In the human body, neuromuscular junctions are the critical points at which motor neurons communicate with muscles, making them vital for nearly all motor functions. But problems with these junctions from diseases can cause a miscommunication between nerves and muscles, potentially leading to incapacitation and even death.

In an effort to better understand neuromuscular junctions and associated diseases, such as amyotrophic lateral sclerosis (ALS), researchers at the Massachusetts Institute of Technology (MIT) have now developed a microfluidic device that is able to replicate the muscle-nerve connection. Described recently in Science Advances, the US quarter-sized devices allow for cell culture in three dimensions (3D) and contain compartments that physically separate muscle fibers and specially developed photo-excitable motor neurons. The device allows scientists to observe and quantitatively measure the response of the muscle cells to the stimulation of the neurons.

“I would say our main achievement is to be able to provide a platform that recapitulates key aspects of neuromuscular junction formation and motor unit function, which includes the 3D nature of the culture and the compartmentalization of both tissue types,” says study first author Sebastien Uzel, who conducted the research as an MIT graduate student and is now a postdoctoral fellow at Harvard University. “It also provides a way to noninvasively, and in a very well-controlled fashion, excite the motor neurons and record a quantitative force readout.”

Over the last few decades, scientists have developed numerous in vitro systems to try to simulate neuromuscular junctions. In many cases, they have used two-dimensional culture platforms in which motor neurons and muscle cells are seeded in the same Petri dish and use the same culture medium—but this does not exactly represent what goes on in the body.

“In fact, physiologically the cell bodies of motor neurons are located a meter away from the muscle cells and experience a different microenvironment than the muscle cells,” says Josep Samitier, director of Barcelona’s Institute for Bioengineering in Catalonia who was not involved in the study. “The ability to tailor cell microenvironments individually can be desirable and useful, and usually mimics more closely the physiological conditions.”

In prior studies, the devices are compartmentalized but two-dimensional. “Two-dimensional substrates poorly emulate in vivo conditions and are often much stiffer than what muscle cells experience in the body,” Uzel says, adding that the mismatch can affect muscle differentiation, among other things. Additionally, previous systems lacked the ability to measure the force generated by the muscles themselves, an important ability if you're looking to test drugs to restore muscle function. “If you want to restore that muscle function—regardless if you do it by targeting the strength of muscle or the neuronal signal—you want to be able to quantitatively compare the rescued muscle contraction to that of a healthy patient.”

To develop their microfluidic neuromuscular device, Uzel and his colleagues used a standard mold-fabrication process. They used AutoCAD to design the molds, which were then produced using silicon wafers patterned with photolithography, transparency masks, and UV exposure. One mold was for the 320-µm-high microfluidic layer of the device and consisted of three gel regions—the left region for neurospheres (cell clusters containing stem cells and differentiated cells), the right region for muscle strips, and the middle region to provide a 0.5 mm buffer between the two—that are surrounded with two medium channels connected to four medium reservoirs. The second mold was for a thin-membrane layer with two sets of capped pillars that allow for the measurement of muscle force.

They mixed polydimethylsiloxane (PDMS) with a base/curing agent, poured the solutions into the silicon molds, and cured them. They then cut off the molds and retrieved their transparent, rubberlike device layers, which they affixed together so that the pillars lined up in the center of the muscle compartment. They filled the channels of the microfluidic device with a plain hydrogel.

The team injected mouse muscle precursor cells into the muscle compartment of the device as a cell-laden collagen solution; the processes of fusion and differentiation gave rise to skeletal muscle fibers. Using a technique called optogenetics, they modified mouse embryonic stem cells to create neural cells that respond to light; they injected neurospheres containing differentiated motor neurons (that are optically excitable) into the neurosphere compartment. Compared with chemical or electrical stimulation techniques researchers previously used, optical stimulation is neat, noninvasive, and allows for more selective control, Uzel says.

Just one day after seeding, the team saw motor neurites growing from the neurospheres. After four days, the axons reached their maximum 3D growth and reached the muscle strips (though, to their surprise, the scientists found that the axons were not actually attracted to the muscle cells). Five days after seeding, the researchers excited the neurons with light, which caused the muscle to contract. They could measure the strength of the contraction by measuring the deflection of the pillars.

“The system could be applied in the short term for drug screening, toxicology studies, and neuromuscular disease models,” says Barcelona’s Samitier. “And in the long term would be a step forward to engineering autonomous robotics systems based on controlled neuromuscular units.”

Julius Steinbeck, a neurologist and stem cell expert with the Sloan-Kettering Institute for Cancer Research in New York, says that the microfluidic device is “rather sophisticated” and has a number of technical strengths, but it remains to be seen how well it can model human development and disease. “In this regard it is a disadvantage that the system, as of now, relies on non-human cells,” says Steinbeck, who was not involved in the study. “Human cells would have to be cultured for 5 to 10 weeks and it is unclear if the device supports such extended culture times.”

Uzel says the next step for the device is actually to use human cells, adding that “translating the seeding and culture framework described in the study to human tissue should be rather straightforward and wouldn’t require drastic design modification, apart from maybe increasing the device throughput.” Using human cells would allow researchers to study ALS and other neuromuscular diseases, and to test the efficacy of drugs for the conditions. The team is also interested in adding different cell types to the hydrogel to better mimic the microenvironment of motor neurons and muscle cells.

“There are a lot of things you could do in the platform that we haven’t even had the chance to look into for time reasons,” Uzel says. “We’re really hopeful that the platform could serve a broad range of applications.”

Read the abstract in Science Advances.