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

The study addresses the challenge of achieving cell-type- and organ-specific manipulation of peripheral neural pathways in awake animals. While optogenetics has transformed circuit dissection in the brain, translation to the periphery has been limited by fiber-optic constraints, lack of stable interfaces, and risks of tissue damage during natural movements. The vagus nerve, the primary neuronal conduit between organs and brain, contains transcriptionally distinct afferents with diverse functions, but organ-specific manipulation has been difficult. Prior methods using fiber optics or non-durable wireless devices were restricted to anesthesia, cumbersome reflex studies, or short functional lifetimes. The purpose of this work is to develop a robust, fully implantable, wireless optogenetic platform that delivers organ-specific light inside hollow organs, operates reliably over long durations in freely moving mice, scales to high-throughput experimentation, and enables functional mapping of peripheral neural circuits. The significance lies in enabling causal, chronic, and organ-specific interrogation of peripheral pathways, exemplified here by stomach vagal afferents and their roles in appetite and behavior.

Existing approaches for peripheral optogenetics include fiber-optic systems that lack stable interfaces and can damage tissues during movement, and early wireless μLED devices affixed to organ surfaces that suffer from light back-scatter, non-specific activation, mechanical strain during organ expansion, and limited durability (typically <8 days). Studies placing μLEDs on the heart allowed pacing but were not suited for long-term behavioral paradigms requiring extended recovery. Organ-surface mounting produces back-scatter that risks activating neighboring tissues. Optogenetic manipulation of vagal afferents with organ specificity has previously been performed under anesthesia, limiting behavioral investigations beyond reflexes. Wireless systems that improve cage coverage often increase RF power (raising tissue exposure and heat) or require complex multiplexing across few antennas with potential interference. Together, these limitations motivate a fully implantable, organ-internal, durable, and scalable wireless optogenetic solution with improved electromagnetic coverage and low thermal/RF exposure.

Device design and fabrication: The fully implantable device comprises an analog front-end RF energy harvester (radius ~5.5 mm, thickness ~1 mm) and a thin, soft tether that delivers current to a μLED positioned internal to the target organ (stomach). The μLED is located mid-tether to permit threading in and out of the organ wall and securing at two points, enhancing stability. Electrical interconnects (12 μm copper) are patterned on an 18 μm polyimide substrate and encapsulated in biocompatible polymers. A unique pre-curved, sandwiched tether design is implemented: the μLED and traces are embedded within multiple polymer layers and overcoated with silicone while held in a pre-bent configuration. This reduces mechanical strain during implantation and operation compared to coating flat and bending post-fabrication. The resulting implant is thin, compliant, and lightweight (<380 mg).

Mechanical/thermal/electromagnetic analyses: 3D finite-element modeling quantified strain distributions, showing large reductions in maximum strain for the pre-curved versus post-curved tether. Mechanical cycling tests across curvatures (0.72, 1.15, 2.87 mm radii) and directions (x, y, z) identified an optimal pre-curved geometry (radius ~1.15 mm) with markedly improved durability. Accelerated saline submersion tests confirmed waterproofing over >2 months. Thermal measurements during typical operating paradigms (10–20 Hz pulses, 5–10% duty) showed minimal temperature rise (~0.2 °C). Electromagnetic simulations and SAR calculations indicated compliance with IEEE safety guidelines.

Wireless power and scalability: A centralized RF generator (13.56 MHz) feeds an RF multiplexer to power eight independent home cages using paired top/bottom antennas per cage. When a given cage is selected, its antenna pair is tuned to 13.56 MHz while others are detuned (e.g., to 100 MHz) via coupling/decoupling circuits, enabling simultaneous and independent control without cross-talk. A dual-coil antenna configuration (driven top coil with a passive bottom coil) improves volumetric wireless coverage within the cage and reduces orientation dependence without increasing RF power.

Multimodal/channel selection: Devices incorporate a reed switch that responds to programmed RF pulse patterns. Long pulses (>100 ms) actuate channel switching between LEDs (e.g., green to blue) on separate tethers, enabling multimodal organ targeting without microcontrollers or high-power radios.

Surgical implantation and targeting: Under sterile conditions, the gastric device is implanted with the μLED positioned within the stomach lumen (corpus region), with the tether secured at two contact points across the gastric wall. The design confines light within the organ, minimizing back-scatter to adjacent tissues.

Optogenetic models and behavioral assays: CalcaCre transgenic mice received AAV9-DIO-ChR2-tdTomato (or control) injections into the nodose ganglion to target Calca+ (Calca/Calcr-related naming in text) gastric vagal afferents. Histology verified transduction specificity (peripheral mucosal endings and central NTS projections). Behavioral assays included: meal pattern analysis (food intake, meal number/size), real-time place preference (RTPP) with RF power restricted to one chamber, open-field exploration with full-coverage wireless power, and a two-bottle sucrose preference/aversion paradigm following paired photostimulation. Device tolerability was assessed against sham surgeries.

• Device durability and mechanics: Pre-curved, sandwiched tethers exhibited dramatically reduced modeled strain (copper/PDMS regions <800 Pa) versus post-curved designs (~6600 Pa). Mechanical cycling showed functionality beyond 200 kJ of work with the optimal pre-curved geometry (~1.15 mm radius), representing ~10-fold improvement over post-curved structures. In vivo, pre-curved devices remained functional for over 1 month, while post-curved devices failed by day 3.
• Environmental robustness and safety: Devices remained operational for >2 months in heated saline. Temperature rise during operation was minimal (~0.2 °C). SAR simulations under localized RF exposure were below IEEE safety limits.
• Wireless scalability and coverage: A single RF transmitter, via an RF multiplexer and coupling/decoupling networks, powered eight cages by tuning the selected antenna set to 13.56 MHz and detuning others (e.g., 100 MHz). A dual-coil antenna (driven top, passive bottom) produced near-uniform cage coverage and reduced dependence on device-antenna orientation, outperforming single-antenna systems without increasing RF power.
• Multimodal operation: Reed-switch-based channel selection enabled remote toggling between multiple LEDs by applying RF pulse patterns (e.g., >100 ms pulses), facilitating multi-organ or multi-opsin actuation without energy-hungry radios.
• Optical confinement and tolerability: μLED placement inside the stomach confined light spread compared to surface-mounted LEDs, which produced significant back-scatter above optogenetic activation thresholds. Device implantation did not alter baseline feeding compared to sham (n = 7 vs 6, p = 0.71).
• Targeting of gastric vagal afferents: CalcaCre mice injected with AAV9-DIO-ChR2-tdTomato showed targeted labeling in nodose ganglion, NTS, and stomach mucosa. Photostimulation of Calca+ gastric vagal afferents suppressed food intake in a frequency-dependent manner (food intake during photostimulation n = 4, p = 0.06 in one analysis), reduced center time in open-field tests (n = 2, p < 0.001), and did not produce RTPP preference/avoidance (n = 2, p = 0.31). A conditioned taste test revealed decreased sucrose preference following pairing with photostimulation (indicative of negative valence/aversive learning). Statistical examples: Fig. 1f device lifetime improvement, two-tailed t-test p < 0.001; Fig. 3a intra- vs extra-gastric light intensity differences p < 0.01 and dependence on RF power p < 0.001.

The platform addresses the central challenge of organ-specific, cell-type-selective manipulation of peripheral neural circuits in freely moving animals. By placing a μLED within the organ via a pre-curved, sandwiched tether, the system achieves durable, confined illumination with minimal thermal load and safe RF exposure. The scalable telemetry (eight cages with a single transmitter) and dual-coil antenna provide reliable, cage-wide power without overexposure, enabling high-throughput behavioral studies. Functionally, selective activation of stomach Calca+ vagal afferents suppressed feeding and produced behavioral signatures consistent with negative valence (reduced center exploration and conditioned avoidance of sucrose), indicating that these chemosensory mucosal afferents can suppress appetite through aversive mechanisms rather than positive satiety alone. Compared to prior fiber-optic or surface-mounted devices, this approach offers chronic stability, reduced back-scatter, organ specificity, and multimodal control. The findings advance understanding of gut-brain mechanisms of appetite and provide a toolset for mapping peripheral circuits in vivo.

This work introduces a durable, fully implantable, organ-specific, multimodal wireless optogenetic platform that enables precise, chronic manipulation of peripheral neural endings in freely behaving mice. Engineering innovations—a pre-curved, sandwiched tether, a dual-coil antenna for comprehensive cage coverage, and RF-multiplexed powering of eight cages—enable high-throughput experimentation with minimal thermal and RF burden. Application to gastric Calca+ vagal afferents uncovers a role in appetite suppression via a negative valence mechanism. The platform is readily adaptable to other hollow organs and opsin wavelengths, supporting multimodal and multi-organ studies. Future directions include chronic activation paradigms over weeks to months to probe adaptation and persistence of circuit effects, testing efficacy in disease models (e.g., obesity), and expanding to additional peripheral targets for therapeutic discovery.

Behavioral experiments were conducted in a single cohort with some small sample sizes (e.g., RTPP and open-field n = 2), limiting statistical power. Some outcomes approached but did not reach conventional significance (e.g., food intake during photostimulation p = 0.06). Organ-specific demonstrations focused on the stomach; generalization to other organs, while feasible, was not empirically shown here. Cell-type targeting relies on transgenic Cre lines and viral transduction; animals without confirmed expression were excluded. Although SAR and thermal rises were within safety limits, long-term biocompatibility beyond the durations tested and potential effects of chronic RF exposure warrant further study.