Creating Interactions between Tissue-Engineered Skeletal Muscle and the Peripheral Nervous System

Abstract

Effective models of mammalian tissues must allow and encourage physiologically (mimetic) correct interactions between co-cultured cell types in order to produce culture microenvironments as similar as possible to those that would normally occur in vivo. In the case of skeletal muscle, the development of such a culture model, integrating multiple relevant cell types within a biomimetic scaffold, would be of significant benefit for investigations into the development, functional performance, and pathophysiology of skeletal muscle tissue. Although some work has been published regarding the behaviour of in vitro muscle models co-cultured with organotypic slices of CNS tissue or with stem cell-derived neurospheres, little investigation has so far been made regarding the potential to maintain isolated motor neurons within a 3D biomimetic skeletal muscle culture platform. Here, we review the current state of the art for engineering neuromuscular contacts in vitro and provide original data detailing the development of a 3D collagen-based model for the co-culture of primary muscle cells and motor neurons. The devised culture system promotes increased myoblast differentiation, forming arrays of parallel, aligned myotubes on which areas of nerve-muscle contact can be detected by immunostaining for pre- and post-synaptic proteins. Quantitative RT-PCR results indicate that motor neuron presence has a positive effect on myotube maturation, suggesting neural incorporation influences muscle development and maturation in vitro. The importance of this work is discussed in relation to other published neuromuscular co-culture platforms along with possible future directions for the field.

Fig. 1

3D co-culture platform and cell population characterization. a A chamber slide (1) fitted with custom-built flotation bars at either end (2). A cell-laden collagen solution is allowed to gel between the flotation bars. The collagen gel is then detached from the edges of the chamber slide so that it floats in the culture medium. Cell-mediated matrix compaction leads to the generation of isometric strain between the flotation bars (white arrow). Compaction of the collagen gel over time produces the characteristic bowing illustrated (3). Scale bar = 10 mm. b Image of MDCs maintained in this culture platform for 2 weeks and stained for desmin (green) and nuclei (blue). Note the parallel alignment of the developing myotubes. Scale bar = 20 μm. c Image of a muscle tissue section taken from a P1 rat pup, cryosectioned, and stained for desmin (green) and nuclei (blue). Scale bar = 20 μm. d Image of ventral horn motor neurons maintained on glass coverslips for 7 days before being stained for MAP-2 (red) and nuclei (blue). Scale bar = 40 μm. e Comparison of fusion efficiency in MDC-only 3D constructs versus motor neuron-muscle co-cultures. p = 0.86.

Fig. 2

Neurite development and synaptic contact within 3D collagen-based co-culture constructs. Longitudinal slices (30 µm) were taken from 3D constructs for immunostaining and imaging. a, b Sections were stained for desmin (green) and MAP-2 (red). Scale bars = 20 μm. c, d Sections were stained for SV-2 (green) and AChRs (red). Scale bars = 10 μm. a Neurites were typically seen tracking along parallel, and in close proximity to, underlying myotubes (white arrows). b Occasionally, neurites were also found wrapped around cultured myotubes (purple arrow). c Pre- and post-synaptic co-localisation (blue arrow) was observed at a frequency of 4.24/mm². d SV-2 tracks followed underlying myotube orientations (yellow arrows) indicating the path of neurite development and axonal transport of synaptic proteins from the cell bodies toward the developing growth cone.

Fig. 3

Gene expression changes in motor neuron-muscle co-cultures. Fold change in mRNA expression levels for genes encoding markers of myotube maturation and post-synaptic membrane development, measured using quantitative RT-PCR. Expression levels of MYH1 (adult fast isoform), MYH3 (embryonic isoform), and MYH8 (neonatal isoform), as well as troponin T1 and AChRε were quantified and expressed relative to levels recorded for 3D constructs without motor neurons at equivalent time points. In all cases, C_T values were normalised to an internal housekeeping gene (RPIIB) before subsequent analysis. For all genes examined, with the exception of MYH1, values were found to be significantly greater than those recorded for MDC-only controls (* p < 0.05). n = 3. Error bars = SEM.

Fig. 4

The effect of motor neuron presence on matrix compaction. a, b Macroscopic images of collagen constructs seeded with 5 × 10⁶ MDCs/ml following 21 DIV. Scale bars = 25 mm. a MDC-only control. b Motor neuron-myotube co-culture. c Graph detailing the measured reduction in surface area of collagen constructs seeded with 5 × 10⁶ MDCs/ml either in co-culture with motor neurons (solid line) or in monoculture (dashed line). n = 5. Error bars = SEM. * p = 0.03. d, e 30-µm-thick cross sections were taken from 3D collagen-based constructs seeded with 5 × 10⁶ MDCs/ml and stained for desmin (green) and MAP-2 (red). Note the difference in shape between the rounded construct co-cultured with 1 × 10⁶ motor neurons (d) and the elongated MDC-only control (e). N.B. The image in e represents half the total width of the construct since the entire gel was too large to section reliably. Similar results were obtained across all cultures examined (n = 3). d, e The widths of the constructs are indicated by red scale bars. These dimensions correspond to those indicated in the macroscale images presented in a and b. Note that the width in muscle-only constructs is far greater than the orthogonal height, whereas the neuron-muscle co-cultures possess a more rounded morphology. Scale bars = 100 μm.