Opsin 3 and 4 mediate light-induced pulmonary vasorelaxation that is potentiated by G protein-coupled receptor kinase 2 inhibition

Cell culture.

Human PASMCs (hPASMCs; PromoCell, Heidelberg, Germany) were maintained in smooth muscle cell growth medium 2 (PromoCell) in a humidified incubator at 37°C and 5% CO₂ and were used for experiments between passages 3 and 7. Cells were passaged every 5–7 days using 0.25% trypsin (GIBCO, Waltham, MA).

Animals.

Adult male Sprague-Dawley and Wistar rats (4–6 mo old; Harlan Laboratories, Indianapolis, IN) and adult male (10–14 wk old) C57BL6/J (The Jackson Laboratory, Bar Harbor, ME) and Opn4 knockout (Opn4^−/−) mice (38) were used. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Johns Hopkins University.

Rat PASMC isolation.

Primary cultures of rat PASMCs (rPASMCs) were obtained as previously described (22). Following anesthesia with sodium pentobarbital (130 mg/kg ip), the heart and lungs were removed en bloc. Intrapulmonary arteries (200–600 µm od) were carefully dissected from the lung in ice-cold HEPES-buffered saline solution (HBSS) containing (in mM): 130.0 NaCl, 5.0 KCl, 1.2 MgCl₂, 10.0 HEPES, 1.5 CaCl₂, and 10.0 glucose with pH adjusted to 7.2 with 5.0 M NaOH. Extraneous connective tissue was removed, and the arteries were opened and the lumen gently scraped with a cotton-tipped swab to remove endothelial cells. The arteries were allowed to recover for 30 min in cold (4°C) HBSS, followed by a 20-min recovery in reduced-Ca²⁺ HBSS (20 μM CaCl₂) at room temperature. The tissue was enzymatically digested for 20–25 min at 37°C in reduced-Ca²⁺ HBSS containing collagenase (type I; 1,750.0 U/ml), papain (9.5 U/ml), bovine serum albumin (2.0 mg/ml), and dithiothreitol (1.0 mM). After digestion, single smooth muscle cells were dispersed by gentle trituration in Ca²⁺-free HBSS. rPASMCs were either used fresh (single-cell contraction studies) or cultured for 24–48 h [membrane potential measurements, diphosphorylated myosin light chain 2 (ppMLC) measures, proximity ligation assays (PLAs)] in smooth muscle basal media (SMBM; Lonza, Walkersville, MD) supplemented with 0.3% fetal calf serum and 1% penicillin/streptomycin.

Membrane potential measurements.

Cells on 25-mm glass coverslips were placed in a heated cell chamber containing HBSS. Cells were incubated under dark conditions with 500 nM DiBAC₄ (3) (Invitrogen, Carlsbad, CA) for 15 min to allow for cell uptake of the dye. In some instances, GRK2-specific inhibitor (1 μM; methyl 5-[2-(5-nitro-2-furyl)vinyl]-2-furoate; Calbiochem, San Diego, CA) was added during the incubation time. Measurement of DiBAC₄ (3) fluorescence was performed in a workstation based on a Nikon TES 100 Ellipse inverted microscope with epifluorescence attachments. The collimated light beam from a xenon arc lamp was filtered by an interference filter at 490 nm and focused onto the PASMCs under examination via a ×20 fluorescence objective (Super Fluor 20, Nikon). Light emitted from the cell (F₄₉₀) was returned through the objective and detected by a CCD imaging camera. An electronic shutter was used to minimize photobleaching of dye and provide a dark environment. Protocols were executed and data collected online with InCyte software (Intracellular Imaging, Cincinnati, OH). Baseline F₄₉₀ was measured for 1 min (5 images/min) before light was turned off (shuttered) for 15 min during addition of phenylephrine (PE) (10 μM) or KCl (80 mM). After 15 min, two images were obtained (maximum depolarization), before the cells were exposed to red or blue light for 3 min using LED diodes positioned directly above the cell chamber. The diodes were then turned off and images captured for 3 min. PE was made fresh as a stock solution (10 mM in diH₂O) on the day of experiments, and KCl was made as a stock solution (4 M in diH₂O) and stored at room temperature. Both drugs were diluted to final concentrations in the cell chamber. Data were collected from 10–30 cells and averaged to obtain a single value per experiment. Each experiment was conducted on cells from a different animal.

Sugen hypoxia rats.

Male Wistar rats were injected with SU5416 (20 mg/kg sc), placed in a hypoxic chamber for 3 wk, and then returned to normoxia for 2 wk. We have shown this protocol induces severe PH in Wistar rats (22). SU5416 was suspended in a mixture of dimethyl sulfoxide and carboxymethyl cellulose [0.5% (wt/vol) carboxymethyl cellulose sodium, 0.9% (wt/vol) sodium chloride, 0.4% (vol/vol) polysorbate 80, and 0.9% (vol/vol) benzyl alcohol in deionized water]. The hypoxic chamber was continuously flushed with a mixture of room air and N₂ (10.0 ± 0.5% O₂) to maintain low CO₂ concentrations (<0.5%). Chamber O₂ concentration was continuously monitored (ProOx 110 oxygen analyzer; BioSpherix, Parish, NY). The rats were exposed to room air for 10 min twice a week to clean the cages and replenish food and water supplies. Normoxic controls were injected subcutaneously with SU5416 vehicle and kept in room air next to the hypoxic chamber for 5 wk. All animals were exposed to the same light/dark cycle and ambient temperatures. After 5 wk, rats were anesthetized with pentobarbital sodium (43 mg/kg ip) and euthanized.

Cell traction microscopy.

The contractile stress arising at the interface between each adherent cell and its substrate was measured with traction microscopy as described elsewhere (1, 9). For these studies, we used collagen-coated elastic gel blocks (Young’s modulus ~8 kPa with Poisson’s ratio 0.48) that were embedded with either red fluorescent (F-8810) or yellow-green fluorescent (F-8811) microspheres (0.2-μm in diameter; Molecular Probes, Eugene, OR). Freshly isolated rPASMCs were plated sparsely and allowed to adhere and spread for 24 h. Fluorescent microbeads embedded near the gel apical surface were excited for 1 min with N2.1 (515- to 560-nm wavelength) and I2 (450- to 490-nm wavelength) filter systems to visualize and image red and yellow-green microspheres, respectively. Images of microspheres were taken at different times (before and after a 30-min treatment with GRK2 inhibitor, or before and after treatment with PE), and the fluorescent images of the same region of the gel after trypsin were used as the reference (traction-free) image. The displacement field between a pair of images was then obtained by identifying the coordinates of the peak of the cross-correlation function (1, 9). From the displacement field, we computed traction field using constrained and unconstrained Fourier transform traction cytometry (FTTC) as described (1, 9).

Vasoreactivity studies.

Vasorelaxation of PAs was measured using force-tension myography in organ bath chambers (Danish Myo Technology, Aarhus N, Denmark) as previously described (38). Rat and mouse PAs (2nd to 3rd branch) were isolated and trimmed to 2 mm in length. Vascular rings were cleaned in ice-cold Krebs-Ringer bicarbonate solution containing the following (in mM): 118.3 NaCl, 4.7 KCl, 1.6 CaCl₂, 1.2 KH₂PO₄, 25.0 NaHCO₃, 1.2 MgSO₄, and 11.1 dextrose. One end of the vascular rings was connected to a transducer and the other to a micromanipulator. The experiment was performed in a bath filled with constantly oxygenated Krebs buffer at 37°C. PAs were dissected under a microscope, ensuring no damage to the smooth muscle or endothelium. After mounting the chambers on to the system, all lights were switched off, and the rest of the experiment was carried out in darkness with minimal red light turned on only for vision when required. The chambers were also covered with aluminum foil to minimize the penetration of light from the monitor. The PAs were passively stretched to an optimal resting tension using the micromanipulator, after which 60 mM KCl was administered, and repeated after a wash with Krebs buffer. After these washes, the vessels were allowed to equilibrate for 1 h. PE (1 μM) was used for preconstriction. Vasorelaxation to light was examined after a stable baseline was achieved, and light doses were delivered via cold light lamp (Opelco20500/06; Schott-Fostec, Auburn, NY) (40,000–200,000 lux), light-emitting diodes [red (620–750 nm), green (495–570 nm), or blue (380–495 nm)], and 455-nm wavelength-specific blue light (Thorlabs, Newton, NJ). Blue light LED exposures were performed at ~20 lux. Sodium nitroprusside (SNP)- or ACh-induced dose response relaxations (1 nM to 10 μM) were carried out as positive controls for smooth muscle cells and endothelium activity, respectively. Responses in the presence of inhibitors were performed after preincubating the vessels with GRK2 inhibitor (1–10 μM) and/or the specified inhibitors for 30 min before constriction was induced. Relaxation responses were calculated as a percentage of tension after preconstriction.

Immunofluorescence staining and imaging.

Samples were processed as described previously (2, 3). Briefly, samples were rinsed with PBS, fixed with 3.7% formaldehyde (Fisher Chemical, Fairlawn, NJ) for 15 min, and permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, St. Louis, MO). Samples were then washed three times with PBS, blocked with 1% BSA (Sigma-Aldrich) for at least 1 h, and incubated overnight with the indicated primary antibodies: rabbit Opn3 (Abcam, Cambridge, MA; ab75285), rabbit Opn4 (Thermo Fisher Scientific, Waltham, MA; PA1-781), mouse GRK2 (Sigma-Aldrich; G7670), and mouse phosphoserine (Sigma-Aldrich; P5747). Samples were rinsed with PBS three times before conjugation for 1 h with the following secondary antibodies: Alexa Fluor 488 Phalloidin (Invitrogen), DyLight 550 anti-rabbit, and DyLight 647 anti-mouse (Thermo Fisher Scientific). Samples were counterstained with DAPI before mounting and imaged using a Nikon eclipse 80i epifluorescence scope (Nikon, Melville, NY).

PLA.

Samples were fixed, permeabilized, and incubated with primary antibodies as described above before being processed with the Duolink in situ PLA kit (Sigma-Aldrich) according to the manufacturer’s instructions. Briefly, samples were incubated for 60 min at 37°C with PLA plus and minus probes (1:5 dilution). Samples were then washed twice and incubated for 30 min at 37°C with ligase (1:40) diluted in the kit’s ligation solution (1:5). Samples were washed twice before the amplification step, in which 1:80 polymerase in 1:5 amplification stock was added to the sample followed by incubation for 100 min at 37°C. Last, samples were rinsed and mounted in the provided mounting medium with DAPI before being sealed and imaged.

PCR and qRT-PCR.

Total mRNA was extracted from isolated rat PAs or cultured hPASMCs using TRIzol and RNeasy Mini Kit (Qiagen, Hilden, Germany) as described previously (2). RNA was then reverse transcribed with oligo dT primers to obtain cDNA using a SuperScript First Strand kit (Invitrogen), and quantitative real-time PCR (Applied Biosystems, Foster City, CA) was performed using SYBR Green or EvaGreen Supermix (Applied Biosystems) following the manufacturer’s instructions (38). PCR was performed using mRNA isolated from rat PA and rat brain for control. The primers as shown in Table 1 were used in this study (IDT, Coralville, IA).

Adenovirus-encoding shRNA.

Ad-sh non-targeted and Ad-shOpn3-encoded viruses were generated using a pAdBLOCK-iT kit (ThermoFisher, Waltham, MA). Briefly, oligonucleotides targeting two different regions of murine Opn3 and nontargeted sequence were designed with proprietary software from Life Sciences and cloned into pU6-ENTR. Sequences used were as follows: NM_010098_shrna_492_top, CACCGCTGGCCTATGAACGTTATATCGAAATATAACGTTCATAGGCCAGC; NM_010098_shrna_492_top, AAAAGCTGGCCTATGAACGTTATATTTCGATATAACGTTCATAGGCCAGC; NM_010098_shrna_1160_top, CACCGGGACAGGCCAAAGAAGAAAGCGAACTTTCTTCTTTGGCCTGTCCC; NM_010098_shrna_1160_top, AAAAGGGACAGGCCAAAGAAGAAAGTTCGCTTTCTTCTTTGGCCTGTCCC.

The resulting pU6-sh-nontargeted and pO6-Opn3shRNA plasmids were tested for function in transient transfection experiments with 293 A cells. The constructs showing the greatest inhibition were LR recombined with pAD/BLOCK-iTDEST (Invitrogen) to generate pAd-shOpn3. Viruses were amplified and purified/concentrated using a Millipore (Billerica, MA) kit.

Myosin light chain phosphorylation.

rPASMCs were isolated as described above and subjected to overnight serum starvation in 0.5% FBS before being exposed to blue light via light diode (380–495 nm) for 90 s in the presence or absence of opsinamide (1 μM, EMD Millipore) or GRK2 inhibitor (1 μM). Cells were immediately lysed in ice-cold lysis buffer consisting of 50.0 mM Tris (pH 7.6), 1.0% Triton, 0.5% DOC, 0.1% SDS, 10.0 µg/ml leupeptin, 10.0 µg/ml aprotinin, 0.5 mM sodium vanadate, and 50.0 nM calyculin A. Samples were then prepped for Western blot analysis with antibodies against diphosphorylated myosin light chain.

Western blot analysis.

Samples were homogenized in 1× radioimmunoprecipitation assay (RIPA) buffer containing protease and phosphatase inhibitors (Roche Diagnostics, Indianapolis, IN), centrifuged at 13,500 g for 5 min at 4°C, and total protein concentration in the supernatants was determined (Bio-Rad Protein Assay Reagent; Bio-Rad, Hercules, CA). Equal amounts of protein from the samples (10–25 μg) were resolved by SDS-PAGE and electrotransferred to a nitrocellulose membrane or PVDF membrane. Western blot analysis was performed using the following primary antibodies as noted: rabbit ppMLC (Cell Signaling Technology, Beverly, MA), GAPDH (Novus Biologicals, Littleton, CO), rabbit Opn3, rabbit Opn4, and mouse GRK2 (same as described above). Secondary antibodies used were horseradish peroxidase-conjugated goat anti-rabbit (Jackson Immunoresearch, West Grove, PA; 111-035-003) or goat anti-mouse (Bio-Rad, 1721011). Chemiluminescent detection was performed with the Bio-Rad Clarity ECL reagent, and samples were imaged with a Bio-Rad ChemiDoc Touch system. Densitometry analysis was performed with ImageJ software. Comparisons between different groups were performed with all groups run on the same gel.

Rat lung perfusion system.

Rat lungs were perfused in situ as previously described (49). Wistar rats (200–400 g body wt) were injected with heparin (1,000 U ip) and anesthetized with pentobarbital sodium (65 mg/kg ip). A tracheostomy was performed, and rats were ventilated with room air at a tidal volume of 10 ml/kg and rate of 30/min. Rats were exsanguinated via the femoral artery, and the ventilating gas switched to 16-5% CO₂. The chest was then opened and cannulas inserted into the main PA and left atrium. The lungs were perfused in situ with a peristaltic pump at a rate of 40 ml·kg⁻¹·min⁻¹ from a heated, recirculating reservoir filled with Krebs solution containing (in mM): 118.00 NaCl, 4.70 KCl, 0.57 MgSO₄, 1.18 KH₂PO₄, 25.00 NaHCO₃, and 10.00 glucose. In addition, Ficoll (4 g/dl) was added to the perfusate to provide oncotic pressure, and sodium meclofenamate (3.1 μM) was added to prevent release of vasodilator prostaglandins. A heat exchanger in line with the PA cannula maintained perfusate at 37°C before entering the lung. P_PA, left atrial pressure, and tracheal pressure were measured relative to the bottom of the lung with pressure transducers (model P10EZ; Spectramed, Oxnard, CA) and recorded with a digital recording system (Powerlab; ADInstruments, Colorado Springs, CO). End-expiratory tracheal pressure was maintained at 3–4 mmHg. Because perfusate flow was constant, and left atrial pressures were maintained <0 mmHg, increases in P_PA were assumed to reflect pulmonary vasoconstriction.

After a 20-min stabilization period, the preparation was shielded with a light-impenetrable cover, and lungs were subjected to repeated cycles of angiotensin II (0.05 μg bolus into the PA cannula) followed 5 min later by hypoxia (4% O₂-5% CO_2; 5 min). During the third and fifth hypoxic exposures, 2 min after beginning hypoxia (at peak increase in P_PA), blue light was turned on for 3 min over the anterior portion of the lung. Between the third and fourth exposures, GRK2 inhibitor (1 µM) was added to the perfusate and allowed to recirculate for 15 min before continuing hypoxic exposures. Maximum P_PA (P_PA max) in response to hypoxia was determined as the difference between baseline P_PA under normoxia and P_PA measured after 2 min of hypoxic exposure, whereas change in P_PA was determined from the difference between P_PA max and P_PA measured after 4 min of hypoxic exposure.

Statistical methods.

All experiments were performed for at least five biological replicates (n = 5); specific numbers of replicates are reported for each experiment. Student’s t-tests and two-way ANOVA analysis were performed as appropriate (GraphPad Prism; GraphPad Software, San Diego, CA), and graphs were plotted as arithmetic mean with SE. Significance levels were determined between samples examined and were set at *P < 0.05, **P < 0.01, and ***P < 0.001.

Opn3 and Opn4, but not Opn5, are expressed in rat PA and in PASMCs.

We previously demonstrated that Opn4 was expressed in mouse aorta (38). As demonstrated in Fig. 1A, rat PAs expressed Opn3 and Opn4, but not Opn5. Because both Opn3 and 4 are classic G protein-coupled receptors, and GRK2 modulates Opn4 activity (6), we evaluated whether GRK2 was also expressed in rat PA. Indeed, rat PA expressed GRK2 (Fig. 1B). The expression of Opn3, Opn4, and GRK2 was further confirmed in isolated rat PASMCs via qRT-PCR (Fig. 1C) and Western blot (Fig. 1D). We next determined the mRNA expression and protein abundance of Opn3 and 4 as well as GRK2 in human PASMCs. As can be seen in Fig. 1E, hPASMCs stained positive for Opn3, Opn4, and GRK2 using immunofluorescence. This expression was confirmed at the mRNA level via PCR and qPCR (Fig. 1, F and G) and at the protein level via Western blotting (Fig. 1H). Consistent with previous results in rat PA, Opn5 was not detected.

PAs relax in response to light exposure in a wavelength-specific manner.

To test whether intact PAs demonstrate the phenomenon of photorelaxation, PAs were mounted in a wire myograph and exposed to light following preconstriction with PE (1 μM). PAs showed an intensity-dependent relaxation to cold white light, a process that was markedly amplified following preincubation with the GRK2-specific inhibitor methyl 5-[2-(5-nitro-2-furyl)vinyl]-2-furoate (1 μM) (Fig. 2A). We previously determined that this inhibitor amplified the vasorelaxation response to light in the aorta, with maximum relaxation achieved between 1 and 10 μM, and used this as our rationale for using 1–10 μM in all of our subsequent experiments.

To test whether this response was wavelength specific, vasorelaxation was measured in response to red (620–750 nm), green (495–570 nm), and blue (380–495 nm) light emitted by RGB LEDs. As shown in Fig. 2B, the response to both red and green light was negligible. However, there was a significant vasorelaxation response to blue light in vessels without GRK2 inhibitor and a markedly amplified response in the presence of GRK2 inhibitor (Fig. 2, B and C). Wavelength-specific blue light exposure (455 nm) similarly produced an amplified effect when combined with GRK2 inhibition (Fig. 2D). To further test the optimal wavelength at which photorelaxation occurs, a monochromator (Ludl Electronic Products, Hawthorne, NY; CM 110, Spectral Products, Putnam, CT) was used to expose PAs to light with increasing wavelengths (from 370 nm to 700 nm in 30-nm increments). Photorelaxation was observed at wavelengths in the blue light spectrum, with a maximum response between 400 nm and 460 nm (Fig. 2, E and F). This is consistent with the maximum absorption spectra of both Opn3 (460 to 500 nm) and Opn4 (420 to 480 nm) (27, 33). To test whether PE-induced contraction could be prevented with blue light, PAs were exposed to blue light in the presence of GRK2 inhibitor before and during exposure to increasing concentrations of PE (Fig. 2, G–I). PE-induced contraction was significantly attenuated by the combination of blue light and GRK2 inhibition.

Photorelaxation signal transduction involves both Opn3 and Opn4 receptors.

To determine the role of the expressed opsin receptors in mediating light-dependent relaxation, rat PAs were preincubated with opsinamide, a small molecule Opn4 inhibitor (Opn4 antagonist, EMD Millipore; 10 µM). Vessels preincubated with the Opn4 inhibitor showed a significant attenuation in blue light response (Fig. 3A). Opsinamide had no effect on endothelial-dependent vasorelaxation in response to increasing doses of ACh (Fig. 3B). The attenuation in photorelaxation following opsinamide preincubation remained significantly different when vessels were preincubated with GRK2 inhibitor (Fig. 3A). To further confirm the role of Opn4 in the response, photorelaxation was tested in PAs isolated from Opn4^−/− mice (Fig. 3C). Although there was a significant attenuation of photorelaxation in the vessels isolated from Opn4^−/− mice in response to blue light both without and with GRK2 inhibition, vasorelaxation induced by the NO donor SNP was unaffected (Fig. 3D). Given that neither vessels preincubated with the Opn4 inhibitor nor vessels isolated from Opn4^−/− mice demonstrated complete inhibition of the light-induced vasorelaxation response, we hypothesized that Opn3, which was also expressed in PA, might contribute to the photorelaxation response.

To test the role of Opn3 in photorelaxation, Opn3 expression was depleted in PAs by adenoviral delivery of Opn3 shRNA. The effectiveness of the Opn3 shRNA construct was confirmed via Western blot analysis of hPASMCs treated with Opn3 shRNA for 72 h, revealing significant knockdown of Opn3. Brain lysate was used as a positive control (data not shown). The combined roles of Opn3 and Opn4 were tested using light from a monochrometer of incremental wavelength (Fig. 3E). Maximum response was observed from 400 to 430 nm, suggesting that Opn3 also plays a role in the photorelaxation phenomenon observed, albeit with a slightly different maximum activation wavelength. SNP-induced vasorelaxation was similar in all samples (Fig. 3F).

GRK2 inhibition prevents desensitization of the photorelaxation response following repeated stimulation.

Given that both Opn3 and Opn4 are classic GPCRs, we tested the hypothesis that GRKs are important in mediating desensitization of the photorelaxation response. GRK2 has been demonstrated to be the GRK important in regulating Opn4 activity (6), whereas the regulation of Opn3 remains unknown. The role of GRK2 in vessel responses to a rapid repetitive blue light stimulus (455 nm) was tested in rat PAs preincubated with and without the GRK2-specific inhibitor (Fig. 4A). In the absence of GRK2 inhibitor, the photorelaxation response of PAs was rapidly attenuated with each subsequent stimulus, demonstrating a desensitization of the PA to repeated light exposure. On the other hand, in vessels preincubated with GRK2 inhibitor, vasorelaxation responses were markedly amplified, and these rings were resistant to desensitization with repetitive photostimulation (Fig. 4, A and B).

To determine whether GRK2 interacts directly with Opn3 and/or Opn4, we performed an in situ PLA using hPASMCs cultured on coverslips. In this assay, a signal is observed only when the two epitopes of interest are in close proximity (<40 nm apart), demonstrating a direct interaction between them. As such, we incubated hPASMCs with primary antibodies against GRK2 and Opn3 or Opn4. As shown in Fig. 4C, GRK2 was found to interact directly with both Opn3 and Opn4, with negative control samples showing no ligation events. To test whether endogenously expressed Opn3 and Opn4 were phosphorylated in human pulmonary vasculature, we performed a similar in situ PLA utilizing primary antibodies against phosphoserine and either Opn3 or Opn4. As seen in Fig. 4D, both Opn3 and Opn4 were found to be substantially phosphorylated in hPASMCs. Control PLA labeling using the primary antibody for phosphoserine alone resulted in no ligation events, and hence no signal was found.

Freshly isolated PASMCs respond to blue light.

To elucidate whether PASMCs alone respond to light, we seeded freshly isolated rPASMCs on collagen-coated elastic gels embedded with either red fluorescent or yellow-green fluorescent beads and measured single-cell traction stress with Fourier transform traction microscopy (1, 9). Red fluorescent beads were excited at 515–560 nm wavelength (i.e., the spectra of green light, which does not stimulate Opn3 or Opn4), whereas yellow-green fluorescent beads were excited at 450–490 nm, which falls within the blue light spectrum that activates Opn3/4 signaling. For these studies, cell traction measurements were performed before and after PE stimulation (10 μM, 5 min). As shown in Fig. 5, A–C, cells seeded on gels with yellow-green fluorescent beads excited with blue light (1 min) exhibited markedly reduced cell traction compared with cells seeded on gels with red fluorescent beads excited with green light. Strikingly, cells exposed to blue light did not constrict the gels in response to PE, whereas cells exposed to green light did (blue light: 4 out of 20 cells contracted; green light: 22 out of 35 cells contracted to PE; P = 0.004, Fisher’s exact test).

To determine whether the mechanism by which photorelaxation occurs involves hyperpolarization, we measured changes in membrane potential (E_m) in rPASMCs. Both KCl (80 mM) and PE (10 μM) depolarized rPASMCs (Fig. 5D). At peak depolarization, cells were exposed to blue or red light in the absence or presence of the GRK2 inhibitor. In the absence of GRK2 inhibition, applying blue light had no significant effect on E_m in rPASMCs depolarized with either KCl or PE. GRK2 inhibition had no effect on the ability of KCl or PE to induce depolarization but resulted in significant blue light-induced reversal of depolarization (i.e., hyperpolarization) in cells depolarized with PE. Application of red light to PE-depolarized cells or blue light in KCl-depolarized cells caused no significant change in E_m.

Because smooth muscle cell relaxation has been linked to the phosphorylation state of myosin light chain (MLC) (41, 47), we further hypothesized that photorelaxation in response to blue light exposure is associated with reduced diphosphorylation of MLC (ppMLC) in rPASMCs. Indeed, we observed a marked reduction in ppMLC in cells exposed to blue light (Fig. 5E). To test whether the decrease in ppMLC was mediated by Opn4 or GRK2, we pretreated cells with GRK2 inhibitor or with the Opn4 inhibitor opsinamide. Interestingly, MLC diphosphorylation was further reduced in blue light-exposed rPASMCs pretreated with GRK2 inhibitor. In contrast, opsinamide effectively ablated the reduction in diphosphorylation of MLC in response to blue light (no significant difference from control samples), suggesting that Opn4 mediates light-induced reduction of MLC diphosphorylation in rPASMCs.

Photorelaxation in PAs occurs by an NO-independent mechanism.

To determine the signal transduction mechanism involved in opsin-mediated photorelaxation, we first explored whether the response occurs through an NO-dependent or -independent mechanism. For this, PAs were preincubated with the NO synthase (NOS) inhibitor nitro-l-arginine methyl ester (l-NAME) (Sigma-Aldrich; 200 μM) before constriction with PE. Although there was a marked attenuation in ACh-mediated vasorelaxation in rat PAs pretreated with l-NAME, there were no significant differences in photorelaxation responses to blue light (Fig. 6, A and B). We next tested whether the photorelaxation effect was due to light-mediated release of NO from thiol-linked stores in the vessel. Preincubation of vessels with the NO scavenger cPTIO (Tocris Bioscience, Avonmouth, UK; 300 μM) significantly attenuated ACh-mediated vasorelaxation but did not affect photorelaxation responses to blue light (Fig. 6, C and D).

Photorelaxation signal transduction involves soluble guanylyl cyclase.

Classic opsin signaling involves cGMP-gated ion channels (14). To test the role of soluble guanylyl cyclase (sGC) as an enzyme responsible for cGMP synthesis, PAs were preincubated with the sGC inhibitors, 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) (Tocris Bioscience; 0.5 μM) or NS2028 (Tocris Bioscience; 0.1 μM). ODQ significantly attenuated vasorelaxation by SNP and, interestingly, markedly attenuated photorelaxation to blue light (Fig. 7, A and B). NS2028 similarly attenuated photo-induced vasorelaxation (Fig. 7C). These findings indicated a key role for cGMP in the photorelaxation signaling cascade.

Photorelaxation signal transduction is independent of protein kinase G.

To evaluate whether cGMP signals through the classic protein kinase G (PKG) pathway, PAs were preincubated with PKG inhibitors Rp-8-Br-PET-cGMPs (Tocris Bioscience; 10 μM) or sc201161 (Santa Cruz Biotechnology, Dallas, TX; 10 μM) (Fig. 8, A–D). Both PKG inhibitors significantly attenuated vasorelaxation mediated by ACh but did not affect vasorelaxation induced by blue light, indicating a noncanonical pathway for cGMP-mediated phototransduction.

Phototherapy can induce vasorelaxation in PAs from rats with PH.

Consistent with reports from other laboratories, we previously demonstrated that rats treated with SU5416 and exposed to hypoxia for 3 wk (SuHx) followed by return to normoxia for 2 additional weeks develop severe PH (22). PAs from SuHx rats showed a significant photorelaxation to blue light, which was markedly amplified by addition of GRK2 inhibitor (Fig. 9A). Notably, ACh treatment produced only a moderate vasorelaxation response in SuHx PAs compared with controls (Fig. 9B). Similar to the effect in normal PAs, GRK2 inhibition and blue light successfully prevented PE-induced constriction in PAs from SuHx rats (Fig. 9C).

Phototherapy reverses hypoxia-induced pulmonary vasoconstriction.

A ventilated, isolated, perfused lung model was used to test photorelaxation during hypoxia-induced pulmonary vasoconstriction in situ. P_PA increased significantly after switching from a normoxic to a hypoxic (4% O₂) gas mixture (Fig. 9D). In control animals, exposure of the anterior portion of the lung to blue light after reaching maximum vasoconstrictive responses induced by hypoxia had no significant effect compared with the slow downward drift in P_PA observed in non-light-exposed lungs (Fig. 9E). Addition of the GRK2 inhibitor had no effect on baseline P_PA and did not alter the maximum contractile response to angiotensin II or hypoxia. However, in GRK2-treated lungs in hypoxic conditions, exposure to blue light resulted in an immediate decline in P_PA of 37% during the first 2 min of exposure, compared with a decrease of only 12% for the same time period in the absence of blue light.