Scalable, High-Density Expansion of Human Mesenchymal Stem Cells on Microcarriers Using the Bach Impeller in Stirred-Tank Reactors

Abstract

This paper describes the results of process developmental experiments to achieve higher cell densities in the manufacturing of hMSCs using the novel Bach impeller in a stirred-tank bioreactor. Engineering experiments have previously shown that the Bach impeller represents an efficient mixing device that suspends particles in fluids at low power inputs. To assess the impeller during biological experiments, the growth performance of Wharton Jelly (WJ)-hMSCs in a 1 L STR equipped with the Bach impeller was evaluated at a variety of culture conditions. The cells attached to Cytodex 1 microcarriers at a concentration of 5.6 g/L and were cultured for 5-7 days. The growth phase was carried out at varying impeller speeds $N$ = 75, 115, and 150 rpm. Cell growth was additionally evaluated at a microcarrier concentration of 11.2 g/L Cytodex 1. Here, a maximum cell density of up to 1.7 × 10⁶ cells/mL and cell viability > 90% was achieved within 5 culture days, which is amongst the highest cell densities ever attained for a hMSC batch culture. Critical cell quality attributes of the WJ-hMSCs were assessed upon completion of the growth phase, that is, FACS to identify stem cell surface markers, tri-lineage differentiation, and capacity of the cells to form colonies. In addition, informed by the previously described engineering characterization, the 1 L process at $N$ = 75 rpm was scaled up to the 5 L scale, where WJ-hMSCs were again confirmed to have retained the relevant cell quality attributes. The reported findings are important to determine the design space to which scale-ups to even larger tank sizes can adhere.

Keywords: bioprocessing; engineering characterization; mesenchymal stem cells; microcarriers; stirred‐tank reactors.

Figure 1

Configuration and detailed views of the UniVessel 1 L bioreactor and Bach impeller. (A) Top and (B) isometric view of the Univessel 1 L bioreactor. The bioreactor is equipped with three probes (temperature, pH, and dissolved oxygen) and three fluid exchange lines. (C) Bach impeller. Adapted from Wyrobnik et al. (2022) and Wyrobnik (2023). D_L, line diameter; D_P, probe diameter; L, line; L_L, line length; L_P, probe length; P, probe.

Figure 2

Power characteristics of the Bach impeller with and without submerged probes and lines: (A) Power number ($N_P$) curve of the Bach impeller ($D/T$ = 0.52, $C$ = 0.33 $T$) plotted against Reynolds number ($\text{Re}$), comparing configurations without submerged probes with six submerged culture internals (three probes and three lines). (B) Volumetric power input ($P/V$) as a function of impeller speed ($N$) with and without internals. Ensemble‐averaged shear stresses, (C) without or (D) with probes, at $\text{Re}$ = 2900 ($N$ = 50 rpm). (E) Maximum shear stresses (${\tau }_{\max }$) as a function of $N$. Data represents the mean (n = 3). C, impeller off‐bottom clearance; $D$, impeller diameter; $T$, vessel tank.

Figure 3

Oxygen transfer characteristics of the Bach impeller with and without submerged probes and lines. (A) Oxygen mass transfer coefficient ($k_{L}a$) and (B) oxygen transfer rate ($\mathit{OTR}$) for the Bach impeller ($D/T$ = 0.52, $C$ = 0.33 $T$) with six submerged culture internals (three probes and three lines). Data represents the mean ± SD (n = 3). C, impeller off‐bottom clearance; D, impeller diameter; T, vessel tank.

Figure 4

Growth kinetics and cell confluence of WJ‐hMSC bioreactor cultures at various impeller speeds. (A–C) Growth kinetics for WJ‐hMSC cultures in bioreactors operated at 75, 115, and 150 rpm. Each curve represents an independent bioreactor run, denoted by the number in parentheses (e.g., 75 rpm (1), (2), etc.). Black‐filled markers indicate viable cell density (VCD), and white‐filled markers indicate cell viability. (D) Growth kinetics in spinner flasks operated at 30 rpm (n = 5 independent runs). Black‐filled markers represent VCD, and white‐filled markers indicate cell viability. (E) Microscopic images of WJ‐hMSCs on Cytodex 1 microcarriers at different impeller speeds. Scale bar = 300 μm. Data represents the mean of technical replicates ± SD (n = 3).

Figure 5

Quality control analyses of WJ‐hMSCs cultured at different impeller speeds. (A) Population doubling time. (B) Colony‐forming unit efficiency across passages. (C) A fold increase in cell numbers at Days 5 and 6. (D) Expression of surface markers. (E) Tri‐lineage differentiation capacity into adipocytes (top, scale bar = 75 μm), osteocytes (bottom left, 24‐well plates), and chondrocytes (bottom right, 96‐well plates). Data represents the mean of technical replicates ± SD (n = 3).

Figure 6

Evaluation of WJ‐hMSC growth and quality metrics at impeller speed of 75 rpm and higher microcarrier concentration (11.2 g/L Cytodex 1). (A) Viable cell density (VCD; black markers) and cell viability (white markers) over the culture period. (B) Metabolite profiles. Data points are color‐coded by condition: green (75 rpm, replicate 1), blue (75 rpm, replicate 2), red (spinner flask control, replicate 1), and black (spinner flask control, replicate 2). (C) Tri‐lineage differentiation capacity of cells harvested from bioreactor and spinner flask in triplicates: adipocytes (left), chondrocytes (top right), and osteocytes (bottom right). Scale bar = 30 μm. (D) Population doubling time for three consecutive passages post‐thawing. (E) Stem cell surface marker expression. Data represents the mean of technical replicates ± SD (n = 3).

Figure 7

Projected scaling conditions for WJ‐hMSC culture from 1 to 5 L bioreactor. (A) Hypothetical scale‐up of the $N$ = 75 rpm process or (B) a $N$ = 115 rpm process from 1 to 5 L bioreactor, using various scaling criteria. The area highlighted in orange represents the broader design space of rotational speed that may be feasible at the 1 L scale (i.e., 50–150 rpm). The area highlighted in green represents conditions that were shown to reliably offer favorable growth conditions for the WJ‐hMSCs (i.e., 75–115 rpm). Impeller speeds above and below these shaded areas were not viable options for the cell culture. $N$, agitation speed; OTR, oxygen transfer rate; $P/V$, power input; $\text{Re}$, Reynolds number; t_M , mixing time; ${\mathit{u}}_{\mathit{tip}}$, impeller tip speed.

Figure 8

Evaluation of WJ‐hMSC expansion and quality attributes in a 5 L bioreactor. (A) Viable cell density (VCD; black markers) and cell viability (white markers) over the culture period. (B) Metabolite profiles over the culture period. Blue markers indicate samples from the 75 rpm bioreactor condition, and black markers represent those from the spinner flask control. (C) Stem cell surface marker expression. (D) Population doubling time for three consecutive passages post‐thawing. (E) Tri‐lineage differentiation capacity of cells harvested from bioreactor and spinner flask in triplicates: adipocytes (left), chondrocytes (top right), and osteocytes (bottom right). Scale bar = 30 μm. Data represents the mean of technical replicates ± SD (n = 3).