At birth, pulmonary vasodilation occurs concomitant with the onset of air-breathing life. Whether and how Rho kinase (ROCK) modulates the perinatal pulmonary vascular tone remains incompletely understood. To more fully characterize the separate and interactive effects of ROCK signaling, we hypothesized that ROCK has discrete effects on both pulmonary artery (PA): 1) endothelial cell (PAEC) nitric oxide (NO) production and contractile state; and 2) smooth muscle cell tone independent of endothelial NO synthase (eNOS) activity. To test these hypotheses, NO production and endothelial barrier function were determined in fetal PAEC under baseline hypoxia and following exposure to normoxia with and without treatment with Y-27632, a specific pharmacological inhibitor of ROCK. In acutely instrumented, late-gestation ovine fetuses, eNOS was inhibited by nitro-l-arginine infusion into the left PA (LPA). Subsequently, fetal lambs were mechanically ventilated (MV) with 100% oxygen in the absence (control period) and presence of Y-27632. In PAEC, treatment with Y-27632 had no effect on cytosolic calcium but did increase normoxia-induced NO production. Moreover, acute normoxia increased PAEC barrier function, an effect that was potentiated by Y-27632. In fetal lambs, MV during the control period had no effect on LPA flow. In contrast, MV after Y-27632 increased LPA flow and fetal arterial Po2 (PaO2) and decreased PA pressure. In conclusion, ROCK activity modulates vascular tone in the perinatal pulmonary circulation via combined effects on PAEC NO production, barrier function, and smooth muscle tone. ROCK inhibition may represent a novel treatment strategy for neonatal pulmonary vascular disease.
- pulmonary hypertension
- nitric oxide
in utero, oxygen tension is low, and pulmonary vascular resistance is greater than systemic vascular resistance (22). At birth, the pulmonary circulation undergoes a biologically imperative transition, as pulmonary blood flow increases 8- to 10-fold and arterial pressure decreases by 50% within 24 h concomitant with an increase in oxygen tension, establishment of an air-liquid interface, and rhythmic distention of the lung (4, 7, 29).
The physiological mediators of perinatal pulmonary vasodilation include ventilation, oxygenation, and an increase in shear stress. Each of these stimuli act, to a significant degree, through an increase in endothelial-dependent nitric oxide synthase (eNOS) activity and elaboration of nitric oxide (NO) (5). NO, in turn, causes activation of pulmonary artery (PA) smooth muscle cell (PASMC) soluble guanylate cyclase and subsequent vasodilation through activation of a calcium-sensitive K+ channel (6).
Recent data demonstrate that RhoA, a member of the Rho family of small GTPases (16), and its effector protein, Rho kinase (ROCK), play a significant role in regulating pulmonary vascular tone through both NO-dependent and -independent mechanisms (21, 33). In the pulmonary vasculature, hypoxia activates RhoA (21, 32). RhoA activation increases Ca2+ sensitivity of the contractile myofilaments in PASMC in part through inhibition of light chain phosphatase (13, 18, 26), thereby increasing vascular tone.
During fetal life, ROCK activity maintains elevated pulmonary vascular resistance (19). However, the role of ROCK in the context of the transition that the perinatal pulmonary circulation undergoes with the onset of air-breathing life remains unknown. Indeed, whether ROCK activity modulates perinatal pulmonary vascular tone primarily through effects on endothelial cells or on smooth muscle cells is similarly unknown. Existing data suggest that ROCK may have separate effects on PA endothelial cells (PAEC) and PASMC. For example, Fagan et al. (8) demonstrated that ROCK inhibition increases eNOS expression in lungs from pulmonary hypertensive mice, whereas others have demonstrated that ROCK inhibition can cause pulmonary vasodilation in a NO-independent manner (15).
Thus the present line of investigation sought to determine whether ROCK modulates tone in the perinatal pulmonary circulation through effects on the PAEC, PASMC, or both. To determine the separate and interactive role of ROCK on PAEC and PASMC in the perinatal pulmonary circulation, we determined the effect of ROCK inhibition: 1) in monolayers of fetal PAEC; and 2) in a pharmacological model of perinatal pulmonary hypertension. The present data demonstrate that ROCK inhibition has direct effects on fetal PAEC NO production and contractile state and causes pulmonary vasodilation even in the presence of eNOS blockade.
MATERIALS AND METHODS
All procedures and protocols performed in this study conformed with the Guide for the Care and Use of Laboratory Animals [DHEW Publication No. (NIH) 85-23, Revised 1985, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20892] and were approved by the Animal Care and Use Committees of the University of Minnesota (Minneapolis, MN) and Stanford University.
Techniques used for cell isolation and culture have been previously described (14, 20, 30). Late-gestation fetal sheep (139 days; term = 147 days) from ewes with time-dated pregnancies were used in this study. Ewes were fasted for 24 h and sedated with pentobarbital sodium. Fetal lambs were partially delivered through a hysterotomy incision, with the head remaining inside the womb to prevent spontaneous breathing, and intracardiac pentobarbital sodium was administered. After thoracotomy, the lung and heart block was isolated. Distal (≥4th generation) PA were quickly excised and placed in MEM (0.2 mM Ca2+). Low-passage fetal PAEC (<8 passages) were isolated from the vasculature (4th generation) of fetal sheep and cultured on acid-washed glass coverslips (25-mm circle; Fisher Scientific, Pittsburgh, PA). Seeded coverslips were placed in 35-mm culture plates and maintained at 37°C in a humidified mixture of 5% O2-5% CO2.
To assess dynamic changes in intracellular calcium concentration ([Ca2+]i) in individual PAEC, the Ca2+-sensitive fluorophore, fura 2-AM (Molecular Probes), was used. Confluent monolayers of fetal PAEC on 25-mm2 glass coverslips were placed on the stage of an inverted microscope (Nikon Diaphot). Cells were loaded with 10 nM fura 2-AM and 2.5 μg/ml pluronic acid (Molecular Probes) and imaged (11, 20). For each experiment, 8–10 cells were visualized, and ratiometric data were acquired from individual cells.
Fluorescent microscopy was used to evaluate the effect of ROCK inhibition on NO production (30) using diaminofluorescein-FM (DAF-FM; Molecular Probes), a single wavelength dye that increases fluorescence emission intensity concomitant with an increase in NO production. Hypoxic PAEC were prepared as outlined above and loaded with the NO-sensitive dye DAF-FM for 30 min under conditions of hypoxia and then washed with hypoxic buffer to allow for deacetylation. Green fluorescence intensity was measured with excitation at 495 nm and emission at 515 nm. After establishment of a stable baseline of fluorescence emission intensity, oxygen tension was acutely increased and emission intensity continuously monitored for the 30-min study period.
To measure NO production under conditions of sustained normoxia, PAEC were grown in MCDB-131 and 10% fetal calf serum and cultured in collagen-coated 12-well plates in hypoxia. After cells reached confluence, fresh, prewarmed, hypoxic solution was placed on the cells. For media blanks, wells without cells were treated in the same manner. After 24 h, the medium was collected, frozen, and replaced with prewarmed normoxic medium. This supernatant was collected and frozen at 24 h for NO determination by determining total nitrate and nitrite using a fluorometric Nitrate/Nitrite Assay Kit (Cayman Chemical, Ann Arbor, MI). Standards were analyzed in medium, and fluorescence values of media blanks were subtracted. The cells were counted in a Coulter counter (Beckman Coulter, Fullerton, CA). Data were calculated as micromole per liter divided by cell number.
Electrical resistance measurements.
The cellular barrier properties were measured using an electrical cell-substrate impedance sensing system (Applied Biophysics, Troy, NY) (3). In brief, cells were grown on small gold microelectrodes (10−4/cm2) in complete culture medium containing 10% fetal bovine serum and growth factor supplement. Two hours before transendothelial electrical resistance (TER) measurements, the culture medium was changed to a medium containing 2% fetal calf serum. The total electrical resistance was measured dynamically across the monolayer, and the effect of an increase in oxygen tension in the presence or absence of Y-27632 was monitored over time. Increased TER was noted as cells adhered and spread over the microelectrode and was maximal at full confluence, whereas cell retraction, paracellular gap formation, rounding, or loss of adhesion were reflected by a decrease in TER. Resistance was normalized to time 0 and expressed as fold change in resistance over time 0.
Whole animal studies.
Eight mixed-breed pregnant ewes between 137 and 141 days of gestation (term = 147 days) were obtained and used in the present study. The fetuses from two separate ewes were used as time controls, leaving six fetuses available for the experimental protocols. The experimental procedures were as described previously (24). In brief, the ewes were fasted, sedated, and anesthetized with 1% tetracaine hydrochloride by lumbar puncture. The ewes breathed spontaneously during surgery and the study period. A midline uterine incision was made, and the fetal lamb's left forelimb was exteriorized. Catheters were placed into the aorta, superior vena cava, main PA (MPA), and left PA (LPA) as previously described (24, 31). Catheter position was confirmed at autopsy. An ultrasonic flow transducer (T201 flowmeter; Transonic Systems, Ithaca, NY) was placed around the LPA to measure blood flow. After instrumentation, the fetus and uterus were returned to the abdominal cavity. Fetal temperature was maintained at 39°C. The fetus was allowed to recover for a minimum of 1 h before initiation of the study. The umbilical circulation remained intact throughout the study.
Table 1 were obtained preventilation (i.e., after study drug infusion but immediately before ventilation), during ventilation, and during administration of the ROCK antagonist.
NG-nitro-l-arginine (l-NNA; Sigma) was suspended in 4.5 M HCl, and the pH was corrected to 7.4 with sodium hydroxide. Acetylcholine (Sigma) and the pharmacological inhibitor of ROCK, Y-27632 (Sigma), were suspended in sterile water.
After a 60-min recovery period, serial MPA blood gas tensions and pH were monitored at 20-min intervals. l-NNA (10 mg/ml) was continuously infused into the LPA at a rate of 0.1 ml/min for 30 min (total dose = 30 mg) immediately after recovery from surgery. To ensure that endogenous NO release was blocked, acetylcholine was administered via the LPA at 2.5 μg/min for 5 min (total = 12.5 μg). In all cases, a single dose of l-NNA was sufficient to inhibit pulmonary vasodilation. After 20 min of l-NNA infusion, pancuronium bromide (0.5 mg; Gensia Laboratories, Irvine, CA) was administered into the superior vena cava to prevent spontaneous fetal respiration. The head was removed from the uterus, a tracheostomy was performed, and a 4.5-mm endotracheal tube was placed. Ventilation was initiated with a volume ventilator (Siemens Servo 900B) with the following initial settings: rate, 20 breaths/min; tidal volume, 40 ml; positive end-expiratory pressure, 4 cmH2O; inspiratory time, 0.5 s; and fraction of inspired oxygen (FiO2), 1.00. Tidal volume was adjusted to maintain a peak inspiratory pressure of 30 cmH2O. Ventilator rate was adjusted to maintain fetal pH and carbon dioxide tension at preventilation values. Sodium bicarbonate was administered to maintain pH >7.30. After 30 min of ventilation with 1.00 FiO2, ventilation was discontinued, and the animal was allowed to recover for 1 h. Acetylcholine was then administered via the LPA at 2.5 μg/min for 5 min (total 12.5 μg) to ensure that pharmacological blockade of endogenous NO persisted. In the presence of vasodilation, the dose of l-NNA was repeated. Y-27632 (10 mg/ml) was continuously infused into the LPA at a rate of 0.1 ml/min for 10 min (total dose = 10 mg). Based on an anticipated weight for each fetal lamb of 4 kg (31), the dose was designed so that the animals received a total dose of 2.5 mg/kg. This dose is consistent with that used in prior studies in the ovine fetus (19) as well as with doses used in studies of the rat pulmonary circulation (8). Following Y-27632 administration, ventilation was resumed with 1.00 FiO2 and ventilator settings identical to those used in the control period. Ventilation was continued for 30 min with physiological and blood gas measurements obtained at 10-min intervals. At the conclusion of the study, the animals were euthanized with a barbiturate-potassium solution (Beuthanasia; Schering-Plough Animal Health, Kenilworth, NJ), and the catheter position was confirmed.
Throughout, results are presented as means ± SE. Statistical significance was determined by one- or two-way ANOVA as appropriate, followed by Bonferroni posttest analysis. P < 0.05 was taken as the threshold level for statistical significance. Experiments were designed to have a statistical power of ≥90% at a probability level of P < 0.05.
ROCK inhibition does not affect PAEC cytosolic calcium.
To determine the role of ROCK in cytosolic calcium ([Ca2+]i) homeostasis in PAEC, measurements of [Ca2+]i were obtained at baseline for 5 min under hypoxic conditions, followed by superfusion of the ROCK inhibitor Y-27632 onto the PAEC. ROCK inhibition had no effect on PAEC [Ca2+]i (P = not significant). An elevation in oxygen tension caused an increase in [Ca2+]i by 17.2 ± 0.5% (P < 0.01 vs. baseline hypoxia; n = 45 cells, 6 coverslips, 5 animals), however, this increase was not different between PAEC treated with Y-27632 vs. vehicle (data not shown).
ROCK inhibition augments normoxia-induced NO production in PAEC.
To determine the effect of ROCK activity on fetal PAEC NO production, PAEC were loaded with the NO-sensitive fluorescent dye DAF-FM diacetate. Under hypoxic conditions, Y-27632 treatment did not affect baseline PAEC NO production (Fig. 1A; 95 ± 6 to 96 ± 2 nm; P = not significant; n = 71 cells). Concomitant with an acute increase in oxygen tension, membrane fluorescence increased in vehicle-treated cells as early as 15 min and progressively increased through the 30-min study period (Fig. 1A; 94 ± 5 to 122 ± 4 nm; P < 0.001 vs. baseline; n = 76 cells). Blocking ROCK activity in the PAEC augmented this effect, causing a significant increase over baseline hypoxia by 10 min of normoxia and increasing NO production over vehicle-treated cells by ∼20% (96 ± 4 to 145 ± 2 nm at 30 min, with P < 0.01 vs. hypoxia and P < 0.01 vs. control cells; n = 52 cells).
We then examined the effect of sustained normoxia on fetal PAEC NO production. In keeping with the results from the prior experiment, there was no difference in NO production between vehicle- and Y-27632-treated cells under hypoxic conditions (0.110 ± 0.074 vs. 0.132 ± 0.05 μM NO per 105 PAEC). However, exposure to normoxia caused a greater increase in NO in Y-27632-treated cells than vehicle-treated PAEC at 24 h (0.874 ± 0.08 vs. 0.749 ± 0.12 μM; Fig. 1B; P < 0.001 vs. Y-27632 hypoxia and vehicle normoxia).
ROCK inhibition enhances the increase in PAEC barrier function induced by acute normoxia.
During the normal transition from the neonatal to the fetal circulation, the pulmonary endothelial cells flatten and spread, resulting in a decrease in luminal diameter and contributing to lowering of the pulmonary vascular resistance (9). To explore the role of ROCK in modulating fetal PAEC shape, the TER of PAEC monolayers was determined in the presence and absence of the ROCK inhibitor Y-27632. Under baseline hypoxia, TER was not different between vehicle- and Y-27632-treated cells. An acute increase in oxygen tension resulted in a sustained increase in resistance over the 10-h study period, corresponding to cell spreading and enhanced endothelial barrier function (Fig. 2). Similar to the effect observed on NO production in the fetal PAEC, Y-27632 treatment significantly augmented the normoxia-induced increase in endothelial barrier function (P < 0.001 vs. control; n = 4 PAEC monolayers).
ROCK inhibition decreases basal fetal pulmonary vascular tone.
To determine the ability of ROCK to modulate basal fetal pulmonary vascular tone independent of its effect on endothelial cell NO production, the ROCK inhibitor Y-27632 was infused into the LPA of fetal lambs after first blocking eNOS activity with l-NNA pretreatment. Despite the presence of NO blockade, Y-27632 caused an almost threefold increase in pulmonary blood flow (47 ± 6 to 139 ± 7 ml/min; Fig. 3A; P < 0.001 vs. baseline, control) and decreased MPA pressure by 25% from 50 ± 3 to 38 ± 2 mmHg (Fig. 3B; P < 0.01 vs. baseline, control). Within 5 min of the onset of Y-27632 infusion, pulmonary blood flow reached a steady state with no further increase in blood flow before the initiation of mechanical ventilation.
ROCK inhibition enhances ventilation-induced perinatal pulmonary vasodilation even in the presence of NO blockade.
We then explored how inhibition of ROCK signaling would modulate the pulmonary vascular response to oxygenation and ventilation in this pharmacological model of pulmonary hypertension induced by NO blockade. In the control period, ventilation with a high concentration of oxygen neither increased perinatal pulmonary blood flow (Fig. 4A) nor decreased PA pressures (Fig. 4B). In contrast, administration of Y-27632 before ventilation significantly increased pulmonary blood flow, decreased pulmonary pressures, and decreased pulmonary vascular resistance (Fig. 4, A–C). Initiation of ventilation with a high concentration of inspired oxygen in combination with Y-27632 resulted in a progressive and sustained increase in pulmonary blood flow (Fig. 4A; P < 0.01 vs. time 0, control) concomitant with a sustained decrease in PA pressure (Fig. 4B; P < 0.01 vs. time 0, control) and a progressive decrease in left pulmonary vascular resistance during ventilation (Fig. 4C). However, the vasodilatory effect of ROCK inhibition was not selective to the pulmonary circulation, as infusion of Y-27632 also decreased aortic pressure (Fig. 4D), an effect that was sustained throughout the study period. Consistent with these effects of pulmonary blood flow, ventilation in the vehicle-treated animals had no affect on arterial oxygenation during the control period (Table 1). In contrast, arterial oxygen tension in the fetus increased significantly at the end of ventilation in the animals receiving the ROCK inhibitor Y-27632 (Table 1).
The present data demonstrate that ROCK activity modulates perinatal pulmonary vascular tone through distinct but interactive effects on both PAEC and PASMC. We report that ROCK activity modulates the response of fetal PAEC to an acute increase in oxygen tension by increasing NO production and augmenting barrier function. Interestingly, these effects on fetal PAEC form and function occur under normoxic, but not hypoxic, conditions. ROCK inhibition causes fetal pulmonary vasodilation, even in the presence of PAEC NO production blockade, thereby implicating a role for ROCK activation in PASMC as a determinant of the elevated pulmonary vascular tone that characterizes the fetal lung. Moreover, whereas blocking eNOS activity prevents pulmonary vasodilation in response to ventilation and oxygenation (1, 5), ROCK inhibition caused marked pulmonary vasodilation in response to ventilation and oxygenation even in the presence of NO blockade. Interestingly, the degree of perinatal pulmonary vasodilation was similar to that caused by oxygenation and ventilation alone (5, 31). The increase in pulmonary blood flow was mirrored by an increase in arterial oxygen tension.
Whereas the oxygen-induced increase in fetal PAEC NO production has been previously described (30), ROCK inhibition augmented NO production in response to normoxia and had no effect on PAEC NO production under hypoxic conditions (Fig. 1). Thus, in fetal PAEC, ROCK activity might serve to constrain fetal PAEC NO production in normoxic, but not hypoxic, conditions. This result stands in contrast to a previous study wherein ROCK activity was found to mediate the decrease in eNOS expression and NO production that occurs in hypoxia (28). However, in the report of Takemoto et al. (28), endothelial cells were taken from the mature pulmonary circulation and exposed to hypoxia, as opposed to the present results in which the cells derived from the fetal circulation were moved from a hypoxic to a normoxic condition. The divergent results may derive from a developmental difference in the eNOS and ROCK interactions and suggest that other mechanisms predominate in the fetal circulation to constrain NO production under hypoxia. ROCK has been demonstrated previously to inhibit eNOS activity by phosphorylation of eNOS by RhoA, thereby limiting NO production, and that inhibition of ROCK prevents this phosphorylation and augments NO production (27). Under hypoxia, PAEC NO production may be sufficiently constrained by the lack of oxygen, a necessary cofactor for eNOS activation (25), so that ROCK activity does not further constrain PAEC NO production.
In this study, we also demonstrate that acute normoxia increases fetal PAEC spreading and enhances barrier function. With the sole stimulus of an acute increase in oxygen tension, the TER of an endothelial monolayer increases (Fig. 2). These in vitro findings possess biological fidelity with observations in the whole animal as successful transition between fetal and neonatal circulation is characterized by a similar spreading of the PASMC and PAEC, resulting in an increase in luminal diameter and a drop in pulmonary vascular resistance (9). Inhibition of ROCK augments the increase in barrier function to a still greater degree, consistent with prior reports that ROCK activity modulates PAEC cytoskeletal structure (34). However, mirroring the results observed with NO production, under hypoxia, ROCK inhibition has no effect on PAEC TER, suggesting that perhaps the effect of ROCK on endothelial barrier function is NO dependent. Furthermore, this observation buttresses the notion that under conditions of hypoxia, the vasodilator actions of ROCK inhibition result primarily from effects on PASMC as opposed to PAEC.
These findings have significant implications for persistent pulmonary hypertension of the newborn, wherein the PAEC fail to spread normally, barrier function is compromised (34), resistance remains high (10), and the production of vasoactive molecules is altered (2). Recent data indicate that the endothelial dysfunction that characterizes congestive heart disease results from an impairment of calcium signaling and cytoskeletal reorganization rather than impaired NO production. This observation underscores a potential role of the Rho-ROCK interactions in pathological alterations of pulmonary vascular tone (12).
The observation that ROCK inhibition caused pulmonary vasodilation of the fetal circulation even in the presence of NO inhibition suggests that ROCK directly modulates the contractile state of the PASMC independent of effects on the PAEC, consistent with observations in the adult pulmonary circulation (8). Whereas prior studies have demonstrated that ROCK inhibition causes pulmonary vasodilation (17), the present report is the first to demonstrate that ROCK inhibition causes vasodilation in the transitional pulmonary circulation. In contrast to the O2-dependent effect of ROCK inhibition on endothelial cells, the effect of ROCK on the fetal PASMC appears to be O2-independent, as Y-27632 treatment caused pulmonary vasodilation both under hypoxia and on transition to normoxia. Furthermore, the observation that ROCK inhibition causes sustained pulmonary vasodilation despite NO blockade suggests a role for ROCK activity in PASMC both in the maintenance of the normally elevated fetal and in the context of pathologically elevated neonatal pulmonary vascular resistance, where PAEC NO production is limited.
In our study, administration of the ROCK inhibitor decreased systemic pressure, evidence of nonselective vasodilatory effects. This effect was observed despite the fact that the pharmacological inhibitor was infused directly into the LPA, distal to the ductus arteriosus, a strategy designed to achieve relatively high concentrations in the pulmonary circulation yet minimize the drug concentration in the systemic circulation. The strategy employed suggests that intravascular administration of a ROCK inhibitor may have limited utility as a therapeutic tool unless targeted delivery can be achieved with a strategy such as a homing peptide (23). However, recent data indicating that aerosol delivery of ROCK inhibitors effectively induce pulmonary vasodilation (17) suggest that further study of aerosol delivery of ROCK inhibition in the context of neonatal pulmonary hypertension might be worthwhile.
The conclusions of the present line of investigation might be further strengthened in the presence of an additional, but mechanistically similar, pharmacological probe. However, as the present experimental design included whole animal- and cell-based systems, use of an additional pharmacological probe would have imposed significant expense on an already relatively labor- and resource-intensive line of investigation. Moreover, in a prior study in the ovine fetus, the effects of both Y-27632 and fasudil were entirely consistent, suggesting that both pharmacological probes act similarly (19). Indeed, there is considerable precedent in the literature for the specificity of Y-27632 in causing ROCK inhibition (8, 13, 15, 16, 18, 19, 21). Thus, while acknowledging the limits of the present experimental design, we nonetheless conclude that the present findings result not from the idiosyncratic properties of the pharmacological probe used, but from effects directly attributable to ROCK inhibition.
In conclusion, we have demonstrated that ROCK activity modulates perinatal pulmonary vascular tone through effects on both the endothelium and smooth muscle. In specific, in the transitional circulation, ROCK inhibition increases PAEC NO production, enhances barrier function, and affects smooth muscle cell tone to cause NO-independent perinatal pulmonary vasodilation. These results indicate that ROCK activity has effects on perinatal pulmonary vascular tone that transcend a single cell or molecule and suggest that normal transition of the pulmonary circulation requires a decrease in ROCK activity. Moreover, as compromised NO production plays a role in persistent pulmonary hypertension of the newborn, ROCK inhibition might provide an additional therapeutic tool to address persistent pulmonary hypertension of the newborn (PPHN), a disease for which no prevention or cure is currently available.
C. M. Alvira is supported by the American Heart Association Fellow to Faculty Award. D. N. Cornfield is an Established Investigator of the American Heart Association. This work was supported by National Heart, Lung, and Blood Institute Grants RO1-HL-60784 and RO1-HL-70628 (both to D. N. Cornfield).
No conflicts of interest, financial or otherwise, are declared by the author(s).
The findings were presented, in part, at the Society for Pediatric Research Meeting on May 5, 2008.
- Copyright © 2010 the American Physiological Society