Very high density of Chinese hamster ovary cells in perfusion by alternating tangential flow or tangential flow filtration in WAVE Bioreactor™-part II: Applications for antibody production and cryopreservation.

A high cell density perfusion process of monoclonal antibody (MAb) producing Chinese hamster ovary (CHO) cells was developed in disposable WAVE Bioreactor™ using external hollow fiber (HF) filter as cell separation device. Tangential flow filtration (TFF) and alternating tangential flow (ATF) systems were compared and process applications of high cell density perfusion were studied here: MAb production and cryopreservation. Operations by perfusion using microfiltration (MF) or ultrafiltration (UF) with ATF or TFF and by fed-batch were compared. Cell densities higher than 10(8) cells/mL were obtained using UF TFF or UF ATF. The cells produced comparable amounts of MAb in perfusion by ATF or TFF, MF or UF. MAbs were partially retained by the MF using ATF or TFF but more severely using TFF. Consequently, MAbs were lost when cell broth was discarded from the bioreactor in the daily bleeds. The MAb cell-specific productivity was comparable at cell densities up to 1.3 × 10(8) cells/mL in perfusion and was comparable or lower in fed-batch. After 12 days, six times more MAbs were harvested using perfusion by ATF or TFF with MF or UF, compared to fed-batch and 28× more in a 1-month perfusion at 10(8) cells/mL density. Pumping at a recirculation rate up to 2.75 L/min did not damage the cells with the present TFF settings with HF short circuited. Cell cryopreservation at 0.5 × 10(8) and 10(8) cells/mL was performed using cells from a perfusion run at 10(8) cells/mL density. Cell resuscitation was very successful, showing that this system was a reliable process for cell bank manufacturing.

Very high density of CHO cells in perfusion by ATF or TFF in WAVE bioreactor™. Part I. Effect of the cell density on the process.

High cell density perfusion process of antibody producing CHO cells was developed in disposable WAVE Bioreactor™ using external hollow fiber filter as cell separation device. Both "classical" tangential flow filtration (TFF) and alternating tangential flow system (ATF) equipment were used and compared. Consistency of both TFF- and ATF-based cultures was shown at 20-35 × 10(6) cells/mL density stabilized by cell bleeds. To minimize the nutrients deprivation and by-product accumulation, a perfusion rate correlated to the cell density was applied. The cells were maintained by cell bleeds at density 0.9-1.3 × 10(8) cells/mL in growing state and at high viability for more than 2 weeks. Finally, with the present settings, maximal cell densities of 2.14 × 10(8) cells/mL, achieved for the first time in a wave-induced bioreactor, and 1.32 × 10(8) cells/mL were reached using TFF and ATF systems, respectively. Using TFF, the cell density was limited by the membrane capacity for the encountered high viscosity and by the pCO2 level. Using ATF, the cell density was limited by the vacuum capacity failing to pull the highly viscous fluid. Thus, the TFF system allowed reaching higher cell densities. The TFF inlet pressure was highly correlated to the viscosity leading to the development of a model of this pressure, which is a useful tool for hollow fiber design of TFF and ATF. At very high cell density, the viscosity introduced physical limitations. This led us to recommend cell densities under 1.46 × 10(8) cell/mL based on the analysis of the theoretical distance between the cells for the present cell line.

Effects of conditioned medium factors and passage number on Sf9 cell physiology and productivity.

The effects of conditioned medium (CM) and passage number on Spodoptera frugiperda Sf9 cell physiology and productivity have been studied. Low passage (LP) cells at passages 20-45 were compared to high passage (HP) cells at passages >100. Addition of 20% CM or 10 kDa filtrated CM to LP cells promoted growth. LP cells passed a switch in growth kinetics, characterized by a shorter lag phase and a higher growth rate, after 30-40 passages. After this point, CM lost its stimulating effect on proliferation. HP cells displayed a still shorter lag phase and reached the maximum cell density 24-48 earlier than LP cells. HP cells also exhibited higher specific productivity of recombinant protein compared to LP cells, when infected with baculovirus during the initial 48 h of culture. The specific productivity of LP cells was decreased by 30-50% by addition of 20% CM or 10 kDa filtrated CM, whereas addition of CM to cells having passed the switch in growth kinetics had no negative effect on productivity. Cell cycle analysis showed that a large proportion of HP cells, >60%, was transiently arrested in G2/M after inoculation. In LP cultures this proportion was lower, 40-45%, and addition of CM decreased the arrested population further. This correlated to the cell size, the HP cells being the largest: HP cells > LP > LP + 20% CM > LP + 20% 10 kDa filtrated CM. Since the degree of synchronization in G2/M correlated to the productivity, yeastolate limitation was used to achieve 85% G2/M synchronized cells. In this culture the specific productivity was maintained during a prolonged production phase and a 69% higher volumetric yield was obtained. The results suggest that a decreasing degree of synchronization during the course of culture partly explains the cell-density-dependent drop in productivity in Sf9 cells.

A homologue of cathepsin L identified in conditioned medium from Sf9 insect cells.

Gelatin zymography revealed the presence of proteolytic activity in conditioned medium (CM) from a serum-free, non-infected Spodoptera frugiperda, Sf9 insect cell culture. Two peptidase bands at about 49 and 39 kDa were detected and found to be proform and active form of the same enzyme. The 49-kDa form was visible on zymogram gels in samples of CM taken on days 4 and 5 of an Sf9 culture, while the 39-kDa form was seen on days 6 and 7. On basis of the inhibitor profile and substrate range, the enzyme was identified as an Sf9 homologue of cathepsin L, a papain-like cysteine peptidase. After lowering the pH of Sf9 CM to 3.5, an additional peptidase band at 22 kDa appeared. This peptidase showed the same inhibitor profile, substrate range and optimum pH (5.0) as the 39-kDa form, indicating that Sf9 cathepsin L has two active forms, at 39 and 22 kDa. Addition of the cysteine peptidase inhibitor E-64c to an Sf9 culture inhibited all proteolytic activities of Sf9 cathepsin L but did not influence the proliferation of Sf9 cells.

Antimicrobial activity of conditioned medium fractions from Spodoptera frugiperda Sf9 and Trichoplusia ni Hi5 insect cells.

Concentrated conditioned medium (CM) fractions from Spodoptera frugiperda Sf9 and Trichoplusia ni cells, eluting from a gel filtration column at around 10 kDa, were found to exhibit strong antibacterial activity against Bacillus megaterium and Escherichia coli. The B. megaterium cells incubated in the CM fraction from Sf9 cells rapidly lost viability: after 8 min the viability had decreased to 0.7%, as compared with the control. Addition of the CM fraction to E. coli cells resulted in a less drastic drop in viability: 65% viability was lost after 60 min of incubation. Further, exposure to the CM fraction caused a substantial leakage of intracellular proteins, as demonstrated by SDS-PAGE analysis. Cell lysis was confirmed by optical density measurements, microscopic investigations and flow cytometry. B. megaterium exposed to a CM fraction from T. ni cells lost 97% of their viability in about 40 min. Ubiquitin, thioredoxin and cyclophilin were identified in the antibacterial fraction from Sf9 cells by mass spectrometry and N-terminal amino acid sequencing. Other proteins in the fraction gave no matches in a database search. Since ubiquitin was shown not to cause the antimicrobial effect and thioredoxin and cyclophilin were likely not involved, the responsible agent may be an unknown protein, not yet registered in databases. The antimicrobial effect of the CM fraction from T. ni cells most probably comes from a lysozyme precursor protein.