Pulmonary Macrophages Attenuate Hypoxic Pulmonary Vasoconstriction via β3AR/iNOS Pathway in Rats Exposed to Chronic Intermittent Hypoxia

Chronic intermittent hypoxia (IH) induces activation of the sympathoadrenal system, which plays a pivotal role in attenuating hypoxic pulmonary vasoconstriction (HPV) via central β1-adrenergic receptors (AR) (brain) and peripheral β2AR (pulmonary arteries). Prolonged hypercatecholemia has been shown to upregulate β3AR. However, the relationship between IH and β3AR in the modification of HPV is unknown. It has been observed that chronic stimulation of β3AR upregulates inducible nitric oxide synthase (iNOS) in cardiomyocytes and that IH exposure causes expression of iNOS in RAW264.7 macrophages. iNOS has been shown to have the ability to dilate pulmonary vessels. Hence, we hypothesized that chronic IH activates β3AR/iNOS signaling in pulmonary macrophages, leading to the promotion of NO secretion and attenuated HPV. Sprague-Dawley rats were exposed to IH (3-min periods of 4–21% O2) for 8 h/d for 6 weeks. The urinary catecholamine concentrations of IH rats were high compared with those of controls, indicating activation of the sympathoadrenal system following chronic IH. Interestingly, chronic IH induced the migration of circulating monocytes into the lungs and the predominant increase in the number of pro-inflammatory pulmonary macrophages. In these macrophages, both β3AR and iNOS were upregulated and stimulation of the β3AR/iNOS pathway in vitro caused them to promote NO secretion. Furthermore, in vivo synchrotron radiation microangiography showed that HPV was significantly attenuated in IH rats and the attenuated HPV was fully restored by blockade of β3AR/iNOS pathway or depletion of pulmonary macrophages. These results suggest that circulating monocyte-derived pulmonary macrophages attenuate HPV via activation of β3AR/iNOS signaling in chronic IH.


Introduction
Intermittent hypoxia (IH) during sleep periods is a distinctive feature in the patients of sleep apnea syndrome (SAS) [1][2][3]. IH exposure to healthy humans and animals causes prolonged activation of the sympathoadrenal system and elevation of daytime blood pressure [1,[4][5][6][7][8]. Therefore, IH accompanying sympathoadrenal activation has been implicated in the pathogenesis of systemic hypertension caused by SAS [9]. However, the effect of an increase in sympathoadrenergic activity on pulmonary vascular tone is not fully elucidated.
Hypoxic pulmonary vasoconstriction (HPV) is an important mechanism for optimizing ventilation/perfusion matching [10] and also inducing pulmonary hypertension [2]. In a previous study, we have reported that the centrally-mediated increase in sympathetic nervous activity following IH acts to blunt HPV via β 1 -adrenergic receptors (β 1 AR) in the brain [11]. In addition, we have also reported that IH-derived activation of β 2 AR, not β 1 AR, in the pulmonary arteries attenuates the HPV [12]. These results demonstrate that IH-derived sympathoadrenal activation attenuates HPV via β 1 AR and β 2 AR. However, the role of β 3 AR in modifying HPV is unknown.
In in vivo and in vitro animal experiments, it has been demonstrated that diseases associated with prolonged increase in catecholamine levels result in β 3 AR upregulation in cardiomyocytes [13][14][15][16]. Furthermore, chronic stimulation of β 3 AR has been shown to induce inducible nitric oxide synthase (iNOS) overexpression and NO secretion in the mouse heart [17]. It is interesting to note that IH exposure per se increased iNOS expression in RAW264.7 macrophages in vitro [18]. Moreover, iNOS/NO signaling has the ability to dilate pulmonary vessels during septic shock [19,20]. These studies suggest a possibility that iNOS in pulmonary macrophages is upregulated by chronic IH directly and/or chronic stimulation of β 3 AR in response to the IHinduced sympathoadrenal activation to release NO and reduce pulmonary vascular tone. To the best of our knowledge, however, there has been no report concerning activation of β 3 AR and iNOS/NO signaling in pulmonary macrophages following chronic IH.
In this study, we hypothesized that 1) β 3 AR in the pulmonary macrophages may be upregulated by IH associated with sympathoadrenal activation, 2) iNOS expression in the pulmonary macrophages may also be increased following stimulation of the upregulated β 3 AR as well as directly by IH per se, and 3) the β 3 AR/iNOS signaling in the pulmonary macrophages may be activated to release substantial NO in response to acute hypoxic exposure and thus, may modify HPV. To verify these hypotheses, we performed the following experiments using rats treated with chronic IH (IH rats) and rats exposed to normoxia (N rats). We first evaluated the expression of β 3 AR and iNOS in the pulmonary macrophages using immunohistochemical and electrochemical techniques (i.e. Western blot). Second, we examined whether stimulation of β 3 AR on the macrophages promotes iNOS-mediated NO secretion in vitro using bronchioalveolar lavage (BALF)-derived macrophages. Third, using synchrotron radiation microangiography for visualizing the pulmonary microvessels in vivo, we revealed the functional contribution of β 3 AR/iNOS signaling in the pulmonary macrophages to modulation of HPV.

Materials and Methods Animals
Experiments were conducted on 7 wk old male Sprague-Dawley rats. All rats were on a 12: 12-h light-dark cycle at 25°C and were provided with food and water ad libitum. Rats were divided into two groups. One was housed in normoxic conditions (N rats). The other was continuously housed in an airtight Plexiglas chamber (27 x 44 x 19 cm, model KYN-370, Bioresearch Center, Tokyo, Japan) with IH exposure for 6 weeks, except for a 10 minute interval every fifth day when chamber was cleaned (IH rats) [21]. IH treatment consisted of alternating 90 seconds cycles of normoxia (21% O 2 ) and hypoxia (reaching 4% O 2 at the nadir). N 2 was delivered to the chamber at a rate of 14 L/min (PSA type N 2 generator, ECONOX Ver. 2.10, ECOTS, Osaka, Japan). Compressed air was delivered at a rate of 54 L/min (Oil free scroll type, Smart Air SLP-15EBD, ANEST IWATA, Yokohama, Japan). The gas flushing the chamber was automatically switched from compressed air to N 2 and back to compressed air with the use of a timed solenoid valve. The O 2 concentration in the chamber was monitored by an O 2 analyzer (Oxygen monitor JKO-25 Ver. 3, JIKCO, Tokyo, Japan). Exposure was performed 8 hours/day (9:00 AM-5:00 PM) for 42 consecutive days. All experiments were approved by the Institutional Animal Care and Use Committee of the University of Tokyo.

Sampling of 24-hour Urine Catecholamine
For 24-hour urine collection, N rats and IH rats were housed in metabolic cages (Natsume, Tokyo, Japan) at the completion of the 6-week IH exposure. The collected urines were consecutively drawn into tubes containing 20 μL of 2.5 mol/L HCl. After urine collection, concentration of catecholamine was measured using high-performance liquid chromatography method (SRL Japan, Tokyo, Japan).

Intratracheal Administration of Liposomal Clodronate
Liposomal clodronate was administered into the trachea as previously described [22]. It was reported previously that the most effective depletion of macrophage was observed after 72 h of clodronate intratracheal administration (i.t.) [22,23]. After 6 weeks exposure of IH, a single dose of liposomal clodronate (Clodrosome, Encapsula NanoSciences, Nashville, TN) was administered to N and IH rats via the i.t. route 3 days before synchrotron radiation (SR) microangiography. Rats were sedated with pentobarbital (70 mg/kg) intraperitoneally (i.p.). Using standard aseptic procedures, the tracheas were exposed by surgical resection and pierced with a 26-gauge needle for i.t. injection of 500 μg of clodronate contained in 100 μL saline. The neck wound was closed with sterile sutures. To confirm that the i.t. injection of clodronate depletes pulmonary macrophages, the lungs were isolated immediately after clodronate treatment in another group of N and IH rats. In this case, the reduction in number of macrophages was evaluated with immunofluorescent staining using anti-ED-1 antibody (see below).

Chronic Repeated Intravenous Administration of Fluorescent Liposomes
Fluorescent liposomes (Fluoroliposome, Encapsula NanoSciences) were injected via the caudal vein on -1 day before IH exposure, and injection was performed every week during the 6 weeks of IH experiment. Each time 0.4 mL of fluorescent liposome contained in 1 mL saline was injected. Lung and liver tissues were frozen in O.C.T., and sliced into 10-μm sections with a cryostat. Images of the unfixed cryosections were captured with fluorescence microscopy BIOREVO BZ-9000 (Keyence, Osaka, Japan). The liver sections were used as a positive control for circulating monocytes-derived macrophages which engulf fluorescent liposomes.

Synchrotron Radiation Microangiography
SR pulmonary microangiography was performed as described previously [24]. Each rat was anesthetized with pentobarbital sodium (70 mg/kg, i.p.) and analgesic agent butorphanol tartrate (0.5 mg/kg, i.p.). Supplementary doses of pentobarbital (~15 mg/kg/hr i.p.) and butorphanol tartrate (0.025 mg/kg/hr i.p.) were periodically administered to maintain a surgical level of anesthesia during microangiography procedure. We used 5-min exposure to 10% O 2 to induce HPV. In this study, 4 types of protocol were performed. 1) To assess the β 3 AR mediated modification of HPV, an angiogram during hypoxic exposure was recorded following the baseline angiogram with room air. These angiograms were repeated after administration of SR59230A (lipophilic selective β 3 -blocker, 7.5 mg/kg, i.v., Sigma-Aldrich) in N and IH rats. 2) To assess β 3 AR/NOS signaling-mediated pulmonary vasodilation, the angiograms were recorded before and after acute administration of CL316243 (lipophilic selective β 3 -agonist, 100 μg/kg, i.v., Tocris Bioscience, Ellisville, MO, USA) with/without L-NAME (non-selective NOS blocker, 50 mg/kg, i.v., Sigma-Aldrich) or L-NIL (selective iNOS blocker, 3 mg/kg, i.v., Cayman Chemical, Ann Arbor, MI). All angiograms in this protocol were taken under ganglion blockade with hexamethonium bromide (autonomic ganglionic blocker, 25 mg/kg, i.v., Wako, Osaka, Japan) to exclude the secondary effect of the β 3 -agonist via central nervous system. 3) To assess the modification of HPV by iNOS, the angiogram during acute hypoxic exposure was recorded following the baseline angiogram with room air, and these angiograms were repeated after an administration of L-NIL in N and IH rats. 4) To assess the modification of HPV by IH-derived accumulated pulmonary macrophage, the angiograms with room air and hypoxic exposure were recorded in N and IH rats following i.t. administration of liposomal clodronate 3 days before angiography.

Image Analysis of SR Microangiograms
Image analysis was performed using Image Pro-Plus ver. 4.1 (Media Cybernetics, Silver Spring, MD) as described previously [24]. The line-profile function of Image Pro-Plus was used as an accurate method for measuring the internal diameter (ID) of individual vessels. A 50 μm thick tungsten wire appeared in all recorded images and was used as a reference for calculating vessel ID. Vessels were categorized according to ID; 100-200 μm, 200-300 μm, 300-500 μm and 500μm. The magnitude of HPV was calculated as % change in diameter relative to baseline.

Immunohistochemistry of Lung Sections
Paraffin blocked lungs from N and IH rats were sliced into 5-μm sections and placed onto clean glass slides. After deparaffinization, the slides soaked in the citrate buffer were heated with microwave for 5 min for antigen retrieval. Then, the slides were incubated in methanol with 3% hydrogen peroxide for 10 min to block endogenous peroxidase activity. Nonspecific protein binding was blocked by treatment with normal bovine serum albumin for 30 min. The sections were incubated overnight with anti-β 3 AR antibody (Santa Cruz Biotechnology, California, CA) at 4°C. The slides were then washed 3 times with PBS and treated with secondary antibodies for 30 min at room temperature. After washing 3 times, the slides were exposed to an ABC horseradish peroxidase (HRP) reagent (Vector Laboratories, Burlingame, CA) in PBS for 30 min. The GFP signal was developed with Peroxidase Substrate Kit AEC (Vector Laboratories), and finally the slides were mounted with water soluble mounting medium. The stained sections were visualized with an Eclipse E400 microscope (Nikon, Tokyo, Japan) attached to a high-resolution digital camera DXM 1200F (Nikon). Images were captured with ACT-1 software (Nikon).

Quantitative Analysis of Pulmonary Arteries
Quantitative image analysis of immunohistochemical stained sections with anti-β 3 AR antibody was performed with Image Pro-Plus ver. 4.1 software as described previously [2,25]. The red stain was selected semi-automatically. Optical density and area of the red stain were obtained. Quantification of the expression level of the protein was estimated as expression level score (ELS): ELS = (mean optical density of positively stained area-mean optical density of background area) x percent area of positively stained.

Immunofluorescence Microscopy of Lung Sections
After deparaffinization, the lung sections were soaked in the citrate buffer and heated with microwave for 5 min. Blocking was performed with bovine serum albumin for 30 min. The sections were exposed to primary antibody overnight followed by appropriate secondary antibody for 60 min. The staining was imaged with fluorescence microscopy BIOREVO BZ-9000 (Keyence). Using primary antibodies were anti-ED1 (CD68) antibody (AbD Serotec, Oxford, UK) and anti-β 3 AR antibody (Santa Cruz Biotechnology).

Bronchioalveolar Lavage
After 6 weeks of IH exposure, N and IH rats were sacrificed by single i.p. injection of pentobarbital and subsequent exsanguination via the abdominal aorta. The trachea was cannulated and bronchoalveolar lavage (BAL) was performed in situ by infusing the lungs with 5 mL aliquots of PBS. The BAL fluid (BALF) was drained passively by gravity and the procedure was repeated four times, giving a total BALF volume of 20 mL.

Positive Control for Pro-inflammatory Macrophage
Pulmonary macrophages obtained from LPS administered rats were used as positive controls for pro-inflammatory macrophages. Rats were sedated by inhalation of 3% isoflurane. BALF was obtained for gathering pulmonary macrophages 24 hours after i.p. administration of LPS (10 mg/kg), after sacrificing the rats by bleeding from the abdominal aorta.

Immunocytochemistry of Pulmonary Macrophages
BALF obtained from LPS-administered rat, N rats, and IH rats was centrifuged at 500 g for 10 min (Kubota 1720, Tokyo, Japan). Two mL of saline was added to the precipitate of BALF and mixed softly. One hundred μL of each solution was dropped onto the slides and dried overnight. The slides were fixed with cold 50% acetone in methanol for 10 min and washed three times with PBS for 5 min. They were then incubated with 1% Triton X-100 for 10 min. After being washed thrice with PBS for 5 min, the slides were blocked with bovine serum albumin for 30 min and incubated with primary antibodies overnight at 4°C. The slides were washed three times with PBS for 5 min. After incubation with fluorescent conjugated secondary antibodies, images were captured with a fluorescence microscopy BIOREVO BZ-9000 (Keyence). The primary antibodies used were anti-ED1(CD68) antibody (AbD Serotec), anti-iNOS antibody (Thermo Fisher Scientific, Waltham, MA), anti-eNOS antibody (Enzo Life Sciences, Farmingdale, NY), anti-nNOS antibody (Enzo Life Sciences), anti-CD11c antibody (AbD Serotec), anti-IL-6 antibody (R&D Systems, Minneapolis, MN), and anti-β 3 AR antibody (Santa Cruz Biotechnology).

Positive Cell Counting in Immunofluorescent Images
For assessment of macrophage infiltration and β 3 AR upregulation, ED1 and β 3 AR positive cell counting was performed in the immunofluorescent images of lung sections. Ten representative images (200 x) were chosen from the left lobe of each animal. The number of stain positive cells was counted automatically by Image Pro Plus ver. 4.1 (Media Cybernetics).

Nitrite Measurement of Macrophage Cultures
BALF obtained from N and IH rats were centrifuged at 500 g for 10 min (Kubota 1720, Tokyo, Japan). The pellet was resuspended in phenol red free RPMI 1640 medium (Gibco Laboratories, Grand Island, NY) with 1% streptomycin at 1.4 x 10 5 cells / mL. The cells were plated at 4.2 x 10 5 macrophages per well in polystyrene tissue culture plates and allowed to adhere for 12 h at 37°C in an atmosphere of 5% CO 2 / 95% O 2 . Then, 100 μM of CL316243 (Tocris Bioscience), 100 μM of isoproterenol (LKT Laboratories, Minneapolis, MN), and 100 μM of CL316243 + 50 μM of L-NIL (Cayman Chemical) were administered. After incubation for 30 h at 37°C, the media were ultrafiltered at 7000 g x 20 min with ultracentrifugal filter units for 10 kD molecules (EMD Millipore, Billerica, MA). NO levels in the cell-free supernatant were determined by analysis of its relative stable metabolite nitrite using the Griess reaction with a fluorometric NO 2 /NO 3 Assay Kit-FX (Dojindo Laboratories, Tokyo, Japan).

Western blot analyses
Frozen lung tissue 0.1 g was homogenized with 1 mL of ice-cold RIPA buffer containing 0.1% SDS, 0.5% DOC, 1% NP-40, 150 mM NaCl, 50 mM Tris-Cl pH 7.4, 50 mM NaF, 1 mM Na 3 VO 4 and Complete Protease Inhibitor Cocktail (Roche Diagnostics, Mannheim, Germany). To remove debris, the homogenate was centrifuged at 1500 g for 5 min, and supernatant was used for analysis, and the rest was frozen at -80°C. For direct detection of protein expression in macrophage, BALF obtained from N and IH-rats was centrifuged at 500 g for 5 min and 1 mL of RIPA buffer was added to the pellet. The macrophage suspension was sonicated sufficiently and centrifuged at 1000 g for 5 min to remove debris. The homogenized samples of lung and macrophage were heated at 95°C for 5 min, and 3 x Laemmli buffer containing 9% mercaptoethanol added. The supernatant was used for analysis and the rest was stored at -80°C. The protein concentration of homogenates was determined by the Bradford assay. The homogenate was subjected to SDS-PAGE on a 4-20% gradient precast gel, and separated protein was then transferred to polyvinylidene fluoride (PVDF) membranes using a transfer system Trans Blot Turbo (BioRad, Tokyo, Japan). Nonspecific antibody binding was blocked using 3% skim milk in TBS-T 0.1%, and the membranes were incubated with primary antibodies. The signals were detected by a luminescent image analyzer Image Quant Las 4000 mini (GE Healthcare, Waukesha, WI) using a secondary antibody coupled to horseradish peroxidase (Promega, Madison, WI). Primary antibodies used were anti-β 3 AR (Santa Cruz Biotechnology), anti-iNOS antibody (Abcam), anti-eNOS antibody (Enzo Life Sciences, Farmingdale), anti-nNOS antibody (Enzo Life Sciences), anti-IL-6 antibody (R&D Systems), and anti-TNFα antibody (Abcam).

RT-PCR analysis
Lung tissue samples were homogenized and used for RNA isolation using ISOGEN (Nippon Gene, Tokyo, Japan). The purified RNA was then reverse transcribed using TaqMan Reverse Transcription Reagents (Applied Biosystems Japan, Tokyo, Japan). Expression levels of mRNA of β 3 AR were assayed quantitatively by real-time RT-PCR using TaqMan Gene Expression Assays (Applied Biosystems Japan). Quantitative mRNA expression data were acquired and analyzed by 7000 Sequence Detection System (Applied Biosystems Japan).

Statistical analysis
All statistical analyses were conducted using GraphPad Prism6 (GraphPad Software, San Diego, CA). The results of relative expression of β 3 AR mRNA and % change in internal diameter in SR angiograms are presented as mean ± standard error of the mean (S.E.M.) and the data analysis was performed using Student's t-test (unpaired) or two-way ANOVA with Sidak's multiple comparison tests. All other data are presented as mean ± standard deviation (S.D.) and the data analyses were performed using Student's t-test (unpaired) or one-way ANOVA with Tukey's multiple comparison tests. A P value of < 0.05 was predetermined as the level of significance for all statistical analysis.

IH induces macrophage accumulation and upregulated β 3 AR expression in the lungs
At first, to confirm the activation of the sympathoadrenal system in IH, the rats' urinary concentrations of dopamine, adrenaline, and noradrenaline were measured. Urine was collected from N and IH rats over 24 hours in normoxic conditions on the day after the last day of IH exposure. The concentrations of these catecholamines were significantly higher in IH rats than in N rats (S1 Fig).
Immunofluorescent staining of pulmonary tissue performed after 6 weeks of IH exposure demonstrated that the number of pulmonary macrophages was significantly increased and the positive ratio of β 3 AR-expressing cells was high (Fig 1A-1D). The number of macrophages in the alveolar space was increased (S2 Fig). Immunocytochemistry showed that β 3 AR was strongly expressed in BALF-derived alveolar macrophages from IH rats (Fig 1E). In IH rats, macrophages accumulated around the small pulmonary arteries and these perivascular macrophages also expressed β 3 AR (Fig 1A, S3 Fig). Western blotting and RT-PCR showed that the β 3 AR was expressed in both the lung tissue and BALF-derived alveolar macrophages of N rats (Fig 1F-1H), and immunohistochemistry demonstrated that the β 3 AR was expressed on the endothelium of the pulmonary arteries (S4A Fig). IH significantly increased the protein and mRNA expression levels of β 3 AR in the lung tissues (Fig 1F and 1G). The β 3 AR protein expression was also elevated in the BALF-derived alveolar macrophages (Fig 1H). In contrast, IH decreased the expression level of β 3 AR in the pulmonary arterial endothelium of vessels with diameters ranging from 50 to 150 μm (S4B Fig). These results indicate that β 3 AR expression was upregulated in macrophages but not in the pulmonary arteries in IH rats.
To identify the origin of accumulated macrophages in the lungs of IH, intravenous administration of fluorescent liposomes was performed during IH experiments. The results of this study demonstrate that the increase in the number of pulmonary macrophages induced by IH stems from the migration of circulating monocytes into the lungs (S5A Fig). As a positive control, the liver was used for observation of fluorescent liposome engulfed monocytes. Interestingly, IH-induced accumulation of macrophages was also observed in the liver (S5B Fig).

Pulmonary macrophages in IH rats expressed pro-inflammatory markers including iNOS
To characterize the phenotype of pulmonary macrophages in the lungs of IH, immunocytochemical staining and western blotting were performed using iNOS, CD11c, and IL-6. LPS administered rats were used for a positive control of inflammatory macrophages (S6 Fig). Proinflammatory markers such as iNOS, CD11c, and IL-6 were detected in IH rat macrophages, but not those of N rats (Fig 2A). Western blotting demonstrated that the protein expression levels of pro-inflammatory markers; i.e., iNOS, IL-6, and TNFα were significantly upregulated in IH-induced macrophages (Fig 2B). These results indicated that the IH stimulation promoted differentiation of the pulmonary macrophages into a pro-inflammatory type.

Pulmonary macrophages release NO via the β 3 AR/iNOS pathway in IH rats
To assess the NO synthesis ability of pulmonary macrophages, BALF-derived macrophages were used for in vitro experiments. In groups without drug administration, the total amount of the macrophage-derived nitrite (chemically stable metabolite of NO) was not different between N and IH rats. In the pulmonary macrophages obtained from IH rats, but not N rats, the administration of the β 3 -agonist CL316243 enhanced the secretion of nitrite, which is indicative of elevated NO synthesis and release (Fig 3). The increase in nitrite synthesis induced by CL316243 was prevented by the simultaneous administration of the iNOS blocker L-NIL. In contrast, the non-selective β 1 and β 2 -agonist isoproterenol decreased nitrite synthesis in both N and IH rats. These results suggest that NO secretion was facilitated in the IH-derived proinflammatory macrophages by the activation of β 3 AR/iNOS signaling, but not by β 1 or β 2 AR activation.

HPV is markedly attenuated in IH rats
The degree of HPV was estimated using synchrotron radiation microangiography. In N rats, acute hypoxic exposure (10% O 2 ) induced marked constriction (HPV) in the small pulmonary arteries with internal diameters (ID) of 100-500 μm, but not in those with ID of more than 500 μm (Fig 4A and 4B). The extent of the HPV (% reduction in ID caused by acute hypoxia) tended to increase as arterial diameter decreased, with the greatest degree of constriction (approximately 24%) occurring in the arteries with ID of between 200 and 300 μm. In IH rats, acute hypoxic exposure only induced significant HPV in the pulmonary vessels with ID of 200-300 μm, and the degree of HPV was approximately half of that seen in N rats (Fig 4B), indicating that the HPV induced by acute hypoxic exposure was greatly attenuated in IH rats.

β 3 AR activation causes iNOS-dependent vasodilation and attenuates HPV in IH
In IH rats, SR59230A, a lipophilic selective β 3 -blocker, restored the attenuated HPV to almost the same level as was seen in N rats (Fig 4A and 4C). In contrast, SR59230A had no significant effect on the HPV seen in N rats. Pretreatment with the selective iNOS inhibitor L-NIL also  restored the attenuated HPV in the same manner as SR59230A (Fig 5). These results suggest that IH activates both a β 3 AR-mediated and iNOS-mediated vasodilatory mechanism to attenuate HPV.
To assess whether peripheral β 3 AR, but not the β 3 AR in the central nervous system, contribute to HPV attenuation in IH, CL316243 was administered under ganglionic blockade with hexamethonium bromide. In IH rats, CL316243 induced extensive dilatation of the small pulmonary arteries, particularly in the arteries with ID of 100-500 μm (Fig 6A and 6B). In contrast, CL316243 had no significant effect in N rats. These results indicate that peripheral β 3 AR were activated to a much greater extent in IH rats than in N rats, and strongly suggest that peripheral β 3 AR contributed to the HPV attenuation observed in IH rats.
Next, we investigated whether NOS is involved in the β 3 AR-mediated pulmonary vasodilation observed in IH rats. Pretreatment with L-NAME or L-NIL completely abrogated the CL316243-mediated pulmonary vasodilation in IH rats (Fig 6A, 6C and 6D). There was no significant difference in the strength of the inhibitory effect between L-NAME and L-NIL. These results suggest that iNOS-derived NO was chiefly responsible for the β 3 AR-induced pulmonary vasodilation observed in IH rats. Taken together, these observations strongly suggest that the β 3 AR/iNOS pathway is activated in IH, resulting in the dilation of pulmonary arteries and the attenuation of HPV.

Acute intra-tracheal administration of clodronate restores HPV in IH rats
To investigate whether the macrophages that accumulated in the lungs of the IH rats contributed to the observed HPV attenuation, HPV was evaluated in pulmonary macrophage depleted rats using microangiography. Clodronate was administered into the lungs of IH rats via the trachea immediately after the 6-week IH exposure period. Clodronate strongly suppressed the number of pulmonary macrophages in IH rats to the control level ( Fig 7A) and enhanced the attenuated HPV to almost the same level as that seen in N rats (Fig 7B and 7C). These results indicate that the intra-alveolar macrophages that accumulate in the lungs in IH contribute to the attenuation of HPV.

Discussion
The present study demonstrated that 1) chronic IH increases pro-inflammatory macrophages with upregulation of β 3 AR and iNOS in the lungs, 2) IH-derived activation of β 3 AR/iNOS signaling promotes NO secretion from pulmonary macrophages, and 3) that the pulmonary macrophages attenuate HPV via the β 3 AR/iNOS signaling pathway in IH rats.
We first confirmed that activation of sympathoadrenal system persists even after chronic IH exposure. Because the 24-hour collection of urine was performed while the rats were maintained in a normoxic atmosphere on the day after the last day of IH exposure, the results of urinary concentrations indicate that the sympathoadrenal system is persistently activated in IH rats even after their release from chronic IH. These results are consistent with the findings of previous reports [8,11,26].
The present angiographic data showed that in N rats, HPV occurs in the small pulmonary arteries of 100-500 μm ID with a maximum constriction in the 200-300 μm range. This IDdependent vasoconstriction pattern is consistent with our previous reports [11,12,27]. HPV was significantly attenuated in the arteries with ID ranging between 100-500 μm and significant HPV was observed only in the arteries of 200-300 μm ID following IH-induced sympathoadrenal activation. This result is also consistent with our previous report [11,12]. The present study has revealed that this attenuation of HPV is abrogated by β 3 AR blockade. Further, after chemical blocking of the sympathetic nerve with a ganglion blocker, stimulation of β 3 AR induced significant dilation in the small pulmonary arteries in IH rats, but not in N rats. Collectively, these results suggest that in IH, peripheral β 3 AR which are distributed within the lungs are activated by increased sympathoadrenal activity, causing the decrease in pulmonary vascular tone and attenuation of HPV. In N rats, β 3 AR was observed in the endothelium of the small pulmonary arteries; however, β 3 -agonist had no significant vasodilatory effect on these vessels. This is consistent with the findings of previous studies in which β 3 AR had no [28,29] or only weak vasodilatory capacity in normoxic pulmonary vessels [30,31]. In addition, we showed that the β 3 AR expression in the small pulmonary arteries is decreased in IH rats. Collectively, pulmonary vascular β 3 AR are likely to play a minimal role in controlling vascular tone in IH. In contrast, we showed that the β3AR expression on the alveolar and the perivascular macrophages is significantly elevated in IH rats. Moreover, the depletion of intra-alveolar macrophages restored the normal level of HPV in IH rats. These results suggest that the β3AR expressed on the more abundant 'alveolar' macrophages contribute to attenuation of HPV in IH rats. In the present study, it was not elucidated whether the 'perivascular' macrophages contributed to attenuation of HPV in the same manner as alveolar macrophages. The high expression of the β3AR in the perivascular macrophages implicates their contribution to HPV modification, although future research is essential to resolve this question. Several subsets of macrophages with distinct functions have been described. M1 macrophages promote inflammation to defend the host from a variety of foreign bodies. M2 macrophages have anti-inflammatory functions and regulate wound healing [32]. Previously, accumulation of macrophages in the lungs [33] and around the pulmonary arteries [34] has been reported in chronic hypoxic experiments. Frid et al. showed that monocyte-derived macrophage infiltration/accumulation in the adventitia of pulmonary arteries was observed after 4 weeks of chronic hypobaric hypoxia (380 mmHg) in rats [34]. Vergadi et al. showed that accumulation of macrophages in the lungs was observed from 2 days of chronic hypoxia in the mouse, and a significant increase in the number of macrophages maintained throughout the 2 weeks of a hypoxic experiment period [33]. Importantly, these macrophages accumulating with chronic hypoxia-induction were M2 type. In our experiments, accumulation of macrophages in the lungs was observed after 6 weeks of IH exposure. The alveolar macrophages expressed pro-inflammatory proteins such as iNOS, CD11c, and IL-6, suggesting that these macrophages were M1 type. However, to elucidate the phenotype of these macrophages, more detail evaluation including characterization of their gene expression profile is needed. Taken together, the present data indicate that the IH-derived stimulation of β 3 AR promoted differentiation of the pulmonary macrophages into a pro-inflammatory type with expression of iNOS. Thus, IH and chronic hypoxia have different effects on the polarization of macrophages. Moreover, chronic intravenous administration of fluorescent liposomes demonstrated that IH induced the migration of circulating monocytes into the lungs. However, the detailed mechanisms responsible for migration of the circulating monocytes are not revealed in the present study. Since the migration ability of circulating monocytes into the local peripheral tissues depends on the expression level of chemokine/chemoreceptors [35], further study is needed to elucidate the migration mechanism of monocytes into the lungs due to chronic IH exposure.
In the present in vitro study, we demonstrated that β 3 AR stimulation increased NO secretion via iNOS activation in IH pulmonary macrophages, but not in N macrophages (Fig 3). In cardiomyocytes, all subtypes of NOS; i.e., eNOS, nNOS, and iNOS, were reported to be downstream of β 3 AR signaling [17,[36][37][38][39]. Most of these studies detected significant correlations between β 3 AR and eNOS expression. To the best of our knowledge, there has been only one report about the relationship between β 3 AR and iNOS [17]. In the present study, eNOS was detected in the alveolar macrophages (S7A Fig); however, the protein expression level of eNOS did not differ between IH and N rats (S7B Fig). eNOS expressed on unactivated macrophages is known to secrete a small amount of NO continuously [40]. Therefore, in the present in vitro study, the baseline level of nitrite secretion from the macrophages in the absence of drug treatment probably originated from eNOS in both N and IH rats. On the other hand, iNOS expressed in IH-derived macrophages is suggested as the most dominant physiologically relevant source of NO from the macrophages [41]. Additionally, this previous report revealed that macrophage-derived NO directly dilates femoral arterial rings in ex vivo experiments [42]. Therefore, considering these findings and our results for angiography (Figs 4-7), it is strongly indicated that pro-inflammatory macrophages attenuate the HPV via β 3 AR/iNOS pathwayderived NO secretion.
In contrast to the results of selective β 3 AR stimulation, isoproterenol inhibited NO secretion in both N and IH macrophages. This result is consistent with a previous report in which catecholamines inhibited the macrophage-mediated production of NO through β 1 and β 2 AR in vitro [43,44]. Recently, it has been revealed that phosphorylation sites for protein kinase A and βAR kinase are found in the β 1 and β 2 AR, whereas the β 3 AR lacks these sites. Thus, the β 1 and β 2 AR undergo functional desensitization after long-term exposure to hypercatecholemia. In contrast, sustained stimulation of the β 3 AR does not modify its functional effects [45][46][47]. In the present study, function of the β 1 and β 2 AR on the pulmonary macrophages was not disrupted after 6 weeks of IH exposure, however, there is a possibility that much longer exposure of IH than 6 weeks decreases the inhibitory effects of NO production from the macrophages. Accordingly, we suggest that β 3 AR signaling probably plays a pivotal role in controlling the pulmonary circulation in the pathogenesis accompanying prolonged sympathoadrenergic activation such as chronic IH.
In our previous study, the β 2 AR dependent activation of PI3kinase/Akt/eNOS signaling in the pulmonary arteries attenuated the HPV, leading to the prevention of the progression of pulmonary arterial hypertension (PAH) [12]. Interestingly, blockade of β 2 AR and β 3 AR exacerbated HPV in IH rats to the same degree as that in control rats. These findings suggest that both β 2 AR and β 3 AR have critical contribution to attenuate HPV in IH. Taking these findings into consideration, the β 3 AR/iNOS pathway in pro-inflammatory macrophages presumably behave in the same manner as the β 2 AR/eNOS pathway in pulmonary arteries to prevent PAH progression in IH. To confirm this hypothesis, further research is required.
In SAS patients, PaCO 2 is increased due to obstruction of the upper airway during sleeping periods [9]. The effect of PaCO 2 on the pulmonary vascular function in SAS has not been elucidated, however it has been reported that supplementation of CO 2 significantly attenuates HPV in rat [48,49]. Therefore, there is a possibility that hypercapnia attenuates IH-induced HPV. To elucidate an effect of blood CO 2 on HPV, additional experiments utilising simultaneous exposure of intermittent hypoxia and hypercapnia are needed. The present study is important with respect to focus on the effect of intermittent hypoxia to pulmonary hemodynamics, which is one of the major factors in the pathophysiology in SAS.
In summary, we demonstrate that pro-inflammatory pulmonary macrophages attenuate HPV via the activation of β 3 AR/iNOS signaling in IH rats. The relationship between IHinduced sympathoadrenal activation and pulmonary circulation is not fully elucidated, although, this study highlights the pivotal role of sympathoadrenal activation and pro-inflammatory macrophages in attenuating the HPV in IH. Moreover, we also highlight the importance of β 3 AR/iNOS signaling pathway in the preservation of the pulmonary circulation under prolonged IH exposure.
Supporting Information S1 Fig. The mean urinary catecholamine concentrations of the IH rats were extremely high compared with those of the control rats. The concentrations of dopamine, adrenaline, and noradrenaline in 24-hour urine samples from N and IH rats (n = 5 each). Urine was collected over 24 hours using metabolic cages under a normoxic atmosphere on the day after the end of a 6-week period of normoxic or intermittent hypoxic exposure. The data are presented as mean ± S.D. values. Ã Significant difference between the N and IH rats ( ÃÃ P<0.01).