Recombinant factor VIIa reduces transfusion requirements in liver transplant patients with high MELD scores.

Patients undergoing orthotopic liver transplantation (OLT) often experience significant coagulopathy and remain at risk for excessive blood loss and massive transfusion. The ability of recombinant factor VIIa (rFVIIa) to reduce transfusion requirements during OLT has not been well established. This retrospective study investigates whether rFVIIa reduces transfusion requirements in liver transplant patients with a significantly prolonged prothrombin time (PT) and a model of end-stage liver disease (MELD) score of > 20. Eleven patients received a single dose of rFVIIa (58 +/- 18 microg kg(-1)) at the time of incision. This group was matched with a selected control group that fulfilled all of the inclusion/exclusion criteria. Patient characteristics, pre-operative PT, HCT, PLT and MELD were identical between groups. Prophylactic application of rFVIIA reduced packed red blood cells (3.9 +/- 2.6 versus 6.9 +/- 2.3 U, P = 0.01) and fresh-frozen plasma (FFP) (12.6 +/- 6 versus 19.8 +/- 7 U, P = 0.018) transfusion requirements when compared with the control group. FFP administration in the first 24 h after surgery was also significantly less in the rVIIa group when compared with the control group (388 +/- 385 versus 1225 +/- 701 mL, P = 0.003). Hospital stay following transplantation tended to be shorter in the rFVIIa group, albeit statistical significance was not achieved (11 +/- 7.3 versus 7.9 +/- 2.7, P = 0.2). All but one patient in the control group survived for 30 days after transplantation. In a selected group of patients with prolonged PT and high MELD score, the prophylactic application of rFVIIa at the start of the OLT may reduce perioperative transfusion requirements.

Protective effects of low tidal volume ventilation in a rabbit model of Pseudomonas aeruginosa-induced acute lung injury.

OBJECTIVE: To determine whether low "stretch" mechanical ventilation protects animals from clinical sepsis after direct acute lung injury with Pseudomonas aeruginosa as compared with high "stretch" ventilation. DESIGN: Prospective study. SETTING: Experimental animal laboratory. SUBJECTS: Twenty-seven anesthetized and paralyzed rabbits. INTERVENTIONS: P. aeruginosa (109 colony forming units) was instilled into the right lungs of rabbits that were then ventilated at a tidal volume of either 15 mL/kg (n = 11) or 6 mL/kg (n = 7) for 8 hrs. Control animals were ventilated at a tidal volume of either 15 mL/kg (n = 4) or 6 mL/kg (n = 5) for 8 hrs, but an instillate without bacteria was used. A positive end-expiratory pressure of 3-5 cm H2O was used for all experiments. Radiolabeled albumin was used as a marker of alveolar epithelial permeability. MEASUREMENTS AND MAIN RESULTS: Hemodynamics, arterial blood gas determination, alveolar permeability, wet-to-dry ratios on lungs, and time course of bacteremia were determined. When final values were compared with the values at the beginning of the experiment, there were significant decreases in mean arterial pressure (from 104 +/- 15 to 57 +/- 20 mm Hg), pH (from 7.46 +/- 0.04 to 7.24 +/- 15), Pao2 (from 528 +/- 35 to 129 +/- 104 torr [70.4 +/- 4.7 to 17.2 +/- 13.9 kPa]), and temperature (from 38.2 +/- 1 to 36.2 +/- 1.2 degrees C) in the high tidal volume group, whereas no significant differences were found in the low tidal volume group. Decreased alveolar permeability was shown in the low tidal volume group, as was decreased extravascular lung water in the uninstilled lung in the low tidal volume group (12.7 +/- 2.5 vs. 4.3 +/- 0.45 g H2O/g dry lung). No noteworthy difference was noted in the time course of bacteremia, although there was a trend toward earlier bacteremia in the high tidal volume group. CONCLUSIONS: In our animal model of P. aeruginosa-induced acute lung injury, low tidal volume ventilation was correlated with improved oxygenation, hemodynamic status, and acid-base status as well as decreased alveolar permeability and contralateral extravascular lung water.

Etiology of acute pulmonary edema during liver transplantation : a series of cases with analysis of the edema fluid.

STUDY OBJECTIVES: To describe the clinical features of a group of patients who acutely developed pulmonary edema during orthotopic liver transplantation and to determine the nature (transudate vs exudate) of the edema. DESIGN: Retrospective review of clinical records and radiographic studies. SETTING: Operating room and ICU of a tertiary-care medical center hospital. PATIENTS: End-stage liver disease patients undergoing orthotopic liver transplantation under general anesthesia. INTERVENTIONS AND MEASUREMENTS: Pulmonary edema fluid obtained from seven patients within 15 min of first appearance was analyzed for protein content and compared with the protein content of a simultaneously obtained plasma sample. Hemodynamic data, fluid administration totals, and length of postoperative intubation and ICU stay were also collected. RESULTS: Eight patients were identified. Six of the seven patients whose edema fluid was analyzed had edema fluid/plasma protein ratios > or =0.75, characteristic of increased permeability pulmonary edema (the one other patient had a ratio of 0.73). Hemodynamic monitoring at the time of onset of the edema effectively ruled out a cardiogenic etiology. One patient died intraoperatively; at autopsy, the cause of death was determined to be pulmonary fat embolization. In the other seven patients, production of edema fluid resolved within 6 h of admission to the ICU. The duration of ventilatory support ranged from 23 to 96 h, with a mean of 49 h. CONCLUSIONS: The most likely cause of the reaction is transfusion-related acute lung injury (TRALI). An incidence of TRALI that is higher than previously reported in this population indicates that other elements, such as reperfusion of the newly implanted liver, may be contributing factors.

Additive inhibition of nicotinic acetylcholine receptors by corticosteroids and the neuromuscular blocking drug vecuronium.

BACKGROUND: Neuromuscular disorders associated with muscular weakness and prolonged paralysis are common in critically ill patients. Acute myopathy has been described in patients receiving a combination therapy of corticosteroids and nondepolarizing neuromuscular blocking drugs for treatment of acute bronchospasm. The cause of this myopathy is not fully established and may involve drug interactions that perturb neuromuscular transmission. To investigate the interaction of corticosteroids with neuromuscular blocking drugs, the authors determined the effects of methylprednisolone and hydrocortisone alone and in combination with vecuronium on fetal (gamma-subunit containing) and adult (epsilon-subunit containing) subtypes of the muscle-type nicotinic acetylcholine receptor. METHODS: Functional channels were expressed in Xenopus laevis oocytes and activated with 1 microM acetylcholine. The resulting currents were recorded using a whole cell two-electrode voltage clamp technique. RESULTS: Both forms of the muscle-type acetylcholine receptor were potently inhibited by methylprednisolone and hydrocortisone, with concentrations producing 50% inhibition in the range of 400-600 microM and 1-2 mM, respectively. The corticosteroids produced noncompetitive antagonism of the muscle-type nicotinic acetylcholine receptor at clinical concentrations. Both receptor forms were also inhibited, even more potently, by vecuronium, with a concentration producing 50% inhibition in the range of 1-2 nM. Combined application of vecuronium and methylprednisolone showed additive effects on both receptor forms, which were best described by a two-site model, with each site independent. CONCLUSIONS: The enhanced neuromuscular blockade produced when corticosteroids are combined with vecuronium may augment pharmacologic denervation and contribute to the pathophysiology of prolonged weakness observed in some critically ill patients.

Pathogenesis of septic shock in Pseudomonas aeruginosa pneumonia.

The pathogenesis of septic shock occurring after Pseudomonas aeruginosa pneumonia was studied in a rabbit model. The airspace instillation of the cytotoxic P. aeruginosa strain PA103 into the rabbit caused a consistent alveolar epithelial injury, progressive bacteremia, and septic shock. The lung instillation of a noncytotoxic, isogenic mutant strain (PA103DeltaUT), which is defective for production of type III secreted toxins, did not cause either systemic inflammatory response or septic shock, despite a potent inflammatory response in the lung. The intravenous injection of PA103 did not cause shock or an increase in TNF-alpha, despite the fact that the animals were bacteremic. The systemic administration of either anti-TNF-alpha serum or recombinant human IL-10 improved both septic shock and bacteremia in the animals that were instilled with PA103. Radiolabeled TNF-alpha instilled in the lung significantly leaked into the circulation only in the presence of alveolar epithelial injury. We conclude that injury to the alveolar epithelium allows the release of proinflammatory mediators into the circulation that are primarily responsible for septic shock. Our results demonstrate the importance of compartmentalization of inflammatory mediators in the lung, and the crucial role of bacterial cytotoxins in causing alveolar epithelial damage in the pathogenesis of acute septic shock in P. aeruginosa pneumonia.

Active and passive immunization with the Pseudomonas V antigen protects against type III intoxication and lung injury.

Pseudomonas aeruginosa is an opportunistic bacterial pathogen that can cause fatal acute lung infections in critically ill individuals. Damage to the lung epithelium is associated with the expression of toxins that are directly injected into eukaryotic cells through a type Ill-mediated secretion and translocation mechanism. Here we show that the P. aeruginosa homolog of the Yersinia V antigen, PcrV, is involved in the translocation of type III toxins. Vaccination against PcrV ensured the survival of challenged mice and decreased lung inflammation and injury. Antibodies to PcrV inhibited the translocation of type III toxins.

Evaluation of antimicrobial and lipopolysaccharide-neutralizing effects of a synthetic CAP18 fragment against Pseudomonas aeruginosa in a mouse model.

CAP18 (cationic antimicrobial protein; 18 kDa) is a neutrophil-derived protein that can bind to and inhibit various activities of lipopolysaccharide (LPS). The 37 C-terminal amino acids of CAP18 make up the LPS-binding domain. A truncated 32-amino-acid C-terminal fragment of CAP18 had potent activity against Pseudomonas aeruginosa in vitro. We studied the antimicrobial and LPS-neutralizing effects of this synthetic truncated CAP18 peptide (CAP18106-137) on lung injury in mice infected with cytotoxic P. aeruginosa. To determine its maximal effect, the CAP18106-137 peptide was mixed with bacteria just prior to tracheal instillation, and lung injury was evaluated by determining the amount of leakage of an alveolar protein tracer (125I-albumin) into the circulation and by the quantification of lung edema. The lung injury caused by the instillation of 5 x 10(5) CFU of P. aeruginosa was significantly reduced by the concomitant instillation of CAP18106-137. However, the administration of CAP18106-137 alone, without bacteria, induced lung edema, suggesting that it has some toxicity. Also, the peptide did not significantly reduce the number of bacteria that had been simultaneously instilled, nor did it significantly improve the survival of the infected mice. The addition of CAP18106-137 to aztreonam along with the bacteria did decrease the level of antibiotic-induced release of inflammatory mediators including tumor necrosis factor alpha, interleukin-6, and nitric oxide and also improved the survival of the mice. Therefore, more investigations are needed to confirm the toxicities and the therapeutic benefits of CAP18106-137 as an adjunctive therapy to antibiotics in the treatment of infections caused by gram-negative bacteria.

In vitro cellular toxicity predicts Pseudomonas aeruginosa virulence in lung infections.

The role of quorum sensing by Pseudomonas aeruginosa in producing cytotoxicity has not been fully investigated. Strains of P. aeruginosa have been characterized as having an invasive or a cytotoxic phenotype (S. M. J. Fleiszig et al., Infect. Immun. 65:579-586, 1997). We noted that the application of a large inoculum of the invasive strain 6294 caused cytotoxicity of cultured epithelial cells. To investigate this dose-related cytotoxicity, we compared the behavior of 6294 to that of another invasive strain, PAO1, and determined whether the cytotoxicity could be related to quorum sensing. Both invasive strains, 6294 and PAO1, appear to have quorum-sensing systems that were operative when large doses of bacteria were applied to cultured lung epithelial cells or instilled into the lungs of animals. Nonetheless, only 6294 was cytotoxic. Cytotoxicity induced by 6294 correlated with increased elastase production. These experiments suggest that there are multiple mechanisms for the induction of cytotoxicity, pathology, and mortality in vivo. However, in vivo cytotoxicity and mortality, but not pathology, could be predicted by quantitative in vitro cellular damage experiments utilizing a range of bacteria-to-cell ratios. It appears that quorum sensing may inversely correlate with virulence in that strains that produced PAI [N-(3-oxododecanoyl) homoserine lactone] also appeared to attract more polymorphonuclear leukocytes in vivo and were possibly eliminated more quickly. In addition, exoproduct production in bacteriological medium in vitro may differ significantly from exoproduct expression from infections in vivo or during cocultivation of bacteria with tissue culture cells.

Induction of cyclooxygenase-2 in alveolar macrophages after acid aspiration: selective cyclooxygenase-2 blockade reduces interleukin-6 production.

BACKGROUND: Gastric acid aspiration can result in acute lung injury. In this study, the authors determined whether alveolar macrophages express cyclooxygenase-2 as a source of inflammatory mediators after acid aspiration. METHODS: Seventy-five microliters of hydrochloric acid solution, pH 1.15, was instilled into one lung in mice. After exposure, alveolar macrophages were harvested, and competitive polymerase chain reaction and enzyme-linked immunosorbent assay were performed to measure expression of cyclooxygenase-1 and -2, interleukin-1beta and -6, tumor necrosis factor-alpha, and inducible nitric oxide synthase (iNOS). The authors used immunocytochemistry to demonstrate expression of cyclooxygenase-2 in alveolar macrophages. Selective cyclooxygenase-2 blockade using N-2(-cyclohexyloxy-4-nitrophenyl) methane-sulphonamide was done to characterize prostaglandin-cytokine interaction. RESULTS: Acid aspiration induced upregulation of cyclooxygenase-2 and interleukin-6. Tumor necrosis factor-alpha and iNOS were not upregulated. Interleukin-1beta was upregulated even with saline instillation but could not be detected in the supernatant of the cell culture. Alveolar macrophages harvested from mice instilled with acid showed a trend toward more production of prostaglandin E2 and produced higher concentrations of interleukin-6 compared with alveolar macrophages from mice instilled with saline. Selective cyclooxygenase-2 blockade significantly decreased release of interleukin-6 from alveolar macrophages harvested from mice instilled with acid. CONCLUSIONS: Acid aspiration induces strong expression of cyclooxygenase-2 and production of interleukin-6 in alveolar macrophages. Selective cyclooxygenase-2 blockade reduced production of interleukin-6 by acid-stimulated alveolar macrophages. These studies suggest that the induction of cyclooxygenase-2 plays an important role in the systemic inflammatory response induced by acid aspiration.

IL-10 improves lung injury and survival in Pseudomonas aeruginosa pneumonia.

Pseudomonas aeruginosa is the most frequent Gram-negative pathogen causing nosocomial pneumonia. Four different strains of P. aeruginosa (including three isogenic transposon mutants) were utilized in experiments in mice to characterize the specific patterns of cytokine generation in response to bacterial products and cytotoxicity. Intratracheal instillation of any of the strains led to the up-regulation of IL-1beta, IL-6, and TNF-alpha mRNA. Instillation of the cytotoxic strains (PA103, PA103tox::omega) led to IL-10 mRNA up-regulation in the lungs and increased concentrations of IL-10 in the blood. In contrast, the instillation of the noncytotoxic strains (PA01, PA103exsA::omega) did not lead to an increase in IL-10 mRNA in the lungs or to an increase of IL-10 concentration in blood. IL-10 production appears to be a response to either cellular injury or to specific cytotoxic exoproducts produced by the bacteria. The systemic administration of rIL-10 significantly decreased the lung injury and the mortality in mice who had received the cytotoxic strains. The improvement in survival induced by administration of rIL-10 required the concomitant presence of IFN-gamma, as blockade of IFN-gamma with a neutralizing Ab led to 100% mortality, despite the administration of rIL-10. These results suggest that IL-10 is produced in response to specific bacterial products and that there is a potential role for IL-10 in the treatment of cytotoxic P. aeruginosa pneumonia.

Quantitative analysis of aquaporin mRNA expression in rat tissues by RNase protection assay.

The RNase protection assay was applied to quantify mRNA expression of five principal mammalian water channels in 18 different rat tissues, and to determine the influence of dehydration on renal water channel expression. Probes consisted of labeled cRNAs transcribed from cDNA fragments of rat CHIP28 (AQP-1, bp 238-575 of coding sequence), AQP-CD (AQP2, bp 53-606), MIWC (AQP4, bp 235-572), GLIP (AQP3, bp 219-604), and AQP5 (bp 56-612). Results were normalized to expression of rat beta-actin by quantitative densitometry of autoradiograms. CHIP28 mRNA was expressed strongly in heart, kidney > placenta, skeletal muscle, and urinary bladder and detected weakly in eye, lung, trachea, spleen, liver, colon, prostate, and skin. AQP-CD was detected only in kidney. MIWC mRNA expression was highest in brain, followed by eye, trachea, lung, stomach, kidney, and skeletal muscle. GLIP was found in eye, trachea, kidney, urinary bladder, skin, prostate, placenta, and skeletal muscle. AQP5 was detected in salivary gland, eye, lung, and trachea. An alternatively spliced form of MIWC (sMIWC) was also identified in lung and kidney by RNase protection assay, corresponding to deletion of exon 2 of MIWC. In response to dehydration (3 days, -15 % body weight), renal expression of CHIP28 and MIWC were unchanged, whereas expression of AQP-CD and GLIP were increased significantly by 2.18 +/- 0.04 and 1.36 +/- 0.11 fold (SE, n = 5), respectively. These results establish quantitative values for aquaporin transcript expression in multiple mammalian tissues. The sensitive RNase protection assay revealed the expression of water channels in several tissues not studied previously or in which mRNA levels were too low to detect by Northern blot analysis. The observation of GLIP up-regulation in kidney by dehydration suggests a role in the urinary concentrating mechanism.

Intraluminal water increases expression of plasmid DNA in rat lung.

Effective gene delivery to specific organs is a major goal for human gene therapy. The lung's structure allows instillation of agents into the airspaces, directly adjacent to the lung epithelium. We hypothesized that the airspace instillation of hypotonic solutions would increase the permeability of the lung epithelium and increase DNA uptake. This hypothesis was tested by instilling plasmid DNA (p4241) encoding the luciferase gene in isotonic and hypotonic solutions. The highest luciferase expression in the lung was achieved after the instillation of this plasmid DNA in distilled water. Aerosolization of water just before the instillation of the plasmid DNA also enhanced the expression level of luciferase in the lung. In addition, an intralobar instillation of the plasmid DNA in water significantly increased the luciferase expression, suggesting that the instillation of the plasmid over a smaller surface area increased expression. Levels of expression could be measured for 3 days. Water increases the permeability of lung epithelial cells transiently and/or enhances gene expression and can be used to achieve gene expression in the lung airspaces for short intervals without toxicity.

Localization of MIWC and GLIP water channel homologs in neuromuscular, epithelial and glandular tissues.

It was shown recently that water channel homologs MIWC (mercurial insensitive water channel) and GLIP (glycerol intrinsic protein) colocalized in basolateral membranes of kidney collecting duct, tracheal and colonic epithelia, and in brain pia mater. We report here an extensive immunolocalization study of MIWC and GLIP in non-epithelial and glandular epithelial tissues in rat. Immunogold electron microscopy confirmed colocalization of MIWC and GLIP in basolateral membrane of principal cells in kidney collecting duct. However, in other epithelia, MIWC but not GLIP was expressed in basolateral membrane of parietal cells in stomach, and in excretory tubules of salivary and lacrimal glands; GLIP but not MIWC was expressed in transitional epithelium of urinary bladder and skin epidermis. In the central nervous system, MIWC was strongly expressed in the ependymal layer lining the aqueductal system, and in astrocytes throughout the spinal cord and in selected regions of brain. MIWC was also expressed in a plasma membrane pattern in skeletal, but not smooth or cardiac muscle. Neither protein was expressed in small intestine, testis, liver, spleen and nerve. The tissue-specific expression of MIWC suggests a role in fluid transport and/or cell volume regulation in stomach and glandular epithelia. The functional role of MIWC expression in the neuromuscular system and of GLIP expression in skin and urinary bladder is uncertain. The specific cellular sites of MIWC expression (astrocytes, trachea, sarcolemma, gastric parietal cells and kidney principal cells) correspond exactly to sites where orthogonal arrays of particles (OAPs) have been visualized by freeze-fracture electron microscopy, suggesting that MIWC may be the OAP protein.

Immunolocalization of the mercurial-insensitive water channel and glycerol intrinsic protein in epithelial cell plasma membranes.

Two water channel homologs were cloned recently from rat kidney, mercurial-insensitive water channel (MIWC) and glycerol intrinsic protein (GLIP). Polyclonal antibodies were raised against synthetic C-terminal peptides and purified by affinity chromatography. MIWC and GLIP antibodies recognized proteins in rat kidney with an apparent molecular mass of 30 and 27 kDa, respectively, and did not cross-react. By immunofluorescence, MIWC and GLIP were expressed together on the basolateral plasma membrane of collecting duct principal cells in kidney. By immunohistochemistry, MIWC and GLIP were expressed on tracheal epithelial cells with greater expression of GLIP on the basal plasma membrane and MIWC on the lateral membrane; only MIWC was expressed in bronchial epithelia. In eye, GLIP was expressed in conjunctival epithelium, whereas MIWC was found in iris, ciliary body, and neural cell layers in retina. MIWC and GLIP colocalized on the basolateral membrane of villus epithelial cells in colon and brain ependymal cells. Expression of MIWC and GLIP was not detected in small intestine, liver, spleen, endothelia, and cells that express water channels CHIP28 or WCH-CD. These studies suggest water/solute transporting roles for MIWC and GLIP in the urinary concentrating mechanism, cerebrospinal fluid absorption, ocular fluid balance, fecal dehydration, and airway humidification. The unexpected membrane colocalization of MIWC and GLIP in several tissues suggests an interaction at the molecular and/or functional levels.

Acute cardiogenic pulmonary edema.

This article examines the pathophysiology, diagnosis, treatment, and outcome of acute cardiogenic pulmonary edema, as well as re-expansion, high-altitude, postobstructive, and neurogenic pulmonary edemas. Acute cardiogenic pulmonary edema most commonly presents as a consequence of congestive heart failure. The other important causes are acute myocardial dysfunction, documented myocardial infarction, postoperative cardiac dysfunction, and pulmonary hypertension. All these entities have in common increased pulmonary vascular pressures that lead to pulmonary edema.

[Estimation of the filtration variables in the rat lung].

The filtration variables, filtration coefficient (K), perimicrovascular pressure (Ppmv) and reflection coefficient (sigma) were estimated independently in previous reports using the Starling Equation or the micropuncture method. We estimated these variables simultaneously. We measured filtration rate by a gravimetric method in isolated rat lung lobes in zone 1 conditions (alveolar pressure = 20 cmH2O) at two vascular pressures, Pvasc = 15 or 8 cmH2O and perfused the lobes with plasma containing a low or high concentration of protein. By extrapolating the log of the rate of weight gain to time = 0, we obtain the initial filtration rate. Assuming that protein filtered into perimicrovascular space only by convection, we substituted into the Starling Equation as follows: Q = K[(Pmv-Ppmv)-sigma 2 pi mv], where Pmv and pi mv are hydrostatic and plasma protein osmotic pressures in microvascular space. pi mv was estimated by Yamada's equation. We obtained K = 26.3 +/- 8.7 mg/(min.cmH2O.g), PPMV = 6.2 +/- 0.7 cmH2O, sigma = 0.46 +/- 0.07. The chief advantages of this method are that it does not require a separate estimation of isogravimetric pressure or a direct measurement of interstitial pressure, and that all variables are obtained simultaneously.

Lung expansion and the perialveolar interstitial pressure gradient.

We have determined the combined effects of lung expansion and increased extravascular lung water (EVLW) on the perialveolar interstitial pressure gradient. In the isolated perfused lobe of dog lung, we measured interstitial pressures by micropuncture at alveolar junctions (Pjct) and in adventitia of 30- to 50-microns microvessels (Padv) with stopped blood flow at vascular pressure of 3-5 cmH2O. We induced edema by raising vascular pressures. In nonedematous lobes (n = 6, EVLW = 3.1 +/- 0.3 g/g dry wt) at alveolar pressure of 7 cmH2O, Pjct averaged 0.5 +/- 0.8 (SD) cmH2O and the Pjct-Padv gradient averaged 0.9 +/- 0.5 cmH2O. After increase of alveolar pressure to 23 cmH2O the gradient was abolished in nonedematous lobes, did not change in moderately edematous lobes (n = 9, EVLW = 4.9 +/- 0.6 g/g dry wt), and increased in severely edematous lobes (n = 6, EVLW = 7.6 +/- 1.4 g/g dry wt). Perialveolar interstitial compliance decreased with increase of alveolar pressure. We conclude that increase of lung volume may reduce perialveolar interstitial liquid clearance by abolishing the Pjct-Padv gradient in nonedematous lungs and by compressing interstitial liquid channels in edematous lungs.