A microfluidic culture platform for CNS axonal injury, regeneration and transport

Abstract

Investigation of axonal biology in the central nervous system (CNS) is hindered by a lack of an appropriate in vitro method to probe axons independently from cell bodies. Here we describe a microfluidic culture platform that polarizes the growth of CNS axons into a fluidically isolated environment without the use of targeting neurotrophins. In addition to its compatibility with live cell imaging, the platform can be used to (i) isolate CNS axons without somata or dendrites, facilitating biochemical analyses of pure axonal fractions and (ii) localize physical and chemical treatments to axons or somata. We report the first evidence that presynaptic (Syp) but not postsynaptic (Camk2a) mRNA is localized to developing rat cortical and hippocampal axons. The platform also serves as a straightforward, reproducible method to model CNS axonal injury and regeneration. The results presented here demonstrate several experimental paradigms using the microfluidic platform, which can greatly facilitate future studies in axonal biology.

Figure 1

The microfluidic-based culture platform directs axonal growth of CNS neurons and fluidically isolates axons. (a) The culture chamber consists of a PDMS mold containing a relief pattern of somal and axonal compartments (1.5 mm wide, 7 mm long, 100 mm high) connected by microgrooves (10 μm wide, 3 μm high). The optically transparent PDMS adheres to a polylysine-coated coverslip. Rat CNS neurons (green) are added to the somal-side reservoir and are drawn into the somal channel (black) by capillary action. Within 3–4 d, axonal growth is guided into the axonal side (yellow) through the microgrooves. (b) A volume difference between the somal side and axonal side (∼50 μl) allows chemical microenvironments to be isolated to axons for over 20 h owing to the high fluidic resistance of the microgrooves. Similarly, the volume difference can be reversed to isolate a chemical microenvironment to the somal side. (c) Fluidic isolation of Texas red dextran (top panel) to the axonal compartment demonstrates that axonal or somatic microenvironments can be independently manipulated using this culture platform. Axonally restricted application of CellTracker Green (middle panel) backtracked neurons from their isolated axons. The bottom image is the merged figure. Scale bar, 100 μm. (d) Counts of radioactivity in samples from somal and axonal compartments after [³⁵S]methionine was localized to the axonal compartment for over 20 h. Counts in the somal compartment (3.7 c.p.m. ± 1.5 s.e.m.) were similar to background levels. Error bars, s.e.m. (n = 3).

Figure 2

Axons are isolated without somata or dendrites. When barrier widths of 450 μm or more were used, axons (red; tau) extended past the barrier at 14 d in vitro without detecting dendrites (green; MAP2). Frames including the longest dendrites in each chamber were imaged. Dashed lines indicate the barrier region.

Figure 3

RNA encoding the presynaptic vesicle protein synaptophysin is localized to CNS axons at 6 d in vitro. RT-PCR analysis of the samples from somal and axonal compartments demonstrate that β actin and synaptophysin mRNAs are localized to axons. H1 histone and calcium/calmodulin-dependent protein kinase II α mRNAs were not detected in isolated axons in chambers with 450 μm barriers at 6 d in vitro.

Figure 4

Axotomy leads to rapid transcription of immediate early genes and regeneration is enhanced by axonal neurotrophin treatment. (a) Axonal side prior to (left), immediately after (middle) and 20 h after (right) axotomy. Arrows indicate axonal regeneration. Scale bar, 50 μm. (b) RNA was isolated from the somal compartment of unlesioned cultures and cultures at 15 min and 2 h after axotomy (9 d in vitro). Semiquantitative RT-PCR was performed for c-fos and GAPDH.PCR products were visualized on a 1% agarose gel and quantified. Error bars, s.e.m. (n = 3; ANOVA, P < 0.0015; Fischer's PLSD P < 0.02). (c) Neurotrophins enhance axonal regeneration following a lesion. Neurofilament immunofluorescence (red) and f actin–labeled growth cones (green, phalloidin), shown approximately 200 μm from the edge of the barrier in the axonal side, reveal extensive axonal arborization and increased ingrowth after axonal BDNF and NT-3 treatment (left), control (right). Data is representative of four experiments. Scale bar, 25 μm.

Figure 5

Axons can be cocultured with other cell types such as oligodendrocytes. Oligodendrocytes were added to the axonal compartment of 7 d in vitro neurons. (a) Low-power confocal image demonstrates the coculture of neurons (red, neurofilament) with oligodendrocytes (green, myelin basic protein) in separate compartments for eight additional days. In some cases, oligodendrocytes align themselves along axons (arrows) in a pattern reminiscent of white matter tracts in vivo. Dashed lines indicate the barrier region. Scale bar, 60 μm. (b) Higher-power views demonstrate the association of oligodendrocyte processes (green) with neuronal axons (red). Scale bars, 30 μm(left) and 10 μm (right). (c) A single confocal optical Z-slice demonstrates myelin basic protein immunoreactive processes surround an axon. Scale bar, 4 μm.