CD8+ T cells stimulate Na-Cl co-transporter NCC in distal convoluted tubules leading to salt-sensitive hypertension

Abstract

Recent studies suggest a role for T lymphocytes in hypertension. However, whether T cells contribute to renal sodium retention and salt-sensitive hypertension is unknown. Here we demonstrate that T cells infiltrate into the kidney of salt-sensitive hypertensive animals. In particular, CD8⁺ T cells directly contact the distal convoluted tubule (DCT) in the kidneys of DOCA-salt mice and CD8⁺ T cell-injected mice, leading to up-regulation of the Na-Cl co-transporter NCC, p-NCC and the development of salt-sensitive hypertension. Co-culture with CD8⁺ T cells upregulates NCC in mouse DCT cells via ROS-induced activation of Src kinase, up-regulation of the K⁺ channel Kir4.1, and stimulation of the Cl^- channel ClC-K. The last event increases chloride efflux, leading to compensatory chloride influx via NCC activation at the cost of increasing sodium retention. Collectively, these findings provide a mechanism for adaptive immunity involvement in the kidney defect in sodium handling and the pathogenesis of salt-sensitive hypertension.

Figure 1. T cells accumulate in the kidney of DOCA mice and are associated with NCC up-regulation.

(a) Immunostaining of pan T cell marker CD3 (brown, upper panels, scale 50 μm) or NCC (green, lower panels, scale 100 μm) on Sham or DOCA-salt treated mouse kidney sections. Nuclei were stained by hematoxyline (upper panels) or DAPI (lower panels). Data are representative of n=8 images in each group. (b) Left panel, isolated mouse splenic pan T cells were stained by CD3 antibody. Flow cytometry confirmed all cells were CD3⁺ T cells. Cells in grey closed area are T cells without CD3 staining as negative control; cells in red dashed line open area are T cells stained by CD3 antibody, indicating CD3⁺ T cells. Right panel, effects of using magnetic beads to remove T cells from the mDCT-T cell co-culture. Control mDCTs (before co-culture, blue closed area) and after co-culture mDCTs (after removal of T cells by using magnetic beads, red dashed line open area) were stained with CD3 antibody and analysed by flow cytometer. Data are representative of three independent tests. (c) Western blot analysis of NCC expression in mDCT cells with (+T cell) or without (Con) mouse splenic T cell-treatment. Two bands were detected for NCC, reflective of mature (upper) and immature (lower) forms (see methods). Quantitative western blot data were normalized using GAPDH as a loading control. n=4–5 in each group. Data are means±s.e. P<0.01 (t-test).

Figure 2. CD8 subtype T cells are involved in the up-regulation of NCC.

(a) Expression of NCC in mDCT cells treated with T cells from spleens of Sham (Sham T+) or DOCA (DOCA T+) mice. After removal of all T cells, whole cell lysates were used for western blot. GAPDH was used as loading control. n=4 in each group. Data are means±s.e. P<0.05 (t-test). (b) Pan T cells isolated from spleens of Sham or DOCA mice were analysed for the proportion (%, red numbers) of CD4 and CD8 subtype by flow cytometry. Data are representative of three independent experiments. (c) mDCT cells with (red) or without (grey, negative control for auto-fluorescence) double staining using MHC-I (Y axis) & MHC-II (X axis) antibodies were analysed by flow cytometry. Data are representative of four independent experiments. (d) Western blot analysis of NCC expression in mDCT cells without or with CD4⁺ or CD8⁺ mouse splenic T cell-treatment. Quantitative western blot data were normalized using GAPDH as a loading control. n=4–5 in each group. Data are means±s.e. P=0.46 Con versus +CD4T; P<0.01 Con versus +CD8T (ANOVA).

Figure 3. Adoptive transfer of CD8⁺ T cells to mice results in salt-sensitive hypertension.

(a) Immunostaining of pan T cell marker CD3 (brown, upper level, scale 50 μm) or NCC (green, lower level, Scale, 100 μm) in kidneys from sham mice or mice receiving CD8⁺ T cell-adoptive transfer (+CD8 T cell). Data are representative of n=10 images in each group. (b) Western blot analysis of p-NCC expression in the kidneys of sham mice and mice receiving CD8⁺ T cell-adoptive transfer (+CD8 T). Quantitative western blot data were normalized using GAPDH as a loading control. n=5–6 in each group. Data are means±s.e. P<0.01 (t-test). (c) Averaged radio-biotelemetry recording of systolic blood pressure in 5 mice with saline injection (sham, black line) and 10 mice receiving adoptive transfer of CD8⁺ T cells (+CD8Ts, red line). All mice were given high-salt diet and HCTZ drinking water as indicated. Blood pressure was recorded hourly. Data are means±s.e. (d) At day 13 (last day of HCTZ treatment) 4 mice from +CD8 T group were removed from the blood pressure recording and their kidneys were used for immunostaining for CD3. Scale, 50 μm.

Figure 4. CD8⁺ T cells form close contacts with distal convoluted tubules in mouse kidney sections.

(a) A wide-field image of a kidney section from a mouse injected with CD8⁺ T cells. Scale 10 μm. Antibodies against NCC (green), CD8 (red), and Na⁺-K⁺-ATPase-α (NKA, imaged far red and pseudo-colored cyan for visual clarity) were used. (b) The same CD8⁺ T cell from (a) processed with 3D structured illumination microscopy (3D-SIM). Scale, 2 μm. Four colour channel images were overlaid: DAPI (blue), NCC (green), CD8 (red), and NKA (cyan). The yellow circle denotes the putative contact point. (c) Orthogonal sections of the 3D image (XZ, XY, YZ) of the same cell. Yellow lines on each section correspond to the position of other orthogonal sections centred around the contact point. (d) A wide-field image of a kidney section from a DOCA-salt mouse. Identically stained as in a. Scale 10 μm. (e) 3D-SIM processed image of (d). Scale 2 μm. Yellow arrow represents the location and direction of intensity profile. (f) Fluorescence intensity profile of CD8 (red) and NKA (cyan) along the yellow arrow in e. (g) Three-dimensional surface rendering of the image shown in e. The yellow circle denotes the putative contact point. Data are representative of 11 images from +CD8 T cell group and 13 images from DOCA group.

Figure 5. TKs (a CD8⁺ T cell line) upregulate NCC in mDCTs via direct cell-cell contact.

(a) Image of TK cells adherent to mDCTs. Scale, 40 μm. After overnight co-culture, cells were washed with PBS × 2. Data are representative of n=10 images in each group. (b) Orthogonal section of the 3D image (XZ, XY, YZ) of individual cell-cell contact of TK (CD8, red) and mDCT (Na-K-ATPase, NKA, green) cells in co-culture for 4 h. Scale, 2 μm. Data are representative of n>15 images. All images were captured and processed with super-resolution 3D-SIM microscopy. (c) Total protein expression of NCC and p-NCC in mDCTs with or without TK co-culture. GAPDH was used as a loading control. n=4–6 in each group. Data are means±s.e. *P<0.01 versus Control (t-test) (d) Upper, effects of 0.4 and 8 μm transwells on TK-mediated up-regulation of NCC and p-NCC in mDCT membrane protein. Lower, effects of transwells on TK-mediated up-regulation of NCC and full length SPAK expression in mDCT total lysate. All groups with or without transwells used the same culture conditions. Na-K-ATPase or GAPDH were used as loading controls of membrane protein or total protein, respectively. n=4 in each group. Data are means±s.e. *P<0.01 versus Control (ANOVA).

Figure 6. TKs induce more mDCTs to exhibit high sodium uptake ability via NCC.

(a) Flow cytometry study using intracellular sodium indicator CoroNa Green. All groups were in presence of ouabain (500 μM), bumetanide (100 μM) and amiloride (100 μM). Negative control without CoroNa Green represents auto-fluorescence of mDCT cells. x-axis, side scatter (SSC); y-axis, CoroNa Green fluorescence. Data are representative of twelve independent experiments. (b) Change (Δ) in the proportion of cells with elevated sodium. n=12 groups. P<0.01 (t-test). (c) Effects of NCC knockdown by siRNA (siNCC) on TK-induced sodium uptake in mDCTs. x-axis, CoroNa Green fluorescence, y-axis, cell number count. Data are representative of four independent experiments. (d) Ratio quantification of TK-induced Na⁺ uptake in mDCTs with or without NCC inhibitor HCTZ (200 μM) in regular PBS incubation (PBS) or in incubation with additional salt (+40 mM NaCl). n=6–8 in each group. Data are means±s.e. P<0.01 (ANOVA). NS, not significant.

Figure 7. Effects of SPAK or ClC-K knockdown on NCC and intracellular chloride.

(a) SPAK knockdown using siRNA (siSPAK) prevented the up-regulation of NCC & p-NCC in TK-treated mDCTs. Total lysate protein loading was normalized by GAPDH. n=4–6 in each group. Data are means±s.e. *P<0.01 versus Control; ^#P<0.01 versus sham+TK (ANOVA). (b) Membrane expression of ClC-K was detected by western blot. The binding of membrane ClC-K to its subunit barttin was detected by immunoprecipitation (IP) of barttin and immunoblot (IB) of ClC-K. Membrane protein loading for both western blot and IP/IB were normalized to Na-K-ATPase (see Methods). n=4–6 in each group. Data are means±s.e. *P<0.01 versus Control (t-test). (c) ClC-K knockdown using siRNAs (siClC-K) prevented the reduction of intracellular chloride in TK-treated mDCTs. Intracellular chloride measured with MQAE. The difference between PBS and HCTZ+Bume of each line reflects NCC/NKCC compensated chloride efflux in each group of mDCTs. TK-induced chloride efflux (in presence of HCTZ+Bume) was compared between the groups without siClC-K (Sham-si, Dashed lines) and with siClC-K (solid lines). Data are means±s.e. n=6–10 in each group. P<0.01 (t-test). (d) NCC-mediated sodium uptake in control or TK-treated mDCTs with or without ClC-K knockdown was measured using the same method as in Fig. 5. Data are representative of nine (sham-si) to ten (siClC-K) experiments. (e) TK-induced change (Δ) in the proportion of cells with elevated sodium. n=9–10 groups. Data are means±s.e. P<0.01 (t-test).

Figure 8. Effects of ClC-K or Kir4.1 knockdown on TK-mediated NCC-regulating pathway in mDCTs.

(a) In membrane proteins (upper), ClC-K knockdown prevented TK-induced membrane expression of p-NCC and NCC but did not affect the up-regulation of Kir4.1. n=4–6 in each group. Data are means±s.e. *P<0.01 versus sham control; ^#P<0.01 versus sham+TK; ^†P<0.01 versus siClC-K Con (ANOVA). In total protein lysates (lower), TK-mediated up-regulations of SPAK and NCC were also diminished by siClC-K administration. n=4–5 in each group. Data are means±s.e. *P<0.01 versus sham control; ^#P<0.01 versus sham+TK (ANOVA). TK-induced (Δ) increase in total NCC expression was compared between sham-si and siClC-K groups (Dashed lined area). n=4–5 in each group. Data are means±s.e. P<0.01 versus sham-si (t-test). Na-K-ATPase or GAPDH were used as loading controls of membrane protein or total protein, respectively. (b) Kir4.1 knockdown prevented TK-induced up-regulation of ClC-K and p-NCC in mDCT cell membrane (upper) and NCC in total lysate (lower). Na-K-ATPase or GAPDH were used as loading controls for membrane protein or total protein, respectively. n=4–6 in each group. *P<0.01 versus Control; ^#P<0.01 versus sham+TK (ANOVA).

Figure 9. Effects of Src knockdown or blocking ROS on TK-induced NCC up-regulation in mDCTs.

(a) Knockdown of Src using siRNA prevented TK-induced up-regulation of Kir4.1 & p-NCC on cell membrane (upper) and SPAK and NCC in total lysate (lower) of mDCT cells. Na-K-ATPase or GAPDH were used as loading controls for membrane protein or total protein, respectively. n=4 in each group. Data are means±s.e. *P<0.01 versus Control; ^#P<0.01 versus sham+TK (ANOVA). (b) intracellular ROS indicator CM-DCF pre-loaded mDCTs with or without TK co-culture. CM-DCF, green; DIC, bright field; Merge, green+bright field. Images were taken using Axio observer microscopy. Scale 40 μm. Data are representative of 9 images in control group and 15 images in +TK group. (c) NADPH oxidase inhibitor Apocynin (100 μM) blocked TK-induced activation of Src (p-srcY419) and up-regulation of SPAK and NCC in mDCTs. Total Src (t-Src) was unaffected. GAPDH was used as loading control. n=4–5 in each group. Data are means±s.e. *P<0.01 versus Control; ^#P<0.01 versus +TK (ANOVA).

Figure 10. A hypothetical mechanism in kidney for the pathogenesis of salt-sensitive hypertension.

Kidney infiltrated CD8⁺ T cells directly contact DCT cells on the basolateral side; stimulate DCTs via ROS-Src signaling; enhance salt retention via the Kir4.1-ClC-K—SPAK—NCC activation pathway and consequently lead to the development of salt-sensitive hypertension. Red solid arrows indicate the suggested pathway in the present study; blue solid arrows show up- or down- regulation of intracellular sodium or chloride.