Blockade of JAK2 protects mice against hypoxia‐induced pulmonary arterial hypertension by repressing pulmonary arterial smooth muscle cell proliferation

Abstract Objectives Hypoxia is an important risk factor for pulmonary arterial remodelling in pulmonary arterial hypertension (PAH), and the Janus kinase 2 (JAK2) is believed to be involved in this process. In the present report, we aimed to investigate the role of JAK2 in vascular smooth muscle cells during the course of PAH. Methods Smooth muscle cell (SMC)‐specific Jak2 deficient mice and their littermate controls were subjected to normobaric normoxic or hypoxic (10% O2) challenges for 28 days to monitor the development of PAH, respectively. To further elucidate the potential mechanisms whereby JAK2 influences pulmonary vascular remodelling, a selective JAK2 inhibitor was applied to pre‐treat human pulmonary arterial smooth muscle cells (HPASMCs) for 1 hour followed by 24‐hour hypoxic exposure. Results Mice with hypoxia‐induced PAH were characterized by the altered JAK2/STAT3 activity in pulmonary artery smooth muscle cells. Therefore, induction of Jak2 deficiency in SMCs protected mice from hypoxia‐induced increase of right ventricular systolic pressure (RVSP), right ventricular hypertrophy and pulmonary vascular remodelling. Particularly, loss of Jak2 significantly attenuated chronic hypoxia‐induced PASMC proliferation in the lungs. Similarly, blockade of JAK2 by its inhibitor, TG‐101348, suppressed hypoxia‐induced human PASMC proliferation. Upon hypoxia‐induced activation, JAK2 phosphorylated signal transducer and activator of transcription 3 (STAT3), which then bound to the CCNA2 promoter to transcribe cyclin A2 expression, thereby promoting PASMC proliferation. Conclusions Our studies support that JAK2 could be a culprit contributing to the pulmonary vascular remodelling, and therefore, it could be a viable target for prevention and treatment of PAH in clinical settings.


| INTRODUC TI ON
Pulmonary arterial hypertension (PAH) is a life-threatening disease manifested by the progressive pulmonary vascular remodelling, which results in persistently increased pulmonary arterial pressure, eventually culminating in right heart failure. 1,2 Sustained pulmonary vasoconstriction and vascular remodelling characterized by the concentric wall thickening and lumen obliteration of small-and medium-sized pulmonary arteries (PAs), are the major causes of elevated pulmonary vascular resistance and pulmonary arterial pressure in patients with PAH. 3 There is compelling evidence that pulmonary arterial smooth muscle cell (PASMC) hyperplasia is a cardinal feature of pulmonary vascular remodelling that underlies the development and progression of PAH. 4,5 A constitutive proliferative phenotype in PASMCs is considered to be a critical feature for patients with PAH or animals with induced PAH. 6,7 As a result, although current therapies against PAH can improve clinical symptoms, but disease progression is inevitable in most patients and mortality remains unacceptably high. 8,9 Therefore, new therapies aimed at ameliorating the irreversible pulmonary arterial remodelling are wanting in clinical settings.
As non-receptor tyrosine kinases, the mammalian Janus kinase (JAK) family comprises four evolutionarily conserved members, JAK1, JAK2, JAK3 and tyrosine kinase 2 (TYK2). Upon engagement of cytokines with cell surface receptors, JAKs undergo autophosphorylation at tyrosine residues, which generates docking sites to phosphorylate signal transducers and activators of transcription (STAT). Once STATs are phosphorylated at highly conserved tyrosine residues (termed p-STAT) by JAKs or other tyrosine kinases, they form dimers and then translocate into the nucleus, where they bind to the specific promoters to transcript the genes involved in cell proliferation, differentiation and apoptosis. 10,11 As an important member of the JAK family, Janus kinase 2 (JAK2) is involved in the regulation of various processes relevant to cell survival, proliferation, activation and differentiation. Particularly, JAK2 gain-of-function somatic mutations have been characterized in most patients with myeloproliferative neoplasms (MPNs) and acute lymphoblastic leukemia (ALL). [12][13][14] As a result, JAK inhibitors have been developed to treat various malignancies and have been shown to be efficacious against solid tumours, psoriasis and rheumatoid arthritis in both preclinical and clinical settings. [15][16][17] The functional importance of JAK2 in systemic arterial vascular smooth muscle cells (VSMCs) has also been noted, 18

but its role
in VSMCs during the course of pulmonary blood vessel remodel- ling is yet to be clarified. To address this question, we generated a SMC-specific Jak2 knockout model and demonstrated that Jak2 deficiency in SMCs protected mice from hypoxia-induced PAH and substantially reduced right ventricular systolic pressure (RVSP), the right ventricle/left ventricle plus septum [RV/ (LV+S)] weight ratio and the median width of pulmonary arterioles. Mechanistic studies revealed that blockade of JAK2 activity inhibited hypoxia-induced HPASMC proliferation by repressing the binding activity of STAT3 to the CCNA2 promoter, thereby attenuating pulmonary blood vessel remodelling. Our data support that strategies aimed at inhibiting JAK2 activity could be a viable treatment for PAH in clinical settings.

Conclusions:
Our studies support that JAK2 could be a culprit contributing to the pulmonary vascular remodelling, and therefore, it could be a viable target for prevention and treatment of PAH in clinical settings.

| Exposure to chronic hypoxia
Eight-week-old male mice were exposed to normobaric normoxia

| Cell culture
Human pulmonary arterial smooth muscle cells were purchased from the ScienCell and cultured with smooth muscle cell growth medium 2 (SMCGM 2, PromoCell) supplemented with 10% foetal bovine serum (FBS), 100 mg/mL penicillin and 100 IU/mL streptomycin at 37°C in a humidified atmosphere of 5% CO 2 in air. Cells at passages 4-9 were used for the experiments. For hypoxia (2% O 2 ) experiments, cells were firstly starved for 12 hours and then placed in a HERAcell vios 160i CO 2 incubator (Thermo Fisher), which was infused with a gas mixture containing 5% CO 2 and 93% N 2 for 24 hours. Normal incubators with 21% O 2 were used for normoxic culture. JAK2 phosphorylation was blocked in human PASMCs by adding 1 µmol/L TG-101348 (fedratinib MedChemExpress) 1 hour prior to hypoxic exposure. JAK2 phosphorylation was then assessed at the indicated time points.

| Hemodynamic measurements
After normoxic or hypoxic exposure, mice were anaesthetized with sodium pentobarbital (60 mg/kg), and hemodynamic measurements were performed. Measurement of RVSP and systemic arterial pressure was performed as described previously. 20 After exsanguination, the left lungs were fixed for histology in 4% neutral buffered formalin, and the right lungs were snap-frozen. The right ventricle (RV) was separated from the left ventricle plus septum (LV+S), and the RV/(LV+S) ratio was calculated as an index of RV hypertrophy. Lung vascular remodelling was assessed by measuring the degree of vessel muscularization as reported. 20

| Western blot analysis
Total protein was isolated from cultured cells using RIPA buffer (Beyotime), and the concentration was quantified using a BCA Protein Assay Kit (Boster). The proteins were subjected to Western blotting with the indicated primary antibodies using the established techniques. 21,22

| Quantitative RT-PCR analysis
Quantitative RT-PCR analysis was conducted using the SYBR Premix Ex Taq (TaKaRa) as previously described. 23 Briefly, total RNA was extracted from HPASMCs using a RNAiso plus kit (TaKaRa) according to the supplier's instructions. Real-time RT-PCR was conducted to assess cyclin A2, cyclin D1, cyclin E1, CDK2 and CDK4 expression using an ABI prisDK1m 7500 Sequence Detection System (Applied Biosystems). β-actin was used for normalization, and the relative expression levels for each target gene were calculated using the 2 −ΔΔCt approach as previously reported. 24 The primers used to amplify each target gene were as

| Histological and immunohistochemical analysis
The lungs were fixed by intratracheal infusion of 4% aqueous buffered formalin. Midsagittal slices from the right lungs were processed for paraffin embedding and sliced into 5-µm sections. The sections were next subjected to haematoxylin and eosin (HE) and elastic van gieson (EVG) staining using the established techniques. 25,26 For immunofluorescence analysis, fresh frozen sections (7 µm) of lung tissues and cells grown on coverslips in the chamber slides were prepared as previously described. 27,28 The slides were co-in-

| Chromatin immunoprecipitation (ChIP) assay and CCNA2 promoter reporter assay
ChIP assays were conducted with a ChIP assay kit (Beyotime) as previously described. 25 The primers used in the ChIP assay were as follows:

5′-CCG CCC CAG CCA GTT T-3′ and 5′CCC GCT CGC TCA CCC A-3′.
A dual luciferase reporter system (Promega) was used for the CCNA2 promoter luciferase reporter assays in which the STAT3 binding site within the CCNA2 promoter was disrupted using the established techniques. 30

| Statistical analysis
All experiments were conducted with at least three independent replications, and the data are presented as the mean ± SEM.
Statistical analysis was performed using the Graph Pad Prism (version 5.0) software (GraphPad Software Inc). Two experimental groups were compared using a Student's t test for paired data or a Student's t test with Welch's correction for unpaired data. Where more than two groups were compared, a one-way ANOVA with Bonferroni's correction was used. In all cases, P < .05 was considered statistically significant.

| Hypoxic insults activate the JAK2/STAT3 signalling in PASMCs
We first sought to examine the expression of JAK2 and STAT3 in the lungs of WT C57BL/6 mice with PAH induced by hypoxic challenge F I G U R E 1 Hypoxia-induced JAK2/STAT3 activation in a PAH mouse model and HPASMCs. Representative results for coimmunostaining of p-JAK2 and α-SMA (A), p-STAT3 and α-SMA (B) in lung sections from WT mice exposed to normoxia or hypoxia (n = 3 per group). Confocal immunofluorescence images for coimmunostaining of p-JAK2 and p-STAT3 in HPASMCs following normoxic or hypoxic exposure (C). All images were taken at an original magnification of ×400. The data are represented as the mean ± SEM. ***P < .001 (10% O 2 , 28 days). Notably, p-JAK2 and p-STAT3 were ubiquitously expressed in the lung sections of mice with PAH. However, this expression was increased mainly in the media layer of the pulmonary arterial region of the PAH lung tissues, as evidenced by the co-staining of p-JAK2 or p-STAT3 with α-SMA, a marker for PASMCs ( Figure 1A,B).
To confirm the above data, we exposed HPASMCs to normoxic or hypoxic conditions. In line with the above observations in animals, a 24 hours of hypoxic challenge (2% O 2 ) stimulated a marked increase of JAK2 phosphorylation at Tyr1007 (p-JAK2) along with an augmented phosphorylated STAT3 at Tyr705 (p-STAT3) in the nuclei of HPASMCs ( Figure 1C; 17.80 ± 1.72 vs 3.80 ± 0.58, n = 5 per group, P < .0001). Collectively, these data support that hypoxic insults activate JAK2/STAT3 signalling in PASMCs, which may contribute to the pathogenesis of pulmonary arterial remodelling during the course of PAH development.

| Generation of mice with SMC-specific Jak2 deficiency
To demonstrate the above assumption, the Jak2 flox/flox mice were bred with SM22-Cre ERT2 mice to generate Jak2 flox/flox SM22-Cre ERT2+ mice ( Figure 2A Immunohistochemical staining of serial lung sections further revealed the absence of Jak2 in PASMCs in Jak2-CKO mice ( Figure 2E).
Consistently, coimmunostaining demonstrated that the expression of Jak2 downstream target, p-STAT3, in PASMCs, was substantially lower in Jak2-CKO mice than that in Jak2-C mice ( Figure 2F), indicating that tamoxifen efficiently induced Jak2 deficiency.

| Jak2 deficiency alleviates pulmonary pressure and right ventricular function in a hypoxiainduced mouse model of PAH
Next, Jak2-CKO mice and Jak2-C mice following last tamoxifen injection were subjected to induction of PAH under normobaric normoxic or hypoxic conditions (10% O 2 ) for 28 days, respectively.
The right ventricular (RV) systolic pressure (RVSP) and RV/(L+S) ratio (the ratio of the right ventricular mass to the sum of the left ventricular and septal masses), a marker of hypertrophy in the right ventricle due to elevated right ventricular pressure and afterload, were employed to assess the development of pulmonary hypertension. After 28 days of hypoxic exposure, the mean RVSP of the F I G U R E 2 Generation of SMC-specific Jak2-knockout mice. Schematic diagram of transgenic mice used to generate Jak2-CKO and Jak2-C mice (A). PCR analysis of tail genomic DNA to determine the presence of the floxed null allele (B). Western blot analysis to confirm Jak2 depletion in pulmonary arteries (C; n = 3 per group) and cardiac muscle (D; n = 3 per group). Immunohistochemistry analysis to confirm Jak2 depletion in PASMCs (E; n = 3 per group). Coimmunostaining results of p-STAT3 and α-SMA in lung sections from Jak2-C and Jak2-CKO mice (F). All images were taken at an original magnification of ×400. The data are represented as the mean ± SEM. ***P < .001 hypoxia group was significantly increased compared with that of the normoxia group in both Jak2-C and Jak2-CKO mice. However, the mean RVSP was significantly lower in Jak2-CKO mice than that in Jak2-C mice ( Figure 3A; 30.80 ± 1.91 vs 21.80 ± 0.97, n = 10 per group, P = .003). In addition, the severity of RV hypertrophy was much lower in Jak2-CKO mice than that in Jak2-C mice after nor-

| Jak2 deficiency attenuates pulmonary blood vessel remodelling and PASMC hyperplasia following chronic hypoxia
Histological studies were next conducted to demonstrate the effect of Jak2 on pulmonary blood vessel remodelling during the course of PAH development. Remarkably, staining with haematoxylin and eosin (HE) demonstrated that loss of Jak2 significantly attenu-

| JAK2 inhibition suppresses hypoxia-induced PASMC proliferation
In general, there is a PASMC phenotype switch from a differentiated state to a proliferative state during the course of PAH development. 31 To further confirm the above data observed in animals, we employed TG-101348 (also named Fedratinib, thereafter referred to as TG), a selective JAK2 inhibitor, 32 to pre-treat HPASMCs for 1 hour, followed by hypoxic insult for 24 hour.
Next, three complementary approaches were employed to assess the impact of JAK2 on hypoxia-induced HPASMC proliferation.  in the lungs of Jak2-C and Jak2-CKO mice after normoxic (n = 8 per group) or hypoxic (n = 10 per group) exposure for 28 days. Ten vessels were analysed per mouse. Coimmunostaining results of α-SMA and Ki67 in lung sections from Jak2-C and Jak2-CKO mice after hypoxic (D; n = 10 per group) exposure for 28 days. All images were taken at an original magnification of ×400. The data are represented as the mean ± SEM. *P < .05; **P < .01. HE, haematoxylin and eosin; EVG, elastic van gieson; PA, pulmonary artery

| JAK2/STAT3 regulates HPASMC proliferation by enhancing cyclin A2 expression
To dissect the mechanisms by which JAK2/STAT3 signalling promotes PASMC proliferation, we first conducted real-time PCR analysis of cyclin A2, cyclin D1, cyclin E1, cyclin-dependent kinase 2 (CDK2) and CDK4 expression in PASMCs, as those molecules are considered to be cell cycle checkpoint regulators for the G0/ G1 and S phases. 33 It was interestingly noted that the expression of cyclin A2, a critical regulator necessary for the initiation of cell cycle and S/G2 phase transition, 34  1.98 ± 0.04 vs 2.10 ± 0.08, n = 4 per group, P = .248; and CDK4: 3.86 ± 0.37 vs 3.44 ± 0.23, n = 4 per group, P = .366). Given that cyclin D1 has been suggested to be a target for STAT3 transcription, 35 we thus embarked on cyclin A2. Indeed, Western blot analysis confirmed a 6-fold reduction of cyclin A2 levels in TGtreated cells (Figure 6F; 0.35 ± 0.04 vs 0.05 ± 0.00, n = 4 per group, P = .0002), suggesting that cyclin A2 could be a novel target downstream of JAK2/STAT3 to regulate HPASMCs proliferation following hypoxic insults.
To address the above question, in silico analysis was conducted and identified 9 potential STAT3-binding sites in the CCNA2 (encoding cyclin A2) promoter ( Figure S1). Chromatin immunoprecipitation (ChIP) was next conducted, and the resulting products were used to amplify the CCNA2 promoter regions flanking the above-indicated 9 potential p-STAT3 binding sites.
Indeed, p-STAT3 manifested binding activity to the CCNA2 promoter in a region 214 bp upstream of the transcriptional start site ( Figure 6G). CCNA2 promoter reporter assays were subsequently employed to confirm that STAT3 transcribes CCNA2 expression.
A mutated CCNA2 promoter reporter (MU) was constructed in which the STAT3-binding motif (ACGCTGGGCAG) was mutated to AGCGGCCCGGC ( Figure 6H, left panel). As expected, disruption of the STAT3-binding site resulted in a 1-fold reduction of hypoxia-induced luciferase reporter activity in HPASMCs ( Figure 6H; right panel, 0.86 ± 0.07 vs 0.44 ± 0.06, n = 4 independent replications, P = .004). These findings prompted us to check cyclin A2 expression in animals induced with PAH. Consistently, Jak2-CKO mice following hypoxic induction manifested remarkably lower cyclin A2 expression in PASMCs ( Figure 6I). Taken together, our data support that JAK2 activates STAT3, thereby transcribing cyclin A2 expression in PASMC to enhance pulmonary blood vessel remodelling during the course of PAH development.

| D ISCUSS I ON
Although a great deal of effort has been recently devoted to dissecting the pathoaetiologies underlying PAH, the exact molecular mechanisms, however, remain poorly understood. The lack of this related information has significantly hampered the development of novel and effective therapies against this devastating disorder.
There is emerging evidence that JAK/STAT signalling is activated during PAH initiation and progression. Studies in a monocrotaline (MCT)-induced PAH model revealed that monocrotaline pyrrole (MCTP), the main metabolite of MCT, targets the pulmonary arteries to induce transforming growth factor β (TGF-β) expression via activating JAK/STAT signalling. 36,37 TGF-β was also noted to induce interleukin 6 (IL-6) production through the JAK/STAT signalling cascade. 38 Together, these data support that JAK/STAT signalling is likely implicated in the pathogenesis of PAH.
In the present study, we investigated the impact of JAK2 on pul- To address the above question, we challenged the mice with chronic hypoxic insult, and then examined JAK2/STAT3 signalling in the lung sections. Remarkably, the enhanced p-JAK2 and p-STAT3 expression was predominantly limited to PASMCs, as evidenced by the α-SMA co-immunostaining. Similar results were also obtained from studies in HPASMCs. Based on the above results, we hypothesized that depletion of Jak2 in smooth muscle cells may confer protection for mice against hypoxia-induced PAH. To prove this assumption, we generated inducible smooth muscle cell specific Jak2 knockout mice, and Jak2 deficiency was induced by tamoxifen as described at 8 weeks of age, followed by hypoxic insult for 28 days.
Indeed, Jak2-CKO mice displayed significantly reduced thickness of the medial layer of small PAs along with attenuated pulmonary vascular remodelling. Specifically, loss of Jak2 in smooth muscle cells F I G U R E 6 JAK2/STAT3 promoted HPASMC proliferation programming by enhancing cyclin A2 expression following hypoxic exposure. Real-time PCR to determine cyclin A2 (A), cyclin D1 (B), cyclin E1 (C), CDK2 (D) and CDK4 (E) expression in HPASMCs. Western blot analysis of cyclin A2 expression in HPASMCs (F). ChIP-PCR results for analysis of p-STAT3 binding activity to the CNNA2 promoter (G). Results for CNNA2 promoter luciferase reporter assays in HPASMCs (H). Coimmunostaining results of α-SMA and cyclin A2 in lung sections from Jak2-C and Jak2-CKO mice after hypoxic (I; n = 10 per group) exposure for 28 days. All images were taken at an original magnification of ×400. The data are represented as the mean ± SEM. *P < .05; **P < .01; ***P < .001. CDK2, cyclin-dependent kinase 2; CDK4, cyclin-dependent kinase 4 repressed hypoxia-induced PASMC proliferation as manifested by the much less number of α-SMA + Ki67 + cells.
To date, many studies have principally focused on the effect of aberrant JAK/STAT signalling on abnormal cancer cell proliferation. In head and neck squamous cell carcinoma, STAT3 phosphorylation is a consequence of increased IL-6 production by tumour cells. 39 Increased expression of the G-CSF receptor is observed in high-grade ovarian epithelial tumours, and cell culture experiments suggest that G-CSF contributes to JAK/STAT activation in this disease. 40 Given that hypoxia stimulates the proliferation of PASMCs, thereby promoting pulmonary blood vessel remodelling, 41 we thus utilized TG101043 (TG), a selective JAK2 inhibitor, to elucidate the effect of JAK2 on PASMC proliferation following hypoxic insult. Excitingly, blockade of JAK2 efficiently suppressed hypoxia-induced HPASMC proliferation, suggesting that blockade of JAK2 could be a viable approach to prevent pulmonary vascular remodelling in clinical settings.
To further address the mechanisms by which JAK2 regulates PASMC proliferation following hypoxic exposure, we first examined the effects of JAK2 inhibition on the cell cycle in HPASMCs after hypoxic challenge. Notably, compared to control cells, TG treatment significantly decreased the number of HPASMCs in G2/M phase by inducing G0/G1 and S phase arrest following hypoxic insult. This result prompted us to examine key regulators relevant to G0/G1 to S phase and S phase to G2/M phase transition after hypoxic challenge.
Cyclin A2 was noted to be significantly downregulated following TG treatment. In general, cyclin A2 is not only required for G0/G1 to S phase transition, but also essential for DNA synthesis in S phase, 42 and therefore, aberrant cyclin A2 expression would arrest the cells in G0/G1 and S phase. 43 In fact, cyclin A2 has been noted to be involved in astrocyte proliferation in a gliosis animal model 44 and trophoblasts in recurrent miscarriage. 45 We thus assumed that cyclin A2 is a downstream target of JAK2/STAT3 signalling. Indeed, ChIP assays demonstrated that p-STAT3 selectively binds to the CCNA2 promoter at a position 214 bp upstream of the transcriptional start site, thereby transcribing cyclin A2 expression to promote PASMC proliferation following hypoxic insult. In line with these findings, JAK2-CKO mice exhibited a substantial decrease in terms of cyclin A2 expression in PASMCs following chronic hypoxic exposure.
Beyond cyclin A2, TG treatment also induced a slight decrease of cyclin D1 expression in HPASMCs following hypoxic exposure, while cyclin D1 has been considered to be a potential STAT3 target to mediates platelet-derived growth factor (PDGF)-induced proliferation in human airway smooth muscle cells. 46,47 In conclusion, we demonstrated evidence that altered of JAK2 activity in PASMCs is a critical manifestation relevant pulmonary vascular remodelling during the course of PAH development.
Therefore, mice with SMC-specific Jak2 deficiency are protected from hypoxia-induced pulmonary vascular remodelling and RV hypertrophy. Upon hypoxia-induced activation, JAK2 phosphorylates STAT3, which transcribes cyclin A2 expression by binding to the CCNA2 promoter, thereby promoting PASMC proliferation.
Collectively, our data suggest that JAK2 could be a viable target to prevent blood vessel remodelling in clinical settings.

ACK N OWLED G EM ENT
This study was supported by the National Natural Science

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

E TH I C A L A PPROVA L
All studies were conducted in accordance with NIH guidelines and approved by the Animal Care and Use Committee (ACUC) of Tongji Hospital. Additionally, all animals were handled with care and euthanized humanely during the study.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.