In healthy subjects, air movement into and out of the lungs throughout respiration produces estimated pressure variations of 5 to 25 cm H₂O, corresponding to functional residual capacity and total lung capacity, respectively. Airway inflation of the lungs induces mechanical strain and in turn causes either smooth muscle relaxation or contraction, which are mediated for the most part by airway epithelium and mechanosensors, such as slowly and rapidly adapting receptors and C-fiber receptors 12. Excessive airway inflation leads to up-regulation of the genes encoding inflammatory protein expression and mediator secretion by airway cells 3. Mechanical ventilation could enhance alteration of mechanical forces in the lungs of patients with asthma, emphysema or acute respiratory distress syndrome, thereby causing ventilator-associated lung injury. Damage to the epithelial cell lining of the airways and alveoli during high tidal volume ventilation have been extensively investigated but the effects of excessive pulmonary ventilation on airway basal tone and bronchial responsiveness are not well known in humans 2456.

Mechanical strain imposed on airway cells induce a cascade of signaling events, primarily mediated by the macromolecular protein complexes associated with the transmembrane integrins that transduce the external forces from the extracellular matrix to the actin cytoskeleton, resulting in the production of various mediators, cytokines, and growth factors, coupled with gene activation 67. The mechanotransduction induced by integrin activation involves a variety of intracellular-signaling pathways mediated by protein kinases, eicosanoids, nitric oxide synthases (NOS), reactive oxygen species, proinflammatory cytokines and stretch-activated channels 589101112. Furthermore, stretch-induced mechanotransduction is dependent on immediate early gene activation, especially the overexpressed cysteine-rich 61/connective tissue growth factor/nephroblastoma (CCN) family 313. WNT (wingless-type mouse mammary tumor virus integration-site family), a large family of secreted glycoproteins with highly conserved cysteine residues, may also be involved in cytoskeletal reorganization after airway stretching 14. However, the pathways involved in the stretch-induced mechanotransduction in freshly isolated human bronchus remain unknown and results extrapolated from cell-culture systems in either static or under cyclic strain conditions have limitations. Similarly, the myogenic response to stretch is not well understood for intact human bronchi. Our study was conducted to establish an experimental model of static acute mechanical stretch that corresponded to airway hyperinflation in isolated human bronchi to investigate the effect of stretch on airway tone and responsiveness.

The study was approved by our local Ethics Committees (Comité de Protection des Personnes se Prêtant à la Recherche Biomédicale de Versailles, France, and Consejo Superior de Investigaciones Científicas, Madrid, Spain) and patients gave informed consent before scheduled surgery. Bronchi were surgically removed from 48 patients with lung cancer (35 men, 13 women, mean age 64 ± 9 years; all patients were smokers or ex-smokers).

The results of this study demonstrated that stretching human isolated bronchi for only five minutes while using a force that corresponded to an inflation pressure of more than 30 cmH₂O increased basal tone, but did not increase tissue responsiveness to ACh. The stretch-induced increase in basal tone was Rho-kinase dependent and in part bronchial epithelium dependent. Some mediators were also implicated, such as leukotrienes and reactive oxygen species, which may have been released from the epithelium. Interestingly, the five minute stretch resulted in the up-regulation of the WNT-signaling, which is known to interact with the Rho-kinase pathway 25. Activation of these intracellular signaling-pathways could have clinical relevance in mechanically ventilated patients, especially those with obstructive pulmonary diseases. Indeed, lung inflation has been shown to alter the airway smooth muscle functions 26. In normal adults with induced bronchoconstriction, a deep inspiration results in bronchodilatation. For unclear reasons, this is impaired in patients with asthma and chronic obstructive pulmonary disease 27. Moreover, a large body of work suggests that imposed tidal fluctuations of airway smooth muscle length relax isolated airway smooth muscle strips 26.

In mechanically ventilated rabbits, the bronchodilating effect of changes in lung volume is caused by the direct effect of stretch on force generation by the airway smooth muscle 28. Lung inflation strains airway smooth muscle and causes a dynamic modulation of cytoskeletal filament organization such as rearrangement of actin and myosin filaments, detachment and attachment of actin filaments and plaques, and polymerization and depolymerization of these filaments 2629. In oscillatory experiments, it is the very first tidal oscillation that provokes most of any force and stiffness reduction in cells or tissues 29. A single transient stretch may acutely fluidize airway smooth muscle which is normally solid, leading to perturbed equilibrium of myosin binding and alteration of airway smooth muscle mechanical properties (the cell becomes much less stiff and much more viscous) 30. When stretch is stopped, the cell resolidifies but the muscle reshortens to a new biophysical state different from the prior under identical loading conditions 30. Moreover, the relaxing effect persists even if stretch oscillation amplitude is reduced 28. Herein, the absence of hyper-responsiveness following stretch could be explained by the stretch-induced perturbation of myosin binding 2930. In addition, we found epithelial removal failed to enhance responsiveness to ACh after stretch, suggesting excessive mechanical strain also alters the epithelial modulation of airway smooth muscle contraction. However, the hypothesis that force fluctuations on airway smooth muscle can diminish airway tone and responsiveness has been recently challenged by LaPrad and colleagues in healthy intact airway 31, questioning the role of cytoskeleton fluidization on airway responsiveness 32.

Consistent with our results using whole intact human tissue, Hernandez and colleagues showed that bovine airway segments also responded similarly to stretch, but only following pretreatment with selected contractile agonists 20. This strain-induced increase in basal tone could reflect length adaptation of the airway smooth muscle: next a stretch and return, the cytoskeleton fluidizes softening, which is then followed by an increase in attachment of actin filaments and increase in tone 2629. Indeed, basal tone is critically influenced by the length at which the airway smooth muscle is adapted 33. After priming the contractile apparatus induced by pretension or contractile agonists such as ACh, quick stretch or very first force oscillations might cause contractile apparatus rearrangements, with many myosin heads interacting with actin filaments, leading to regeneration of a force exceeding prestretch levels 5203435. The strain-stiffening behavior of human airway smooth-muscle cells observed in response to sustained stretch or reconstituted cross-linked actin gels could also explain in part the increase in basal tone 2936. However, the myogenic response to stretch did not entirely explain the effects of various inhibitors or receptor antagonists on the stretch-induced spontaneous airway-smooth muscle contraction observed previously and herein 912192037.

In addition to direct mechanical injury, lung injury may be also caused by cell-mediated inflammation and soluble mediators such as TNF-α, IL-1β, IL-6, IL-8 and substance P 38394041424344. Cytokine release can be related to the degree of lung stretch and inhibition of cytokine actions decreases stretch-induced injury 4546. However, as suggested by the present study, the mechanisms of stretch-induced injury in intact human bronchi appear to be more complex involving neural, mechanical, and mediator effects. The mechanotransduction involved in the force generation in the immediate response to stretching airways should involve pathways with short activation times, for example, such as mediator release or neurogenic mechanisms or cellular Ca²⁺ influx via membrane depolarization caused by the stimulation of acid-sensing ion and stretch-activated channels 122047. Herein, we showed that the rapid rise in basal tone following stretch of human bronchi was abolished in the presence of a Rho-kinase inhibitor or attenuated after epithelium removal, blocking of the Cys-LT₁ receptor or inhibition of NOS_1,2,3. These results differ from those reported previously when using precontracted tissue from other species 122048. Differences among bovine and rodent species models of airway response to stretch have been reported previously 122048. Therefore, the discrepancies between our results and those from animal models are likely to be due to interspecies differences rather than differences among experimental protocols. For example, in the airways of guinea pigs, basal tone is dependent on COX products and epithelium, and stretch-induced spontaneous contraction is independent of LT 1248. Conversely, the intrinsic tone of human airways results from a balance of contractile LT and, to a lesser extent, histamine and bronchodilating prostanoids, like PGE₂ 154849. Stretch modulates prostanoid synthesis by inhibiting COX-1 and activating COX-2, probably via stimulation of the inducible Ca²⁺-independent NOS₂ and production of reactive oxygen species in airway epithelial cells 48101112. It was also shown that airway epithelial cells could produce LT during inflammatory conditions or mechanical stretch 1050. Moreover, lung epithelial response to stretch appears to be extremely rapid and sensitive 4. In accordance with these observations, our results confirmed the role of bronchial epithelium in enhancing the myogenic response to stretch and in particular the stretch-induced imbalance between prostanoid and LT production in human bronchi. However, mechanical stretch may induce LT release from others cells like bronchial smooth muscle and further investigations would examine stretch-induced prostanoid release in epithelium-denuded bronchial tissues. Taken together, these findings suggest that the immediate rise in basal tone following stretch of human bronchi involves increased epithelial LT production due to activation of Ca²⁺-independent inducible NOS₂. In addition, the Rho-kinase inhibitor Y27632 abolishes the Ca²⁺-independent delayed contraction of airway smooth muscle by modulating COX-2 and NOS₂ activation 122429. In isolated guinea-pig airways, stretch caused delayed contraction attributable to Ca²⁺ influx into the muscle cells mediated by stretch-activated channel stimulation 12. In rats exposed to injurious high airway pressure, substance P, which acts on the NK-1 and transient receptor potential ion channels, is implicated in ventilator-induced lung injury 43. In our study gadolinium and the tachykinin NK₁-, NK₂-, and NK₃-receptor antagonists had no significant effect on either the immediate or sustained phase of spontaneous basal tone provoked by stretch. Therefore, we cannot confirm that stretch-activated channels or substance P are involved in airway response to acute mechanical stress in intact isolated human bronchi.

In rodents, cyclical stretch induced by high tidal volume ventilation quickly increased transcription of several genes that regulate transcription like EGR1, GADD45A or JUN, whereas the induction of TNF mRNA expression was not significantly elevated, consistent with our results 33951. Mechanical stress also provoked vascular distension in rat trachea that may lead to activation of endothelial cells causing expression of P-selectin and endothelin-1 4052. Interestingly, we showed herein that P-selectin ligand gene expression was increased by stretching human bronchi, whereas ET_A- and ET_B-receptor antagonists did not reduce airway response to stretch suggesting that endothelin-1 is not implicated in this phenomenon in human airways. The mechanotransduction leading to mechanical regulation of gene expression in contractile tissues has principally been investigated in bladder smooth muscle cells, where the Rho-kinase pathway seems to play a major role in the stretch-induced modulation of CCN-family genes 135354, but it is not well known in human bronchi. To elucidate the signal transduction pathways that are activated in human bronchi following five minutes of stretch, we used the human signal transduction pathway finder PCR array, which analyses the expression of 84 genes representative of 18 different signal transduction pathways. The results of this assay showed that changes in early gene expression provoked by mechanical stretch involved genes implicated in cell proliferation-apoptosis or migration-adhesion but not cell inflammation in isolated human bronchi. Notably, we observed that stretching up-regulated the mRNA of genes implied in the WNT-signaling pathways. The WNT gene family includes 19 members encoding glycoproteins known as WNTs, which can activate two distinct signaling pathways (canonical and noncanonical), responsible for several cellular processes, including cell movement and polarity, proliferation and differentiation 14255556. Our experiments with the human WNT-signaling pathway finder PCR array confirm that stretch modulates the mRNA levels of different genes of the WNT-signaling pathway, particularly WNT7B and WNT2 expression. WNT7b and WNT2 are expressed in the distal mesenchyme and in airway epithelium, and act via the seven membrane-spanning frizzled Wnt receptor (FZD) cell-surface receptors 1425. In addition, WNT7b activates the canonical but not the noncanonical pathway and exerts autocrine-signaling activity on airway epithelial cells 14. Depending on the cellular context, WNTs stimulate the canonical signaling pathway via FZD receptors and nuclear translocation of β-cathenin, thereby up-regulating genes such as WISP1, MYC, CCND1, CCND3, CSNK1N1, DIXDC1, EP300, FDZ7 or PORCN 1425, or activate the noncanonical signaling pathway through the Fzd receptors and ROR2/RYK coreceptors as suggested by the up-regulation of JUN, FOSL1, and FRAT1 observed herein. It is also known that WNT-signaling might induce cytoskeletal reorganization via up-regulation of MYC, CCND1, CCND3 2557, or via the activation of Rho-mitogen-activated protein kinase pathways elicited by the up-regulation of GADD45A, SLC9A3R1 and TP53 58, as suggested by our results. Nevertheless, further investigations will be needed to confirm the functional role of the stretch-induced changes in mRNA expression, especially WNT.

Collectively, our results indicate that the immediate component of myogenic response to stretch for human bronchi is mediated mostly by epithelial LT release via NOS activation and that the sustained component is dependent on Rho-kinase and WNT-signaling pathways. Moreover, the role of the bronchial epithelium in enhancing the myogenic response to stretch in our study illustrates the airway cells interactions during excessive pulmonary inflation. In this way, freshly human isolated bronchus represents a relevant tool to test the stretch-effects on complex networks of multiple airway cells. From a clinical perspective, our experimental model highlights the response of human airway to acute mechanical stress during excessive pulmonary inflation. Future work based on the results of the present study will be needed to determine the impact of long-term cyclic stretch on human airway mechanics, especially in mechanically ventilated patients with obstructive pulmonary diseases.

• The consequences of mechanical stress during excessive pulmonary ventilation on human bronchi are unknown.

• Airway response to acute mechanical stretch (using a force that corresponded to airway-inflation pressure > 30 cmH₂O) in isolated human bronchi involves LT release, NOS activation and Rho-kinase and WNT-signaling pathways.

• This experimental model could help to appraise the impact of excessive pulmonary ventilation on human airway mechanics, especially in mechanically ventilated patients with obstructive pulmonary diseases.

ACh: acetylcholine; CCN: cysteine-rich 61/connective tissue growth factor/nephroblastoma; COX: cyclooxygenase; ELISA: enzyme-linked immunosorbent assay; FZD: frizzled Wnt receptor; IL: interleukin; L-NAME: L-nitroarginine methyl ester; LT: leukotriene; NOS: nitric oxide synthase; PG: prostaglandin; RT-PCR: reverse transcriptase quantitative polymerase chain reaction; TNF: tumor necrosis factor; WNT: wingless-type mouse mammary tumor virus (MMTV) integration site family.

The authors declare that they have no competing interests.

CF conceived the study, performed functional experiments, interpreted the data and drafted the manuscript. FMP performed RT-PCR experiments, interpreted the data and drafted the manuscript. MLG performed statistical analyses, interpreted the data and revised the manuscript. EN collected the tissue samples, interpreted the data and revised the manuscript. SGD performed ELISA analysis of organ-bath fluid and revised the manuscript. ES collected the tissue samples and revised the manuscript. MLC performed RT-PCR experiments, interpreted the data and revised the manuscript. PD participated in the design and coordination of the study and revised the manuscript. All authors read and approved the final manuscript.