The variability in expression of recombinant proteins has been analyze
d with regard to (a) comparison of clones from the same transfection e
xperiment; (b) comparison of the same genetic construct in different c
ell lines; (c) the effect of the culture system used (free suspension,
aggregate suspension, and microcarrier); and (d) physicochemical para
meters in long-term (100d) culture in a macroporous fixed bed bioreact
or (FBR). Differences in product expression between clones were accomp
anied by differences in growth rates, metabolic kinetics, and ability
to grow in suspension as opposed to attached culture. The single most
important factor affecting product expression when comparing construct
s (for SEAP and IgG), cell lines (BHK 21 and myeloma), and culture sys
tems was whether cells were grown in an attached or suspension mode. T
hus key factors could be related to cell morphology (suspension versus
monolayer), the presence of microenvironments and physiological stres
s to control growth rate. The relationship of key process parameters t
o volumetric and specific rAb productivity of the FBR was investigated
in a partial factorial experiment with a rBHK cell line. The highest
productivity levels are associated with a combination of the highest v
alues tested for re-cycle (195 ml min(-1)) and dilution rates (1 d(-1)
) and glutamine concentration (2.5 mmol l(-1)), plus the lowest values
for bead size (2 mm) and inoculum density (10(7) ml(-1)). Together wi
th data from fluidised bed cultures, these results suggest that higher
productivity is not primarily the result of greater cell numbers with
in the system but more the physicochemical definition of the system.