Abstract
Femtosecond laser nanosurgery is a standard axonal injury model for nerve regeneration studies in Caenorhabditis elegans. However, manual worm handling is a significant bottleneck. This paper reports a fully automated microfluidic platform that performs laser axotomies of fluorescently tagged neurons in living worms. The system utilizes a unique T-shaped chip design with 3D microfluidic interconnects and automated image processing to achieve a processing rate of ~17 s/worm—at least an order of magnitude faster than manual approaches. With a success rate of 67.4%, this platform enables high-throughput, genome-wide screening of nerve regeneration.
Introduction
The nematode Caenorhabditis elegans is an essential model organism in neurobiology due to its transparent body and well-mapped nervous system. Researchers frequently use laser axotomy—severing a single axon with a high-precision laser—to study nerve regeneration in vivo. Despite its utility, the process is traditionally labor-intensive. To target a 300 nm-wide axon, worms must be perfectly immobilized, usually via manual gluing or anesthetics. These methods are slow, potentially harmful to the biological recovery process, and limit throughput to a few animals per hour.
To address these limitations, we developed a fully automated microfluidic platform that integrates mechanical immobilization with real-time image processing and femtosecond laser surgery. This system minimizes human intervention and significantly accelerates the axotomy process, paving the way for large-scale genetic screenings.
System Design and Methodology
The platform consists of a custom two-layer polydimethylsiloxane (PDMS) microfluidic chip and a LabVIEW-based automation suite.
Microfluidic Architecture
The chip features a T-shaped architecture designed for serial transport and immobilization. Key innovations include:
- Staging and Trapping: Gate valves isolate individual worms, while a "T" junction uses pressure differentials to straighten and center the worm against a microfluidic sieve.
- 3D Interconnects: To maintain the high optical clarity required for 1.4 NA oil-immersion objectives, the control valves are located on a separate layer and connected via vertical thru-holes. This prevents valve structures from interfering with the laser path or imaging quality.
Automated Image Processing
The automation sequence follows four critical steps:
- Centroid Identification: The system locates the worm at low magnification.
- Soma Detection: High-magnification fluorescence imaging identifies the neuronal cell body (soma) using circular object detection.
- Neuron Verification: The software distinguishes between specific neurons (e.g., ALM vs. AVM) based on their anatomical orientation.
- Axon Targeting: A closed-loop proportional control algorithm aligns the femtosecond laser spot with the target axon, compensating for mechanical stage hysteresis.
Results and Discussion
The system was evaluated using the SK4005 strain of C. elegans, which expresses GFP in touch receptor neurons. Across a population of 350 worms, the platform achieved an average processing rate of 17.0 ± 2.4 seconds per worm. This is approximately 11 times faster than a skilled human operator.
The overall success rate for automated axotomies was 67.4 ± 3.2%. Crucially, the mechanical immobilization did not negatively impact the animals; lifespan (17.6 days for chip-processed vs. 19.3 days for control) and nerve reconnection probabilities showed no statistically significant differences compared to manual methods. The system successfully targeted 300 nm axons using a 620 nm laser spot, demonstrating sub-micron precision.
Limitations and Future Directions
The primary operational challenge is chip clogging. Over extended periods (approx. 20 hours), debris can accumulate in the microfluidic sieves, requiring cleaning cycles with bleach.
Future improvements could include:
- Enhanced Actuators: Replacing open-loop piezo actuators with strain-gauge-based controllers to further reduce timing variability.
- End-to-End Automation: Integrating the platform with upstream multiwell delivery systems and downstream automated confocal imaging for a completely "hands-free" research pipeline.
This platform establishes a robust methodology for high-throughput nerve regeneration studies, potentially accelerating the discovery of genetic factors involved in neural repair.
Summary updated