Paracrine purinergic signaling determines lung endothelial nitric oxide production

Abstract

Although the vascular bed is a major source of nitric oxide (NO) production, factors regulating the production remain unclear. We considered the role played by paracrine signaling. Determinations by fluorescence microscopy in isolated, blood-perfused rat and mouse lungs revealed that a brief lung expansion enhanced cytosolic Ca²⁺ (Ca²⁺cyt) oscillations in alveolar epithelial (AEC) and endothelial (EC) cells, and NO production in EC. Furthermore, as assessed by a novel microlavage assay, alveolar ATP production increased. Intra-alveolar microinfusion of the purinergic receptor antagonist, PPADS, and the nucleotide hydrolyzing enzyme, apyrase, each completely blocked the Ca²⁺cyt and NO responses in EC. Lung expansion induced Ca²⁺cyt oscillations in mice lacking the P2Y1, but not the P2Y2, purinergic receptors, which were located in the perivascular interstitium basolateral to AEC. Prolonged lung expansion instituted by mechanical ventilation at high tidal volume increased EC expression of nitrotyrosine, indicating development of nitrosative stress in lung microvessels. These findings reveal a novel mechanism in which mechanically induced purinergic signaling couples cross-compartmental Ca²⁺cyt oscillations to microvascular NO production.

because of its vast endothelial (EC) surface area, the lung is a major producer of EC nitric oxide (NO) (6). However, although the molecular mechanisms of EC-derived NO production are well understood (reviewed in Ref. 18), the mechanisms by which the lung triggers EC NO production remain unknown. Here, we consider the possibility that the trigger mechanism might be determined by lung mechanics. Thus lung expansion, as during a sigh, or during high-tidal volume mechanical ventilation, might induce vectorial signaling from alveoli to capillaries to initiate EC NO production consequent to alveolar stretch.

Our findings in intact alveoli indicate that lung expansion, even if held briefly for a few seconds, increases cytosolic Ca²⁺ (Ca²⁺cyt) oscillations in alveolar epithelial cells (AEC) (2). Studies in cultured AEC indicate that cell stretch releases ATP (1, 10, 15, 20). Hence, we considered the possibility that these mechanisms might be linked in EC NO production. Thus, ATP could induce vectorial signaling by ligating purinergic receptors of the P2Y and the P2X families that include eight isoforms each (Y1, 2, 4, 6, 11–14; X^1–7 and X^M) (1, 3). P2Y receptors induce Ca²⁺ signaling through inositol triphosphate (IP3)-mediated endoplasmic reticular Ca²⁺ release, while P2X receptors are Ca²⁺ channels. Although P2Y and P2X receptors are expressed in pulmonary vascular and airway epithelial cells (reviewed in Ref. 1), direct evidence for alveolar expression is scant (reviewed in Ref. 23). Hence, no understanding exists as to whether this receptor group plays a role in EC NO production.

We considered these issues in the context of real-time fluorescence imaging of the lung that provides a unique approach for defining cross-compartmental signaling mechanisms between alveoli and juxta-alveolar microvessels (13). We report below that lung expansion sets up P2Y2-driven Ca²⁺cyt oscillations that in turn induce NO production in juxta-alveolar capillaries. Thus, we show for the first time that in lung, a purinergic mechanism translates the mechanical effects of alveolar expansion to induction of microvascular NO production.

METHODS

Fluorescent dyes and reagents.

Fura 2-AM and 4-amino-5-methylamino-2′,7′-difluorofluorescein (DAF-FM diacetate) were purchased from Molecular Probes (Eugene, OR). ATP, UTP, 2-methylthio-ATP tetrasodium salt (MeSATP), apyrase, 4-{4-formyl-5-hydroxy-6-methyl- 3-[(phosphonooxy)methyl]-2-pyridinylazo}-1,3-benzenedisulfonic acid (PPADS) tetrasodium salt, N^G-nitro-l-arginine methyl ester (l-NAME), S-nitroso-N-acetylpenicillamine (SNAP), and t-BHQ (tert-butylhydroquinone) were purchased from Sigma-Aldrich (St. Louis, MO). Saponin was purchased from Calbiochem-Novabiochem (San Diego, CA). Vehicle for dyes and other agents was HEPES buffer (150 mmol/l Na⁺, 5 mmol/l K⁺, 1.0 mmol/l Ca²⁺, 1 mmol/l Mg²⁺, and 20 mmol/l HEPES, pH 7.4) containing 4% dextran (70 kDa) and 1% FBS, pH 7.4, and osmolarity of 295 mosM. Polyclonal rabbit anti-nitrotyrosine antibody was purchased from Upstate (Temecula, CA). Polyclonal rabbit anti-P2Y1 and P2Y2 receptor antibodies were purchased from Chemicon International. Alexa Fluor 546 F(ab′)₂ fragment of goat anti-rabbit IgG was purchased from Invitrogen (Carlsbad, CA).

Lung preparation.

All animal procedures were approved by the Institutional Animal Use and Care Committee of St. Luke's-Roosevelt Hospital Center. Using our reported methods (2), lungs were isolated from Sprague-Dawley rats (anesthetized with 3.5% halothane inhalation and 35 mg/kg intraperitoneal pentobarbital sodium), and blood was perfused (14 ml/min) at 37°C. Lungs were constantly inflated through an airway cannula at baseline airway pressure (Palv) of 5 cmH₂O. Pulmonary artery and left atrial (P_LA) pressures were held at 10 and 2 cmH₂O, respectively. For intra-alveolar infusions of fluorescent probes and drugs, a single alveolus was micropunctured. The microscope objective was adjusted until the maximum diameter of the alveolus came into focus. For intracapillary infusion of fluorescent probes and drugs, a microcatheter (PE 10; Baxter Diagnostics, Deerfield, IL) was introduced through the left atrium and wedged in a pulmonary vein draining a small area on the lung surface. In mouse experiments, lungs were isolated from C57BL wild-type (WT)-mice purchased from Jackson Labs (Bar Harbor, ME) as well as from C57BL mice lacking P2Y1 [P2Y1^(−/−)] or P2Y2 [P2Y2^(−/−)] receptors. The P2Y1^(−/−) and P2Y2^(−/−) mice were kindly provided by Dr. B. H. Koller (Univ. of North Carolina at Chapel Hill, Chapel Hill, NC). Fluorescent probes and drugs given to the perfusate consisted of 4% dextran and 1% BSA dissolved in HEPES buffer at hematocrit of 20%.

Real-time lung imaging.

Juxta-alveolar capillaries were viewed with a charge-coupled device camera (Photometrics) and an epifluorescence microscope (Nikon ×40 water-immersion objective, Nikon, Japan). A two-photon microscope (Radiance 2100 MP; Zeiss, Melville, NY; ×40 water-immersion objective, Nikon; Chameleon laser; Coherent, Santa Clara, CA) was used for localization of P2Y1 and P2Y2 receptors. During imaging, we infused the capillaries with buffer to establish blood-free conditions. Capillary fluorescence was quantified using image analysis software (Metafluor; Molecular Devices, Downington, PA).

Ca²⁺ and NO imaging.

Ca²⁺ imaging and analysis methods have been reported (11). For epithelial and endothelial Ca²⁺ determinations, fura 2-AM (10 μM) was given for 30 min by intra-alveolar or intracapillary infusion, respectively. Alveoli and capillaries loaded with fura 2 were excited at 340 nm and 380 nm, and Ca²⁺cyt was determined from the computer-generated 340:380 ratio based on a K_d of 224 nmol/l and appropriate calibration parameters (11, 12). NO production was detected by EC fluorescence following a 20-min intracapillary infusion of DAF-FM diacetate (10 μM; excitation at 490 nm). Background levels were determined in capillaries not loaded with fluorescent dyes. Imaged gray levels are reported after background subtraction.

Indirect immunofluorescence in situ.

To determine endothelial nitrotyrosine formation, anti-nitrotyrosine antibody (10 μg/ml) was infused into capillaries for 10 min, and then secondary Alexa 546 MAb (1:200) was infused for 10 min. Unbound fluorescence was removed by perfusing the capillaries with HEPES for 5 min. For abluminal staining of ECs, we permeabilized microvessels by infusion of 4% paraformaldehyde and 0.1% Triton X-100 for 10 and 4 min, respectively. We then infused the capillaries with rabbit anti-rat P2Y1 or P2Y2 antibody (8 μg/ml) followed by infusion of Alexa 546 (1:200), and HEPES for washing. For determination of abluminal localization of the P2Y1 and P2Y2 receptors, EC nuclei were stained by capillary infusion of Hoechst 33342 (Molecular Probes, Eugene, OR) followed by a wash.

Experimental procedures.

Palv was elevated by increasing airflow through the airway cannula (2). We gave intra-alveolar injections by alveolar micropuncture according to our reported methods (8, 21). Each microinfusion was given for 10 min to fill seven to ten alveoli. The micropunctured alveolus was excluded from the imaged field. Agents given by alveolar microinfusion were: ATP (100 μM), the ATP inhibitor apyrase (50 U/ml), and the purinergic receptor antagonist PPADS (100 μM). Agents given through the venous microcatheter included ATP (100 μM), UTP (100 μM), PPADS (100 μM), the specific P2Y1 receptor antagonist MRS2179 (100 μM), t-BHQ (10 μM), and l-NAME (250 μM). These agents were each given for 10 min. For some infusions, PLA was varied as stated. Ca²⁺-free conditions were established by infusion of Ca²⁺-free HEPES solution containing 4% dextran, 1% FBS, and EGTA (0.5 mmol/l).

ATP assay in alveolar microlavage.

An assay that generates NADPH in the presence of ATP was used (5). The assay is based on the reduction of NADP to fluorescent NADPH by glucose-6-phosphate dehydrogenase (G6PD) in the presence of glucose and ATP. For alveolar microlavage, we introduced two micropipettes in two neighboring alveoli. Through one micropipette we infused buffer containing glucose (10 mM), NADP (5 mM), hexokinase, and G6PD (8 U/ml each). After 2 min of infusion, we collected ∼0.5 μl of the infused solution through the second micropipette. We determined ATP content in the collected sample by quantifying NADPH fluorescence (excitation 340 nm; emission 430 nm) in the micropipette tip, which we viewed by fluorescence microscopy (IX 81; Olympus America, Melville, NY). In the inhibitor group, alveoli were pretreated with apyrase (50 U/ml).

Statistics.

All data are represented as means ± SE. Statistical data analysis was performed using SigmaStat (Systat Software, San Jose, CA). Comparisons between the groups were tested using ANOVA on ranks by the Dunn method. Repeated measurements were tested using the Mann-Whitney Rank Sum Test. Statistical significance was accepted at P < 0.05.

RESULTS

Ca²⁺cyt responses.

As shown in the images in Fig. 1, A and D, a 15-s lung expansion caused by an increase of Palv from 5 to 15 cmH₂O induced marked increases in the amplitude of Ca²⁺cyt oscillations in both AEC of the intact alveolus and EC of juxta-alveolar capillaries. The tracings show that although the AEC oscillations were larger than the EC oscillations, in each case the mean Ca²⁺cyt remained unchanged (Fig. 1, C and E).

Fig. 1.

Effects of lung expansion. A: images in pseudocolor show the 340, 380, and 340/380 ratio for fura 2 fluorescence in an alveolus. In each image set, the region enclosed by the rectangle in the low-power image (left) is shown at high power in the 3 right images that were obtained sequentially, 5 min after a 15-s lung expansion. B: tracings show inflation-induced changes in pseudocolor intensity in 340 and 380 nm fura 2 excitation channels. C: tracing shows inflation-induced Ca²⁺ oscillations in alveolar epithelial cells. D: images in pseudocolor show the 340:380 ratio for fura 2 in a capillary. The region enclosed by the rectangle in the low-power image (left) is shown at high power in the 3 right images that were obtained sequentially, 5 min after a 15-s lung expansion. E: tracing shows inflation-induced Ca²⁺ oscillation in pulmonary endothelium. Images (F) and group data (G) show NADPH fluorescence in microlavage fluid held in micropipette tips. Each bar is mean for 6 samples obtained from 6 lungs. *P < 0.05 compared with preinflation control. H: epithelial Ca²⁺ oscillation amplitudes for the indicated interventions. Means ± SE, n = 4 each bar, *P < 0.05 vs. baseline, ^#P < 0.05 vs. inflation effect. I: endothelial Ca²⁺ oscillation amplitudes are shown for the indicated interventions. Inf, inflation. Agent abbreviations and doses are as in methods. Means ± SE, n = 4 for each bar, *P < 0.05 vs. baseline, ^#P < 0.05 vs. inflation alone, †P < 0.05 vs. ATP alone.

To determine whether these Ca²⁺cyt responses resulted from alveolar ATP release, we assayed ATP content in terms of NADPH fluorescence in alveolar washes collected by microlavage as described in methods. A 15-s lung expansion markedly increased lavage ATP (Fig. 1F). Pretreating alveoli with the nucleotide hydrolyzer apyrase blocked the increase of NADPH fluorescence in the microlavage (Fig. 1G). Apyrase did not affect the control fluorescence, indicating that under baseline conditions the fluorescence did not reflect ATP content. These findings affirmed that lung expansion induced alveolar ATP release.

With Palv held at the baseline level of 5 cmH₂O, intra-alveolar ATP injections replicated the lung expansion-induced Ca²⁺cyt oscillations in both AEC and EC (Fig. 1, H and I). Pretreatment with intra-alveolar apyrase blocked both expansion- and ATP-induced responses in AEC and EC (Fig. 1, H and I), whereas similar pretreatment with the purinergic receptor antagonist PPADS blocked the inflation- and ATP-induced EC Ca²⁺cyt responses (Fig. 1I). Furthermore, the lung expansion-induced response was blocked by t-BHQ, the inhibitor of Ca²⁺ release from the endoplasmic reticulum (ER) (Fig. 1I). The inhibitors given alone were without effect (data not shown). Together, these findings suggested that the AEC and EC Ca²⁺cyt oscillations resulted from ligation of an alveolar purinergic receptor.

Receptor type.

To define receptor expression, we permeabilized EC membranes by capillary microinfusions of Triton X-100. Then, we sequentially microinfused a nuclear-staining dye (Hoechst 33342), a primary MAb against the P2Y1 or the P2Y2 receptor, and a fluorescent secondary IgG directed against the primary MAb. Finally, we removed unbound fluorescence from the vessel lumen by buffer infusion and imaged the alveolo-capillary region by two-photon microscopy. As exemplified by the images in Fig. 2, A and B, P2Y2 distribution was evident on both the luminal and interstitial aspects of AEC, but only on the interstitial aspect of EC. By contrast, P2Y1 expression was not detectable (not shown). These findings suggest that EC of lung capillaries specifically expressed P2Y2.

Fig. 2.

Purinergic receptor distribution in the alveolo-capillary region. ALV, alveolus; CAP, capillary. A: low-power image shows immunofluorescence of the P2Y2 receptor shown in red juxtaposed with Hoechst-stained nuclei shown in blue. B: image at high power shows P2Y2 fluorescence in the luminal aspect of alveolar epithelial cells (single arrows) and in the interstitial space (double arrows) between alveolar and capillary lumens. The interstitial space also contains Hoechst-stained nuclei (N) of nonspecific cells. C: responses are shown for the indicated conditions. INF, inflation; P_LA, capillary pressure. Agent abbreviations and doses are as described in methods. Means ± SE, n = 4 for each bar, *P < 0.05 vs. P_LA = 3 cmH₂O. #P < 0.05 vs. ATP at P_LA = 8 cmH₂O.

Since the receptors were expressed abluminally on EC, for a blood-borne ligand to access the receptors, the transcapillary liquid flow would have to be in the direction of microvascular filtration, namely outward from the vascular lumen. We previously established that outward filtration occurs at left atrial pressure (P_LA) of 8 cmH₂O, whereas at P_LA of 3 cmH₂O transcapillary flow is in the opposite direction (24). Accordingly, intracapillary ATP injections at P_LA of 8 cmH₂O doubled the Ca²⁺cyt oscillation amplitude in EC, although injections at P_LA of 3 cmH₂O were without effect (Fig. 2C). At P_LA of 8 cmH₂O, intracapillary injection of PPADS inhibited the lung expansion-induced increase of the EC Ca²⁺cyt oscillations (Fig. 2C), but not the AEC oscillations (not shown). These Ca²⁺cyt data supported the morphological evidence that P2Y2 expression is strategically located between the basolateral regions of AEC and EC such that following inflation, nucleotides released basolaterally by AEC might ligate interstitial purinergic receptors.

To determine the purinergic receptor type responsible for the present EC responses, we gave intracapillary infusions of several ligands at P_LA of 8 cmH₂O (Fig. 3A). The P2Y2 agonist UTP increased Ca²⁺cyt oscillation amplitude in EC. However, an equimolar infusion of the P2Y1 agonist 2 MeSATP failed to induce EC Ca²⁺cyt responses. Furthermore, the P2Y1 inhibitor MRS2179 failed to block the ATP- and UTP-induced EC Ca²⁺ responses (shown for UTP only). Infusion of MRS2179 alone was without effect.

Fig. 3.

Effects of purinergic receptor ligation on endothelial Ca²⁺ responses. Data show responses to agents given by intracapillary infusion at P_LA of 8 cmH₂O (A) and to a 15-s lung inflation in wild-type (WT) and the indicated knockout mice (B). EC, endothelial cells; AEC, alveolar epithelial cells. Agent abbreviations and doses are given in methods. Means ± SE, n = 4 for each bar. *P < 0.05 vs. baseline.

Since these findings suggested the involvement of P2Y2 receptors, we determined inflation-induced responses in mice lacking P2Y1 [P2Y1^(−/−)] or P2Y2 [P2Y2^(−/−)] receptors. As similar to rat lungs, inflation-induced increases in Ca²⁺cyt oscillations occurred in both AEC and EC of WT and P2Y1^(−/−) mice. However, the oscillations were completely absent in P2Y2^(−/−) mice (Fig. 3B). Hence, ligation of P2Y2 receptors was specifically responsible for the inflation-induced EC Ca²⁺cyt responses.

NO.

Since increase of EC Ca²⁺cyt increases EC NO, we imaged capillary fluorescence of the NO-detecting intracellular dye DAF in capillaries perfused with RBC-free buffer (Fig. 4A) and confirmed that the NO donor SNAP increased the fluorescence (Fig. 4B). At baseline Palv, no significant changes in DAF fluorescence occurred in 20 min (Fig. 4D), indicating that basal NO production was below our detection limit. However, as shown by the image (Fig. 4C), the tracing (Fig. 4D) and the group data (Fig. 4E), following a 15-s lung expansion, DAF fluorescence increased linearly for more than 30 min at a rate that was markedly higher than baseline. This response was completely blocked by pretreating the capillaries with the NOS inhibitor l-NAME, by intra-alveolar PPADS and by capillary infusion of t-BHQ in Ca²⁺-free buffer (Fig. 4E), but not by infusion of Ca²⁺-free buffer alone (not shown). These findings indicate that lung expansion-induced EC NO production occurred secondary to intra-alveolar ligation of alveolar purinergic receptors followed by ER Ca²⁺ release in EC.

Fig. 4.

Endothelial nitric oxide (NO) production. A–C: representative images of lung venular capillaries loaded with DAF-FM. Images were obtained at baseline (A), 5 min after intracapillary infusion of 1 mM SNAP (B), and 30 min after airway pressure elevation from 5 to 20 cmH₂O for 15 s (C). D: tracing of DAF-FM fluorescence in a single endothelial cell of a lung venular capillary before and after a 15-s inflation (INF). E: group data show responses under indicated conditions. Agent abbreviations and doses are as described in methods. Means ± SE, n = 4 for each bar, *P < 0.05 vs. baseline, #P < 0.05 vs. inflation alone.

To determine the extent to which lung expansion-induced EC NO production caused EC stress, we determined the presence of nitrotyrosine in venular capillaries by means of in situ immunofluorescence. We mechanically ventilated lungs at high tidal volume (HV) for 20 min by cycling Palv at inspiratory and expiratory pressures of 20 and 5 cmH₂O (12 cycles/min). As shown by the images (Fig. 5, A–E) and the group data (Fig. 5F), ventilation caused an increase of nitrotyrosine staining on the EC luminal surface. This increase was blocked by preinjecting intra-alveolar PPADS, or intracapillary l-NAME, or intracapillary PEG-catalase. Buffer injections given by the intra-alveolar or intra-capillary routes were without effect. These findings indicate that in HV, alveolar purinergic receptor ligation led to EC NO production, hence nitrotyrosine formation on the EC membrane.

Fig. 5.

Endothelial nitrotyrosine expression. Images show immunofluorescence of nitrotyrosine in lung venular capillaries. Fluorescence intensities are color coded. Vessel margins are depicted by arrowheads (A). Images are obtained at baseline (A) and following 20 min of ventilation either with no treatment (B) or after the indicated treatments (C–F). BL, baseline; UN, untreated; PP, PPADS; LN, l-NAME; PG, PEG-catalase. Agent abbreviations and doses are as described in methods. Means ± SE, n = 4 for each bar, *P < 0.05 vs. baseline.

DISCUSSION

We show here that lung mechanics plays a critical role in NO production from the lung's vascular bed. A single lung expansion held for 15 s augmented Ca²⁺cyt oscillations not only in the alveolus, as previously reported from our group (2), but also in juxta-alveolar capillaries. These responses were ATP induced as indicated by several key results, namely, first, the AEC and EC Ca²⁺cyt oscillations could not be induced in mice lacking the P2Y2 receptor. Second, intra-alveolar injections of the ATP inhibitor apyrase and the purinergic receptor antagonist PPADS each inhibited both sets of Ca²⁺cyt oscillations. Third, lung expansion caused alveolar ATP release, as detected in microlavages sampled directly from the alveolus. Fourth, intra-alveolar injection of ATP elicited Ca²⁺cyt oscillations similar to those resulting from lung expansion. Together, these findings are the first evidence that the mechanical effects of lung expansion induce fast, ATP-dependent alveolo-capillary communication of Ca²⁺ signals.

An important consequence of this communication was NO production in juxta-alveolar microvessels, as determined by DAF fluorescence. Nitrosylation of DAF by NO forms a highly fluorescent benzotriazole (12). DAF imaging allows quantification of nanomolar NO, providing a highly sensitive approach for NO detection (27). Since the fluorophore is retained in the cell, cellular NO production results in a progressive increase of DAF fluorescence. A 15-s lung expansion induced a steady increase of EC DAF fluorescence for up to 40 min, corresponding to the duration for which the increases in AEC and EC Ca²⁺cyt oscillations were maintained. These findings indicate that a major effect of the lung expansion was to increase capillary NO production. Importantly, the effect was inhibited by infusion of t-BHQ in Ca²⁺-free buffer, implicating ER Ca²⁺ release in the NO increase. Since ER Ca²⁺ release determines Ca²⁺cyt oscillations, we interpret that the induced Ca²⁺cyt oscillations were responsible for the EC NO increase.

Direct evidence by real-time lung imaging as in the present studies, in our previous reports (9, 19), and by the Kuebler group (11, 28), indicates that lung capillaries constitutively produce low levels of NO under nonstressed conditions. Although multiple reports implicate capacitative Ca²⁺ entry (CCE) in EC NO production, CCE associates with lung stresses such as pulmonary edema (14) and is therefore unlikely to underlie nonstressed lung microvascular NO production. Here, enhancement of EC Ca²⁺cyt oscillations enhanced EC NO production, and the durations of these two responses matched. Further, the NO increase occurred under external Ca²⁺-free conditions. These findings rule out a causal role for CCE in the present NO responses. We suggest that for the relatively low levels of NO that are produced in lung capillaries during unstressed conditions, constitutive EC Ca²⁺cyt oscillations (14, 29) might provide a sufficient stimulus both at baseline as well as during lung inflation. The main function of the capillary NO might be to inhibit platelet aggregation, hence hemostasis in lung vessels (7).

Methodological considerations include the possibility that microinjected agents leaked across the alveolar wall. For example, if apyrase or PPADS leaked across the alveolar wall, the agents might have inhibited the induced Ca²⁺cyt oscillations by directly blocking EC purinergic receptors. However, this possibility may be ruled out since intra-alveolar apyrase or PPADS did not inhibit the increase of EC Ca²⁺cyt oscillations induced by intra-capillary infusion of ATP at P_LA of 8 cmH₂O, indicating absence of alveolar leak. Moreover, we confirmed our previous findings (data not shown) that following intra-alveolar microinjection, agents such as fura 2 are retained in the alveolar compartment and do not passively leak across to capillaries (13), such that the fura 2 fluorescence is exclusively alveolar.

Circulating blood cells may be excluded as a source of NO because the imaged capillaries were perfused with cell-free HEPES buffer instead of blood. Constant EC fluorescence intensities at baseline and following l-NAME showed that photo bleaching did not take place under our experimental conditions. Previous reports suggest that increase of DAF fluorescence can also be obtained by superoxide anions (O₂^·−) or by peroxynitrite formed by the reaction of NO with O₂^·−. The possibility that the present increase of DAF fluorescence was induced by O₂^·− alone may be ruled out because the fluorescent response was inhibited by the NO inhibitor, l-NAME, which to our knowledge is not a superoxide scavenger.

Inhibition of the communicated Ca²⁺ responses in P2Y2^(−/−) mice, but not P2Y1^(−/−) mice, indicated that the P2Y2 receptor was critical for this signaling. Consistent with this conclusion was the immunofluorescence evidence for the P2Y2, but not the P2Y1, receptor on the basolateral aspects of AEC and EC. The somewhat patchy and nonuniform fluorescence distributions in (Fig. 2, A and B) suggest that the membrane permeabilization procedures might have caused receptor loss. Morphological characterization by ultra-structural analyses might better resolve the basolateral localization of the receptor.

However, in support were the ligational responses to vascular injections of P2Y2 ligands that were evoked only at vascular pressures that caused ligand flux across the capillary wall to basolateral sites. Our detection of ATP in the alveolar microlavage following lung expansion confirms that the nucleotide was apically secreted by AEC. However, it is unlikely that apically secreted ATP was transported across the restrictive alveolar epithelial barrier. Since AEC might secrete ATP basolaterally, as shown for airway epithelial cells (4), it appears that the location of P2Y2 receptor at the EC basolateral aspect appropriately places the receptor on the vectorial route of Ca²⁺ communication between and alveoli and capillaries.

Our ATP assay in micropipette collections of the alveolar microlavage provides a new approach for analyzing fluid recovered directly from the pulmonary acinus. The advantage is that acinar collections lack contaminations from the upper airway as might be the case for conventional bronchoalveolar lavage. Since micropuncture does not cause alveolar leaks, as we have shown previously (13), the acinar collections are not contaminated by liquid from the extra-alveolar interstitium. Given that acinar volume in the rat is ∼2 μl (25) and that each micropipette collection was ∼0.5 μl, we conclude that the collections were obtained entirely from a single acinus.

The cellular source of the present ATP secretion remains undetermined. Cultured A549 cells and airway epithelial cells secrete ATP in response to cell distortion induced by hypotonic exposure, or cyclic compression (4, 26). In a study of live alveoli by confocal microscopy, we determined that the bulk of the alveolar distortion during lung expansion occurs in type 1, and not type 2, epithelial cells (21). It is possible therefore that here, type 1 cells were the primary acinar sources of exogenous ATP. Further understanding is required as to whether loss of alveolar elastic recoil in elastin-degrading diseases increases type 1 cell stretch, thereby increasing ATP secretion.

Thus our findings reveal a novel role for extracellular ATP, namely as an alveolo-capillary signal transmitter that is activated by lung mechanical distortion. Thus, in addition to being a secretagogue in the lung parenchyma (17, 20), ATP might be relevant to alveolo-capillary function under normal and stressed conditions. As we show here, an important consequence of this ATP-based communication is an increase in EC NO production, an effect that might explain the regional vasodilatation that follows lung expansion (16). However, in high-tidal volume mechanical ventilation (HV), the NO effects might be deleterious. Our findings support this possibility since HV increased nitrotyrosine formation in capillaries, indicating development of nitrosative stress. Hence, since HV induces lung ATP secretion (22), the associated lung injury might be due to nitrosative stress caused by increased NO formation.

In conclusion, our findings provide the first direct evidence that EC NO production following lung expansion is regulated by alveolar nucleotide release, which induces vectorial alveolo-capillary signaling through cascades of Ca²⁺cyt oscillations in AEC and EC. To the extent that this signaling induces vascular nitrosative stress, alveolar ATP secretion might be an appropriate therapeutic target for prevention of ventilator-induced ALI.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants HL-57556, HL-64598, HL-78645, and HL-69514 (to J. Bhattacharya) and HL-75503 (to K. Parthasarathi). R. Kiefmann was partly supported by a Research Fellowship from the Deutsche Forschungsgemeinschaft.