Figure 1.
Transduction and transmission of stimuli by axons in MFC.
(A) Right, calcium imaging was used to characterize DRG axonal responses in MFC cultures. Schematic shows the assay for axonal function. A stimulus applied to the axonal compartment activates isolated axons and produces an action potential (AP) which propagates along the axon. The subsequent depolarization of the soma membrane facilitates opening of calcium channels, leading to Ca2+ influx that can be monitored by the recording setup. Left, representative images of Ca2+ responses demonstrating DRG cell soma activation after axonal stimulation with capsaicin (100 nM) or KCl (30 mM). The total number of neurons is determined by responders to KCl applied to the somal compartment (bottom left), while the number of axonal crossings is revealed by the fluorescent tracer uptake (bottom right). (B) Representative traces of Ca2+ increases in the neuron soma, following axonal or somal activation with 100 nM capsaicin or 30 mM KCl. (C) Comparison of capsaicin dose-response curves after axonal or somal stimulation. Left, quantification of Ca2+ signal; right, percentage of neuronal responders (values represent normalized mean ± SEM, n = 3 per condition).
Figure 2.
Using MFC cultures to selectively screen drug effects on axonal excitability and conduction.
(A) DRG cultures in triple-compartment MFCs. Axons (stained here for β3tubulin) traverse two compartments growing through two sets of microgrooves; axons on the far right compartment (distal) can then be stimulated, while drugs are perfused through the middle compartment to study their effect on axonal conduction. Scale bar = 50 μm. (B) Left, representative calcium imaging traces illustrating axonal activation after capsaicin stimulation of axonal endings, which is completely blocked with concurrent perfusion of lidocaine (middle). After washing out the lidocaine, axons respond to a second stimulation (right). (C) Left, axonal responses to KCl stimulation are blocked by lidocaine in the proximal chamber. The presence of lidocaine in the axonal compartment does not inhibit KCl responses after somal application (right), illustrating the fluidic isolation property of microfluidic chambers. (D) Quantification of magnitude of Ca2+ response and percentage of responders to capsaicin with or without lidocaine (n = 3 independent cultures).
Figure 3.
Local treatment of axons with NGF enhances capsaicin-evoked responses and TRPV1 expression in the axons.
(A) Chronic NGF applied locally to axons induces a significant increase in capsaicin-induced excitability. Quantification of calcium imaging responses to axonal stimulation with capsaicin or KCl, in control cultures and cultures subjected to axonal NGF treatment. Left, magnitude of response (**p < 0.01, Fisher’s exact test, n = 70 and 89 from three independent experiments); right percentage of responders (**p < 0.01, one-way ANOVA with Bonferroni’s correction, n = 17 - 63 cells, three independent experiments,). (B) Axons treated with high NGF show increased local TRPV1 expression. Image is from a triple-compartment MFC in which neurons were seeded in middle channel and axons crossed on both sides. The axons on the left side were deprived of NGF while the right side was subjected to high NGF axonal treatment. Right, quantification of axonal TRPV1 immunoreactivity with or without NGF treatment (***p < 0.001, Student’s t-test, n = 47 and 37, three independent experiments). Scale bar = 100 μm.
Figure 4.
An in vitro model of traumatic nerve injury using adult mouse DRG cultures in MFCs.
(A) Sensory neurons in MFC (left) can be axotomized in vitro using a suction pipette (middle), following which axons regenerate through the microgrooves (right). All images show staining for β3tubulin. (B-C) Axonal Nav1.8 expression is upregulated 3d after axotomy in the regenerating axons. Images show Nav1.8 staining. Right graphs show quantification of staining as fold change in Nav1.8 expression in the cell soma, microgroove-contained axons and nerve endings, before and after axotomy (*p < 0.05, Student’s t-test, n = 3). (D) Axotomized neurons are sensitized compared to control neurons, evidenced by the increased percentage of responders after axonal stimulation with capsaicin or KCl (left, *p < 0.05, **p < 0.01 vs naïve, Student’s t-test, n = 3 independent experiments). The magnitude of Ca2+ responses to axonal stimulation following axotomy was not significantly different compared to control neurons (right). Scale bars: A = 200 μm, B = 50 μm.
Figure 5.
Electrical stimulation to generate axonal APs in the MFC.
(A) Electrical stimuli applied to the axonal chamber consisted of trains of constant current stimuli of 2 ms duration applied at 5.6 Hz for 5 s, with 20 s inter-train intervals. Stimuli produced somal APs which increased in fidelity as the stimulus intensity was increased. Stimulus intensities for each train are shown on the left. (B) The configuration of the triple-compartment electrophysiological setup. Whole-cell current clamp recordings are made from cell bodies in the somal compartment. Perfusion is applied through the middle axonal channel with electrical stimulation applied in the far axonal channel. (C) Application of lidocaine blocks axonally stimulated responses. Control responses are shown (top trace), where each stimulus resulted in a somal AP. Application of 5 mM lidocaine in the middle axonal channel abolished somal spikes, whereby somal stimulation could still elicit an AP (see inset). Washout reversed the inhibitory effects of lidocaine. All stimuli were suprathreshold at 1.2 mA.
Figure 6.
Using patch clamp electrophysiology to study impact of sodium channel blockers on axonal function.
(A) The top trace shows control somal responses to electrical stimulation of axons (1 mA, 2 ms duration stimuli, applied in 5 s, 5 Hz trains, with 20 s inter-train intervals). Addition of 0.5 μM TTX to the middle axonal channel completely inhibited somal APs, however increasing stimulus intensity to 2 mA restored the somal response. Subsequent application of lidocaine to the middle channel in this cell abolished responses which could not be restored with higher stimulus intensities. Washout reversed the inhibitory effects of lidocaine. Application of lidocaine to the far stimulation channel (bottom trace) had no effect on the somal response. (B) The top trace shows control responses to electrical stimulation of axons (5 mA, 2 ms duration stimuli, applied in 10 s, 2.2 Hz trains, with 10 s inter-train intervals). The addition of the Nav1.8 blocker A-803467 (1 μM) to the middle axonal channel in this cell increased the somal spike failure rate. Washout reversed the effects of A-803467 and the lower trace shows complete abolition of somal responses following axonal application of lidocaine (middle channel).
Figure 7.
MFCs can be used to examine the electrophysiological effects of in vitro axotomy.
(A) Following in vitro axotomy there was no significant difference in the somal AP threshold to axonal stimulation, however a new population of cells with higher thresholds appeared (upper panel). Axotomy produced a significant increase in the axonally stimulated somal spike amplitude (*p < 0.05, lower panel). (B) In vitro axotomy caused no significant difference in the somal AP rheobase or spike amplitude following direct somal stimulation (control, n = 6; axotomy, n = 5 for all data).
Figure 8.
Using MFCs to study axonal function in DRG keratinocyte co-cultures.
(A) Rat neonatal keratinocytes culture in MFC. Immunocytochemistry for basal keratinocyte marker cytokeratin 5 and differentiation marker cytokeratin 10, illustrating low levels of differentiation after 8 div in MFC. (B) A co-culture of rat neonate sensory neurons (stained for β3tubulin, left compartment) and keratinocytes (stained for cytokeratin5, right compartment) in MFC at 6 div. (C) Example traces of axonal activation in the presence or absence of co-cultured keratinocytes in the axonal compartment, as measured by Ca2+ imaging in response to capsaicin stimulation of axons. Scale bars = 100 μm. (D) Co-cultured keratinocytes did not impede crossing of DRG axons. There was no difference in tracer (DiO) uptake between cultures grown with or without keratinocytes, indicating similar levels of axonal crossing.