|
Research - Multiphoton Excited Nano/MicroFabrication
|
My group has been using multiphoton excited (MPE) photochemistry to fabricate synthetic tissue engineering scaffolds. The method is analogous to two-photon excited fluorescence imaging microscopy, but rather than use the MPE for optical sectioning, we exploit the confinement of the excitation to the focal volume to fabricate 3 dimension structures one plane at a time. Specifically, we use MPE to crosslink proteins together in virtually any desired size or shape. Shown below are fluorescence images of two such objects. The left image is a channel structure created from the polymer TMPTA, and the right is a high aspect ratio cylinder (60 microns high, 1 micron thick) created from the protein BSA.
| The current paradigm in tissue engineering requires the fabrication of multi-scale structures that can reproduce the complex topography and biochemical composition of the extracellular matrix (ECM). For example, the dermal layer of skin consists of elastin and collagen fibers with widths ranging from a few hundred nanometers to a few microns. The figure shows the SHG image from the collagen fibers in ex vivo skin. Fibroblasts require this complex 3D structure to properly spread, differentiate, migrate, and synthesize new matrix. A fundamental issue in the design of tissue engineering scaffolds is an understanding of the topographic and ECM cues that underlie these cellular processes. To shed light on this problem, we have been systematically examining the cell spreading response on simple geometrical patterns we use to emulate the native ECM structure. |
|
 |
We used MPE photochemistry to fabricate linear patterns of crosslinked fibers from the ECM proteins collagen (types I and II), fibronectin, and fibrinogen, as well as the topographic control of crosslinked BSA. A phase contrast image of fibroblasts on crosslinked type I collagen fibers is shown in the left panel below. We see that the cells preferentially spread on and across the patterns. We observe similar behavior from the other ECM proteins, whereas cells on BSA display quantitatively different spreading. The right panel further shows the importance of ECM cues on cell morphology. The cells were triple stained, and the cells on patterns (white dashes) show highly directed stress fibers (green), whereas the cells in the periphery of the image show only diffuse actin staining. Similarly the cells on the crosslinked collagen show punctate focal adhesion staining (red), indicative of strong cell-matrix interactions.
 |
|
We are also examining the materials properties of model scaffolds created through MPE crosslinking. If we are to use our scaffolds as in vivo devices, we must have a thorough understanding of their properties such as diffusion and susceptibility to degradation. To measure diffusion we use the method of Fluorescence Recovery After Photobleaching (FRAP). Here we introduce a tracer dye into the crosslinked protein matrix and bleach out a small region to create a “hole” in the fluorescence. The characteristic recovery of the hole is then indicative of the diffusion coefficient within the matrix. Left show the recovery data for 10 and 70 kD Rhodamine dextrans in a MPE crosslinked BSA matrix. The recovery times for these two species scale the same as in solution, suggesting that we can fabricate such structures with a priori knowledge of the resulting diffusional properties. The corresponding diffusion coefficients are ~10-9 cm2/s, or about 3 orders of magnitude slower than in solution. These coefficients are however, approximately the same as in native ECM. An important result is that through control of the fabrication conditions, we can carefully tune the crosslink density and resulting diffusion coefficients over a range of about 20 fold. We have performed these measurements for the ECM proteins fibrinogen and fibronectin as well. |
 |
|
An additional important aspect in the fabrication of scaffolds is to understand how they degrade with proteases, as this provides insight into how the assembly compares to native ECM, as well as how it would degrade in vivo. We have characterized the degradation of MPE crosslinked collagen matrices with a series of proteases. An example of the resulting data is shown in the figure. Here we degraded fluorescently-labeled type II collagen, with collagenase (100 kD) and pepsin (20kD). These proteases are fairly non-specific and thus cleave similar peptide bonds. The difference in decay rate is related to the difference in diffusion into and throughout the matrix of the higher molecular weight collagenase relative to pepsin. In both cases, the data is well-fit to a single exponential as would be expected for enzymatic degradation. This result gives us confidence in the ability to fabricate scaffolds with knowledge of the crosslinked assembly. |
|
|
|
|
