Thomas Dschietzig is employee of Immundiagnostik which holds a patent (DE102005040492A1) on “Relaxin in solutions for the preservation of organs”. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and materials.
Conceived and designed the experiments: TD KM. Performed the experiments: KA MW. Analyzed the data: KA MW TD. Contributed reagents/materials/analysis tools: TD. Wrote the paper: KA MW KM TD.
Early allograft dysfunction following lung transplantation is mainly an ischemia/reperfusion (IR) injury. We showed that relaxin-2 (relaxin) exerts a protective effect in lung IR, attributable to decreases in endothelin-1 (ET-1) production, leukocyte recruitment, and free radical generation. Here, we summarize our investigations into relaxin’s signalling.
Isolated rat lungs were perfused with vehicle or 5 nM relaxin (n = 6–10 each). Thereafter, experiments were conducted in the presence of relaxin plus vehicle, the protein kinase A inhibitors H-89 and KT-5720, the NO synthase (NOS) inhibitor L-NAME, the iNOS inhibitor 1400W, the nNOS inhibitor SMTC, the extracellular signal-regulated kinase-1/2 (ERK-1/2) inhibitor PD-98059, the phosphatidylinositol-3 kinase (PI3K) inhibitor wortmannin, the endothelin type-B (ETB) antagonist A-192621, or the glucocorticoid receptor (GR) antagonist RU-486. After 90 min ischemia and 90 min reperfusion we determined wet-to-dry (W/D) weight ratio, mean pulmonary arterial pressure (MPAP), vascular release of ET-1, neutrophil elastase (NE), myeloperoxidase (MPO), and malondialdehyde (MDA). Primary rat pulmonary vascular cells were similarly treated.
IR lungs displayed significantly elevated W/D ratios, MPAP, as well as ET-1, NE, MDA, and MPO. In the presence of relaxin, all of these parameters were markedly improved. This protective effect was completely abolished by L-NAME, 1400W, PD-98059, and wortmannin whereas neither PKA and nNOS inhibition nor ETB and GR antagonism were effective. Analysis of NOS gene expression and activity revealed that the relaxin-induced early and moderate iNOS stimulation is ERK-1/2-dependent and counter-balanced by PI3K. Relaxin-PI3K-related phosphorylation of a forkhead transcription factor, FKHRL1, paralleled this regulation. In pulmonary endothelial and smooth muscle cells, FKHRL1 was essential to relaxin-PI3K signalling towards iNOS.
In this short-time experimental setting, relaxin protects against IR-induced lung injury via early and moderate iNOS induction, dependent on balanced ERK-1/2 and PI3K-FKHRL1 stimulation. These findings render relaxin a candidate drug for lung preservation.
Ischemia-reperfusion (IR) injury
We have recently provided experimental evidence that the peptide human relaxin-2 (relaxin), a member of the insulin superfamily, exerts a protective effect in IR-induced lung injury
Here, we summarize our investigations into the signal transduction of this relaxin effect: The peptide is shown to exert its protective effect via early and moderate iNOS induction dependent on balanced ERK-1/2 and PI3K stimulation.
The study conforms to the European Commission Directive 2010/63/EU. According to German animal welfare regulations, the killing of laboratory animals for mere excision of organs, without prior experiments performed, does not pose an animal experiment and an approval is therefore not required. Male Wistar rats, weighing 300 to 350 g, were selected for this study. For excision of the lung lethal anesthesia with thiopental sodium (80 mg/kg body weight (BW) i. p.) was performed. After cessation of the corneal reflex, animals were exsanguinated by cutting the carotid artery and the jugular vein. A tracheotomy permitted positive pressure ventilation with a small animal respirator at 60 strokes/min, tidal volume 8–10 ml/kg BW, 1 mm Hg positive end-expiratory pressure, gas mixture 95% O2 and 5% CO2. A median sternotomy was performed, a cannula was placed into the pulmonary artery, and the heart was removed to allow passive drainage of the pulmonary effluent from the pulmonary veins. Perfusion was carried out with Krebs-Henseleit buffer containing composition in mmol/l: NaCl, 127; KCl, 3.7; CaCl2, 2.5; KH2PO4, 1.2; MgSO4, 1.1; NaHCO3, 25.0; glucose, 10; pyruvate, 1.8; and
Each lung was perfused with 40 ml of buffer in recirculatory mode, the entire perfusate was rapidly frozen in liquid nitrogen and stored at −70°C for determination of ET-1, NE, and MPO. Different lungs were homogenized (for protocol, see citrulline assay) for the immediate determination of NOS activities and for measurement of NOS gene expression and tissue MDA. The IR lungs were exposed to ischemia by stopping the perfusion and the ventilation for 90 min. The subsequent reperfusion lasted also 90 min after exchanging the complete perfusate.
The following drugs were administered during the entire IR cycle in combination with 5 nM human relaxin, which was the dose used in our previous work
the non-selective nitric oxide synthase (NOS) inhibitor L-NAME (100 µM) (n = 8)
the inducible NOS (iNOS) inhibitor 1400W (1 µM) (n = 6)
the neuronal NOS (nNOS) inhibitor S-methyl-L-thiocitrulline (l-SMTC) (10 µM) (n = 6)
the extracellular signal-regulated kinase-1/2 (ERK-1/2) inhibitor PD-98059 (50 µM) (n = 6)
the phosphatidylinositol-3 kinase (PI3K) inhibitor wortmannin (100 nM) (n = 6)
the glucocorticoid receptor (GR) antagonist RU-486 (500 nM) (n = 3)
the endothelin type-B receptor (ETB) receptor antagonist A-192621 (500 nM) (n = 4)
the protein kinase A (PKA) inhibitors H-89 (5 µM) and KT-5720 (20 µM) (n = 3)
In general, all drugs were also tested alone in the various settings (which is not always shown to assure clarity of figures) and, if not otherwise stated, were found to have no effect.
After IR experiments without any intervention (n = 6), lungs were homogenized as described below to test for the selectivity of the different NOS inhibitors. In another subset of lungs (n = 6 each), L-NAME, 1400W, SMTC, or relaxin were exclusively administered either during ischemia or during reperfusion.
In order to record the time course of NOS regulations and FKHRL1 phosphorylation lungs were also processed at baseline, after ischemia, or after reperfusion (n = 6 each).
Wet-to-dry weight ratio (W/D); NE, MPO, and ET-1in perfusate; as well as tissue MDA were determined as described in detail elsewhere
Lungs were homogenized over 30 min in ice-cold RIPA buffer (1% NP-40) (4 ml/mg tissue) containing 50 mmol/l NaF, 15 mmol/l Na4P2O7, 2 mmol/l Na3VO4,10 µg/ml trypsin inhibitor, 1 mmol/l PMSF, 2 µg/ml aprotinin, 5 µg/ml leupeptin, 0.1 mmol/l okadaic acid, 5 mmol/l EDTA, and 1 mmol/l EGTA. The homogenates were centrifuged at 10.000 rpm at 4°C over 15 min; then, the supernatants were collected, and the protein concentration was determined using the Bradford method.
Enzyme reactions (final volume 200 µl) were carried out in the presence of 10 µl homogenization supernatant, 1 mmol/l NADPH, 15 µmol/l tetrahydrobiopterin, 1 µmol/l FAD, 1 µmol/l FAM, 1 mmol/l DTT, 10 µmol/l 3H-L-arginine, 1 mmol/l MgCl2, 100 nmol/l calmodulin, 300 µmol/l calcium, 0.2 mmol/l EDTA, and 0.2 mmol/l EGTA in HEPES buffer (50 mmol/l). Assays were also run in the presence of 1 mmol/l of the NOS inhibitor L-NG-nitro-L-arginine to detect non-specific production of 3H-citrulline as well as in the presence of 5 mmol/l EDTA to calculate calcium/calmodulin-independent NOS activity. Activity of constitutive NOS (eNOS plus nNOS) was then calculated by subtracting values for non-specific and for calcium/calmodulin-independent activity from the originally obtained values. In additional experiments, we also aimed at determining the contribution of nNOS which was regarded as the part of constitutive NOS sensitive to inhibition by SMTC (10 µmol/l).
In order to validate the selectivity of 1400W lung homogenates were pre-incubated over 30 min with 1400W prior to being diluted into the above-mentioned assays.
Assays were incubated for 60 min at 37°C; the reaction was stopped by adding ice-cold stop buffer (HEPES, pH 5.5, 10 mmol/l EDTA) and then transferred to columns containing 1 ml of Dowex 50w resin. After passage through the Dowex columns, 3H-citrulline was quantified in a scintillation counter.
Total RNA was extracted from lungs using an RNeasy kit (Quiagen). The reverse transcriptase reaction contained 5 ng per µl total RNA, M-MLV reverse transcriptase (800 U), RNAseOUT (40 U), reverse primer (4 pM), dNTPs (0.5 mM), and supplied optimal buffers (all from Invitrogen). PCR was performed with 1 ng of cDNA template, 200 nM of iNOS or eNOS primers, and SYBR Green PCR master mix (Applied Biosystems). The eNOS (GeneBank accession: AJ011116) and iNOS primers (GeneBank Accession: D44591) were designed as described by Okada et al.
Lungs were homogenized over 30 min in ice-cold RIPA buffer (1% NP-40) (4 ml/mg tissue) containing 50 mmol/l NaF, 15 mmol/l Na4P2O7, 2 mmol/l Na3VO4,10 µg/ml trypsin inhibitor, 1 mmol/l PMSF, 2 µg/ml aprotinin, 5 µg/ml leupeptin, 0.1 mmol/l okadaic acid, 5 mmol/l EDTA, and 1 mmol/l EGTA. The homogenates were centrifuged at 10.000 rpm at 4°C over 15 min; then, the supernatants were collected, and the protein concentration was determined using the Bradford method. After heating, equal amounts of protein (20 µg per lane) were separated by 8% SDS-PAGE and transferred to PVDF membranes (Hybond, Amersham). These membranes were blocked over 2 h at room temperature with RotiBlock (Roth).
Total FKHRL1 and phospho(Ser253)-FKHRL1 were detected with rabbit polyclonal IgG antibodies (sc-11351 and sc-12897, respectively) from Santa Cruz (dilution 1∶1000 and 1∶500, respectively). Horseradish peroxidase-conjugated antibodies (Santa Cruz) served as secondary antibodies. The signals were visualized by the ECL Plus chemoluminiscence system (abcam) and quantified using Quantity One (Bio-Rad Versadoc). For semi-quantitative analysis, densitometric data for phosphor-FKHRL1 were normalized to the total FKHRL1 protein.
Primary rat pulmonary artery endothelial cells (RPAEC) were prepared from rat pulmonary arteries by collagenase incubation, see
Primary rat pulmonary artery smooth muscle cells (RPASMC) were prepared according to Dahan et al.
Cells were transiently transfected with SilencerR Select siRNA vectors targeting rat FKHRL1 or rat FOXB2 and scrambled siRNA using the Lipofectamine method according to the supplier’s instructions (Invitrogen). After 24 h of transfection, RPAEC and RPASMC were washed with DMEM supplemented with 4% horse serum plus 2% FCS and with 5% FCS, respectively, kept at rest for 8 h and then used for the following experiments: Cells were subjected to hypoxia (5% oxygen) for 90 min followed by another 90 min of normoxia. Control cells were kept under normoxia for 180 min. Concomitantly, cells were treated with vehicle, relaxin (5 nM), wortmannin (100 nM), PD-98059 (50 µM), and combinations thereof (n = 5 each).
The knock-downs were ascertained by 2 different antibodies against the N-terminus (sc-34897) and the C-terminus (sc-34894) of rat FKHRL1 and against rat FOXB2 (sc-132299) (Santa Cruz).
For NOS assays, cells were homogenized in RIPA buffer as described above for isolated lungs; 20 µl of extracts were then diluted into NOS assay buffer.
The data are presented as means ± S.E.M. Differences between groups were analysed using the Kruskal-Wallis ANOVA on ranks or, in the case of two independent variables (time and group), a non-parametric two-way ANOVA. After global testing, individual groups were compared using the Mann-Whitney rank sum test with Bonferroni-Holm adjustement of p. An error probability of p<0.05 was regarded as significant.
As recently shown
In contrast, the selective nNOS inhibitor, SMTC; the 2 different PKA inhibitors, KT-5720 and H-89; the ETB antagonist, A-192621; and the GR antagonist, RU-486, did not show any effect on relaxin’s beneficial action during IR (data not shown).
In NOS assays conducted after IR experiments without intervention (
Isolated rat lungs were perfused with buffer in recirculatory mode and underwent an IR cycle comprised of 90(A) NOS assays conducted in lung homogenates after IR experiments without intervention (n = 6 each) showing that 1400W inhibited iNOS (Ca/CaM-independent activity) with sufficient selectivity over constitutive NOS. L-NAME, a non-selective NOS inhibitor, was given at 100 µM, 1400W, a selective iNOS inhibitor, at 1 µM, and SMTC, an nNOS inhibitor, at 10 µM. (B) In another subset of lungs (n = 6 each), L-NAME, 1400W, SMTC, or relaxin (5 nM) were exclusively administered either during ischemia or during reperfusion. Here, the wet-to-dry ratio is shown which was representative for all other readouts. While L-NAME affected both ischemia and reperfusion 1400W impacted only on reperfusion; relaxin (when not applied during the full IR cycle) had no effect. There was no relevant effect of nNOS inhibition at all. P<0.05; *vs. control.
In other experiments (
Wet-to-dry weight ratio, mean pulmonary arterial pressure (MPAP), neutrophil elastase, myeloperoxidase (MPO), and endothelin-1 in perfusate, as well as tissue malonyldialdehyde (MDA) in isolated rat lungs perfused with buffer in recirculatory mode and subjected to 90 min ischemia followed by 90 min reperfusion. Experiments (n = 6 each) were carried out in the presence of vehicle (control), 5 nM of relaxin (Rlx), the selective iNOS inhibitor 1400W (1 µ), and combinations thereof. Both 1400W (IR+1400W) and relaxin (IR+Rlx) exerted beneficial effects; when applied in the presence of 1400W relaxin did not add protection (IR+Rlx+1400W). P<0.05; *vs. control; #vs. IR.
Wet-to-dry weight ratio, mean pulmonary arterial pressure (MPAP), neutrophil elastase, myeloperoxidase (MPO), and endothelin-1 in perfusate, as well as tissue malonyldialdehyde (MDA) in isolated rat lungs perfused with buffer in recirculatory mode and subjected to 90 min ischemia followed by 90 min reperfusion. Experiments (n = 6 each) were carried out in the presence of vehicle (control), 5 nM of relaxin (Rlx), the ERK-1/2 inhibitor PD-98059 (PD) (50 µmol/l), the PI3K inhibitor wortmannin (WM) (100 nM), and combinations thereof. Inhibition of either ERK-1/2 or PI3K did not change IR (IR+PD and IR+WM) but completely prevented the effects of relaxin therein (IR+Rlx+PD and IR+Rlx+WM). P<0.05; *vs. control; #vs. IR.
Isolated rat lungs perfused with buffer in recirculatory mode were processed at baseline, after 90(n = 6 each) in order to determine gene expression (upper panel) and activity (lower panel) of eNOS and iNOS. Experiments were carried out in the presence of vehicle (control), 5 nM of relaxin, the ERK-1/2 inhibitor PD-98059 (PD) (50 µmol/l), the PI3K inhibitor wortmannin (WM) (100 nM), and combinations thereof. The natural course of IR, down-regulation of eNOS and up-regulation of iNOS in reperfusion, is altered by relaxin, into eNOS maintenance in reperfusion and early moderate iNOS induction in ischemia. Relaxin’s action is suppressed both by ERK-1/2 and PI3K inhibition. P<0.05; *vs. baseline; #vs. relaxin.
The use of a two-way ANOVA allowed for interaction testing. Whereas no significant overall interaction “time×group” was found for eNOS mRNA it was detectable for iNOS mRNA and activity (p for overall testing <0.00001 either). Then, a significant interaction between relaxin and wortmannin towards iNOS could be inferred from the finding that upon pairwise testing, “relaxin+wortmannin” was different from both “relaxin alone” (p<0.00001) and “wortmannin alone” (p = 0.0004) while there was no difference between the two latter ones (p = 0.29, relaxin vs. wortmannin). Such an interaction was not determined for relaxin and the ERK inhibitor, PD-98059.
Since PI3K-related phosphorylation of the forkhead transcription factor, FKHRL1, has been shown to inhibit iNOS induction in lung cells in inflammatory states
Isolated rat lungs perfused with buffer in recirculatory mode were processed at baseline, after 90(n = 6 each) in order to determine protein expression (panel A) and Ser-253 phosphorylation (panel B) of the forkhead transcription factor, FKHRL1. Experiments were carried out in the presence of vehicle (control), 5 nM of relaxin, the ERK-1/2 inhibitor PD-98059 (PD) (50 µmol/l), the PI3K inhibitor wortmannin (WM) (100 nM), and combinations thereof. (A) Representative Western blot from lung homogenates produced after ischemia. (B) Quantification of Western blot data (n = 6 per group). Relaxin causes FKHRL-1 phosphorylation during ischemia in a PI3K (wortmannin)-dependent fashion; there is no dependence of ERK-1/2. P<0.05; *vs. baseline; #vs. relaxin.
Using a similar protocol in RPAEC and RPASMC (except for the fact that IR was mimicked here as hypoxia) (
Primary rat pulmonary artery endothelial cells (RPAEC, left side) and rat pulmonary artery smooth muscle cells (RPASMC, right side) were processed at baseline, after 90 min hypoxia (5% oxygen), or after another 90 min of normoxia (n = 5 each) in order to determine iNOS activity. Experiments were carried out in the presence of vehicle (control), 5 nM of relaxin, the ERK-1/2 inhibitor PD-98059 (PD) (50 µmol/l), the PI3K inhibitor wortmannin (WM) (100 nM), and combinations thereof. Prior to experiments, cells had been transfected with scrambled siRNA (control), FOXB2 siRNA, or FKHRL1 siRNA. While both transfection with scrambled siRNA and knock-down of non-related forkhead factor, FOXB2, had no influence knock-down of FKHRL-1 abolished the susceptibility of relaxin’s effect towards PI3K inhibition in ischemia and equalized the extent of iNOS induction in reperfusion. P<0.05; *vs. baseline; #vs. relaxin.
Corresponding to the data in
In control cells kept under constant normoxia for 180 min, there was no relevant regulation of iNOS (data not shown).
The rationale behind studying the effects of relaxin in lung IR is given by its broad spectrum of anti-inflammatory, vasodilatory, and endothelium-protecting properties
In general, relaxin has been found to target a variety of pathways via its receptor, RXFP1, including cAMP/PKA, NO via all different NOS enzymes, ETB receptors, ERK-1/2, and PI3K
Under pathological conditions such as pulmonary IR, the increased vascular tone may be partially explained by an attenuation of NO synthesis from eNOS and/or decreased NO bioactivity
Studies on vascular endothelial and smooth muscle cells have clearly demonstrated that relaxin can promote vascular NO generation, either by activating eNOS or by increasing, in a cell type-dependent manner, the expression of the three different NOS isoforms
The current study demonstrates that the mitigation of IR-induced lung injury by relaxin is promoted by an early and moderate induction of iNOS. This effect mainly consists of an up-regulation of iNOS mRNA and Ca/calmodulin-independent NO production during
The conclusion of iNOS being prevailingly responsible for relaxin’s effect herein is based on the following findings: Both non-specific NOS inhibition by L-NAME and specific iNOS inhibition by 1400W completely suppressed the relaxin-related protection; the nNOS-specific inhibitor SMTC did not show any effect; and modulating the course of relaxin-mediated iNOS induction by parallel ERK-1/2 or PI3K inhibition also abolished relaxin’s beneficial action.
Our findings (
The selectivity of 1400W for iNOS over eNOS/nNOS was validated in our experimental setting: First, 1400W proved sufficiently selective in NOS assays from lung homogenates. Second, and in contrast to L-NAME, it affected only the reperfusion events when iNOS was evidently induced but was neutral during ischemia when eNOS function was still normal.
Moreover, relaxin seems to influence what is called eNOS-iNOS cross-talk, a kind of reciprocal regulation involving iNOS stimulation at the cost of eNOS expression and activity
Furthermore, the mechanism of action of relaxin towards iNOS seems to be mediated by parallel activation of the ERK-1/2 and PI3K cascades. Causal implication of these cascades was confirmed by the finding that both PD-98059 and wortmannin completely suppressed the protective relaxin effect. Whereas ERK-1/2 inhibition prevented the early iNOS up-regulation during ischemia PI3K inhibition by wortmannin enhanced this effect. Accordingly, relaxin-related ERK-1/2 activation, which is known to target NF-κB
With regard to iNOS induction in IR, relaxin and PI3K inhibition by wortmannin enhance each other. An interaction was identified by statistical analysis and supports the fact that PI3K activation represents an essential part of relaxin’s signaling. This interaction should be considered in the design of relaxin-based therapies since enhanced iNOS induction may be harmful.
Endothelin-1, which acts as vasoconstrictory, permeability-increasing, and pro-inflammatory mediator
Upper panel: In a short-term IR cycle in rat isolated lungs, relaxin induced iNOS in a manner (early in ischemia but moderately in level) distinct from the natural course (later but marked in reperfusion). This moderately elevated NO generation suppresses the IR-related ET-1 surge, both effects act in concert to inhibit cell activation and subsequent tissue damage. Lower panel: Early and moderate iNOS induction by relaxin is dependent on a subtle balance between stimulatory ERK-1/2-NF-κB and inhibitory PI3K-FKHRL-1 pathways.
As to the negative results of pharmacological interventions, relaxin has been demonstrated to stimulate endothelial and epithelial ETB expression via a Ras-independent Raf-1–MEK-1–ERK-1/2 pathway that activates NF-κβ
At last, relaxin-mediated ERK-1/2 activation has been reported to occur downstream of cAMP/PKA but PKA-independent effects on ERK have also been established depending on the cell type investigated
As to the limitations to the present study, we investigated IR effects in a short-time setting, which may not be accurately comparable to real clinical IR states. However, isolated lungs exhibit key characteristics of IR-related pulmonary alteration, in particular edema, pulmonary hypertension, cell activation, and ROS generation
Finally, lungs were perfused with saline buffer, which excludes many cell-mediated events that may affect pulmonary perfusion pressure and vascular permeability. Blood-free perfusion, on the other hand, does not affect the number of neutrophils adhering to the endothelium of isolated lungs
In conclusion, in this short-time experimental setting, human relaxin-2 exerts its protective effect on IR-induced lung injury via early and moderate iNOS induction, dependent on balanced ERK-1/2 and PI3K stimulation (as summarized in