Direct demonstration of tubular fluid flow sensing by macula densa cells

Abstract

Macula densa (MD) cells in the cortical thick ascending limb (cTAL) detect variations in tubular fluid composition and transmit signals to the afferent arteriole (AA) that control glomerular filtration rate [tubuloglomerular feedback (TGF)]. Increases in tubular salt at the MD that normally parallel elevations in tubular fluid flow rate are well accepted as the trigger of TGF. The present study aimed to test whether MD cells can detect variations in tubular fluid flow rate per se. Calcium imaging of the in vitro microperfused isolated JGA-glomerulus complex dissected from mice was performed using fluo-4 and fluorescence microscopy. Increasing cTAL flow from 2 to 20 nl/min (80 mM [NaCl]) rapidly produced significant elevations in cytosolic Ca(2+) concentration ([Ca(2+)](i)) in AA smooth muscle cells [evidenced by changes in fluo-4 intensity (F); F/F(0) = 1.45 ± 0.11] and AA vasoconstriction. Complete removal of the cTAL around the MD plaque and application of laminar flow through a perfusion pipette directly to the MD apical surface essentially produced the same results even when low (10 mM) or zero NaCl solutions were used. Acetylated α-tubulin immunohistochemistry identified the presence of primary cilia in mouse MD cells. Under no flow conditions, bending MD cilia directly with a micropipette rapidly caused significant [Ca(2+)](i) elevations in AA smooth muscle cells (fluo-4 F/F(0): 1.60 ± 0.12) and vasoconstriction. P2 receptor blockade with suramin significantly reduced the flow-induced TGF, whereas scavenging superoxide with tempol did not. In conclusion, MD cells are equipped with a tubular flow-sensing mechanism that may contribute to MD cell function and TGF.

Fig. 1.

Direct stimulation of the macula densa (MD) by increased fluid flow. Differential interference contrast (DIC) image of an afferent arteriole (AA)-attached glomerulus and MD preparation in which the tubule segments surrounding the MD were completely removed. The apical surface of MD cells was directly accessible from the bath with a perfusion pipette (PP). Compared with 2 nl/min control perfusion (A), the application of 20 nl/min lateral fluid flow from the PP (0 mM NaCl solution) directly to the MD apical surface produced vasoconstriction of the AA (arrows) and elevations in vascular smooth muscle cell VSMC cytosolic Ca²⁺ concentration ([Ca²⁺]_i; B). Note that in addition to the AA the whole glomerulus contracted as well. Scale is 20 μm. C: representative [Ca²⁺]_i recordings from different VSMCs (indicated by different shades of gray) in the AA during flow-induced tubuloglomerular feedback (TGF). The flow stimulus (20 nl/min) was applied using 0 mM NaCl solution. When the same maneuver was repeated 5 times within 4 min, the response showed no signs of desensitization.

Fig. 2.

The relationship between fluid flow rate at the MD and AA smooth muscle cell [Ca²⁺]_i changes. Data points represent the average ± SE of 3–5 individual measurements at flow rates of 2, 5, 7.5, 10, 12.5, 15, 20, 25, and 30 nl/min from 6 different open MD plaque preparations. The stimulating flow rate with the half-maximal effect was 10.4 nl/min.

Fig. 3.

Summary of the changes in AA smooth muscle cell [Ca²⁺]_i and AA internal diameter (ID) during MD fluid flow-induced TGF using the open MD plaque preparation (A) or the intact cortical thick ascending limb (cTAL)-double perfusion model (B). Flow at the MD was increased from 2 to 20 nl/min. The tissue was bathed in 135 mM NaCl containing Ringer's solution in all preparations. Numbers are shown in each column. All changes are significant (P < 0.05) if not indicated (ns, nonsignificant). A: significant AA vasoconstriction and [Ca²⁺]_i responses were observed regardless of the applied fluid [NaCl]. Even the application of 0 mM [NaCl] perfusion solution induced TGF. B: using 10 mM [NaCl] solution, the application of increased fluid flow from 2 to 20 nl/min evoked significant [Ca²⁺]_i elevations in the AA VSMCs and AA vasoconstriction. When 80 mM [NaCl] solution was used, significantly higher flow-induced [Ca²⁺]_i elevations were observed compared with the effect with 10 mM [NaCl] (*P < 0.05).

Fig. 4.

Immunofluorescence labeling of α-tubulin in the mouse kidney. Each MD cell has one 5- to 8-μm-long primary cilium (red) at the apical membrane. Autofluorescence (green) was used as a background to visualize tissue structure. Scale is 10 μm. G, glomerulus.

Fig. 5.

Stimulation of the MD cells by directly bending their apical cilia with a glass micropipette under no-flow conditions. Representative DIC (A) and fluorescent [Ca²⁺]_i images (fluo-4 labeling: green, D–E) before (A and D) and after (E) TGF was induced by moving a large micropipette (P; 10 μm in diameter) parallel with and close to the MD luminal membrane without touching the apical surface of the cells, as indicated by the arrow in A. B: direct, frontal poking of the MD cell's apical surface was performed using a small PP (2–3 μm in diameter). C: changes in fluo-4 intensity in VSMCs of the AA after repeated mechanical stimuli in the same preparation by cilia bending (as shown in A). In contrast, direct, repeated poking of MD cells with a perfusion pipette produced no changes in VSMC [Ca²⁺]_i. F: summary of the cilia bending data. *P < 0.05, stimulated vs. control; n = 6. Scale is 20 μm.

Fig. 6.

Signaling mechanism of flow-induced TGF. Summary of the changes in AA smooth muscle cell [Ca²⁺]_i and AA ID during flow-induced TGF (increasing fluid flow rate from 2 to 20 nl/min) using the open MD plaque preparation. A: the addition of tempol (100 μM) to the bathing solution did not alter [Ca²⁺]_i and vasoconstrictor responses using either the 10 or 80 mM [NaCl] solution, confirming that these results were independent of the flow-induced generation of reactive oxygen species. B: evidence for the involvement of purinergic signaling in flow-induced TGF. The bath application of suramin (50 μM) significantly decreased the flow-induced [Ca²⁺]_i elevations in AA VSMCs and blocked AA vasoconstriction using either 0 or 80 mM [NaCl] perfusate. All changes are significant (P < 0.05) if not indicated. Numbers are shown in each column.