InterPore2025

Can asymmetric pores rectify fluid flow in the brain?

20 May 2025, 11:50

15m

Oral Presentation (MS20) Biophysics of living porous media MS20

Speaker

Douglas Kelley (University of Rochester)

Description

Water-like fluids move around and through brain tissue, sweeping away metabolic wastes whose accumulation correlates with diseases like Alzheimer's and Parkinson's. Flow is driven, at least in part, by the dilation and constriction of arteries that lie within annular perivascular spaces filled with water-like cerebrospinal fluid. Naively, we would expect the oscillatory motion of artery walls to drive oscillatory flows, pushing cerebrospinal fluid out of the perivascular space when the artery dilates and drawing it back in when the artery constricts. In vivo experiments reveal just such an oscillatory flow -- but accompanied by a fast, directed motion in the same direction as blood flow, deeper into brain tissue. The mechanism driving that directional flow remains unknown. In other bodily fluid transport systems, directional flow is ensured by valves, as in the heart, lymph vessels, and veins. Several researchers have speculated that artery-driven pulsation could be rectified by valve-like structures. More specifically, it may be that the narrow gaps between the astrocyte endfeet, through which fluid probably passes as it transits from perivascular spaces into the surrounding tissue, have asymmetric structures and rectify flow.

We have explored that possibility through simulations and experiments. Experiments and corresponding theory confirm that a conical pore in an elastic membrane favors one flow direction over the other. Simulations confirm that a wedge-shaped gap likewise rectifies flow, that realistic sizes and stiffnesses might produce the flows observed in vivo, and that such a valve would likely work well for frequencies associated with respiration, heart beats, and rerouting of blood (functional hyperemia), but not for frequencies associated with ultrasound. Rectification is most efficient when endfeet are neither too thick, nor too thin. Simulations also show that wedge-shaped gaps are not the only possible valves. Overlapping endfeet of unequal lengths can rectify flow, as can the convex shape of the endfoot wall. The three mechanisms are not mutually exclusive. I will close with brief comments about open problems in brain fluid flow related to pores and poroelasticity.