Macroporous microcarriers such as Cytopore entrap mammalian cells in a mesh network allowing growth to high cell concentrations in a protected environment within a stirred culture. Chinese hamster ovary (CHO) cells producing recombinant human β-interferon (IFN-β) and grown in Cytopore microcarriers showed a 2.6- to 2.8-fold increase in the volumetric product titer compared with cells grown in an equivalent suspension culture. In an attempt to maximize production of IFN-β, microcarrier cultures were subjected to a low temperature regime. Low temperature culture conditions (32°C) have been shown previously to enhance cell specific productivity in suspension cultures although at reduced cell growth rates. These conditions can be optimized by a timely shift from physiological to hypothermic conditions during the culture run to maximize volumetric protein production. In the case of IFN-β production the lower temperature has the added advantage of stabilizing the product and reducing intramolecular aggregation. Using a biphasic temperature-shift regime from 37 to 32°C the volumetric production of IFN-β was enhanced to 4.2-fold compared with a single temperature suspension culture in a controlled bench-top bioreactor. Furthermore, the degree of intramolecular aggregation of IFN-β was reduced significantly (59%) compared with control cultures, largely due to the lower temperature but also partially due to the presence of microcarriers. These results indicate that the hypothermic conditions in a Cytopore culture had a combined and possibly synergistic effect of increasing volumetric production of the recombinant protein.
The use of glycoproteins as biopharmaceuticals has increased enormously in the last few years putting considerable pressure on process development to enhance the efficiency of production.1 Chinese hamster ovary (CHO) cells have been adopted as an industrial standard for the production of most recombinant proteins, as the cells are well characterized and shown to be capable of stably producing proteins with a similar glycosylation profile to equivalent human forms.2–4 However, the production of each recombinant protein may present some specific difficulties requiring bioprocess design.
Recombinant human β-Interferon (IFN-β) is used for the treatment of relapsing and remitting multiple sclerosis.5–7 IFN-β is a 166 amino acid glycoprotein produced naturally in human cells in response to viral infection or exposure to other biologics. There are two recombinant forms of IFN-β available clinically: IFN-β-1a and IFN-β-1b. IFN-β-1a is expressed in CHO cells and is a glycosylated protein similar to that found in humans.8 IFN-β-1b is non-glycosylated with a single point mutation at Cys-17 to Ser produced in E. coli.9 Runkel10 established that IFN-β-1a is 10–15 times more active compared with IFN-β-1b and this was attributed to the presence of a glycan structure.
The glycosylation profile of IFN-β from mammalian cells appears to be fairly robust with respect to changes of culture conditions.11, 12 However, the greatest challenge for the production of IFN-β is a tendency for intramolecular aggregation because of the hydrophobic nature of the peptide backbone.13 Several additives and culture conditions have been suggested to stabilize the molecule and minimize aggregation in culture.14
Although microporous (solid) microcarriers such as Cytodex have been found to be beneficial for the growth of anchorage-dependent cells,15–20 they are unsuitable for suspension cells. On the other hand, macroporous microcarriers can enhance the growth of non-anchorage-dependent cells following entrapment within the mesh of the microcarrier. Cells are able to enter the pores of these carriers and grow to high cell densities. Several articles report the use of commercially available macroporous microcarriers such as Cytoline,21–23 Cultispher G,15, 18, 24, 25 and Cytopore26–29 for the culture of mammalian and insect cells.
For our work, we chose Cytopore microcarriers because of their robustness and ability to be used in simple stirred tank bioreactors. These are fabricated as a transparent macroporous crosslinked cellulose microcarrier with positively charged N,N-diethyl-aminoethyl groups. The microcarriers have a sponge-like network structure in which almost 95% of the internal volume is available for cell growth. The diameter of the microcarriers is within the range 200–280 μm with an average diameter pore size of 30 μm and a total surface area of 1.1 m2/g.
The use of hypothermic conditions to enhance recombinant protein production has been well described. Lowering culture temperatures from the traditional 37°C to hypothermic culture conditions (30–35°C) has been reported to enhance the specific productivity and volumetric productivity of a culture. There have been many reports of the enhancement of recombinant protein production using this culture strategy. Examples include the following. Chen et al.30 reported a 40% increase in pro-urokinase with a lowered culture temperature to 31°C. Shi et al.31 showed a 3.2-fold increase in antibody-IL-2 fusion protein at 30°C. Fogolin et al.32 described a 2.1-fold increase in recombinant human granulocyte macrophage colony stimulating factor.
In the present study, we investigated the growth and productivity of CHO cells producing recombinant human IFN-β in Cytopore 2 microcarrier cultures, in temperature shifted suspension cultures, and in Cytopore 2 cultures combined with a temperature shift. Our objective was to attempt to enhance recombinant protein production as well as to minimize the intramolecular aggregation of IFN-β.
Cell line and medium
A CHO cell line transfected with the gene for human IFN-β (CHO 674) was provided by Cangene Corporation (Winnipeg, Canada). The cells were grown in serum free medium, CHO-SFM (Biogro Technologies, Winnipeg), under a 10% CO2 atmosphere. Cells were passaged every 4 days in 75-cm2 T-flasks.
Cell culture in spinner flasks
Suspension cultures were established in 100-mL spinner flasks following inoculation with 1 × 105 cells/mL and stirred at 45 rpm. Microcarrier cultures were established in spinner flasks by the addition of Cytopore 1 or 2 (GE Healthcare) at a concentration of 1.0 mg/mL. Cells were found to enter microcarriers after 2 h at 45 rpm. The temperature was adjusted based on the culture condition. A temperature shifted or biphasic culture was grown at 37°C until the day of the shift and then decreased to 32°C.
Cell culture in bioreactor
Batch cultures were established in a 3-L Applikon bioreactor with a working volume of 2 L. The initial inoculum was 1.0 × 105 cells/mL. Cultures were maintained at pH 7.1, dissolved oxygen of 50%, temperature of 37°C, and agitation speed of 100 rpm (except in the case of Cytopore) with a marine impeller. In Cytopore cultures, after cell inoculation, the stir rate was held constant at 40 rpm for 2 h, raised to 60 rpm for 2 h, 100 rpm for 2 h, and finally 120 rpm for the duration of the culture. Cells were found to enter microcarriers after 2 h of stirring at 40 rpm. The final stir rate ensured a homogenous mixture of microcarriers throughout the bioreactor.
Cytopore microcarriers (GE Healthcare) were prepared for cultures by hydrating (1 g/100 mL PBS) and autoclaving for 20 min at 121°C in accordance with the manufacturer's instructions. The microcarriers were then washed twice with fresh PBS and resuspended in serum free media with several washings.
Suspension cultures were monitored by determination of viable cell concentrations from daily samples by the trypan blue exclusion method. A cell suspension (0.15 mL) was treated with 0.15 mL trypan blue reagent (0.5% trypan blue in phosphate buffered saline). The cells were removed from the microcarriers by treating the culture samples (0.5 mL) with 0.5 mL of crystal violet reagent (0.2 M citric acid, 0.2% w/v crystal violet, 2% w/v Triton-X-100) for up to 3 h at 37°C to lyse the cells. Then aspiration (25×) through a 25-gauge needle allowed the recovery of stained nuclei which were counted in a haemocytometer. This procedure was sufficient to remove all cells from the microcarriers, which were routinely checked for residual nuclei by microscopic examination following staining.11 To compare equivalent data, nuclei counts were also determined in suspension cultures.
The IFN-β concentration in the medium was determined by an enzyme-linked immunosorbent assay (ELISA). Media samples from cultures were centrifuged to remove cells and microcarriers, and the cell-free samples were stored immediately at −20°C for later IFN-β analysis. Assay plates (96-well) were coated with polyclonal rabbit anti-human IFN-β antibody (Biogenesis) and incubated overnight. Plates were washed with Tris-buffered saline (TBS) (3×) between all additions. Samples were diluted and incubated at room temperature. The second antibody was mouse monoclonal anti-human IFN-β (Chemicon) and was followed by goat anti-mouse IgG alkaline phosphatase conjugate (Sigma). The assay was developed with p-nitrophenyl phosphate (Sigma) and read at 405 nm. The values were compared to a standard curve of IFN-β (United States Biological), and were reported as relative units of IFN-β. The calculated coefficients of variation from each sample assay were ∼ 10%. Calibration of the assay with standards showed that 1 μg IFN-β was equivalent to 5 × 105 units.
Denaturation of IFN-β
A second set of culture samples was pretreated prior to ELISA to dissociate any aggregates. Media samples (100 μl) were mixed with 1 μL SDS (10%) and 1 μL of 2-mercaptoethanol and boiled for 5 min. The IFN-β content was determined by ELISA assay as described earlier.
Specific rate of production of recombinant proteins
Qp values of IFN-β were calculated from plots of product concentration during the growth phase against the integral of values of the growth curve (IVC).
Cell growth and productivity of IFN-β in Cytopore spinner flasks
Suspension and Cytopore (1 mg/mL) cultures of CHO cells were established by inoculation of 1 × 105 cells/mL in 100 mL CHO-SFM media in spinner flasks. For the microcarrier cultures, all cells were found to be entrapped within the microcarriers at least 2.5 h after inoculation.
The growth profile of the cultures over a 7-day period is shown in Figure 1. Cell densities were determined in relation to the total volume of the culture. The maximum specific growth rate was significantly higher for the suspension culture (0.032 h−1) compared with the Cytopore cultures (0.019 h−1 and 0.016 h−1, Cytopore 1 and 2, respectively). The maximum cell yield in the suspension culture was 4.0 × 106 cells/mL at day 6. This compared to a cell yield of 3.2 × 106 cells/mL which was attained in the microcarrier cultures at day 7. The cell/microcarrier ratio at this point was 947 and close to the saturation of the internal space offered by the microcarrier. Beyond this time point significant cell growth occurred in suspension and outside the microcarriers.
The production of recombinant IFN-β was monitored from culture samples by an ELISA assay. Previous work showed that the ELISA response of culture samples could be enhanced by a step of protein denaturation with detergent at high temperature.14 This caused a breakdown of aggregates formed by hydrophobic interaction of IFN-β. The breakdown of aggregates causes the exposure of the epitopes recognized by the specific antibody used in the ELISA. The ELISA response was determined before and after sample denaturation with the difference attributed to the degree of aggregation of IFN-β.
Samples from the cultures shown in Figure 1 were analyzed on days 2, 4, and 6. The suspension cultures showed a maximum yield of 4.04 × 106 units/mL at day 6 with a level of aggregation determined to be 34% (Table 1). However, the yield in the microcarrier cultures was >twofold higher at 11.4 × 106 units/mL (Cytopore 1) and 10.6 × 106 units/mL (Cytopore 2). Furthermore, the level of aggregation was significantly lower in the microcarrier cultures, particularly with Cytopore 2 (2%). The decision to pursue further experiments with Cytopore 2 was based on this higher yield of monomeric IFN-β.
Table 1. Volumetric Production of IFN-β from CHO Cells Grown from Microcarrier and Suspension Cultures
Native protein (units/mL)
Denatured protein (units/mL)
2.67 × 106
4.04 × 106
8.22 × 106
11.42 × 106
10.36 × 106
10.57 × 106
Growth and productivity of IFN-β in biphasic cultures
Exposure of cultures to low temperature can enhance cell specific productivity.33 In the following experiment our objective was to maximize volumetric productivity by combining the beneficial effects of enhanced specific productivity with substantial cell growth. Preliminary work showed that a low temperature of 32°C allowed limited growth following a shift from the standard temperature of 37°C.14
We investigated the effects of a biphasic culture involving a shift to the lower temperature after a period of cell growth at 37°C (Figure 2). A temperature shift was induced at days 1, 2, and 3 in an attempt to determine the optimal time to enable a sustainable high cell density and high product titer. The growth profile of these cultures was compared to those maintained at either 37 or 32°C.
At 37°C the cells grew with a maximum specific growth rate of 0.032 h−1 during the exponential growth phase and reached a maximum density of 2.5 × 106 cells/mL after 4 days of culture. Cells that were incubated for an equivalent period at 32°C showed negligible growth. After day 12 at this lower temperature the population increased to just over 1.3 × 105 cells/mL. Reducing the culture temperature to 32°C after 1, 2, or 3 days at 37°C resulted in intermediate cell yields. Analysis of IFN-β showed a significantly higher production from the day 2 temperature culture with a yield of 14.4 × 106 units/mL (Table 2).
Table 2. Volumetric Production of IFN-β from CHO Cells Grown at 37°C and in Biphasic (37–32°C) Cultures
Native protein (units/mL)
Denatured protein (units/mL)
1.81 × 106
4.64 × 106
1.39 × 106
1.98 × 106
7.08 × 106
8.14 × 106
11.38 × 106
14.4 × 106
12.8 × 106
12.8 × 106
Temperature shift in microcarrier cultures
Given that the maximum yield of IFN-β was over twofold higher in Cytopore 2 cultures (Table 1) and about threefold higher in biphasic temperature cultures (Table 2) it was decided to combine the two strategies in a controlled bench top bioreactor.
Figure 3 outlines the growth profiles of cells cultured in suspension and on Cytopore 2 microcarriers in a 3-L Applikon bioreactor either at 37°C or in a biphasic temperature-shift (day 2) regimen. At 37°C the suspension culture reached a maximum cell yield of 5.5 × 106 cells/mL at day 5 (Figure 3a). In comparison, the equivalent biphasic culture showed a maximum yield of 2 × 106 cells/mL at day 9 after which the density decreased.
The microcarrier cultures in the bioreactor showed a steady increase in cell concentration to a maximum of 4.4 × 106 cells/mL after 7 days at 37°C (Figure 3b). After this point the cells appeared to detach from the microcarriers and grow in suspension. In comparison, the cell concentration of the equivalent biphasic culture reached over 8.4 × 106 cells/mL after 20 days. Only at this point did cells appear to detach from the microcarriers in the culture. The growth corresponded with a constant specific glucose uptake rate of 115 pg glucose/cell/day (r2 = 0.88) between day 1 and 20 (data not shown).
As expected the maximum yield of IFN-β in the suspension cultures occurred in the stationary phase with a significantly higher value for the biphasic culture (Figure 4). For the 37°C culture the maximum yield of the denatured samples was 5.4 × 106 units/mL which was significantly higher than the non-denatured samples suggesting a degree of product aggregation of up to 70%. For the biphasic cultures, a minimal degree of aggregation (8%) was shown up to day 8 with a significantly higher product yield of IFN-β at 8.6 × 106 units/mL. Beyond this time point the IFN-β titer increased to 10.6 × 106 units/mL at day 10, although the degree of protein aggregation also increased significantly to 40%. The product titers in the bioreactor corresponded approximately to those shown for the equivalent cultures in spinner flasks (Table 2).
For the microcarrier cultures at 37°C, the production of native IFN-β was enhanced fivefold (to 8.1 × 106 units/mL) compared to the equivalent suspension culture (Figure 4b). The temperature shift further enhanced the production of native IFN-β by 9.9-fold yielding a maximum of 16.2 × 106 units/mL by the end of the culture at day 20. In this culture, there was a constant rate of production of IFN-β from days 1 to 20 which followed the constant growth of the cells (0.72 units/cell/day, r2 = 0.96). Up to day 16 there appeared to be no protein aggregation but after this point the titer responses suggested increased aggregation up to 50% at day 22.
This shows that in these cultures the temperature shift regime increased the product yield 2.0-fold in suspension culture and by 2.3-fold in microcarrier culture at 37°C. The enhancement in yield due to both the use of microcarriers and the temperature shift was determined to be 4.2-fold which is greater than the product of the individual effects. This suggests a synergy in the enhancement of IFN-β titers from the use of microcarriers and the temperature shift regime in these cultures.
As well as the increased overall product titer, a further advantage of these culture strategies is a decrease in aggregation of IFN-β. The lower temperature decreased the degree of aggregation of IFN-β from 70 to 8% in suspension culture. The microcarriers also decreased the degree of aggregation to 33% but at a higher concentration. The combined effect was a reduction of aggregation to 29% but at a substantially enhanced product yield of 22.9 × 106 units/mL. This product yield could be further enhanced toward the end of the culture (26.8 × 106 units/mL) but at the expense of increased aggregation (44%).
The use of Cytopore microcarriers to increase recombinant protein production in cell culture has been well documented. The viability of cells appears to be extended in the cultures and this increases the amount of recombinant protein obtained. Several articles show this effect in culture systems incorporating Cytopore, for example; t-PA,29 u-PA,28 pro-UK,27 and prothrombin.26 Our lab previously reported a 30% increase in the titer of recombinant human IFN-β produced in CHO cells grown in Cytopore 1 microcarriers in a bioreactor compared to a suspension culture.11
There have also been other macroporous microcarriers used in mammalian cell culture. These include Cytoline, Cultispher G, and collagen gel particles. The use of Cytoline in culture was reported to increase the production of erythropoietin from CHO cells by twofold compared to a suspension culture.22 Furthermore, the production of an osteoblast derived antiviral protein from CHO cells was enhanced by 5.5-fold compared to a suspension culture.23 Yamaguchi et al.34 showed a 30- to 60-fold increase in volumetric monoclonal antibody production from BHK cells in an immobilized cell culture in collagen gel particles compared to a suspension culture.
However, the enhancement in production of recombinant proteins from microcarrier cultures is cell-line dependent. Nam et al.35 reported a difference between two CHO cell lines. In one case there was no enhancement in accumulated tPA from Cytopore 1 cultures compared to suspension cultures, whereas a second cell line produced 53% more SEAP in suspension cultures. Mignot et al.25 showed that there was no enhancement in production of Von Willebrand Factor from CHO cells grown in Cultispher G cultures.
Our study showed that Cytopore microcarriers and low temperature conditions can be combined to increase the production of IFN-β by up to 4.2-fold compared to a suspension 37°C culture. Varying the temperature alone doubled total protein production and resulted in a 4.8-fold increase in monomeric IFN-β rather than the aggregated form. An increase in volumetric productivity is attributed to an increase in specific productivity which under appropriate conditions can off set a lower cell growth rate.30–32, 36 Fogolin et al.32 reported a 2.3-fold increase in the volumetric production of recombinant human granulocyte macrophage colony stimulating factor with a 2-day temperature shift culture from 37 to 33°C. Yoon et al.33 indicated a 4-fold increase in the volumetric production of recombinant erythropoietin from CHO cells. A similar increase in volumetric productivity was shown by Chen et al.,30 for pro-urokinase. Fox et al.37 reported a 2-fold increase of interferon-γ in a culture at 32°C compared to 37°C.
However, an increase in specific protein production at low temperature also appears to be cell line specific. Yoon et al.38 found that the production of an anti-4-I-BB antibody was not enhanced at 33°C. Furthermore, they reported that 12 parental cell lines producing the anti-4-I-BB antibody showed different levels of recombinant protein expression at low temperature. Clark et al.39 reported no difference in the production of tissue plasminogen activator in 37°C and 34°C cultures.
It has been well documented that there is a decreased growth rate associated with CHO and other cells producing recombinant proteins at decreased temperature.30, 38, 40 This decreased growth rate is explained by a shift in the cell cycle from S phase to G1/G0 phase.30–32 A lower temperature may also be associated with reduced specific glucose and glutamine uptake rates and consequently reduced specific lactate and ammonia production rates.32, 41
Cells grow within Cytopore microcarriers at a slower rate than in suspension. However, our results show that that a high viable density of cells can be maintained for a prolonged period under hypothermic conditions compared with equivalent suspension cultures. The cell population within the Cytopore microcarriers reached a concentration of 2 × 108 cells/mL by the end of the culture. The maintenance of such a high cell density for a prolonged period is clearly advantageous to enhanced protein production. Similar high viable cell densities within microcarriers have been reported in the literature. Chen et al.26 reported a maximum cell density of 2.3 × 108 cells/mL bead during a culture of CHO cells producing prothrombin. CHO cells producing prourokinase grew to a final cell density of 1.6 × 108 cells/mL bead.27 Hu et al.28 reported a cell density of 2.2 × 108 cells/mL bead volume during the production of urokinase-type plasminogen activator from CHO cells.
IFN-β is prone to aggregation in culture because of its hydrophobic characteristics. This aggregation can lead to insoluble precipitates, to the loss of activity, low bioavailability, injection site reactions, and immunogenicity.42 A previous report from our lab showed that the aggregation of IFN-β produced from cell culture is temperature-dependent.14 The aggregation in culture may be reduced by the addition of glycerol14, 43–45 or DMSO.14, 45 In this study, we show that the presence of Cytopore microcarriers and hypothermic conditions independently decrease intramolecular aggregation. Furthermore, the combination of these two factors decreased the degree of intramolecular aggregation significantly, largely due to the lower temperature but also partially due to the presence of microcarriers.
In conclusion, the present report shows that the use of Cytopore microcarriers and a temperature shift regime independently resulted in a 2-fold increase in total secreted recombinant protein compared to a controlled suspension culture. There was also a 5-fold enhancement in the production of monomeric, nonaggregated protein, thus indicating that these two culture strategies were effective in reducing intramolecular aggregation of the IFN-β. The combined effects of Cytopore and temperature shift showed a 4.2-fold increase in total protein production and a 9.9-fold increase in monomeric protein. Thus the combined effect of a lowered temperature in the presence of Cytopore 2 microcarriers was additive and synergistic for the enhanced production of IFN-β.
The work was supported by the Natural Science and Engineering Council (NSERC) of Canada with a network grant (Cellnet).