Mechanism and function of heat shock-dependent IkappaBalpha expression.

OBJECTIVE: Heat shock is known to inhibit activation of the NF-kappa B pathway. One potential mechanism of this effect is de novo expression of the intracellular NF-kappaB inhibitor, Ikappa Balpha. Herein we sought to elucidate the mechanisms by which heat shock induces Ikappa Balpha gene expression and the functional consequences of heat shock-mediated Ikappa Balpha gene expression in A549 cells. METHODS: Nuclear run-on assays demonstrated that heat shock had a small effect on transcription of the Ikappa Balpha gene relative to the level of steady state Ikappa Balpha mRNA that is seen following heat shock. Accordingly, we determined the effect of heat shock on Ikappa Balpha mRNA stability by treating cells with actinomycin D to induce transcriptional arrest. RESULTS: The half-life of IkappaBalpha mRNA was 36 +/- 7.2 min in control cells and 101 +/- 3.7 min in cells subjected to heat shock. These data were consistent with heat shock-mediated increased stability of Ikappa Balpha mRNA. Heat shock induced activation of p38 MAP kinase and inhibition of p38 MAP kinase substantially reduced heat shock-dependent expression of Ikappa Balpha mRNA. After a 4 h recovery period from heat shock, there was inhibition of tumor necrosis factor-alpha-mediated NF-kappaB activation. The introduction of an Ikappa Balpha anti-sense oligonucleotide reversed this inhibitory effect of heat shock. CONCLUSIONS: We conclude that heat shock increases IkappaBalpha gene expression primarily by increasing Ikappa Balpha mRNA stability and this effect is partially dependent on p38 MAP kinase. The functional consequence of heat shock-mediated Ikappa Balpha gene expression is inhibition of NF-kappaB activation.

Absence of endogenous interleukin 10 enhances early stress response during post-ischaemic injury in mice intestine.

BACKGROUND: Interleukin 10 (IL-10) exerts a wide spectrum of regulatory activities in immune and inflammatory responses. AIMS: The aim of this study was to investigate the role of endogenous IL-10 on modulation of the early inflammatory response after splanchnic ischaemia and reperfusion. METHODS: Intestinal damage was induced by clamping the superior mesenteric artery and the coeliac trunk for 45 minutes followed by reperfusion in IL-10 deficient mice (IL-10(-/-)) and wild-type controls. RESULTS: IL-10(-/-) mice experienced a higher rate of mortality and more severe tissue injury compared with wild-type mice subjected to ischaemia and reperfusion. Splanchnic injury was characterised by massive epithelial haemorrhagic necrosis, upregulation of P-selectin and intercellular adhesion molecule 1, and neutrophil infiltration. The degree of oxidative and nitrosative damage was significantly higher in IL-10(-/-) mice than in wild-type littermates, as indicated by elevated malondialdehyde levels and formation of nitrotyrosine. Plasma levels of the proinflammatory cytokines tumour necrosis factor alpha and interleukin 6 were also greatly enhanced in comparison with wild-type mice. These events were preceded by increased immunostaining and activity of the stress regulated c-Jun NH(2) terminal kinase and activation of the transcription factor activator protein 1 in the cellular nuclei of damaged tissue. CONCLUSIONS: These data demonstrate that endogenous IL-10 exerts an anti-inflammatory role during reperfusion injury, possibly by regulating early stress related genetic response, adhesion molecule expression, neutrophil recruitment, and subsequent cytokine and oxidant generation.

The serine/threonine phosphatase, PP2A: endogenous regulator of inflammatory cell signaling.

We have investigated the regulation of kinases and phosphatases in early gene activation in monocytes because these cells are implicated in the pathogenesis of acute inflammatory states, such as sepsis and acute lung injury. One early gene up-regulated by endotoxin is c-Jun, a member of the activating protein (AP) family. C-Jun is phosphorylated by c-Jun N-terminal kinase (JNK) and associates with c-Fos to form the AP-1 transcriptional activation complex that can drive cytokine expression. Inhibition of the serine/threonine phosphatase, PP2-A, with okadaic acid resulted in a significant increase in JNK activity. This finding was associated with increased phosphorylation of c-Jun, AP-1 transcriptional activity, and IL-1beta expression. Activation of PP2A inhibited JNK activity and JNK coprecipitated with the regulatory subunit, PP2A-Aalpha, supporting the conclusion that PP2A is a key regulator of JNK in the context of an inflammatory stimulus.

Regulation of chemokine expression by IL-10 in lung inflammation.

We have been interested in understanding the mechanisms regulating the inflammatory process underlying acute lung injury. The current studies have employed a model of acute lung inflammation in mice triggered by bacterial lipopolysaccharide. The development of this injury was associated with increased expression of the chemokines, MIP-1alpha and MIP-2, that coordinate recruitment of neutrophils to the lung. IL-10 is a potent, endogenous anti-inflammatory molecule that has been shown to decrease lung inflammation partly on the basis of TNF-alpha and IL-1beta inhibition. In these studies we tested the hypothesis that endogenous IL-10 modulates chemokine expression using the IL-10 knock-out mouse, and then explored the molecular mechanisms by which IL-10 might do so. The results demonstrate that significant elevations in both chemokines were observed in the absence of IL-10 and that these findings were associated with significant increases in lung neutrophil accumulation. In vitro studies defined two, gene-specific, mechanisms by which IL-10 regulated chemokine expression: mRNA destabilization and NF-kappaB inhibition. These results suggested that IL-10 is an important, endogenous regulator of chemokine expression in acute lung inflammation.

Recovery after cardioplegia in the hypertrophic rat heart.

BACKGROUND: Enhanced recovery after cardioplegic arrest has been observed in rat hearts with hypertrophy induced by hemodynamic overload. We hypothesize that this is related to altered characteristics of hypertrophied myocardium-reflected by increased V(3) isomyosin and glycolytic potential-other than increased left ventricular mass. MATERIALS AND METHODS: Isolated hearts from age-matched nonoperated and sham-operated control rats and from aortic-banded, hyperthyroid, and hypothyroid rats-groups in which hypertrophy and V(3) as a percentage of left ventricular myosin vary independently-underwent 2 h of multidose cardioplegic arrest at 8 degrees C followed by reperfusion at 37 degrees C. Left ventricular V(3) isomyosin was evaluated after separation by gel electrophoresis. RESULTS: Moderate left ventricular hypertrophy was produced by aortic banding or hyperthyroidism and atrophy by hypothyroidism. V(3) isomyosin was increased in banded (28%) and hypothyroid (75%) rats compared to control (12%) and hyperthyroid rats (7%). Myocardial glycogen content closely paralleled %V(3). At 30 min of working reperfusion, functional recovery (assessed as percentage prearrest cardiac output) was 66 +/- 4 and 68 +/- 5% in control and hyperthyroid hearts and 81 +/- 2 and 80 +/- 5% in hearts from banded and hypothyroid rats (each P < 0.05 vs controls), respectively. At 30 min, hearts from banded and hypothyroid rats were also more efficient (as indexed by cardiac output at constant mean aortic pressure/myocardial oxygen consumption) than control and hyperthyroid hearts. CONCLUSIONS: The data suggest that recovery is related not to increased mass but to other changes in overload hypertrophy. Increased percentage V(3) isomyosin and glycogen reflect these changes and may themselves contribute to improved functional recovery after cardioplegic arrest, as may increased postischemic efficiency.

Increased expression of heat shock protein-70 protects A549 cells against hyperoxia.

Acute and chronic lung injury secondary to hyperoxia remains an important complication in critically ill patients, and, consequently, there is interest in developing strategies to protect the lung against hyperoxia. Heat shock proteins (HSPs) confer protection against a broad array of cytotoxic agents. In this study, we tested the hypothesis that increased expression of the 70-kDa HSP (HSP70) would protect cultured human respiratory epithelium against hyperoxia. Recombinant A549 cells were generated in which human HSP70 was increased by stable transfection with a plasmid containing human HSP70 cDNA under control of the cytomegalovirus promoter (A549-HSP70 cells). A549-HSP70 cells exposed to hyperoxia had greater acute survival rates and clonogenic capacity compared with wild-type A549 cells and with control cells stably transfected with the empty expression plasmid. Hyperoxia-mediated lipid peroxidation and ATP depletion were also attenuated in A549-HSP70 cells exposed to hyperoxia. Increased expression of HSP70 did not detectably alter mRNA levels of the intracellular antioxidants manganese superoxide dismutase, catalase, and glutathione peroxidase. Collectively, these data demonstrate a specific in vitro protective role for HSP70 against hyperoxia and suggest that potential mechanisms of protection involve attenuation of hyperoxia-mediated lipid peroxidation and ATP depletion.

Poly(ADP-ribose) synthetase activation mediates increased permeability induced by peroxynitrite in Caco-2BBe cells.

BACKGROUND & AIMS: Peroxynitrite induces cytotoxicity by generating DNA single-strand breaks and activating poly(ADP-ribose) synthetase (PARS), a nuclear enzyme that consumes oxidized nicotinamide adenine dinucleotide (NAD+) and depletes cellular adenosine triphosphate (ATP). The aim of this study was to examine this mechanism of injury in an intestinal epithelial cell model after exposure to exogenous peroxynitrite (ONOO-) and nitric oxide (NO). METHODS: Caco-2BBe cell monolayers exposed to donors of peroxynitrite (3-morpholino-sydnonimine [SIN-1], 3 mmol/L) or NO (S-nitroso-N-acetyl penicillamine [SNAP]; 3 mmol/L) were analyzed for DNA strand breaks, [NAD+], [ATP], and transepithelial flux of fluorescein sulfonic acid. RESULTS: SIN-1 but not SNAP induced DNA single-strand breakage. Both SIN-1 and SNAP reduced [ATP], but only SIN-1 reduced [NAD+]. Inhibition of PARS activity by the PARS inhibitors 5-iodo-6-amino 1,2-benzopyrone or 3-aminobenzamide prevented the SIN-1-induced reduction in [NAD+] and [ATP] but had no effect on the SNAP-induced reduction in [ATP]. PARS inhibition reduced SIN-1-but not SNAP-induced hyperpermeability. CONCLUSIONS: Peroxynitrite but not NO increases transepithelial permeability by inducing DNA strand breaks that activate the PARS pathway and cause the depletion of intracellular energy stores. Inhibition of PARS activity may represent a novel strategy in ameliorating peroxynitrite-mediated epithelial injury during intestinal inflammation.

Superior recovery of hypertrophied rat myocardium after cardioplegic arrest.

BACKGROUND: Although cardioplegic protection of the hypertrophied heart remains a clinical challenge, we have previously observed enhanced recovery in rat hearts with pressure-overload hypertrophy induced by aortic banding. We investigated whether this unexpected result is found in other models of hypertrophy. METHODS: Hearts with hypertrophy induced by aortic banding or administration of desoxycorticosterone acetate were each compared with age-matched sham-operated and nonoperated controls. Spontaneously hypertensive rats and Wistar-Kyoto controls were also compared. We evaluated left ventricular isomyosin distribution by gel electrophoresis and recovery of isolated working rat hearts arrested at 8 degrees C for 2 hours. RESULTS: The percentage of V3 isomyosin in hearts with hypertrophy from aortic banding or administration of desoxycorticosterone acetate was increased compared with the control groups. Recovery of aortic flow in all three groups of hypertrophied hearts was at least as good or better than their respective controls. There were no significant differences in ATP or glycogen between hypertrophied and control hearts before or after arrest. CONCLUSIONS: Enhanced recovery of hypertrophied hearts is not specific to a single model. This level of recovery may be supported by induction of a "fetal genetic program," exemplified in the rat by the shift in isomyosin from predominantly V1 to the more efficient V3 isoform, which occurs in pressure-overloaded hearts.

Bacterial induction of inducible nitric oxide synthase in cultured human intestinal epithelial cells.

BACKGROUND & AIMS: Enterocytes play a major role in the mucosa as a source of proinflammatory cytokines and cytotoxins. We tested the hypothesis that bacteria induce expression of the inducible nitric oxide synthase (iNOS) in cultured human enterocytes. METHODS: DLD-1 and Caco-2BBe cell monolayers exposed to Salmonella dublin were analyzed for iNOS up-regulation and nitric oxide production (NOx) in the presence of various proinflammatory cytokines. RESULTS: S. dublin augmented NOx in interferon gamma (IFN-gamma)-primed cells but had no independent effect on iNOS expression. S. dublin-induced NOx was not mediated by endotoxin and was augmented by an enteroinvasive phenotype. In DLD-1 cells, S. dublin-mediated NOx was blocked by inhibitors of nuclear factor kappa B (NF-kappa B) and tyrosine kinase activation and was steroid resistant. Cis-acting elements in the human iNOS promoter responsive to endotoxin and S. dublin stimulation of IFN-gamma-treated DLD-1 cells were identified between 10.9 and 8.7 kilobases upstream of the transcription initiation site. CONCLUSIONS: S. dublin alters the regulation of iNOS messenger RNA in IFN-gamma-treated intestinal epithelial cells via a steroid-resistant pathway involving NF-kappa B and tyrosine kinase activity. Because bacterial interaction with cytokine-primed epithelial cells induces the synthesis of NO, an endogenous antimicrobial agent, these findings may have implications for the regulation of mucosal immunity.

Circulating methemoglobin and nitrite/nitrate concentrations as indicators of nitric oxide overproduction in critically ill children with septic shock.

OBJECTIVES: To examine the relationship between circulating methemoglobin and nitrite/nitrate concentrations and to compare these markers of nitric oxide overproduction with clinical variables in children diagnosed with septic shock. DESIGN: Prospective, controlled, clinical study. SETTING: Pediatric intensive care unit and outpatient clinic in a children's hospital. PATIENTS: Twenty-two children diagnosed with septic shock and ten age-matched healthy control patients. INTERVENTIONS: Patients diagnosed with septic shock had blood specimens taken on study entry and every 6 hrs for 72 hrs for methemoglobin and nitrite/nitrate determinations. Single blood specimens were obtained from controls. MEASUREMENTS AND MAIN RESULTS: Circulating methemoglobin and nitrite/nitrate concentrations were significantly higher in children diagnosed with septic shock in comparison with healthy control children (p = .01 and .05, respectively). Peak nitrite/nitrate concentrations correlated with serum creatinine (r2 = .19; p = .04) and were inversely correlated with arterial pH (r2 = .28; p = .01) and urine output (r2 = .21; p = .03) when analyzed by log-linear regression. There were no significant relationships between methemoglobin and nitrite/nitrate or between methemoglobin and any other clinical variable. CONCLUSIONS: Circulating methemoglobin and nitrite/nitrate concentrations are increased in children diagnosed with septic shock. Plasma nitrite/nitrate values correlate with selected clinical variables in these children. Circulating methemoglobin measurements are not superior to plasma nitrite/nitrate concentrations as an indicator of endogenous overproduction of nitric oxide in children diagnosed with septic shock. A need remains to develop markers of endogenous nitric oxide activity that have greater accuracy and reliability.

Heat shock protein induction protects human respiratory epithelium against nitric oxide-mediated cytotoxicity.

Induction of heat shock proteins (HSPs) confers protection against a variety of cytotoxic agents. We hypothesized that induction of HSPs would protect cultured human respiratory epithelium against nitric oxide (NO)-mediated injury. Incubation of a human bronchial epithelial cell line (BEAS-2B cells) at 43 degrees C for 1.5 h induced expression of several HSPs. Prior induction of HSPs was associated with protection against the NO-donors S-nitroso-N-acetyl penicillamine and 3-morpholinsydnonimine. Protection was evident as improved short term survival and improved ability of cells to recover and proliferate after exposure to NO. Prior induction of HSPs also attenuated NO-mediated decreases in cellular ATP levels, but did not decrease nitrotyrosine formation. Specific overexpression of HSP-70 by plasmid-directed gene transfer protected murine respiratory epithelial cells against S-nitroso-N-acetyl penicillamine. We conclude that in cultured human respiratory epithelium induction of HSPs confers protection against NO-mediated cytotoxicity, possibly by preservation of cellular energetics. We also suggest that HSP-70 may play a specific role in protection.

K(ATP)-channel inhibition improves hemodynamics and cellular energetics in hemorrhagic shock.

We tested whether activation of K(ATP) channels contributes to vasodilatation and end-organ hypoperfusion in severe hemorrhagic shock (HS). Anesthetized juvenile pigs were hemorrhaged to a portal blood flow of 45% of baseline for 45 min and then resuscitated with Ringer lactate (RL; 100% volume of shed blood; n = 10) or RL in combination with the K(ATP)-channel antagonist glibenclamide (10 mg/kg iv bolus injection; n = 10). Addition of glibenclamide to the resuscitation fluid caused a sustained recovery of systemic blood pressure, cardiac index, portal blood flow, renal blood flow, renal cortical ATP concentration, and ileal mucosal P(CO2). Treatment with RL alone caused only a partial and transient hemodynamic and metabolic benefit. Glibenclamide treatment of sham-shocked control pigs (n = 6) transiently increased mesenteric and systemic vascular resistance. Inhibition of K(ATP)-channel activity in HS, which effectively and safely restores systemic hemodynamics, regional perfusion, and tissue metabolism, is a potentially novel therapeutic approach to the management of severe HS.

Induction and activity of nitric oxide synthase in cultured human intestinal epithelial monolayers.

We have examined the induction and activity of inducible nitric oxide (NO) synthase (iNOS) in monolayers of DLD-1 cells, an epithelial cell line derived from a human intestinal adenocarcinoma. Induction of iNOS transcription?by a combination of the cytokines interferon-gamma and IL-1 beta was inhibited by genistein, pyrrolidine dithiocarbamate, or dexamethasone and unaffected by pretreatment with ethylene glycol-bis(beta-aminoethyl ether)-N, N,N',N'-tetraacetic acid, basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), or the isoflavone daidzein. iNOS activity and NO synthesis were inhibited by nitro-L-arginine methyl ester, NG-monomethyl-L-arginine, S-methyl-isothiourea sulfate, or aminoethyl-isothiourea, but not by dexamethasone. NO synthesis was potently inhibited by N-alpha-p-tosyl-lysine chloromethyl ketone and hypoxia. In the absence of cytokines no iNOS induction was observed with oxidant stress (H2O2), growth factors (bFGF, EGF), hypoxia or hypoxia reoxygenation. We conclude that in this model of the human intestinal epithelium 1) cytokine-mediated induction of iNOS is Ca2+ independent, weakly steroid sensitive, and may involve the activation of nuclear factor-kappa B and a tyrosine kinase, and 2) iNOS activity is Ca2+ -independent and inhibited by hypoxia, NG-substituted L-arginine analogues, and isothioureas.

Endotoxin-induced ileal mucosal hyperpermeability in pigs: role of tissue acidosis.

Administration of lipopolysaccharide (LPS) to experimental animals leads to diminished mesenteric perfusion, increased ileal mucosal [H+] , and increased gut epithelial permeability to hydrophilic solutes. We sought to determine whether these phenomena are causally related. Experiments were performed in anesthetized pigs. Permeability was assessed by measuring the plasma-to-lumen clearance of fluorescein isothiocyanate dextran (4,000 Da; FD-4) by a segment of ileum perfused with Ringer lactate solution. Mucosal perfusion (Qmuc) and [H4+] were estimated using laser-Doppler flowmetry and tonometry, respectively. In an initial series of experiments, we showed that mucosal permeability was linearly correlated with mucosal [H+] in animals subjected to graded degrees of mechanically induced mesenteric ischemia (n = 14, R2 = 0.58, P < 0.002) or injected with graded doses of LPS (n = 11, R2 = 0.93, P < 0.0001). In a second series of experiments, we induced mucosal acidosis in normal pigs by mechanical ventilation with either a hypoxic (n = 7) or a hypercapnic (n = 5) gas mixture. In both groups, ileal mucosal permeability to FD-4 increased significantly (P < 0.05), although transmesenteric release of lactate increased significantly only in the hypoxic group. Qmuc was unchanged in both groups. These data suggest that mucosal acidosis, even in the absence of tissue ischemia or hypoxia, increases intestinal permeability to a macromolecular hydrophilic solute. Tissue acidosis may be an important factor contributing to LPS-induced gut mucosal hyperpermeability.

Protection of the hypertrophied myocardium by crystalloid cardioplegia.

Patients with left ventricular hypertrophy (LVH) have a worse outcome after cardiac surgery than those without hypertrophy. We studied protection of hearts with LVH in an isolated rat heart model using multidose, cold, oxygenated cardioplegia. LVH was produced by banding the abdominal aorta in young rats. Six weeks after banding, this produced a 31% increase in the left ventricular dry weight/body weight ratio compared to two age-matched control groups comprising sham-operated and nonoperated animals. The recovery of cardiac output after arrest was higher in LVH (82 +/- 4% of prearrest) than in sham-operated (69 +/- 4%) or nonoperated (66 +/- 3%) control groups. The improved functional recovery in LVH occurred although there were no differences among the groups in myocardial adenosine triphosphate (ATP) and phosphocreatine (PCr) prior to arrest, at the end of arrest, or after reperfusion. Glycogen levels were also similar among the three groups prior to arrest and after reperfusion but were highest in LVH after arrest. Myocardial oxygen consumption (MVO2) and efficiency, expressed as cardiac output/MVO2, were similar among the groups prior to arrest. Myocardial efficiency after reperfusion declined in all groups but was best preserved in LVH. We also compared the sensitivity of hypertrophied and control hearts to the deleterious effects of calcium in cardioplegia. Calcium in the cardioplegia increased myocardial lactate production during arrest in a dose-related fashion and depressed myocardial levels of ATP, PCr, and glycogen at end arrest in all groups. Cardiac output recovery was also depressed by calcium but was still best in LVH. We conclude that the hypertrophied myocardium is well protected by standard cardioplegia and that calcium in cardioplegia does not preferentially depress recovery in LVH.

Oxygenated cardioplegia: the metabolic and functional effects of glucose and insulin.

Reports differ as to the efficacy of glucose and insulin as cardioplegic additives. Although deliberate oxygenation of crystalloid cardioplegic solutions improves myocardial protection, little is known about the protection afforded by glucose and insulin in such oxygenated solutions. In the isolated working rat heart, we studied the addition of oxygen, glucose, and insulin, separately and together, to a cardioplegic solution. The solution was equilibrated with O2 or N2, with glucose added as a substrate or sucrose as a nonmetabolizable osmotic control, with or without insulin. Hearts were arrested for 2 hours at 8 degrees C by multidose infusions. Oxygenation decreased lactate production and improved high-energy phosphate and glycogen preservation during arrest, prevented ischemic contracture, and improved functional recovery. The addition of glucose to the oxygenated solution increased the level of adenosine triphosphate at end-arrest from 10.5 +/- 0.5 to 13.9 +/- 0.6 nmol/mg dry weight and glycogen stores from 18.7 +/- 2.5 to 35.7 +/- 5.5 nmol/mg dry weight. The further addition of insulin did not better preserve these metabolites. Improvements in functional recovery due to glucose or insulin in the oxygenated solution attained statistical significance when both additives were included. Glucose increased lactate production significantly only when the solution was nitrogenated. Insulin added to the nitrogenated glucose-containing solution increased adenosine triphosphate and glycogen levels after 1 hour of arrest; and, although insulin did not prevent ischemic contracture from developing during the latter part of arrest with profound depletion of these metabolites, functional recovery was improved. The mechanism of improved functional recovery by insulin is not clear.(ABSTRACT TRUNCATED AT 250 WORDS)

Maximal oxygenation of dilute blood cardioplegic solution.

The content of dissolved O2 (the major source of O2 for the myocardium) of dilute blood cardioplegic solution (dBCS) varied widely when oxygenated at 4 degrees C by surface flow of O2 in a Bentley BCR-3500 cardiotomy reservoir. We have modified the system to consistently deliver maximally oxygenated dBCS to the heart. Laboratory studies indicated that bubbling O2 through a 16-gauge intravenous catheter in a central Luer-Lok port of the cardiotomy reservoir provided contents of dissolved O2 that were consistently near maximal. We then studied 17 patients in the operating room. The first 6 patients received dBCS oxygenated with 100% O2 with a high dissolved O2 content of 3.2 +/- 0.2 ml/dl. However, the pH of the dBCS became highly alkaline (7.83 +/- 0.11 at 37 degrees C). Therefore, in the remaining 11 patients, 2% CO2 was added to the O2. The dissolved O2 content remained high (3.3 +/- 0.1 ml/dl), and the pH was in a more physiological range (7.35 +/- 0.09 at 37 degrees C). We conclude that consistently maximal oxygenation of a dBCS at a more physiological pH can be achieved by this method.

Microsphere reference flow samples during systemic flow adjustment.

Regional myocardial blood flow measurements in the right heart bypass preparation can be particularly valuable, since this preparation provides control of the main hemodynamic determinants of coronary blood flow. We examined the validity of aortic reference flow samples in relation to coronary samples during continuous systemic flow adjustment for aortic pressure control in six dogs on right heart bypass, anesthetized with chloralose and urethan. Microsphere concentrations were compared in paired reference flow samples drawn from the aortic arch and from a coronary artery for 119 left atrial microsphere injections. During left subclavian artery infusion and during femoral artery infusion at rates above 2,000 ml/min, there were high percentage errors in microsphere concentration between paired samples, consistent with aortic sample dilution by systemically infused blood. In 52 injections during withdrawal or femoral infusion below 2,000 ml/min, at cardiac outputs of 390-4,800 ml/min, the percentage error was 0.001 +/- 1.18% (SE); the absolute value of this error was below 20% in 96%, and below 10% in 77% of these injections. Linear regression related these coronary to aortic microsphere concentrations by the equation Y = 1.005X - 1.64, r = 0.997, Sy.x = 13.2 (5.9%). (Sy.x represents the standard deviation from regression.) These data indicate that valid aortic reference flow samples can be obtained within specific hemodynamic conditions during systemic flow adjustment in the right heart bypass preparation.