Vacuum-stabilized lung window enables real-time and simultaneous imaging of vascular and calcium responses to hypoxia in vivo Authors: Jie Liu, Junyan Shao, Pu Liao, Xinyu Song, Wenbo Yang, Chenlong Yu, Rui Xiao, Liping Zhu, Qinghua Hu Published: 2026 ABSTRACT AbstractBackgroundThis study introduced a vacuum-stabilized lung window technique for intravital imaging, which for the first time enabled the simultaneous, high-resolution, real-time monitoring of both pulmonary arterial vascular reactivity and dynamic changes in intracellular calcium signaling smooth muscle cells during acute hypoxic pulmonary vasoconstriction (HPV) in spontaneously breathing mice.MethodsBy integrating two-photon microscopy with intravascular fluorescent tracers (FITC, Evans Blue) and a smooth muscle-specific genetically encoded calcium indicator (GCaMP6f), this technique enabled stable imaging of pulmonary arteries within a diameter range of 13.95–80.16 μm and supported the systematic extraction and analysis of calcium signals from multiple regions of interest along the longitudinal axis of the vascular wall.ResultsAcute hypoxia (10% O2, 30 min) elicited a time-dependent vasoconstriction, with an average 8% reduction in diameter of 30–50 μm pulmonary arteries. Hypoxia triggered PASMC calcium signals characterized by a distinct latency period and marked spatial heterogeneity, with the earliest calcium responses observed in 30–40 μm arteries. Pharmacological intervention with dasatinib prolonged the latency of the calcium response and enhanced the amplitude of calcium oscillations, with effects preferentially targeting vessels in the 30–50 μm range. Simultaneous recordings confirmed that the rise in intracellular calcium consistently preceded vasoconstriction, and revealed that dasatinib treatment modulated calcium-contraction coupling in a vessel size-dependent manner.ConclusionsThe technological platform established in this study overcomes the limitations of traditional approaches, simultaneously monitoring pulmonary artery function and cellular Ca2+ signaling mechanisms in HPV.Keywords: Hypoxic pulmonary vasoconstriction, In vivo imaging, Vacuum-stabilized lung window, Pulmonary artery, Calcium signaling Introduction Hypoxic pulmonary vasoconstriction (HPV) is a rapid, pronounced, and reversible response of the pulmonary circulation, wherein intrapulmonary arteries constrict in response to reduced alveolar oxygen tension (PO2). By diverting blood flow away from poorly ventilated lung regions, HPV maintains ventilation‑perfusion matching, optimizes pulmonary gas exchange, and supports systemic oxygen delivery [47]. Although adaptive in the short term, excessive or sustained hypoxic vasoconstriction elevates pulmonary vascular resistance and right ventricular afterload, contributing to the pathobiology of pulmonary hypertension and right ventricular dysfunction [21, 30, 37, 49, 50]. The principal site of HPV is believed to reside in small resistance-sized pulmonary arteries, where pulmonary arterial smooth muscle cells (PASMCs) act as both oxygen sensors and effectors [2, 19, 48]. The early phase of HPV can be detected within 100 ms of hypoxia onset, peaks within 5–20 min, and is characterized by a rapid rise in pulmonary artery pressure-typically beginning within approximately 5 min, and reaching a plateau around 10 min [49]. Although the complete signaling cascade is not yet fully elucidated, the response is closely associated with an increase in the intracellular calcium concentration ([Ca2+]i) within PASMCs [6, 8, 12, 23, 44, 46, 55, 60]. Both mobilization of intracellular calcium stores and calcium influx pathways have been implicated, making dynamic Ca2+ signaling a key driver of hypoxia-induced constriction in the resistance pulmonary circulation. However, this mechanistic framework is largely derived from studies utilizing ex vivo preparations, including arterial ring myography [3, 56], buffer-perfused lung preparations where HPV is inferred from pulmonary artery pressure changes [15, 24, 45], precision-cut lung slices [4, 34], and cultured PASMCs [7, 33]. These widely employed models inevitably disrupt the native alveolar-vascular architecture, mechanical boundary conditions, and oxygen gradients that collectively govern pulmonary vascular tone in vivo. Consequently, they may alter vascular reactivity and limit the precise tracking of subtle diameter changes. While high-resolution X-ray imaging [39] and intravital microscopy via pleural windows [14, 51] have significantly improved visualization of the pulmonary vasculature, motion artifacts from respiration and cardiac pulsation, along with the need for stabilization strategies, continue to constrain the simultaneous, quantitative longitudinal tracking of vessel diameter and intracellular signaling. Crucially, these approaches remain inadequate for directly and synchronously quantifying real-time vasomotor responses and the underlying Ca2+ dynamics in intact, perfused, and spontaneously breathing lungs during acute hypoxic transitions. Consistent with these limitations, a technique capable of simultaneously measuring HPV-induced vascular reactivity and PASMC Ca2+ signaling dynamics in vivo is still lacking. Consequently, the coupling mechanism between local oxygen sensing and the rapid, spatially coordinated constriction of resistance pulmonary arteries within the living lung remains incompletely understood. To address this unmet need, this study developed an in vivo vacuum-stabilized lung window technique. This approach gently stabilizes the lung surface while preserving spontaneous respiration and physiological tissue compliance in mice. By integrating two-photon microscopy with intravascular fluorescent tracing, this technique enables prolonged, continuous intravital imaging. Utilizing the genetically encoded calcium indicator (GCaMP6f) and intravascular fluorescent tracers (FITC and Evans Blue) to delineate the vascular lumen and GCaMP6f to report PASMC Ca2⁺ activity, this method allows for the synchronous, high-resolution quantitative analysis of pulmonary artery diameter changes and intracellular Ca2+ dynamics during acute hypoxia. Employing this technique, we directly coupled calcium signaling to functional vasomotor responses in vivo to investigate whether dynamic Ca2+ responses correspond to diameter-dependent vascular reactivity. This facilitates the exploration of the spatiotemporal relationship between arterial constriction and PASMC Ca2+ signaling across different pulmonary artery sizes during HPV, thereby providing a valuable platform for mechanistic study and assessment of the HPV response. Methods Ethics approval and animal housing A total of 30 C57BL/6J male mice, aged 8–9 weeks, weighing 23–28 g, were obtained from Bainter Biotechnology Co., Ltd. (Wuhan, China) and housed under specific pathogen-free (SPF) conditions. Mice were group-housed (5 per cage) in 0.01 m3 cages under a 12-h light/dark cycle with ad libitum access to standard laboratory chow and water. Bedding consisted of autoclaved wood shavings, which was replaced twice weekly. This study was approved by the Institutional Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology (IACUC No. [2025] 4899) and conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Construction and expression of AAV5-SM22α-GCaMP6f The AAV5 vector carrying the GCaMP6f gene under the control of the smooth muscle-specific promoter SM22α was constructed at Vigene (Vigene Bio, Co., Ltd., China). Viral quality was validated prior to use. For in vivo expression, anesthetized mice underwent tracheal intubation followed by intratracheal instillation of 40 µL of AAV5-SM22α-GCaMP6f (1 × 1011 vg). Mice recovered on a 37 °C heating pad and were maintained under SPF conditions for two weeks to allow for stable transgene expression in pulmonary arterial smooth muscle cells [9, 38, 62]. Pre-treatment before in vivo imaging Following disinfection with 75% ethanol, mice were anesthetized via intraperitoneal injection of tribromoethanol (HY-B1372, 26,589, MCE, USA) at a dose of 250 mg/kg (2.5% solution). The trachea was exposed in the cervical region, a small inverted V-shaped incision was made, and an 18G tracheal cannula was inserted and secured to enable controlled gas delivery via a mechanical ventilator (PS5677, Kent Scientific, USA) during hypoxia experiments. The right jugular vein of the mouse was exposed and isolated. A PU-50 soft catheter (inner diameter 0.5 mm, outer diameter 0.7 mm) was inserted into the jugular vein to establish a venous route for drug administration. Vacuum-stabilized lung window and intravital two-photon imaging After removing the fur from the left thorax and axillary region, the mouse was positioned on a temperature-controlled stage within the main imaging unit of the two-photon microscopy platform. The skin and overlying muscle were dissected to expose the ribs. A small circular incision (≤ 5 mm in diameter) was made in the intercostal space between the third and fourth ribs to access the thoracic cavity. The vacuum-stabilized lung window, connected to a negative-pressure system, was then carefully placed over the incision. Gentle suction was applied to establish a stable imaging field while minimizing tissue damage and preserving physiological lung movement [35, 36, 61]. Specifically, the vacuum stabilization system was set to 970 mbar, generating a suction pressure of approximately 30 mmHg, a level that ensures a stable imaging field while minimizing tissue damage and preserving physiological lung motion [35, 36, 61]. Ventilatory parameters for anesthetized mice were as follows: respiratory rate of 120–130 breaths/min, positive end-expiratory pressure (PEEP) of 2 cm H2O, and peak inspiratory pressure (PIP) of 15 cm H2O. Physiological stability was continuously verified using a pulse oximeter (MouseSTAT, Kent Scientific, USA) attached to the hind paw, which recorded oxygen saturation (95–100%) and heart rate (300–600 bpm). Intravital lung imaging was performed using a two-photon microscopy platform [27, 61]. The system comprised a main imaging unit (IVM-M Series III, Revvity, USA), a tunable-wavelength two-photon laser (Chameleon Vision-S, Coherent, USA), a vacuum stabilization system (IVM-TMS, Revvity, USA), and a specialized lung imaging chamber (Revvity, USA). This platform enables ultra-high-speed imaging (up to 30 fps/second) with fully automated, high-precision motion compensation. Identification and diameter quantification of pulmonary arteries Pulmonary arteries were identified by flow direction and branching pattern: vessels with flow from larger to smaller caliber and diverging flow at branch points were classified as arteries, whereas vessels with converging flow patterns were classified as veins. Pulmonary artery diameter was measured manually at 20-s intervals using Fiji/ImageJ software. The inner boundary of the pulmonary artery wall, as delineated by intravascular fluorescent tracers (Evans Blue or FITC) and the genetically encoded calcium indicator GCaMP6f, was used to determine vascular diameter. Vasoreactivity assay during acute hypoxia Intravascular tracers were administered via tail vein injection to facilitate subsequent imaging: Fluorescein isothiocyanate (FITC) (150 kDa, BCCB8276, Sigma-Aldrich, Germany) (1%, 0.2 mL) [32, 52, 57] and Evans Blue (E2129, Sigma-Aldrich, Germany) (0.5%, 0.2 mL) [20, 41]. Heparin sodium (9041–08-1, Biofroxx, China) was injected (45 U/mL, 0.1 mL) to prevent coagulation. Following mechanical ventilation and the establishment of the vacuum-stabilized lung window, images were acquired using a 25 × water immersion objective lens (Nikon) to locate pulmonary arteries within the field of view and capture fluorescence signals. After a 60-s stabilization period under normoxia (21% O2), acute hypoxia was induced by switching the inspired gas to 10% O2; control mice were maintained at 21% O2. Video recording continued throughout the entire experimental protocol. Ca2+ imaging In GCaMP6f-expressing mice, a 25 × water-immersion objective was used to position the field of view over an area exhibiting strong fluorescence and well-defined vascular morphology for calcium signal acquisition. Following this, heparin (45 U/mL, 0.1 mL) was slowly administered via the jugular vein, followed by Evans Blue injection as an intravascular flow-direction marker, which does not spectrally interfere with the GCaMP6f signal. After a 90-s baseline recording under normoxia, hypoxia was induced by switching to 10% O2 for 12 min; control animals remained under room air ventilation. Video data and synchronized calcium signals were collected continuously throughout the recording period. A blinded study dasatinib (5 mg/kg) [43, 59] was administered by oral gavage 6 h before Ca2+ measurements in drug groups. Dasatinib is rapidly absorbed after oral administration, achieving peak plasma concentration (Cmax) between 0.5 and 6 h [10]. At a dose of 5 mg/kg, the effective concentration (EC90) is maintained for approximately 7 h, longer than the 6 h observed with the 2.5 mg/kg dose [31]. The clinically relevant doses of dasatinib (1 mg/kg/day) do not induce pulmonary hypertension, whereas doses exceeding the clinical range can influence HPV over a shorter time course [16]. Consequently, dasatinib at 5 mg/kg was administered via oral gavage to mice 6 h prior to hypoxic exposure. This timing ensured that plasma concentrations had already reached a plateau phase (within the Cmax window) at the onset of hypoxia, thereby enabling an effective assessment of dasatinib’s impact on HPV in a manner consistent with clinical dosing regimens and preclinical study paradigms. Ca2+ signal processing and ΔF/F calculation Raw fluorescence values from calcium-labeled pulmonary arteries were extracted using Fiji/ImageJ software (version 2.15.1, National Institutes of Health, USA) and processed in MATLAB (version R2017a, MathWorks, USA) with a low-pass filter (cut-off frequency 1 Hz). Calcium imaging experimental results were evaluated in terms of ΔF/F, ΔF/F = (F-F0)/F0 [42, 63]. For each selected region of interest, fluorescence values recorded during a 90-s normoxic baseline period were sorted in ascending order. The baseline fluorescence (F0) was defined as the average of the middle 50% of these sorted values. Calcium imaging results were then evaluated by calculating the normalized fluorescence change ΔF/F as (F-F0)/F0, where F represents the fluorescence intensity at each time point. Statistical analysis Data are presented as mean ± standard error of the mean (SEM). Statistical analyses were performed using SigmaPlot (version 14.0). Normality was assessed using the Brown-Forsythe test. For comparisons among multiple groups, data following a normal distribution were analyzed by two-way ANOVA if variance homogeneity was confirmed; post-hoc multiple comparisons were conducted using Tukey's HSD test. If variance homogeneity was not met, Welch's ANOVA was applied. For non-normally distributed data among multiple groups, a rank-based non-parametric two-way analysis (Scheirer-Ray-Hare test) was used. The threshold for statistical significance was set at P < 0.05. Results The vacuum-stabilized lung window constitutes a platform for pulmonary artery imaging in vivo We established an in vivo imaging system employing a vacuum-stabilized lung window (Fig. 1A). Through the use of intravascular fluorescent tracers and the vascular wall GCaMP6f signal, the pulmonary arterial lumen was visualized with high contrast, enabling its localization along the axis of blood flow and permitting quantitative analysis across a broad range of diameters. Specifically, FITC labeling detected pulmonary arteries measuring 13.95 μm to 62.49 μm in diameter (Fig. 1B); Evans Blue delineated vessels ranging from 18 μm to 80.16 μm (Fig. 1C); and the GCaMP6f signal identified arteries with diameters between 17.98 μm and 76.96 μm (Fig. 1D). Consequently, the vacuum-stabilized lung window enables the imaging of pulmonary arteries within a comprehensive range of 13.95 μm to 80.16 μm in diameter. The vacuum-stabilized lung window for simultaneous imaging of pulmonary artery and calcium signal in vivo. A Schematic overview of the vacuum-stabilized lung window imaging system, with a representative field of view illustrating the pulmonary arterial structures accessible via this window. B-D Representative images of pulmonary arteries visualized via intravascular fluorescent tracers (FITC, Evans Blue) (B and C) and the vascular wall GCaMP6f (D), demonstrating the imaging range of pulmonary artery diameters and Ca2+ signal achievable with this technique, with the direction of blood flow indicated by arrows. A representative GCaMP6f imaging field displaying anatomically registered regions of interest (ROIs) delineated along the vascular wall for spatially resolved analysis of Ca2⁺signals (D) To enable calcium signal monitoring within the same pulmonary artery, we imaged the GCaMP6f signal and delineated multiple regions of interest (ROIs) along the vascular wall, aligned with the direction of blood flow (Fig. 1D). This approach facilitated the extraction of spatially resolved Ca2+ signals both along the vessel and across distinct sites. By maintaining anatomical registration with the imaged pulmonary artery, this ROI-based strategy established an experimental framework for the simultaneous in vivo quantification of dynamic changes in mice PASMCs Ca2+ signaling and corresponding vascular reactivity as reflected by diameter alterations. Real-time in vivo capturing vascular reactivity Utilizing the vacuum-stabilized lung window in conjunction with intravascular fluorescent tracers (Evans Blue or FITC), we clearly visualized and differentiated pulmonary arteries of varying diameters (> 50 μm and < 50 μm) in live mice under normoxic conditions (Fig. 2A, left and middle). Continuous diameter tracking over a 30-min imaging period revealed minimal fluctuations in pulmonary artery caliber, irrespective of vessel size (< 50 μm or > 50 μm). These findings demonstrate the stable in vivo pulmonary artery imaging capability and reliable quantification potential of this technique (Fig. 2A, right). The continuous and dynamic monitoring of pulmonary artery response in vivo by vacuum-stabilized lung window. A Representative in vivo fluorescence images of mice pulmonary arteries with diameters both greater and less than 50 μm visualized through the vacuum-stabilized lung window under normoxic conditions, labeled with intravascular tracers (Evans Blue and FITC). B Representative time-lapse imaging of pulmonary arteries with diameters both greater (upper) and less than 50 μm (lower) under normoxia followed by acute hypoxia (10% O2) at 10 and 30 min, showing progressive vessel narrowing. C Quantitative analysis of acute hypoxia-induced changes in pulmonary artery diameter, stratified by vessel size (30–50 μm vs. 50–65 μm) at the 10- and 30-min time points. * P < 0.05, *** P < 0.001; normoxia group: n = 6 (3 mice for vessels > 50 μm, 3 mice for vessels < 50 μm); hypoxia group: n = 8 (4 mice for vessels > 50 μm, 4 mice for vessels < 50 μm). For the statistical analyses presented in Fig. 2C, the Brown-Forsythe test was first used to confirm normality and homogeneity of variances. For comparisons involving multiple groups, data following a normal distribution with confirmed homogeneity of variance were analyzed using two-way ANOVA, followed by Tukey’s HSD test for post-hoc multiple comparisons To further investigate whether the vacuum-stabilized lung window accurately captured HPV, mice were subjected to acute hypoxia (10% O2), and the pulmonary arteries of varying diameters (> 50 μm and < 50 μm) were imaged. The results revealed progressive narrowing at 10 min of hypoxia compared to the normoxic baseline (1 min), with a more pronounced reduction observed at 30 min (Fig. 2B). These findings demonstrated that acute hypoxia induces a time-dependent vasoconstriction, detectable at the single-vessel level. In pulmonary arteries stratified by diameter, acute hypoxia resulted in significant diameter reductions at both 10 and 30 min relative to normoxic controls, with a greater magnitude of constriction evident at the 30-min time point (Fig. 2C). These findings demonstrated that the vacuum-stabilized lung window enables accurate in vivo discrimination of pulmonary artery size categories, facilitates continuous and dynamic monitoring of diameter changes, and reliably detects acute hypoxia-induced vasoconstriction in a time- and vessel-size-dependent manner. Real-time in vivo monitoring of calcium signaling within pulmonary arteries To investigate whether the vacuum-stabilized lung window supports time-resolved measurements of calcium signaling during HPV, we performed in vivo imaging of pulmonary arteries in mice expressing the smooth muscle cell calcium indicator GCaMP6f. Calcium signals (ΔF/F) were continuously recorded from multiple ROIs bilaterally along the vascular wall and aligned with the direction of blood flow. Under normoxic conditions, ROI signals remained at baseline levels and showed no alteration in the temporal profile following the administration of dasatinib (Fig. 3A and C), indicating that dasatinib treatment alone does not interfere with calcium signaling. The multi-ROI quantification of pulmonary artery Ca²⁺ signaling2 dynamics in vivo by vacuum-stabilized lung window. The representative GCaMP6f image field showing multiple ROIs delineated along the pulmonary arterial wall with direction of blood flow indicated by arrow(s) (left), and the color-coded ΔF/F traces from individual ROIs and the corresponding average trace (red) demonstrating near-baseline Ca2⁺ activity under normoxic conditions (A) and acute hypoxia (10% O2, B), n = 3 for each group. And the effects of the Src kinase inhibitor, dasatinib, on Ca2+ signaling under normoxia (C) and hypoxia (D), n = 5 for each group Following a measurable but brief delay after hypoxia exposure (10% O2), Ca2+ levels within multiple ROIs distributed along the same vascular segment increased significantly. The amplitude and kinetics of the ΔF/F curves displayed heterogeneity across individual ROIs, with some exhibiting pronounced calcium oscillations. This observation indicates spatially heterogeneous activation of calcium signaling along the pulmonary artery. The average ΔF/F rose progressively during hypoxia, reflecting the rate of Ca2+ response at the whole-vessel level (Fig. 3B). Compared to acute hypoxia alone, dasatinib treatment resulted in a further enhanced Ca2+ signal following a prolonged latency. This was characterized by larger-amplitude calcium oscillations in responsive ROIs and more pronounced Ca2+ spikes in the average ΔF/F curve (Fig. 3D). Collectively, these data demonstrated that the vacuum-stabilized lung window is capable of capturing kinetic features of acute hypoxic signaling including the latency of Ca2+ response, Ca2+ spikes, and heterogeneity at the ROI level while also enabling the detection of pharmacological modulation of hypoxia-induced Ca2+ activation. Real-time in vivo simultaneous monitoring of pulmonary arterial reactivity and Ca2+ signaling dynamics To achieve real-time, simultaneous in vivo measurement of pulmonary arterial reactivity and calcium dynamics, we employed the vacuum-stabilized lung window to concurrently quantify vascular diameter and wall GCaMP6f ΔF/F in pulmonary arteries of varying sizes (stratified as > 50 μm and < 50 μm). Under normoxia in the absence and presence of dasatinib, pulmonary artery diameters remained constant, and no Ca2+ activation was detected (Fig. 4A and C). In contrast, acute hypoxia (10% O2) triggered Ca2+-constriction coupling in pulmonary arteries across all sizes, with elevated Ca2+ signal consistently preceding progressive vasoconstriction (Fig. 4B). Dasatinib treatment prolonged latency of the Ca2⁺ response and markedly enhanced Ca2⁺ activation in pulmonary arteries > 50 μm in diameter (Fig. 4D). Pulmonary artery size-stratified pulmonary artery Ca2⁺ dynamics and vascular reactivity using the vacuum-stabilized lung window. Paired time courses of simultaneous measurement of GCaMP6f fluorescence intensity (ΔF/F, black traces) and diameter (red traces) in pulmonary arteries of different sizes (> 50 μm and < 50 μm) under normoxia and acute hypoxia (10% O2) in the absence (A, B) or presence of dasatinib (C, D) The integrated analysis of calcium signal and size-resolved pulmonary artery diameter during HPV To quantify the dynamics of calcium signaling during HPV, we employed the vacuum-stabilized lung window and performed a vessel size-resolved analysis of hypoxia-induced Ca2+ kinetics by stratifying pulmonary arteries according to diameter (20–30, 30–40, 40–50, and 50–60 μm). Under hypoxic conditions, the latency of the Ca2⁺ response varied significantly across diameter groups, with arteries of 30–40 μm exhibiting the shortest latency, while vessels in the remaining diameter groups demonstrated delayed activation (Fig. 5A). This finding indicated vessel diameter-dependent heterogeneity in the initiation of hypoxic Ca2+ signaling. The dasatinib treatment prolonged latency in the 30–40 μm and 40–50 μm vessel groups, with a more pronounced effect observed in the latter (Fig. 5A). Furthermore, we quantified the rate of hypoxia-triggered Ca2+ signal change using Ca2⁺ spikes (ΔF/F). Under acute hypoxia without and with dasatinib treatment, no significant differences were detected in the rate of Ca2+ signal change across diameter groups (Fig. 5B). Pulmonary artery size-resolved Ca2⁺ latency and Ca2⁺ spike amplitude induced by hypoxia via the vacuum-stabilized lung window. Ca2⁺ latency (A) and Ca2⁺ spike amplitude (B) during acute hypoxia in the absence and presence of dasatinib, stratified by pulmonary artery diameter (20–30, 30–40, 40–50, 50–60 μm). ** P < 0.01, *** P < 0.001; n = 5 arteries or ROI for each group. The hypoxia and hypoxia + dasatinib groups each comprised 5 mice. One pulmonary artery was selected from each animal, and multiple regions of interest (ROIs) were sampled along its length. In the hypoxia group, the number of ROIs per artery was 8, 12, 15, 17, and 24, respectively, yielding a total of 76 ROIs. These ROIs were distributed across the following diameter brackets: 20–30 μm (5 ROIs), 30–40 μm (29 ROIs), 40–50 μm (23 ROIs), and 50–60 μm (19 ROIs). In the hypoxia + dasatinib group, the number of ROIs per artery was 4, 10, 12, 12, and 14, respectively, giving a total of 52 ROIs, with the following diameter distribution: 20–30 μm (13 ROIs), 30–40 μm (19 ROIs), 40–50 μm (14 ROIs), and 50–60 μm (6 ROIs). For the statistical analyses presented in Fig. 5, the Brown‑Forsythe test was first used to confirm normality and homogeneity of variances. Given that the assumptions of homogeneity of variance and normality were violated, a rank-based non-parametric two-way analysis (Scheirer-Ray-Hare test) was employed for comparisons involving multiple groups These data demonstrated that the vacuum-stabilized lung window enables an integrated analysis quantifying calcium signaling dynamics alongside vessel size-resolved pulmonary artery assessment during HPV, as well as its pharmacological modulation. Discussion Current methodologies for investigating HPV are diverse, each with distinct advantages and limitations. While ex vivo approaches utilizing PASMCs and isolated vascular rings can elucidate detailed hypoxic signaling pathways, they fail to capture the integrated morphological and functional properties of the intact pulmonary vasculature [54]. The perfused/ventilated ex vivo lung model, although capable of simulating global pulmonary vascular responses and assessing reactivity via changes in pulmonary artery pressure (ΔPAP), cannot discriminate the specific contributions of vessels of different diameters to HPV [22, 58]. Precision-cut lung slices preserve the lung's native cellular architecture and allow for observation of vessels across a range of calibers; however, inherent shortcomings including the absence of circulatory perfusion, declining viability over time, and a lack of standardized preparation and culture protocols compromise their reproducibility and widespread application [4, 34]. Commonly employed in vivo pulmonary vascular imaging techniques primarily consist of intravital microscopy (IVM) and 4D-CT X-ray imaging. Conventional IVM, which often employs a vinyl acetate membrane to create a lung window and restore negative pressure, is constrained by limited imaging stability and shallow penetration depth, currently permitting reliable observation only of murine pulmonary vessels smaller than 50 μm in diameter [14, 51]. Conversely, 4D-CT X-ray imaging is restricted by spatial resolution, capturing only vessels larger than 60 μm [39]. More critically, both in vivo imaging modalities struggle to visualize key molecular signals, such as calcium flux, during HPV. Consequently, the development of an in vivo methodology capable of simultaneous, high-resolution observation of pulmonary vascular responses across different diameters alongside detection of relevant molecular signals represents a critical imperative for advancing our understanding of HPV mechanisms. Given that HPV predominantly occurs in small resistance pulmonary arteries (< 100 μm) [2, 19, 48], existing in vivo imaging methods present limitations in both observable vascular range and molecular mechanistic resolution. To address this, the present study introduces a vacuum-stabilized lung window technique, an established in vivo imaging system previously validated for pulmonary immunology research [17, 28, 40]. In this work, the technique was successfully adapted to achieve stable imaging of pulmonary arteries ranging from 13.95 to 80.16 μm in diameter. This enabled, for the first time in vivo, the simultaneous recording of both vascular reactivity and dynamic changes in smooth muscle intracellular calcium signaling during HPV. Considering that alterations in pulmonary artery luminal diameter during HPV serve as a functional correlate of vascular resistance, this study employed a consistent and operational definition for pulmonary artery caliber measurement: internal diameter was measured at the luminal boundary defined by intravascular tracer (Evans Blue or FITC) and GCaMP6f-labeled smooth muscle signal. For calcium detection, we utilized GCaMP6f-a genetically encoded calcium indicator renowned for its high sensitivity and rapid kinetics, widely adopted in neuroscience [5, 26]. We achieved stable expression of GCaMP6f in PASMCs and confirmed its accurate reporting of hypoxia-induced calcium dynamics. Dasatinib is a broad-spectrum tyrosine kinase inhibitor that targets Src family kinases and PDGFR. It may modulate the temporal dynamics of Ca2+ signaling in pulmonary arterial smooth muscle cells (PASMCs) through multiple convergent mechanisms. First, PDGFR inhibition by dasatinib delays PLCγ-IP3-mediated Ca2+ release from the sarcoplasmic reticulum, thereby prolonging the latency of Ca2+ signal onset [11, 18, 64]. Second, dasatinib-mediated Src inhibition can alter the kinetics of Ca2+ influx and excitation–contraction coupling by affecting voltage-gated Ca2+ channels, store-operated Ca2+ entry (SOCE), and cytoskeletal architecture [1, 13]. Moreover, dasatinib-induced modulation of RhoA/ROCK signaling and actin cytoskeletal remodeling may influence mechanotransduction and Ca2+ sensitization, thereby further affecting the spatiotemporal propagation of Ca2+ signals [25, 29, 53]. Collectively, these effects converge to produce the observed prolongation of Ca2+ latency in PASMCs. The principal findings of this study are as follows: First, the vacuum-stabilized lung window technique enables stable intravital imaging of pulmonary arteries within the 13.95–80.16 μm range, allowing simultaneous quantification of vascular reactivity and spatiotemporally resolved PASMC calcium signals (Fig. 1). Second, acute hypoxia induces time-dependent vasoconstriction. Specifically, 30 min of acute hypoxia (10% O2) elicited an average constriction of 8% in 30–50 μm pulmonary arteries, consistent with previously reported values of 6%−8% [14, 51]. Third, acute hypoxia triggers marked spatial heterogeneity in PASMC calcium signaling; pharmacological intervention prolongs the latency period and enhances the amplitude of calcium oscillations (Fig. 3). Fourth, the latency of the hypoxia-induced calcium rise exhibits vessel diameter dependence, with the earliest responses observed in 30–40 μm arteries, and drug effects preferentially targeting vessels within the 30–50 μm range (Fig. 5). Fifth, simultaneous recordings confirm that the calcium rise consistently precedes vasoconstriction, supporting its role as an early initiating event in HPV, with drug treatment modulating calcium-contraction coupling in a vessel size-dependent manner (Fig. 4). This study has several limitations. First, it does not encompass the entire spectrum of intrapulmonary resistance vessels, leaving unexamined the vascular reactivity and PASMC calcium dynamics in pulmonary arteries larger than 80.16 μm. Second, the investigation is currently confined to the responses elicited by acute hypoxia (10% O2) within a 30-min window; the effects of prolonged hypoxia or varying degrees of hypoxic exposure remain unexplored. Furthermore, while GCaMP6f effectively reports changes in intracellular calcium concentration, it cannot discriminate the source of these calcium transients, such as release from internal stores versus influx from the extracellular space. Consequently, future investigations should aim to expand the sample size, establish chronic hypoxia models, and integrate optogenetic or pharmacological tools to further elucidate the intricate signaling network underlying HPV. Although the vacuum-stabilized lung window platform may be adaptable to chronic hypoxia models and potentially to larger pulmonary vessels in future studies, further work will be required to optimize long-term window stability, repeated imaging performance, and imaging depth/field coverage to support reliable application in chronic models and in larger vascular structures. In summary, the vacuum-stabilized lung window technique established in this study has, for the first time, enabled the in vivo, synchronous, and high-resolution monitoring of both pulmonary vascular reactivity and PASMC calcium signaling dynamics during HPV. This approach overcomes the multiple constraints inherent in traditional methodologies regarding vascular scope, signal detection, and systemic integration. This novel platform not only provides a new tool for elucidating the spatiotemporal regulatory mechanisms of HPV but also lays a methodological foundation for future drug screening and mechanistic research targeting pulmonary vascular diseases. Acknowledgements None. Abbreviations Intravital microscopy Hypoxic pulmonary vasoconstriction Pulmonary artery pressure Pulmonary artery smooth muscle cell Authors’ contributions QH originated the conception and conceived this study; QH, JL, JS, PL, XS, RX, and LZ designed the experiments; JL, JS, WY, and CY performed the experiments; JL, JS, PL, XS, WY, CY, RX, LZ, and QH executed the data analyses and interpretation. JL, JS, PL, XS, LZ, and QH wrote the manuscript, QH finalized the work. All authors approved the final version of the manuscript. Funding The study was supported by Noncommunicable Chronic Diseases-National Science and Technology Major Project-2024ZD0528600, grants from the National Natural Science Foundation of China (82130002, 82325001, 82241014, 82170068 and 82270060), and Shenzhen Medical Research Fund-C2404001. Data availability No datasets were generated or analysed during the current study. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Jie Liu and Junyan Shao contributed equally to this work. Contributor Information Pu Liao, Email: luoshen166@163.com. Xinyu Song, Email: songxinyu@ctgu.edu.cn. Liping Zhu, Email: 16654396@qq.com. Qinghua Hu, Email: qinghuaa@mails.tjmu.edu.cn. References Associated Data Data Availability Statement No datasets were generated or analysed during the current study.