We also compared the transfection efficiency of our new electroporation technique with that of Lipofectin (Life Technologies), a reagent that has been used to successfully transfect chick embryo wing bud mesenchyme in suspension (12, 13). Lipofectin-mediated transfection resulted in a much lower efficiency (~11%) and significantly decreased the number of cells present in micromass (Figure 1A, B). Peister and colleagues (14) used square wave pulse electroporation and hypoosmolar buffer (Eppendorf, Hauppauge, NY) to introduce plasmid DNA into hard-to-transfect human bone marrow MSCs. However, when we substituted hypoosmolar buffer for SEB, we could not achieve greater than 10% transfection efficiency of wing bud mesenchyme under our various electroporation conditions (data not shown).
We settled on 3 – 400 V pulses 150 μs in length, 100 ms intervals, and 1 μg of plasmid DNA as our optimized electroporation parameters. We wished to determine whether electroporation-mediated plasmid transfection performed according to these parameters would affect the extent of chondrogenic differentiation in micromass culture. We stained 3-day micromasses with Alcian blue, a histochemical dye that binds proteoglycans present in the cartilage ECM. Micromasses composed of electroporated cells accumulated levels of Alcian blue-positive cartilage ECM that were visually indistinguishable from unelectroporated controls (Figure 1E). Next, we transfected limb bud mesenchyme with a human SOX9 promoter-GFP reporter plasmid consisting of a 750 bp fragment of the human SOX9 promoter ligated into pEGFP-1 (Clontech, Mountain View, CA) and driving expression of the g fp gene (15). Under fluorescence microscopy, cartilage nodules were easily identifiable by their strong GFP expression, whereas internodular regions composed of muscle and tendon cells (16, 17) were GFP-negative (Figure 1F).
Next, we wanted to demonstrate the usefulness of our technique for inducing observable changes in micromass culture morphology. We therefore transfected limb mesenchyme with an expression vector encoding full-length human SOX9 (OriGene Technologies Inc., Rockville, MD). We carried out real-time PCR with primers directed against human SOX9 (18) to confirm that the mRNA was indeed expressed in the chick cells (data not shown). Collagen type II immunostaining (12) revealed that SOX9 overexpression resulted in the formation of a uniform sheet of cartilaginous tissue, without the internodular, non-cartilaginous regions found in controls transfected with the corresponding empty vector (Figure 2A). Western blotting (12) with an antibody against sarcomeric myosin (MF 20; Developmental Studies Hybridoma Bank, Iowa City, IA) demonstrated that, as expected, cultures overexpressing SOX9 form less differentiated muscle than controls (Figure 2B). In addition, when we co-transfected the SOX9 expression vector with the SOX9 promoter-GFP reporter plasmid, we found that the uniform sheet of cartilage was observable by fluorescence microscopy (Figure 2C).
Finally, we wished to determine whether introduction of the SOX9 gene via our new electroporation protocol could block the effects of some recently described chondro-inhibitory treatments. For example, an intact vimentin cytoskeleton is required for normal cartilage-characteristic gene expression in human bone marrow MSCs (18) and bovine chondrocytes (19). Acrylamide-induced collapse of the vimentin network causes decreased collagen type II production in both of these cell types. We found that SOX9 overexpression could block the decrease in collagen type II accumulation caused by 5 mM acrylamide (Sigma-Aldrich) treatment in micromass cultures of wing bud cells (Figure 2D). ten Berge and colleagues (20) have shown that treatment of micromass cultures with a combination of FGF8 and WNT3A maintains wing bud mesenchymal cells in an undifferentiated state via synergistic repression of Sox9. We show that transfection of wing bud mesenchyme with a SOX9 expression plasmid can overcome the chondro-inhibitory effects of FGF8/WNT3A treatment (Figure 2E).
In conclusion, we have described a novel electroporation-based protocol for highly efficient transfection of embryonic limb mesenchyme with plasmid DNA, and we show that these transfected cells retain their chondrogenic potential when plated in micromass. This technique should serve as an effective tool for unraveling the molecular mechanisms regulating embryonic cartilage formation, and the potential exists to apply this new method to other hard-to-transfect mesenchymal cell types.Author contributions
BEB, PGA, and RST conceptualized the study; BEB and PGA designed and performed the experiments; BEB, PGA, and RST analyzed and discussed the results; BEB and RST wrote the manuscript.