1 Introduction
Traditionally, bioreactor performance is quantified by considering the
critical transport processes and the biochemical conversion process
itself [1]. Since most industrial bioprocesses are aerobic, oxygen
is a limiting nutrient for cell growth and metabolic production; thus,
oxygen transport into the liquid phase is a limiting factor in
bioreactor and overall process performance [2]. It is therefore
critical to maintain adequate transport of oxygen from the gas phase
into the culture medium. Commonly, characterization of bioreactor oxygen
transport is presented as the volumetric mass transfer coefficient,
kLa which can be used to compare oxygen transfer
performance from the microscale to industrial volumes [2,3] and is
an established criterion for predictive process scale-up [4,5].
With the recent development of cell-based therapies to treat a wide
range of unmet clinical needs, pharmaceutical companies have significant
pressure to introduce new products to the market as quickly as possible
[6]. When developing new therapies, early stage process development
and optimization requires extensive screening of culture conditions to
maximize cell growth and viability, with thousands of experiments
optimized across hundreds of variables and the best performing
conditions transferred to downstream production. This screening process
is time consuming and expensive, due to the high cost of media
components and often reduced cell growth of the cellular therapy cell
types [7]. As a result, it is advantageous to implement a scaled
development process wherein testing of the broadest parameter space
takes place in the smallest possible volumes with promising candidates
then transferred to larger volumes for additional testing.
Microplates present an ideal format for early process development due to
their low working volumes (µL-nL) and throughput and automation
capabilities. However, the utility of microplates for high throughput
process development is limited due to poor control of oxygen transfer or
kLa (h-1) in these formats. Ideally,
it would be possible to control the kLa in microplates
to match those seen in larger volume mammalian cell culture systems,
such as the Sartorius ambr® 15/250 (15-250mL) and Cytiva’s Wave Cellbag
(2L-200L). For the former, kLa value ranges are of 0.18
- 7.90 h-1 and 2.15 -11.52 h-1 are
predicted while the latter can achieve kLa values that
range from 1-60+ h-1 [8,9].
Mechanical shaking is the only commercially deployed solution for
microplate mixing of cultures, and its utility is limited. Microplate
shaking [10] induces splashing motion by overcoming surface tension.
Splashing is an inertial effect, and as microplate well size decreases,
the inertia of the droplet decreases, while surface tension remains the
same. Thus, smaller wells require greater shaking energy (amplitude,
frequency) to achieve the same mixing performance. For 96 well plates
this critical threshold has been observed to be approximately 800 RPM
(for a 3mm shake amplitude) [10,11]. Achieving the desired results
from mechanical shaking is surprisingly challenging. While
kLa in a conventional bioreactor improves smoothly with
increasing impellor speed [12], shaking performance depends on
whether the shaking is in-phase or out-of-phase. Increased frequency has
little impact on out-of-phase mixing, and the harmonic frequency of
microwells varies dramatically with well size and shape [11]. Due to
these limitations, as well as concerns related to shear stress induced
damage at high shake frequencies, commercially available incubator
shakers are not marketed for use with mammalian cell culture. Shake-less
methods such as acoustic mixing exists, but adoption has been limited by
concerns about heating and scalability. In short, gentle microplate
agitation remains a major unmet need for bioprocess discovery and
development, especially for mammalian cell culture.
This research reports the development and characterization of a novel
microliter scale culture plate with novel surface-attached agitation
methodology. The microplate is designed for high throughput process
development applications due to its operation in a 96-well culture plate
with culture volume of 150-300 µL allowing for adequate sampling for
bioanalysis. Evaporation of culture media during operation, which has
been shown to be an issue in small volume culture reactors [13], is
compensated for by operation in a high humidity environment. The oxygen
transfer performance of the microreactor is benchmarked against
conventional small-scale culture systems.