WNK-SPAK-NKCC1/KCC2 pathway disrupts chloride homeostasis and drives hyperactivation of PVN-CRH neurons in essential hypertension

Summary

Hyperactivity of hypothalamic corticotropin-releasing hormone (CRH) neurons contributes to elevated sympathetic outflow in hypertension, but the underlying mechanisms remain unclear. Here, we demonstrate that activation of the WNK-SPAK-NKCC1/KCC2 pathway disrupts chloride homeostasis in paraventricular nucleus (PVN) CRH neurons, diminishing GABAergic inhibition and driving neuronal hyperactivity in spontaneously hypertensive rats (SHRs). Chemogenetic inhibition of PVN-CRH neurons reduced blood pressure and renal sympathetic nerve activity in SHRs. We found depolarizing shifts in GABA reversal potential (E_GABA) and elevated intracellular chloride concentrations ([Cl⁻]_i) in PVN-CRH neurons of SHRs, associated with increased phosphorylation of NKCC1 (Thr212/Thr217) and KCC2 (Thr1007). Upstream WNK1/4 and phospho-SPAK (Ser373) were upregulated in the PVN of SHRs. Pharmacological inhibition of WNK or SPAK shifted E_GABA toward more hyperpolarized values, reduced PVN-CRH neuronal excitability, and lowered blood pressure in SHRs. These findings identify the WNK-SPAK-NKCC1/KCC2 pathway as a promising therapeutic target for neurogenic hypertension.

Graphical abstract

Highlights

• •GABAergic signaling shifts from inhibitory to excitatory in SHR PVN-CRH neurons
• •Chloride imbalance underlies PVN-CRH neuronal hyperexcitability in hypertension
• •WNK-SPAK-NKCC1/KCC2 signaling drives chloride dysregulation in PVN-CRH neurons
• •Targeting WNK-SPAK signaling reduces sympathetic outflow and blood pressure in SHRs

Cardiovascular medicine; Molecular biology; Neuroscience; Cell biology

Introduction

Essential hypertension affects over one billion people worldwide and remains a leading cause of cardiovascular morbidity and mortality. Excessive sympathetic nerve activation is a hallmark of hypertension, with sympathetic signals originating from critical brain regions.^1,2 The paraventricular nucleus of the hypothalamus (PVN) serves as a key integrative center for sympathetic regulation, sending direct projections to the rostral ventrolateral medulla (RVLM) and the intermediolateral cell column of the spinal cord to modulate sympathetic outflow.^3,4,5

Corticotropin-releasing hormone (CRH) neurons in the PVN play a pivotal role in blood pressure regulation. While traditionally recognized for releasing CRH into the hypophyseal portal system, recent evidence demonstrates that PVN-CRH neurons also modulate sympathetic activity through local paracrine mechanisms, potentiating the excitability of neighboring presympathetic neurons.^6,7 Central administration of exogenous CRH elevates mean arterial pressure (MAP) and heart rate (HR),^8,9,10 and CRH knockdown in the PVN significantly reduces neuronal hyperexcitability.⁷ However, the cellular and molecular mechanisms underlying the hyperactivity of PVN-CRH neurons in essential hypertension remain poorly understood.

Neuronal excitability is critically regulated by GABAergic inhibition, primarily mediated by GABA_A receptors that conduct chloride ions across the cell membrane.^11,12,13,14 The efficacy and direction of this chloride transport depend on the intracellular chloride concentration ([Cl⁻]_i), which is predominantly controlled by cation-chloride cotransporters (CCCs): the Na⁺-K⁺-Cl^- co-transporter (NKCC) and the K⁺-Cl^- co-transporter (KCC).^15,16,17 NKCC1 and KCC2 are the major isoforms of CCCs expressed in brain neurons,¹⁸ and dysregulation of these transporters disrupts chloride homeostasis and contributes to impaired GABAergic inhibition in various neurological conditions, including spinal cord injury, stroke, stress, and hypertension.^{19,20,21,22,23,24,25,26,27,28} However, whether and how CCCs affect the GABAergic inhibition of PVN-CRH neurons in essential hypertension still needs to be substantiated.

With-no-lysine kinase (WNK) and the downstream Ste20-related proline-alanine-rich kinase (SPAK)/oxidative stress response kinase 1 (OSR1) kinases regulate activities of multiple ion transporters and channels via protein phosphorylation, including NKCCs and KCCs.^19,29,30 WNK proteins function as chloride sensors, detecting changes in [Cl⁻]_i to activate or inhibit the SPAK-NKCC1/KCC2 pathway.^31,32 Four WNK isoforms (WNK1, 2, 3, and 4) are expressed in the brain,^33,34 and the WNK-SPAK pathway has been shown to regulate NKCC1 and KCC2 phosphorylation during the sleep-wake cycle and following stroke.^17,19 However, whether the WNK-SPAK-NKCC1/KCC2 pathway contributes to PVN-CRH neuronal hyperactivity in essential hypertension has not been investigated.

In this study, we investigated whether WNK-SPAK-related regulation of NKCC1/KCC2 is associated with altered intracellular chloride regulation and enhanced excitability of PVN-CRH neurons in spontaneously hypertensive rats (SHRs). Using chemogenetic manipulation, electrophysiological recordings, molecular analyses, and pharmacological interventions, we demonstrate that the WNK-SPAK-NKCC1/KCC2 signaling is associated with the activation of PVN-CRH neurons and heightened sympathetic outflow in SHRs. Thus, our findings identify this pathway as a promising therapeutic target for neurogenic hypertension.

Results

Chemogenetic inhibition of the excitability of PVN-CRH neurons reduces sympathetic vasomotor tone in SHRs

Hyperactivation of the sympathetic vasomotor plays a pivotal role in the pathogenesis of hypertension.³⁵ We selectively inhibited the excitability of PVN-CRH neurons using a chemogenetic approach to observe the changes of arterial blood pressure (ABP), HR, and renal sympathetic nerve activity (RSNA) in SHRs. Adeno-associated virus (AAV)-CRH promoter-Cre and AAV-hSyn-DIO-hM4Di-mCherry vectors were microinjected into the PVN of SHRs to selectively express inhibitory hM4Di in PVN-CRH neurons. For control, AAV-CRH promoter-Cre and its control vector AAV-hSyn-DIO-hM4Di-mCherry were microinjected into the PVN of SHRs. (Figure 1A). Bath application of clozapine-N-oxide (CNO) (10 μM), a designer receptors exclusively activated by designer drugs (DREADDs) agonist, significantly decreased the firing rate (t₅ = 11.26, p < 0.0001) of hM4Di-mCherry PVN neurons, while it did not affect the firing rate of mCherry PVN neurons (n = 6 neurons per group, Figures 1B and 1C).

Figure 1.

Chemogenetic inhibition of PVN-CRH neuronal excitability decreases the ABP and RSNA in SHRs

(A) Schematic illustrating microinjection of AAV1/2-CRH-Cre and AAV (PHP.eB)-hSyn-DIO-hM4Di-mCherry or its control vector AAV (PHP.eB)-hSyn-DIO-mCherry into the PVN and representative microscopic image showing the virus transfection sites in the PVN of SHRs. Scale bars, 500 μm.

(B and C) Representative traces (B) and summary data (C) show that bath application of CNO (10 μM) decreases the spontaneous firing of hM4Di-mCherry CRH neurons but does not affect mCherry neurons (n = 6 neurons from 4 rats per group).

(D) Original recording traces show the effect of intravenous administration of CNO (1.0 mg/kg) on ABP, HR, RSNA, and Int. RSNA of SHRs.

(E–G) Summary data show changes in MAP (E), HR (F), and Int. RSNA (G) in response to CNO administration in SHRs (n = 6 rats).

CRH, corticotropin-releasing hormone; PVN, paraventricular nucleus; CNO, clozapine N-oxide; ABP, arterial blood pressure; HR, heart rate; RSNA, renal sympathetic nerve activity. Data are presented as mean ± SEM. ∗∗p < 0.01, ∗∗∗p < 0.001. Two-tailed Student’s t test.

Next, we determined whether chemogenetic inhibition of PVN-CRH neuronal activity affects sympathetic outflow and ABP in SHRs. CNO administration (1.0 mg/kg, i.v.) markedly decreased MAP (t₅ = 21.22, p < 0.0001), HR (t₅ = 7.750, p < 0.0006), and RSNA (t₅ = 4.531, p = 0.0062) in SHRs with hM4Di-mCherry-expressing PVN-CRH neurons (n = 6 rats, Figures 1D–1G). However, chemogenetic inhibition of PVN-CRH neurons had no significant effect on the sympathetic outflow or ABP in Wistar-Kyoto (WKY) rats (n = 6 rats, Figures S1A–S1D).

To further confirm whether chemogenetic activation of PVN-CRH neurons influences ABP in WKY rats, we microinjected rAAV-hSyn-DIO-hM3Dq-mCherry-WPRE-hGH polyA together with a CRH promoter-driven Cre viral vector into the PVN. Fourteen days after viral injection, telemetric transmitters were implanted to continuously record ABP before and after CNO administration. Chemogenetic activation of PVN-CRH neurons by CNO increased the MAP in WKY rats (n = 6 rats, Figures S2A–S2D).

Collectively, these data suggest that inhibiting the excitability of PVN-CRH neurons reduces the hyperactivation of the sympathetic vasomotor in hypertension.

Depolarizing shift in the E_GABA and disrupted chloride homeostasis impairs GABAergic inhibition of PVN-CRH neurons in SHRs

PVN-CRH neurons were selectively labeled 21 days after the microinjection of a CRH-promoter driven by enhanced green fluorescent protein (eGFP) virus vector (Figure 2A) and identified using fluorescent microscope (Figure 2B). We assessed the GABAergic regulation of the excitability of eGFP-labeled PVN-CRH neurons in WKY rats and SHRs using gramicidin-perforated patch recordings, which are impermeable to chloride and therefore, maintain the intracellular chloride concentration ([Cl⁻]_i) unchanged. Puff application of GABA (100 μM) hyperpolarized the membrane potential (−6.886 ± 0.3232 mV) in eGFP-labeled PVN-CRH neurons of WKY rats. In contrast, GABA depolarized the membrane potential (+3.9 ± 0.9774 mV) of PVN-CRH neurons in SHRs (n = 7 neurons, p < 0.001, Figures 2C and 2D).

Figure 2.

GABAergic inhibition is impaired and E_GABA is shifted toward depolarization in PVN-CRH neurons of SHRs

(A) Schematic showing the CRH-promoter driven eGFP AAV vector microinjected to the PVN of rats.

(B) Experimental schematic and cell images illustrating patch-clamp recording of eGFP-labelled PVN neuron with an attached recording electrode under fluorescence illumination and bright light. Scale bars, 25 μm.

(C) Original recordings show changes of membrane potential (ΔV) induced by puff GABA to PVN-CRH neurons from WKY rats and SHRs.

(D) Summary data of ΔV changes induced by puff GABA to PVN-CRH neurons from WKY rats and SHRs.

(E and F) Original traces of perforated recordings (E) and I-V plots (F) show the E_GABA of GABA-induced currents at a series of membrane potentials from −90 to −30 mV of PVN-CRH neurons from WKY rats and SHRs. Note that the GABA-induced currents could be totally blocked by GABA receptor antagonist picrotoxin.

(G and H) Mean changes in E_GABA (G) and derived [Cl⁻]_i (H) of labeled PVN-CRH neurons from WKY rats and SHRs. The arrows indicate the time points of puff GABA (100 μM).

CRH, corticotropin-releasing hormone. PVN, paraventricular nucleus. n = 7 neurons from 4 rats per group. Data are presented as mean ± SEM. ∗∗∗p < 0.001. Two-tailed Student’s t test.

A depolarizing shift in the GABA reversal potential (E_GABA) causes chloride imbalance and attenuates GABAergic inhibition in neurons of the brain^21,36 and spinal cord.³⁷ Next, we conducted chloride-impermeable gramicidin-perforated recordings in eGFP-labeled PVN-CRH neurons to assess the changes in E_GABA in the brain slices from WKY rats and SHRs. The E_GABA of PVN-CRH neurons was −67.94 ± 2.649 mV in WKY rats, compared with −47.61 ± 2.029 mV in SHRs, revealing a depolarizing shift of about 20.33 mV in PVN-CRH neurons in SHRs relative to that in WKY rats (n = 7 neurons per group, t₁₂ = 6.093, p < 0.001, Figures 2E–2G). Bath application of picrotoxin (20 μM), a GABA_A receptor channel blocker, completely abolished GABA-evoked currents, confirming that these responses were mediated by GABA_A receptors. The [Cl⁻]_i calculation through the Nernst equation³⁸ revealed markedly elevated [Cl⁻]_i in PVN-CRH neurons of SHRs compared with that in WKY rats (t₁₂ = 5.988, p < 0.0001, Figure 2H). These data suggest that depolarizing shift in the E_GABA and increased [Cl⁻]_i contribute to reduced GABA inhibition of PVN-CRH neurons in SHRs.

WNK-SPAK-NKCC1/KCC2 pathway-related proteins expressions are upregulated in the PVN of SHRs

The WNK family detects changes in [Cl⁻]_i concentration, activating the kinase SPAK, which either enhances chloride influx via phosphorylated NKCC1 (p-NKCC1, Thr212/Thr217) or inhibits chloride efflux via phosphorylated KCC2 (p-KCC2, Thr1007) in the brain.³⁹ We measured the protein expression levels of p-NKCC1, NKCC1, p-KCC2, and KCC2 in the PVN of SHRs through western blot analysis. The results showed a 50% increase in p-NKCC1 (Thr212/Thr217)/NKCC1 expression ratio (t₅ = 4.482, p = 0.0065, Figures 3A and 3B) and a 1-fold increase in p-KCC2 (Thr1007)/KCC2 ratio (t₅ = 5.431, p = 0.0029, Figures 3D and 3E) in the PVN of SHRs compared with that in WKY rats (n = 6 rats per group). Furthermore, immunofluorescence staining revealed an increased number of CRH and p-NKCC1 (Thr212/Thr217) double-positive neurons (Figure 3C) and CRH and p-KCC2 (Thr1007) double-positive neurons (Figure 3F) in the PVN of SHRs relative to that in WKY rats. These data indicate enhanced phosphorylation of NKCC1 and KCC2 in the PVN-CRH neurons of SHRs.

Figure 3.

The WNK-SPAK-NKCC1/KCC2 pathway-associated proteins is upregulated in the PVN of SHRs

(A and B) Representative blots images (A) and qualification (B) of p-NKCC1 (Thr212/Thr217) and NKCC1 protein levels in the PVN of WKY rats and SHRs.

(C) Representative immunofluorescence staining images of CRH and p-NKCC1 immunopositive neurons in the PVN of WKY rats and SHRs. Scale bars for merge images: 100 μm; scale bars for enlarge images: 30 μm.

(D and E) Representative blots images (D) and qualification (E) of p-KCC2 (Thr1007) and KCC2 protein levels in the PVN of WKY rats and SHRs.

(F) Representative immunofluorescence staining images of CRH and p-KCC2 immunopositive neurons in the PVN of WKY rats and SHRs. Scale bars for merge images: 100 μm; scale bars for enlarge images: 30 μm.

(G and H) Representative blots images (G) and qualification (H) of p-SPAK (Ser373), SPAK and MO25 protein levels in the PVN of WKY rats and SHRs.

(I and J) Representative blots images (I) and qualification (J) of WNK1, WNK2, and WNK4 protein levels in the PVN of WKY rats and SHRs. n = 6 rats per group.

Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Two-tailed Student’s t test.

Increased expression of SPAK and/or phosphorylated SPAK (p-SPAK, Ser373) promotes the phosphorylation of NKCC1 and KCC2.⁴⁰ MO25, a calcium-binding protein, regulates the phosphorylation of SPAK protein by binding to it and influencing its activity.⁴¹ Therefore, we measured the expression levels of SPAK, p-SPAK (Ser373), and MO25 in the PVN of WKY rats and SHRs. We showed that both the expression levels of SPAK (t₅ = 16.30, p < 0.0001) and p-SPAK (Ser373) (t₅ = 4.327, p = 0.0075) in the PVN of SHRs were significantly higher than those observed in WKY rats (n = 6 rats per group, Figures 3G and 3H). However, there was no significant difference in MO25 protein expression between the PVN of SHRs and WKY rats (t₅ = 1.602, p = 0.1701, n = 6 rats per group, Figures 3G and 3H).

The WNK family was found to regulate the expression and phosphorylation levels of the SPAK protein in epilepsy, cerebral ischemia, and spinal cord injury.^42,43 Next, we determined the protein expression levels of WNK1, WNK2, and WNK4 in the PVN of WKY rats and SHRs. The protein levels of WNK1 (t₅ = 8.981, p = 0.0003) and WNK4 (t₅ = 4.737, p = 0.0052) in the PVN of SHRs were significantly increased compared with those in WKY rats, while WNK2 protein level did not change (t₅ = 1.838, p = 0.1255, n = 6 rats per group, Figures 3I and 3J). Together, these biochemical changes are consistent with enhanced WNK-SPAK-related signaling associated with increased phosphorylation of NKCC1 and KCC2 in the PVN of SHRs. Although phosphorylation status alone does not fully define pathway hierarchy, these findings support the involvement of WNK-SPAK signaling in the regulation of chloride transporter function under hypertensive conditions.

NKCC1 inhibition or KCC2 activation normalizes GABA inhibition and reverses the hyperactivity of PVN-CRH neurons in SHRs

The chloride homeostasis of neurons in the central nervous system is mainly maintained by two kinds of CCCs—NKCC1 and KCC2.^36,44 To determine the role of NKCC1 or KCC2 in the chloride homeostasis disturbance of PVN-CRH neurons, we pre-incubated the PVN brain slices of SHRs with bumetanide (a selective NKCC1 inhibitor) or CLP257 (a selective KCC2 activator). Then, we recorded the puff GABA-induced currents to observe the E_GABA of PVN-CRH neurons in brain slices treated with bumetanide or CLP257. Pre-incubation with bumetanide (20 μM, 1 h) or CLP257 (50 μM, 1 h) effectively restored the depolarizing shift of E_GABA (F_3,24 = 18.67, p < 0.0001, Figures 4A–4C; F_3,24 = 18.59, p < 0.0001; Figures 5A–5C) and [Cl⁻]_i (F_{3, 24} = 25.25, p < 0.0001, Figure 4D; F_3,24 = 26.49, p < 0.0001, Figure 5D) in PVN-CRH neurons of SHRs to levels comparable to WKY rats (n = 7 neurons per group). However, bumetanide or CLP257 incubation did not affect the E_GABA and [Cl⁻]_i of PVN-CRH neurons in WKY rats (n = 7 neurons per group, Figures 4A–4D and 5A–5D). Bumetanide or CLP257 incubation also reversed the GABA-induced depolarization membrane potential to hyperpolarization of PVN-CRH neurons in SHRs (F_{3, 24} = 84.49, p < 0.0001, Figures 4E and 4F; F_{3, 32} = 30.14, p < 0.0001, Figures 5E and 5F) but had no obvious effect on the membrane potential of PVN-CRH neurons in WKY rats. Meanwhile, bumetanide or CLP257 incubation normalized the increased firing rate (F_3,28 = 22.27, p < 0.0001, Figures 4G and 4H; F_3,29 = 34.36, p < 0.0001, Figures 5G and 5H) and depolarized membrane potential (F_3,28 = 17.08, p < 0.0001, Figures 4G and 4I; F_3,29 = 23.20, p < 0.0001, Figures 5G and 5I) of PVN-CRH neurons in SHRs to levels comparable to WKY rats. These findings suggest that restoring chloride homeostasis through NKCC1 inhibition (decreasing chloride influx) or KCC2 activation (increasing chloride outflux) may normalize GABA inhibition and reduce the hyperactivity of PVN-CRH neurons in hypertension.

Figure 4.

Blocking NKCC1 normalizes E_GABA, GABAergic inhibition, and firing activity of PVN-CRH neurons in SHRs

(A and B) Original traces of perforated recordings (A) and I-V plots (B) show the E_GABA of GABA-induced currents at a series of membrane potentials from −90 mV to −30 mV in PVN-CRH neurons from brain slices of WKY rats and SHRs incubated with or without bumetanide (20 μM).

(C and D) Mean changes show E_GABA (C) and derived [Cl⁻]_i (D) of labeled PVN-CRH neurons in the above groups (n = 7 neurons from 4 rats per group).

(E and F) Original recordings (E) and summary data (F) show membrane potential changes (ΔV) induced by puff GABA to PVN-CRH neurons in brain slices from WKY rats and SHRs incubated with or without bumetanide (20 μM) (n = 7 neurons from 4 rats per group).

(G) Original traces show the spontaneous firing of eGFP-labelled PVN-CRH neurons in brain slices from WKY rats and SHRs treated with or without bumetanide.

(H and I) Summary data show the firing rate (H) and membrane potential (I) of spontaneous firing of PVN-CRH neurons in the aforementioned groups (n = 7 neurons from 4 rats in WKY, SHR, and SHR+BUM; n = 11 neurons from 4 rats in WKY+BUM). The arrow indicates the time point of puff GABA (100 μM).

BUM, bumetanide. Data are presented as mean ± SEM. ∗∗∗p < 0.001. One-way ANOVA followed by Tukey’s post hoc test.

Figure 5.

Activation of KCC2 restores E_GABA, GABAergic inhibition and firing activity of PVN-CRH neurons in SHRs

(A and B) Original traces of perforated recordings (A) and I-V plots (B) show the E_GABA of GABA-induced currents at a series of membrane potentials from −90 mV to −30 mV in PVN-CRH neurons from brain slices of WKY rats and SHRs incubated with or without CLP257 (50 μM).

(C and D) Mean changes show E_GABA (C) and derived [Cl⁻]_i (D) of labeled PVN-CRH neurons in the above groups (n = 7 neurons from 4 rats in WKY, WKY+CLP257, and SHR+CLP257; n = 8 neurons from 4 rats in SHR).

(E and F) Original recordings (E) and summary data (F) show membrane potential changes (ΔV) induced by puff GABA to PVN-CRH neurons in brain slices from WKY rats and SHRs incubated with or without CLP257 (50 μM) (n = 9 neurons from 4 rats in WKY, n = 12 neurons from 5 rats in WKY+CLP257, n = 8 neurons from 4 rats in SHR, and n = 7 neurons from 4 rats in SHR+CLP257).

(G) Original traces show the spontaneous firing of eGFP-labelled PVN-CRH neurons in brain slices from WKY rats and SHRs incubated with or without CLP257.

(H and I) Summary data show the firing rate (H) and membrane potential (I) of spontaneous firing of PVN-CRH neurons in the aforementioned groups (n = 7 neurons from 4 rats in WKY, n = 11 neurons from 5 rats in WKY+CLP257, n = 8 neurons from 4 rats in SHR, and n = 7 neurons from 4 rats in SHR+CLP257). The arrow indicates the time point of puff GABA (100 μM).

Data are presented as mean ± SEM. ∗∗∗p < 0.001. One-way ANOVA followed by Tukey’s post hoc test.

Inhibition of WNK-SPAK pathway normalizes GABA inhibition by restoring chloride homeostasis in PVN-CRH neurons of SHRs

Inhibition of WNK-SPAK pathway may enhance chloride efflux and promote GABAergic inhibition in neurons.³⁹ We demonstrated the upregulation of the WNK-SPAK-NKCC1/KCC2 pathway in the PVN of SHRs (Figure 3). However, its impact on the firing activity and GABAergic inhibition of PVN-CRH neurons in hypertension remains unknown. Incubation with ZT-1a (an SPAK inhibitor, 1 μM, 1 h) or WNK463 (a WNK inhibitor, 5 μM, 1 h) significantly reversed the depolarizing shift of E_GABA (F_3,28 = 20.57, p < 0.0001, Figures 6A–6C; F_3,28 = 21.38, p < 0.0001; Figures 7A–7C) and normalized the [Cl⁻]_i (F_3,28 = 19.14, p < 0.0001, Figure 6D; F_3,28 = 46.61, p < 0.0001; Figure 7D) in PVN-CRH neurons of SHRs to levels similar to those observed in WKY rats (n = 8 neurons per group). However, administration of ZT-1a or WNK463 did not affect the E_GABA and [Cl⁻]_i in PVN-CRH neurons of WKY rats (Figures 6A–6D and 7A–7D). The aforementioned data suggest that inhibition of WNK or SPAK restores chloride homeostasis of PVN-CRH neurons in hypertension.

Figure 6.

Blockade of the SPAK with ZT-1a normalizes GABA inhibition by restoring the chloride homeostasis of PVN-CRH neurons in SHRs

(A and B) Original traces (A) and I-V plots (B) show the effect of ZT-1a (1 μM) incubation on perforated recordings of GABA-induced currents at a series of membrane potentials from −90 mV to −30 mV in PVN-CRH neurons from brain slices of WKY rats and SHRs (n = 8 neurons from 4 rats per group).

(C and D) Mean changes show E_GABA (C) and derived [Cl⁻]_i (D) of labeled PVN-CRH neurons in the above groups of rat brain slices (n = 8 neurons from 4 rats per group).

(E and F) Original recordings (E) and summary data (F) show the effect of ZT-1a (1 μM) incubation on membrane potential changes (ΔV) induced by puff GABA to PVN-CRH neurons in brain slices from WKY rats and SHRs (n = 7 neurons from 4 rats per group).

(G) Original traces show the effect of ZT-1a incubation on spontaneous firing of eGFP-labelled PVN-CRH neurons in brain slices from WKY rats and SHRs.

(H and I) Summary data show the firing rate (H) and membrane potential (I) of spontaneous firing of PVN-CRH neurons in the above groups (n = 7 neurons from 4 rats in WKY, SHR, and SHR+ZT-1a; n = 10 neurons from 4 rats in WKY+ZT-1a). The arrow indicates the time point of puff GABA (100 μM).

Data are presented as mean ± SEM. ∗∗∗p < 0.001. One-way ANOVA followed by Tukey’s post hoc test.

Figure 7.

Blockade of the WNK with WNK 463 normalizes GABA inhibition by restoring the chloride homeostasis of PVN-CRH neurons in SHRs

(A and B) Original traces (A) and I-V plots (B) show the effect of WNK463 (5 μM) incubation on perforated recordings of GABA-induced currents at a series of membrane potentials from −90 mV to −30 mV in PVN-CRH neurons from brain slices of WKY rats and SHRs (n = 8 neurons from 4 rats per group).

(C and D) Mean changes show E_GABA (C) and derived [Cl⁻]_i (D) of labeled PVN-CRH neurons in the aforementioned groups of rat brain slices (n = 8 neurons from 4 rats per group).

(E and F) Original recordings (E) and summary data (F) show the effect of WNK463 (5 μM) incubation on membrane potential changes (ΔV) induced by puff GABA to PVN-CRH neurons in brain slices from WKY rats and SHRs (n = 7 neurons from 4 rats per group).

(G) Original traces show the effect of WNK463 incubation on spontaneous firing of eGFP-labelled PVN-CRH neurons in brain slices from WKY rats and SHRs.

(H and I) Summary data show the firing rate (H) and membrane potential (I) of spontaneous firing of PVN-CRH neurons in the above groups (n = 7 neurons from 4 rats in WKY, n = 10 neurons from 4 rats in WKY+WNK463, n = 8 neurons from 4 rats in SHR, and n = 7 neurons from 4 rats in SHR+WNK463). The arrow indicates the time point of puff GABA (100 μM).

Data are presented as mean ± SEM. ∗∗∗p < 0.001. One-way ANOVA followed by Tukey’s post hoc test.

Next, we assessed the impact of ZT-1a or WNK463 on the response of PVN-CRH neurons to GABA (100 μM) and their spontaneous firing activity. Administration of ZT-1a or WNK463 significantly reversed puff GABA-induced depolarizing changes in membrane potential in PVN-CRH neurons of SHRs (F_3,24 = 37.84, p < 0.0001, Figures 6E and 6F; F_3,24 = 66.25, p < 0.0001, Figures 7E and 7F). However, ZT-1a or WNK463 did not affect the response of PVN-CRH neurons to puff GABA in PVN-CRH neurons of WKY rats. Meanwhile, incubation with either ZT-1a or WNK463 reduced the firing rate (F_3,27 = 21.00, p < 0.0001, Figures 6G and 6H; F_3,28 = 30.12, p < 0.0001, Figures 7G and 7H) and hyperpolarized membrane potential (F_3,27 = 23.13, p < 0.0001, Figure 6I; F_3,28 = 29.60, p < 0.0001, Figure 7I) of PVN-CRH neurons in SHRs but not in WKY rats. Together, these findings support a functional link between WNK-SPAK signaling and altered chloride homeostasis in PVN-CRH neurons of SHRs.

Activation of the WNK-SPAK pathway results in increased sympathetic vasomotor tone in hypertension

Finally, we determined whether increased activation of the WNK-SPAK pathway contributes to the heightened sympathetic outflow in hypertension. Microinjection of ZT-1a (100 μM, 50 nL) into the PVN resulted in a notable reduction in MAP (t₅ = 7.287, p = 0.0008), HR (t₅ = 5.693, p = 0.0023), and Int. RSNA (t₅ = 6.182, p = 0.0016) in SHRs (n = 6 rats per group, Figures 8A–8D). Moreover, microinjection of WNK463 (200 μM, 50 nL) similarly decreased the MAP (t₅ = 7.483, p = 0.0007), HR (t₅ = 4.006, p = 0.0103), and Int. RSNA (t₅ = 10.05, p = 0.0002) in SHRs (n = 6 rats per group, Figures 8E–8H). However, neither ZT-1a nor WNK463 affected the sympathetic outflow or ABP in WKY rats (Figures S3A–S3H). We also verified the injection site and excluded the data from animals with microinjections outside the PVN (Figures 8I–8K). Together, these findings strongly support a functional link between WNK-SPAK signaling within the PVN and elevated sympathetic vasomotor activity in SHRs. Although derived from pharmacological interventions, the consistent effects on neuronal excitability and sympathetic output highlight a significant role for this pathway in the central regulation of sympathetic activity during hypertension.

Figure 8.

Inhibition of WNK-SPAK pathway reduces sympathetic vasomotor tone and blood pressure in SHRs

(A) Representative recordings show the effect of bilateral microinjection of ZT-1a (100 μM, 50 nL) into the PVN on the ABP, HR, RSNA, and Int. RSNA in SHRs.

(B–D) Summary data show changes in MAP (B), HR (C), and Int. RSNA (D) in response to microinjection of ZT-1a (100 μM, 50 nL) into the PVN of SHRs (n = 6 rats per group).

(E) Representative recordings show the effect of bilateral microinjection of WNK463 (200 μM, 50 nL) into the PVN on the ABP, HR, RSNA, and Int. RSNA in SHRs.

(F–H) Summary data show changes in MAP (F), HR (G), and Int. RSNA (H) in response to microinjection of WNK463 (200 μM, 50 nL) into the PVN of SHRs (n = 6 rats per group).

(I) Representative micrographs show the microinjection sites of ZT-1a or WNK463 in the PVN of rats under bright-field (a) and fluorescence (b) microscopy. Representative image shows the microinjection sites in the PVN of SHRs. Scale bars, 500 μm.

(J) Brain atlas depicts the microinjection sites of ZT-1a (Δ in blue) in the PVN of SHRs.

(K) Brain atlas depicts the microinjection sites of WNK463 (○ in red) in the PVN of SHRs.

ABP, arterial blood pressure; MAP, mean arterial pressure; HR, heart rate; RSNA, renal sympathetic nerve activity. Data are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001. Two-tailed Student’s t test.

Discussion

This study provides new evidence that dysregulation of the WNK-SPAK-NKCC1/KCC2 signaling pathway is closely associated with disrupted chloride homeostasis in PVN-CRH neurons, reduced GABAergic inhibition, and enhanced neuronal excitability in SHRs. Chemogenetic inhibition of PVN-CRH neurons reduces blood pressure and RSNA, supporting a functional contribution of CRH neuronal excitability to elevated sympathetic outflow in hypertension. In addition, pharmacological inhibition of WNK-SPAK signaling improved chloride regulation, restored excitability of PVN-CRH neurons, and reduced blood pressure in SHRs, providing a promising therapeutic strategy for neurogenic hypertension.

GABAergic inhibition plays a critical role in regulating the activity of PVN-CRH neurons.^36,45 Notably, GABA_A-mediated inhibition of these neurons is markedly reduced during stress, depression, anxiety, and post-traumatic stress disorder.^36,46,47,48 Under normal conditions, activation of GABA_A receptors causes chloride influx and leading to neuronal inhibition. However, when [Cl⁻]ᵢ is elevated, the chloride driving force is reversed, transforming GABA responses from inhibitory to excitatory. Remarkably, our study revealed that GABA_A-mediated inhibition of PVN-CRH neurons is similarly diminished during hypertension. Specifically, we observed that in SHRs, GABA application depolarized the membrane potential and enhanced neuronal excitability, whereas in normotensive WKY rats, GABA hyperpolarized the membrane potential and suppressed excitability. Furthermore, we identified a depolarizing shift in E_GABA in PVN-CRH neurons of SHRs. These findings suggest that chloride imbalance, resulting from the depolarizing shift in E_GABA, contributes to the attenuation of GABAergic inhibition and the heightened excitability of PVN-CRH neurons in hypertension.

Our data further indicate that dysregulation of NKCC1 and KCC2 is associated with this altered chloride balance. Enhanced NKCC1 activity and/or reduced KCC2 activity leads to higher [Cl⁻]_i and a shift of E_GABA from inhibition to excitation.^49,50 We further confirmed the roles of NKCC1 and KCC2 on E_GABA by applying the NKCC1 antagonist bumetanide and the KCC2 agonist CLP257, which both restored the depolarized E_GABA and reduced GABAergic inhibition in PVN-CRH neurons of SHRs to levels observed in WKY rats. Notably, pharmacological interventions targeting NKCC1 or KCC2 had no effect on neuronal excitability or chloride homeostasis in normotensive WKY rats. Therefore, therapeutic targeting of this pathway would not adversely affect normal PVN-CRH neuronal function.

Ye et al. reported that enhanced NKCC1 function (including increased glycosylation), rather than KCC2 downregulation, mediates the depolarizing shift in spinally projecting PVN presympathetic neurons in SHRs.²¹ Consistent with their findings, we observed increased total NKCC1 expression, whereas total KCC2 expression was unchanged in SHRs. Importantly, our study extends this prior work by showing that phosphorylation of both NKCC1 (Thr212/Thr217) and KCC2 (Thr1007) is significantly increased in SHRs. These results support a dual mechanism in which upregulated/glycosylated surface NKCC1 is further activated via phosphorylation, while KCC2-mediated chloride extrusion is concurrently impaired by phosphorylation at Thr1007. We propose that these parallel phosphorylation events converge to drive the depolarizing shift in E_GABA and disrupt chloride homeostasis. The importance of these two co-transporters in GABAergic signaling in SHRs makes them valuable targets for the treatment of neurogenic hypertension.

Our study also reveals that the WNK-SPAK pathway acts as an upstream signaling mechanism for activating NKCC1 or inhibiting KCC2 in PVN-CRH neurons. It has been demonstrated that WNK1 and WNK4 play an important role in pseudohypoaldosteronism type II disease, which is characterized by hypertension and hyperkalemia.⁵¹ More recent evidence suggests that WNK1 may also be important to the pathogenesis of salt-sensitive essential hypertension in the general population.⁵² However, these studies mainly focus on the regulation of renal WNK-SPAK-NKCC1/KCC2 in salt homeostasis, blood volume, and long-term blood pressure regulation. Despite its abundance in the brain, the central role of WNK-SPAK-NKCC1/KCC2 pathway,^17,19 particularly in the hyperactivation of the sympathetic system, is largely overlooked. Our biochemical analysis revealed elevated expression of WNK1 and WNK4, accompanied by increased p-SPAK (Ser373), p-NKCC1 (Thr212/Thr217), and p-KCC2 (Thr1007) in the PVN of SHRs. SPAK and OSR1 phosphorylate NKCC1 at Thr212/Thr217 to enhance its activity, whereas phosphorylation of KCC2 at Thr1007 impairs its function.^53,54,55 Therefore, the elevated expression of WNK1 and WNK4 promotes the phosphorylation of p-SPAK (Ser373), which subsequently increases the phosphorylation of p-NKCC1 (Thr212/Thr217) and p-KCC2 (Thr1007), enhancing NKCC1 activity while reducing KCC2 activity in the PVN of SHRs. Notably, pharmacological inhibition of WNK or SPAK with WNK463 or ZT-1a normalized GABAergic inhibition in PVN-CRH neurons ex vivo. Therefore, the WNK-SPAK-NKCC1/KCC2 pathway appears to be involved in the hyperactivation of PVN-CRH neurons in hypertension.

The selective efficacy of ZT-1a or WNK463 in SHRs is most consistent with state-dependent activation of the WNK-SPAK/OSR1-CCCs axis in hypertension. In SHRs, this pathway appears chronically upregulated, as evidenced by elevated WNK1/4 expression and increased phosphorylation of SPAK, NKCC1, and KCC2, which increased sensitivity to pathway inhibition. Conversely, the lower basal activity of this signaling cascade in WKY rats likely limits the functional impact of blockade. Similar state-dependent amplification of WNK-SPAK-CCCs signaling has been reported in other pathological conditions.^39,56,57 Although WNK-SPAK kinases are established upstream regulators of NKCC1 and KCC2 phosphorylation,^39,58 the present study did not directly test the pathway linking WNK-SPAK activation to CCCs-dependent shifts in E_GABA. Therefore, the effects of WNK or SPAK inhibition cannot be attributed exclusively to NKCC1 or KCC2 modulation.

Importantly, when these inhibitors were microinjected directly into the PVN in vivo, they reduced MAP, HR, and RSNA in SHRs. These findings establish the WNK-SPAK pathway as not merely a correlative finding but a mechanistically important target for therapeutic intervention. Although systemic administration of SPAK inhibitors has shown limited effects on blood pressure in previous studies,¹⁹ our direct central delivery demonstrates the importance of central WNK-SPAK signaling for cardiovascular regulation, highlighting the necessity of targeting this pathway within the brain for therapeutic efficacy.

While CRH neurons in the PVN are classically recognized for their neuroendocrine role in regulating the hypothalamic-pituitary-adrenal (HPA) axis, increasing evidence indicates that a subset of PVN neurons expressing CRH also participates in central autonomic regulation. Recent anatomical and functional studies suggest that CRH neurons can influence these presympathetic populations through local circuit interactions.⁶ CRH signaling within the PVN has been shown to modulate neuronal excitability and synaptic transmission, thereby affecting sympathetic nerve activity and cardiovascular function.⁷ Our finding that chemogenetic inhibition of PVN-CRH neurons reduces RSNA and blood pressure in SHRs confirms their role in autonomic regulation beyond endocrine functions.

PVN-CRH neurons projecting to brainstem autonomic centers, such as the nucleus of the solitary tract (NTS) and the RVLM,⁵⁹ constitute a critical pathway for central cardiovascular regulation. Indeed, activation of PVN-CRH terminals in the NTS has been shown to increase blood pressure, underscoring the role of CRH signaling in descending autonomic control.⁶⁰ However, it remains to be determined whether this preautonomic PVN-CRH population overlaps with the specific CRH neurons exhibiting disrupted chloride homeostasis in the present study.

In conclusion, this study demonstrates that WNK-SPAK-NKCC1/KCC2 pathway activation is associated with the disruption of chloride homeostasis in PVN-CRH neurons, which contributes to the reduction of GABAergic inhibition, leading to hyperactivity of these neurons and heightened sympathetic outflow in SHRs. Our study suggests that targeting the WNK-SPAK-NKCC1/KCC2 pathway in PVN-CRH neurons could be an effective strategy for treating neurogenic hypertension.

Limitations of the study

While the present study identifies a role for WNK-SPAK-NKCC1/KCC2 signaling in chloride dysregulation and PVN-CRH neuronal hyperexcitability during hypertension, several limitations should be considered. First, the cell-type-specific mechanisms linking pathway activation to altered chloride homeostasis were not directly investigated. Second, as our findings rely primarily on the SHR model, it remains to be determined whether similar mechanisms are conserved in other forms of neurogenic hypertension. Future studies using genetic and circuit-specific approaches will help further define the role of chloride homeostasis dysfunction in central autonomic regulation.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Huijie Ma (huijiema@hebmu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •All data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

Acknowledgments

This study was supported by the National Natural Science Foundation of China (31971044 and 82571813); Natural Science Foundation of Hebei Province (H2019206325 and C2021106015); Key laboratory of Neural and Vascular Biology, Ministry of Education of China (NV20210004); Natural Science Foundation of Hebei Province for Innovative Research Group Project (H2025206897); and Hebei Postgraduate Innovation Grant Program (CXZZBS2022081 and XCXZZB202317). The authors thank the Core Facilities and Centers of Hebei Medical University for their assistance with confocal image acquisition.

Author contributions

H.M. designed the study; X.G., H.M., Q.Z., Y.Z., P.Y., Q.Z., and L.J. performed the experiments and data analysis; X.G., L.Z., Y.G., Y.Z., and H.M. interpreted the findings; X.G. and H.M. wrote the first draft of the manuscript; Y.Z., S.M., and H.M. revised and commented on the manuscript.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Antibodies Rabbit anti-CRH antibody	Abclonal	Cat#A1122; RRID:AB_2758463
Antibodies Rabbit anti-NKCC1 antibody	Proteintech	Cat#13884-1-AP; RRID:AB_2188522
Antibodies Rabbit anti-phospho-NKCC1 antibody	Sigma-Aldrich	Cat#ABS1004; RRID:N/A
Antibodies Rabbit anti-KCC2 antibody	Proteintech	Cat#19565-1-AP; RRID:AB_10638485
Antibodies Rabbit anti-phospho-KCC2 antibody	Cayman Chemical	Cat#29292; RRID:N/A
CY3-Tyramide-conjugated secondary antibody	Servicebio	Cat#G1233; RRID:N/A
iF488-Tyramide-conjugated secondary antibody	Servicebio	Cat#G1231; RRID:N/A
Antibodies Rabbit anti-SPAK antibody	Abclonal	Cat#A2275; RRID:AB_2862985
Antibodies Rabbit anti-phospho-SPAK antibody	Sigma-Aldrich	Cat#07-2273; RRID:AB_11205577
Antibodies Rabbit anti-WNK1 antibody	Proteintech	Cat#68149-1-Ig; RRID:AB_2923673
Antibodies Rabbit anti-WNK2 antibody	Thermo Fisher Scientific	Cat#PA5-37108; RRID:AB_2553911
Antibodies Rabbit anti-WNK4 antibody	Novus Biologicals	Cat#NB600-284SS; RRID:AB_922775
Antibodies Mouse anti-MO25 antibody	Santa Cruz Biotechnology	Cat#sc-517655; RRID:N/A
Antibodies Rabbit anti-β-actin antibody	Abclonal	Cat#AC026; RRID:AB_2768234
Goat anti-rabbit IgG-HRP	Proteintech	Cat#SA00001-2; RRID: AB_2722564
Goat anti-mouse IgG-HRP	Proteintech	Cat#SA00001-1; RRID: AB_2722565
Bacterial and virus strains
AAV1/2-CRH-eGFP	Genedetect	N/A
AAV1/2-CRH-Cre	Genedetect	N/A
AAV(PHP.eB)-hSyn-DIO-hM4Di-mCherry	GeneChem	N/A
AAV(PHP.eB)-hSyn-DIO-mCherry	GeneChem	N/A
rAAV-hSyn-DIO-hM3Dq-mCherry-WPRE-hGH polyA	BrainVTA	Cat#PT-0019
Chemicals, peptides, and recombinant proteins
CNO	Cayman Chemical	Cat#25780
WNK463	CSNpharm	Cat#CSN19078
ZT-1a	GLPBIO	Cat#GC39826
Bumetanide	Cayman Chemical	Cat#14630
CLP257	Cayman Chemical	Cat#30948
Tetrodotoxin (TTX)	Cayman Chemical	Cat#14964
CGP55845	Cayman Chemical	Cat#149184-22-5
Picrotoxin	Cayman Chemical	Cat#20771
Colchicine	Cayman Chemical	Cat#9000760
DiI	Sigma-Aldrich	Cat#468495
Gramicidin	GENTAUR BIOVISION	Cat#B1624-5
CNQX	Cayman Chemical	Cat#14618
AP5	hellobio	Cat#HB0252
GABA	Sigma-Aldrich	Cat#A2129
Critical commercial assays
ECL Western Blot Detection Kit	Report Bio & Technology	Cat#RP-WA0601
RIPA Lysis Buffer	Report Bio & Technology	Cat#RP-WA0002
Protease/Phosphatase Inhibitor Cocktail	Report Bio & Technology	Cat#RP-WA0102
MOPS-SDS-PAGE	ACE Biotechnology	Cat#ET15420Gel
Experimental models: Organisms/strains
Spontaneously Hypertensive Rat (SHR)	Beijing Vital River	SHR/NCrl
Wistar Kyoto Rat (WKY)	Beijing Vital River	WKY/NCrl
Software and algorithms
Dataquest ART	Data Sciences Internetional	https://support.datasci.com/hc/en-us/sections/115002864548-Dataquest-ART
ZEISS ZEN software	ZEISS	https://www.zeiss.com/microscopy/us/products/software/zeiss-zen.html
Prism 9	GraphPad	https://www.graphpad.com/
Image J (FIJI)	GNU General Public License	https://imagej.net/software/fiji/
Clampfit 11	Molecular Devices	https://www.moleculardevices.com/
Multiclamp Commander	Molecular Devices	https://support.moleculardevices.com/s/article/Axon-MultiClamp-700B-CommanderDownload-page
Spike2	Cambridge Electronic Design	https://ced.co.uk/downloads/latestsoftware

Experimental model and study participant details

Animals

Male normotensive WKY rats and SHRs (15–16 weeks old) were obtained from Beijing Vital River Laboratory Animal Technology Company (Beijing, China). The rats were housed in a facility at a controlled temperature of 24–25 °C, with ad libitum access to food and water, and a 12 h/12 h light/dark cycle. The surgical procedures and experimental protocols were approved by the Ethics Committee for the use of experimental animals at Hebei Medical University (IACUC-Hebmu-2024030) and conformed to the US National Institutes of Health guidelines for the Care and Use of Laboratory Animals.

Method details

Viral vectors microinjection

The CRH neurons in the rat PVN were identified by an AAV 1/2 vector-mediated expression of eGFP driven by rat CRH promoter (Genedetect, Florida, USA).³⁶ Inhibitory hM4Di was selectively expressed in PVN-CRH neurons through bilateral microinjection of a 50 nL mixture of two viral vectors: AAV1/2-CRH-Cre (Genedetect, Florida, USA) and AAV (PHP.eB)-hSyn-DIO-hM4Di-mCherry or its control vector AAV (PHP.eB)-hSyn-DIO-mCherry (GeneChem, Shanghai, China) into the PVN of rats for chemogenetic experiment. The expression of hM4Di-mCherry in PVN-CRH neurons was confirmed using a confocal microscope (ZEISS LSM 900 with Airyscan2).

To further confirm whether chemogenetic stimulation of PVN-CRH neurons lead to hypertension in WKY rats, we microinjected rAAV-hSyn-DIO-hM3Dq-mCherry-WPRE-hGH polyA together with a CRH promoter–driven Cre viral vector into the PVN to enable chemogenetic activation of PVN-CRH neurons. Fourteen days after viral injection, we implanted telemeters to measure arterial blood pressure before and after CNO administration.

Rats were anesthetized with pentobarbital (60 mg/kg, i.p.), and viral vectors were microinjected bilaterally into the PVN through a drilled hole in the skull using a glass micropipette connected to Nanoject II Microinjectors (Drummond Scientific Company, Broomall, PA, USA) mounted on a stereotaxic apparatus (RWD Life Science, Shenzhen, China). The tip of the glass micropipette (tip diameter, 20–30 μm) was advanced into the PVN according to the stereotaxic coordinates: 1.8–2.0 mm caudal to the bregma, 0.4-0.5 mm lateral to the midline, and 7.5–7.8 mm ventral to the dura. We delivered two separate 50 nL injections of the AAV vector into each side of the PVN within 2 min. After the injection of viral vectors, the glass pipette was then held in place for an extra 8 min before being slowly withdrawn. After surgery, the rats were returned to their home cages after recovering, and analgesics and antibiotics were administered intraperitoneally for 3 days. The vector was allowed to be expressed in the PVN for 21 days before electrophysiological experiments.

Telemetry and measurement of arterial blood pressure in awake WKY rats

ABP was measured in freely moving, conscious WKY rats using wireless telemeters (Data Sciences International, DSI; St. Paul, MN, USA). Briefly, rats were anesthetized with 2% isoflurane. The catheter connected to the telemetry transmitter was inserted into the abdominal aorta via an incision in the right femoral artery. Then the transmitter body was placed subcutaneously and secured with sutures. The incision was closed in two layers with interrupted sutures. Postoperatively, rats received analgesics (meloxicam, 2 mg/kg) and antibiotics (enrofloxacin, 5 mg/kg) for 3 days and were housed individually for a 7-day recovery period. ABP signals were continuously acquired and analyzed by DSI acquisition hardware and Dataquest ART software, and HR were derived from the arterial pressure pulse waveform.

Arterial blood pressure and renal sympathetic nerve activity recording

Rats were anesthetized by intraperitoneal injection of urethane (800 mg/kg) and α-chloralose (60 mg/kg). ABP was monitored with a pressure transducer through a catheter into the right femoral artery. HR was derived from the pulsatile blood pressure. Left renal sympathetic nerve branch was isolated under a surgical microscope through a retroperitoneal incision. The nerve signal was amplified and bandpass filtered using an AC amplifier (model NL125/NL126; NeuroLog System, USA). The RSNA and ABP were recorded using a 14-bit analog-to-digital converter and analyzed by Spike2 software (Cambridge Electronic Design, UK). At the end of each experiment, the nerve’s proximal end was crushed, and the noise level was measured. Integrated RSNA was obtained by subtracting the noise level from total RSNA and expressed as percentage of the baseline. The basal level was set as 100%, and the percentage change from the baseline value was calculated.

CNO, (1.0 mg/kg, #25780, Michigan, USA) was administrated intravenously to assess the impact of activating hM4Di-expressed or hM3Dq-expressed PVN-CRH neurons on BP, HR and RSNA. To evaluate the effect of WNK-SPAK pathway on sympathetic vasomotor tone, WNK463 (200 μM, 50 nL, #CSN19078-10 mg, Chicago, USA) or ZT-1a (100 μM, 50 nL, #GC39826, Montclair, USA) was microinjected into the PVN using a microinjection pipette following the aforementioned stereotactic coordinates. We added 5% DiI fluorescent tracer to the microinjection solution to verify injection site accuracy at the end of each experiment. Rats were excluded from data analysis if microinjections were misplaced outside the PVN.

Immunofluorescence staining

To enhance somatic CRH immunoreactivity, rats received an intracerebroventricular (i.c.v.) injection of colchicine (25 μg in 1 μL sterile saline) 24h prior to tissue fixation. For perfusion, the rats were anesthetized with pentobarbital (60 mg/kg, i.p.) and transcardially perfused with 200 ml of chilled saline, followed by 4% phosphate-buffered (500 ml, 0.1 mol/L, pH 7.4) paraformaldehyde and 10% phosphate-buffered sucrose solutions (200 ml, 0.1 mol/L, pH 7.4). Then the rat was decapitated, and the brain was removed and stored in 4% paraformaldehyde at 4 °C for 1 h. Subsequently, the brains were immersed in 30% phosphate-buffered sucrose solutions at 4 °C for at least 2 days. Brain tissues were then embedded in paraffin, sectioned at 4 μm thickness using a microtome. After deparaffinization, the sections underwent high-pressure antigen retrieval with sodium citrate buffer (G1202, pH 6.0, Servicebio, Wuhan, China).

For CRH and p-NKCC1 co-localization, tissue sections were incubated overnight at 4 °C with CRH primary antibody (1:50, #A1122, Abclonal, Wuhan, China), followed by a 2-hour incubation at 25 °C with CY3-Tyramide-conjugated secondary antibody (1:500, #G1233, Servicebio, Wuhan, China). Subsequently, the sections were incubated overnight with p-NKCC1 primary antibody (1:100, #ABS1004, Sigma, Burlington, MA) at 4 °C, followed by a 2-hour incubation at 25 °C with iF488-Tyramide-conjugated secondary antibody (1:500, #G1231, Servicebio, Wuhan, China). For the co-localization of CRH (1:50, A1122, Abclonal, Wuhan, China) and p-KCC2 (1:100, 29292, Cayman, Michigan, USA), the same protocol was employed.

Images were acquired with an orthogonal fluorescence microscope (Nikon Eclipse C1, Nikon, Japan) and Scanning was performed using a Pannoramic MIDI scanner (Pannoramic MIDI, 3DHISTECH, Hungary).

Brain slice preparation

The whole brain of rat was rapidly removed under 3% isoflurane anesthesia and placed in an ice-cold artificial cerebrospinal fluid (aCSF) containing (in mmol/L) 126.0 NaCl, 3.0 KCl, 1.2 NaH₂PO₄, 11.0 Glucose, 26.0 NaHCO₃, 1.5 MgCl₂, 2.4 CaCl₂. Coronal slices (300 μm) containing the PVN were obtained from the rat injected with viral vectors using vibrating microtome (VT1000S; Leica Biosystems Inc., Buffalo Grove, IL, USA). Then the slices were incubated in HEPES holding aCSF (in mmol/L) containing 92 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 30 NaHCO₃, 20 HEPES, 25 glucose, 2 thiourea, 5 Na-ascorbate, 3 Na-pyruvate, 2 CaCl₂·2H₂O, and 2 MgSO₄·7H₂O, saturated with 95% O₂ and 5% CO₂ at 34 °C for 1 h before electrophysiological recording.

Electrophysiological recordings

The eGFP labeled neurons located in the medial third of the PVN area between the third ventricle and fornix were selected for electrophysiological recording. The selected neuron was briefly identified using epifluorescence illumination combined with differential interference contrast optics on an upright microscope (BX51WI, Olympus Optical).

Perforated recordings were used to measure the spontaneous firing activity of labeled PVN-CRH neurons under current-clamp mode, minimizing the interference from the intracellular recording solution. The internal pipette solution contained (in mmol/L): 130.0 K-gluconate, 15.0 KCl, 5.0 NaCl, 1.0 MgCl₂, and 10.0 HEPES (pH adjusted to 7.2 with KOH to 300 mOsmol/L). The tip of the recording pipette was backfilled with an internal solution containing gramicidin, which was freshly dissolved in dimethylsulfoxide (DMSO) and diluted to a final concentration of 40 μg/ml. Access resistance was continuously monitored throughout recordings and data were included only when access resistance stabilized between 50–80 MΩ and changed by <20% during recording. Recordings of the firing activity of the labeled neurons started when they reached a steady state for 3–6 min.

Cl^--impermeable gramicidin-perforated recordings were performed using an internal pipette solution containing (in mmol/L) 140 CsCl, 5 EGTA, and 10 HEPES, pH 7.4.^21,36 To determine the E_GABA, GABA currents were elicited through puff application of GABA under voltage-clamp mode and recorded at a range of holding potentials (from −90 mV to −30 mV in 10 mV increments). GABA (100 μM) was applied using pressure ejection (3–5 psi) through a glass pipette positioned approximately 100 μm from the soma, with a puff duration of 15 ms. To isolate GABA_A receptor–mediated responses, ionotropic glutamate receptor blockers CNQX (10 μM, #14618, Cayman Chemical, Michigan, USA) and AP5 (50 μM, #HB0252, hellobio, Dunshaughlin, Ireland) were included in the bath solution during E_GABA measurements. In addition, tetrodotoxin (TTX, 1 μM, #14964, Cayman Chemical, Michigan, USA) and CGP55845 (2 μM, #149184-22-5, Cayman Chemical, Michigan, USA) and were added to the bath solution to block action potential-dependent synaptic activity and GABA_B receptors, respectively. Picrotoxin (20 μM, #20771, Cayman Chemical, Michigan, USA), a GABA_A receptors antagonist, was dissolved in aCSF to confirm that the induced currents were mediated by GABA_A receptors.

Bumetanide (20 μM, #14630, Cayman Chemical, Michigan, USA), CLP 257 (50 μM, #30948, Cayman Chemical, Michigan, USA), ZT-1a (1 μM, #GC39826, GLPBIO, Montclair, USA) and WNK463 (5 μM, #CSN19078-10 mg, CSNpharm, Chicago, USA) were dissolved in aCSF (continuously saturated with 95% O₂ and 5% CO₂) for brain slices incubation. The concentrations of bumetanide and CLP257 were selected based on previous studies demonstrating effective functional modulation of NKCC1 and KCC2 in brain slices.^16,18,61

Western immunoblotting

Rats were decapitated under 5% isoflurane anesthesia in 100% O₂. Brains were quickly removed and placed in ice-cold aCSF saturated with 95% O₂ and 5% CO₂. Hypothalamic slices containing the PVN were sectioned 1.08–2.12 mm caudal to the bregma. Tissue samples were collected from rat brain slices using the micro-punch technique as previously described.⁶² Total proteins were extracted from the PVN tissue using a RIPA lysis/extraction buffer (#RP-WA0002, Report Bio & Technology Co., Ltd, Hebei, China) supplemented with a protease and phosphatase inhibitor cocktail (#RP- WA0102, Bio & Technology Co., Ltd, Hebei, China). The brain tissue was homogenized and then centrifuged at 16,000 g for 15 min to obtain the supernatant. The samples were subjected to 4%–20% MOPS-SDS-PAGE (#ET15420Gel, ACE, Changzhou, Jiangsu, China) and then transferred to a polyvinylidene difluoride membrane (Millipore, Sigma, Burlington, MA, USA). The immunoblots were incubated with primary antibodies (Table S1) overnight at 4 °C. After washing with TBST, the membranes were incubated with a goat anti-rabbit IgG/HRP secondary antibody (1:10,000, #SA00001-2, Proteintech, Wuhan, China) or goat anti-mouse IgG/HRP secondary antibody (1:100,000, #SA00001-1, Proteintech, Wuhan, China) for 1 h at room temperature. Protein bands were visualized with an ECL kit (#RP-WA0601, Report Bio & Technology Co., Ltd, Hebei, China), and captured with ChemiScope S6 (CLINX, Shanghai, China). The intensity of target protein bands was normalized to the β-actin bands on the same blot.

Quantification and statistical analysis

Data are presented as means ± standard error of the mean (SEM). Data collection was randomized, and electrophysiological data and immunoblotting images were analyzed by investigators who were blinded from treatment groups. For electrophysiological recordings, only one neuron in each brain slice was recorded, and at least three rats were used for each group. The membrane potential and firing rate of neurons were analyzed using Clampfit 11 software (Molecular Devices, San Jose, CA, USA). E_GABA was determined by using linear regression to calculate a best-fit line for the voltage dependent GABA-induced currents. The intercept of the current–voltage line with the abscissa was taken as E_GABA, and then [Cl^-]_i was calculated using the Nernst equation.⁶³ The MAP, HR and RSNA were analyzed using Spike2 software (Cambridge Electronic Design, UK). Two-tailed Student t-test was used for comparison of two groups. One-way ANOVA followed by Tukey’s post hoc test was used for comparison of more than two groups. Statistical analyses were performed using Prism 9 software (GraphPad Software Inc., La Jolla, CA, USA). Statistical significance was denoted as ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P< 0.001.

Published: June 24, 2026