Real-time imaging of adherent and non-adherent cell interactions: utility of an automated microfluidic trap platform to recapitulate in vivo cell culture microenvironment

Rekha Kannan, Victor Yeh, James Helton, Amedeo Cappione, Jun Park* MilliporeSigma, Research Solutions, Danvers, MA & Hayward, CA, USA

The study of dynamic cell processes and their interactions is of crucial importance to understand the complexities of tissue microenvironments. Simulation of in vivo cell culture microenvironments can greatly improve biologic relevance of cell-cell interaction studies, and so there is an ongoing demand for heuristic optimization of cell co-culture devices and protocols. One particularly challenging task is establishing real-time image analysis while co-culturing non-adherent cells in contact with adherent cell layers under precisely controlled environmental parameters. To address this challenge, (A) we present novel microfluidic approaches to trap two different cell types in precise locations in order to image and study their interactions over time. The platform utilizes a well plate format that contains multiple microfluidic units. Each unit consists of a 6 mm x 3 mm x 40 µm (L x W x H) chamber for culture, and an array of microscopic trap areas tailored to the dimensions of the cells of interest, typically ~15 µm. (B) The micro-scale geometries of the traps physically confine non-adherent cells, and co-cultures can be achieved by trapping the non-adherent cells in close proximity of adherent cells. The microfluidic plate is integrated with a system which enables (C) perfusion-based nutrient supply along with gas and temperature control for long-term cell culture. Each chamber within the plate can be addressed by (D) programmable and on-demand perturbation of up to 6 reagents, enabling uninterrupted real-time live cell imaging assays. We present here the use of this microfluidic platform to replicate the tumor microenvironment by maintaining adjacent co-cultures of cancer and immune cells. We (E) have successfully imaged and cultured monolayers of tumor cell lines for 6 days followed by subsequent loading and trapping of immune (non-adherent) cells. In conclusion, the microfluidic platform enables unique co-cultures with environmental control and real-time imaging to facilitate the investigation of cell-cell interactions in a wide range of applications such as drug response and screening in tumor microenvironments, invasion, evasion pathways and other mechanisms governing cell-cell interactions.

Introduction

Figure 1. CellASIC® ONIX2 live cell imaging platform and associated microfluidics plates. A) Image shows plate sealed with manifold and mounted on inverted microscope stage of Lionheart™ FX automated microscope (BioTek®). Microfluidics culture chamber within the microfluidics plate is indicated in blue. Also indicated are first 6 inlet wells from which various media can be added and then perfused into the culture chamber, as well as 2 outlet wells (wells 7 and 8) either used for cells loading or collection of waste. The environmental controller provides pressure driven control of liquid as well as delivery of gas and temperature regulations. B) Schematic showing the features of manifold with sealed plate (upper panel) and corresponding cross sectional image (lower panel). The manifold lines allow pressure driven flow control, as well as delivery of premixed gas to the incubation chamber. Plates with various chamber designs are currently available commercially.

Figure 2. CellASIC® mammalian trap plates for live imaging during co-culturing. The overall design features of "trap array" frame within the culture chamber designed to trap non-adherent suspension cells in close proximity to underlying adherent cells (lower panel). Due to the physical dimensions of the traps, suspension cells contacting or coming in close proximity to adherent cells can be viewed within focal plane during live cell imaging facilitating more detailed imaging of cellular interactions and micro-environment particularly during long term culture (upper panel). Prototype plates having two different types of traps (open and closed) were investigated in this reports (see figure 4).

Figure 3. Live MDA-MB-231 cell imaging under static condition vs. dynamic physiological conditions. A) Cells cultured under static room temperature condition show very little cellular movement and noticeable loss of cell attachment even during 3 hrs incubation (arrows). Cells culture under two different 37'C dynamic conditions by varying the flow rate of culture media from 2.5 psi (B) to gravity flow (C). Under 2.5 psi, cells exhibit highly motile behavior characterized by triangular morphology with rapidly advancing leading edges. Under gravity flow (<0.1psi), motility of the cells are significantly reduced and their shape transition into more of spread elliptical morphology commonly associated with MDA-MB-231 cells. All images taken from Lionheart™ FX. It will be interesting to compare these morphological features with invasiveness of different tumors.

Figure 4. Live cell co-culture imaging of MDA-MB- 231 and CD4 T cells. MDA cells were seeded into culture chamber from outlet well 8, and after removing remaining cells from the outlet were allowed to adhere for 1hr. Culture chamber was perfused with complete culture media from inlet wells (2.5psi, 8 hrs) every two days with fresh media and waste removal. At day 5, MDAs were labeled with CellTracker TM Red (LifeTech) from well 8, and cultured for additional day after removing the excess dye. At day 6, purified human CD4 T cells (Lonza) labeled with CellTracker TM Green were loaded, followed 1hr later by anti-CD3 coated particles (5 µm, Spherotech). Co-cultures were now perfused for 2 minutes at 0.5 psi with fresh media after removing the excess T cells and beads and imaged for 4 days with intermittent perfusion twice a day. A) and B) Live cell imaging of T cell/MDA co-culture using "open" traps. While the in focusing advantage of the trap is clearly visible (arrows), open traps enabled T cells to escape, rendering the visual study of the microenvironment as function of time difficult. C) "Closed" trap design enabled long term maintenance of localized micro environment (arrow) with up to 4 days of co-culture at present report. Also shown are regular 96 well co-culture of the same system for visual comparison (D). Notice how perfusion greatly remove non interacting T cells, enabling clear visualization within µfluidics culture chamber.

Figure 5. Live cell images from 3 day co-culture of MDA-MB-231 and CD4 T cells. Image of the trap microenvironment and surrounding area taken from the same location shown in previous figure (4C). Images were originally taken at 10 min intervals with environmental control under ONIX2. Notice fluorescence has greatly diminished compared to the original image shown. A) Closed trap still maintained the microenvironment composed of T cells and MDAs. B) Examples of time lapse live imaging from the same area showing the efficiency of closed trap in maintaining T cells trapped while keeping the them in focus. For comparison, arrow indicates population of un-trapped activated T cells clustering around the CD3 beads. Since these T cell population are from outside of the original imaging area, they still maintained fluorescence. However, notice how they move considerably within the imaging field of view over time and are extremely difficult to focus during live imaging. These population of T cells eventually exited the filed of view completely at later time points.

Conclusions

  • Real time live cell imaging with environmental control offers unique platform through which much deeper insights into cellular dynamics can be gained.
  • Current closed trap design offers potential to create and maintain complex cellular micro-environment and enables real time imaging while maintaining image focus of the micro-environment over time.
  • Future design iterations for better trap is currently underway.

*Corresponding Authors: jun.park@emdmillipore.com

     
Related Links