Microfluidic control over millimeter scale alignment of collagen fibers

ByAdeel Ahmed
Type I collagen is a fibrous protein and the largest constituent of the ECM by mass. Collagen fibers are aligned across several millimeters to centimeters within load-bearing tissues, like the musculoskeletal tissue, cardiac tissue, and the cornea. Millimeter scale regions of fiber alignment are also observed in the tumor stroma and fibrotic
tissue. At the cell scale, alignment guides cell behavior such as motility, gene expression, and morphology. At the millimeter scale, fiber alignment regulates tissue morphogenesis, facilitates cell-cell communication via force transmission through aligned fibers, and results in anisotropic mechanical properties in tissues. The degree of fiber alignment is also an essential property of type I collagen matrices. Changes in the degree of alignment have been shown to regulate focal adhesion formation, cell motility, and polarization. Given the length scale and wide range of alignment observed in vivo, microfluidic platforms have been used to establish control over fiber alignment in vitro. Early studies to control fiber alignment within microfluidic channels have relied upon shear that is induced during laminar flow. Lee et al. first demonstrated the alignment of a self-assembling collagen solution flowing through microchannels up to 100μm in width and 40μmin height. They showed that collagen fiber alignment was a function of channel width, and no alignment was observed for channels wider than 100μm. While shear has been a common method to induce alignment, the resulting hydrogel constructs are <100μm in width, and <50μm in thickness. Other studies have used microscale flow physics to fabricate free-standing, 3D collagen structures with fiber alignment such as sheets, or to extrude fibers as in bioprinting. However, the resulting structures are free-standing and are not integrated within a microfluidic system, and thus do not offer the versatility and control that is afforded by microfluidics over the microenvironment.
In this study we describe a simple microfluidic approach to i) develop millimeter-scale constructs of 3D aligned collagen fibers within a microfluidic device, and ii) provide regions of defined degrees of alignment within the same 3D (>130μm) construct. In our previous study, we showed that millimeter-scale control over collagen fibers could be established using extensional strain applied to a self-assembling collagen solution. The extensional strain rate in a flow is described as the rate of change of velocity (δv) along the flow direction (δx). Recent studies have also shown that extensional strain can be used to induce fiber self-assembly in a semi-dilute collagen solution, and align fibers during microextrusion. To generate millimeter-scale regions with distinct degrees of alignment, we use a channel that has discrete reductions in the crosssection area, along the flow direction to control the extensional strain rate. To provide control over the cellular microenvironment, we demonstrate the integration of our channel into a reversibly sealed, easy access, modular microfluidic device (SEAM). Using the SEAM platform, we are able to directly access the collagen construct without the need for auxiliary ports, individually address regions with different degrees of alignment, and add secondary ECM materials in a layer-by-layer fashion. We are therefore able to combine the long-range control over collagen alignment as offered by extrusion-based methods, with microfluidic control over the culture environment

 

 

 

Figure 1. Examples of long-range biophysical guidance via aligned COL1 fibers. A) Cell-traction forces induced by hMSC spheroids separated by ~800 um, with cells migrating along the aligned fibers (inset). Scale bar 100 um. B) Aligned collagen fibers promote the long-range organization of HUVECs, with the vessel length increasing in proportion to fiber alignment. Scale bar = 100um. C) Image of biologically patterned COL1 fiber architecture in the mouse mammary end bud. Fiber orientation analysis at branch sites (inset) revealed that COL1 epithelial branch directions were co-oriented with the fibers. Scale bar = 50 um. D) MDA- MB-231 spheroid in a magnetically aligned COL1 gel with cells moving preferentially along the fibers. Scale bar = 100 um

 

Shear flow did not induce 3D, long-range fiber alignment

 

Figure 2: (A) Schematic showing the dimensions of the microchannel (B) Plot shows the coefficient of alignment (CoA) of collagen fibers for shear rates ranging from 50/s to 1000/s in the microchannel after self-assembly. (C) Representative images of collagen fibers injected at different shear rates. Scale bar = 25µm

 

 

Fabrication of segmented channel to induce extensional strain in a self-assembling collagen solution

Figure 3: Schematic showing segmented channel design (A)The segmented channel is 25mm in length, comprising of 5 discrete segments with widths ranging from 10mm to 0.75mm. The channel thickness is 130μm. Outset is a schematic showing the principle of collagen fiber alignment. Collagen subunits experience an extensional strain, at a rate defined as the change of flow velocity in the streamwise direction. (B) PIV measurements of the extensional strain rate in the 4 constrictions of the segmented channel.

 

Extensional strain induced, long-range collagen fiber alignment in 3D hydrogels

 

Figure 4: (A) Schematic showing the different segments of the segmented channel and the
uniform channel. Panels (a-e) show representative images of collagen fibers in the corresponding
section of the segmented channel. The degree of alignment of collagen fibers is seen to increase
across the different segments. (B) Bar plot showing the mean COA ±SD in each segment, at a
flow rate of 50μL min-1. The COA was found to be greater than 0.5 at . (C) Bar plot showing
COA in zero strain conditions ( ̇ε =0) , and in combined extensional and shear flows ( ̇γ + ̇ε ).
The COA in the two conditions is not statistically different at lower shear rates and strain rates,
however, the mean COA increases rapidly at ̇ε = 2.7s−1. (D) representative images of collagen
fibers at a ̇γ =257s-1, in the zero-strain condition and at ̇ε =9.08s-1. Fiber anisotropy can be
visually observed to be higher with extensional strain. n=3. data represented as mean±SD. *
p<0.05, ** p<0.01, Scale bar = 25µm
Figure 5: Figure shows CRM images of collagen fibers across the length of different segments in the channel. The COA within each segment was observed to be constant and the alignment persisted across the entire length of the channels. Scale bar = 25µm

 

Integration with SEAM for direct access to collagen

 

Figure 6: Schematic showing the integration of the segmented channel into a reversibly sealed, easy access modular microfluidic device (SEAM). The SEAM consists of a PMMA base in which the channel cutout is placed. A BSA treated PDMS cover is placed on the channel to inject collagen. (2) Peeling off of PDMS cover to expose collagen (3) Directly accessible collagen. (B) Bar graph showing that fiber alignment is not affected by the cover removal process. (C) Representative images of collagen before and after cover removal. Scale bar = 25µm

 

SEAM integration for cell culture and layer by layer fabrication

 

Figure 7: Figure shows the modular capabilities of the SEAM platform. New modules can be easily attached using magnetic attachment. (A) Cell culture module was used to seed endothelial cell populations in aligned and randomly oriented fibers. (B) Confocal images of HUVECs on aligned and random collagen fibers (C) Module to enable layer by layer deposition of ECM materials. A PDMS plug was inserted into the topmost layers of the cell culture module and collagen with beads was injected as shown in the figure. (D) 3D view showing distinct ECM layers after injection. The PDMS plug could be removed for further processing. Scale bar = 25µm. Scale bar in D = 50µm

Similar Posts