Organ-specific, multimodal, wireless optoelectronics for high-throughput phenotyping of peripheral neural pathways

Abstract

The vagus nerve supports diverse autonomic functions and behaviors important for health and survival. To understand how specific components of the vagus contribute to behaviors and long-term physiological effects, it is critical to modulate their activity with anatomical specificity in awake, freely behaving conditions using reliable methods. Here, we introduce an organ-specific scalable, multimodal, wireless optoelectronic device for precise and chronic optogenetic manipulations in vivo. When combined with an advanced, coil-antenna system and a multiplexing strategy for powering 8 individual homecages using a single RF transmitter, the proposed wireless telemetry enables low cost, high-throughput, and precise functional mapping of peripheral neural circuits, including long-term behavioral and physiological measurements. Deployment of these technologies reveals an unexpected role for stomach, non-stretch vagal sensory fibers in suppressing appetite and demonstrates the durability of the miniature wireless device inside harsh gastric conditions.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1. Development of a soft, wireless gastric optogenetic implant with a pre-curved, sandwiched tether.

a Illustration of a soft, wireless gastric optogenetic implant with a pre-curved, sandwiched tether. b Image of wireless LED operation in the stomach; scale bar 5 mm. c Procedures for device fabrication; scale bar 5 mm. d Three-dimensional modeling of the mechanics for the pre-curved, sandwiched structure. e Plots of output power vs. curvatures of a tether. The legend numbers represent the curvature of the device as the length of the arc for a radius of 2 mm (see Supplementary Fig. 3 for equation); smaller numbers represent sharper U-shape angle. f Measurement results of device lifetime cycling test for both structures when strain applied in the horizontal (top) and vertical direction (bottom). g Measurements of device lifetime for the pre- and post-curved structure when implanted (pre-curved, n = 8; post-curved, n = 8). Bar graphs are mean ± SEM. Statistical comparison was made using two-tailed t test; ***p < 0.001.

Fig. 2. Electrical characteristics of multiple cage wireless power TX system and multiple organ stimulation devices.

a Schematic illustration of the proposed wireless power TX system for high-throughput phenotyping of neural pathways. b Functional block diagram of the proposed wireless power TX system. c Electromagnetic simulation of wireless coverage for the proposed TX system; scale bar 10 cm. d Picture of the TX system; scale bar 10 cm. e Representative magnetic field distributions in antenna set 6 (top) and comparisons of wireless coverage for the proposed system and other wireless power TX systems (bottom); scale bar 10 cm. f Illustration of wireless operation of a scalable, multimodal wireless gastric optogenetic implant (left), images of an animal with the device implanted (middle), and image of the device (right); scale bar 1 cm. MUX; multiplexer, Decou.; decoupling, SW; switch, Cap; capacitor, Mat. B.; RF matching board, Ant.; antenna set.

Fig. 3. Optogenetic targeting of Calca+ vagal afferents in the stomach.

a Light intensity measurements comparing LED implantation inside vs. outside the stomach (n = 5, p < 0.01), with varying RF powers (p < 0.001). Dashed horizontal lines indicate light intensity needed for 10 and 50% maximal activation of channelrhodopsin2. b Comparison of total food intake, number of meals, and meal size in mice implanted with LED device (n = 7) or sham operated (n = 6) (p = 0.71). cCalcaCre transgenic mice received nodose ganglion injection of AAV9-DIO-ChR2:tdTomato. Images show fluorescence in situ hybridization of tdTomato and Calca mRNA, demonstrating the cell-type specificity of transgenic/viral approach; scale bars 25 µm. d tdTomato fluorescence labeling of central Calca+ vagal afferent endings in the nucleus of the solitary tract (NTS); scale bar 25 µm. e Fluorescence labeling of peripheral Calca+ vagal afferent endings in the stomach mucosal layer; scale bar 50 µm. Behavioral experimental results are from one cohort of animals. Bar graphs are mean ± SEM. Statistical comparisons were made using two-way repeated-measures ANOVA, Tukey’s post hoc; ***p < 0.001.

Fig. 4. Activation of Calca+ stomach vagal afferents suppress appetite via negative valence mechanism.

aCalcaCre transgenic mice received a left nodose ganglion injection of AAV9-DIO-ChR2:tdTomato or AAV9-DIO-tdTomato control virus. The LED was implanted in the stomach corpus-function junction. b Frequency-dependent suppression of food intake in the ChR2:tdTomato group (n = 8). c The tdTomato control group did not suppress food intake during photostimulation (n = 4) (p = 0.06). d Illustration of real-time place-preference (RTPP) box. The RF antenna powered the device only in the right chamber. e Activation of LED device (20 Hz light pulses) did not induce a place preference nor avoidance in both ChR2 and tdTomato groups (n = 7 per group) (p = 0.31). f Representative traces for RTPP assay. g Illustration of a large open-field box; the antenna delivered wireless power throughout the entire arena (20 Hz light pulses). h Photoactivation of Calca+ gastric vagal afferents decreased time spent in center (n = 7 per group). i Representative traces from open-field test. j Mice were exposed to a novel sucrose solution on Day 1 followed by optogenetic activation of vagal sensory fibers (20 Hz). On Day 5, mice were water-restricted overnight and then given simultaneous access to a bottle of sucrose and a bottle of water. The graph is the sucrose preference score (ChR2, n = 7; tdT, n = 5). Experimental results are from one cohort of animals. Bar graphs are mean ± SEM. Statistical comparisons were made using two-way repeated-measures ANOVA, Tukey’s post hoc, except for (h) and (j), which were two-tailed t test; **p < 0.01, ***p < 0.001.