AQP4-independent TRPV4 modulation of plasma membrane water permeability

Abstract

Despite of the major role of aquaporin (AQP) water channels in controlling transmembrane water fluxes, alternative ways for modulating water permeation have been proposed. In the Central Nervous System (CNS), Aquaporin-4 (AQP4) is reported to be functionally coupled with the calcium-channel Transient-Receptor Potential Vanilloid member-4 (TRPV4), which is controversially involved in cell volume regulation mechanisms and water transport dynamics. The present work aims to investigate the selective role of TRPV4 in regulating plasma membrane water permeability in an AQP4-independent way. Fluorescence-quenching water transport experiments in Aqp4^-/- astrocytes revealed that cell swelling rate is significantly increased upon TRPV4 activation and in the absence of AQP4. The biophysical properties of TRPV4-dependent water transport were therefore assessed using the HEK-293 cell model. Calcein quenching experiments showed that chemical and thermal activation of TRPV4 overexpressed in HEK-293 cells leads to faster swelling kinetics. Stopped-flow light scattering water transport assay was used to measure the osmotic permeability coefficient (Pf, cm/s) and activation energy (Ea, kcal/mol) conferred by TRPV4. Results provided evidence that although the Pf measured upon TRPV4 activation is lower than the one obtained in AQP4-overexpressing cells (Pf of AQP4 = 0.01667 ± 0.0007; Pf of TRPV4 = 0.002261 ± 0.0004; Pf of TRPV4 + 4αPDD = 0.007985 ± 0.0006; Pf of WT = 0.002249 ± 0.0002), along with activation energy values (Ea of AQP4 = 0.86 ± 0.0006; Ea of TRPV4 + 4αPDD = 2.73 ± 1.9; Ea of WT = 8.532 ± 0.4), these parameters were compatible with a facilitated pathway for water movement rather than simple diffusion. The possibility to tune plasma membrane water permeability more finely through TRPV4 might represent a protective mechanism in cells constantly facing severe osmotic challenges to avoid the potential deleterious effects of the rapid cell swelling occurring via AQP channels.

Keywords: AQP4; TRPV4; astrocytes; calcium ions; cell swelling; ion channels; water channels.

FIGURE 1

Graphical summary of the design protocols. Control condition in the absence (A) and in the presence (B) of calcium in the external solution. No intracellular calcium rises are promoted. Activation of TRPV4 by chemical agonist 4αPDD in external medium without (C) and with external calcium (D). Activation of TRPV4 by chemical agonist 4αPDD in external medium without (E) and with (F) external calcium upon thapsigargin stimulation.

FIGURE 2

Water transport assay in Aqp4^–/– and WT astrocyte primary cultures. (A) Representative time courses of calcein-AM loaded astrocytes upon exposure to 60 mOsm/L hypotonic gradient at 20°C in the presence or absence of external calcium with 4αPDD and/or thapsigargin (TG) incubation. (B) Bar graph showing the time constants (τ) of cell swelling measured in the conditions reported in (A). Data are reported as mean ± SEM and representative of three independent experiments (Table 1). Different letters were used to indicate significant differences between different cell lines under the same conditions (in italics: comparison between controls; in bold: comparison between activated cells). One-way ANOVA with Newman-Keuls multiple comparison tests for WT (ns, p > 0.05) and Kruskal-Wallis with Dunn’s multiple comparison test for Aqp4^–/– astrocytes for comparisons between conditions within each genotype; Kruskal-Wallis with Dunn’s multiple comparison test (in italics, p < 0.0001) and One-way ANOVA with Newman-Keuls multiple comparisons test (in bold, p < 0.0001) for comparison between genotypes under control and activating conditions, respectively. Detailed statistically significant differences are shown in Supplementary Tables 3–5. (For Aqp4^–/–: Ctrl with Ca²⁺ n = 10; Ctrl w/o Ca²⁺ n = 10; 4αPDD with Ca²⁺ n = 12; 4αPDD w/o Ca²⁺ n = 11; TG + 4αPDD n = 16; TG w/o Ca²⁺ n = 10. For WT: Ctrl with Ca²⁺ n = 12; Ctrl w/o Ca²⁺ n = 11; 4αPDD with Ca²⁺ n = 9; 4αPDD w/o Ca²⁺ n = 9; TG + 4αPDD n = 9; TG w/o Ca²⁺ n = 9).

FIGURE 3

Water transport assay in Aqp4^–/– astrocytes transfected with TRPV4-siRNA. (A) Representative time courses of calcein-AM loaded astrocytes knockdown for TRPV4 upon exposure to 60 mOsm/L hypotonic gradient at 20°C in the presence or absence of external calcium with 4αPDD and/or thapsigargin (TG) incubation. (B) Bar graph showing the time constants (τ) of cell swelling measured in the conditions reported in (A). Data are reported as mean ± SEM (Table 2). Kruskal-Wallis with Dunn’s multiple comparison test for comparison between conditions. Detailed statistically significant differences are shown in Supplementary Table 6. (C) Western Blot showing TRPV4 expression (band at ∼95 kDa) in WT and TRPV4-overexpressing cells revealed by anti-TRPV4 antibody and GFAP (band at ∼50 kDa) as astroglial marker. On the right, bar graph summarizing the densitometric analysis of TRPV4 relative quantification in Aqp4^–/– astrocytes knockdown for TRPV4 with siRNA and a scramble siRNA used as control. Data are expressed as means ± SEM and representative of three independent experiments (unpaired t-test, ****p < 0.0001). (For Aqp4^–/– with scramble: Ctrl with Ca²⁺ n = 10; 4αPDD with Ca²⁺ n = 9. For Aqp4^–/– with TRPV4-siRNA: Ctrl with Ca²⁺ n = 9; 4αPDD with Ca²⁺ n = 11; TG w/o Ca²⁺ n = 10).

FIGURE 4

Characterization of AQP4-mCherry and TRPV4-eGFP stably transfected HEK-293 cells. (A) Epifluorescence images of AQP4-mCherry (red) and TRPV4-eGFP (green) expression in stably transfected HEK-293 cells. DAPI is in blue (scale bar 50 μm). (B) Western Blot analysis of AQP4 and TRPV4-overexpressing cells and WT, as indicated, revealed by anti-AQP4, anti-GFP, and anti-TRPV4 antibodies (Ab), respectively. Left: AQP4 expression in AQP4-overexpressing cells revealed as two bands at ∼60 kDa for recombinant AQP4 and at ∼32 kDa for untagged. No AQP4 expression was detected in WT cells. Right: TRPV4 expression revealed as one band at ∼130 kDa for recombinant TRPV4 in TRPV4-overexpressing cells with anti-GFP and anti-TRPV4 antibodies and one band at ∼95 kDa for native TRPV4 in both transfected and WT cells detected with anti-TRPV4 antibody. (C) Superimposed kinetics of calcium responses expressed as a ratio of fluorescence at 340 and 380 nm (F340/F380) under hypotonic stimulus performed by FlexStation3 in Fura2-AM loaded cells at 37°C. (D) Bar graph showing calcium amplitude [Δ(F340/F380)] for the three cell lines. Different letters indicate significant differences between the indicated cell lines. Data are reported as mean ± SEM and representative of three independent experiments (AQP4 = 0.2315 ± 0.01639, n = 25; TRPV4 = 0.2012 ± 0.01187, n = 22; WT = 0.09547 ± 0.01245, n =8). Brown-Forsythe ANOVA test and Tamhane’s T2 multiple comparisons test (p < 0.0001). Detailed statistically significant differences are shown in Supplementary Table 7.

FIGURE 5

Calcein-AM quenching water transport assay in WT, AQP4, or TRPV4-overexpressing HEK-293 cells. (A) Representative water transport kinetics, showing swelling and RVD phase at 20°C under control conditions or stimulation by 10 μM 4αPDD. (B) Bar graph showing the time constant (τ) for the swelling phase at 20°C. Data are reported as mean ± SEM and representative of four independent experiments. For comparisons within each cell line: unpaired t-test with Welch’s correction (ns, p > 0.05; **p < 0.01) and Brown-Forsythe ANOVA test and Tamhane’s T2 multiple comparisons test between cell lines for TRPV4 at 20°C (p < 0.001) for comparisons per condition between cell lines: Brown-Forsythe ANOVA test and Tamhane’s T2 multiple comparisons test for controls at 20°C and (in italics, ns, p > 0.05; p < 0.0001) and activated cells (in bold, p < 0.0001). Detailed statistically significant differences are shown in Supplementary Tables 8–10 (WT: Ctrl n = 18; 4αPDD n = 19. TRPV4: Ctrl n = 24; 4αPDD n = 24; RN-1734 n = 9. AQP4: Ctrl n = 13; 4αPDD n = 9). (C) Representative water transport kinetics, showing swelling and RVD phase at 37°C. (D) Bar graph showing the time constant (τ) for the swelling phase at 37°C. Data are reported as mean ± SEM and representative of three independent experiments (WT: n = 18; TRPV4: n = 29; AQP4: n = 33). Brown-Forsythe ANOVA test and Tamhane’s T2 multiple comparisons test (p < 0.0001). Detailed statistically significant differences are shown in Supplementary Table 11.

FIGURE 6

Biophysical parameters (Pf and Ea) associated with plasma membrane water transport in TRPV4-overexpressing cells measured by stopped-flow light scattering. (A–C) Representative curves for the time course of scattered light intensity at 10, 20, and 37°C in response to a 100 mOsm/L outwardly directed osmotic gradient resulting in cell swelling (decreased scattered light intensity). Data show AQP4 and TRPV4-expressing cells, where the TRPV4 channel was activated by 4αPDD agonist (A,B) or thermally (C). Each of the time courses is plotted within the same y-range. (D) Osmotic water permeability coefficient (Pf) computed from kinetics at 20°C. The ranges of Pf values that predict facilitated water movement by molecular pores (>0.01 cm/s) or consistent with water diffusion across a lipid portion of a membrane (<0.005 cm/s) are evidenced in the bar plot with blue and gray lines, respectively. Data are represented as the mean ± SEM and representative of three independent experiments. Brown-Forsythe ANOVA test and Tamhane’s T2 multiple comparisons test (p < 0.0001). Detailed statistically significant differences are shown in Supplementary Table 12. (E) Arrhenius plot of temperature-dependence data for the three cell lines, as indicated.

FIGURE 7

Western-Blot analysis of aquaporins in mouse astrocytes and HEK-293 cells. Western blot analysis of aquaporins (from 1 to 9) in WT and TRPV4-overexpressing HEK-293 cells and WT and Aqp4^–/– astrocytes. Protein samples (10 μg) were loaded as follows: lane 1-2, HEK-293 WT; lane 3-4, TRPV4-overexpressing HEK-293; lane 5, WT astrocytes; lane 6 Aqp4^–/– astrocytes. For WB of AQP1, AQP3, AQP4, AQP6-AQP9: lane 7, mouse brain; lane 8, mouse kidney; lane 9, mouse liver. For WB of AQP2: lane 7, mouse kidney; lane 8 (2.5 μg) and lane 9 (5 μg), M1 mouse kidney cortical collecting duct cells stably transfected with human AQP2 (MCD4) for AQP2 (Milano et al., 2018). For WB of AQP5: lane 7, mouse lung; lane 8, mouse kidney; lane 9, mouse liver. Coomassie Blue staining was used as loading control for Western Blot (Supplementary Figure 4).