Figure 6. AQP1 is localized in proximity to mechanosensitive cation channels, and its deletion or mechanosensitive cation channel blockade attenuates extracellular calcium entry, PS exposure, and microvesiculation.

(A) Human platelets were stained with antibodies against AQP1, TRPC6, and Piezo1, and images were evaluated for colocalization. Pearson’s correlation analysis demonstrates AQP1 to be in proximity to TRPC6 and Piezo1 but not in significant proximity to TRPC1. (B) AQP1 interaction with TRPC6 in human platelets after evaluation by Förster resonance energy transfer (FRET). (C and D) Washed mouse platelets in modified Tyrode supplemented with 1 mM calcium were labeled with the calcium indicator Fluo-4 (green) and Alexa Fluor 568–annexin V (AnxV, red) and allowed to adhere to collagen-coated surfaces. (C and D) Platelets superimposed for AnxV and Fluo-4 at time points after adherence to collagen corresponding to the numbered arrows in the graphs below. Data are for AQP1^+/+ (magenta) and AQP1^–/– (blue) platelets. Box-and-whisker plots of peak change and area under the curve (AUC__{15 min}) of calcium responses in untreated platelets (NoTrm) and in platelets pretreated with the cytosolic calcium chelator BAPTA-AM (1 μM), extracellular calcium chelator EGTA (1 mM), and sarco/endoplasmic reticulum Ca²⁺ ATPase (SERCA) inhibitor thapsigargin (1 μM) in AQP1^+/+ and AQP1^–/– platelets. Similarly, the peak change and AUC of annexin V binding were quantified and shown. (E) Mouse platelets were preincubated with the mechanosensitive channel inhibitor Grammostola spatulata mechanotoxin 4 (GsMTx-4) or vehicle as a control; outcome is shown as in C and D. (F) Microvesiculation was quantified in control (untreated) and GsMTx4-treated AQP1^+/+ mouse platelets. Data analysis was performed by Wilcoxon signed-rank test. *P < 0.05, **P < 0.01 were considered significant. Scale bar: 1 μm (A and B). Data were from 4 independent experiments.