Reply to Davis: Nitric oxide regulates lymphatic contractions

We appreciate the opportunity to respond to Michael J. Davis’s Letter (1) regarding our paper “Mechanobiological Oscillators Control Lymph Flow” (2).

Our model was developed to study phasic contractions and the dynamics of NO and calcium rather than the control of basal tone. Many factors—in addition to NO—affect lymphatic baseline diameter, including inflammatory cytokines and hormones, and the mechanical properties of the surrounding tissue. On the other hand, there are few mechanisms that operate at the short timescales required for the phasic contractions. Our simulations show that NO and Ca²⁺ are sufficient to drive these phasic contractions and that mechanobiological regulation of their levels provides intrinsic feedback control.

Davis’s primary criticism is directed at data published by Liao et al. (3). As with any field, there are discrepancies in the lymphatic literature. For example, some studies show an increase in contraction frequency with NO inhibition (4), whereas others report a decrease (5). Comparing eNOS^−/− mice to wild-type mice in vivo, Liao et al. show an increase in end diastolic diameter of the popliteal lymphatic vessel (PLV). A similar trend, although apparently not statistically significant, was observed for isolated lymphatic vessels ex vivo (figure 5A of ref. 5). The study by Liao et al. was the first analysis of lymphatic contractions in vivo in eNOS^−/− mice and showed general agreement with ex vivo data. However, PLV diameter in vivo is 35–40 μm (3), but 60–70 μm when this vessel is removed from its supporting tissue and cannulated (5). Thus, removing the vessel for ex vivo manipulation introduces differences in tone that may account for some of the discrepancies pointed out by Davis.

Regarding the dependence of pumping on pressure in the absence of NO production, in vivo experiments have not yet been performed, so the issue has not already “been tested and refuted.” However, in excised vessel studies, there is general agreement that when NO is blocked, pump performance is biphasic, with the highest output at intermediate transwall pressures (5, 6). Our simulations predict this qualitatively (figure 3F of ref. 2). The pumping observed in excised vessels at low and high pressures (where our simulations predict no contractions) may be explained by extrinsic random perturbations that trigger the action potentials in the absence of stretch-induced calcium release. This can be shown by investigating the stability of the static solutions.

Finally, the suggestion that our model only “recapitulates primarily the in vivo behavior upon which it is based” is not accurate. The model was not developed by curve fitting experimental data nor does it comprehensively match all of the data reported by Liao et al. It was formulated from the bottom-up based on known biochemical mechanisms underlying lymphatic physiology. With no parameter adjustments, it reproduces the general dynamics in response to relevant physical perturbations reported in the literature over the past 40 y. The model provides a unifying, analytical framework for lymphatic contractions and can be further refined as more information about lymphatic physiology in vivo becomes available.