Human Motor Assembloids Show Different Energetic Responses to Three NAD⁺-Pathway Interventions
A human neuron–muscle construct developed functional neuromuscular connections and hypoxia-related deficits. In a proof-of-concept drug comparison, 78c and NMN improved different NAD⁺/ATP measurements, while the NAMPT activator NAT had little effect; restoration of motor function was not demonstrated by those biochemical results.
Connecting Human Motor-Neuron Spheroids with Skeletal Muscle
A Fudan University-led study combines human motor-neuron spheroids with engineered skeletal muscle to examine neuromuscular injury under intermittent hypoxia. The Nature Communications report uses a patterned open device to organize the tissues, then follows electrical activity, muscle movement and cellular energy metabolism. Its drug comparison tests whether different ways of supporting NAD⁺ metabolism produce the same biochemical response.
The neuronal component was generated from a commercially obtained human induced pluripotent stem-cell line. The muscle component used purchased human skeletal-muscle cells. Three to four motor-neuron spheroids were distributed along each developing muscle bundle. These were assembled laboratory tissues; the study did not use disease-patient-specific neuron–muscle pairs.
Regional surface treatment created anchoring points and a less adhesive central region, guiding the muscle into aligned bundles. Neuronal projections extended over the muscle and formed structures consistent with neuromuscular junctions. The study also compared surface patterning with frame and micropillar anchoring: average formation success was not significantly different, although the patterned approach showed lower batch variability.
Functional Controls and Hypoxia Reveal Different Levels of Injury
The investigators used targeted neuronal stimulation to test whether muscle responses depended on the neural component. Optogenetics, which makes selected neurons responsive to light, elicited synchronized activity and muscle contraction. Curare blocked the muscle response while preserving neuronal output, supporting functional transmission across the junction. Microelectrode arrays recorded activity, and image-based movement analysis measured contraction.
Intermittent hypoxia alternated 40 minutes at 1% oxygen with 20 minutes at 21%, against a continuously normoxic comparison. Four days of this exposure reduced muscle contraction and altered muscle-fiber structure. Repeated contraction comparisons used four assembloids, while molecular and metabolic experiments had their own separately reported replicate counts.
Neuronal and structural measurements did not always change together. Mature motor-neuron spheroids showed lower firing activity and synchrony under hypoxia while subsequent axonal outgrowth remained relatively unchanged. Within the assembled tissues, muscle and neuronal recording regions also showed different electrical responses. The model therefore exposed functional deficits that a single structural image would miss.
Transcriptomic analysis identified changes in oxidative-phosphorylation and neuromuscular programs. Whole-assembloid sequencing compared five hypoxia-exposed constructs with four controls. Follow-up protein and enzyme measurements showed impaired mitochondrial complex I activity, increased oxidative stress and disruption of the NAD⁺/NADH balance, providing the rationale for the pharmacological comparison.
Three Approaches to NAD⁺ Metabolism Produce Unequal Results
The study compared three interventions: nicotinamide mononucleotide, or NMN, as an NAD⁺ precursor; NAT, described by the authors as an activator of the NAD⁺-synthesis enzyme NAMPT; and 78c, an inhibitor of the NAD⁺-consuming enzyme CD38. The comparison examined NAD⁺, NADH, their ratio and ATP, with three independent experiments reported for these metabolic readouts.
NAT had a negligible effect under the tested conditions. The other two agents produced distinct responses: 78c increased NAD⁺ and corrected the NAD⁺/NADH ratio, whereas NMN mainly restored ATP. An increase in an energy-related measurement therefore did not imply uniform correction of the entire redox and energetic profile.
The drug experiments were a biochemical proof of concept. The report does not show that the same treatments restored the hypoxia-impaired contraction or electrical-network outcomes. Connecting these metabolic changes with functional rescue would require those endpoints to be measured after treatment, rather than inferring recovery from NAD⁺ or ATP alone.
Architecture Also Shapes What Can Be Measured
The construct preserves coupled neuronal and muscle responses, but axons penetrated deeper muscle regions only incompletely. Neural stimulation could also influence fibers beyond directly innervated sites. These features limit how precisely a recorded muscle response maps onto an individual neuromuscular connection.
The microelectrode arrays were designed primarily for monolayer cultures, leaving some three-dimensional activity undetected. Recordings also required transfer from the original device onto the electrode plate. The study thus pairs a useful human neuromuscular injury model with practical measurement constraints. Its clearest pharmacological result is the endpoint-specific response to NMN, NAT and 78c, including the weak NAT finding and the distinction between metabolic improvement and demonstrated motor recovery.