We have investigated the effect of kLa on CHO cell growth and
recombinant protein yield in small- and large-scale OSRs. Using a
1-L OSR, the minimal kLa value which ensured good cell culture
performance was determined for three CHO cell lines. As the
working volume increased, the kLa decreased and pH and DO
limitations occurred in the cultures. The minimal kLa required to
avoid these limitations was cell-line dependent and ranged from
7 to 9 hour1
. Surprisingly, oxygen did not appear to be the most
important factor limiting cell growth. Instead, the culture pH
appeared to be the most crucial factor. In cultures of CHO-DG44
and CHO-TNFR cells for which the pH fell below 6.6, the DO only
decreased to 30 air sat. As previously reported, in suspension
cultures of mammalian cells a DO of 30 air sat is not limiting
while a pH below 6.6 is [13]. CO2 accumulation in cultures with
an insufficient kLa may be the main reason for the low pH values
observed in these cultures. The cell densities of cultures experiencing
limitations were lower than those of cultures with a
sufficient kLa. However, the biomass was similar in different
cultures in the same OSR independently of the kLa. This suggests
that the average cell size increased in the cultures experiencing
pH or DO limitations, a phenomenon reported before and suspected
to be linked to cell stress [14]. Cell-specific productivity
has been reported to increase with cell size [15]. This could
explain why recombinant protein titers were not affected in
those cultures experiencing pH or DO limitations relative to
those that did not.
In 250-mL, 500-mL and 1-L OSRs, the same kLa values led to
similar cell growth, culture conditions (pH and DO), and recombinantprotein
yields. These results suggest that minimal kLa values for
a bioprocess can be determined at small scale for a given cell clone
and that these values remain valid upon scale-up. CHO-IgG cells
cultivated without pH or DO controllers in a 200-L OSR grew
similarly, had equivalent recombinant protein titers, and generated
the same pH profile as cells in 1- and 5-L OSRs. Because the three
OSRs were operated at the same kLa, these results suggest that largescale
bioprocesses can be operated without pH or DO controllers as
long as a sufficient kLa (7–10 hour1
) is maintained through appropriate
cultivation conditions (e.g. working volume, agitation rate
and geometry of the vessel). Furthermore, this confirms that for
cultures having the same kLa, the same trends and values can be
expected for cell density, pH and recombinant protein production
independently of the scale. Thus, the kLa appears to be a good
predictor for scalability in OSRs. These probe-independent bioprocesses
are appealing as they are cost-effective and simple. They do
not require a controller or the addition of base or pure oxygen to
keep suitable cultivation conditions. The robustness of such bioprocesses
is also higher as they are independent of probe and
controller failures.