Identifying Criteria and Indicators for Assessing External
Forcing and Internal Variability-Generated Uncertainly in Stem Cell
Bioprocesses
Understanding the prevalence of pre-existing intrinsic and extrinsic
variability over external noise is important for developing key
indicators to identify and predict the bioprocess variability. The
phenomenon of “intrinsic disorder” has been widely studied in
bioprocess conditions, and understanding and controlling the sources of
variability are anticipated to play important roles in improving
bioprocess control and optimization (Kim and Kino-oka, 2020b;
Misener et al., 2018). Indeed, it
has been shown that the phenotypic variability among stem cells can
contribute to the inherent differences resulting from donor-dependent
variability as well as changes induced during cellular reprogramming and
expansion and cell handling processes such as passaging, freezing, or
thawing (Kim and Kino-oka, 2018). The causes of cell variability
occurring in a series of bioprocessing steps and the related mechanisms
are often still a “black box.”
While the impact of forces and mechanical environment on the structure
and function of stem cells has long been appreciated (Ivanovska et al.,
2015; Liu et al., 2010; Maul et al., 2011; Mendelson and Frenette, 2014;
Naqvi and McNamara, 2020), the importance of mechanical forces in
bioprocessing has only recently been investigated. The potential sources
of cell-to-cell variability to bioprocesses—described here as
intrinsic (intra-cellular) noise, extrinsic (inter-cellular) noise, or
external (environmental) noise—have been difficult to investigate
because the evidence assessing the effect of external environmental
fluctuations on a system is limited (Hilfinger and Paulsson, 2011). The
intrinsic noise often refers to the inherent stochasticity of
biochemical processes, such as transcription and translation (Raser and
O’shea, 2005; Lei, 2009; Soltani et al., 2016; Swain et al. 2002;
Thomas, 2019). In contrast, the extrinsic noise and external noise are
generated form cellular processes, such as cell-cell and cell-matrix
contacts, and cell migration or form environmental fluctuation,
respectively (Lei, 2009; Hilfinger and Paulsson, 2011; Swain et al.
2002). Thus, it is essential to identify all factors causing intrinsic
disorder within a bioprocess, based on the noise sources associated with
a process in the locations and interactions among different sources,
both within and outside the cells.
The adaptation mechanisms involve a multistep cellular
mechanotransduction process, including: (i) mechanosensing —
conversion of mechanical forces into local mechanical signals, such as
fluid shear stresses, which initiate a cellular response and (ii)
mechanosignaling transduction of an intracellular signaling event
occurring in response to a mechanical force and gene expression or
protein activation, which ultimately alters cell phenotype and function
(Argentati et al., 2019; Ingber, 2018; Mammoto et al., 2013; Sun et al.,
2012) (Figure 3 ). Cells probe and respond to the forces of
their surroundings through cytoskeletal networks composed of actin
filaments (F-actin), myosin motors (Jégou and Romet-Lemonne, 2021;
Lenormand et al., 2007; Matthews et al., 2006). F-actin filaments are
assembled with the assistance of actin nucleators, such as formins and
Arp2/3, which are involved in generating long unbranched and branched
actin filaments, respectively (Rotty et al., 2013; Zalevsky et al.,
2001). Furthermore, the myosin-driven contraction of the actin
cytoskeleton plays a central role in the ability of stress fibers to
sense matrix mechanics and generate contractile forces inside cells
(Weirich et al., 2021). Both externally-applied and cell-generated
mechanical forces are transmitted across the cytoskeleton to the
nucleus, where they alter the epigenetic state and expression of genes
relating to cellular homeostasis to maintain self-renewal and
pluripotency (Li et al., 2020;
Vining and Mooney, 2017). Based on these mechanisms, the intracellular
processes favoring the interpretation of mechanical cues can be
classified here. In this context, the first step is universally
recognized as the change in cell behaviors, such as cell-cell
interaction, cell-substrate interaction, and cell migration, which
serves as the central trigger, converting these input signals into
cellular outputs, which in turn manifest as changes to the actin
cytoskeleton of the cells. In addition, the actin-myosin cytoskeleton is
a major “integrator” and “organizer” of mechanical and biochemical
signaling inputs that continuously senses, organizes, and integrates
these signals using its control “effector” protein belonging to the
Rho family of GTPases (Arnold et al., 2017; Kimura et al., 1996).
Specially, the antagonistic activities of RhoA and Rac1 GTPases play a
role in the dynamics of myosin-mediated contraction and relaxation
during cell migration in several cellular settings
(Thanuthanakhun et al., 2021).
This confirms that the cells adapt to culture environment through the
alteration of the Rho-Rho kinase-phospho-myosin pathway, influencing the
epigenetic modifications and the transcription factors that contribute
to their inherent states (Thanuthanakhun et al., 2021). Studies using
ESCs and iPSC cultures revealed that the time-dependent regulation of
two key bivalent epigenetic marks, histone H3 trimethylation at lysine 4
(H3K4me3) and histone H3 trimethylation at lysine 27 (H3K27me3), could
contribute to the maintenance of the stem cell state and differentiation
potential (Harikumar and Meshorer, 2015; Li et al., 2018; Thanuthanakhun
et al., 2021). This regulation can take place through epigenetic changes
and may be manifested as a “memorizer.” Memories acquired during
culture are initialized by the disruption of the actin cytoskeleton
during enzymatic digestion-based passage culture while maintaining cell
homeostasis.
Thus, by adopting these definitions, we can describe the intracellular
processes favoring the interpretation of mechanical cues and concentric
groups acting to deliver the signals from cell surface receptors in the
plasma membrane to the nucleus via cytoskeleton (Figure 4 ).
This chain reaction can eventually lead to epigenetic changes and may be
manifested as “memorizer.” These multiple correlations can be
described using the key factors “trigger,” “effector,” and
“memorizer,” and the impact of mechanical force application to the
cells may be best described using the terms “sensory,” “short-term,”
and “long-term” memory, implying that any of the various definitions
or various components can be used to describe different coordinating
systems. The short-term memory induced during growth processes is
transient and repeatedly undergoes initialization and re-formation
during the passage culture. However, the acquired long-term memory is
permanent and is stored or consolidated after passage culture. This
simple conceptual framework and the comprehensive classification
approach described here provides potential mechanistic insights into the
formation and maintenance of epigenetic memory for lasting changes in
cell behavior during different growth phases, and indicates that the
appropriate seeding density and time required to maintain cellular
homeostasis are important from a technological perspective for
developing and optimizing bioprocess design and operation.
Mechanobiological Conceptual Framework for Assessment and
Prediction of Stem Cell Bioprocess: A Case Study
To understand how cells control and exploit the extent of intrinsic
disorder in bioprocesses, we must develop a conceptual framework to
define the objective of the assessment. Here, we illustrate a conceptual
framework with example case studies and
discuss about future outlooks and
challenges on the use of data-driven approaches (Figure 5 ). The
comparison between intrinsic disorder sources highlights how the
consensus on biological indicators of intrinsic disorder captures the
main features of intrinsic disorder and correlates well with growth
kinetic profiles. Growth kinetics studies are often used to optimize
bioprocesses because cells respond to external disturbances by
maintaining a homeostatic level of mechanical tension, and these studies
lead to the establishment of specific criteria for cell and tissue
production (Ehrig et al., 2019; Kato et al., 2018; Nath et al., 2018).
In many stem cell studies, it has been reported that the ability to
adjust growth kinetics in response to unpredictable environmental
fluctuations is associated with the impacts of mechanical force changes
(Borys et al., 2021; Kato et al., 2018; Nath et al., 2018). Four basic
shapes are hypothesized as plausible to represent the dominant growth
curve shape experienced by bioengineers during the culture period. There
are four different types of growth: time-dependent, time-delay, tardive,
and uncertainty. The identification of phenomena related to the quality
of the growth kinetics analysis depends directly on the processing
applied to the initial and subsequent data. The use of growth kinetics
analysis in the estimation of the impact of study interventions not only
improves the statistical quality of the estimates but also enables
rational process optimization alongside the dynamic changes of cell
requirements throughout the process progression.
For successful development and optimization of stem cell bioprocesses,
we must study the fluctuations in current bioprocesses and the
variability in stem cells and derived products allowed by regulators and
their applications to obtain an understanding of improvement areas (Kim
and Kino-oka, 2020a; Kim and Kino-oka, 2020b). The main problem is the
lack of biological indicators that use formulas to facilitate their
assessment and comparability of bioprocess effectiveness for various
external stimuli. Thus, the
designed classifier based on mechanical signals will help understand the
relationship between externally forced and internally generated
variations and identify the time-varying noise generation within and
between processes. Based on mechanobiological conceptual frameworks,
identifying and quantifying the relationship between inputs, outputs,
and changes in bioprocess properties can provide deep insights into
bioprocess interactions and would allow the opportunity to balance the
costs and benefits of cell production while explicitly acknowledging and
internalizing the unintended outcomes of bioprocess strategies. The
proposed indicators for assessing a fluctuating interaction network and
time-varying stability and consistency are used together with target or
threshold values of the metrics employed for determining the degree of
stability of a bioprocess responding to externally applied forces. The
discussion on which biological indicators are more suitable for
describing the cause of intrinsic disorders in stem cell bioprocesses
and predicting the impact of mechanical forces in nature and engineered
systems is still ongoing. The processes should fit within an overall
stem cell-based bioprocess, and where current automation is capable of
replicating manual processes, if we consider future production systems
there must be a move towards machine tools that are more advanced than a
human manual process (Ratcliffe et al., 2011). The determination of
growth kinetics, thus, paved a way for the optimal design of the
operating conditions and process for optimal product formation.
Acceptable ranges for a subset of these indicators were established
based on a combination of mechanical, biological, and clinical studies
and prior knowledge. The determination of the boundary conditions of
applied mechanical forces is important to ascertain robust and
reproducible cell culture processes that result in the right product
quality and suitable product yields. The proposed conceptual framework
here will help reduce process variability and increase within- and
between-process reproducibility by understanding bioprocess forces and
cellular responses of interest and will ensure that the calibrated
apparatus performance checks are carried out and operation limits are
set adequately.