Microfluidic technologies have become a standard tool in cell biological studies, offering unprecedented control of the chemical and physical environment of cells grown in microdevices, the possibility of multiplexing assays, as well as the capacity to monitor the behavior of single cells in real time while dynamically manipulating their growth medium. However, the properties of the materials employed for the fabrication of microchips that are compatible with live-cell imaging has limited the use of these techniques for a broad range of experiments. In particular, the strong absorption of a large panel of small molecules by these materials prevents the accurate delivery of compounds of interest. Here we describe a novel microsystem dedicated to live-cell imaging that (1) uses alternative materials devoid of absorptive properties, and (2) allows for dynamic in-chip control of sample temperature. Based on a proof-of-concept design that we have routinely used with non-adherent fission yeast cells, this chapter details all the steps for the fabrication and utilization of these microdevices.

BACKGROUND: Real-time monitoring of cellular responses to dynamic changes in their environment or to specific treatments has become central to cell biology. However, when coupled to live-cell imaging, such strategies are difficult to implement with precision and high time resolution, and the simultaneous alteration of multiple parameters is a major challenge. Recently, microfluidics has provided powerful solutions for such analyses, bringing an unprecedented level of control over the conditions and the medium in which cells under microscopic observation are grown. However, such technologies have remained under-exploited, largely as a result of the complexity associated with microfabrication procedures. RESULTS: In this study, we have developed simple but powerful microfluidic devices dedicated to live-cell imaging. These microsystems take advantage of a robust elastomer that is readily available to researchers and that presents excellent bonding properties, in particular to microscopy-grade glass coverslips. Importantly, the chips are easy-to-build without sophisticated equipment, and they are compatible with the integration of complex, customized fluidic networks as well as with the multiplexing of independent assays on a single device. We show that the chips are re-usable, a significant advantage for the popularization of microfluidics in cell biology. Moreover, we demonstrate that they allow for the dynamic, accurate and simultaneous control of multiple parameters of the cellular environment. CONCLUSIONS: While they do not possess all the features of the microdevices that are built using complex and costly procedures, the simplicity and versatility of the chips that we have developed make them an attractive alternative for a range of applications. The emergence of such devices, which can be fabricated and used by any laboratory, will provide the possibility for a larger number of research teams to take full advantage of these new methods for investigating cell biology.

Maintenance of long-term cultures of yeast cells is central to a broad range of investigations, from metabolic studies to laboratory evolution assays. However, repeated dilutions of batch cultures lead to variations in medium composition, with implications for cell physiology. In Saccharomyces cerevisiae, powerful miniaturized chemostat setups, or ministat arrays, have been shown to allow for constant dilution of multiple independent cultures. Here we set out to adapt these arrays for continuous culture of a morphologically and physiologically distinct yeast, the fission yeast Schizosaccharomyces pombe, with the goal of maintaining constant population density over time. First, we demonstrated that the original ministats are incompatible with growing fission yeast for more than a few generations, prompting us to modify different aspects of the system design. Next, we identified critical parameters for sustaining unbiased vegetative growth in these conditions. This requires deletion of the gsf2 flocculin-encoding gene, along with addition of galactose to the medium and lowering of the culture temperature. Importantly, we improved the flexibility of the ministats by developing a piezo-pump module for the independent regulation of the dilution rate of each culture. This made it possible to easily grow strains that have different generation times in the same assay. Our system therefore allows for maintaining multiple fission yeast cultures in exponential growth, adapting the dilution of each culture over time to keep constant population density for hundreds of generations. These multiplex culture systems open the door to a new range of long-term experiments using this model organism.