A Macroscopic Diffusion-Based Gradient Generator to Establish Concentration Gradients of Soluble Molecules Within Hydrogel Scaffolds for Cell Culture

Abstract

Concentration gradients of soluble molecules are ubiquitous within the living body and known to govern a number of key biological processes. This has motivated the development of numerous in vitro gradient-generators allowing researchers to study cellular response in a precise, controlled environment. Despite this, there remains a current paucity of simplistic, convenient devices capable of generating biologically relevant concentration gradients for cell culture assays. Here, we present the design and fabrication of a compartmentalized polydimethylsiloxane diffusion-based gradient generator capable of sustaining concentration gradients of soluble molecules within thick (5 mm) and thin (2 mm) agarose and agarose-collagen co-gel matrices. The presence of collagen within the agarose-collagen co-gel increased the mechanical properties of the gel. Our model molecules sodium fluorescein (376 Da) and FITC-Dextran (10 kDa) quickly established a concentration gradient that was maintained out to 96 h, with 24 hourly replenishment of the source and sink reservoirs. FITC-Dextran (40 kDa) took longer to establish in all hydrogel setups. The steepness of gradients generated are within appropriate range to elicit response in certain cell types. The compatibility of our platform with cell culture was demonstrated using a LIVE/DEAD® assay on terminally differentiated SH-SY5Y neurons. We believe this device presents as a convenient and useful tool that can be easily adopted for study of cellular response in gradient-based assays.

Keywords: agarose; biopolymer; chemotaxis; collagen; neuron.

Figure 1

Design of the diffusion-based gradient generator (A) The diffusion chamber inverse master mold is designed in AutoCAD software and 3D-printed to the specified dimensions; all values are expressed as millimeters. (B) PDMS is cast into the mold and the chambers attached to a glass microscope slide. The device is placed in a petri dish prior to placing in the incubator. (C) Schematic illustration depicting the principle behind gradient formation in the diffusion-based gradient generator. A center chamber containing the hydrogel separates outer source and sink wells connected by a small 0.25 mm high gap beneath the dividing walls. The source solution diffuses through the hydrogel into the sink chamber, establishing a concentration gradient across the hydrogel. Replenishing the source and sink solutions every 24 h maintains this concentration difference between the two compartments. The glass coverslip “lid” is placed on top of the chambers to minimize evaporation. (D) Acquired fluorescent image of concentration gradient across hydrogel chamber visualized using FITC-Dextran (40 kDa) at 96 h. White rectangle illustrates the region where the gradient profiles are obtained.

Figure 2

Rheological behavior of 0.2% collagen (Col), 1% agarose (AG) and 0.03% collagen+1% agarose co-gel (AG-Col) determined through parallel-plate rheometry. (A) Storage (shaded symbols) and loss (open symbols) moduli represent the elastic and viscous properties of the material. For all hydrogels the storage modulus exceeds the loss, indicating that the hydrogel is present in its gel state. (B) The complex modulus is used to represent the stiffness of the hydrogels. A statistically significant difference is observed between all groups, yet the stiffness for all remains within appropriate range for neuronal culture. Each data point represents the mean (n = 3) ± SD. (**p < 0.01 and ***p < 0.001).

Figure 3

FTIR Spectra of (A) AG-Col, AG, Col (B) AG, AG + NaFl and AG + FITC-Dextran 10 kDa, and (C) AG-Col, AG-Col + NaFl and AG-Col + FITC-Dextran 10 kDa. Shaded “fingerprint” areas represent regions where changes and potential interactions are visible between the different spectra.

Figure 4

Concentration gradient profiles of NaFl (376 Da), FITC-Dextran (10 kDa) and FITC-Dextran (40 kDa) measured at various time points after setup. Concentration profile was determined across the central hydrogel chamber where 0–5,000 μm corresponds with the central 2,500–7,500 μm region of the hydrogel. The outer 2,500 μm from either end was excluded to avoid interference by the meniscus. Each data point represents the mean ± SD error bands.

Figure 5

Steepness of the concentration gradient (ng mL⁻¹ mm⁻¹) as determined between the linear 2,500–7,500 μm region of the hydrogel chamber where cells will be cultured. Each data point represents the mean (number of replicate devices: n = 3; where indicated by “--” n = 2) ± SD.

Figure 6

The biocompatibility of the gradient-generator and hydrogels used was investigated using a LIVE/DEAD® viability/cytotoxicity assay on SH-SY5Y neurons after 96 h in culture. (A) LIVE/DEAD® assay fluorescence and brightfield images visualized using a 10× objective. (B) The viability of the neurons was calculated by determining the number of live cells (dual labeled with Hoechst nuclear stain and calcein) as a percentage of the total number of cells. Cells grown on polystyrene and glass were used as the positive control. Each bar represents the mean (two separate regions, three independent replicates n = 6) ± SD. (C) Morphology of neurons cultured on Matrigel-coated polystyrene (20× objective). (D) Morphology of neurons cultured on the Matrigel-coated gradient-generator overlaid with thin agarose hydrogel (20× objective).