Disposable microbial and animal cell bioreactors for use in a disposable biopharmaceutical manufacturing process

This thesis concerns the concept of disposables for use in the manufacture of biopharmaceuticals. A disposable process is where all the equipment in contact with the process is disposable. Specifically disposable bioreactors for microbial fermentation and animal cell culture were investigated. For the disposable microbial bioreactor, the plunging jet reactor was proposed as the best option where the thesis details the design and manufacture for a disposable plunging jet bioreactor for use in a microbial fermentation. The 14 L disposable plunging jet bioreactor was compared with a 5 L stirred tank bioreactor. The comparison was made on both the physical performance of the bioreactors and the resultant performance of the fermentation. The maximum kLa in the plunging jet bioreactor was 0.13 to 0.15 s-1, which was comparable to kLa of 0.12 s-1 in the stirred tank bioreactor. This oxygen transfer was achieved with the stirred tank operating at a maximum power per unit volume of 11.2 to 12.7 kW m-3 whilst the plunging jet bioreactor operated at a lower maximum power per unit volume of 1.4 kW m-3. The plunging jet bioreactor incurred a higher maximum shear rate of 85000 s-1, compared to the stirred tank bioreactor with a maximum shear rate of 13000 to 135000 s-1. The mode of operation to maintain the DOT above 10 to 20 [percent] for the two bioreactors was different. For the plunging jet bioreactor the kLa increases with the OUR so that bioreactor was operated with a fixed power per unit volume. Operating the stirred tank bioreactor at a fixed power per unit volume results in a kLa that is fixed whilst the OUR increases. Thus for the stirred tank the power per unit volume was periodically increased resulting in an instantaneous increase in the kLa. For the plunging jet bioreactor the kLa was affected by the position of nozzle with respect to the outlet. For the plunging jet bioreactor operating at a power per unit volume of 1.4 kW m-3, with the Fab fermentation at an OUR of 5 mol L-1 hr-1; the kLa increased from 0.005 s-1 to 0.045 s-1 as a result of moving the nozzle angle from vertical (0°) to 5° away from the outlet (i.e. to an of angle of -5°). For both the Wild Type and Fab fermentation there were distinct differences between the resultant performance of the fermenations in the stirred tank bioreactor and the plunging jet bioreactor. For the Wild Type fermentation in the plunging jet bioreactor, lysis occurred within the first 140 minutes of growth, whilst in the stirred tank bioreactor the fermentation was grown as per the protocol. Unlike with the Wild Type fermentation, the Fab fermentation was grown in both the bioreactors without lysis, where the growth in the plunging jet bioreactor was attributed to the greater strength of Fab microbial cells compared to the Wild Type microbial cells. Despite the growth of the Fab fermentation in both bioreactors, there were distinct differences in the resultant performance of the fermentation. In the plunging jet bioreactor the Fab fermentation had both a lower maximum OUR and a lower final dry weight. In addition the RQ varied between 0.5 to 1.0, where addition of glycerol resulted in an increase in the RQ. In the stirred tank bioreactor the final concentration of the Fab product was 75 mg L-1 in the solid fraction and 14 to 22 mg L-1 in the supernatant. In the plunging jet bioreactor, the final concentration of the Fab product was 0.005 mg L-1 in the supernant and zero concentration in the solid fraction. This difference in the Fab concentration between the two bioreactors was attributed to shear in the plunging jet bioreactor stripping away the outer polysaccharide layer, so that the Fab product was released and subsequently degraded in the fermentation broth. The thesis suggests that these differences in the resultant performance of the Wild Type and Fab fermentation are attributed to the higher shear rate in the plunging jet bioreactor. Further work must be performed to determine whether reducing the shear rate of the nozzle whilst maintaining the mass transfer and mixing will result in a bioreactor whose resultant performance of the fermentation is comparable with the stirred tank bioreactor. The high turbulence nozzle, developed by Kenyres has been suggested as a potential method of reducing the shear rate in the bioreactor. The plunging jet reactor, bubble column reactor and the airlift reactor were all evaluated as potential disposable animal cell bioreactors. The comparison was made on the basis of measuring the kLa, CO2 stripping and mixing versus the power per unit volume for the three reactors. The targets for the kLa, CO2 stripping and mixing were measured in the airlift at a superficial velocity of 0.01 and 0.015 ms-1 which is the typical maximum superficial velocity range that a large scale industrial airlift operates during a biopharmaceutical process. The thesis concludes that the plunging jet reactor is not suitable as a disposable animal cell bioreactor. A 6 L plunging jet reactor was operating with a maximum power per unit volume of 0.01 kW m-3 which corresponded to a shear rate and stress of 15000 s-1 and 15N m-2 respectively. Whilst the upper kLa and mixing targets of 8.9 hr-1 and 13 seconds respectively were achieved below this maximum power per unit volume, operating at the maximum power per unit volume resulted in CO2 stripping of 0.030 hr-1, which was just within the range of the lower target. Whilst the 6 L plunging jet reactor performance was limited by its CO2 stripping, the principle reason for rejecting the reactor was that there were perceived to be problems in scaling up the reactor. Increasing the scale of the reactor would probably require an increase in the nozzle diameter and the use of multiple nozzles in order to maintain a low shear rate and prevent a 'foam' in the upper part of the bioreactor. Measuring the kLa, mixing and CO2 stripping for an airlift and bubble column both of 26 L scale, showed that at small scale the two reactors had comparable performance. In terms of construction the only difference between the bubble column reactor and the airlift reactor is the presence of a divider in the latter. The proposed design for the disposable airlift shown in section 7.3 has a square cross section so that the divider is diagonal. This insures that the divider is taut and is welded along the main seams of the disposable bag. Thus the airlift's construction is not sufficiently more complex than the bubble column and thus is the preferred option for a disposable animal cell bioreactor. This is because at large scale, published work has shown that the airlift has better mixing and aeration than the bubble column so that the airlift is more scaleable. Further work is required to determine whether the proposed design will result in a 7 day animal cell culture that is comparable with the same animal culture performed in a conventional airlift or stirred tank bioreactor.