Mechanical Stretch Alters Alveolar Type II Cell ... - ATS Journals

Mechanical Stretch Alters Alveolar Type II Cell Mediator Release toward a Proinflammatory Pattern Stefan Hammerschmidt*, Hartmut Kuhn*, Ulrich Sack, Anke Schlenska, Christian Gessner, Adrian Gillissen, and Hubert Wirtz Department of Respiratory Medicine and Critical Care, and Department of Immunology, University of Leipzig; and Robert-Koch-Klinik, Leipzig, Germany

Increased mechanical stretch of alveolar type II (ATII) cells occurs during mechanical ventilation. The effects of three patterns of stretching rat ATII cells (frequency [min-1]-⌬surface area [%]: S40-13, S60-13, S40-30) were compared with those in static cultures at 12, 18, and 24 h. Cell viability and expression of cyclooxygenase-2, 5-lipoxygenase, inducible nitric oxide synthase (iNOS), and endothelial nitric oxide synthase (eNOS) were characterized. Supernatants were analyzed for eicosanoids, nitrite, cytokines, and stimulatory effects on rat lymphocytes. S40-13 simulates normal breathing; the other patterns increased amplitude and frequency. There were no significant differences between S40-13 and static cultures. S60-13 only significantly increased the supernatant nitrite (11.2 ⫾ 1.6 versus 3.9 ⫾ 0.4 ␮M at 24 h). S40-30 significantly reduced the number of trypan blue–excluding cells, increased the supernatant concentration of TXB2 (4.1 ⫾ 0.61 versus 2.2 ⫾ 0.36 pg/ml), 6-keto-PGF1␣ (8.7 ⫾ 1.0 versus 6.7 ⫾ 0.52 pg/ml), cysteinyl-LT (12.2 ⫾ 2.0 versus 6.1 ⫾ 0.75 pg/ml) and nitrite (7.2 ⫾ 1.7 versus 3.9 ⫾ 0.4 ␮M). S40-30 did not alter the release of tumor necrosis factor-␣ and monocyte chemotactic protein-1, but significantly reduced the concentration of the anti-inflammatory interleukin-10 (20.8 ⫾ 13.3 versus 130 ⫾ 21.5 pg/ml). Expression of cyclooxygenase-2/5-lipoxygenase was increased/decreased; expression of iNOS/eNOS was unchanged by high-amplitude stretch. Supernatants from S40-30 experiments caused lymphocyte activation measured by CD71 and CD54 surface expression. Continuing mechanical distension of ATII cells contributes to an inflammatory response by a shift in the balance of proand anti-inflammatory mediators. Keywords: alveolar type II cell; mechanical stretch; mediator release

Low in comparison with high tidal volume ventilation has been shown to improve the prognosis of patients with acute respiratory distress syndrome (1). Clinical data (2) and experimental studies using animal models (3) or isolated organs (4, 5) provided evidence that high tidal volume ventilation induces the release of proinflammatory mediators into the alveolar and/or vascular compartment, such as cytokines (tumor necrosis factor [TNF]-␣, interleukin [IL]-1␤, IL-6, macrophage inflammatory protein [MIP]-2, monocytes chemotactic protein [MCP]-1), eicosanoids (prostacyclin, 6-keto-PGF1␣, TXB2, PAF), and nitric oxide (NO). The alveolar epithelium is subjected to increased distension during high tidal volume ventilation and, therefore, may be involved in the pulmonary response to high tidal volume ventilation.

The alveolar epithelial A549 cells have been shown to constitutively secrete cytokines (MCP-1 and transforming growth factor [TGF]-␤) and eicosanoids (LTB4) (6). In response to humoral inflammatory stimuli, e.g., IL-1␤, TNF-␣, or bradykinin, alveolar epithelial type II (ATII) cells or A549 cells overexpress IL-8 (7) and release IL-6, IL-8, G-CSF, interferon, MCP-1, and TGF-␤ (7–9). Eicosanoids, such as 6-keto-PGF1␣, TXB2, and LTC4 are also released by alveolar epithelial cells in response to injurious stimuli (10). Alveolar epithelial cells evoke chemotactic activity by the release of LTB4 (11). Cytokines may also perpetuate the inflammatory response by the induction of cyclooxygenase (COX)-2, the inducible COX isoenzyme in ATII cells (12). Mechanical stretch plays an important role in the regulation of ATII cell functions. Mechanical stimuli associated with physiologic ventilation have been shown to stimulate ATII cell calcium signaling followed by surfactant phospholipid secretion (13) and surfactant protein expression (14). Mechanical stretch has also been reported to cause adverse effects. It induces apoptosis (15, 16) and cell membrane stress failure (17) resulting in cell death (18). Cyclic mechanical stretch has also been demonstrated to upregulate IL-8 and TGF-␤ mRNA expression (19) and to induce release of IL-8, but not that of GM-CSF or TNF-␣, in A549 cells (20). Although the involvement of prostanoids in clinical (21) and experimental settings of acute lung injury has been well established, the effect of mechanical stretch on ATII cell eicosanoid metabolism has not yet been investigated. We hypothesized that cyclic mechanical stretch acts as an injurious stimulus in ATII cells altering the pattern of ATII cell mediator release toward a proinflammatory pattern. In the present study we used defined stretching patterns to characterize the influence of stretch frequency, amplitude, and duration (1 ) on the secretion of typical eicosanoids and gene expression of key enzymes of the eicosanoid metabolism (COX and 5-lipoxygenase [5-LO]), (2 ) on the release of cytokines, and (3 ) on the synthesis of NO and the gene expression of inducible and endothelial nitric oxide synthases (iNOS, eNOS). Cysteinyl-LT, and the stable products of prostacyclin and TXA2, were investigated as prominent metabolites of eicosanoid metabolism. Two proinflammatory cytokines, TNF-␣ and MCP-1, as well as the anti-inflammatory cytokine IL-10, were measured. NO synthesis was assessed by the measurement of nitrite. The proinflammatory activity of the supernatants of stretched cells was functionally demonstrated by its stimulatory activity on rat lymphocytes.

(Received in original form February 15, 2005 and in final form May 6, 2005)

MATERIALS AND METHODS

This work was supported by a grant from Deutsche Forschungsgemeinschaft (Ha 3263/1-1).

ATII Cell Preparation

* Both authors contributed equally to the study. Correspondence and requests for reprints should be addressed to PD Dr. Stefan Hammerschmidt, Medizinische Universita¨tsklinik I, Pneumologie, Universita¨t Leipzig, Johannisallee 32, 04103 Leipzig, Germany. E-mail: stefan.hammerschmidt@ t-online.de Am J Respir Cell Mol Biol Vol 33. pp 203–210, 2005 Originally Published in Press as DOI: 10.1165/rcmb.2005-0067OC on June 9, 2005 Internet address: www.atsjournals.org

ATII cells were isolated as described previously (13, 16). In brief: ATII cells were isolated from male Sprague Dawley rats (150–200 g) by elastase digestion and differential adherence on IgG-coated dishes. ATII cells were 88 ⫾ 4.1% pure at the time of plating, as proven by modified Papanicolaou staining. ATII cells were placed on the central area (ⵑ 1.5 cm diameter) of fibronectin-coated silicon membranes (Bioflex, coated additionally with 150 ␮M bovine fibronectin for at least 3 h at 4⬚C; Flexcell International,

204

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 33 2005

Hillsborough, NC) of 6-well plates (106 per well with Dulbecco’s modified Eagle’s medium with 10% fetal calf serum [FCS], 1% wt/vol gentamicin, and 1% glutamine). After 22 h of adherence, medium was replaced by fresh medium containing 2% FCS. These plates were used for experiments.

Experimental Protocol ATII cells on flexible membranes were exposed to cyclic stretch using the FX 4000T Flexercell Tension Plus system (Flexcell International). The cells were randomly subjected to one of three cyclic stretching patterns. The stretching pattern was applied over the entire experimental time. Cells and supernatants were collected at 0 h (immediately before starting the stretching pattern) and at 12, 18, and 24 h. Stretching patterns were defined by frequency and amplitude. Amplitude was defined by the increase in surface area of the calibrated silicon membranes. Membrane distension was calibrated with negative pressure for each instrument and type of membrane and monitored during the experiment. The numbers shown represent the extent of membrane distension as constantly indicated and calculated directly from negative pressure readings during each pressure swing on the monitor within the Flexercell software program. ATII cells, on identical fibronectincoated silicon membranes but not subjected to cyclic stretch, served as controls (static culture).

Stretching Pattern We used stretching patterns that were reported in detail previously (16). They were designed to evaluate the influence of stretch frequency and stretch amplitude on eicosanoid, cytokine, and NO release in cultured isolated rat ATII cells. Increases in epithelial basement membrane surface area have been reported to occur predominantly with lung inflations (residual together with tidal volume) exceeding 40% of total lung capacity (TLC) in the rat (22). Volume changes below this level presumably involve unfolding and collapse rather than stretch and relaxation. We therefore chose changes in epithelial basement membrane surface area of 13% and 30% relating to inflations of ⵑ 75 and 100% TLC (with 40% TLC, i.e., functional residual capacity, as the point of initiation of real tissue distension; equation in Figure 3A of Ref. 22). Thus the amplitudes chosen ranged from tidal volume to approximately a maximum inspiration. To characterize the influence of frequency we chose both 40 and 60 min-1. Due to technical limitations the combination of 30% distension and a frequency of 60/min was not possible. In summary, we used three patterns of cyclic stretch [S] characterized by frequency [min-1] and amplitude [%]: S40-13, S60-13, and S40-30.

Lactic Acid Dehydrogenase Release Supernatants of stretched cells and controls were analyzed for lactic acid dehydrogenase (LDH) activity. LDH activity was measured by use of a Cytotoxicity Detection Kit (Roche Diagnostics GmbH, Mannheim, Germany). The kit was calibrated using LDH standards (Sigma-Aldrich GmbH, Deisenhofen, Germany) between 0.016 and 2 ␮g/ml. Each data point represents n ⫽ 4 cell isolations with measurements in triplicate.

Trypan Blue Staining: Changes in Trypan Blue Exclusion Trypsin-EDTA (Invitrogen, Paisley, Scotland, UK) was used to harvest the cells for analysis. Cells were stained with trypan blue and the number of trypan blue excluding cells was counted. Both adherent and nonadherent cells were included in the analysis. The determination of trypan blue exclusion was performed at 0, 12, 18, and 24 h in static controls and at 12, 18, and 24 h in stretched cell. Each data point represents n ⫽ 4 cell isolations with measurements in triplicate.

Nitrite in the Supernatant Griess reagent (400 ␮l) was added to 400 ␮l supernatant. The mixture was incubated for 20 min at room temperature. Extinction was read at 550 nm. Sodium nitrite was used as a standard LV5P group that was similar to that of the HVZP group (5). Each data point represents n ⫽ 4 cell isolations with measurements in duplicate.

Cytokine and Eicosanoid Concentrations in the Supernatant Rat MCP-1 BD OptEIA-ELISA, rat TNF-␣ BD OptEIA-ELISA, and rat IL-10 OptEIA-ELISA were purchased from BD Biosience Pharmigen (San Diego, CA). The ELISA tests were performed according to

the instructions of the manufacturer. Standards were run from 15.6– 1,000 pg/ml (IL-10) or 31.2–2,000 pg/ml (TNF-␣, MCP-1), respectively. Samples for MCP-1 measurements were diluted 50-fold before use in the assay. Cysteinyl-Leuktriene EIA Kit, Thromboxane B2 EIA Kit, and 6-keto Prostaglandine F1␣ EIA Kit were purchased from Cayman Chemical (Ann Arbor, MI). The standard curves were measured between 3.9 and 500 pg/ml (TXB2 and 6-keto PGF1␣) and between 7.8 and 1,000 pg/ml (cysteinyl-LT). Each data point represents n ⫽ 4 cell isolations with measurements in duplicate.

Quantitative Polymerase Chain Reaction for COX-2, 5-LO, iNOS, and eNOS Total RNA was isolated from harvested cells using the RNeasy Mini Kit (Quiagen, Hilden, Germany). One microgram of RNA was incubated with DNase (GIBCO Invitrogen, Karlsruhe, Germany) at room temperature (RT) for 15 min. DNase purified RNA was reverse transcribed at 42 ⬚C for 30 min (Reverse Transcription System; Promega, Mannheim, Germany). Primers, annealing temperatures, and number of PCR cycles of COX-2, 5-LO, iNOS, eNOS, and GAPDH (as control) are shown in Table 1. PCR was performed in a final volume of 20 ␮l, with 1 ␮l cDNA, 10 ␮l iQ SYBR Green Supermix (Bio-Rad, Mu¨nchen, Germany), and 9 ␮l H2O. PCR was performed with an initial denaturation at 94 ⬚C for 3 min, cycling times of 30 s denaturing at 94 ⬚C, 30 s annealing, and 30 s extension at 72 ⬚C and final extension at 72 ⬚C for 7 min. PCR products were detected during PCR with the MyiQ Single-Color Real-Time PCR Detection System (Bio-Rad). Real-time PCR data were automatically calculated with the data analysis module. The results were analyzed according to the ⌬⌬Ct method (23). Real-time PCR was performed in triple runs with cDNA from n ⫽ 4 experiments (i.e., 4 rats). The difference in the number of cycles (⌬Ct) between the gene of interest and GAPDH was used for statistical comparisons among groups. If differences in ⌬Ct between groups were significant, the difference between these values (⌬⌬Ct) was used to quantify the differences in gene expression.

Rat Lymphocyte Activation Lymphocytes were isolated from heparinized rat blood by density gradient centrifugation (Ficoll Serva, Heidelberg, Germany). After washing trice, lymphocytes were seeded in 6-well plates (1 ⫻ 106 cells/well) and incubated with 50% RPMI 1640 supplemented with 10% FCS and 50% supernatant of static or stretched cell cultures. Cells incubated in 100% supplemented RPMI 1640 or cells stimulated with phytohemagglutinin (PHA; Sigma) were used as controls. At 48 h cells were harvested, immunostained with FITC-labeled anti–rat-CD54 or anti–rat-CD71 antibody (Beckman Coulter, Krefeld, Germany) for 30 min at RT. The lymphocytes were analyzed for expression of CD54 or CD71 on the surface by flow cytometry. The experiment was performed with n ⫽ 7 (CD54) and n ⫽ 10 (CD71) cell isolations with single measurements.

Statistics The differences in viable cells, LDH release, and mediator release among the groups were first analyzed by ANOVA with respect to time and stretching protocol (static, S40-13, S60-13, or S40-30). If there was a significant different among the groups, post hoc Bonferroni analysis was performed to detect significant differences between individual groups. The ⌬Ct values between GAPDH and gene of interest were used for statistical analysis of RT-PCR data. The differences in these ⌬Ct values among groups were first analyzed by ANOVA with respect to time and stretching protocol. If there was a significant different among the groups, post hoc Bonferroni analysis was performed to detect significant differences between static and stretched samples. Paired Student’s t test was used to compare the surface expression of lymphocyte activation markers (CD54 and CD17) between static controls and S40-30. SPPS 11.0 for Windows (SPPS Inc., Chicago, IL) software was used. Results are expressed as mean ⫾ SD. Statistical significance was assumed if P ⬍ 0.05.

Hammerschmidt, Kuhn, Sack, et al.: Stretch-Induced ATII Cell Mediator Release

205

TABLE 1. PRIMER, ANNEALING TEMPERATURE AND PCR PRODUCT (bp) OF THE GENES ANALYZED Gene

Primer

Annealing

PCR Product (bp)

Cycle

eNOS

gct ctt tcg gaa ggc gtt tga aca gaa gtg cgg gta tgc tc cga cac cct tca cca ca ttc ctc tat ttt tgc ctc ttt agg ctg cag tga gaa gca tc gcc agt ggt tct tga ctc tc gga tga cga gcg act gtt c caa tgc ggt tct gat act gga gtc atc atc tcc gcc cct tcc Ccg cct gct tca cca cct tct

59 ⬚C

364

45

54 ⬚C

193

40

56 ⬚C

591

35

56 ⬚C

164

35

60 ⬚C

442

23

iNOS 5-LO COX-2 GAPDH

Definition of abbreviations: 5-LO, 5-lipoxygenase; COX-2, cyclooxygenase-2; eNOS, endothelial nitric oxide synthase; GAPDH, glyceraldehydes phosphate dehydrogenase; iNOS, inducible nitric oxide synthase.

RESULTS Trypan Blue Exclusion and LDH Release

A decrease was observed in the number of trypan blue–excluding cells in static controls and in all stretch-groups over time (Figures 1A–1C). The number of trypan blue–excluding cells did significantly differ with respect to stretching pattern and time. Post hoc comparison revealed significant differences between S40-30 and each of the other groups. LDH activity in the supernatants of all experimental groups is summarized in Figures 1D–1F. LDH activity in supernatants of static cell cultures tended to increase with time to a maximum of 0.048 ⫾ 0.010 ␮g/ml at 24 h. The supernatant LDH activity did significantly differ with respect of stretching pattern and

time. Post hoc analyis revealed that LDH activities of the S40-30 (maximum 0.165 ⫾ 0.048 ␮g/ml at 24 h) were significantly different in comparison with all other groups, whereas LDH activities of the S60-13 (maximum 0.092 ⫾ 0.014 ␮g/ml at 24 h) were significantly different from those of static controls and the S40-30 but not S40-13 pattern. The LDH activities of the S40-13 group did not differ from those of static controls (maximum 0.061 ⫾ 0.011 ␮g/ml at 24 h). Cyclic stretching with the larger amplitude (S40-30) resulted in the greatest LDH release. Also, the S60-13 pattern induced significant LDH release in comparison with static controls. The S40-13 stretching pattern, however, did not induce LDH release exceeding that of static controls. NO Release

The time course of nitrite concentrations in the supernatant of cells stretched with the different stretch patterns was compared with that in static controls (Figure 2). There was a continuous increase in supernatant nitrite over time in controls reaching a maximum of 3.9 ⫾ 0.4 ␮M at 24 h. The time course of supernatant nitrite concentration did significantly differ among all groups. Both increased amplitude (S40-30) and increased frequency (S60-13) resulted in increased nitrite concentrations at 18 h (5.3 ⫾ 0.7 and 6.8 ⫾ 1.6 ␮M) and 24 h (7.2 ⫾ 1.7 and 11.2 ⫾ 1.6 ␮M). Post hoc analysis showed that the nitrite concentrations of the S60-13 group did significantly differ from those of static controls

Figure 1. Trypan blue–excluding cells and LDH. The number of trypan blue–excluding cells at the start of the experimental period, at 12, 18, and 24 h, is presented in A (group S40-13), B (group S60-13), and C (group S40-30). Static controls are depicted as open squares, experiments with cyclic stretch as closed symbols. LDH activity in supernatants of cells subjected to cyclic stretching (closed symbols) and cells in static culture (open squares) is shown in D (group S40-13), E (group S60-13), and F (group S40-30). The differences among the groups were significant as proven by ANOVA (*P ⬍ 0.05 versus all other groups and †P ⬍ 0.05 versus control as proven by post hoc Bonferroni analysis).

Figure 2. NO release. NO release is measured as supernatant nitrite concentration. Nitrite concentration in supernatants of cells subjected to cyclic stretching (closed symbols) and cells in static culture (open squares) are shown in A (group S40-13), B (group S60-13), and C (group S40-30). The differences among the groups were significant as proven by ANOVA (*P ⬍ 0.05 versus control and versus S40-13, and † P ⬍ 0.05 versus control as proven by post hoc Bonferroni analysis).

206


and from those of the S40-13 group. Nitrite concentrations of the S40-30 were only significantly different from those of static controls. The differences between S60-13 and S40-30 did not reach the level of significance. Eicosanoid Release

The concentration of TXB2 (the stable product of TXA2), 6-ketoPGF1␣ (the stable product of prostacyclin) and cysteinyl-LT was determined in supernatants of all experimental groups at baseline (t ⫽ 0) and at 12, 18, and 24 h (Figure 3). The concentration of the three eicosanoids measured from supernatants of static cell cultures increased slightly with time to a maximum mean of 2.2 ⫾ 0.36 pg/ml for TXB2, 6.7 ⫾ 0.52 pg/ml for 6-ketoPGF1␣, and 6.1 ⫾ 0.75 pg/ml for cysteinyl-LT at 24 h. The supernatant eicosanoid concentrations were significantly different among all groups as proven by ANOVA for the three eicosanoids analyzed. Post hoc analysis revealed significant differences between static controls and the high-amplitude stretching pattern (S40-30). The maximum concentrations in the supernatant were 4.1 ⫾ 0.61 pg/ml TXB2, 8.7 ⫾ 1.0 pg/ml 6-keto-PGF1␣, and 12.2 ⫾ 2.0 pg/ml cysteinyl-LT. This corresponds with an almost 2-fold increase in TXB2 and cysteinyl-LT, and an increase by ⵑ 40% in the concentration of 6-keto PGF1␣.

ternal standard. Gene expression of COX-2, 5-LO, iNOS, and eNOS was performed for static controls and for those stretching patterns that induced a significant difference of the corresponding mediator with static controls. That means COX-2 and 5-LO gene expression was analyzed for static controls and for the S40-30 pattern (Table 2); iNOS and eNOS gene expression was analyzed for static controls, S60-13 and S40-30 pattern (Table 3). Gene expression was analyzed after the complete experimental time of 24 h and to detected early changes at 6 h. COX-2 expression (Table 2) in static and S40-30 samples at 6 h and at 24 h was significantly different as demonstrated by ANOVA. Post hoc analysis demonstrated significant differences between static controls and S40-30 samples at 6 h and 24 h (increased 2.23 and 3.4 times). The expression of 5-LO (Table 2) in S40-30 and static samples after 6 h and after 24 h was also significantly different as demonstrated by ANOVA. Post hoc analysis revealed a decreased 5-LO expression in response to the S40-30 stretching pattern after 24 h (0.4 fold of the static control) but no change after 6 h. Increased amplitude (S40-30) and increased frequency (S60-13) stretch did not induce significant changes in the expression of eNOS and iNOS as demonstrated by ANOVA analysis (Table 3).

Gene Expression of COX-2, 5-LO, iNOS, and eNOS

Cytokine Release

Expression of COX-2, 5-LO, iNOS, and eNOS genes was investigated using quantitative real-time PCR and GAPDH as an in-

Culture supernatants were analyzed for cytokine concentrations (Figure 4). MCP-1 concentrations exhibited a comparable increase with time in all experimental groups. The ANOVA did not indicate significant differences among the groups. TNF-␣ concentrations remained almost stable at all time points and in all experimental groups and were not different among groups as indicated by ANOVA. IL-10 concentrations, however, increased in the static controls, and in the S40-13 and the S60-13 groups, but evidently failed to increase in the high-amplitude stretch group (S4-30). This resulted in great differences between static controls and the S40-30 group values at 18 and 24 h. IL-10 concentration in static culture supernatants (130 ⫾ 21.5 pg/ml) were increased more than 6-fold over concentrations in supernatants of the S40-30 (20.8 ⫾ 13.3 pg/ml) stretched cultures. The differences in supernatant IL-10 concentrations were proven to be significant by ANOVA with respect to time and stretching pattern. Post hoc analysis revealed a significant difference between static controls and the S40-30 stretching pattern. Rat Lymphocyte Activation

The effect of supernatant collected from cells stretched with high amplitude (S40-30 pattern) on rat lymphocyte expression of activation markers CD71 and CD54 is demonstrated in Figure 5. The intensity of staining of both CD71 and CD54 was increased at 48 h after incubation with supernatant from stretched cells in comparison with that from static cells. PHA added to control lymphocyte cultures resulted in increased expression of CD71 and CD54, whereas supplemented culture medium did not alter the expression of these lymphocyte activation markers (data not shown).

DISCUSSION

Figure 3. Eicosanoid release. The concentrations of TXB2, 6-keto PGF1␣ and cys-LT in supernatants of cells subjected to cyclic stretching (closed symbols) and cells in static culture (open squares) are presented. A, D, and G show results of group S40-13; B, E, and H those of group S60-13; and C, F, and I those of group S40-30. The differences among the groups were significant as proven by ANOVA (*P ⬍ 0.05 versus control as proven by post hoc Bonferroni analysis).

This study investigates the influence of different stretching patterns on the release of NO, eicosanoids, and cytokines, as well as iNOS/eNOS, COX-2, and 5-LO gene expression in rat ATII cells. The stretching patterns were first chosen to reflect the mechanical stimulation associated with normal breathing. Resting breathing frequency in the rat is ⵑ 40/min. The mechanical distension associated with quiet breathing translates into an ⵑ 13% increase in alveolar surface area (S40-13; 22). Additional stretching patterns were chosen to reflect the influence of in-


207

TABLE 2. RESULTS OF QUANTITATIVE RT-PCR FOR COX-2 AND 5-LO S40-30 versus Static Control Time 6h 24 h

Static Control ⌬Ct versus GAPDH

Gene COX-2 5-LO COX-2 5-LO

6.66 12.02 7.94 13.00

⫾ ⫾ ⫾ ⫾

0.81 1.01 0.34 1.13

S40-30 ⌬Ct versus GAPDH 5.50 11.81 6.17 14.36

⫾ ⫾ ⫾ ⫾

⌬⌬Ct versus Static ⫺1.16 ⫺0.21 ⫺1.77 1.36

1.52* 0.62 0.51* 0.30*

⫾ ⫾ ⫾ ⫾

Change (fold)

0.99 0.40 0.23 1.05

2.23 1.16 3.40 0.39

Definition of abbreviations: 5-LO, 5-lipoxygenase; COX-2, cyclooxygenase-2; GAPDH, glyceraldehyde phosphate dehydrogenase. Gene expression of COX-2 and 5-LO was analyzed at 6 h and 24 h for static controls and for those stretching patterns that induced a significant difference of supernatant eicosanoid concentrations with static controls (S40-30). The difference in the number of cycles between the gene of interest and GAPDH (⌬Ct versus GAPDH) is compared between static controls and S40-30. The difference between the ⌬Ct of a stretching experiment and the ⌬Ct of the static control is used to calculate the ⌬⌬Ct and the corresponding change in gene expression. Bold type indicates that the differences between the ⌬Ct were significant. The differences among the groups were significant as proven by ANOVA. * P ⬍ 0.05 versus control as proven by post hoc Bonferroni analysis.

creased frequency (S60-13) and increased amplitude (S40-30) in a fashion similar to that reported previously (16). The 30% increase in surface area roughly corresponds to total lung inflation (22). This increase in alveolar surface area may occur during deep breaths or breathing under heavy exertion. It is likely to be exceeded in damaged lungs during mechanical ventilation with ”conventional” (i.e., not low) or even low tidal volumes because of derecruitment of alveolar units. Cyclically stretching cells with a pattern that is assumed to be physiologic (S40-13) did not result in increased LDH release compared with cells cultured on identical static membranes, nor did it alter the number of trypan blue–excluding cells. Increasing the stretching frequency (S60-13) and even more so increasing the amplitude of cyclic stretch (S40-30) resulted in a significant increase in LDH release, as has been reported previously. The reduction in the number of trypan blue–excluding cells in the high-amplitude stretch group (S40-30) was also similar to that observed in a previous study using a propidium iodide stain and annexin V binding (16). This may indicate that the trypan blue–excluding cells as well as the propidium iodide–negative/ annexin V–negative cells represent viable cells. The stretching pattern designed to reflect the level of physiologic mechanical stress in the alveolus did not alter the release of proinflammatory mediators or gene expression of iNOS, eNOS, COX-2, or 5-LO compared with static cells. High-amplitude and high-frequency stretching, however, resulted in distinct effects such as increases in NO release, eicosanoid generation, and secretion of cytokines.

While the secretion of cytokines and eicosanoids was affected by high-amplitude cyclic stretching (S40-30) only, the release of NO was stimulated by high frequency (S60-13) as well as by high amplitude (S40-30). Thus, mechanically induced NO release from ATII cells may be regulated differently from the release of eicosanoids and cytokines. Known NO donors such as SNAP as well as L-Arginine, the substrate of NO synthase, have been reported to prevent apoptosis in ATII cells subjected to increased mechanical distension (16, 24). Alveolar macrophages may be a source of NO in the alveoli (24). Our previous data (16) and the findings of this study indicate that ATII cells may also be able to synthesize NO in response to mechanical stretch and thus avert mechanically induced apoptosis. This mechanism may be equally or even more effective in high-frequency mechanical stimulation, which is imitated by the S60-13 stretching pattern. Accordingly, this pattern led to the greatest increase in NO release and at the same time did not cause a significant decrease in trypan blue–excluding cells. ATII cells do release NO in response to mechanical stretch and may be an important source of NO in the alveolar compartment. The physiologic role of NO release remains to be defined, but NO release has been shown to protect lung epithelial cells from mechanically induced injury and apoptosis. However, NO may also have adverse effects in the lung, because it has been shown to aggravate lung damage through generation of reactive molecules (25). The finding of an increase in 6-keto-PGF1␣ and TXB2 the stable products of prostacyclin and TXA2, respectively, as well

TABLE 3. RESULTS OF QUANTITATIVE RT-PCR FOR iNOS AND eNOS. S60-13 Time

Gene

6h

iNOS eNOS iNOS eNOS

24 h

Static ⌬Ct versus GAPDH 13.06 12.40 12.33 11.37

⫾ ⫾ ⫾ ⫾

0.70 0.58 0.49 0.27

S40-30

⌬Ct versus GAPDH

⌬⌬Ct versus Static

Change (fold)

⫾ ⫾ ⫾ ⫾

⫺0.14 ⫾ 0.25 ⫺0.24 ⫾ 0.33 0.088 ⫾ 0.27 ⫺0.38 ⫾ 0.36

1.10 1.18 0.94 1.30

12.92 12.16 12.42 10.99

0.62 0.75 0.33 0.61

⌬Ct versus GAPDH

⌬⌬Ct versus Static

Change (fold)

⫾ ⫾ ⫾ ⫾

⫺0.47 ⫾ 0.79 ⫺0.27 ⫾ 0.35 0.18 ⫾ 0.072 ⫺0.34 ⫾ 0.45

1.38 1.21 0.88 1.27

12.59 12.13 12.51 11.03

1.25 0.71 1.07 1.32

Definition of abbreviations: eNOS, endothelial nitric oxide synthase; GAPDH, glyceraldehyde phosphate dehydrogenase; iNOS, inducible nitric oxide synthase. Gene expression of iNOS and eNOS was performed at 6 h and 24 h for static controls and for those stretching patterns that induced a significant difference of the corresponding mediator with static controls (S60-13 and S40-30). The difference in the number of cycles between a gene of interest and GAPDH (⌬Ct vs. GAPDH) is provided for all groups and used for statistical comparison. The difference between the ⌬Ct of a stretching experiment and the ⌬Ct of the static control is used to calculate the ⌬⌬Ct and the corresponding change in gene expression. The differences were not significant with respect to time and stretching pattern, as proven by ANOVA.

208


Figure 4. Cytokine release. The concentrations of MCP-1, TNF-␣, and IL-10 in supernatants of cells subjected to cyclic stretching (closed symbols) and cells in static culture (open squares) are presented. A, D, and G show results of group S40-13; B, E, and H those of group S60-13; and C, F, and I those of group S40-30. If the differences among the groups were significant as proven by ANOVA, asterisk indicates P ⬍ 0.05 versus all other groups as proven by post hoc Bonferroni analysis.

as increased amounts of cys-LT in the supernatant of the cells from the S40-30 group, demonstrates the stimulated secretion of eicosanoids by high-amplitude stretching. The difference between static and stretched cells may appear moderate with an ⵑ 2-fold increase in concentrations. However, these differences reached the level of significance despite a reduction in the number of cells in the high-amplitude stretch group. Cys-LTs are produced by 5-LO and lead to an increase in pulmonary vascular permeability and to contraction of vascular and bronchial smooth muscle cells (26). They will therefore aggravate acute lung injury or contribute to ventilator-induced lung injury. Our data suggest that ATII are a source of leukotrienes. This is in fact supported by other experiments demonstrating that cyclic mechanical stretch increased the amount of free arachidonic acid by activating phospholipase A2 in rabbit renal proximal tubule cells (27). Because 5-LO gene expression was found to be reduced, altered 5-LO gene expression did not contribute to stretch-induced release of leukotrienes. 5-LO activity is regulated by 5-LO activating protein (FLAP) and by known 5-LO inhibitors (28). Stretch-induced changes of FLAP or inhibitory proteins may contribute to the elevated cys-LT concentration without concomitant changes in 5-LO gene expression. The effects of prostacyclin and TXA2 on pulmonary artery pressure and on platelet function are opposite. The functional consequences of the increased release of prostacyclin and TXA2

Figure 5. Rat lymphocyte activation. Rat lymphocytes were incubated with the supernatants of ATII cells after 24 h cyclic stretch with high amplitude (S40-30) or after 24 h under static conditions (static control). The lymphocytes were analyzed for expression of CD54 and of CD71. A and B show representative flow cytometry plots of CD54 and CD71 analysis. The ordinate indicates the number of events, the abscissa the fluorescence intensity. The results of n ⫽ 7 experiments with CD54 analysis and of n ⫽ 10 experiments with CD71 analysis are presented in C and D. The mean intensity of all cells analyzed is given in arbitrary units. The surface expression of the activation markers was found to be increased due to incubation with supernatant of stretched cells in 6 of 7 single experiments with analysis CD54 and in 8 of 10 single experiments with analysis of CD71. *P ⬍ 0.05 (paired t test) versus static controls.

in response to mechanical stretch are therefore difficult to predict. However, both prostacyclin and TXA2 have been suggested to be involved in the pathogenesis of acute lung injury (21). TXA2 is a very potent constrictor of the pulmonary circulation and contributes to pulmonary hypertension in several models of acute lung injury (29). TXA2 also stimulates platelet aggregation and may disturb the pulmonary circulation through this effect. MCP-1, a proinflammatory cytokine, is secreted constitutively from rat ATII cells in our experiments, but the concentration in the supernatant is not altered by either an increase in frequency or amplitude. Constitutive secretion of MCP-1 from A459 cells has been reported previously (6). Paralleling the findings of MCP-1, TNF-␣ remained unchanged by alterations of stretch amplitude and frequency. Two potent proinflammatory cytokines were thus observed to remain remarkably stable after mechanical stretch. This was different for the anti-inflammatory cytokine IL-10. The interesting observation here was that IL-10 accumulated similarly in the supernatant of static cells and cells stretched with a “physiologic” stretch pattern, and finally in cells stretched with physiologic amplitude but increased frequency. However, in cells stretched with increased amplitude, IL-10 concentration failed to increase. Thus, the secretion of the antiinflammatory IL-10 seems to be attenuated greatly by a highamplitude stretch pattern. Considering the entire response of


cytokines and arachidonic acid metabolites following mechanical stretch, it is likely that the balance is shifted not by increasing the standard set of fast and powerful proinflammatory cytokines, but rather by a combination of slight increases on the proinflammatory side and a stronger decrease of an important antiinflammatory cytokine. In accordance with these findings, the supernatants of ATII cells subjected to high-amplitude cyclic stretch exhibited a stimulatory effect on rat lymphocytes indicating the biological relevance. We studied the effect of supernatant of static cells and cells stretched with the high-amplitude stretch pattern (S40-30) on two distinct lymphocyte markers of activation. CD71 represents the transferrin receptor. It is widely used as a marker for activation or proliferation of lymphocytes. CD71 expression is virtually undetectable in resting T- and B-lymphocytes, but increased in response to activation (30). The second marker of activation, CD54, is an adhesion molecule (ICAM-1), which mediates lymphocyte adhesion and migration from the vascular bed into the interstitium as well as the cell–cell interactions of lymphocytes (31). It is only weakly expressed on resting lymphocytes and its expression is considered as a marker of activation (32). Both markers were found to be induced by the supernatants of high-amplitude stretched ATII cells compared with static controls. Because stimulated lymphocytes may well contribute to aggravation or initiation of lung injury, this confirms a proinflammatory effect of high-amplitude stretched ATII cell supernatant on rat lymphocytes. We used freshly prepared rat ATII cells in primary culture. Compared with the human alveolar epithelial A549 cell line, which lacks essential functional and morphologic properties of ATII cells, these native ATII cells are more appropriate for the study of proprietary ATII cell functions. The use of this rat cell model required consideration in the selection of the selected cytokines. The conclusions for the situation in humans, of course, is somewhat limited, as in all animal models. Rat ATII cells in primary culture start to undergo phenotypic changes (33). Our experimental time frame including 22 h for adherence followed by 6–24 h of experimental time does not appear long enough for ATII cells to dedifferentiate completely. In addition, it is not known, whether dedifferentiation known to occur rather rapidly on plastic culture dishes will occur at the same frequency on flexible and stretched membranes. However, the phenotypic changes that may occur in this time frame may represent another limitation of this primary rat ATII cell model. The mechanisms that mediate changes in the secretory behavior of ATII cells were not addressed in this study. Other recent studies using isolated lung models demonstrated phosphoinositide 3-OH kinase–akt-kinase pathway involvement in lung overinflation (34). Activated akt-kinase may induce nuclear translocation of NF-␬B via interaction of Akt and I␬B-kinase and may also change gene expression of several early response and other genes. In contradiction with our findings, these authors reported an increased release of proinflammatory cytokines in response to overventilation. Furthermore, activated akt-kinase may phosphorylate and activate eNOS (35). It may therefore be speculated that these mechanisms may contribute to findings observed in our study. High-amplitude mechanical stretch in comparison with physiologic mechanical stimuli and static controls modifies the release of mediators from ATII cells. Increased production of eicosanoids and a decreased generation of the anti-inflammatory IL-10 with unchanged proinflammatory cytokines were found in response to cyclic mechanical stretch with an amplitude simulating lung inflation to ⵑ 100% TLC. The generation of NO from ATII cells at the same time may represent an anti-apoptotic and protective mechanism. The maximum of this protective effect

209

was seen in response to high-frequency cyclic stretch and is markedly smaller following high-amplitude mechanical stretch. These findings suggest that continuing mechanical distension of the alveolar epithelium to the approximate limit, comparable to TLC, contributes to a systemic inflammatory response and to reduced ATII cell survival by a shift in the balance of pro- and anti-inflammatory mediators. Conflict of Interest Statement : None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Acknowledgments : The expert technical assistance of Ms. Konstanze Bu¨ttner and Ms. Kirsten Wrabetz is greatly acknowledged.

References 1. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000;342:1301–1308. 2. Ranieri VM, Suter PM, Tortorella C, De Tullio R, Dayer JM, Brienza A, Bruno F, Slutsky AS. Effect of mechanical ventilation on inflammatory mediators in patients with acute respiratory distress syndrome: a randomized controlled trial. JAMA 1999;282:54–61. 3. Tremblay L, Valenza F, Ribeiro SP, Li J, Slutsky AS. Injurious ventilatory strategies increase cytokines and c-fos m-RNA expression in an isolated rat lung model. J Clin Invest 1997;99:944–952. 4. von Bethmann AN, Brasch F, Nusing R, Vogt K, Volk HD, Muller KM, Wendel A, Uhlig S. Hyperventilation induces release of cytokines from perfused mouse lung. Am J Respir Crit Care Med 1998;157:263–272. 5. Hammerschmidt S, Schiller J, Kuhn H, Meybaum M, Gessner C, Sandvoss T, Arnold K, Wirtz H. Influence of tidal volume on pulmonary NO release, tissue lipid peroxidation and surfactant phospholipids. Biochim Biophys Acta 2003;1639:17–26. 6. Koyama S, Sato E, Nomura H, Kubo K, Nagai S, Izumi T. Type II pneumocytes release chemoattractant activity for monocytes constitutively. Am J Physiol 1997;272:L830–L837. 7. Standiford TJ, Kunkel SL, Basha MA, Chensue SW, Lynch JP III, Toews GB, Westwick J, Strieter RM. Interleukin-8 gene expression by a pulmonary epithelial cell line: a model for cytokine networks in the lung. J Clin Invest 1990;86:1945–1953. 8. Crestani B, Cornillet P, Dehoux M, Rolland C, Guenounou M, Aubier M. Alveolar type II epithelial cells produce interleukin-6 in vitro and in vivo: regulation by alveolar macrophage secretory products. J Clin Invest 1994;94:731–740. 9. Koyama S, Sato E, Nomura H, Kubo K, Miura M, Yamashita T, Nagai S, Izumi T. Bradykinin stimulates type II alveolar cells to release neutrophil and monocyte chemotactic activity and inflammatory cytokines. Am J Pathol 1998;153:1885–1893. 10. Cott GR, Westcott JY, Voelkel NF. Prostaglandin and leukotriene production by alveolar type II cells in vitro. Am J Physiol 1990;258:L179– L187. 11. Koyama S, Sato E, Masubuchi T, Takamizawa A, Kubo K, Nagai S, Izumi T. Alveolar type II-like cells release G-CSF as neutrophil chemotactic activity. Am J Physiol 1998;275:L687–L693. 12. Mitchell JA, Belvisi MG, Akarasereenont P, Robbins RA, Kwon OJ, Croxtall J, Barnes PJ, Vane JR. Induction of cyclo-oxygenase-2 by cytokines in human pulmonary epithelial cells: regulation by dexamethasone. Br J Pharmacol 1994;113:1008–1014. 13. Wirtz HR, Dobbs LG. Calcium mobilization and exocytosis after one mechanical stretch of lung epithelial cells. Science 1990;250:1266–1269. 14. Sanchez-Esteban J, Cicchiello LA, Wang Y, Tsai SW, Williams LK, Torday JS, Rubin LP. Mechanical stretch promotes alveolar epithelial type II cell differentiation. J Appl Physiol 2001;91:589–595. 15. Edwards YS, Sutherland LM, Power JH, Nicholas TE, Murray AW. Cyclic stretch induces both apoptosis and secretion in rat alveolar type II cells. FEBS Lett 1999;448:127–130. 16. Hammerschmidt S, Kuhn H, Grasenack T, Gessner C, Wirtz H. Apoptosis and necrosis induced by cyclic mechanical stretching in alveolar type II cells. Am J Respir Cell Mol Biol 2004;30:396–402. 17. Vlahakis NE, Hubmayr RD. Invited review: plasma membrane stress failure in alveolar epithelial cells. J Appl Physiol 2000;89:2490–2496.

210


18. Tschumperlin DJ, Oswari J, Margulies AS. Deformation-induced injury of alveolar epithelial cells: effect of frequency, duration, and amplitude. Am J Respir Crit Care Med 2000;162:357–362. 19. Yamamoto H, Teramoto H, Uetani K, Igawa K, Shimizu E. Stretch induces a growth factor in alveolar cells via protein kinase. Respir Physiol 2001;127:105–111. 20. Vlahakis NE, Schroeder MA, Limper AH, Hubmayr RD. Stretch induces cytokine release by alveolar epithelial cells in vitro. Am J Physiol 1999;277:L167–L173. 21. Deby-Dupont G, Braun M, Lamy M, Deby C, Pincemail J, Faymonville ME, Damas P, Bodson L, Lecart MP, Goutier R. Thromboxane and prostacyclin release in adult respiratory distress syndrome. Intensive Care Med 1987;13:167–174. 22. Tschumperlin DJ, Margulies SS. Alveolar epithelial surface area-volume relationship in isolated rat lungs. J Appl Physiol 1999;86:2026–2033. 23. Winer J, Jung CK, Shackel I, Williams PM. Development and validation of real-time quantitative reverse transcriptase-polymerase chain reaction for monitoring gene expression in cardiac myocytes in vitro. Anal Biochem 1999;270:41–49. 24. Edwards YS, Sutherland LM, Murray AW. NO protects alveolar type II cells from stretch-induced apoptosis: a novel role for macrophages in the lung. Am J Physiol Lung Cell Mol Physiol 2000;279:L1236–L1242. 25. Beckman JS, Koppenol WH. Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly. Am J Physiol 1996;271:C1424–C1437. 26. Busse WW. Leukotrienes and inflammation. Am J Respir Crit Care Med 1998;157:S210–S213. 27. Alexander LD, Alagarsamy S, Douglas JG. Cyclic stretch-induced cPLA2

28. 29.

30.

31. 32.

33.

34.

35.

mediates ERK 1/2 signaling in rabbit proximal tubule cells. Kidney Int 2004;65:551–563. Peters-Golden M, Brock TG. 5-lipoxygenase and FLAP. Prostaglandins Leukot Essent Fatty Acids 2003;69:99–109. Grimminger F, Wahn H, Kramer HJ, Stevens J, Mayer K, Walmrath D, Seeger W. Differential influence of arachidonic vs. eicosapentaenoic acid on experimental pulmonary hypertension. Am J Physiol 1995;268: H2252–H2259. Galbraith RM, Galbraith GM. Expression of transferrin receptors on mitogen-stimulated human peripheral blood lymphocytes: relation to cellular activation and related metabolic events. Immunology 1981; 44:703–710. Stanciu LA, Djukanovic R. The role of ICAM-1 on T-cells in the pathogenesis of asthma. Eur Respir J 1998;11:949–957. Chirathaworn C, Tibbetts SA, Chan MA, Benedict SH. Cross-linking of ICAM-1 on T cells induces transient tyrosine phosphorylation and inactivation of cdc2 kinase. J Immunol 1995;155:5479–5482. Dobbs LG, Williams MC, Gonzalez R. Monoclonal antibodies specific to apical surfaces of rat alveolar type I cells bind to surfaces of cultured, but not freshly isolated, type II cells. Biochim Biophys Acta 1988; 970:146–156. Uhlig U, Fehrenbach H, Lachmann RA, Goldmann T, Lachmann B, Vollmer E, Uhlig S. Phosphoinositide 3-OH kinase inhibition prevents ventilation-induced lung cell activation. Am J Respir Crit Care Med 2003;24:24. Kuebler WM, Uhlig U, Goldmann T, Schael G, Kerem A, Exner K, Martin C, Vollmer E, Uhlig S. Stretch activates nitric oxide production in pulmonary vascular endothelial cells in situ. Am J Respir Crit Care Med 2003;168:1391–1398.