A novel real-time system to monitor cell aggregation and trajectories in rotating wall vessel bioreactors

Abstract

Rotating wall vessel bioreactors (RWVs) constitute dynamic suspension culture venues for tissue engineering. Quantitative real-time assessment of the kinetics of cell–cell aggregation in RWVs can yield mechanistic information about the initial steps leading to the assembly of individual cells into tissue-like constructs. In our imaging system, fluorescently labeled cells suspended in a HARV-type RWV were irradiated by a laser-beam. Emission was recorded by a camera mounted at 90° to the excitation plane. Using macro lenses, the system identified ∼5 μm particles from a 5 cm working distance, distinguished aggregated 20 μm microspheres from larger (45 and 90 μm) microspheres, and plotted local trajectories of microspheres and cells. Sizes of PC12 cells assessed by our system matched conventional measurements. We validated the system's ability to follow HepG2 and PC12 aggregation in real time over 24 h of RWV culture. Taken together, our system provides the means to measure and analyze in real time the processes that lead to the 3D tissue-like assembly of diverse cell types into spheroids. Future studies include development of intelligent feedback algorithms, allowing automatic control over RWV rotational speed required to maintain aggregating cells and nascent tissue in continual free fall.

Introduction

The RWV bioreactor is a widely acknowledged venue for three dimensional tissue engineering and simulated microgravity on earth (Goodwin et al., 1993, Unsworth and Lelkes, 1998). It has been proved to facilitate aggregation and been used to support chondrogenesis for bone repair and replacement (Duke et al., 1993), and to assist in the aggregation of embryonic limb cells (Duke et al., 1996). The RWV provides solid body rotation of the enclosed fluid medium (Schwarz et al., 1992), and particles are continually suspended because the gravity-induced sedimentation is balanced by the upward forces produced by rotating the vessel (Nickerson et al., 2004).

Cells aggregating in bioreactors over time experience increasing mass, resulting in enhanced gravitational pull which in turn leads to increased shear stress and viscous drag acting on the aggregates. Thus for increasing aggregate sizes the rotation speed of the vessel has to be adjusted to meet solid body rotation requirements. Currently, aggregates are kept in continual free fall by empirically/manually making minor increments to the rotational speed at regular time intervals, based on experience and visual examination of the way in which the samples “tumble” in the rotating bioreactors (Duke et al., 1996). An imaging system that could detect particle motion in real time would be useful for both understanding processes leading to cell–cell aggregation as well as for designing a feedback loop that automatically can adjust the rotational speed of the RWV bioreactor based on the real-time analysis of images of particle movement captured by time lapse photography.

To date, no real-time systems exist for 3D cell aggregation measurement because of optical limitations. Experimental models of aggregation have been developed for cells seeded on micro-carrier beads in RWVs (Muhitch et al., 2000) as well as other cell systems. Traditionally the kinetics of cell aggregation are modeled based on Smoluchowski's (1917) population-balance equations (Nguyen and O’Rear, 1990). There is however, limited experimental verification for these models (Enmon et al., 2002) and no real-time system that provides instantaneous information of cell aggregate size. Gao et al. (1997) studied the motion of microcarrier particles in HARV type RWV bioreactors numerically, but their work does not include experimental verification in real-time but is based on analysis of samples extracted from the bioreactor.

In order to study cell aggregation in HARVs we focus on two different cell types, PC12 rat pheochromocytoma cells and HepG2 human hepatocellular carcinoma cells. PC12, and to a greater extent HepG2 cells, form large aggregates with histotypic epitheloid morphology in RWV culture (Lelkes et al., 2004, Khaoustov et al., 2001), unlike other cells, such as the insect ovary cell line SF-9 (Cowger et al., 1999) for example, which do not aggregate in RWV conditions.

In this study, we describe a newly designed imaging system which we used to record and analyze the motion and aggregation of PC12 and HepG2 cells in real time. Using a similar optical system with significantly lesser resolution, Pollack et al. (2000) measured trajectories of large (1 mm diameter) microcarrier particles. Qiu et al. (1999) used the same system to evaluate different kinds of microcarrier beads and selected the most suitable ones for use in RWV based cell culture. Significantly, in this study we now further advance the system developed by Pollack et al., into a true high-resolution (<2.5 μm/pixel) cell imaging device.