LETTER

Top Abstract Letter References

Shear stress-induced dilation of arterioles

These findings support the idea that arterioles are sensitive to changes in wall shear stress and that this mechanism could be important in the negative feedback regulation of wall shear stress and the regulation of blood flow in the microcirculation. This idea was further tested in arterioles (20-200 µm in diameter) in vivo by increases in viscosity (8) and in vitro by increases in flow (5, 6) and viscosity (5) of the perfusate (no occlusions have been used in these studies), providing further evidence for the presence of a flow-/shear stress-sensitive mechanism in arterioles.

McGahren et al. (7) attempted to repeat the in vivo studies, employing parallel occlusion "to test the hypothesis that flow-dependent changes dominate control of the microvasculature" [emphasis added]. The authors reported that they were unable to increase wall shear stress (increased flow velocity before a change in diameter) in the nonoccluded vessels investigated, yet they observed dilation in these vessels which reduced calculated wall shear stress. Thus they concluded that the arteriolar dilations in response to occlusion are not related to increases in shear stress. Providing no room for speculation of possible causes for being unable to increase wall shear stress in their experimental conditions, the authors imply that flow-/shear stress-dependent regulation in microvessels is nonexistent and, without offering any evidence, postulate that the dilation after occlusion is caused by a conducted response generated by hypoxia or ischemia in downstream regions.

It is well known that several important, redundant mechanisms govern the diameter of microvessels. Therefore, it is obvious that if one mechanism is not activated, others may come into play. It is likely that placing an occluder in a microcirculatory network can cause several divergent changes in pressure and flow, eliciting a variety of responses dependent on the site of the occlusion and whether the various branch vessels studied are upstream or downstream, serial or parallel connected, etc. It follows that a dilation or constriction is not necessarily achieved by identical mechanisms in all experimental situations. Thus, if one wishes to uncover a mechanism that may have a role in the regulation of vascular diameter, one should select optimal conditions for its demonstration.

On the basis of published data, we point out some of the differences between previous studies and those of McGahren et. al. (7) which may suggest reasons for their inability to demonstrate flow (shear stress)-dependent dilation of arterioles.

First, because McGahren et al. (7) did not increase blood flow velocity prior to a change in diameter, they could not test the idea of whether an increase in wall shear stress is followed by an arteriolar vasodilation. Thus the cornerstone for drawing a conclusion regarding flow velocity-/shear stress-induced dilation is lacking. Had wall shear rate/stress increased and had there still been no increases in vessel diameter, the authors could have concluded that the arterioles are not sensitive to increases in shear stress. Also, no one, as far as we know, has suggested that flow-dependent responses "dominate" the control of the microcirculation, a hypothesis the authors set out to disprove.

Second, a lack of the necessary time resolution in the bar graphs presented makes it impossible to discern which parameter changed first, flow velocity or diameter, thus rendering the data unsuitable for testing the authors' hypothesis. It has been shown previously that arterioles respond within the first 5-20 s to increases in flow velocity/shear stress (2, 3, 10). Thus data pertaining to the first 20 s of the responses should have been analyzed. At 2 or 3 min after occlusion, wall shear stress is already regulated by a change in diameter and one cannot ascertain the sequence of events, i.e., which changed first, diameter or shear stress. Consequently, data included in Figs. 2 and 5-7, depicting changes that occurred after 2 min of occlusion, are irrelevant with regard to wall shear stress-induced dilation. Similarly, data presented in Fig. 3 and Table 2 cannot be evaluated, because correlative flow velocity data are not given.

Third, the only data set that has the time resolution suitable for any evaluation is depicted in Fig. 8. We believe that all data should have been analyzed in this time-dependent manner, as was done previously (3, 4, 10). Nevertheless, there are intriguing questions that arise as one examines the raw data in Fig. 8. The data reveal that, although the diameter of a "branch" vessel (Fig. 8A) increased by an estimate of 60% and blood flow velocity nearly doubled, blood flow essentially did not change. There does not seem to be any apparent explanation for this finding. In addition, the calculated wall shear stress in this vessel was 160 dyn/cm², an inordinately high, nonphysiological level. Furthermore, according to the recordings shown, blood flow velocities, even before occlusion, are not any different in a 7-µm "capillary-size" vessel than in a 90-µm "first order-size" feed vessel. These data are not only incomprehensible but also challenge elementary hemodynamic principles.

Fourth, the authors also did not provide any evidence to indicate whether, in their experiments, the endothelial cell layer of the vessels studied was intact and functional, a condition necessary for the expression of shear stress-dependent responses. McGahren et al. (7) suggest a role for conducted vasodilation in the observed responses, initiated by "hypoxia or ischemia in downstream regions normally perfused by the occluded vessels." If so, they should have been able to attenuate or eliminate the response by the use of uncouplers of gap junctions, as the senior author's laboratory has previously demonstrated (9).

Fifth, it is important to emphasize that even if, in a given experimental condition, dilation of a vessel in response to occlusion cannot be assigned to an increase in wall shear stress, this would not necessarily preclude the presence of shear stress-sensitive mechanisms in the vessel. For example, administering adenosine into an arteriole would elicit a great increase in diameter, which can cause a decrease in calculated wall shear stress. This finding, however, by no means would negate the presence of shear stress-sensitive mechanisms in this vessel. Also, arterioles of different sizes of a variety of tissues have been shown, by now, to be sensitive to changes in shear stress (1).

Finally, we believe, on the basis of an entire body of published experimental data (1), that arterioles, similar to large conduit-type vessels, are sensitive to changes in shear stress, providing for flow-/shear stress-dependent regulation of microvascular resistance.