Sepsis remains the leading cause of death on noncardiologic intensive care units with mortality rates of approximately 30% to 50% (1). It is increasingly recognized that endothelial barrier disruption in sepsis and systemic inflammation is a hallmark preceding breakdown of the microcirculation followed by metabolic dysbalance, organ failure, and death in patients (2). Despite its pathophysiologic relevance, a therapy to stabilize endothelial barrier disruption in septic patients is not available.
Meanwhile, it is well established that breakdown of endothelial barrier functions following application of bacterial lipopolysaccharide (LPS) or proinflammatory cytokines (e.g., tumor necrosis factor α [TNF-α]) is caused by reduced cAMP levels in endothelial cells (3). Accordingly, application of cAMP-increasing agents blocked inflammation-induced endothelial barrier breakdown in vivo and in vitro. Recently, we demonstrated in a complex in vivo model of LPS-induced systemic inflammation that systemic i.v. application of cAMP-increasing phosphodiesterase 4 inhibitors (PD-4-I) rolipram and roflumilast was effective to prevent endothelial barrier disruption and breakdown of the microcirculation with significantly metabolic improvement and increased survival rates (4).
In support of that, it was observed that intraperitoneal application of PD-4-I resulted in attenuation of plasma loss and transvascular fluid exchange in volume-expanded mice (5). Furthermore, PD-4-I application reduced TNF-α levels and inhibited neutrophil recruitment in a model of acute lung injury, which led to the assumption that PD-4-I predominately exerts protective effects by immunmodulation (6).
Because LPS-induced systemic inflammation resulted in rapid mortality, long-term observations of beneficial and adverse effects of i.v. use of PD-4-I are absent. Moreover, the LPS model does not reflect the complex situation of a patient with polymicrobial sepsis. In addition, a correlation between PD-4-I–mediated endothelial barrier stabilization and the immunomodulatory effects of PD-4-I is not clear at present.
Here we tested the effects of continuous i.v. application of PD-4-I rolipram at different doses in a model of polymicrobial sepsis in the time course of 26 h. The present study provides further evidence that PD-4-I rolipram may be suitable therapeutic agents to stabilize the endothelial barrier in sepsis.
MATERIALS AND METHODS
After animal care committee approval (Laboratory Animal Care and Use Committee of the district of Unterfranken, Germany), experiments were performed on male 32 Sprague-Dawley rats (355 ± 34 g BW; Janvier [Le Genest St Isle, France]). Animals were kept under conditions that conformed to the National Institutes of Health “Guide for the Care and Use of Laboratory Animals,” approved by the Government of Unterfranken as well as for those of the US National Institutes of Health. All rats were maintained on a standard diet and water ad libitum at 12-h day-and-night cycles. Animals were not fasted before the procedure.
Phosphodiesterase 4 inhibitor rolipram (Sigma-Aldrich, Taufkirchen, Germany) was dissolved in 0.5 mL 100% dimethyl sulfoxide (DMSO) and used to increase cAMP levels by continuous i.v. administration of 3 mg/kg body weight (BW) per hour or 1 mg/kg BW per hour as indicated below. In the following, 3 mg/kg BW per hour was defined as “high-dose” and 1 mg/kg BW per hour as “low-dose” application of PD-4-I. Fluorescein isothiocyanate (FITC)–albumin was applied i.v. at 5 mg/100 g (Sigma, Schnelldorf, Germany). Dose selection was based on our previous experiences (4).
We established a modified model of polymicrobial sepsis using the previously described colon ascendens stent peritonitis (CASP) model (7, 8). To enable macrohemodynamic and microhemodynamic monitoring, we modified the model based on or previously described model in which we used LPS for induction of systemic inflammation as follows (4) (Fig. 1): first, animals were anesthetized using isoflurane (Forene; Abbott, Wiesbaden, Germany)/nitrous oxide inhalation (Air Liquide, Düsseldorf, Germany). To apply i.v. medication, the right jugular vein was cannulated. In addition, the left carotid artery was cannulated for continuous blood pressure and heart rate measurements (model 88S; Hewlett Packard, Hamburg, Germany) as well as for taking arterial blood samples for repeated blood gas analyses. Thereafter, anesthetized animals underwent median laparotomy; the cecum was identified and gently pulled out. Twenty-millimeter aboral from the ileocecal valve, the wall of the ascending colon was pierced with a suture at the antimesenteric side. The colon was perforated with a 14-gauge needle. Thereafter, a 10-charriere plastic tube (tip of a suction catheter, type “Ideal”; B.Braun, Melsungen, Germany) was inserted and fixed with the prepared suture at the colonic wall. This served to maintain the perforation of the colon during the duration of the experiments. Stool was milked from the cecum toward the colonic stent by the use of the cotton swabs until stool appeared on top of the stent. Then, the gut was put back into the peritoneal cavity, and the stent was flushed with 2 mL NaCL 0.9% (Fresenius Kabi, Bad Homburg, Germany) to distribute the feces into peritoneal cavern. Afterward, the peritoneum and the skin were closed with continuous two-layer sutures. Sham-operated animals received only median laparotomy without perforation of the colon.
After the preparation procedure, animals were allowed to reconstitute for 30 min and woke up, and then samples for the first blood gas analyses were taken using ABL505 blood gas analyzer (Radiometer, Copenhagen, Denmark), and hemodynamic parameters were measured. Each collected amount of blood was substituted by an equal volume of 0.9% sodium chloride (Fresenius Kabi). In the following, each animal received 2 µg/100 g BW per hour fentanyl (Fagron, Barsbüttel, Germany), and all animals had free access to water and food. At the end of the experimental procedures, all animals were killed by barbiturate overdose.
Randomization of animals
Twelve hours after the operation, animals were randomized in the following groups. Sham-operated animals (n = 7), and one of the CASP groups received 0.5 mL DMSO and 4.5 mL H2O/12 h (n = 7). Another CASP group received 5 mL H2O/12 h (n = 7). Two other CASP groups were treated continuously with PD-4-I rolipram/12 h. As indicated above, 1 group received a high dose of 3 mg/kg body (n = 5) weight per hour and the other group a low dose of 1 mg/kg BW per hour (n = 6).
Animal preparation for monitoring of the microcirculation and microvascular leakage
Twenty-four hours after the operation, animals were reanesthetized, and tracheotomy was performed. Rats were then mechanically ventilated with using a rodent ventilator (type: 7025; Hugo Sachs Elektronic KG, March-Hugstetten, Germany) in a fixed ventilation regimen (FIO2 0.28 and respiratory rate of 75 breaths/min). After tracheotomy, anesthesia regimen was changed using midazolam (Midazolam-ratiopharm; Ratiopharm, Ulm, Germany) 0.7 mg/100 g BW per hour and fentanyl 7 µg/100 g BW per hour, and for a sufficient anesthetic depth 0.7 vol% isoflurane (without N2O) was adapted. Animals’ body temperature was always kept at 37°C using a heating plate.
Thereafter, relaparotomy was performed; the mesentery was gently taken out and spread over a pillar as has been described before (4). The whole experimental setup was then carefully placed under a modified inverted Zeiss microscope (Axiovert 200; Carl Zeiss, Göttingen, Germany) equipped with different lenses (Achroplan ×10 NA0.25/×20 NA0.4/×40 NA 0.6). This allowed continuous observation of microcirculatory flow within postcapillary venules in the mesenteric windows as well as observation of microcascular leakage following i.v. injection of FITC-albumin at 5 mg/100 g as described below. Images and videos were captured using digital camera ColorSnap CF driven by MetaMorph analysis software and digitally recorded for off-line analysis as described below. In the time course of experiments, the upper surface of the mesentery was continuously super perfused with 37.5°C crystalloid solution (Sterofundin; B.Braun Melsung AG).
Macrohemodynamic monitoring and blood gas analyses
Mean arterial pressure (MAP) and heart rate were continuously measured throughout the experiments and documented after the first operation (baseline); after 24 h, i.e., before relaparotomy (24 h); and before the end of the experiments (26 h). At similar time points blood gas samples were taken, at the end of the experimental procedures whole-blood samples were taken for laboratory investigation. A fluid resuscitation was not applied regardless of the MAP. Furthermore, an adaptation of fixed ventilation protocol was not performed to detect any possible changes, e.g., pulmonary failure and the effects of PD-4-I application.
Intravital measurement of capillary endothelial barrier properties
To assess changes of microvascular permeability during the experimental procedures, fluorescent digital images were taken after single i.v. injection of FITC-albumin 5 mg/100 g BW (Sigma, Deisenhofen, Germany) as described previously (4). In brief, fluorescent images were taken using a 100-W mercury lamp and a filter set consisting of a 450- to 490-nm excitation and a 520-nm emission filter within the inverted microscope (Zeiss microscope, Axiovert 200; Carl Zeiss). Microvascular permeability was then estimated by determining the extravasation of FITC-albumin by measurements of integrated optical intensity as described previously (9). Using Image J software, 10 randomly selected intravascular and interstitial areas near the postcapillary venules were selected for the measurements by a blinded observer. To compare changes between experimental groups, values measured from sham-operated animals were set to 100%.
Evaluation of microcirculatory flow
Analysis of microvascular blood flow was carried out in straight segments of venular microvessels (20–35 µm in diameter) as described before (4). Velocity of erythrocytes was measured by a blinded investigator using Software MetaMorph V 6.1r4. The velocities of 18 randomly chosen erythrocytes within postcapillary venules were measured for each animal, and six vessels per animal were analyzed. For each of these vessels, diameter was measured, and from these data, volumetric flow was calculated, as described previously (4); volumetric flow (Q V) was derived from the average velocity of free-passing red blood cells (V RBC; µm/ms) of each vessel assuming a cylindrical vessel geometry (Q V [pl / s] = [V RBC / 1.6] × [phi] × r 2).
Evaluation of blood samples and cytokine measurement
After vital microscopy plasma samples were drawn for determination of creatinine, potassium levels, leukocytes, coagulation parameters, and platelet count by using routine laboratory methods.
As described in detail previously (10) for cytokine evaluation, blood was centrifuged at 3,400g for 10 min at 4°C. Afterward, samples were frozen at −80°C. For quantification, a commercially available solid phase sandwich immunoassay for simultaneously quantifying multiple biomarkers with the Luminex method (“Rat 10-Plex”; Invitrogen, Karlsruhe, Germany) was used according to the manufacturer’s recommendations. Analyses were done in triplicate with a Luminex100 instrument (Luminex Corporation, Austin, Tex).
Histopathologic analysis of lungs
Lungs were removed after experimental procedures for histopathologic studies. The tissues were fixed in formaldehyde 3.5% (Otto Fischar GmbH, Saarbruecken, Germany) for more than 24 h. Tissues were then embedded in paraffin, and subsequently, sections were stained with hematoxylin-eosin for analyses of morphological alterations within the tissues as described previously. Sections were photographed using a Keyence BZ-9000 Microskop and BZ-II Analyzer software from Keyence (Keyence Corporation, Neu-Isenburg, Germany) was used to measure thickness of alveolar septa. Analysis was performed as described previously (4).
Values throughout are expressed as mean ± SEM. Statistical analyses were performed using SPSS 19.0 (IBM SPSS Software, Munich, Germany). For parametric parameters, possible differences were assessed with analysis of variance followed by post hoc Duncan test. Statistical significance is assumed for P < 0.05. For nonparametric data, Kruskal-Wallis test following Mann-Whitney U test with Bonferroni correction was used for significant differences.
All animals survived the experimental procedures, and therefore no animal had to be excluded in the time course of experiments. After relaparotomy, it could be visually confirmed that all animals that underwent CASP procedure showed signs of general fecal peritonitis. Concerning the extent of peritonitis, there were no macroscopic differences observed in CASP animals and CASP + PD-4-I animals. Animals that were treated with high dose of rolipram showed bloody tears at the end of the experimental procedures. Therefore, we stopped high-dose PD-4-I treatment after five animals in view of this severe adverse effect. At low-dose PD-4-I treatment, no obvious adverse effects were detected. After the end of the experimental procedures, all animals were killed as outlined above.
DMSO application had no adverse effects as revealed by comparison of CASP-treated animals
First, we compared CASP-treated animals that were given H2O or DMSO as vehicle (Fig. 1) to test for unexpected adverse effects of DMSO in which PD-4-I rolipram had to be dissolved. We found that CASP-treated animals that received DMSO or H2O, respectively, showed no differences in either macrohemodymic parameters or measurements of vascular permeability and microcirculatory flow in the time course of experiments. Similarly, no differences were observed when cytokine measurements were performed. This indicated that possible adverse effects of DMSO as used in our experimental setup were negligible. Therefore, in the following, CASP-treated animals were taken together and considered as 1 group.
Low-dose PD-4-I treatment led to stabilization of MAP and heart rate in CASP animals
At the beginning of the experimental procedure when baseline measurements were taken, no difference of MAP was observed within the different groups (Fig. 2). After 24 h, all animals that underwent CASP procedure showed a decrease in MAP indicating the onset of sepsis in these animals. After 26 h when animals had undergone 2 h of reanesthesia, tracheotomy, and mechanical ventilation, surgical procedures, i.e., relaparotomy and intravital microcirculation measurements, macrohemodynamic parameters of all animals including sham-operated showed a significant loss of MAP compared with baseline levels and with the situation after 24 h. This can be explained by systemic reactions because of the repeated surgical procedures and anesthesia. In addition, at this stage of the experiment, shift of the intestines to extracorporeal with mesenteric traction syndrome as well as fluid loss by surgery and perspiratio insensibilis may contribute to decreased MAP. Furthermore, it has to be considered that no fluid resuscitation was applied regardless of macrohemodynamic changes, which is in contrast to our previous experiments (4).
Animals with CASP treated with high-dose PD-4-I showed no significant differences compared with untreated CASP after 24 h, although these animals tended to have lower MAP compared with untreated CASP animals. Interestingly, animals treated with low-dose PD-4-I showed stabilization of MAP, which was not different compared with sham animals.
In sham animals, heart rate did not change significantly within the first 24 h of experiments. However, only at the end of the experimental procedures, we observed a significant drop of the heart rate in sham animals. In CASP animals, heart rate dropped earlier, i.e., at the time point of relaparotomy, whereas both high-dose and low-dose treatment with PD-4-I did not change heart rate in the time course of experiments.
Low-dose application of PD-4-I stabilized mesenteric microvascular barrier properties
We tested microvascular barrier properties in mesenteric postcapillary venules by measurement of FITC-albumin efflux. In the visual analysis of postcapillary venules of sham-operated animals, no obvious efflux following injection of FITC-albumin was observed (Fig. 3Aa). In contrast, CASP animals showed a significant extravasation of FITC-albumin in dicating microvascular leakage, i.e., loss of microvascular endothelial barrier functions after 24 h (Fig. 3Ab). Phosphodiesterase 4 inhibitor led to reduced extravasation of FITC-albumin at high dose (Fig. 3Ac) and low-dose treatment (Fig. 3Ad). Quantification of efflux in all animals confirmed the visual impression of significantly increased efflux of FITC-albumin in CASP animals. This was blocked by treatment of animals with PD-4-I at both doses. Unexpectedly quantification of FITC-albumin extravasation revealed that the barrier-stabilizing effect of PD-4-I treatment was more pronounced in animals treated with low-dose PD-4-I.
Sepsis-induced lung edema and interstitial cellular infiltration were reduced following PD-4-treatment
Analysis of hematoxylin-eosin–stained lung tissue sections showed severe lung edema as revealed by significantly increased thickness of alveolar septa in the CASP group (Fig. 4, Ab and B). This was accompanied with considerable interstitial cellular infiltration of neutrophils and hyperemia suggesting microvascular dysfunction. In contrast, PD-4-I application attenuated thickening of alveolar septa in CASP-induced sepsis. In addition, interstitial cellular infiltration and hyperemia were reduced (Fig. 4, Ac and d). Accordingly, oxygen parameters of spontaneously breathing animals were significantly improved in PD-4-I–treated animals compared with CASP and even to sham group 24 h after first operation (Table 1).
Microcirculation was significantly improved by low-dose treatment of PD-4-I
Next we performed measurements of microcirculatory flow in mesenteric postcapillary venules to test the hypothesis whether stabilization of microvascular endothelial barrier functions resulted in improvement of the microcirculation (Fig. 5). Volumetric flow in CASP animals was significantly reduced to 157 ± 7 pl/s compared with 220 ± 14 pl/s in sham animals, i.e., 71% of sham. Unexpectedly, treatment of animals with high-dose PD-4-I resulted in further reduction of microcirculatory flow to 93 ± 7 pl/s. In contrast, low-dose application of PD-4-I was effective to significantly improve microcirculation to 217 ± 10 pl/s, which was significantly higher when compared with CASP animals.
Metabolic changes in different experimental groups were most obvious at the end of the experimental procedures
We tested whether alterations of the microcirculation resulted in changes of parameters in blood gas analyses. We compared blood gas analyses taken immediately after the first operation, after 24 h, and after measurements of microvascular barrier functions and microcirculatory flow, i.e., after 26 h (Table 1). pH values were not altered in CASP-treated and in CASP + PD-4-I–treated animals when compared with sham after the first operation. After 24 h, all groups especially PD-4-I–treated animals shifted toward basic pH values, which was significant in low-dose PD-4-I–treated animals. In the latter groups, Pco2 values were decreased, suggesting hyperventilation in these animals. In addition, Po2 values decreased in sham and in CASP-treated animals, whereas PD-4-I–treated animals showed significantly improved oxygenation.
This changed after the second operation when all groups shifted to acidic pH values. The acidic shift with increasing lactate levels and HCO3 − consumption was most pronounced in the CASP group, although the differences reached no significance. Phosphodiesterase 4 inhibitor–treated animals were able to compensate better the metabolic acidosis by improved lung function. Furthermore, it must be emphasized that at this time point all animals were mechanically ventilated in a fixed ventilation regimen, and no adaptation of ventilation on blood gas parameters was performed to investigate the sole influence of PD-4-I treatment as far as possible. Therefore, Pco2 values shifted into a normal range, and oxygenation was improved in all animals, which reached significance for PD-4-I–treated animals.
In addition, we estimated kidney values, inflammation, and coagulation parameters in blood samples that were obtained at the end of the experimental procedures. In sham animals, creatinine was 0.45 ± 0.04 mg/dL, whereas CASP animals had augmented creatinine levels of 0.64 ± 0.04 mg/dL, indicating the manifestation of sepsis-induced acute kidney injury (AKI). This was substantiated by significantly elevated potassium levels in CASP animals (4.8 ± 0.12 vs. 5.7 ± 0.24 mmol/L). In high-dose PD-4-I–treated animals creatinine was augmented to 0.7 ± 0.06 mg/dL, whereas low-dose-PD-4-I–treated animals displayed no differences compared with sham (creatinine 0.47 ± 0.08 mg/dL). Accordingly, potassium was still increased in high-dose PD-4-I–treated animals (6.38 ± 0.43 mmol/L), whereas potassium in low-dose-PD-4-I–treated animals was 4.26 ± 0.18 mmol/L. Platelets were significantly reduced in CASP animals but not in animals treated with PD-4-I. Leukocytes and coagulation parameters were not different between experimental groups.
PD-4-I showed differential modulation of cytokine levels
Finally, we tested whether the previously described immunomodulatory effects of PD-4-I were also evident in our model. The early-phase cytokines, which are generally observed at the beginning of systemic inflammation, namely, interleukin 1α (IL-1α), IL-1β, and TNF-α, were differentially modulated (Fig. 6): IL-1α and IL-1β (Fig. 6, A and B) were elevated in CASP animals compared with sham animals. Although results did not reach significance, PD-4-I treatment appeared to decrease IL-1β secretion at high and low doses compared with CASP, whereas suppression of IL-1α appeared to occur only in high-dose PD-4-I–treated animals. Unexpectedly, TNF-α levels (Fig. 6C) were elevated and appeared to be higher in sham animals than in CASP animals. Compared with sham and CASP animals, TNF-α was significantly reduced in high-dose and low-dose PD-4-I–treated animals. Interleukin 6 (Fig. 6C) appeared increased in all groups and showed a trend to be augmented by PD-4-I treatment. Anti-inflammatory cytokine IL-10 (Fig. 6E) was significantly increased in the CASP group compared with sham animals. High-dose PD-4-I showed a slight increase, whereas low-dose PD-4-I appeared to induce a stronger increase in IL-10 levels. IL-12 (Fig. 6F), which induces T-helper cell 1 differentiation, was significantly reduced in PD-4-I–treated animals but not different between CASP and sham animals.
It is increasingly recognized that disruption of the endothelial barrier in septic patients is a critical event that determines breakdown of the microcirculation, organ failure, and death (2, 11–13). A specific therapy to stabilize microvascular barrier functions is not available at present. Previously, we suggested a clinically applicable approach in which systemic application of PD-4-I rolipram and roflumilast was effective to stabilize endothelial barrier functions and microcirculation followed by dramatically augmented survival in a model of systemic inflammation of rats (4). The present study extends our investigations in a long-term model of polymicrobial sepsis. We found that application of PD-4-I led to stabilization of the microvascular barrier and reduced lung injury. Interestingly, only low-dose application was effective to restore microcirculatory flow in CASP animals, whereas high-dose therapy had adverse effects and showed severe adverse effects such as bloody tear drop. Improved microcirculatory flow by low-dose application of PD-4-I prevented sepsis-induced AKI. Furthermore, PD-4-I application appeared to have immunomodulatory effects, which did not correlate directly with effects on microvascular barrier functions. The present data support the notion that systemic application of PD-4-I could be suitable for therapeutic microvascular barrier stabilization and improvement of microcirculatory flow in sepsis.
CASP-induced sepsis represents a clinically relevant animal model
In general, there are numerous models that were previously applied in sepsis research with several advantages and limitations for each of these models. In our own previously published study, we applied a defined dose of bacterial LPS intravenously (4). This model leads to a fast and severe systemic hyperinflammation that is comparable to the clinical situation of septic shock. Whereas the advantage of this model was that it was highly reproducible because a defined dose of LPS was applied, the disadvantage was that the rapid mortality of the animals made long-term observations impossible. Beside this, LPS-induced hyperinflammation does not represent a sepsis model in its clinically defined sense because there is a lack of a bacterial focus. Therefore, the cecum ligation and puncture model has been proposed as a suitable model to mimic polymicrobial sepsis (14). However, the cecum ligation and puncture model represents a model of intra-abdominal abscess formation with minor signs of systemic inflammation (14–16). Therefore, we chose the CASP model, which closely mimics the clinical course of diffuse peritonitis with continuously increasing systemic inflammation (15, 17). This is also documented in our present data, where CASP led to diffuse peritonitis in all animals after 24 h and induced severe critical illness as revealed by increased cytokine levels, hypotension, and loss of microvascular barrier properties followed by decreased microcirculatory flow. Furthermore, the CASP model reflects a two-hit model, in which reanesthesia and relaparotomy are crucial events for cardiopulmonary performance. However, a severe limitation of the model is the interindividual range of inflammatory response within the animals. Although the surgical procedures were highly standardized, it is impossible to predict the extent of critical illness. This may explain that some of the data obtained in this model showed only trends instead of significant differences.
Systemic long-term application of PD-4-I resulted in stabilization of microvascular endothelial barrier properties in the CASP model in a dose-dependent manner
There is a large body of evidence showing that increased cAMP levels in endothelial cells effectively stabilize endothelial barrier properties under resting conditions as well as in acute inflammation in vitro and in vivo (3). Based on preliminary work from in vitro and in vivo experiments (18, 19), the systemic application of PD-4-I to augment endothelial cell cAMP in a systemic model appeared to be a promising attempt to suggest a clinically applicable approach. In the present study, it was not tested again whether intravenously applied PD-4-I augments microvascular cAMP levels because this was done in our previous study in a comparable model with comparable doses of PD-4-I. In view of the similar biologic effect, i.e., microvascular barrier stabilization following PD-4-I application, the assumption that systemic application of PD-4-I in the CASP model also stabilized microvascular barrier functions by augmented endothelial cAMP levels is warranted.
The stabilization of the endothelial barrier was observed at both doses used in our experimental setup, although the low-dose application appeared to be even more effective than the high-dose treatment of CASP animals. Similarly, low-dose PD-4-I treatment resulted in increased microvascular flow. This can be interpreted as a consequence of cAMP-mediated endothelial barrier stabilization, which in turn augmented intravascular volume. However, improved microvascular flow resulted in only a trend for less pronounced metabolic acidosis in animals treated with low-dose PD-4-I. Nonetheless, the beneficial effects of low-dose PD-4-I are underlined by the fact that early signs of sepsis-induced multiple organ failure with AKI and acute respiratory distress syndrome in CASP animals were abolished in these animals.
In general, the previous experimental use of PD-4-I rolipram to stabilize endothelial barrier functions was restricted to local, inhalative, or intraperitoneal application (5, 18–23). Rolipram administered orally has already been tested in clinical trials in order to treat depression, neuroimmunologic diseases or chronic obstructive pulmonary disease (24, 25). However, to our knowledge, experiences using PD-4-I intravenously are not available to date. The clinical use of phosphodiesterase inhibitors in sepsis such as PD-3-I inhibitor enoximon or other cAMP-increasing medications such as β-adrenoreceptor mimetica (i.e., dobutamine) was considered unsuitable because these agents induced severe adverse effects such as excessive vasodilatation, tachyarrhythmia, and loss of blood pressure (26, 27). Adverse effects such as vasodilation followed by excessive hypotension was not observed in low-dose PD-4-I treatments. The stabilizing effect of low-dose PD-4-I on MAP after 26 h can be interpreted as a result of higher intravascular volume in response to improved microvascular barrier functions. As outlined above, the reason why this observation was evident only at the end of the experimental procedure may be explained by the fact that anesthesia in addition to the inflammatory response and surgery may have led to additional loss of MAP. Therefore, the presence of intravascular volume in the groups may be more distinct at this time point. In addition, positive inotropic effects by PD-4-I application cannot be excluded (27).
It must be emphasized that high-dose application of PD-4-I resulted in bloody tear drop of the animals as a severe adverse effect so that we ended this trial. Also, after 26 h, we observed an additional loss of MAP in the animals treated with high-dose PD-4-I, which may argue for a vasodilative effect at higher doses. In our previous study, high-dose rolipram did not lead to severe adverse effects and was highly effective to stabilize endothelial barrier properties and microcirculation (4). This difference may be explained by the duration of PD-4-I application at this dose. In principle, it must be considered that toxic effects of DMSO in which rolipram had to be dissolved could contribute to these adverse effects. However, because CASP animals that were given water or DMSO, respectively, showed no differences at all, it is unlikely that DMSO made a significant contribution to the adverse effects of high-dose PD-4-I. Rather, it must be concluded that the therapeutic index of intravenously applied rolipram is small. Therefore, in case of clinical application, detailed dose effects relations of intravenously applied rolipram will have to be compiled. Furthermore, a suitable chemical way for possible i.v. application of rolipram in human remains to be developed.
PD-4-I–induced immunomodulatory effects did not correlate with endothelial barrier alterations in the CASP model
It is known that PD-4-I rolipram and other phosphodiesterase inhibitors have significant effects on cytokine expression in different models in vitro and in vivo (6, 21, 28, 29). In this context, it was shown that rolipram effectively stabilized immune cells, which resulted in reduced levels of proinflammatory cytokines such as TNF-α. Therefore, the previously described beneficial effects of PD-4-I treatment in different models of inflammation have primarily been attributed to these immunomodulatory effects. The aspect of immunomodulation by rolipram is clearly confirmed by our present data in a model polymicrobial sepsis. At first glance, it may be concluded that endothelial barrier stabilization and improved microcirculation in our study can be explained by immunomodulatory effects rather than direct effects on microvascular endothelial barrier functions alone. However, in view of the fact that LPS decreased cAMP in intact microvessels in vivo and loss of cAMP-levels was abrogated by systemic application of rolipram, a direct effect must be assumed (4).
In the present study, we decided to look at cytokines that are known to be associated with endothelial barrier breakdown and CASP-induced sepsis (8, 30, 31). Interestingly, the effects on endothelial barrier protection and improved microcirculation did not correlate directly with immunomodulatory effects of rolipram. This was especially the case for early-phase cytokine TNF-α, which is well known to induce endothelial barrier disruption (32–34) because TNF-α was even higher in sham animals in which endothelial barrier properties were intact. Tumor necrosis factor α was blocked by PD-4-I treatment. However, IL-1α suppression was observed only in high-dose PD-4-I–treated animals where PD-4-I treatment appeared to be less effective to stabilize endothelial barrier properties. Interleukin 6, which is also a potent inductor of microvascular leakage (30), tended to be higher in CASP- and PD-4-I–treated animals. A lack of correlation between levels of cytokines and endothelial barrier properties in our study can also be drawn for the other cytokines estimated. This interesting observation may be explained in part by the fact that bacterial cell wall components such as LPS that occur in the blood at different levels during sepsis are known to induce endothelial activation and barrier disruption without the need for cytokine release (18). Furthermore, other factors such as significant release of barrier-comprising angiopoietin 2 in sepsis may play a more important role for endothelial barrier disruption than cytokines (35). In conclusion, although there is a large number of factors that were described to affect endothelial barrier functions in inflammation, the predominant factor that induces disruption of the endothelial barrier in sepsis remains unclear. Nonetheless, a potentially beneficial role of altered cytokine release by rolipram cannot be negotiated from our data.
In summary, these findings warrant further studies to establish the potential clinical use of PD-4-I to treat microvascular endothelial barrier disruption in sepsis.
The authors thank Alexia Witchen, Veronika Heimbach and Mariola Dragan for skilful technical assistance.
1. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, et al.: Surviving Sepsis
Campaign: international guidelines for management of severe sepsis
and septic shock: 2012. Crit Care Med 41 (2): 580–637, 2012.
2. Lee WL, Slutsky AS: Sepsis
and endothelial permeability. N Engl J Med 7: 689–691, 2010.
3. Schlegel N, Waschke J: cAMP with other signaling cues converges on Rac1 to stabilize the endothelial barrier
- a signaling pathway compromised in inflammation. Cell Tissue Res 355 (3): 587–596, 2014.
4. Schick MA, Wunder C, Wollborn J, Roewer N, Waschke J, Germer CT, Schlegel N: Phosphodiesterase-4 inhibition as a therapeutic approach to treat capillary leakage in systemic inflammation. J Physiol 590 (Pt 11): 2693–2708, 2012.
5. Lin YC, Adamson RH, Clark JF, Reed RK, Curry FR: Phosphodiesterase 4 inhibition attenuates plasma volume loss and transvascular exchange in volume-expanded mice. J Physiol 590 (Pt 2): 309–322, 2012.
6. Irie K, Fujii E, Ishida H, Wada K, Suganuma T, Nishikori T, Yoshioka T, Muraki T: Inhibitory effects of cyclic AMP elevating agents on lipopolysaccharide (LPS)–induced microvascular permeability change in mouse skin. Br J Pharmacol 133 (2): 237–242, 2001.
7. Zantl N, Uebe A, Neumann B, Wagner H, Siewert JR, Holzmann B, Heidecke CD, Pfeffer K: Essential role of gamma interferon in survival of colon ascendens stent peritonitis, a novel murine model of abdominal sepsis
. Infect Immun 66 (5): 2300–2309, 1998.
8. Lustig MK, Bac VH, Pavlovic D, Maier S, Grundling M, Grisk O, Wendt M, Heidecke CD, Lehmann C: Colon ascendens stent peritonitis—a model of sepsis
adopted to the rat: physiological, microcirculatory and laboratory changes. Shock 28 (1): 59–64, 2007.
9. Bekker AY, Ritter AB, Duran WN: Analysis of microvascular permeability to macromolecules by video-image digital processing. Microvasc Res 38 (2): 200–216, 1989.
10. Schick MA, Isbary JT, Stueber T, Brugger J, Stumpner J, Schlegel N, Roewer N, Eichelbroenner O, Wunder C: Effects of crystalloids and colloids on liver and intestine microcirculation
and function in cecal ligation and puncture induced septic rodents. BMC Gastroenterol 12: 179, 2012.
11. Cinel I, Dellinger RP: Advances in pathogenesis and management of sepsis
. Curr Opin Infect Dis 20 (4): 345–352, 2007.
12. Lee WL, Liles WC: Endothelial activation, dysfunction and permeability during severe infections. Curr Opin Hematol 18 (3): 191–196, 2011.
13. Goldenberg NM, Steinberg BE, Slutsky AS, Lee WL: Broken barriers: a new take on sepsis
pathogenesis. Sci Transl Med 3 (88): 88ps25, 2011.
14. Dejager L, Pinheiro I, Dejonckheere E, Libert C: Cecal ligation and puncture: the gold standard model for polymicrobial sepsis
? Trends Microbiol 19 (4): 198–208, 2011.
15. Traeger T, Koerner P, Kessler W, Cziupka K, Diedrich S, Busemann A, Heidecke CD, Maier S: Colon ascendens stent peritonitis (CASP
)—a standardized model for polymicrobial abdominal sepsis
. J Vis Exp (46): e2299, 2010.
16. Schick M, Isbary T, Schlegel N, Brugger J, Waschke J, Muellenbach R, Roewer N, Wunder C: The impact of crystalloid and colloid infusion on the kidney in rodent sepsis
. Intensive Care Med 36 (3): 541–548, 2010.
17. Maier S, Traeger T, Entleutner M, Westerholt A, Kleist B, Huser N, Holzmann B, Stier A, Pfeffer K, Heidecke CD: Cecal ligation and puncture versus colon ascendens stent peritonitis: two distinct animal models for polymicrobial sepsis
. Shock 21 (6): 505–511, 2004.
18. Schlegel N, Baumer Y, Drenckhahn D, Waschke J: Lipopolysaccharide-induced endothelial barrier
breakdown is cAMP dependent in vivo
and in vitro
. Crit Care Med 37 (5): 1735–1743, 2009.
19. Lin YC, Samardzic H, Adamson RH, Renkin EM, Clark JF, Reed RK, Curry FR: Phosphodiesterase 4 inhibition attenuates atrial natriuretic peptide-induced vascular hyperpermeability and loss of plasma volume. J Physiol 589 (Pt 2): 341–353, 2011.
20. Adamson RH, Liu B, Fry GN, Rubin LL, Curry FE: Microvascular permeability and number of tight junctions are modulated by cAMP. Am J Physiol 274 (6 Pt 2): H1885–H1894, 1998.
21. Krause W, Kuhne G: Anti-inflammatory activity of rolipram
in a rat ear edema model. Arzneimittel Forschung 44 (2): 163–165, 1994.
22. Spindler V, Waschke J: beta-Adrenergic stimulation contributes to maintenance of endothelial barrier
functions under baseline conditions. Microcirculation
18 (2): 118–127, 2011.
23. de Visser YP, Walther FJ, Laghmani EH, van Wijngaarden S, Nieuwland K, Wagenaar GT: Phosphodiesterase-4 inhibition attenuates pulmonary inflammation in neonatal lung injury. Eur Respir J 31 (3): 633–644, 2008.
24. Houslay MD, Schafer P, Zhang KY: Keynote review: phosphodiesterase-4 as a therapeutic target. Drug Discov Today 10 (22): 1503–1519, 2005.
25. Bielekova B, Richert N, Howard T, Packer AN, Blevins G, Ohayon J, McFarland HF, Sturzebecher CS, Martin R: Treatment with the phosphodiesterase type-4 inhibitor rolipram
fails to inhibit blood-brain barrier disruption in multiple sclerosis. Mult Scler 15 (10): 1206–1214, 2009.
26. Lehtonen LA, Antila S, Pentikainen PJ: Pharmacokinetics and pharmacodynamics of intravenous inotropic agents. Clin Pharmacokinet 43 (3): 187–203, 2004.
27. Bangash MN, Kong ML, Pearse RM: Use of inotropes and vasopressor agents in critically ill patients. Br J Pharmacol 165 (7): 2015–2033, 2012.
28. Holthoff JH, Wang Z, Patil NK, Gokden N, Mayeux PR: Rolipram
improves renal perfusion and function during sepsis
in the mouse. J Pharmacol Exp Ther 347 (2): 357–364, 2013.
29. Goncalves de Moraes VL, Singer M, Vargaftig BB, Chignard M: Effects of rolipram
on cyclic AMP levels in alveolar macrophages and lipopolysaccharide-induced inflammation in mouse lung. Br J Pharmacol 123 (4): 631–636, 1998.
30. Henrich M, Gruss M, Weigand MA: Sepsis
-induced degradation of endothelial glycocalix. Sci World J 10: 917–923, 2010.
31. Wei Z, Yang J, Xia YF, Huang WZ, Wang ZT, Dai Y: Cardamonin protects septic mice from acute lung injury by preventing endothelial barrier
dysfunction. J Biochem Mol Toxicol 26 (7): 282–290, 2012.
32. Koga S, Morris S, Ogawa S, Liao H, Bilezikian JP, Chen G, Thompson WJ, Ashikaga T, Brett J, Stern DM, et al.: TNF modulates endothelial properties by decreasing cAMP. Am J Physiol 268 (5): C1104–C1113, 1995.
33. Seybold J, Thomas D, Witzenrath M, Boral S, Hocke AC, Brger A, Hatzelmann A, Tenor H, Schudt C, Krll M, et al.: Tumor necrosis factor-alpha-dependent expression of phosphodiesterase 2: role in endothelial hyperpermeability. Blood 105 (9): 3569–3576, 2005.
34. Schlegel N, Waschke J: Impaired cAMP and Rac 1 signaling contribute to TNF-alpha–induced endothelial barrier
breakdown in microvascular endothelium. Microcirculation
16 (6): 521–533, 2009.
35. Moss A: The angiopoietin:Tie 2 interaction: a potential target for future therapies in human vascular disease. Cytokine Growth Factor Rev 24 (6): 579–592, 2013.
Keywords:© 2014 by the Shock Society
Endothelial barrier; microcirculation; sepsis; phosphodiesterase inhibitor; rolipram; sepsis; cytokines; CASP