On-chip functional neuroimaging with mechanical stimulation in Caenorhabditis elegans larvae for studying development and neural circuits

Abstract

Mechanosensation is fundamentally important for the abilities of an organism to experience touch, hear sounds, and maintain balance. Caenorhabditis elegans is a powerful system for studying mechanosensation as this worm is well suited for in vivo functional imaging of neurons. Many years of research using labor-intensive methods have generated a wealth of knowledge about mechanosensation in C. elegans, and the recent microfluidic-based platforms continue to push the boundary for this field. However, developmental aspects of sensory biology, including mechanosensation, are still not fully understood. One current bottleneck is the difficulty in assaying larvae because they are much smaller than adult worms. Microfluidic devices with features small enough for larvae, especially actuators for the delivery of mechanical stimulation, are difficult to design and fabricate. Here, we present a series of automatic microfluidic platforms that allow for in vivo functional imaging of C. elegans responding to controlled mechanical stimulation at different developmental stages. Using a novel fabrication method, we designed highly deformable pneumatically actuated on-chip structures that can deliver mechanical stimulation to larval worms. The PDMS actuator allows for quantitatively controlled mechanical stimulation of both gentle and harsh touch neurons, by simply changing the actuation pressure, which makes this device easily translatable to other labs. We validated the design and utility of our systems with studies of the functional role of mechanosensory neurons in developing worms; we showed that gentle and harsh touch neurons function similarly in early larvae as they do in the adult stage, which would not have been possible previously. Finally, we investigated the effect of a sleep-like state on neuronal responses by imaging C. elegans in the lethargus state.

Fig. 1. A key engineering challenge is the scaling-down of the device for adult worms to match the size of the imaging channel to the size of developing worms. (A) Schematic diagrams of the PDMS actuator for the delivery of mechanical stimulation for adult (left) and L2 stage worms (right). All scales of the device for adult worms have to be scaled down except the thickness of the PDMS membrane (25 μm) for L2 stage imaging to minimize the worm movement and deliver localized mechanical stimulation. (B) Example images of worms in the device made by using a standard ratio of the PDMS mixture (10 : 1). 10 psi (left) and 50 psi (right) are applied by using anterior touch valves (red arrows). Worms were cultured 20 hours after hatching. Scale bar: 25 μm.

Fig. 2. Simulation results for the displacement of the actuated membrane by applying various pressures. (A) Examples of PDMS membrane deformation using different applied pressures from the results of COMSOL simulation. The color bar represents the scale of membrane displacement. Scale bar: 10 μm. (B) Red and purple represent simulation results from COMSOL Multiphysics (Young's modulus – red (30 : 1 PDMS mixture used in this study: 110 kPa) and purple (10 : 1 standard PDMS mixture: 1100 kPa)).

Fig. 3. Schematic of the new device fabrication method and results of membrane displacement measurements. (A) Schematic of the new fabrication method. (B) Example images of worms in the device made by using a high ratio of the PDMS mixture (30 : 1). 10 psi (left), 30 psi (middle), and 50 psi (right) are applied by using anterior touch valves (red arrows). Worms were cultured 20 hours after hatching. Scale bar: 25 μm. (C) Displacement of the actuated membrane by applying pressures. Experimental (blue) and simulation results (gray, the same as red in Fig. 2B). The R-square value is 0.9892. (D) Sample frames from an activated ALM neuron show changes in fluorescence due to mechanical stimuli (scale bar: 5 μm, used 100× magnification).

Fig. 4. Both gentle and harsh touch receptor neurons can distinguish the magnitude of applied mechanical stimulation. (A) ALM responses to anterior touch with 1 s and diverse pressures (15 psi: n = 9, 20 psi: n = 7, 25 psi: n = 10). (B) PVD responses to posterior touch with 1 s and diverse pressures (25 psi: n = 5, 35 psi: n = 13, 45 psi: n = 5). (C) Maximum calcium responses of ALM and PVD to a variety of applied pressures. (D) AVM responses of wild-type and mec-4 (e1611 and u253) mutants to anterior gentle touch (1 s and 15 psi, WT: n = 11, mec-4 (e1611): n = 7, mec-4 (u253): n = 15). (E) AVM responses of wild-type and mec-4 (e1611 and u253) mutants to anterior harsh touch (1 s and 35 psi, WT: n = 7, mec-4 (e1611): n = 5, mec-4 (u253): n = 13). (F) Maximum calcium responses of wild-type and mec-4 mutant animals (Kruskal–Wallis test, *p < 0.05, **p < 0.01, n.s. is non-significant). (A–F) All worms in these experiments were cultured 18–22 h after hatching.

Fig. 5. Worms in the L2 lethargus state show drastically reduced neural responsiveness to mechanical stimulation. (A) Simplified circuit diagram showing two mechanosensory neurons connecting to AVA command interneuron to backward locomotion behavior. (B–D) Average traces of calcium responses of L2 and L2 lethargus worms in (B) ALM (L2: n = 19, L2 lethargus: n = 9), (C) AVM (L2: n = 19, L2 lethargus: n = 15), and (D) AVA (L2: n = 15, L2 lethargus: n = 13) to 1 s anterior touch with 50 psi. (E) Maximum calcium responses of L2 and L2 lethargus worms (Kruskal–Wallis test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001). (B–E) All worms in these experiments were cultured either for 18–22 h for L2 or 24–25 h for L2 lethargus. Error bars represent SEM.