ALDH2 rs671 Variant Lowers Blood Pressure Via the Endothelial HSPA5–PIEZO1–eNOS Pathway

Abstract

Background

The inactivating ALDH2 rs671 polymorphism, present in ≈8% of the global population and up to 30% to 50% of East Asian individuals, has been negatively associated with hypertension. However, the direct role of ALDH2 (aldehyde dehydrogenase 2) dysfunction in this association, as well as the underlying mechanisms by which ALDH2 rs671 influences blood pressure regulation, remains unclear.

Methods

Using Aldh2 ^E506K (equivalent to the human ALDH2 rs671 polymorphism), Aldh2 global knockout, and endothelial cell‐specific Aldh2 knockout mice, we assessed changes in blood pressure (BP) using both telemetry and tail‐cuff methods. Mass spectrometry, coimmunoprecipitation, and other biological techniques were used to elucidate underlying molecular mechanisms.

Results

Aldh2 ^E506K or Aldh2 global knockout mice displayed a significant decrease in BP under both physiological conditions and angiotensin II–induced hypertension, independent of changes in cardiac output and renal function. Mechanistic studies revealed that endothelial ALDH2 dysfunction inhibited the lysosomal degradation of PIEZO1 (piezo‐type mechanosensitive ion channel component 1), resulting in its upregulated protein level and activation of the PIEZO1–endothelial nitric oxide synthase BP regulatory axis. PIEZO1 knockdown or channel inhibition by Grammostola spatulata mechanotoxin 4 peptide reversed the reduced BP phenotype exhibited by Aldh2 ^ECKO mice. Using coimmunoprecipitation combined with mass spectrometry, the molecular chaperone HSPA5 (heat shock protein family A member 5) was identified as a novel interactor with PIEZO1, binding to its C‐terminus. HSPA5, induced by aldehyde accumulation, functionally enhanced PIEZO1 protein stability upon ALDH2 loss‐of‐function mutation. Endothelial‐specific overexpression of HSPA5 could alleviate angiotensin II–induced hypertension in mice.

Conclusions

Our study provides definitive experimental evidence that ALDH2 rs671 variation or endothelial Aldh2 deficiency plays a protective role in BP regulation by activating the HSPA5–PIEZO1–endothelial nitric oxide synthase axis. We also revealed the previously unrecognized role of HSPA5 in maintaining PIEZO1 stability and BP regulation, providing a potential novel intervention target for hypertension therapy.

Nonstandard Abbreviations and Acronyms

4‐HNE

4‐hydroxy‐2‐nonenal

AAV9

adeno‐associated virus

ALDH2

aldehyde dehydrogenase 2

eNOS

endothelial nitric oxide synthase

GsMTx4

Grammostola spatulata mechanotoxin 4

HSPA5

heat shock protein family A member 5

HUVEC

human umbilical vein endothelial cell

ICAM2

intercellular adhesion molecule 2

MS

mass spectrometry

NO

nitric oxide

PIEZO1

piezo‐type mechanosensitive ion channel component 1

WT

wild‐type

Research Perspective.

What Is New?

• This study provides novel evidence that the ALDH2 rs671 polymorphism, prevalent in East Asian individuals, lowers blood pressure by activating the HSPA5 (heat shock protein family A member 5)–PIEZO1 (piezo‐type mechanosensitive ion channel component 1)–endothelial nitric oxide synthase axis in endothelial cells.

• The research identifies HSPA5 as a new molecular chaperone that stabilizes PIEZO1, a key ion channel involved in blood pressure regulation, in the context of ALDH2 (aldehyde dehydrogenase 2) dysfunction.

What Question Should Be Addressed Next?

• Further elucidation of the mechanisms underlying HSPA5’s regulation of PIEZO1 homeostasis and function in response to ALDH2 dysfunction, and future exploration and validation of the therapeutic potential of targeting the ALDH2–HSPA5–PIEZO1 pathway in hypertension treatment are warranted.

Hypertension remains a paramount concern owing to its extensive prevalence and the inherent risk of substantial cardiovascular complications and fatalities. Globally, its associated complications, including ischemia, myocardial infarction, heart failure, and stroke, contribute to ≈8.5 million deaths annually. ¹ In the United States, hypertension is particularly concerning, with a prevalence rate of 49.6%, indicating that ≈115 million individuals have hypertension. ² The widespread incidence and serious consequences of hypertension underscore the pressing demand for improved treatment strategies.

Genetic factors play a pivotal role in the onset of hypertension. Understanding the complex interplay between genetic variations and blood pressure (BP) regulation is critical for developing precise diagnostic and therapeutic approaches. Epidemiological studies have identified multiple genetic loci associated with hypertension across various populations, including those of European, African, Chinese, and Japanese ancestry. ³ , ⁴ , ⁵ Among these, the inactivating ALDH2 rs671 polymorphism, detected in ≈8% of the global population and up to 50% of East Asian individuals, has garnered considerable attention in recent years. ⁶ , ⁷ However, the association between ALDH2 rs671 polymorphism and hypertension remains inconclusive.

While several studies suggest that the rs671 variant may exert a protective effect against elevated BP, results have been inconsistent. Epidemiological research, ⁸ , ⁹ , ¹⁰ , ¹¹ including a large‐scale prospective study using data from the China Kadoorie Biobank, indicated that individuals carrying the ALDH2 rs671 G>A variant exhibit lower BP levels. Similarly, genome‐wide association studies ³ , ¹² , ¹³ support a potential negative effect of ALDH2 (aldehyde dehydrogenase 2) polymorphisms on BP. In contrast, a cross‐sectional study from Japan found no definitive causal relationship between Aldh2 and BP, owing to the presence of confounding factors. ¹⁴ Accordingly, the causal role and underlying mechanisms of ALDH2 rs671 in BP regulation remain unclear, and elucidating the specific role of ALDH2 rs671 in BP regulation may offer valuable insights for the precise diagnosis and personalized management of hypertension.

ALDH2 is a critical aldehyde‐metabolizing enzyme that reduces oxidative stress by detoxifying reactive aldehydes, such as 4‐hydroxy‐2‐nonenal (4‐HNE). ⁷ , ¹⁵ Dysfunction of ALDH2 leads to the accumulation of these aldehydes, triggering cellular stress responses. Endothelial ALDH2 plays a crucial role in maintaining vascular homeostasis. ¹⁶ , ¹⁷ , ¹⁸ For example, ALDH2 dysfunction preserves endothelial cell barrier function through the LIN28B–ELK3 signaling pathway. ¹⁹ However, how endothelial ALDH2 regulates vascular tone and influences BP remains unclear. Mechanosensation, the ability of endothelial cells to detect and respond to mechanical forces, is essential for vascular function and BP regulation. ²⁰ Endothelial cells rely on mechanosensitive ion channels, such as PIEZO1 (piezo‐type mechanosensitive ion channel component 1), to detect shear stress from blood flow and regulate vascular tone. Endothelium‐specific Piezo1 knockout mice exhibit arterial hypertension with impaired endothelial nitric oxide synthase (eNOS) activity, demonstrating PIEZO1’s crucial role in endothelial function. ²¹ , ²² , ²³ PIEZO1 was recently found to be sensitive to endothelial cell oxidative stress, serving as a key mediator that connects oxidative signaling to changes in mechanosensory function. ²⁴ Whether ALDH2 dysfunction could regulate BP by affecting the PIEZO1 mechanosensory function remains to be determined.

In this study, we aimed to elucidate the influence of ALDH2 and the ALDH2 rs671 variant on BP and hypertension and the underlying mechanism. We confirmed that Aldh2 mutation or endothelial Aldh2 knockout reduces BP and alleviates hypertension. Mechanistically, ALDH2 dysfunction suppressed the degradation of mechanosensory PIEZO1 protein, thereby upregulating PIEZO1–eNOS axis to reduce BP. Chaperone protein HSPA5 was found to interact with the C‐terminus of PIEZO1, thus enhancing its stability and suppressed degradation in response to ALDH2 dysfunction. Therefore, ALDH2 and HSPA5 could serve as novel PIEZO1 regulators and potential therapeutic targets for treating hypertension.

Methods

The methods and data used in this study will be made available by the corresponding author upon reasonable request.

Genetic Mouse Models

All animal experiments were approved by the Ethics Committee and Scientific Investigation Board of Qilu Hospital, Shandong University, and conducted in accordance with institutional guidelines and the US National Institutes of Health Guide for the Care and Use of Laboratory Animals (Publication No. 85–23, revised in 2011). We used 3 lines of ALDH2 genetic mice: Aldh2 global knockout (Aldh2 ^KO), Aldh2 ^E506K mice, and endothelial cell‐specific Aldh2 knockout (Aldh2 ^ECKO) mice. All animals were maintained at a constant temperature of 25 °C under a 12/12‐hour day/night cycle, with free access to water and standard chow. The sample size for each animal experiment is described in the figure legends. Mice used in this study were 8 weeks old at the start of the experiments.

Aldh2 ^KO was constructed by the University of Occupational and Environmental Health (Fukuoka, Japan) on a C57BL/6J (wild‐type [WT]) background and backcrossed for ≥10 generations with C57BL/6J mice to generate a congenic strain. WT littermates were used as controls. Aldh2 ^E506K mice carrying the E506K mutation corresponding to the human E504K mutation (ALDH2 rs671 G>A) were generated using clustered regularly interspaced short palindromic repeats/Cas9 nickase. The genomic RNA sequence flanking exon 12 with the mutation site was as follows: CACCATCACCACCTTATATGG and CACAGCCTTCAGACTCCGTGG. The donor oligo with the E506K mutation was coinjected into the C57BL/6J zygotes to introduce an E506K point mutation into the mouse genome. The Aldh2 ^E506K targeting allele was screened using a sequencing polymerase chain reaction (PCR) product. WT littermates were used as controls. Aldh2 ^f/f mice, characterized by the presence of floxP sites flanking exons 2 to 4 of Aldh2, with Tek‐creERT2 mice, were developed using the WT strain at the Animal Experimental Center of Beijing Viewsolid Biotech. By crossbreeding these Aldh2 ^f/f mice with Tek‐creERT2 strains, we generated Aldh2 ^ECKO mice. For comparison purposes, we used Aldh2 ^f/f mice as the control group.

Hypertension Mouse Model

To induce hypertension in animal models, 8‐week‐old mice were subcutaneously implanted with an osmotic minipump (2004; ALZET, Cupertino, CA) that infused angiotensin II (450 ng/kg per min) for indicated days. All hypertensive mice met the standard that systolic BP (SBP) was >130 mm Hg.

Disulfiram–Ethanol Reaction Mouse Model

To induce the disulfiram–ethanol reaction mouse model, ethanol was administered intraperitoneally to mice as a 20% solution in saline at a dose of 2 mg/kg. Disulfiram (HY‐B0240, MedChemExpress, Shanghai, China), suspended in corn oil, was administered intraperitoneally at a dose of 200 mg/kg, 24 hours before ethanol administration. Control mice received intraperitoneal injections of saline and corn oil in corresponding volumes to match the treatment groups. BP was measured using the tail‐cuff method 20 minutes after ethanol injection.

BP Measurements

Male or female 8‐week‐old mice were subjected to BP monitoring. BP was measured using tail‐cuff (BP2010A; Softron, Tokyo, Japan) and telemetric BP systems (HD‐X11 Transmitter; DataSciences International, St. Paul, MN). For the tail‐cuff technique, the mean value was derived from ≥3 successive measurements for each mouse. All mice underwent a 1‐week preconditioning phase. BP was monitored daily during the afternoon (2:00 pm to 5:00 pm) until the experiment was concluded. For the telemetric measurement, a radio‐telemetry system was used to monitor BP in conscious, unrestrained mice. Following general anesthesia, the catheter of the telemetric transducer was surgically inserted into the left common carotid artery of the mouse and carefully advanced to the thoracic aorta, consistent with methods previously outlined. ²³ , ²⁵ , ²⁶ The transducer body was securely positioned within a subcutaneous pouch along the right flank of the mouse, accessed via the initial ventral neck incision. Postsurgical recovery spanned 1 week before initiating radio‐telemetry assessments. BP readings were captured over a 1‐minute interval every 30 minutes using a 1‐kHz acquisition rate, covering daytime (7:00 am to 7:00 pm) and nighttime (7:00 pm to 7:00 am).

Echocardiography

Echocardiographic parameters were determined using a high‐resolution ultrasound imaging system, a Vevo3100 system (Fujifilm Visual Sonics, Toronto, ON, Canada), in a blinded manner. Mice were anesthetized with 1.2% to 1.5% isoflurane. Measurements encompassed the following parameters: left ventricular (LV) ejection fraction, LV fractional shortening, LV internal diameter, LV posterior wall thickness, and LV volume.

Histological Staining

To identify pathomorphological alterations in the heart and kidneys, tissues were harvested from mice and preserved in 4% paraformaldehyde for 12 hours. Subsequently, the specimens were sectioned into 4‐μm‐thick slices for histopathological evaluation. Hematoxylin and eosin (G1120; Solarbio, Beijing, China) staining was conducted to observe the alterations in structural morphology of the heart tissue. Wheat germ–agglutinin (29 022; Biotium, Fremont, CA) staining was performed in accordance with the manufacturer’s instructions to analyze the cardiomyocytes cross‐sectional area. Periodic acid–Schiff (G1281, Solarbio) staining was performed to evaluate the morphology alterations in kidney tissue. Quantification was performed using the ImageJ software (National Institutes of Health, Bethesda, MD) for digital image analysis.

RNA Microarray

RNA quantity and quality were measured using a NanoDrop ND‐1000. RNA integrity was assessed by standard denaturing agarose gel electrophoresis. Arraystar Mouse Microarray version 3.0 was used for global profiling of mouse protein‐coding transcripts. Sample labeling and array hybridization followed the Agilent One‐Color Microarray‐Based Gene Expression Analysis protocol (Agilent Technologies, Santa Clara, CA) with minor modifications. mRNA was purified from total RNA after the removal of rRNA (MRNA‐ONLYrm Eukaryotic mRNA Isolation Kit, Epicenter). Subsequently, each sample was amplified and transcribed into fluorescent cRNA. Then, 1 μg of labeled cRNA was fragmented with 5 μL 10 × blocking agent and 1 μL 25 × fragmentation buffer, heated at 60 °C for 30 minutes, and diluted with 25 μL 2 × GE Hybridization buffer. This was followed by the application of 50 μL of hybridization solution to the gasket slide and assembly onto the RNA microarray slide. Slides were incubated for 17 hours at 65 °C in an Agilent Hybridization Oven. Finally, hybridized arrays were washed, fixed, and scanned using the Agilent DNA Microarray Scanner (G2505C). All mRNA expression levels are listed in Data S1.

Enzyme‐Linked Immunosorbent Assay

Plasma samples were collected from the WT and Aldh2 ^KO 8‐week‐old male mice; levels of angiotensin II (catalog no. CEA005Mu), renin (catalog no. SEA889Mu), and aldosterone (catalog no. CEA911Ge) were measured using corresponding ELISA kits purchased from Cloud‐Clone Corporation (Shanghai, China) according to the manufacturer’s protocols.

Adeno‐Associated Virus Delivery

To achieve specific knockdown or overexpression of a specific gene in vivo, mice were administered via the tail vein injection of a recombinant virus. For the endothelial PIEZO1 gene knockdown in vivo, mice were administered via the tail vein injection of a recombinant adeno‐associated virus (AAV) serotype 9 system, using an ICAM2 (intercellular adhesion molecule 2) promoter. The AAV9 short hairpin RNA (shRNA) targeted PIEZO1 was incorporated into the pAV‐ICAM2‐GFP‐mir30‐shRNA (pAV, adeno‐associated virus vector backbone; GFP, green fluorescent protein; miR30, microRNA‐30‐based shRNA scaffold) vector to generate the PIEZO1‐RNAi virus vector. The backbone vector pAV‐ICAM2‐GFP‐mir30 was used as a control. For HSPA5 overexpression, an AAV9 construct containing the HSPA5 coding sequence was inserted into the pAV‐ICAM2‐GFP vector to create the HSPA5 overexpression virus. All viral constructs were produced by Wzbio (Shandong, China). Mice were administered with the aforementioned viruses (5 × 10¹¹ pfu/kg) via tail vein injection and allowed a 4‐week period for sufficient gene overexpression or knockdown.

Cell Culture

Human umbilical vein endothelial cells (HUVECs) were obtained from Procell Life Science and Technology with the catalog number CL‐0675 (Wuhan, China). The HUVECs were cultured in RPMI 1640 medium from Sigma (8758; Sigma). The mouse aortic endothelial cells purchased from Procell Life Science and Technology (CP‐M075; Wuhan, China) were primary cultured cells isolated from the mouse aortic tissue. Mouse aortic endothelial cells were cultured using specific culture medium obtained from Procell (CM‐M075; Wuhan, China). Human embryonic kidney 293T cells were sourced from the American Type Culture Collection (Manassas, VA) and were cultured in high‐glucose DMEM, supplemented with 10% FBS, 100 units/mL penicillin, and 100 μg/mL streptomycin. All cells were maintained at 37 °C in a humidified atmosphere with 5% CO₂.

Reagents

Reagents including angiotensin II (HY‐13948), Grammostola spatulata mechanotoxin 4 (GsMTx4; HY‐P1410), chloroquine (HY‐17589A), MG132 (HY‐13259), cycloheximide (HY‐12320), and rapamycin (HY‐10219) were all obtained from MedChemExpress (Shanghai, China).

Plasmid Construction and Cell Transfection

The human PIEZO1, ALDH2, calnexin, and HSPA5 gene cDNA sequenced were amplified and cloned into the backbone vector pcDNA3.1 purchased from Genechem (Shanghai, China). Verification of all constructs was achieved through DNA sequencing. For plasmid cell transfection, cells were seeded onto the 6‐well plates and cultured for 24 hours to reach 60% to 80% density. Transfection was performed using lipofectamine 2000 (11 668 030; Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s instruction. The transfection medium was replaced 6 hours after transfection, and the cells were harvested after an additional 24‐ to 72‐hour culture for subsequent experimental assay.

RNA Interference

Small interfering RNAs (siRNAs) targeting ALDH2, PIEZO1, calnexin, and a nontargeting scramble control were acquired from GenePharma (Shanghai, China). The scrambled siRNA was used as control. All siRNA sequences were provided in Table S1. Lipofectamine RNAiMAX (13 778 150; Thermo Fisher Scientific) was used to achieve the transient transfection of siRNA duplexes into the HUVECs or 293T. Cells were seeded onto the 6‐well plates and cultured to reach 60% to 80% density. The transfection medium was replaced 6 hours after transfection, and the cells were harvested after an additional 24‐ to 72‐hour culture for subsequent experimental assay.

Cycloheximide Chase Experiment

HUVECs were seeded at the same density in culture dishes and cultured for 24 hours to reach 40% to 50% confluency. Cells were transfected with siALDH2 or a control Scramble sequence using Lipofectamine RNAiMAX (13 778 150; Thermo Fisher Scientific). After 48 hours of transfection, HUVECs were treated with cycloheximide (20μM). Cells were collected at 0, 2, 4, and 10 hours after cycloheximide treatment, and total protein was extracted to assess PIEZO1 protein levels using western blot analysis.

Immunoprecipitation and Coimmunoprecipitation

For immunoprecipitation, cells were washed 3 times, harvested into ice‐cold PBS, and reconstituted using cell lysis buffer (9803; CST). The subsequent protein supernatant was incubated with 2 μg of an antibody targeting specific proteins, using immunoglobulin G as a negative control, under 4 °C with overnight incubation and continuous rotation. The formed immune complexes were incubated with 40 μL of protein A/G agarose for 3 hours of rotation, followed by 3 washes with the lysis buffer. Thereafter, the immunoprecipitates were collected. For coimmunoprecipitation, cells were lysed with immunoprecipitation lysis buffer, as described previously. The proteins were merged with the primary antibody and incubated overnight at 4 °C. Subsequently, protein A/G beads were added to each sample, followed by a 3‐hour incubation of the lysate beads mixture at 4 °C under rotary agitation. After immunoprecipitation, the samples were washed with immunoprecipitation lysis buffer 3 times. Proteins were eluted by resuspending with 1× SDS‐PAGE loading buffer and boiling for 5 minutes. The resulting immunoprecipitates and cell lysates were subjected to western blot analysis.

Western Blot Analysis

Briefly, cells and tissues are lysed with radioimmunoprecipitation assay buffer, integrating protease and phosphatase inhibitors. Equal protein concentrations, verified using bicinchoninic acid assay, underwent separation via SDS‐PAGE and were transferred to polyvinylidene fluoride membranes. The membranes, once blocked in 5% nonfat milk, were probed overnight at 4 °C with specific primary antibodies, followed by incubation with horseradish peroxidase–conjugated secondary antibodies. The band signal intensity was captured and analyzed using a chemiluminescence instrument (General Electric Company). Subsequent quantification of these immunoreactive bands was conducted using ImageJ software. The results were normalized to β‐actin and expressed as a percentage of the control. To validate the consistency and reliability of the results, each experiment was replicated at least 3 times. All antibodies used in this study are listed in the Table S2.

Quantitative Reverse Transcription PCR

Total RNA from tissues or cells was initially extracted using Trizol Reagent (T9424; Sigma) following mechanical homogenization. Extracted RNA (1 μg) was reverse‐transcribed into cDNA using the Reverse Transcription System kit (R223‐01; Vazyme). Subsequent quantitative PCRs were conducted using specific primers and ChamQ Universal SYBR Quantitative PCR Master Mix (Q711‐02; Vazyme). GAPDH mRNA served as an internal control and was coamplified with target sequences for normalization purposes. The analysis used the 2^−∆∆Ct method to determine relative changes in mRNA expression levels, quantifying fold changes about a control. All primer sequences used in reverse transcription quantitative PCR are listed in Table S3.

Tandem Mass Spectrometry

In the mass spectrometry (MS) experiment, samples were thawed on ice and pH adjusted using triethylammonium bicarbonate. Proteins were extracted for SDS‐PAGE. During trypsin digestion, proteins were reduced, alkylated, diluted, and digested twice with trypsin. In liquid chromatography‐MS/MS analysis, the peptides, which were dissolved in solvents, were separated using a gradient on an ultra‐performance liquid chromatography system and analyzed using the Q Exactive Plus MS. The MS/MS data were processed using the MaxQuant search engine, searching against a specific human protein database with set parameters for ion tolerance, modifications, and false discovery rate control. Protein expression levels are listed in Data S2.

Statistical Analysis

The normality of data distribution was assessed using the Shapiro–Wilk test, followed by the Brown–Forsythe test to evaluate the equality of variances. Results are presented as mean±SEM, with each experiment conducted at least 3 times to ensure significance and reproducibility. For comparisons between the 2 groups, we used unpaired 2‐tailed Student t tests. For multiple group analyses, depending on the experiment’s design and variables involved, we applied 1‐way or 2‐way ANOVA with Bonferroni’s post hoc correction. The specific tests applied for each analysis are detailed within the figure legends. Statistical significance was set at P<0.05. All statistical analyses were conducted using Prism version 9.0 (GraphPad Software, Inc., La Jolla, CA).

Results

Aldh2 ^E506K and Aldh2 ^KO Mice Exhibited Lower BP Under Physiological Conditions

We first assessed BP differences between Aldh2 ^E506K male mice and their WT controls over a 14‐day continuous and high‐resolution telemetry monitoring period after confirming the genetic background of mice by sequencing (Figure S1A). Aldh2 ^E506K mice exhibited lower SBP (Figure 1A), lower diastolic BP (DBP) (Figure 1B), and lower mean arterial pressure (MAP) (Figure S1B) without alterations in the heart rates (Figure S1C) and activity levels (Figure S1D).

Figure 1. ALDH2 deficiency resulted in a reduction in baseline blood pressure.

A, (SBP and (B) DBP in Aldh2 mutant (Aldh2 ^E506K) male mice and the WT littermates over 14 days of telemetry monitoring (n=6 per group). C, SBP and (D) DBP in Aldh2 knockout male mice (Aldh2 ^KO) and WT littermates over 14 days of telemetry monitoring (n=6 per group). E, SBP and (F) DBP in Aldh2 ^KO male mice and WT littermates over 24 hours of telemetry monitoring (n=6 per group). All BP measurements were performed on 8‐week‐old mice. Data are presented as mean±SEM with P values indicated; repeated measures 2‐way ANOVA with Bonferroni’s post hoc test was used for statistical analysis. ALDH2 indicates aldehyde dehydrogenase 2; BP, blood pressure; DBP, diastolic blood pressure; SBP, systolic blood pressure; and WT, wild‐type.

Since the Aldh2 ^E506K is a loss‐of‐function mutation with a substantial loss of its enzymatic activity, we speculated that the loss‐of‐function of ALDH2 would lead to a reduction in BP. We subsequently examined whether Aldh2 ^KO conferred a similar phenotype in male mice. The genotypes of Aldh2 ^KO mice and their littermate WT controls were first confirmed (Figure S2A). Similar to Aldh2 ^E506K mice, over a 14‐day telemetry monitoring period, Aldh2 ^KO male mice maintained a lower SBP (Figure 1C), DBP (Figure 1D), and MAP (Figure S2B) without alterations in heart rates (Figure S2C) and activity counts (Figure S2D). Moreover, male Aldh2 ^KO mice demonstrated consistently lower BP during dark and light periods (Figure 1E and 1F; Figure S2E), showing no variation in heart rates (Figure S2F) and activity levels (Figure S2G).

After confirming the BP phenotype in male mice, we sought to determine whether the phenotype was consistent in female mice or exhibited sex‐specific differences. Considering the cost and technical complexity of radiotelemetry, BP in female mice was measured using the tail‐cuff method, a practical, noninvasive, and cost‐effective alternative for assessing BP. Aldh2 mutated (Figure S3A through S3D) or Aldh2 ^ECKO (Figure S3E through S3F) female mice demonstrated lower BP, suggesting that ALDH2 could regulate BP in a sex‐independent manner. These findings are consistent with human cohort studies ¹⁰ , ¹¹ and suggest that ALDH2 might be a therapeutic target in regulating BP.

Aldh2 ^ECKO Alleviated Angiotensin II–Induced Hypertension in Mice

After confirming the BP‐reducing effect of ALDH2 loss‐of‐function at baseline level, we further evaluated whether this BP‐reducing effect still exists under hypertensive conditions via the angiotensin II–induced hypertension mouse model (Figure 2A). In this model, angiotensin II administration effectively increased SBP (Figure 2B), DBP (Figure 2C), and MAP (Figure 2D) without impacting the heart rate (Figure S4A) and activity levels (Figure S4B). Consistent with the results in baseline and clinical observations, ¹⁰ , ¹¹ Aldh2 ^KO mice exhibited significant attenuated levels of SBP (Figure 2B), DBP (Figure 2C), and MAP (Figure 2D) compared with WT mice under the treatment of angiotensin II. Therefore, both at baseline and under hypertensive conditions, Aldh2 ^KO consistently led to lower BP levels. Compared with baseline levels, the extent of BP reduction caused by Aldh2 ^ECKO appears to be enhanced after angiotensin II stimulation. However, this difference did not reach statistical significance in our measurement (Figure S4C and S4D). The overall reduced level of BP upon ALDH2 loss‐of‐function suggests that the ALDH2 rs671 variant might be a potential protective factor for hypertension.

Figure 2. ALDH2 deficiency alleviated angiotensin II–induced hypertension.

A, Schematic diagram showing the timeline of telemetry monitoring, angiotensin II minipump implantation, and the end point. B, SBP, (C) DBP, and (D) MAP monitored via telemetry in WT and Aldh2 ^KO male mice during angiotensin II administration (n=6 per group). Data are presented as mean±SEM with P values indicated; repeated‐measures 2‐way ANOVA with Bonferroni’s post hoc test was used for statistical analysis. ALDH2 indicates aldehyde dehydrogenase 2; DBP, diastolic blood pressure; MAP, mean arterial pressure; SBP, systolic blood pressure; and WT, wild‐type.

ALDH2 Regulated BP Independent of Cardiac and Renal Functions

The regulation of BP involves multiple organ interplay between cardiac output, peripheral vascular resistance, the renin–angiotensin–aldosterone system, and renal function. ²⁷ , ²⁸ To further elucidate the mechanism of ALDH2 in BP regulation, we subsequently investigated its impact on the aforementioned organs. All analyses were conducted on 8‐week‐old male mice without any prior interventions to ensure baseline physiological conditions. No significant differences in the heart and kidney morphology were detected between WT and Aldh2 ^KO mice (see Figure S5A for heart histology and Figure S6A for kidney morphology). Cardiomyocyte size was comparable between both groups (Figure S5B). Echocardiographic analysis showed similar LV ejection fraction, LV fractional shortening, LV internal diameter, LV posterior wall thickness, and LV volume in WT and Aldh2 ^KO mice (Figure S5C through S5K). RNA microarray analysis of the kidney revealed no significant differences in gene expression related to BP regulation pathways between the genotypes (Figure S6B and S6C). Additionally, plasma levels of renin–angiotensin–aldosterone system components, including renin, aldosterone, and angiotensin II, did not differ significantly between the 2 groups (Figure S6D through S6F). Collectively, ALDH2 deficiency lowered BP without altering cardiac and renal function. This observation has redirected our focus toward vascular resistance, a critical component of BP regulation.

Endothelial Cell–Specific Aldh2 ^ECKO Reduced BP in Mice

As the endothelium plays a pivotal role in sensing and fine‐tuning vascular BP, we examined the role of the endothelium in the ALDH2‐mediated regulation of BP by developing a mouse model with Aldh2 ^ECKO using a Tek‐CreERT2–driven Cre‐LoxP recombination (Figure 3A). Genotyping of Aldh2 ^ECKO is shown in Figure S7A and S7B. We validated the efficiency of the endothelial cell–specific Aldh2 ^ECKO by detecting the protein level of ALDH2 in stripped endothelium (Figure 3B and 3C). Aldh2 ^ECKO mice demonstrated reduced SBP (Figure 3D), DBP (Figure 3E), and MAP (Figure S7C) compared with their respective controls. No changes in heart rate and activity levels were observed (Figure S7D and S7E). We further analyzed female mice and found that Aldh2 ^ECKO lowered BP in a sex‐independent manner (Figure S7F through S7I). Overall, these findings underscored the crucial role of endothelial ALDH2 in BP regulation.

Figure 3. Endothelial cell–specific Aldh2 knockout phenocopied global Aldh2 knockout.

A, A schematic of endothelial cell‐specific Aldh2 knockout (Aldh2 ^ECKO) mice generated by crossing Tek‐CreERT2 mice with Aldh2 ^f/f mice. B, C, Representative western blot images and quantification of ALDH2 protein levels in aortic intima stripped from Aldh2 ^f/f and Aldh2 ^ECKO male mice (n=6 per group). D, E, Telemetry monitoring of SBP and DBP over a 14‐day period in Aldh2 ^f/f and Aldh2 ^ECKO male mice (n=6 per group). Data are presented as mean ± SEM with P values indicated; Student t tests were used for analysis in (C), repeated‐measures 2‐way ANOVA with Bonferroni’s post hoc test was used for statistical analysis in (D) and (E). ALDH2 indicates aldehyde dehydrogenase 2; DBP, diastolic blood pressure; and SBP, systolic blood pressure.

Aldh2 ^ECKO Activated the PIEZO1–eNOS Axis in Vascular Tissues

We further explored the molecular pathway involved in ALDH2 deficiency–induced BP alteration in blood vessels. Major determinants of BP include vasodilatory factors such as nitric oxide (NO) generated from the eNOS under the influence of fluid shear stress exerted by the flowing blood. ²³ Accordingly, we examined eNOS phosphorylation (at serine 1177) and the upstream Akt (at serine 473) in both the aortic and mesenteric arteries of Aldh2 ^KO mice. As was shown in the results, Aldh2 ^KO mice exhibited upregulated levels of phosphorylated eNOS and phosphorylated Akt in the aorta (Figure 4C and 4D) and mesenteric artery (Figure 4H and 4I) when compared with WT mice. Consequently, plasma NO levels were elevated in Aldh2 ^E506K and Aldh2 ^KO mice when compared with those in WT control mice (Figure 4K and 4L). These results indicated that ALDH2 deficiency could promote Akt‐mediated eNOS phosphorylation and NO release to fine‐tune vascular tone homeostasis. Endothelial cells are endowed with an array of mechanosensors that regulate cellular homeostasis and vascular function, with PIEZO1 identified as a critical component. ²⁹ The PIEZO1–eNOS axis is central to the physiological modulation of BP. ²¹ , ²² , ²³ Given its importance, we then examined PIEZO1 levels in vascular tissues derived from WT and Aldh2 ^KO mice. Western blotting analysis revealed the upregulated expression of PIEZO1 in the aortic tissues of Aldh2 ^KO mice when compared with those in WT controls (Figure 4A, 4B, and 4E). Consistently, alterations in PIEZO1 expression were observed in the mesenteric arteries, enriched in endothelial cells, further supporting the negative association between ALDH2 and PIEZO1 (Figure 4F, 4G, and 4J). We further validated the upregulation of PIEZO1‐eNOS signaling in Aldh2 ^ECKO mice. Western blot analysis showed increased PIEZO1 levels and a higher phosphorylated eNOS (S1177)/eNOS ratio (Figure S8A through S8D) in the aortic tissues of Aldh2 ^ECKO mice compared with Aldh2 ^f/f mice, indicating enhanced PIEZO1–eNOS signaling. Therefore, upregulation of the PIEZO1–eNOS axis underscored its potential role as a key downstream of ALDH2 in modulating BP.

Figure 4. ALDH2 deficiency upregulated PIEZO1 expression and downstream Akt/eNOS signaling in vascular tissues.

A, Representative western blot images showing the expression levels of PIEZO1, p‐eNOS, total eNOS, p‐Akt, total Akt, and ALDH2 in aorta tissues from WT and Aldh2 ^KO male mice. B–E, Quantitative analysis of PIEZO1, p‐eNOS/eNOS ratio, p‐Akt/Akt ratio, and ALDH2 in aortic tissues (n=8 per group). F, Representative western blot images of proteins in mesenteric artery tissues from WT and Aldh2 ^KO male mice. G–J, Quantitative analysis of PIEZO1, p‐eNOS/eNOS ratio, p‐Akt/Akt ratio, ALDH2 in mesenteric arteries (n=8 per group). K, Plasma nitrate and NOx levels in WT and Aldh2 ^E506K male mice (n=8 per group). L, Plasma nitrate and NOx levels in WT and Aldh2 ^KO male mice (n=8 per group). Data are presented as mean ± SEM; P values as indicated; statistical significance assessed using unpaired Student t tests. ALDH2 indicates aldehyde dehydrogenase 2; eNOS, endothelial nitric oxide synthase; p‐Akt, phosphorylated Akt; NOx, nitrite; p‐eNOS, phosphorylated endothelial nitric oxide synthase; PIEZO1, piezo‐type mechanosensitive ion channel component 1; and WT, wild‐type.

Upregulated PIEZO1 Contributed to ALDH2 Deficiency–Induced BP Reduction

Next, we examined whether upregulated PIEZO1 contributes to ALDH2 dysfunction–induced BP reduction. We first used AAV9–shRNA–PIEZO1 to silence the expression of PIEZO1 in vivo, and examined the BP alteration in Aldh2 ^f/f and Aldh2 ^ECKO mice. Initially, we screened the effective siRNA targeting Piezo1 mRNA in mouse aortic endothelial cells (Figure S9A), selecting the most efficient candidate (siRNA1) for constructing AAV9–shRNA–PIEZO1 vectors for in vivo studies. PIEZO1–shRNA was administered via tail vein injection, followed by a 4‐week period for effective gene knockdown, after which daily BP measurements commenced. After angiotensin II administration, BP increased steadily in both Aldh2 ^f/f and Aldh2 ^ECKO mice; heart rates were not affected across groups (Figure S9B). Piezo1 silencing could increase BP compared with the scramble group in both the Aldh2 ^f/f and Aldh2 ^ECKO mice (Figure 5A through 5C), suggesting that PIEZO1 plays a role in attenuating BP. Consistent with our previous results, Aldh2 ^ECKO mice exhibited significantly lower BP levels compared with Aldh2 ^f/f mice, confirming the BP‐lowering effects of ALDH2 deficiency under angiotensin II–induced hypertension (Figure 5A through 5C). However, following Piezo1 silencing, increased BP levels in Aldh2 ^ECKO mice were no longer statistically different compared with the Aldh2 ^f/f mice.

Figure 5. PIEZO1 inhibition could reverse the blood pressure reduction in ALDH2‐deficient mice.

A–C, Tail‐cuff measurements of SBP (A), DBP (B), and MAP (C) in Aldh2 ^f/f and Aldh2 ^ECKO male mice treated with Scramble or AAV9‐shPiezo1 (n=8 per group). (*: P<0.05, Aldh2 ^flox + Scramble vs Aldh2 ^ECKO + Scramble; ns: P>0.05, Aldh2 ^f/f+shPIEZO1 vs Aldh2 ^ECKO+shPIEZO1). D–F, Tail‐cuff measurements of SBP (D), DBP (E), and MAP (F) in Aldh2 ^f/f and Aldh2 ^ECKO male mice treated with vehicle or GsMTx4 (n=8 per group). (*P<0.05, Aldh2 ^f/f+vehicle vs Aldh2 ^ECKO+vehicle; ns: P>0.05, Aldh2 ^f/f +GsMTx4 vs Aldh2 ^ECKO+GsMTx4). Data are presented as mean ± SEM; repeated‐measures 2‐way ANOVA with Bonferroni’s post hoc test was used for analysis. AAV9 indicates adeno‐associated virus serotype 9; ALDH2, aldehyde dehydrogenase 2; Ang II, angiotensin II; DBP, diastolic blood pressure; GsMTx4, Grammostola spatulata mechanotoxin 4; MAP, mean arterial pressure; ns, nonsignificant; PIEZO1, piezo‐type mechanosensitive ion channel component 1; and SBP, systolic blood pressure.

As a mechanosensitive cation channel, PIEZO1 was reported to sense the shear stress and regulate BP, relying on its channel activity. ²¹ Thereby, we then used the PEIZO1 channel inhibitor to further validate the activity of PIEZO1 channel in ALDH2‐related BP regulation. GsMTx4, a mechanosensitive ion channel inhibitor, was used to assess the functional role of PIEZO1 channel activity in ALDH2‐mediated BP regulation. GsMTx4 was administered intraperitoneally at 10 mg/kg every 2 days for 14 days, with daily BP measurements starting on day 7 of GsMTx4 treatment. Like the effect of Piezo1 silencing, PEIZO1 inhibition by GsMTx4 could also reverse the BP reductions in Aldh2 ^ECKO mice (Figure 5D through 5F; Figure S9C).

Taken together, these findings demonstrated that upregulated PIEZO1 expression and function is critical for the BP‐lowering effects induced by ALDH2 dysfunction, highlighting the mechanistic link between ALDH2 and PIEZO1 in endothelial cells.

ALDH2 Dysfunction Suppressed the Lysosomal Degradation of PIEZO1 Protein

We then investigated how ALDH2 deficiency upregulates PIEZO1 in the endothelium. Silencing ALDH2 in HUVECs was shown to increase the PIEZO1 protein level (Figure 6A through 6C), and this effect was most pronounced 72 hours post siRNA transfection (Figure S10A through S10C). The mRNA level of Piezo1 was found unaltered, prompting us to speculate that ALDH2 modulates PIEZO1 via posttranscriptional mechanisms (Figure 6D). The cycloheximide chase experiment demonstrated a prolonged half‐life upon Aldh2 knockdown, suggesting PIEZO1 degradation might be compromised (Figure 6E and 6F). We further determined the degradation pathway of PIEZO1. We found that suppressing the lysosomal degradation pathway by using chloroquine, a lysosomal inhibitor, dramatically elevated the PIEZO1 protein level in the scramble groups, suggesting that the lysosomal degradation pathway is responsible for PIEZO1 degradation. The Aldh2 knockdown by Aldh2‐siRNA could effectively increase PIEZO1 protein level, with a similar effect to that of the chloroquine treatment. However, when blocking the lysosomal degradation pathway by chloroquine treatment, silencing Aldh2 failed to further increase the PIEZO1 level in comparison with the scramble group. This suggests that the lysosomal degradation pathway mediates the elevation of PIEZO1 caused by ALDH2 dysfunction. On the other hand, MG132, a proteasome pathway inhibitor, did not impact PIEZO1 levels in endothelial cells. And suppressing the proteasomal degradation pathway by MG132 could not block the elevation of PEIZO1 induced by Aldh2 silencing (Figure 6G through 6J). This differential response suggested that the lysosomal pathway, rather than the proteasomal pathway, may impact PIEZO1 protein degradation. Collectively, these results revealed that ALDH2 deficiency could suppress the lysosomal degradation of PIEZO1.

Figure 6. Aldh2 knockdown elevated PIEZO1 level via inhibiting its lysosomal degradation.

A–C, Representative western blot images and quantitative analysis for PIEZO1 and ALDH2 protein levels following treatment with Scramble or siALDH2 in HUVECs (n=7 per group), with Student t tests used for statistical analysis. D, Relative mRNA level of Aldh2 and Piezo1 following Aldh2 knockdown (n=12 per group); 1‐way ANOVA with Bonferroni’s post hoc test was used for analysis. E, F, Representative western blot images and quantification of PIEZO1 levels following CHX chase assay to assess protein stability (n=3 per group); 2‐way ANOVA with Bonferroni’s post hoc test was used for analysis. G, H, Representative western blot images and quantification of PIEZO1 expression following CQ treatment to evaluate the involvement of lysosomal degradation pathway (n=5 per group); 1‐way ANOVA with Bonferroni’s post hoc test was used for analysis. I, J, Representative western blot images and quantification of PIEZO1 expression following MG132 treatment to assess the involvement of proteasomal degradation pathway (n=5 per group); 1‐way ANOVA with Bonferroni’s post hoc test was used for analysis. Data are presented as mean ± SEM; P values from statistical tests are indicated. ALDH2 indicates aldehyde dehydrogenase 2; CHX, cycloheximide; CQ, chloroquine; HUVEC, human umbilical vein endothelial cell; and PIEZO1, piezo‐type mechanosensitive ion channel component 1.

Enhanced Binding of PIEZO1 to the Molecular Chaperone HSPA5 Increased PIEZO1 Stability in the Presence of ALDH2 Deficiency

We next examined the interplay between ALDH2 and PEIZO1 to further uncover how ALDH2 deficiency increased PIEZO1 protein homeostasis. We conducted coimmunoprecipitation experiments to assess the interaction between ALDH2 and PIEZO1. When flag‐tagged ALDH2 was immunoprecipitated, subsequent western blotting did not detect PIEZO1 in the precipitated complex (Figure S10D). Consistently, immunoprecipitation of myc‐tagged PIEZO1 did not coprecipitate ALDH2 (Figure S10E), excluding the possibility of direct interaction between ALDH2 and PIEZO1. To identify intermediate molecules that enhanced PIEZO1 stability, we examined the proteomic landscape of PIEZO1 interactome upon Aldh2 knockdown (Figure 7A). Myc‐tagged PIEZO1 was immunoprecipitated, and the coprecipitated protein samples were subjected to silver staining, followed by subsequent proteomic analyses (Figure S11A). Clusters of Orthologous/EuKaryotic Orthologous Groups analysis indicate that, among proteins interacted with PIEZO1, those involved in posttranslational modification, protein turnover, and chaperoning were profoundly altered upon ALDH2 knockdown (Figure 7B). The heatmap analysis indicated potential enhanced interaction of molecular chaperones including calnexin and HSPA5 with PIEZO1 (Figure 7C). The MS/MS tandem mass spectrum for HSPA5 and calnexin detailing the identified peptide fragments was shown (Figure S11B and S11C). Since chaperone proteins play important roles in maintaining cellular protein homeostasis, we thus speculated that enhanced interaction between chaperone protein calnexin or HSPA5 with PIEZO1 protein might contribute to upregulated PIEZO1 upon ALDH2 loss of function.

Figure 7. Enhanced binding of PIEZO1 to molecular chaperone HSPA5 increased its protein stability upon ALDH2 dysfunction.

A, Flow chart of MS/MS assay and representative western blot images for PIEZO1 and ALDH2 protein levels following transfection with ALDH2‐siRNA or myc‐PIEZO1 plasmid in 293T cells. B, COG/KOG analysis of proteins differentially coimmunoprecipitated with myc‐PIEZO1 upon Aldh2 knockdown. C, Heatmap displaying proteins differentially coimmunoprecipitated with PIEZO1 upon Aldh2 knockdown, with molecular chaperones calnexin and HSPA5 highlighted in red. D, Coimmunoprecipitation assay of exogenous expressed flag‐HSPA5 and myc‐PIEZO1 in 293T cells. Myc‐PIEZO1 was immunoprecipitated with anti‐myc antibody and the co‐precipitated flag‐HSPA5 was subjected to western blot assay. E, Coimmunoprecipitation assay of endogenous HSPA5 and PIEZO1 in HUVECs. Endogenous HSPA5 was immunoprecipitated by anti‐HSPA5 and coprecipitated PIEZO1 was detected by western blot. F–H, Representative western blot images and quantitative analysis for HSPA5 and PIEZO1 protein levels following HSPA5 overexpression (n=5 per group). I, Structural representation of the interaction between human PIEZO1 and HSPA5 proteins obtained through molecular docking and molecular dynamics simulation. PIEZO1 is shown in green, cyan, and magenta representing its 3 subunits, while HSPA5 is depicted in blue. The binding interface between the 2 proteins is shown, with a calculated binding energy of −92.301 kcal/mol. J, Schematic representation of the truncated PIEZO1 constructs used for mapping HSPA5 interaction. K, Coimmunoprecipitation assay showing the interaction between HSPA5 and different PIEZO1 constructs. Data are presented as mean±SEM; P values from statistical tests are indicated; Student t tests were used for analysis. ALDH2 indicates aldehyde dehydrogenase 2; COG/KOG, Clusters of Orthologous/Eukaryotic Orthologous Groups; HSPA5, heat shock protein family A member 5; HUVEC, human umbilical vein endothelial cell; IgG, immunoglobulin G; IP, immunoprecipitation; MS, mass spectrometry; and PIEZO1, piezo‐type mechanosensitive ion channel component 1.

We then verified the interaction and functional role of chaperone protein calnexin or HSPA5 on PIEZO1. Even coimmunoprecipitation experiments showed that exogenous calnexin could interact with PlEZO1 in 293T cells (Figure S11D). However, contrary to our initial hypothesis, calnexin knockdown in HUVECs led to an increase in the protein level of PIEZO1 (Figure S11E through S11G). Thus, we performed further analysis to investigate the interaction and regulatory effects between HSPA5 and PIEZO1. Considering HSPA5, we found that both exogenous and endogenous HSPA5 coimmunoprecipitated with ALDH2 (Figure 7D and 7E). Additionally, enhanced interaction between HSPA5 and PIEZO1 was observed upon Aldh2 knockdown in HUVECs (Figure 7E). Functional analysis confirmed that HSPA5 overexpression could increase PIEZO1 protein level (Figure 7F through 7H). To determine whether HSPA5 overexpression reduces lysosomal degradation of PIEZO1, we treated HUVECs with rapamycin, a well‐established autophagic lysosomal activator, and assessed PIEZO1 protein levels. As expected, HSPA5 overexpression led to a significant increase in PIEZO1 expression. However, upon rapamycin treatment, PIEZO1 protein levels were markedly reduced in HSPA5‐overexpressing HUVECs (Figure S12A through S12C), suggesting that HSPA5 enhances PIEZO1 protein levels by inhibiting its autophagic lysosomal degradation. Hence, Aldh2 knockdown could upregulate PIEZO1 through enhancing HSPA5–PIEZO1 interaction. Using molecular docking and molecular dynamics simulation techniques, we further analyzed the potential interaction patterns between PIEZO1 and HSPA5. The study found that a good binding interface was formed between PIEZO1 and HSPA5 proteins, with HSPA5 being able to effectively bind to the C‐terminal region of the PIEZO1 protein (Figure 7I). By constructing various truncated forms of PIEZO1 (N‐terminus 1‐730aa, middle domain fragment 731‐1959aa, and C‐terminus 1960‐2520aa), we confirmed that HSPA5 interacted with the C‐terminus of PIEZO1, and may be significant for PIEZO1 regulation (Figure 7J through 7K). Collectively, these findings revealed the presence of a regulatory network wherein ALDH2 modulates PIEZO1 levels via intermediary chaperone molecules, emphasizing the role of HSPA5 as a pivotal contributor by potentially improving the stability of PIEZO1.

ALDH2 Dysfunction Elevated HSPA5 in Response to 4‐HNE Accumulation to Reduce BP

HSPA5 is an inducible chaperone protein assisting protein folding and quality control. ³⁰ , ³¹ ALDH2 works as aldehyde‐metabolizing enzyme in cells. HSPA5 expression is reportedly induced by ALDH2 dysfunction upon the accumulation of aldehyde, such as 4‐HNE, in neutrophils. ³² Reanalysis of published MS data ³³ revealed that inhibiting ALDH2 using daidzin could upregulate HSPA5 expression (Figure S12D and S12E). Herein, we found that ALDH2 silencing increased the HSPA5 level in HUVECs (Figure 8A through 8C). Further western blotting analysis of the aortic tissue confirmed elevated levels of 4‐HNE and HSPA5 in Aldh2 ^KO mice when compared with those in WT mice (Figure 8D through 8G). The above results indicate a potential role for HSPA5 in BP regulation.

Figure 8. ALDH2 dysfunction elevated HSPA5 in response to 4‐HNE accumulation to reduce BP.

A–C, Representative western blot images and quantitative analysis for ALDH2 and HSPA5 protein levels following Aldh2 knockdown in HUVEC (n=5 per group), Student t tests were used for analysis. D–G, Representative western blot images and quantitative analysis for ALDH2, 4‐HNE, and HSPA5 protein levels in the aorta of WT and Aldh2 ^KO male mice (n=10 per group), Student t tests were used for analysis. H, Tail‐cuff measurements of SBP in WT and Aldh2 ^E506K male mice with or without AAV–HSPA5 in angiotensin II–induced hypertensive mouse model. One‐way ANOVA with Bonferroni’s post hoc test was used for analysis (n=10 per group). I–K, Representative western blot images and quantification of PIEZO1, HSPA5, and ALDH2 protein levels in WT mice treated with angiotensin II and AAV‐HSPA5. One‐way ANOVA with Bonferroni’s post hoc test was used for analysis. L–O, Effects of disulfiram–ethanol treatment on blood pressure in WT and Aldh2 ^E506K male mice. Two‐way ANOVA with Bonferroni’s post hoc test was used for analysis (n=10 per group). Data are presented as mean±SEM; P values from statistical tests are indicated. 4‐HNE indicates 4‐hydroxy‐2‐nonenal; AAV, adeno‐associated virus; ALDH2, aldehyde dehydrogenase 2; Ang II, angiotensin II; BP, blood pressure; HSPA5, heat shock protein family A member 5; HUVEC, human umbilical vein endothelial cell; SBP, systolic blood pressure; and WT, wild‐type.

We then validated the BP reduction function of HSPA5 in angiotensin II–induced hypertension mouse model. We found that the addition of AAV–HSPA5 effectively moderates the angiotensin II–induced increase in BP in both WT and Aldh2 ^E506K mice, providing evidence that HSPA5 overexpression can counteract the hypertensive effects of angiotensin II (Figure 8H, Figure S13A through S13C). Western blot validated that AAV–HSPA5 delivery significantly increased HSPA5 expression and upregulated PIEZO1 in angiotensin II–treated WT mice (Figure 8I through 8K).

Disulfiram reaction is a frequently encountered clinical emergency characterized by significant aldehyde accumulation, and severe cases can lead to a reduction in BP. Studies have demonstrated that the disulfiram‐alcohol reaction may lead to profound hypotension. ³⁴ Previous research has suggested that aldehydes play a role in reducing BP during disulfiram–ethanol reactions ³⁴ , ³⁵ ; however, the underlying mechanism remained unclear. According to our results, both SBP and DBP decrease in response to disulfiram–ethanol treatment, with Aldh2 ^E506K mice showing a greater reduction compared with WT mice (Figure 8L through 8O), indicating then involvement of ALDH2–HSPA5–PIEZO1 signaling in this BP reduction.

DISCUSSION

The widespread incidence and severity of hypertension highlight the urgent need for more advanced and accurate treatment strategies. ¹ Multiple genetic variations associated with hypertension have been identified, although the specific roles of several of these variations remain unclear. ³ Translating genetic information from genome‐wide association studies into causal mechanisms and clinical applications is of considerable importance but remains largely unexplored. ALDH2 rs671 polymorphism, present in ≈8% of the global population and up to 50% of East Asians, has been associated with hypertension ⁶ , ⁷ ; however, conclusions in human populations remain debatable. ⁸ , ⁹ , ¹⁰ , ¹⁴ Experimental evidence confirming its role and mechanism in BP regulation has been ambiguous. Herein, we found that the ALDH2 rs671 polymorphism or ALDH2 loss of function can lower BP by upregulating the classical PIEZO1–eNOS axis, while HSPA5, in response to ALDH2 dysfunction and aldehyde accumulation, binds to the C‐terminus of PIEZO1, stabilizing the protein and preventing its lysosomal degradation, thereby reducing the BP. To the best of our knowledge, we, for the first time, unraveled the causal role of ALDH2 in BP reduction by using Aldh2 genetic mice (Aldh2 ^E506K, Aldh2 ^KO, and Aldh2 ^ECKO). Thus, targeting ALDH2 or the downstream HSPA5–PIEZO1 interaction could be a novel approach for treating hypertension (Figure 9).

Figure 9. Study summary.

Compared with WT mice, Aldh2 ^E506K mice exhibit elevated PIEZO1 expression due to impaired protein degradation. Shear stress activates PIEZO1 on endothelial cells, leading to the phosphorylation of Akt and eNOS, which promotes NO production and vasodilation. In the scenario of Aldh2 ^E506K mutation, increased levels of 4‐HNE lead to elevated HSPA5. This enhances the binding of HSPA5 to PIEZO1, thereby promoting PIEZO1 stability by suppressing its degradation via the lysosomal pathway. This results in elevation of PIEZO1 and the subsequent activation of downstream Akt/eNOS vessel dilation pathway, leading to reduced blood pressure. eNOS indicates endothelial nitric oxide synthase; NO, nitric oxide; p‐Akt, phosphorylated Akt; PIEZO1, piezo‐type mechanosensitive ion channel component 1; and WT, wild‐type.

Although the association between ALDH2 rs671 polymorphism and hypertension has been extensively explored, it remains inconclusive. In our study, we confirmed the BP‐lowering effect of ALDH2 loss of function in both the baseline level and the angiotensin II–induced hypertensive condition. However, whether the BP‐lowering effect would be more enhanced in angiotensin II–induced hypertension and its protective role in this pathological condition requires further clarification through larger sample sizes or more meticulous experiments. ALDH2 has been widely reported to regulate endothelial function, thereby participating in cardiovascular disease regulation. ¹⁹ , ³⁶ , ³⁷ However, no study has uncovered the mechanism through which ALDH2 regulates vascular tone and BP. BP is mechanically regulated by 2 principal cardiovascular factors: peripheral vascular resistance and cardiac output. Echocardiogram results showed no significant changes in cardiac function in ALDH2‐deficient mice, ruling out the possibility that ALDH2 could influence BP through cardiac output changes. Additionally, given the significance of renin–angiotensin–aldosterone system in BP regulation, ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ plasma concentration measurements of key endogenous renin–angiotensin–aldosterone system components, including renin, aldosterone, and angiotensin II, were measured in Aldh2 ^KO mice, revealing no substantial upregulation upon ALDH2 deficiency. Arterial BP is controlled by vasodilatory factors, such as NO, which are released from the endothelium under the influence of fluid shear stress exerted by flowing blood. ²² Endothelium PIEZO1 has been identified as the critical sensor of fluid shear stress, thereby regulating NO formation, vascular tone, and BP through activating the Akt/eNOS pathway. ²¹ , ²² , ²³ In our study, we observed substantial PIEZO1 accumulation and the phosphorylation of Akt and eNOS in mouse aortic tissues and mesenteric artery of Aldh2 ^KO mice and HUVECs with Aldh2 knockdown. shRNA‐mediated Piezo1 knockdown and Grammostola spatulata mechanotoxin 4–mediated PIEZO1 inhibition consistently reversed the reduced BP in Aldh2 ^ECKO mice, functionally confirming that PIEZO1 is the key downstream molecule of ALDH2 involved in BP regulation. However, it is important to note that GsMTx4 is not a PIEZO1‐specific inhibitor, as it broadly targets mechanosensitive ion channels by interacting with the lipid bilayer and altering membrane tension. ⁴² , ⁴³ While this nonspecificity is a limitation, the consistency between GsMTx4‐mediated functional inhibition and shRNA‐mediated genetic knockdown strongly supports the central role of PIEZO1 in the ALDH2‐mediated BP regulatory pathway. Herein, our study uncovered the relationship between ALDH2 and the PIEZO1/eNOS vascular tone modulation axis, thereby clarifying the mechanism underlying the regulatory effect of ALDH2 on BP.

The PIEZO family of mechanosensitive ion channels, consisting of 2 main members, PIEZO1 and PIEZO2, is activated by various mechanical stimuli and plays a key role in regulating fundamental physiological functions. Unlike PIEZO2, which is predominantly expressed in mechanosensory neurons, PIEZO1 is mainly expressed in the cardiovascular system and plays an important role in Ca²⁺ influx, endothelial cell function, and endothelium‐dependent vasorelaxation. ⁴⁴ Mice with induced endothelium‐specific PIEZO1 deficiency reportedly lost the ability to induce NO formation and vasodilation in response to blood flow, consequently developing hypertension. ²¹ The notable role of PIEZO1 in diverse physiological and pathological processes makes it an important therapeutic target for cardiovascular disease. ²⁹ , ⁴⁴ , ⁴⁵ However, a comprehensive understanding of the interacting molecules of the PIEZO1 protein and its biological function is lacking. Reportedly, cartilage oligomeric matrix protein can interact with PIEZO1 via its C terminus, and its interaction with PIEZO1 regulated endothelium‐dependent relaxation and BP by increasing PIEZO1 activity and NO production, thereby alleviating angiotensin II–induced hypertension. ⁴⁶ Our study confirmed that ALDH2 acts as an upstream factor affecting the protein turnover of PIEZO1. ALDH2 deficiency resulted in upregulated PIEZO1 expression and reduced BP in mice. Using coimmunoprecipitation and MS, we further identified HSPA5 as a novel interactor of PIEZO1, functionally acting as a pivotal molecular chaperone regulating PIEZO1 stability upon ALDH2 depletion. HSPA5 is a chaperone protein responsible for protein folding and quality control. ³⁰ , ³¹ , ³² HSPA5 plays a crucial role in stabilizing key proteins and maintaining cellular homeostasis in response to stress conditions. It has been suggested that HSPA5 may be involved in hypertension, although the underlying mechanisms remain poorly clarified. ¹⁸ , ⁴⁷ , ⁴⁸ , ⁴⁹ Our study confirmed that HSPA5 could reduce BP in angiotensin II–induced hypertension. The newly uncovered interaction between HSPA5 and PEIZO1 leads to the perspective that HSPA5 could reduce BP and alleviate hypertension by enhancing the endothelial PIEZO1–eNOS axis.

Like several other integral membrane proteins, PIEZO1 channels undergo biosynthetic quality control in the endoplasmic reticulum. To date, few studies have examined the protein biosynthesis and quality control of PIEZO1. Herein, we showed that ALDH2 deficiency could prolong the half‐life of the PIEZO1 protein. This suggests that the homeostasis of PIEZO1 and its mutated form may be orchestrated through a diverse quality control system, with degradation of WT PIEZO1 through the lysosomal pathway and that of mutants mainly via the ubiquitin–proteasome system. ⁵⁰ In this study, we confirmed that impaired lysosome‐dependent degradation activity facilitates PIEZO1 accumulation, consistent with previous reports that PIEZO1 is degraded through the lysosomal pathway. ⁵⁰ In our previous study, we reported the regulatory role of ALDH2 on the lysosomal–autophagy pathway, wherein the ALDH2 downregulation was shown to impair autophagy lysosomal activity. ⁵¹ This supports our hypothesis that ALDH2 deficiency leads to compromised PIEZO1 degradation through the autophagic‐lysosomal pathway. HSPA5 is an endoplasmic reticulum chaperone that plays a key role in protein folding and quality control, whose expression would be substantially enhanced in response to oxidative stress. ⁵² , ⁵³ HSPA5 was upregulated in the aorta of Aldh2 ^KO mice, and the interaction of HSPA5 with PIEZO1 was also enhanced upon Aldh2 knockdown, thereby leading to upregulated PIEZO1 protein levels. To the best of our knowledge, HSPA5 is the first chaperone protein reported to facilitate PIEZO1 homeostasis.

ALDH2 plays a crucial role in detoxifying both the alcohol‐derived acetaldehyde and endogenous aldehydes. Reduction or loss of the ALDH2 enzyme function would result in the accumulation of endogenous aldehydes, such as 4‐HNE. Elevated levels of 4‐HNE can induce severe oxidative stress. Hence, ALDH2 is not only a critical enzyme involved in ethanol metabolism but also an important factor against intracellular oxidative stress. HSPA5 can be induced by various endogenous stress conditions. Pharmacological inhibition of ALDH2 activity has been shown to upregulate HSPA5 due to the elevated toxic aldehydes. ⁵⁴ , ⁵⁵ Moreover, ALDH2 knockout mice exhibited elevated HSPA5 levels and accumulation of the endogenous aldehyde 4‐HNE in neutrophils. ³² Consistently, in our study, we found elevated HSPA5 and 4‐HNE in the aorta of Aldh2 ^KO mice. Therefore, although multiple reports have suggested that ALDH2 could regulate cholesterol biosynthesis and high‐density lipoprotein biogenesis in hepatocytes, ⁵⁶ , ⁵⁷ as well as foam cell formation ⁵⁸ and efferocytosis ⁵⁹ in macrophages via nonenzymatic pathways, here in our study, ALDH2 regulated the HSPA5–PIEZO1 axis in the arteries through an enzymatic mechanism. In clinical settings, disulfiram‐like reactions induced by alcohol are frequently lethal, leading to a dangerous drop in BP and potentially resulting in shock or death. Our findings suggest that the accumulated aldehyde 4‐HNE triggers the upregulation of HSPA5 and PIEZO1‐eNOS signaling may represent the underlying molecular mechanism.

Identifying ALDH2 as a direct modulator of BP opens up avenues for therapeutic applications in hypertension. However, it is important to acknowledge the potential dual impact of ALDH2 inhibition or knockdown. Although ALDH2 inhibition appears to be beneficial for treating hypertension and aortic aneurysm/dissection, ⁶⁰ this could simultaneously pose a risk of promoting ischemic cardiomyopathy, ⁶¹ which is intriguing because hypertension is well recognized as a pivotal risk factor for the development of cardiomyocyte hypertrophy and acute myocardial infarction. ⁶² , ⁶³ Therapies targeting ALDH2 for BP reduction were expected to reduce the development of acute myocardial infarction. However, the ALDH2 loss of function associated with the ALDH2 rs671 variant was shown to increase the incidence of acute myocardial infarction. ⁶⁴ This finding suggests that ALDH2 contributes to both hypertensive scenarios and ischemic cardiomyopathy through multiple and diverse pathways. For example, ALDH2 could suppress atherosclerosis by impacting Rac2‐mediated macrophage efferocytosis, ⁵⁹ reducing myocardial injury, ⁶⁵ or manipulating neutrophil extracellular traps. ⁶⁶ Consequently, pursuing direct ALDH2 inhibition may present challenges as a therapeutic goal. In light of these considerations, alternative strategies, such as developing cell/tissue‐specific drug delivery systems, may offer more nuanced and targeted approaches. Additionally, exploring downstream pathways like the HSPA5 pathway could unveil potential therapeutic targets with fewer off‐target effects. These avenues hold promise for developing more precise and effective treatments for hypertension while mitigating potential adverse effects associated with broad ALDH2 inhibition.

Sources of Funding

This work was supported by grants from the National Natural Science Foundation of China (82030059, U23A20485, 82102290, 82472215), Noncommunicable Chronic Diseases–National Science and Technology Major Project (2023ZD0505504, 2023ZD0505500), the Key R&D Program of Shandong Province (2022ZLGX03), and the Natural Science Foundation of Shandong Province (ZR2022QH014).

Disclosures

None.

Supporting information

Data S1

Data S2

Data S3

Data S4