2Institute of Biomaterials and Biomedical Engineering, University of Toronto, Toronto, Canada
##S.J. and K.J.L. contributed equally to this work
In native tissues, different cell types are organized into defined structures and architectures that are critical for correct tissue function. In vitro cellular patterning methods enable control over the spatial organization of cells, permitting, to some extent, the reproduction of native tissue structures and the generation of a more “in vivo–like” culture platform. While this is advantageous for applications such as drug screening, existing patterning methods are time-consuming, labor-intensive, and low-throughput. Here, we describe a novel medium-throughput patterning strategy for generating spatially controlled co-cultures of two cell types based on differential deposition of BSA solution in a tilted plate. Our method allows generation of homotypic and heterotypic co-cultures that are stable for at least seven days in culture. The reproducibility and consistency of this patterning technique, together with its low cost and ease of use, make it a promising cell culture platform for medium- to high–throughput screening using high-content imaging.
Appropriate cellular organization is vital for healthy tissue function. Consequently, structural disruption within tissues is a hallmark of many diseases (1, 2). Patterned co-cultures of cells can be used to study interactions between different cell types and the effect of cell organization on tissue function in vitro. Cellular patterning is possible using a set of techniques that control cell localization on a culture substrate in vitro (3-5). In patterned co-cultures, the localization of different cell types is controlled, allowing for the generation of “tissue” with a pre-defined composition and architecture. There are typically two approaches to generating co-cultures: (i) micropatterning, which allows direct cell-cell contact, as occurs in vivo, and (ii) co-culture in a multi-compartment system, such as a transwell plate, that physically separates the two cell populations. While the latter approach is easier, the interaction between different cell types is limited to diffusible signals. Direct cell-to-cell contact produces significantly different cell behaviors from those observed in multi-compartment co-culture (6-8). For this reason, methods to produce direct contact co-cultures may provide more relevant in vitro models for understanding cell and tissue function. Such in vitro models could be particularly useful for screening applications. Indeed, integration of direct co-culture patterning techniques into current high-throughput drug screening protocols is predicted to increase the success rate of drug discovery by promoting a more in vivo–like cell culture system (9). More specifically, patterned co-cultures can be valuable systems for screening changes in cell-cell interactions. For example, screens of patterned co-cultures could be used to search for compounds that inhibit infiltration of tumor cells into sheets of other cell types. Another example is in regenerative therapy, where infiltration of therapeutic cells into damaged tissue is necessary for regeneration (10). Despite the promise of direct contact co-culture, there are currently no medium- to high–throughput methods for generating patterned 2-D co-cultures.
Our method allows the generation of simple, spatially controlled co-cultures of two cell types based on differential deposition of BSA in a tilted multi-well plate. The procedure is simple, allows medium- to high–throughput pattern generation, is compatible with high-content microscopy, and allows patterning of homotypic and heterotypic co-cultures that are stable for at least seven days.
Most current co-culture patterning techniques involve the use of manual stamps (11, 12) or stencils (13, 14) to prevent cell attachment to certain locations on the substrate. Significant optimization is required to generate reliable and consistent patterning using these methods. Several attempts have been made to adapt existing patterning strategies for higher throughput. For example, microstamping techniques allow stamping of multiple well surfaces in parallel (15). While this method could potentially be adapted to generate patterned co-cultures, in its current form it is only suitable for generation of patterned mono-cultures. Moreover, this method is difficult to optimize and currently only allows 6 to 12 wells in a 96-well plate to be patterned in parallel. An alternative is high-throughput patterning using microarray spotting technologies (16, 17). This approach allows high control and precision, but is labor-intensive and requires specialized equipment. Moreover, the use of microarray spotting for patterning direct contact cellular co-cultures remains to be demonstrated. Microfluidic methods are also available, but these (18, 19) are often complex to fabricate, can compromise cell viability during extended culture, and are not easily compatible with high-throughput screening methods such as high-content microscopy.
Here we describe a simple method for generating patterned co-cultures with predictable cell organization that is suitable for medium-to high–throughput applications. Our method allows limited spatial control over the localization of two cell types in multi-well plates and is compatible with high-content screening platforms. The ratio between cell types in each well can be varied on demand, which is useful for probing the effect of diffusible signals in combination with direct cell-cell contact. Cell patterning is controlled by sequential seeding of two cell populations onto substrates using selective deposition of cell-repellent protein (BSA). This is achieved by placing a set volume of BSA solution in the well while tilting the plate at a specific angle. We demonstrate the versatility of our co-culture method by generating co-cultures of various cell combinations in wells of different sizes. Further, we characterize the reproducibility and consistency of our patterning technology in order to confirm its compatibility with automated high-content imaging. Materials and methods Cell culture
We used the ARPE-19 human retinal epithelial cell line, the MDCK dog kidney epithelial cell line, the NIH 3T3 and BJ (ATTC, Manassas, VA) fibroblast cell lines at passages P10–P20, and human umbilical vein endothelial cells (HUVECs) (Lonza, Basel, Switzerland) at passages P4–P8 grown in medium recommended by the supplier. Populations of eGFP (green fluorescing) or mCherry (red fluorescing) overexpressing cells were prepared by lentivirus transduction. The eGFP lentiviral construct was a gift from J Axelrod at Stanford and we cloned in the mCherry gene (Clontech, Mountain View, CA, USA) into the backbone to generate the mCherry lentiviral construct. Plate pre-patterning with BSA
To generate pre-patterned plates with differential BSA deposition, we added 200 μL of 0.5% w/v BSA (Sigma Aldrich, Oakville, Ontario, Canada) to cover only the bottom half of each well in a 24-well tissue-culture treated plate (BD Biosciences, Mississauga, Ontario, Canada). The plate was tilted using a stand and incubated for 2 h at 37°C and 5% CO2. For 6-, 12-, 48-, and 96-well plates, 800, 400, 100, and 50 μl BSA was added to each well respectively. After incubation, the BSA solution was removed carefully and the plate was air-dried to ensure that half the well remained BSA-free. Cell patterning
The first cell type to be seeded was suspended in serum-free medium (DMEM-F12; Gibco, Gaithersburg, MD), seeded on the plate, and incubated. The seeding densities, volumes, and incubation times for different well sizes are listed in Table 1. After gently removing excess cells, we covered the entire well with 500 μL of 10 μg/mL fibronectin (Biomedical Techonologies Inc., Stoughton, MA) and incubated the plate for 30 min. We suspended the second cell type in complete medium at a density of 2 × 106 cells/mL. The fibronectin solution was removed and 500 μL (for a 24-well plate) of the second cell type suspension was added. The plate was incubated for 3–4 h until the cells were attached before washing off unadhered cells to reveal a co-culture with a robust interface.