Hypercapnic acidosis attenuates pressure-dependent increase in whole-lung filtration coefficient (Kf).

Hypercapnic acidosis (HCA) has beneficial effects in experimental models of lung injury by attenuating inflammation and decreasing pulmonary edema. However, HCA increases pulmonary vascular pressure that will increase fluid filtration and worsen edema development. To reconcile these disparate effects, we tested the hypothesis that HCA inhibits endothelial mechanotransduction and protects against pressure-dependent increases in the whole lung filtration coefficient (Kf). Isolated perfused rat lung preparation was used to measure whole lung filtration coefficient (Kf) at two levels of left atrial pressure (PLA = 7.5 versus 15 cm H2O) and at low tidal volume (LVt) versus standard tidal volume (STVt) ventilation. The ratio of Kf2/Kf1 was used as the index of whole lung permeability. Double occlusion pressure, pulmonary artery pressure, pulmonary capillary pressures, and zonal characteristics (ZC) were measured to assess effects of HCA on hemodynamics and their relationship to Kf2/Kf1. An increase in PLA2 from 7.5 to 15 cm H2O resulted in a 4.9-fold increase in Kf2/Kf1 during LVt and a 4.8-fold increase during STVt. During LVt, HCA reduced Kf2/Kf1 by 2.7-fold and reduced STVt Kf2/Kf1 by 5.2-fold. Analysis of pulmonary hemodynamics revealed no significant differences in filtration forces in response to HCA. HCA interferes with lung vascular mechanotransduction and prevents pressure-dependent increases in whole lung filtration coefficient. These results contribute to a further understanding of the lung protective effects of HCA.

Background

Low tidal volume ventilation has been shown to improve survival in acute respiratory distress syndrome (ARDS)[1] and is a key component of lung-protective ventilatory strategies. The concomitant increase in CO2 accumulation leads to hypercapnic acidosis (HCA), which has both beneficial and harmful effects depending on the disease state, e.g. attenuating pulmonary edema in mechanical models of lung injury but inhibiting neutrophil function and worsening outcomes in infectious models of lung injury.[2,3] HCA has been studied extensively in non-septic models of mechanical lung injury where high-tidal volume ventilation has been used to induce ventilator-induced lung injury (VILI). A principle mechanism to account for the protective effects of HCA in VILI is the inhibition of NF-kB,[4] a key regulatory protein in many inflammatory pathways. HCA also attenuates the inflammatory activation of p42/44 MAP-kinase, an enzyme that has numerous effects on cytoskeletal and focal adhesion proteins that are directly involved in endothelial permeability.[5,6] We have shown that airway pressure is synergistic with vascular pressure in altering Kf;[7] thus, the ability of HCA to mitigate alveolar injury during VILI prompted us to hypothesize that HCA may also influence endothelial barrier function during mechanical stress.

The acute increase in PaCO2 and resulting decrease in pH have effects on pulmonary hemodynamics that are complex and dependent upon pre-existing vascular tone. In general, HCA increases pulmonary artery pressure (PAP)[8,9] by vasoconstriction of both pre- and post-capillary vessels[10] and stimulates sympathetic activity as evidenced by increases in plasma epinephrine and norepinephrine concentrations.[11] Collectively, these responses increase heart rate, mean pulmonary arterial pressure (mPAP), and mean pulmonary capillary pressure (PPc). Thus, we found it curious that HCA could reduce pulmonary edema in models of lung injury when vascular permeability and hydrostatic pressures were both increased. While HCA may attenuate increases in endothelial permeability induced by inflammatory pathways, the elevated hydrostatic pressure would increase fluid filtration and, therefore, increase edema. One might expect, therefore, that increased filtration induced by elevated hydrostatic pressure would nullify the anti-inflammatory effect(s) of HCA on edema development.

In addition to the simple increase in filtration pressure that promotes edema development, increased hydrostatic pressure activates endothelial mechanotransduction and leads to enhanced endothelial permeability. For example, Czarny and Schnitzer demonstrated that increased pressure and flow activated neutral sphingomyelinase, resulting in the release of ceramide, a potent activator of endothelial nitric oxide synthase.[12,13] As we have previously reported, pressure-dependent nitric oxide production is associated with increased endothelial permeability and acute barrier failure as assessed by the whole lung filtration coefficient[7] and an increase in hydraulic conductivity of lung microvascular endothelial cells.[14] To explain the beneficial effects of HCA on reducing pulmonary edema during experimental lung injury, we tested the hypothesis that HCA inhibits pressure-induced endothelial mechanotransduction.[7]

Methods

Isolated perfused lung preparation (IPL)

All animal experiments were approved by the University of Illinois at Chicago Animal Care Committee, and were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council). Adult male Sprague-Dawley rats (300–400 g) were anesthetized with ketamine/xylazine (90/10 mg/kg). A tracheotomy was performed and the animals were mechanically ventilated with a pressure-controlled ventilator (Kent Scientific, Torrington, CT, USA) at a respiratory rate of 60/min. Low tidal volume (LVt) ventilation with low peak inspiratory pressure (PIP = 6–8 cm H2O) was compared with standard tidal volume (STvt) ventilation with standard peak inspiratory pressure (PIP = 10–13 cm H2O) as described previously.[7] Following sternotomy, the pericardial sac was opened, ligatures were placed around the aorta and pulmonary artery, and heparin (200U) was injected into the right ventricle and allowed to circulate for 2 min. The left atrium was cannulated via a small incision in the left ventricle and the pulmonary artery was cannulated via the right ventricle; lungs were perfused with Krebs-Ringers bicarbonate solution (Sigma Chemical, St. Louis, MO, USA) containing 3% bovine serum albumin (BSA; Fraction V; Proliant, Ankeny, IA, USA).

PAP and left atrial pressures (PLA) were measured continuously via in-line pressure transducers (P-75, Harvard Apparatus, Natick, MA, USA) connected to an A/D board (NI-DAQ USB-6211 National Instruments, Austin, TX, USA). In-line solenoid valves (20PSI, 12V DC, Cole Parmer, Mount Vernon, IL, USA) on both the arterial and venous tubing could be closed simultaneously for measuring double occlusion pressures (PDO). An in-line ultrasonic flow probe (Harvard Apparatus, Natick, MA, USA) on the pulmonary artery cannula recorded flow data in real time. The perfusion rate was 6 mL/min. Lungs were excised and suspended from a force transducer (Radnoti, Minrovia, CA, USA) and lung weight was continuously measured. Lung weight, vascular pressures, and flow were recorded using a custom-written program (LabVIEW 8.2, National Instruments, Austin, TX, USA).

Pressure–volume relationship

Tidal volumes (Vt) in isolated perfused rat lungs during pressure-controlled ventilation were measured as previously described.[7] Briefly, a differential pressure transducer (DP-45, Validyne, Northridge, CA, USA), driven by a carrier frequency bridge amplifier (Type 677, Hugo Sachs Elektronik-Harvard Apparatus) was used to record airflow (mL/s) at a sampling rate of 200 Hz using a custom-written program (LabVIEW 8.5, National Instruments, Austin, TX, USA). Each respiratory cycle was integrated to obtain Vt. A calibration curve was created by applying airflow (air pressure = 1 bar) from negative 16.16 mL/s (vacuum) to positive flow at 16.16 mL/s, through the DP-45 and recording the corresponding voltage output. Flow was regulated with a calibrated flowmeter (Dwyer, model UA1049; accuracy ± 2%) connected to a pressure regulator. The lungs were ventilated with a range of PIPs and positive end-expiratory pressures (PEEPs) to record the resulting VT values. Flow/Vt was recorded for 1 min for each ΔP (ΔP = PIP–PEEP). We used inflation pressures that produced VT identical to previous literature values[15,16] during standard tidal volume (STVt). Low tidal volume ventilation (Low Vt) inflation pressures were chosen to mimic clinically relevant lung-protective strategies. Lung-protective ventilation strategies are typically described in mL/kg; thus, despite using a pressure-controlled ventilator, we present the results as Low Vt and Standard Vt based on the pressure-volume relationship in Fig. 1a.

Fig. 1.

(a) Pressure–volume (P-V) relationship: P-V relationship for ventilation strategies used for low tidal volume group and standard tidal volume group. See “Methods” for details. (b) Experimental protocol. Isolated perfused lungs were suspended from a force transducer and real-time changes in weight were recorded (g) as a function of left atrial pressure (PLA, cm H2O). Isogravimetric (IG) conditions were achieved at basal PLA (2 cm H2O). Double occlusion pressures (PDO) were performed before each increase in PLA and at the end of the experiment. Whole lung filtration coefficient (Kf) was measured at PLA = 7.5 cm H2O and taken to be baseline; Kf was then measured after Step 2 (PLA = 15 cm H2O); the ratio of Kf2/Kf1 was used to derive change in whole lung permeability.

The pressure–volume curve presents ΔP (PIP–PEEP) versus Vt (normalized to body weight [kg]) (Fig. 1a). The low Vt group was ventilated with 6–8 cm H2O providing a Vt per body weight of approximately 3–5 mL/kg. The ST Vt group was ventilated with 10–12 cmH2O yielding Vt per body weight of 6–8 mL/kg.

Calculation of Kf

The rate of change in lung weight (g) between 18–20 min of the left atrial pressure step was divided by PPc, where PPc = (PAP + PLA)/2, to yield mL/min/cm H2O (17). This value was normalized to 100 g of predicted lung weight (PLW) using the equation, PLW = 0.0053* (rat weight) – 0.48.

Zonal characteristic (ZC)

Kf measurements are dependent on vascular surface area available for fluid filtration and differences in vascular recruitment can result in spurious increases in Kf. To determine if HCA altered vascular recruitment, ZCs, i.e. percentage of West Zone 2 and 3 conditions were derived for each test group. ZCs were calculated according to ZC = ΔPAP/ΔPLA as described in the references.[18,19]

Experimental protocols

Protocol 1

LVt: Isolated perfused lungs were ventilated at LVt and kept at isogravimetric pressure for 20 min. The PLA was then increased to 7.5 cmH2O for 20 min. After a second 20-min isogravimetric period, the lungs were again subjected to a PLA step of 7.5 cmH2O (Group C), 15 cmH2O (Group P), or 15 cmH2O + HCA (Group P+HCA). Finally, the lungs were returned to a third isogravimetric condition for 20 min (Fig. 1b). Kf1 was derived during the first PLA step; Kf2 was derived during the second pressure step; PAP and PLA were measured during both pressure steps, and PDOs were measured after each pressure step.

Protocol 2

STVt: Isolated, perfused lungs were ventilated with STVt and subjected to two left atrial pressure steps and three isogravimetric periods as described above. The second left atrial pressure step was 7.5 cmH2O (Group C), 15 cmH2O (Group P), and 15 cmH2O with HCA (Group P+HCA).

Hypercapnic acidosis

HCA was achieved by bubbling CO2 into the perfusate reservoir to establish a pH of 7.10–7.15 that equates to a PCO2 of approximately 65–70 mmHg. pH was monitored throughout the experiment and CO2 infusion rate was adjusted to maintain appropriate pH. Lungs were exposed to HCA starting 10 min into the second isogravimetric period and continuing through the second increase in left atrial pressure. Total time for exposure to HCA was 30 min.

Neutral sphingomyelinase

Neutral sphingomyelinase was inhibited by adding 20 μM GW4869 to the perfusate 10 min into the second isogravimetric period. GW was only tested using STVt at 15 cm H2O (Group P+GW).

Statistics

Data are presented as box plots showing median and interquartile range (IQR) as well as whiskers drawn to the most extreme values that fall within 1.5 times the IQR. Groups were compared using the Kruskal–Wallis rank sum test followed by Dunn’s test using R 3.3.1 with the data.table, dunn.test, ggplot2, and cowplot packages.[20-24] Significant difference between groups was determined at P < 0.05.

Results

Hemodynamics

All lungs were exposed to a PLA = 7.5 cm H2O to determine baseline hemodynamic variables and baseline Kf, before exposure to either high PLA (Group P: PLA = 15 cm H2O) or high PLA + HCA and are presented in Fig. 2. In the LVt group, PAP was slightly higher (by 1.5 cm H2O) in P versus P+HCA (P = 0.042) while there were no differences in Ppa during STVt conditions.

Fig. 2.

Effect of HCA on pulmonary hemodynamic variables. Left column: hemodynamic variables for low tidal volume group. PAP and PPc increased significantly between C vs. P (*P = 0.0014). PPa increased slightly in P + HCA vs. P (**P = 0.042); PPc was significantly increased during HCA vs. P (**P = 0.011). PDO was significantly higher in P + HCA vs. P (*P = 0.003). ZC was significantly increased in C vs. P. No differences in ZC were observed between P vs. P+HCA. Right column: hemodynamic variables for STVt group. PAP and PPc were higher in P vs. C (P = 0.0002 and 0.003, respectively). There were no differences in PAP, PPc, PDO, or ZC for P vs. P + HCA.

During LVt, PPc was slightly increased from 17.50 to 18.30 cm H2O (P = 0.011) but during STVt no differences existed (P = 0.38). The PDO for LVt Group P+HCA versus Group P was 3.4 versus 4.5 cm H2O (P = 0.003); during STVt there was no difference in PDo between Group P versus Group P+HCA (P = 0.432).

Zonal characteristics

To rule out pH-dependent changes on vascular tone, recruitment, and vascular surface area that could affect measurement of Kf, ZC was derived as an indicator of zone II and zone III conditions. During LVt, increasing PLA2 from 7.5 cm H2O to 15 cm H2O (Group C versus Group P) resulted in an increase in ZC to 0.76 ± 0.061 from 0.59 ± 0.066 (P = 2.0e-04). Group P + HCA had no effect on ZC compared with Group P (Group P+HCA 0.73 ± 0.044 versus Group P 0.76 ± 0.061; P = 0.44) (Fig. 2, left column).

During STVt, increasing PLA2 from 7.5 cm H2O to 15 cm H2O increased ZC to 0.78 ± 0.074 from 0.45 ± 0.044 (P = 2.3e-03). Group P + HCA had no effect on ZC compared with Group P (Group P+HCA 0.80 ± 0.053 versus Group P 0.77 ± 0.074, P = 0.23) (Fig. 2, right column). The absence of any differences in ZC between Group P versus Group P + HCA indicates that changes in vascular recruitment and surface area were not responsible for the observed differences in Kf.

Baseline Kf

Each preparation was subjected to an initial pressure step in PLA to 7.5 cmH2O for 20 min. The Kf measured during the initial pressure step was termed Kf1. Each Kf1 served as an individual control to account for the natural variation in baseline permeability between isolated lung preparations; filtration is reported as Kf during the second pressure step, Kf2, normalized to Kf1 (Kf2/Kf1). Kf1 was compared across experimental groups to ensure any differences observed in Kf2/Kf1 could not be attributed to differences in starting conditions. There was no difference in Kf1 across groups at Lvt (P = 0.053) or STVt (P = 0.19; Fig. 3).

Fig. 3.

Baseline Kf1 across groups. All groups had Kf1 measured at PLA = 7.5 cm H2O to assess baseline conditions. Kf1 for all groups during both LVt and STVt are statistically similar indicating that baseline Kf did not affect final Kf2/Kf1.

Increased PLA and Kf2/Kf1

To characterize the influence of increased vascular pressure on Kf, lungs were exposed to a second step increase in PLA2 to 15 cmH2O for 20 min (Group P). During LVt, increasing PLA2 to 15 cm H2O (n = 11) increased Kf2 to 0.14 ± 0.047 from 0.049 + 0.016 mL/min/cmH2O per 100 g wet lung weight and resulted in a Kf2/Kf1 of 3.4 ± 1.8, a 4.9-fold increase compared to the Kf2/Kf1 of 0.69 ± 0.38 for Group C (n = 8) where PLA2 was set to 7.5 cmH2O for 20 min (P = 2.9e-5; Fig. 4).

Fig. 4.

Effect of HCA on Kf2/Kf1. Kf2/Kf1 was derived during control conditions (Group C) vs. high pressure (Group P) vs. high Pressure + HCA (Group P + HCA) during both LVt (left) and STVt (right). In the LVt group, Kf2/Kf1 increased 3.4 (C vs. P). When HCA was induced prior to PLA = 15 cm H2O (Group P + HCA), Kf2/Kf1 was maintained at baseline values; *Group P > Group C, P = 2.9e-05; **Group P > Group P + HCA, P = 1.9e-03. During STVt, Kf2/Kf1 increased by 4.9-fold between Group C vs. Group P. HCA prevented increase in Kf2/Kf1 during high pressure (Group P+HCA): *Group P > Group C, P = 1.3e-03; **Group P > Group P + HCA, P = 5.8e-03.

During STVt, increasing PLA2 to 15 cm H2O (Group P; n = 7) increased Kf2 to 0.41 ± 0.39 from 0.082 ± 0.038 mL/min/cmH2O per 100 g wet lung weight (P = 1.3e-3) and resulted in a Kf2/Kf1 of 6.7 ± 1.6, a 4.8-fold increase compared with the Group C (n = 8) Kf2/Kf1 of 1.4 ± 0.50 (P = 6.9e-4; Fig. 4).

Hypercapnic acidosis

HCA eliminated the effects of increased PLA on Kf2/Kf1 in the LVt group (n = 6). At LVt, HCA reduced Kf2 to 0.089 ± 0.097 from 0.14 ± 0.047 mL/min/cm H2O per 100 g wet lung weight (Group P + HCA versus Group P, P = 0.067), resulting in a Kf2/Kf1 of 1.25 ± 0.95. This value was significantly different from group P (Kf2/Kf1 = 3.4 ± 1.8; P = 1.9e-3) but not significantly different from Group C (Kf2/Kf1 = 0.70 ± 0.38; p = 0.23; Fig. 4), indicating a return to baseline permeability.

At STVt, HCA reduced Kf2 to 0.12 ± 0.042 from 0.41 ± 0.39 mL/min/cm H2O per 100 g wet lung weight (Group P + HCA versus Group P, P = 0.038) and returned the Kf2/Kf1 ratio to 1.3 ± 0.38 (versus Group P, P = 5.8e-4; versus Group C, P = 0.44; Fig. 4).

Neutral sphingomyelinase inhibition

GW4869, a neutral sphingomyelinase inhibitor, was tested at PLA2 = 15 cmH2O in the STvt group (n = 3). GW4869 administration during PLA2 = 15 cmH2O reduced Kf2 to 0.24 ± 0.16 from 0.41 ± 0.36 mL/min/cm H2O per 100 g wet lung weight (versus Group P; P = 0.014), resulting in a Kf2/Kf1 of 2.4 ± 2.5. This was significantly different from the Group P (P = 0.014) and not significantly different from Group C (Kf2/Kf1 of 1.4 ± 0.50; P = 0.29; Fig. 5).

Fig. 5.

Neutral sphingomyelinase inhibition attenuates pressure-mediated increase in Kf2/Kf1. Rat lung Kf was derived during control conditions (Group C, PLA = 7 cm H2O) and after a step increase to PLA = 15 cm H2O (Group P) during STVt conditions. Kf2/Kf1 was significantly increased in Group P vs. Group C (*P = 0.014) but was significantly attenuated by inhibition of neutral sphingomyelinase (Group P + GW, **P = 0.014). Thus, inhibition of neutral sphingomyelinase prevents pressure-dependent increase in at lung Kf.

Discussion

We investigated the effect of HCA on Kf using the isolated perfused rat lung model and investigated the hemodynamic consequences of HCA on pressure-dependent endothelial barrier function. The major findings of the present study are that HCA, within a clinically acceptable range (pH 7.10–7.15): (1) inhibited pressure-induced increase in the whole lung filtration coefficient (Kf) at both low tidal volume ventilation and standard tidal volume ventilation; (2) had minimal effects on PAP and PPc; and (3) had no effect on ZCs that could explain the change in Kf. In addition, we demonstrate that inhibition of neutral sphingomyelinase attenuates pressure-dependent changes in Kf during STVt. Collectively, these data suggest that HCA enhances barrier function in lung vascular endothelium by attenuating pressure-dependent mechanotransduction.

In our analysis of HCA on Kf, it is important to distinguish the direct versus indirect effects of HCA on the pulmonary vascular pressure. In an intact animal, HCA increases pulmonary vasculature pressures by activating sympathetically mediated reflexes[8,11] that shift blood volume from the splanchnic circulation to the pulmonary circulation.[25] The lung vasculature is a low-compliance system, possessing only one-seventh the compliance of the overall systemic circulation,[26,27] meaning that small changes in blood volume distribution create large changes in hydrostatic pressure. In the isolated perfused rat lung preparation, HCA caused minor vasoconstriction leading to a small increase in PAP, presumably via actions on both pre- and post-capillary vessels.[9,10] An increase in pre-capillary resistance would protect the downstream exchange vasculature from elevated pressures and reduce Kf, while an increase in post-capillary resistance would increase capillary pressure and increase Kf. An increase in both pre- and post-capillary resistances would likely offset each other resulting in no change in Kf. We measured PAP, PPc, and PDO after HCA exposure and noted a small (1–2 cm H2O) increase in both PAP, PPc, and PDO (Fig. 2). These increases in pressure during HCA should have increased Kf and not reduce Kf as we observed. Thus, changes in hydrostatic pressure and ZC cannot explain the effect of HCA on Kf.

In the lung vasculature, hydrostatic pressure-dependent mechanotransduction involves several known effector proteins including: heparan sulfate proteoglycans;[7] nitric oxide synthase;[7,14] TRPV4;[28] neutral sphingomyelinase;[12,13] and PECAM.[29] Despite this list of known entities, a unifying model for pressure-sensing and coordinated downstream signaling that modulates permeability has yet to be validated. The influence of hydrostatic pressure on endothelial permeability via heparan sulfate-[7,14] and TRPV4-[28] dependent changes on lung permeability has been described in detail and both likely involve downstream activation of NOS. Increased concentration of intra-cellular nitric oxide is associated with nitrosylation of adherens junction proteins and subsequent disassembly of the adherens junction leading to increased permeability.[30-32]

Czarny et al.[13] were the first to demonstrate that lung endothelial membrane fractions isolated after exposure to increased pressure or flow exhibited increased neutral sphingomyelinase (N-SMase) activity. Follow-up studies by Czarny et al.[12] reported that inhibition of N-SMase prevents lung vascular mechanotransduction via a reduction in endogenous ceramides, while exogenous ceramide mimics the effects of mechanotransduction. High pressure- and flow-induced ceramide release and exogenous ceramide stimulated the phosphorylation of MEK1/2, AKT, eNOS(ser1177) and Src-like kinases. This work, however, did not demonstrate a direct link between changes in neutral sphingomyelinase activity and changes in permeability. The inhibition of N-SMase with GW4869 and the attendant protective effect on pressure-induced whole lung permeability represents direct evidence for the role of N-SMase in lung vascular barrier regulation. Activated N-SMase rapidly generates ceramide, a director activator of Enos;[33,34] thus, it seems highly likely that heparan sulfate proteoglycans, directly or indirectly, activate N-SMAse to release ceramide which in turn activates eNOS. As we have shown in endothelial monolayer studies and the isolated perfused lung preparation, increased NO production is associated with an increase in endothelial permeability by promoting adherens junction disassembly.

Acute increases in hemodynamic forces, either by pressure or flow, stimulate the disassembly of the adherens junction, resulting in increased permeability. Activation of nitric oxide synthase is an ubiquitous response to increases in pressure and shear stress, having been demonstrated in HUVECs,[35] BLMVEC,[14] BAEC,[36] and rat mesenteric vessels in vivo.[37,38] For example, DeMaio et al.[39] demonstrated that an acute increase in hydrostatic pressure (from 0 to 10 cm H2O) was associated with a biphasic response by junctional proteins. Initially, ZO-1 was observed to undergo translocation from the cytoplasm to the adherence junction where it contributed to the “sealing response,” whereby cells adapt to the sudden increase in pressure by creating tighter junctions. However, further increases in hydrostatic pressure activates secondary mechanosensitive signals that stimulate NOS activation, NO production, and junctional disassembly leading to increased hydraulic conductivity and increased Kf in monolayers and whole lungs, respectively.[7,14,36] eNOS-dependent nitration of p190RhoGap-A is a principle mechanism for adherence junction disassembly.[30] In addition, oxidant signaling induces the interaction Gα13 with VE-cadherin at the p120-catenin binding site leading to VE-cadherin removal, internalization, and degradation resulting in hyperpermeability.[32]

Despite the aforementioned discussion of our interpretation of the current data and the extensive literature citations that support the idea of a pressure-eNOS mechanism that regulates barrier function, one publication has reported that NO preserves barrier function during elevated hydrostatic pressure. Yin et al.[40] reported that inhibition of eNOS using LNAME increased rat lung Kf and resulted in worse edema formation. We have reviewed the difference in experimental approach of Yin et al. in our paper and Dull et al.[7] Briefly, a major difference that may explain the results of Yin et al. is the role of cyclic ventilation in maintaining endothelial cGMP levels. Yin et al. used a model where the lungs were held in a statically inflated state and not subjected to cyclic mechanical ventilation. Lung cGMP levels are positively influenced by mechanical ventilation and non-ventilated lungs show a significant reduction in cGMP levels.[41-43] Kuebler,[44] in “The Janus-faced regulation of endothelial permeability by cGMP,” reviewed the disparate effects of cGMP on endothelial barrier function, where cGMP can be either barrier-protection or barrier-disruptive. As Kuebler presented in this review, intra-cellular concentrations of cyclic nucleotides, the ratios of cGMP/cAMP, sub-cellular compartmentalization of cyclic nucleotides, and the activities of PDE2A and PDE3, will determine the ultimate effects of cGMP on endothelial barrier properties and Kf. We assume that the use of cyclic ventilation in our model and the prevailing hemodynamic forces within the pulmonary endothelium maintained cyclic nucleotide concentrations such that NO produced barrier-disrupting processes.

Clinical relevance

Lung vascular mechanotransduction is most likely to be operational during states of acute heart failure (HF) where pulmonary vascular pressure increases rapidly. This can occur as a result of left ventricular dysfunction or sympathetically induced blood volume shifts from the periphery to the central compartment. The hemodynamic changes required to activate mechanotransduction and lead to barrier failure are quite modest. An increase in PPc of 8 cmH2O for 20 min resulted in a fivefold increase in Kf[7], while Kuebler et al.[45] showed that an increase in ΔPLA = 10 cm H2O resulted in a 4.6-fold increase in lung NO-dependent fluorescence.

Limitations

The main limitation in these studies was the use of a buffer-perfused ex vivo lung preparation that is devoid of circulating blood cells. Red blood cells may have effects on endothelial cell permeability during control, pressure, and HCA conditions that were not accounted for. These studies demonstrate that HCA can prevent pressure-induced increase in Kf but we did not assess if HCA can reverse the effect of pressure on whole lung permeability once it has increased. Such studies are the topic of future investigations.

Conclusion

We have demonstrated that HCA does not affect pulmonary hemodynamics in a manner that could account for the observed reduction in Kf and suggests that HCA inhibits endothelial mechanotransduction. We have previously demonstrated that an acute increase in hydrostatic pressure increases Kf via a heparan sulfate-nitric oxide dependent pathway[7] and we now extend this mechanism to show that inhibition of neutral sphingomyelinase activity results in barrier protection. This places neutral sphingomyelinase between heparan sulfate proteoglycans and eNOS in the mechanotransduction pathway. Additional studies are underway to elucidate the effects of HCA on heparan sulfate proteoglycan-dependent signaling and on neutral sphingomyelinase to achieve a more detailed understanding of endothelial mechanotransduction as it relates to barrier function.

Availability of data

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.