular adaptation at each step-wise increase in dissolved oxygen. Adaptation to a DO of 100% was associated with an increase in the specific activities of glutathione peroxidase (x18), glutathione S-transferase (x11) and superoxide dismutase (x6) which are all known antioxidant enzymes. At DO above 100%, the activities of GPX and GST decreased possibly as a result of inactivation by reactive oxygen radicals.The increase in dissolved oxygen concentration caused changes in energy metabolism. The specific rate of glucose uptake increased at higher dissolved oxygen concentrations with a higher proportion of glucose metabolized anaerobically. Short-term radioactive assays showed that the relative flux of glucose through glycolysis and the pentose phosphate pathway increased whereas the flux through the tricarboxylic acid cycle decreased at high DO. Although the specific glutamine utilization rate increased at higher DO, there was no evidence for a change in the pattern of metabolism. This indicates a possible blockage of glycolytic metabolites into the TCA cycle, and is compatible with a previous suggestion that pyruvate dehydrogenase is inhibited by high oxygen concentrations.Analysis of the oxygen uptake rate of cell suspensions at steady state under all conditions showed a pronounced Crabtree effect which was manifest by a decrease (up to 40%) in oxygen consumption on addition of glucose. This indicates that the degree of aerobic metabolism in these cultures is highly sensitive to the glucose concentration.

Unsaturated fatty acids enhance cell yields and perturb the energy metabolism of an antibody-secreting hybridoma.

Growth of the murine B-lymphocyte cell line CC9C10 and the myeloma SP2/0 was enhanced significantly by the presence of the unsaturated fatty acids, oleic and linoleic acids in serum-free culture. The cellular content of linoleic and oleic acids gradually increased during continuous culture passage, with no evidence of regulatory control. Over 10 culture passages in the presence of these fatty acids, the unsaturated/saturated fatty acid ratio of all cellular lipid fractions increased substantially. Most of the fatty acid accumulated in the polar lipid fraction (more than 74%) and only a small proportion was oxidized to CO2 (0.5%). Linoleic acid caused a decrease to one-eighth in the rate of metabolism of glutamine and a 1.4-fold increase in the rate of metabolism of glucose. There was no change in the relative flux of glucose through the pathways of glycolysis, pentose phosphate or the tricarboxylic acid cycle. The changes in energy metabolism were reversed when the cells were removed from fatty acid-supplemented medium. The most plausible explanation for these effects is the observed decrease in the rate of uptake of glutamine into cells loaded with linoleic acid. Growth of the CC9C10 cells in linoleic acid caused the Km of glutamine uptake to increase from 2.7 to 23 mM, whereas glucose uptake was unaffected.

Novel DNase I hypersensitive sites in the 3'-flanking region of the human c-myc gene.

DNase I hypersensitivity regions correlate with genetic regulatory loci and binding sites for sequence-specific DNA-binding proteins. We present data supporting the presence of novel DNase 1 hypersensitive sites (which we have designated sites VI-IX) in both the body of the human c-myc gene downstream from exon 2 and the 3'-flanking region of the c-myc gene in HL-60 cells. All of these novel DH sites are markedly decreased when HL-60 cells are treated with either dimethyl sulfoxide (DMSO) or retinoic acid. Moreover, a similar pattern of DNase I hypersensitive sites in this region of c-myc was present in MCF-7 human breast cancer cells growing in culture. Our results suggest a potential role for these sites in transcriptional regulation of the human c-myc gene.

The effect of fatty acids on hybridoma cell growth and antibody productivity in serum-free cultures.

A murine B-lymphocyte hybridoma (CC9C10) was adapted for growth in a serum-free medium. Supplementation of the medium with cis-unsaturated fatty acids (10-50 microM) improved the cell yield in the order oleic/linoleic > linoleic > oleic. Initial supplementation with the fatty acids also caused a significant increase (58%) in the volumetric Mab titre. Continued growth of the cells in the fatty acid supplemented media over five culture passages resulted in a gradual deterioration of the Mab yield concomitant with the appearance of lipid inclusions in the cytosol. The higher Mab yield could be restored by a limited period of growth of the lipid-loaded cells in fatty acid-free medium. These effects were independent of growth rate. This suggests that the optimal intracellular lipid content is finely balanced between a reduced and an overloaded state. Specific glucose and glutamine utilisation rates were unaffected by the presence of fatty acids. Also, the optimal glucose and glutamine concentrations for growth were independent of the fatty acids.

C-myc gene chromatin of estrogen receptor positive and negative breast cancer cells.

Expression of the c-myc protooncogene is estrogen regulated in estrogen receptor (ER) positive, hormone-dependent human breast cancer cells, but it is constitutively active in ER negative, hormone-independent breast cancer cells. To determine whether these differences are reflected in c-myc chromatin, DNase I hypersensitive sites (DHS) were mapped. Six DHS were detected in all cell lines studied, with DHS 3(2) being more prominent than DHS 3(1). The accessibility of DHS 2 was markedly greater in ER negative cells than in ER positive cells, and this relative accessibility remained unchanged when cells were grown in estrogen free medium. DHS 2, 3(1) and 3(2) map near the P0, P1 and P2 promoters, respectively. An analysis of promoter usage demonstrated that P2 was the preferred promoter. Thus, the differences in the accessibility of DHS 2 in c-myc chromatin of ER positive and negative cells likely reflects alterations in DNA-protein interactions in this region.

ATPase-inhibitor proteins of brown-adipose-tissue mitochondria from warm- and cold-acclimated rats.

1. A group of male Sprague-Dawley rats (5-6 weeks old) was cold-acclimated at 4 degrees C for 4 weeks. Warm-acclimated controls remained at 24 degrees C. Total protein content of brown adipose tissue (BAT) increased more than 3-fold and total uncoupling protein (UCP) content increased more than 6-fold upon cold-acclimation. The concentration of UCP in isolated BAT mitochondria almost doubled. 2. Specific ATPase activity of the non-thermogenic BAT mitochondria (from warm-acclimated controls) was low and increased about 6-fold on addition of 1 microM-Ca2+, which raised free Ca2+ levels (measured by Fura-2) in the incubation media from 1.32 +/- 0.28 microM (mean +/- S.E.M.) to 2.29 +/- 0.39 microM [at which the Ca(2+)-binding ATPase-inhibitor protein (CaBI) is inactivated]. Correspondingly, the specific ATP synthetase activity of the non-thermogenic BAT mitochondria was high and was decreased by 74% by addition of 1 microM-Ca2+. 3. In contrast, specific ATPase activity of thermogenic BAT mitochondria (from cold-acclimated rats) was 5 times that of the control group, and addition of Ca2+ had only a small stimulatory response. Correspondingly, the specific ATP synthetase activity of the thermogenic BAT mitochondria was low, and the decrease by Ca2+ was small, albeit significant. 4. Extracts of BAT mitochondria from both groups of animals contained significant amounts of the ATPase-inhibitor protein of Pullman and Monroy (PMI) as well as of CaBI, as shown by gel electrophoresis. Kinetic studies of inhibition of mitochondrial ATPase activity showed that PMI activity was unaltered in extracts from the thermogenic BAT mitochondria, whereas CaBI activity was slightly but significantly increased. 5. The presence of active ATPase-inhibitor proteins in BAT mitochondria was shown for the first time. We conclude that uncoupling of oxidative phosphorylation occurs in thermogenic BAT mitochondria, even in the presence of the ATPase-inhibitor proteins.

Distribution of the ATPase inhibitor proteins of mitochondria in mammalian tissues including fibroblasts from a patient with Luft's disease.

The mitochondrial ATPase inhibitor proteins--the Pullman-Monroy inhibitor (PMI) and the Ca(2+)-binding protein (CaBI)--have a wide distribution, both being present in mitochondria of bovine heart and kidney, rat liver and brain, two mitochondrial populations of rabbit skeletal muscle, and mitochondria from human fibroblasts and the human breast cancer cell line T-47-D. The ratio of CaBI to PMI was highest in heart and skeletal muscle mitochondria. The subsarcolemmal fraction of skeletal muscle had 2.6-times as much CaBI as did the intermyofibrillar. The ratio of CaBI to PMI in the mitochondria of the other normal tissues and fibroblasts was close to 1. In contrast, mitochondria from T-47D cells had 1.5-times as much PMI as CaBI whilst mitochondria from fibroblasts from a patient with Luft's disease showed a virtual lack of PMI. The specific ATPase, ATP-synthetase and succinate dehydrogenase activities of the Luft's mitochondria were, however, in the normal range. The specific ATP synthetase activity of the T-47D cells was significantly higher than normal. We conclude that tissues like heart and skeletal muscle which experience wide fluctuations in intracellular Ca2+ have a greater need for CaBI. Why lack of PMI could lead to 'loose' coupling of oxidative phosphorylation in skeletal muscle of Luft's patients, but not in fibroblasts is discussed.

Antigenic reactivity of ribosomal protein S6 and the calcium-binding ATPase inhibitor protein of mammalian mitochondria.

Phosphorylation of ribosomal protein S6 of mammals precedes activation of cell growth in numerous biological systems. We have cloned a cDNA for ribosomal protein S6 from T-47D human breast cancer cells by immunoscreening a lambda gt11 expression library with antibody raised against the mitochondrial Ca(2+)-binding ATPase inhibitor protein (CaBI) of bovine heart mitochondria (Yamada & Huzel: J Biol Chem 263: 11498-11503, 1988). Similar clones were obtained by the immunoscreening of a rat heart expression library. In agreement with others, the open reading frames of the cDNAs from the two species coded for the same amino acid sequence. No difference in S6 of the human neoplastic cells compared to that of non-neoplastic cells was found. However, common antigenic determinants in S6 and CaBI were indicated. Accordingly, S6 was purified from rat liver ribosomes and antiserum prepared. Immuno-dot blot and Western blot analyses showed high specific reactivity between S6, the cloned chimeric beta-galactosidase fusion protein from a cDNA clone, and CaBI with anti-S6 and anti-CaBI antibodies. The antibodies also showed a high degree of discrimination for S6 and CaBI. Neither interacted with the other ribosomal proteins nor with another ATPase inhibitor protein from bovine heart mitochondria. Neither interacted with the Ca(2+)-binding proteins, calmodulin, oncomodulin, Protein C, or Factor X. Prothrombin was weakly reactive with anti-CaBI but not with anti-S6. Thus, the results fulfill the specific criteria for the concept and operational definition of common protein epitopes in S6 and CaBI. However, neither prothrombin nor S6 fusion protein inhibited mitochondrial ATPase activity even at 20 times the concentrations at which CaBI gave 97% inhibition.

Isolation of histone-like proteins from mitochondria of bovine heart.

Two methods for isolating and purifying histone-like proteins from mitochondria of bovine heart are described. In the first, a sonicated extract of the mitochondria was fractionated in three chromatography steps, including affinity chromatography on DNA-cellulose, to purify a protein that resembles very closely the histone-like protein (HM) of yeast mitochondria. In the second method, an acid extract of the heart mitochondria was the starting material; two other histone-like proteins were separated. Thus, as in mitochondria of Xenopus laevis, several histone-like proteins are present in mitochondria of bovine heart.

Calcium-binding ATPase inhibitor protein of bovine heart mitochondria. Role in ATP synthesis and effect of Ca2+.

Submitochondrial particles (A particles) and phosphorylating electron-transport particles (ETPH) were prepared from bovine heart mitochondria. The A particles either were supplemented with or were depleted of the mitochondrial calcium-binding ATPase inhibitor protein (CaBI). The CaBI-depleted A particles still retained the Pullman-Monroy ATPase inhibitor protein (PMI), and the other particles all contained both CaBI and PMI. ATP synthase and ATPase activities of the particles were measured in similar reaction mixtures by luminescence of firefly luciferin-luciferase. Succinate was the respiratory substrate, and the adenylate kinase inhibitor P1, P5-di(adenosine-5') pentaphosphate was obligatory. The ATP synthase activity of CaBI-depleted A particles was 30-40% of that of the A and ETPH particles, and its ATPase activity was 7-8 times greater. Reconstitution of the CaBI-depleted A particles with CaBI restored the original ATP synthase and ATPase activities. ATP synthase activity rose about 1.7-fold when A particles were supplemented with additional CaBI and ATPase activity dropped to 9% of the original. Varying Ca2+ levels had little or no effect on the ATP synthase and ATPase activities of the CaBI-depleted A particles. In contrast, ATP synthase activity of the other particles was decreased by as much as 70% at the optimal Ca2+ concentration of 1 microM, and the ATPase activity of the A and EPTH particles rose concomitantly by 7-8-fold. The ATP synthase and ATPase activities of all the particles in microM Ca2+ became like those of the CaBI-depleted A particles. These changes were reversible; normal activities were restored as Ca2+ concentrations were raised above 1 microM.(ABSTRACT TRUNCATED AT 250 WORDS)

The calcium-binding ATPase inhibitor protein from bovine heart mitochondria. Purification and properties.

Two ATPase inhibitor proteins were isolated together from bovine heart mitochondria by a new procedure; each was purified further. The one inhibitor is a Ca2+-binding protein. It was found to contain 2 cysteine residues/mol as well as threonine and proline residues, all of which the other inhibitor (first isolated by Pullman and Monroy (Pullman, M.E., and Monroy, G. C. (1963) J. Biol. Chem. 238, 3762-3769] lacks. Its minimal molecular weight was 6390 with 62 amino acid residues/mol, and its isoelectric point was 4.6. Besides differences in size, composition, and response to Ca2+, the two inhibitor proteins also differed in response to sulfhydryl compounds, pH, KCl, and cardiolipin. Inhibition by the two inhibitor proteins was additive. Both cross-reacted with mitochondrial ATPase from rat skeletal muscle. Calmodulin, with or without Ca2+, had no effect on the activity of either inhibitor protein. Antibody to the Ca2+-binding inhibitor protein did not interact with the Pullman-Monroy inhibitor or have any effect on its activity. The antibody interacted with intact submitochondrial particles that contained both inhibitor proteins but not with particles from which only the Ca2+-binding inhibitor had been removed. Clearly, the two inhibitors are distinct immunologically as well as in other properties. The two types of inhibitor protein were also isolated from rat skeletal muscle mitochondria by the new procedure.

Ca2+-binding properties of a unique ATPase inhibitor protein isolated from mitochondria of bovine heart and rat skeletal muscle.

Previous studies showed that Ca2+ induced monomer to active dimer interconversion of a mitochondrial ATPase inhibitor protein from bovine heart or rat skeletal muscle (Yamada, E.W., Huzel, N.J. and Dickison, J.C. (1981) J. Biol. Chem. 256, 10203-10207). Initial equilibrium dialysis measurements of Ca2+ binding showed that this unique protein possesses three binding sites of high affinity with a maximum of one mol of Ca2+ bound/mol of protein monomer. Magnesium (1 mM) did not affect the first association constant but increased the second and third by about 1.2 and 1.5 fold, respectively. That the apparent association constants varied with concentration of protein monomer was in agreement with the self-associating nature of the protein. Scatchard plots at three concentrations of protein intersected at a molar ratio of about 0.5 (Ca2+/monomer). Ka1 and Ka2 values of 4.2 microM and 12.1 microM, respectively, were estimated by extra-polation of apparent constants to infinite dilution of protein. Ka3 (51.3 microM) was estimated by extrapolation of double reciprocal plots of apparent constants versus protein concentration to infinite levels of protein. A model for Ca2+ binding by this self-associating protein is described. Trifluoperazine had no effect on the activity of the inhibitor protein from either tissue.

Isolation of two ATPase inhibitor proteins from mitochondria of rat skeletal muscle.

An ATPase inhibitor protein was isolated from mitochondria of rat skeletal muscle by alkaline extraction and then was purified. It differed in definitive ways from the ATPase inhibitor protein isolated previously by Ca2+-stripping of submitochondrial particles of rat skeletal muscle. The two ATPase inhibitor proteins were shown to be present together in intact mitochondria.

Stripping of proteins from submitochondrial particles of rat skeletal muscle or bovine heart by chemical uncouplers.

Proteins of similar molecular weights were stripped from submitochondrial particles (A particles) of rat skeletal muscle or bovine heart by treatment with classical chemical uncouplers at 0 degrees C as with Ca2+. Proteins released included two of high molecular weight (about 43 000 and 30 000), an ATPase inhibitor protein (IF1) as well as the Ca2+-binding lipoprotein that has previously been shown to protect the mitochondrial ATPase complex against inhibition by N,N'-dicyclohexylcarbodiimide (DCCD). The latter two proteins were purified to a high degree. The crude fraction obtained by stripping with chemical uncouplers also contained traces of an additional protein (relative mass (Mr) approximately 13 000) which was also found upon aging of the crude fraction stripped by Ca2+. It was not found in aged preparations of either purified IF1 or the lipoprotein, but appeared when IF1 and the lipoprotein were mixed and aged together. Pretreatment of the mixture with 2-mercaptoethanol prior to electrophoresis did not remove the hybrid. More phospholipid was stripped from A particles by chemical uncouplers than by Ca2+ but less protein was stripped. Phosphatidylcholine, phosphatidylethanolamine, lysophosphatidylcholine, and cardiolipin were identified in the phospholipid fractions.