Three-dimensional cell culture microarray for high-throughput studies of stem cell fate

Abstract

We have developed a novel three-dimensional (3D) cellular microarray platform to enable the rapid and efficient tracking of stem cell fate and quantification of specific stem cell markers. This platform consists of a miniaturized 3D cell culture array on a functionalized glass slide for spatially addressable high-throughput screening. A microarray spotter was used to deposit cells onto a modified glass surface to yield an array consisting of cells encapsulated in alginate gel spots with volumes as low as 60 nL. A method based on an immunofluorescence technique scaled down to function on a cellular microarray was also used to quantify specific cell marker protein levels in situ. Our results revealed that this platform is suitable for studying the expansion of mouse embryonic stem (ES) cells as they retain their pluripotent and undifferentiated state. We also examined neural commitment of mouse ES cells on the microarray and observed the generation of neuroectodermal precursor cells characterized by expression of the neural marker Sox-1, whose levels were also measured in situ using a GFP reporter system. In addition, the high-throughput capacity of the platform was tested using a dual-slide system that allowed rapid screening of the effects of tretinoin and fibroblast growth factor-4 (FGF-4) on the pluripotency of mouse ES cells. This high-throughput platform is a powerful new tool for investigating cellular mechanisms involved in stem cell expansion and differentiation and provides the basis for rapid identification of signals and conditions that can be used to direct cellular responses.

Figure 1

3D cellular microarray platform for mouse ES cell culture. A: Cell encapsulation on a functionalized glass slide. The cells are spotted on PS–MA modified glass slides using a microarray spotter. B: Light microscope image of a portion of the cellular microarray (60 nL spots) depicting the spot diameter and center-to-center distance between adjacent spots. C: Confocal microscopy images showing the three-dimensional distribution of mouse ES cells on day 0 (I), day 2 (II), and day 5 (III) of incubation. Shown in each panel is the top view and side view of the spot, from a z-stack obtained with 4 µm sections. Scale bar is 100 µm. Each spot is 30 nL resulting in a diameter of 560 µm and a height of 150 µm. D: Correlation of fluorescence intensity (RFU) of calcein AM staining integrated over the spot area with the number of viable cells present in the spot. [Color figure can be seen in the online version of this article, available at www.interscience.wiley.com.]

Figure 2

Mouse ES cell expansion in the 3D cellular microarray platform. A: Schematic representation of cell encapsulation on functionalized glass slides, followed by expansion in serum-free medium, and staining of viable cells in the spot microenvironment. B: Growth curves in terms of viable cell number of 46C mouse ES cells expanded in alginate gel spots. Cells were inoculated at 50,100,200, and 400 cells/spot on 60 nL spots in ESGRO^® complete medium (black diamonds) and DMEM/SR medium supplemented with LIF (white triangles). Insets show a fluorescence-scanning image of representative alginate spots (800 µm in diameter) after expansion for 5 days at each seeding density used. Calcein AM was used to detect viable cells through green fluorescence. C: Comparison between the four initial cell densities tested. Cell expansion in terms of fold increase in total cell number and maximum specific growth rates were calculated for cells inoculated at 50,100,200, and 400 cells/spot on 60 nL spots in ESGRO^® complete medium (black bars) and DMEM/SR medium supplemented with LIF (white bars). All error bars show the standard error of the mean (n = 3 slides, 144 spots in total for each data point). [Color figure can be seen in the online version of this article, available at www.interscience.wiley.com.]

Figure 3

Mouse ES cells retain their undifferentiated state after expansion in the alginate spots. A: Mouse ES cells express high levels of Oct-4 and Nanog prior to spotting in the microarray, as can be seen by flow cytometry. B: Following expansion for 5 days in the alginate spots in serum-free medium containing LIF, mouse ES cells stain positively for Oct-4. Cells were inoculated at 100 cells/spot on 60 nL spots in DMEM/SR medium supplemented with LIF (I: bright field; II: fluorescence; scale bar: 100 µm). C: Histogram plots showing the distribution of Oct-4 and Nanog (normalized for β-actin) signals for cells expanded during 5 days in serum-free medium in the presence of LIF (+LIF) or absence of LIF (−LIF). D: Quantification of Oct-4 and Nanog using the in-cell, on-chip immunostaining method. Cells were inoculated at 50 or 100 cells/spot on 60 nL spots and were expanded in serum-free medium in the presence of LIF (+LIF) or absence of LIF (−LIF). Oct-4 and Nanog signals were normalized for β-actin and are relative to +LIF conditions. [Color figure can be seen in the online version of this article, available at www.interscience.wiley.com.]

Figure 4

Neural commitment of mouse ES cells on the cellular microarray. A: Neural committed cells express high levels of Sox1, a marker of primitive neuroectoderm, as can be seen by flow cytometry after culture on tissue culture plates. B: After incubation for 6 days in the alginate spots in RHB-A serum-free medium, mouse ES cells differentiate into Sox1-GFP⁺ neural precursor cells. The image shows a magnification of the center of a cell-containing alginate spot. Cells were inoculated at 400 cells/spot on 60 nL spots (I: bright field; II: fluorescence; scale bar: 100 µm). C: Histogram plot showing the distribution of Sox1-GFP (normalized for β-actin) signal for cells cultured for 6 days in RHB-A serum-free medium. D: Quantification of Oct-4, Nanog, and Sox1-GFP for cells expanded in serum-free medium in the presence of LIF (DMEM/SR + LIF) and cells cultured for 6 days in RHB-A serum-free medium. Oct-4 and Nanog signals were normalized for β-actin and are relative to cells expanded in DMEM/SR + LIF. Sox1-GFP signals were normalized for β-actin and are relative to cells cultured in RHB-A medium. [Color figure can be seen in the online version of this article, available at www.interscience.wiley.com.]

Figure 5

Cellular microarray platform is compatible with high-throughput studies. A: Schematics depicting the dual slide incubation procedure. FGF-4 (2 ng/mL) and retinoic acid (10µM)were printed onto MTMOS-functionalized glass slides and stamped on top of alginate spots containing mouse ES cells. LIF was used as positive control. The dual slide system was incubated overnight and cells were incubated for additional 3 days before analysis. B: Percentage of viable cells after stamping using a dual slide system and incubation for specific times at 37°C in a humidified chamber. After stamping, the cellular microarray was incubated in expansion medium (ESGRO^® complete) for 3 days, and finally stained using the calcein AM viability dye. The results are normalized for the levels of the control (no stamping) cells. C: Quantification of Oct-4 and Nanog levels on the 3D cellular microarray after stamping with FGF-4 (2 ng/mL) and retinoic acid (10µM), or LIF, and subsequent incubation in LIF-containing medium for 3 days. The results are normalized for the levels obtained with LIF. β-Actin was used as the internal control.