Yves Ouellette

Endotoxin increases intercellular resistance in microvascular endothelial cells by a tyrosine kinase pathway.

Gap junction communication between microvascular endothelial cells has been proposed to contribute to the coordination of microvascular function. Septic shock may attenuate microvascular cell‐to‐cell communication. We hypothesized that lipopolysaccharide (LPS) attenuates communication between microvascular endothelial cells derived from rat hindlimb skeletal muscle. Endothelial cells grown in monolayers expressed mRNA for connexin 37, 40, and 43. The expression of connexin 43 protein was confirmed, but connexin 40 protein was not detected by immunocytochemistry or immunoblot analysis. Intercellular resistance between cells of the monolayer, calculated using a Bessel function model, was increased from 3.3 to 5.3 MΩ by LPS. The effect was seen after 1 h of exposure and required a minimum concentration of 10 ng/ml. Intercellular resistance returned to normal 1 h following removal of LPS. Neither the response to LPS, nor its reversal, was blocked by the protein synthesis inhibitor cycloheximide (10 μg/ml). Pretreatment of monolayers with the tyrosine kinase inhibitors PP‐2 (10 nM), lavendustin‐C (1 μM), and geldanamycin (200 nM) prevented this LPS response; geldanamycin was also able to reverse the response. Inhibitors of MAP kinases, PD 98059 (5 μM) and SB 202190 (5 μM), and PKC (500 nM bisindolylmaleimide I) were unable to block the LPS response. We propose that LPS attenuates cell‐to‐cell communication through a signaling pathway that is tyrosine kinase dependent. J. Cell. Physiol. 185:117–125, 2000.

Lipopolysaccharide-induced reductions in cellular coupling correlate with tyrosine phosphorylation of connexin 43

We have previously shown in cultured rat microvascular endothelial cells (RMEC) that lipopolysaccharide (LPS) stimulates a protein tyrosine kinase (PTK)‐dependent reduction in cellular coupling. We hypothesized that connexin 43 (Cx43) becomes phosphorylated following exposure to LPS. Cx43 was immunoprecipitated from control and LPS‐treated RMEC monolayers. Tyrosine phosphorylation of Cx43, detected by immunoblot, was found only in the LPS treatment. To verify these results, Cx43 was radiolabeled with [32P]‐orthophosphate. Radiolabeled Cx43 exhibited a slight increase in phosphorylation in response to LPS; phosphoamino acid analysis displayed equivalent amounts of phosphoserine in control and LPS treatments, but detected phosphotyrosine only in the LPS treatment. The PTK inhibitors PP‐2 (10 nM) and geldanamycin (200 nM) were found to block the response to LPS in terms of Cx43 tyrosine phosphorylation and cellular coupling. The phosphatase inhibitor BpV (1 μM) accentuated the effect of LPS, while the putative phosphatase activator C6‐ceramide prevented it. When measuring cell communication, phosphatase inhibition also blocked the reversal of the LPS response following LPS washout. We conclude that Cx43 is tyrosine phosphorylated following exposure to LPS and suggest that the LPS‐induced increase in intercellular resistance may be mediated by tyrosine phosphorylation of this connexin. Altering tyrosine kinase and phosphatase activities can modulate the LPS‐induced tyrosine phosphorylation of Cx43 and reductions in cellular coupling. J. Cell. Physiol. 193: 373–379, 2002.

Conducted Vasoconstriction Is Reduced in a Mouse Model of Sepsis

The ability of an arteriole to conduct vasomotor responses along its length contributes to the control of organ perfusion. Sepsis, a systemic inflammatory response to infection, may compromise this control. We aimed to determine whether sepsis, induced by cecal ligation and perforation (CLP), reduces conducted vasoconstriction 24 h post-CLP. We locally stimulated mouse cremaster arterioles with KCl, measured the resulting local and the conducted constriction (500 µm upstream) and, based on these measurements, determined the communication ratio (CR500) as an index of the conducted response. Sepsis significantly reduced the CR500 from 0.75 to 0.20. Based on a mathematical model, this reduction was predicted to have a significant impact on blood flow control. In septic mice, either a 1-hour washout of the cremaster muscle with physiological saline or a treatment of this muscle with the tyrosine kinase inhibitor PP-2 (100 nM) restored the CR500 to the control level. Treatment of septic arterioles with the nitric oxide synthase inhibitor Nω-nitro-L-arginine methyl ester (100 µM) partially restored the CR500 from 0.2 to 0.4. In control mice, lipopolysaccharide (LPS; 10 µg/ml) superfused over the cremaster muscle for 1 h reduced the CR500; the nitric oxide (NO) donor S-nitroso-N-acetyl-penicillamine (50 µM) also reduced the CR500. Thus, LPS and NO could be two factors mediating reduced conduction of vasoconstriction in sepsis. We conclude that sepsis reduces the KCl-induced conducted vasoconstriction in the mouse cremaster muscle by a tyrosine kinase- and nitric oxide- dependent mechanism.

Abrupt reoxygenation following hypoxia reduces electrical coupling between endothelial cells of wild-type but not connexin40 null mice in oxidant- and PKA-dependent manner

Although electrical coupling along the arteriolar endothelium is central in arteriolar conducted response and in control of vascular resistance, little is known about the pathophysiological effect of hypoxia and reoxygenation (H/R) on this coupling. We examined this effect in a monolayer of cultured microvascular endothelial cells (ECs) derived from wild‐type (WT) or connexin (Cx)40−/− mice (Cx40 is a key gap junction protein in ECs). To assess electrical coupling, we used a current injection technique and Bessel function model to compute the monolayer intercellular resistance. Hypoxia (0.1% O2, 1 h) followed by abrupt reoxygenation (5–90 min) reduced coupling (i.e., increased resistance) in WT but not in Cx40−/− monolayer. H/R increased superoxide production and reduced protein kinase A (PKA) activity in both monolayers. Activation of PKA by 8‐bromo‐cAMP prevented the reduction in coupling. Preloading of the WT monolayer with the antioxidant ascorbate prevented reductions in both PKA activity and cell coupling. Inhibition of PKA with 6–22 amide during normoxia mimicked the reduction in coupling. Finally, hypoxia followed by slow reoxygenation caused no change in superoxide level, PKA activity, or coupling. Using intravital microscopy, we assessed the physiological relevance of these findings in terms of KCl‐induced conducted vasoconstriction in arterioles of WT mouse cremaster muscle in vivo. Ischemia (1 h) followed by abrupt reperfusion (15–30 min) reduced conduction. 8‐bromo‐cAMP prevented this reduction, while 6–22 amide mimicked this reduction in control nonischemic arterioles. We propose that abrupt reoxygenation reduces interendothelial electrical coupling via oxidant‐ and PKA‐dependent signaling that targets Cx40. We suggest that this mechanism contributes to compromised arteriolar function after H/R.

Communication of Agonist-Induced Electrical Responses along ‘Capillaries’ in vitro Can Be Modulated by Lipopolysaccharide, but Not Nitric Oxide

Communication of agonist-induced membrane potential changes along blood vessels has been proposed to contribute to the coordination of microvascular function. Factors mediating septic shock may compromise this coordination. Using electrophysiology in a simplified in vitro model of endothelial cells grown as capillary-like structures, we aimed to determine (i) the effect of lipopolysaccharide (LPS) on endothelial cell membrane potential responses to ATP and KCl and (ii) the effect of LPS and nitric oxide (NO) on cell-to-cell communication. Treatment of ‘capillaries’ with LPS (10 µg/ml for 1 h) did not affect local responsiveness to ATP or KCl, but reduced cell communication by a tyrosine-kinase-dependent mechanism. Treatment of ‘capillaries’ with the NO donor DETA (100 µM) or the NO synthase inhibitor L-NAME (100 µM) had no effect on cell communication or the response to LPS. Endogenous NO production, stimulated by LPS + interferon-γ (100 U/ml) treatment, also had no effect on cell communication beyond that of LPS alone. We conclude that LPS, but not NO, can modulate conduction of agonist-induced electrical responses along endothelial capillary-like structures in vitro.

Implementation of patient-centered bedside rounds in the pediatric intensive care unit.

Implementation of effective family-centered rounds in an intensive care unit environment is fraught with challenges. We describe the application of PDSA (Plan, Do, Study, Act) cycles in a quality improvement project to improve the process of rounds and increase family participation and provider satisfaction. We conducted pre-/postintervention surveys and used 5 process measures for a total of 1296 daily patient rounds over 7 months. We were successful in conducting family-centered rounds for 90% of patients, with 40% family participation and a 64.6% satisfactory rating by pediatric intensive care unit providers.

Abstract P-585: IMPROVEMENT IN OXYGENATION UTILIZING TRANSPULMONARY PRESSURE MONITORING IN PEDIATRIC-ARDS

Case: A 2 year-old female (10.2 kg) with Trisomy 21 developed severe P-ARDS secondary to Human Metapneumovirus pneumonia, complicated by superimposed bacterial pneumonia. She required intubation on Hospital Day (HD) 2. Her oxygenation was poor (FiO2 60–70%), with an elevated Oxygenation Index (OI) of 18. PEEP was incrementally increased from 8 to 12 cmH2O with constant tidal volumes of 6 mL/kg. Her oxygenation remained labile on HD 3–4. On HD 5, the OI worsened to 30.

Effect of Ondansetron on QT Interval in Patients Cared for in the PICU.

Objectives: There is no evidence regarding the effect of ondansetron on the QT interval in pediatric patients in the ICU. This study aimed to describe the effect of ondansetron on the corrected QT interval in patients cared for in the PICU. Design: Retrospective cohort, consecutive enrollment study. Setting: Single-center, tertiary-level, medical/surgical PICU. Patients: All patients less than 8 years old who received ondansetron over an 11-month period were included. Exclusion criteria were atrial arrhythmia, bundle-branch block, known congenital long QT syndrome, and concomitant administration of proarrhythmic antiarrhythmic agents. Interventions: None. Measurements and Main Results: Overall, 210 doses of ondansetron were administered to 107 patients, with a mean age 10.5 ± 4.8 years; 49% were men. Corrected QT interval increased to 460–500 ms in 29% and to more than 500 ms in 11% of events of ondansetron administration. The mean baseline corrected QT interval even before ondansetron administration was higher for these groups (460–500 and > 500 ms; 457 ± 33 and 469 ± 45, respectively; p ⩽ 0.05). In multivariate analysis, both groups were associated significantly with underlying electrolyte abnormalities (odds ratio, 2.2; 95% CI, 1.1–4.4 and odds ratio, 5.1; 95% CI, 1.8–15.7, respectively); the group with corrected QT interval more than 500 ms was also significantly associated with organ dysfunction (odds ratio, 3.2; 95% CI, 1.1–9.4). As the numbers of risk factors increased from only ondansetron to three additional QT aggravating factors (electrolyte abnormalities, administration of other QT-prolonging drugs, and organ dysfunction), the likelihood of being associated with corrected QT interval more than 500 ms increased. Conclusions: Prolonged QT interval is observed commonly in PICUs following the administration of ondansetron. Underlying risk factors, such as electrolyte abnormalities and organ dysfunction, seem to pose the highest risk of prolongation of QT interval in these patients. The awareness of prevalent risk factors for increased corrected QT interval may help identify patients at high risk for arrhythmias.