3D Organotypic Cell Cultivation
A Micro-Bioreactor for Actively Perfused Tissue Cultivation
- DAPI/TRITC stain (merged) of 3D-cultivated HepG2 cells. Laser Scanning Microscopy
- Fig. 1: Schematic overview of the single unit micro-bioreactor
- Fig. 2: Long-term cultivation of HepG2 cells. % of G2/M-phase in cell cycle
- From left to right, front row: Christian Hildmann, Andreas Schober (group leader), Anette Läffert, Maren Klett; back row: Jörg Hampl, Frank Weise, Uta Fernekorn.
The development of new, reliable systems for the testing of new compounds in pharmaceutical industry is presently a great challenge. Three-dimensional, tissue-like, organotypic cell cultures have become a serious alternative to traditional two-dimensional cell cultures or animal testing in this field. In the context of developing artificial systems, fluidic devices and a biochemical environment must be constructed and arranged in such a way that living cells can survive in a three-dimensional, technical structure. With such systems we are aiming to increase reliability of experimental toxicity tests and metabolic data compared to those from conventional cell culture systems and thereby reduce the cost and number of experimental animals in drug screening.
The most common investigation in preclinical development of new drugs is the conventional cultivation of cells on flat plasticware, the ‘two-dimensional cell culture' (2D-culture). New substances are tested in such systems in terms of ADME/Tox parameters. Unfortunately, these culture conditions do not display the nature of tissue. Differentiated cells with their organ-like function and the typical 3D-architecture cannot be achieved by cultivating in a 2D-manner. 2D-cultivation, therefore, does not create an in vivo cellular environment, where cells interact with extracellular matrix (ECM) and neighboring cells by membranes, fibrous layers and adhesion proteins. Furthermore, 2D-cell monolayer cultures are generally exposed to a uniform environment whereas cells in tissues, such as solid tumors, are exposed to pH- and concentration gradients of soluble factors, nutrients and (toxic) metabolites. Animal experimentation on mice or rats cover all of these cellular requirements, but they are number-, time- and cost intensive and do not accurately represent a human being.
3D-cell culture is an alternative to these two techniques . Here we present the development of a single unit micro-bioreactor system for 3D-cultivation of cell lines, primary and stem cells which resembles the in vivo situation. The system (fig. 1) consists of a microfluidic bioreactor housing, a cell carrier substrate (MatriGrid) and a perfusion unit (internal or external).
Sensoric devices can be implemented into the system.
The micro-bioreactor housing is made of FDA-certified biocompatible poly-carbonate (PC). Its components are fully autoclavable and media/solvent resistant. The total volume of media in an assembled micro-bioreactor is 1,350 µl. Separate inflow and outflow channels, connected to the media chambers above and below the MatriGrid, are implemented for media exchange and probe extraction. In the upper part of the micro-bioreactor a media reservoir with an aeration port on top has been constructed. This serves as oxygen enrichment for the media and therefore optimal oxygen support for the cell culture.
A central element of our 3D cell culture equipment is the MatriGrid, which is a microstructured polymeric scaffold consisting of up to more than 200 microcavities which contain the cells and allow three-dimensional tissue-like aggregation. The active area of this cell carrier substrate is 5 mm x 5 mm on which the cells are seeded. Before seeding, the scaffold is treated with ECM factors, e. g. collagen. The MatriGrid is a further development and miniaturisation based on the CellChip-Technology of the Karlsruhe Institute of Technology (KIT) . With microstructuring processes such as hot embossing or micro thermoforming, the MatriGrids are produced.
Microcavities with dimensions of about 300 µm in diameter and 270 µm in depth were created by these techniques. The design and size of the micro cavities of the MatriGrid are adapted to the specific requirements of different cellular species. Applications for cell-lines, primary cells and stem cells, are possible. With 106 pores per cm2 and a diameter of 2 to 4 μm per pore, an optimal perfusion of the microcavities is achieved.
Media circulation and cell perfusion is achieved by active pumping in the bioreactor system. According to the active microstructured area of the MatriGrid, optimal pump rates for cell culture perfusion were found to be in the range of 25 µl/min. One option is to operate with external syringe or flexible tube pumps, which can be connected to the housing of the micro-bioreactor. The second variant is the usage of a commercially available micro-pump, driven by two piezo actuators. This micro-pump is nearly integrated in the micro-bioreactor with all required connections and fittings. (These two variants can be classified as external pumps.) For an internal pump, we are in process of developing a pneumatically driven type of micro-pump. The working principle of this miniaturized membrane pump is the deflection of the membrane, influenced by compressed air and vacuum which operates properly in the range of 10-1,000 µl/min.
The integration of sensors is within the construction principle. Novel AlGaN/GaN nanosensors are preferred for our applications based on the possibility to analyze reactions of cells attached to the sensor surface non-destructively and label-free . This class of sensors is highly sensitive and biocompatible to cells without any cytotoxic effect. The principle is the detection of ionic potential differences, which can be, for instance, a method of pH-measurement. Online detection of oxygen content by commercially available fiberoptical sensors ensures optimal monitoring of the cell culture and gives the opportunity to calculate oxygen consumption rates.
The main difference between 2D- and 3D-cell cultures is that 3D-grown cells show a more differentiated than proliferated state. To verify hepatic differentiation, FACS analysis in terms of cell cycle analysis and life/dead staining (CFDA/PI) can be carried out. A panel of different assays has been established to monitor metabolism and ADME/Tox parameters. Cell viability (AlamarBlue), apoptosis (caspase), albumin secretion, urea production and liver specific parameters e.g. EROD, UGT, can be determined. Immunohistochemical staining is used for visualization of differentiation markers, e.g. cd29 (cell adhesion) and cytokeratin (intracytoplasmic cytoskeleton of epithelial tissue). These probes can be analyzed in a 3D-resolution with a laser scanning microscope (LSM).
In comparison to 2D-culture we figured out that 3D-cultivated cells show a comparable or even increased secretion of albumin (as an indicator for differentiation) and a lower percentage of cells in G2/M-phase in the cell cycle (fig. 2).
For future experiments, a family of single unit micro-bioreactors for diverse and variable applications, according to the experimental setup desired, is planned. Furthermore, the development of a 24-well parallel system with 24 independent working micro-bioreactors is also in progress. Here we are adapting the standard 24-well microtiterplate footprint in our system to create a system compatible with throughput- and automated pipetting lines. Another approach is the construction of a pharmaceutical test platform, where a gradient dependent dosage of new pharmaceutical entity correlated with the 3D-cell culture mimics the in vivo situation of ADME parameters in tissue/organs. All these developments lead to more informative and representative data and reduce the need for animal experimentation in pharmaceutical research and development.
 Justice, B.A., et al.: Drug Discov Today 14, 102-107 (2009)
 Gottwald, E., et al.: Lab Chip 7, 777-785 (2007)
 Lübbers, B., et al.: physica status solidi (c) 6, 2361-2363 (2008)
Christian Hildmann, Uta Fernekorn, Jörg Hampl, Frank Weise, Maren Klett, Annette Läffert, Caroline Augspurger, Andreas Schober