In this work a microfluidic cell cultivation device for perfused hypoxia assays as well as a suitable controlling unit are presented. The device features active components like pumps for fluid actuation and valves for fluid direction as well as an oxygenator element to ensure a sufficient oxygen transfer. It consists of several individually structured layers which can be tailored specifically to the intended purpose. Because of its clearness, its mechanical strength and chemical resistance as well as its well-known biocompatibility polycarbonate was chosen to form the fluidic layers by thermal diffusion bonding. Several oxygen sensing spots are integrated into the device and monitored with fluorescence lifetime detection. Furthermore an oxygen regulator module is implemented into the controlling unit which is able to mix different process gases to achieve a controlled oxygenation. First experiments show that oxygenation/deoxygenation of the system is completed within several minutes when pure nitrogen or air is applied to the oxygenator. Lastly the oxygen input by the pneumatically driven micro pump was quantified by measuring the oxygen content before and after the oxygenator.
Because of its direct and indirect participation in many metabolic processes in cells, tissues and organisms oxygen is one of the most important parameters. Especially in mammalian tissue the adequate supply of oxygen is an essential prerequisite influencing cell viability, proliferation and differentiation . Oxygen shortage is called hypoxia and leads to changes in cellular behaviour. Among others it induces the release of hypoxia inducible factors (HIF)  and vascular endothelial growth factor (VEGF) which affect cell metabolism, wound healing , angiogenesis  and tumor metastasis . Therefore a deeper understanding of cell response on oxygen limitation is of particular interest in medical basic research. For this issue lab-on-a-chip (LOC) devices have become more and more important for in-vitro cell cultivation . Especially perfused LOCs are able to mimic physiological environment inducing mechanical stimulation and convective mass transfer. Currently most LOC systems are externally perfused. In contrast to that the Fraunhofer IWS developed a multilayer based LOC with on chip actuation . This offers the possibility to create micro circulation systems with a high tissue to cell culture media ratio (volumes in the range of several microliters). Moreover the effect of dilution which appears in flow through LOC is negligible. The mentioned micro circulation approach is superior for the examination of metabolites and their effect on several tissue types. Due to its good optical properties established imaging technologies can be applied throughout the complete microfluidic system. The on-chip micro pump was characterized using micro-Particle-Image-Velocimetry and a mathematic model was developed . In the present study an additional oxygenator element is implemented into the IWS microfluidic platform, allowing automated and reproducible hypoxia assays on the chip. In Figure 1 the principle of the perfusion controlled hypoxia is shown.
One of the main benefits of the presented microfluidic system is the possibility to integrate microfluidic actuators to create completely closed and perfused microfluidic circuits and microvascular systems with small volumes in the range of several μL. External pumping devices are not needed as it is mostly the case in microfluidic cell cultivation devices presented in literature . The microfluidic system is produced with a layer-by-layer manufacturing technology of laser-cut polymer foils . The mentioned fluidic actuators are pneumatically driven systems . Therefore a flexible membrane has to be integrated into the microfluidic system, which can be easily displaced with pressure rates up to 100 kPa. A commercially available silicone foil (SILPURAN® FILM, Wacker Chemie AG) with a thickness of 200 microns and high gas permeability is perfectly suited for this application. In Figure 2 an exploded view of the microfluidic system with mounted reservoirs on top, the pneumatic part, the elastomeric membrane in the middle and the fluidic part at the bottom is shown.
The complete system is mounted in a chipholder with two connection elements which include the pneumatic fittings. In Figure 3 the developed microfluidic layout is shown. It features a micro pump for fluid actuation and an in- and outlet valve to operate the system as totally closed microfluidic circuit. In previous fluidic designs the oxygen exchange was solely performed by the gas-permeable elastomeric pump membrane. So the pump acted as a combined flow source and oxygenator . This may be problematic especially when gas bubbles are produced, as reported by Goldowsky and Knapp . To enhance the oxygen transport capability an additional helical oxygenator element was placed at the center of the microfluidic system utilizing the same gas-permeable membrane also used for fluid actuation whereby the gas exchange area of this oxygenator is ten times larger than that of the pump. The process gas flow at the top of the membrane can be adjusted to avoid oxygen depletion over the length of the oxygenator. Three oxygen sensing chambers (A, B, C, 3 mm diameter) are implemented to characterize both the oxygenator and the cell culture segment (oxygen consumer).
To control the oxygen content in the cell culture segment a controlling unit based on an embedded Linux device is used. Figure 4 shows a block diagram of the system.
It switches up to 24 pneumatic outputs to actuate the pumps and valves on the chips. Furthermore it is capable to ensure stable temperature conditions and can mix three different process gases (oxygen O2, nitrogen N2 and carbon dioxide CO2) to control the oxygenator. Several digital interfaces (I2C, ETHERNET, CAN …) give the opportunity to integrate the controller in other laboratory infrastructure or laboratory information management systems.
In presence of oxygen, the fluorescence lifetime of several fluorescent dyes is decreased due to quenching . This decay in the fluorescence lifetime can be utilized to measure the oxygen content in the microfluidic system as shown in Figure 5 . The experimental setup was described earlier in detail . Commercial available CPOx-beads (Colibri photonics, Potsdam) are immobilized with Polydimethyl-siloxane (PDMS) in each sensing chamber (A, B, C) and the fluorescence signal was measured using a compact optic block coupled to a photomultiplier. The OPAL digital lock-in amplifier (Colibri photonics, Potsdam) calculates the fluorescence decay signal and communicates to a host-PC via USB. Each measurement spot can be automatically evaluated using a 3-axis positioning system (Nanoplotter 2.1, GeSiM mbH, Großerkmannsdorf) controlled by the measurement software.
For calibration purpose the microfluidic system was flushed with three different process gases with varying oxygen contents wO2 compared to the volume fraction of oxygen in atmosphere under standard conditions (ϕO2 = 20.9 vol.%). Those gases were: Air (wO2 = 100 %), reference gas (wO2 = 47 %) and nitrogen (wO2 = 0 %). Afterwards the microfluidic system was filled with DI water for the oxygenation and deoxygenation experiments.
First of all the oxygenation coefficient of the oxygenator elements has to be measured:
Whereby KO2 is the oxygenation coefficient, cB is the oxygen concentration at the oxygenator outlet, α is the oxygen solubility in water and pO2 is the oxygen partial pressure applied to the oxygenator. This oxygenation coefficient can be determined by measuring the step response of the system when the process gas is changed. Therefore all channels were firstly flushed with DI water and afterwards pure nitrogen was applied to the oxygenator and the oxygen content at the outlet of the oxygenator was measured until it reaches a lower boundary. Afterwards the process gas was changed to air until the oxygen content of wO2 = 100 % was reached again. Figure 6 shows the up- and downstream oxygen contents (of the oxygenator) with a pumping speed of 1.2 Hz, a pumping pressure of 500 mbar and a filling vacuum of -500 mbar and compressed air as pump gas.
As mentioned earlier the pump is gas permeable too. Therefore a parasitic oxygen input Kp can be observed which influences the transient oxygenation/deoxygenation behaviour. Both parameters can be calculated with the measured stationary up- and downstream oxygen contents (wA, wB). For the deoxygenation process and the pump operated with compressed air they are:
This leads to an oxygenation coefficient of KO2 = 0.4 for the oxygenator and Kp = 0.25 for the pump. Knowing those parameters gives the opportunity to develop an adapted oxygen regulator simply by varying the process gas.
A microfluidic cell cultivation device with integrated oxygenator and micro pump is presented which can be operated with different process gases to generate well-defined hypoxic conditions in one or more cell culture segments. Both, the pneumatic micro pump as well as the oxygenator is operated by a controlling unit. Due to the clearness of the used polymer foils (in the visible wavelength range) which were joined together to form the microfluidic system established imaging technologies can be applied everywhere at the microfluidic system. Therefore the impact of hypoxia can be studied online in perfused 3D cell culture models. Possible applications could be the observation of metastatic spread or wound healing processes under hypoxic conditions. Commercial available oxygen sensitive particles are immobilized at different measurement spots to characterize the oxygenation coefficients of the micro pump and the oxygenator. Firstly the oxygenator was operated with pure nitrogen as process gas resulting in a media deoxygenation. Afterwards the process gas was switched to compressed air which induces the oxygenation of the pumped fluid. Subsequent to the characterisation of the oxygenator an oxygen regulator should be implemented in the controlling unit allowing the precise adjustment and regulation of oxygen levels even for different oxygen consumers. Moreover the implementation of the previously developed substance transport model of the microfluidic system in the controller is possible, allowing a “model-in-the-loop” based regulation of the system .
Research funding: The authors want to express great appreciation to the Free State of Saxony and the European Union (SAB project “UNILOC”) as well as to the BMWi (ZIM project “Plasmafügen”) for the financial support. Conflict of interest: Authors state no conflict of interest. Material and Methods: Informed consent: Informed consent is not applicable. Ethical approval: The conducted research is not related to either human or animal use.
 Hung S, Ho JH, Shih YV, Lo T, Lee OK. Hypoxia promotes proliferation and osteogenic differentiation potentials of human mesenchymal stem cells. J Orthop Res. 2012;30:260–6. Search in Google Scholar
 Richter S, Qin N, Pacak K, Eisenhofer G. Role of hypoxia and HIF2 α in development of the sympathoadrenal cell lineage and chromaffin cell tumors with distinct catecholamine phenotypic features. Adv Pharmacol. 2013;68:285–317. Search in Google Scholar
 Krock BL, Skuli N, Simon MC. Hypoxia-induced angiogenesis: good and evil. Genes Cancer. 2012;2:1117–33. Search in Google Scholar
 Eltzschig HK, Eckle T. Ischemia and reperfusion–from mechanism to translation. Nat Med. 2011;17:1391–401. Search in Google Scholar
 Hockel M, Vaupel P. Tumor Hypoxia: definitions and current clinical, biologic, and molecular aspects. J Natl Cancer Inst. 2001;93:266–76. Search in Google Scholar
 Baker M. Tissue models: a living system on a chip. Nature. 2011;471:661–5. Search in Google Scholar
 Klotzbach U, Sonntag F, Grünzner S, Busek M, Schmieder F, Franke V. Multilayer-based lab-on-a-chip systems for perfused cell-based assays. Advanced Optical Technologies 2014;3:515–21. Search in Google Scholar
 Busek M, Nötzel M, Polk C, Sonntag F. Characterization and simulation of peristaltic micropumps. J. Sens. Sens. Syst. 2013; 2: 165–169. Search in Google Scholar
 Bhatia SN, Ingber DE. Microfluidic organs-on-chips. Nat Biotechnol. 2014;32:760–72. Search in Google Scholar
 Sonntag F, Grünzner S, Schmieder F, Busek M, Klotzbach U, Franke V. Multilayer based lab-on-a-chip-systems for substance testing. In: Klotzbach U, Washio K, Arnold CB, editors. SPIE LASE: SPIE; 2015. p. 93510C (SPIE Proceedings). Search in Google Scholar
 Huang C, Huang S, Lee G. Pneumatic micropumps with serially connected actuation chambers. J Micromech Microeng. 2006;16:2265–72. Search in Google Scholar
 Busek M, Rudolph A, Grünzner S, Schmieder F, Sonntag F, Hofmann K, et al. Design and regulation of complex microfluidic systems with simulationX. 18th ITI Symposium 2015. Search in Google Scholar
 Goldowsky J, Knapp HF. Gas penetration through pneumatically driven PDMS micro valves. RSC Adv. 2013;3:17968. Search in Google Scholar
 Schmälzlin E, van Dongen JT, Klimant I, Marmodée B, Steup M, Fisahn J, et al. An optical multifrequency phase-modulation method using microbeads for measuring intracellular oxygen concentrations in plants. Biophys J. 2005;89:1339–45. Search in Google Scholar
 Schmieder F, Grünzner S, Winkelmann C, Sonntag F. Hollow fiber-based Lab-on-a-chip perfusion system with integrated fluorescence-based oxygen monitoring. 2nd International Conference on Microfluidic Handling 2014. Search in Google Scholar
 Busek M, Hofmann K, Grätz U, Rudolph A, Grünzner S, Schmieder F, et al. Regulating microfluidic-based cell culture systems with the use of network models. Bionection. 2015;2015. Search in Google Scholar