Pharmacokinetic Modeling of Acetylcholine‐Induced Contraction of Mouse Trachea

Abstract

Autonomic and sensory nerves containing a variety of neurotransmitters innervate the lungs and control airway diameter. Pulses of electric field stimulation (EFS) cause release of all neurotransmitters where their effects on smooth muscle tone can be measured. The primary neurotransmitter driving airway contraction is acetylcholine (ACh). We therefore developed a pharmacokinetic (PK) model which treats EFS pulses as doses of ACh being administered to the trachea. This PK model builds outward from first principles using computational methods to describe this system. Understanding the functional role of the multiple neurotransmitter systems in the airways may inform the rational use of drugs in the treatment of asthma, COPD, and other airway diseases. We hypothesized that ACh‐induced tracheal contraction caused by electric stimulation of nerves can be modeled using a simple one‐compartment PK model.Isolated tracheal rings from CF‐1 wild‐type mice were exposed to EFS pulses known to activate neurotransmitter release (1.0 ms duration, 80 V, 1–30 Hz). Rings were pre‐treated with 100 μM capsaicin and 10 μM indomethacin to eliminate the effects of sensory nerves and endogenous prostaglandins, respectively. Contractile responses were blocked by the muscarinic ACh receptor antagonist atropine. These responses were modeled using the open‐source R package RxODE. The one‐compartment PK model was fit using three unknown parameters: absorption constant (ka,), elimination constant (ke), and dose.EFS caused atropine‐sensitive contraction, consistent with activation of airways by ACh released from parasympathetic nerves. Iterative solving of the differential equations defined by the model produced best‐fit parameters that accurately described the experimental data. From these data, the model was re‐run using global average values calculated to fit all samples at all frequencies to determine model accuracy (ka = 0.1597 s−1, ke = 0.8242 s−1, dose = 1.5 × 10−6 M). The average error across all samples and frequencies was 23.5 ± 3.87%. Stimulation at 15 Hz had the lowest average error of 5.91 ± 2.23%. A separate model incorporating presynaptic inhibition at muscarinic‐2 receptors and saturation of ACh metabolism had a lower average error 21.7 ± 3.52%, but was not statistically (p > 0.05) different from the simple one‐compartment model.The one‐compartment PK model, based only on first‐order absorption and elimination constants, was a reasonable model to approximate the experimental data. Future in vitro and computational experiments will expand the model to incorporate sensory nerve activity. A more complete model may provide new insights into potential mechanisms of neuronal control of airways, allow for the rational use of drugs in treating lung diseases, and enable in silico assessment of novel therapeutics.Support or Funding InformationState of Nebraska Research Fund Grant LB595This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.