Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Microvascular Endothelial Cell Control of Peripheral Vascular Resistance During Sepsis

Microvascular Endothelial Cell Control of Peripheral Vascular Resistance During Sepsis ObjectiveTo determine the endothelial-dependent control of decreased peripheral vascular resistance in skeletal muscle microvessels during evolving sepsis.Materials and InterventionsAcute (4 hours, n=7), established (24 hours, n=7), or chronic (72 hours, n=8) infection was induced in Sprague-Dawley rats (150-175 g) by injecting Escherichia coliand Bacteroides fragilis(1×109colony-forming units for both) into a subcutaneous sponge. Control animals were injected with an isotonic sodium chloride solution and analyzed at the same time points: (n=6-8 per group). Dilation in response to the topically applied endothelial-dependent agonist acetylcholine (ACH) (1×10−9to 1×10−5mol/L) was measured in inflow first-order (A1) and precapillary fourth-order (A4) arterioles in cremaster muscle in vivo with videomicroscopy. Acetylcholine dose-response curves were used to determine vascular reactivity by calculating the concentration of ACH necessary to elicit 50% of the maximal dilator response.Main Outcome MeasuresIn vivo reactivity of striated muscle microvessels to the dilation agonist ACH during acute, established, and chronic infection.ResultsA1 vessels were unresponsive to all doses of ACH at all time points. A4 vessels showed an increased dilator response during short-term treatment, which deteriorated over time to depressed dilation during chronic infection.ConclusionsPrecapillary A4 vessels have increased dilator reactivity during early sepsis, which progresses to depressed levels with chronic infection. A1 microvessels remain dilated and are not substantially influenced by endothelial dilator mechanisms initiated by ACH. Maximum dilation of the large A1 vessels appears to contribute to the decrease in peripheral vascular resistance noted during systemic infection.ACUTE SEPSIS manifests a hemodynamic state in which the cardiac output (CO) and heart rate are increased in an attempt to adequately supply peripheral perfusion. This high-output state occurs concomitantly with a decrease in systemic vascular resistance and oxygen consumption. These cardiovascular and metabolic changes reflect a failure to deliver or use oxygen at the cellular level, resulting in metabolic acidosis.Microvascular blood flow abnormalities, such as an increased number of capillaries with no flow, occur in striated muscle during sepsis.These metabolic, cardiovascular, and blood flow alterations indicate a need to assess skeletal muscle microvascular function to understand changes in peripheral resistance during the evolution of sepsis.Endothelial cells control the contractility of vascular smooth muscle cells primarily by the synthesis and release of nitric oxide (NO).Endothelial release of NO is altered during sepsis.The question remains whether endogenous NO activity in microvascular beds leads to the hemodynamic changes of sepsis. In the acute phase of sepsis, large arterioles are constricted and small terminal arterioles are dilated in rat cremaster muscle.This constriction of large-inflow arterioles reduces the overall blood volume delivered to capillaries and may ultimately contribute to metabolic acidosis.Sepsis typically begins with a hyperdynamic clinical presentation that includes tachycardia, tachypnea, acidosis, temperature elevation, leukocytosis, and early organ system dysfunction. As the infection persists, organ dysfunction, hypoperfusion, and hypotension ensue.During chronic sepsis, inducible nitric oxide synthase (iNOS) levels are increased, and constitutive NOS (cNOS) levels are decreased.These and many similar findings are the basis for iNOS inhibition as a treatment of sepsis.Recent studieshave shown, however, that a nonspecific inhibition of NOS is detrimental during the later stages of endotoxemia or septic shock. The effectiveness of iNOS inhibition may depend on the stage of sepsis, the capability of endothelial cells to release NO, and the selectivity of the antagonist. In vivo microvascular reactivity throughout the development of sepsis has not been studied. To determine the capacity of endogenous NO release by the striated muscle microvascular bed, we chose to analyze endothelial-dependent relaxation in response to acetylcholine (ACH) during the progression of infection from the acute to the chronic state. We hypothesized that the development of chronic sepsis would depress the endothelial-dependent relaxation because of a diminished capacity of the endothelial cell to release NO.MATERIALS AND METHODSANIMAL MODELMale Sprague-Dawley rats (125-135 g) were acclimated for 2 weeks in a facility approved by the American Association for Accreditation of Laboratory Animal Care. All studies were approved by the Institutional Animal Care and Use Committee and conformed to the National Institutes of Health Guidelines for Care and Use of Laboratory Animals. Animals were given an intraperitoneal injection of sodium pentobarbital (45 mg/kg). Under sterile conditions, a 2-cm2gauze sponge was placed in the subcutaneous space at the base of the tail. Animals were then housed individually and observed for 4 days to document good health and weight gain. On the fifth postoperative day, all of the animals demonstrated a weight gain (150-175 g) and normal food and water intake. On this day, the subcutaneous sponge was inoculated with either 2 mL of isotonic sodium chloride solution (control groups) or 1 mL of 1×109colony-forming units of Bacteroides fragilisplus 1 mL of 1×109colony-forming units of Escherichia coli(septic groups). This inoculation was repeated after 48 hours for animals studied at 72 hours. At 4, 24, or 72 hours, all animals were studied, respectively representing the study groups with acute, established, or chronic sepsis.The E coliwas obtained from a hospital-acquired pathogenic strain and kept frozen until used. Periodically, a colony was grown in 25 mL of tryptose-soy agar broth, streaked onto an agar plate, and incubated until individual colony growth appeared. Preceding each experiment, 1 colony was extracted from the agar plate, suspended in tryptose-soy agar broth (25 mL), and incubated at 37°C for 16 hours. The suspended bacteria were washed and suspended in isotonic sodium chloride solution. Serial dilutions were performed and measured spectrophotometrically at 620-nm wavelength to provide 1×109colony-forming units of bacteria in 1 mL of isotonic sodium chloride solution. The B fragiliswas obtained (American Type Culture Collection, Rockville, Md) in freeze-dried pellet form and suspended in Schaedler broth (40 mL) (Difco Laboratories, Detroit, Mich), which includes Oxyrase (0.8 mL) (Oxyrase Inc, Mansfield, Ohio) to maintain an anaerobic environment. Preceding each experiment, 100 µL of B fragilissuspension was added to 40 mL of fresh culture medium and incubated for 46 hours. The bacteria were then washed and calibrated in the same manner as used for E coli.EXPERIMENTAL PROTOCOLAt 4, 24, or 72 hours after initial sponge injection, each animal's cremaster muscle microvascular reactivity was analyzed in vivo by videomicroscopy. Animals were anesthetized with an intraperitoneal injection of sodium pentobarbital (45 mg/kg), and a tracheotomy tube was positioned for airway control. Transpulmonary thermodilution CO measurements were made with a small-diameter thermistor positioned in the aortic arch through the right carotid artery. Each CO measurement was made by injecting a bolus of room temperature isotonic sodium chloride (40 µL) into the superior vena cava by a jugular vein catheter. All CO measurements were made in triplicate every 10 minutes throughout the experiment. Mean arterial pressure and heart rate measurements were obtained continuously through a right femoral artery cannula.The cremaster muscle videomicroscopy preparation was originally described by Baez.Briefly, the scrotal sac was cut longitudinally, and the cremaster-covered testicle was exposed. Layers of connective tissue were gently peeled off. The cremaster muscle, with neurovascular bundle intact, was cut longitudinally and secured in a custom-made tissue bath over an optical port. Throughout the rest of the surgical procedure, the muscle tissue was kept moist with Krebs solution (sodium bicarbonate, 15.5 mmol/L; sodium chloride, 112.9 mmol/L; potassium chloride, 4.7 mmol/L; and calcium chloride diluted 1:2 parts in water, 2.55 mmol/L). The bath was filled with Krebs solution, continuously bubbled with nitrogen and carbon dioxide to maintain the pH between 7.35 and 7.45, and heated to 34°C. Pancuronium bromide, 0.1 mL of 1×10−6mol/L, was added to the bath to prevent muscle spasms.The cremaster microcirculation was directly observed by a trinocular transillumination microscope (model ACM, Carl Zeiss, Inc, Thornwood, NY), and microvascular images were transmitted by videocamera (model 5100 HS, Panasonic, Secaucus, NJ) to a monitor and videocassette recorder at ×1246 magnification. Vessel intraluminal diameters were measured on the monitor and converted to micrometers. Vascular images were read from the videocassette recording using still images to ensure accuracy.The branch order of the cremaster microvascular anatomy was determined by assigning the major inflow arteriole as the first-order arteriole (A1) and the parallel adjacent venule as the first-order venule (V1). Perpendicular branches from A1 microvessels were termed second-order arterioles(A2), and successive arteriolar branches from the second-order vessels were designated third-order(A3) and fourth-order(A4) vessels. These smaller vessels occur without adjacent venules and are terminal precapillary arterioles.Following the surgical procedure, the cremaster muscle was allowed to equilibrate in the tissue bath for 1 hour. Baseline measurements were taken of mean arterial pressure, heart rate, CO, and microvascular diameters of A1, V1, A3, and A4 microvessels. A second set of baseline measurements was repeated 10 minutes after the first set and was required to be within 10% of the initial measurements for entry into the experimental protocol. After baseline measurements were obtained, serial doses of ACH were added to the bath (final concentrations, 1×10−9to 1×10−5mol/L). Each dose of ACH remained in the bath for 10 minutes. Hemodynamic measurements were taken during the first minute of exposure, and vessel diameters were recorded at 1, 3, 5, 7, and 9 minutes for each dose of ACH. Sodium nitroprusside was added after the final ACH dose, and hemodynamic and vessel diameter measurements were taken 5 minutes later to determine maximal vessel dilation.Dose-response curves for dilation were generated by normalizing the measured vessel diameters with reference to the maximal vessel dilation. The mean baseline value was set at 0, and the maximal vessel diameters with sodium nitroprusside application were set at 100% maximal dilator capacity. The following equation was used to calculate the percentage of dilator capacity: (X−BL)/(NP−BL)×100%, where X indicates vessel diameter; BL, vessel diameter at baseline; and NP, maximal vessel diameter. From the ACH dose-response curve of each animal, vascular sensitivity was determined by the concentration of ACH necessary to elicit 50% of the maximal dilator response (EC50). The pD2value was determined as the negative log EC50, which represents the changes in microvascular sensitivity to ACH.STATISTICAL ANALYSISWithin each experimental group, the baseline values were compared with ACH doses using repeated-measures analysis of variance, followed by the Tukey-Kramer least significant difference test. Experimental groups were compared using the 1-way analysis of variance, followed by the Tukey-Kramer least significant difference test. The null hypothesis was rejected when P<.05.RESULTSHEMODYNAMICSBaseline hemodynamic values are reported in Table 1. The mean arterial pressure was similar in all experimental groups, except at 24 hours in the established infection group, which showed a 24% increase (P<.05) when compared with controls that received isotonic sodium chloride, and a 14% increase in heart rate (P<.05). Although there was a trend toward a hyperdynamic cardiovascular state in the animals with chronic infection, these changes were not significant (P>.05). All hemodynamic variables were stable during intravital microscopy and dose-response protocols.Hemodynamic Baseline Values for All Experimental Groups&ast;Infection GroupMean Arterial Pressure, mm HgHeart Rate, Beats/minCardiac Output, mL ⋅ min−1⋅ 100 g−1Acute4-h isotonic sodium chloride solution (n = 8)107.0 ± 3.5414.0 ± 14.017.5 ± 2.84-h bacteria (n = 7)106.0 ± 3.5400.0 ± 11.519.3 ± 3.8Established24-h isotonic sodium chloride solution (n = 7)106.0 ± 2.2382.0 ± 13.522.9 ± 3.024-h bacteria (n = 7)144.0 ± 7.9†457.0 ± 12.2†16.6 ± 2.9Chronic72-h isotonic sodium chloride solution (n = 6)112.0 ± 2.6371.0 ± 9.028.5 ± 3.572-h bacteria (n = 8)118.0 ± 4.5398.0 ± 8.636.9 ± 3.7&ast;Data are reported as mean ± SEM.†P<.05 vs isotonic sodium chloride control (analysis of variance followed by the Tukey-Kramer test).MICROVASCULAR RESPONSESFirst-order arterioles and venules were unresponsive to increasing concentrations of ACH and sodium nitroprusside. These first-order microvessels maintained diameters within 10% of baseline values in all experimental groups, and dose-response curves were flat. In the group showing chronic sepsis after 72 hours, the baseline diameters of A1 and V1 microvessels were slightly larger than in the groups with acute or established infection.Although the A3 microvessels in the groups with either acute or established infection showed no change in reactivity compared with those in controls, there was a shift in the EC50to the right in the group with chronic sepsis, indicating a substantial depression of dilator function (Figure 1). Precapillary A4 microvessels demonstrated a differential shift in reactivity to dilator stimulation, with an increased sensitivity and shift to the left of the EC50noted in the groups with acute sepsis (Figure 2) and a definite depression of dilator function and shift to the right of the EC50after 72 hours of infection (Figure 3). There was no change in reactivity of A4 microvessels in the groups with established infection compared with control animals. This differential shift in microvascular sensitivity to dilator stimulation is reflected in the calculated pD2values for A3 and A4 microvessels in each group. The pD2value was depressed in A3 vessels during chronic infection at 72 hours (Figure 4), whereas A4 vessels showed an early increase in the pD2value during acute infection at 4 hours, indicating a shift over time in these small precapillary microvessels.Figure 1.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A3 arterioles after 72 hours of bacterial infection. Microvascular responses are reported as the percentage (mean±SEM) of maximum dilator capacity (for the method of calculation, see the "Experimental Protocol" subsection in the "Materials and Methods" section). Open circles indicate control animals (n=6); closed squares, the group with chronic infection (n=8); BL, baseline; NP, maximal vessel diameter; and asterisk, P<.05 (by analysis of variance, followed by the Tukey-Kramer test) for isotonic sodium chloride control vs bacterial treatment.Figure 2.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A4 arterioles after 4 hours of bacterial infection. Microvascular response reporting is as given in the legend for Figure 1. No significant differences were found between the 2 experimental groups. Open circles indicate control animals (n=8); closed squares, the group with acute infection (n=7); BL, baseline; and NP, maximal vessel diameter.Figure 3.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A4 arterioles after 72 hours of bacterial infection. Microvascular response reporting is as given in the legend for Figure 1. Open circles indicate control animals (n=6); closed squares, the group with chronic infection (n=8); BL, baseline; NP, maximal vessel diameter; and asterisk, P<.05 (by analysis of variance, followed by the Tukey-Kramer test) for isotonic sodium chloride control vs bacterial treatment.Figure 4.pD2values of A3 and A4 arterioles in control animals (light-shaded bars) and animals with acute, established, or chronic infection (dark-shaded bars). pD2indicates −log EC50, or dose required to elicit a 50% response of maximal dilation; asterisk, P<.05 (by analysis of variance) vs isotonic sodium chloride control.COMMENTSeveral studieshave documented that the synthesis and release of NO—the primary dilator mechanism in microvessels—is altered during sepsis. Opposing results exist,however, regarding circulating NO levels during endotoxemia or sepsis. These results are most likely caused by differences in cNOS and iNOS at various stages of sepsis. The overproduction of NO is thought to be caused by the activity of iNOS in macrophages, hepatocytes, and neutrophils, which are activated during sepsis and partially released into the systemic circulation.A decreased production of NO has been attributed to a down-regulation of cNOS, which in the initial stages of sepsis is the major endothelial-dependent control mechanism of dilator function. The quantification of microcirculatory reactivity has not been studied in the face of changing levels of NO production by the endothelial cell. We conducted the present study to determine changes in the endothelial-dependent relaxation response during the developmental phases of sepsis.Our results indicate that A3 and A4 striated muscle arterioles become less sensitive to endothelial-dependent relaxation during the development of chronic sepsis. These results correlate with the common finding of diminished constrictor and dilator responses during late-stage sepsis.These data also support the recent finding that endothelial cell–derived NO is decreased in late sepsis.Because ACH stimulates endogenous NO release primarily through cNOS, the decrease in microvascular dilator response is indicative of either a diminished or an impaired capability of endothelial cells to release NO or a decreased responsiveness of the smooth muscle cell.Parker and Adamshave reported that the process of relaxation by an endothelial-dependent, receptor-independent agonist was unaffected in guinea pig aortic and coronary vascular rings exposed to endotoxin, but was diminished for receptor-dependent agonists (ACH and adenosine diphosphate). In addition, the stimulation of a different endothelial receptor with substance P elicited a relaxation response similar to that in controls, leading to the conclusion that endotoxin inhibits endothelial cell cNOS activity and microvascular reactivity by disrupting receptor-coupled activation mechanisms shared by ACH and adenosine diphosphate.These results, coupled with our present findings, indicate that the signal transduction pathway for receptor activation of cNOS by ACH becomes disabled over time in the small precapillary microvessels when continually stimulated by sepsis. The present study documents depressed endothelial-dependent dilator function that develops over time. These endothelial cell findings thus support a diminished ACH or NOS signal transduction pathway.The possibility that physical damage to the endothelium in the small arterioles may be responsible for the diminished response was not supported in this study. According to Furchgott and Zawadzki,the presence of intact and functional endothelial cells is necessary for smooth muscle cell relaxation in response to ACH application. If endothelial cells are denuded or damaged, ACH application will result in smooth muscle cell constriction.Acetylcholine-stimulated constriction was not observed at any point during this study; thus, the endothelial lining of the cremaster microvasculature remained structurally intact. Other investigatorshave also ruled out endothelial cell damage as the cause of decreased sensitivity of the endothelial cell–dependent relaxation and constriction responses.Increasing amounts of the cytokines tumor necrosis factor α and interleukin 1β have been shown to directly inhibit endothelial-dependent relaxation by reducing NO production.Myers et alalso show that tumor necrosis factor α and interleukin 1β do not elicit the release of endothelium-derived relaxing factor from endothelial cNOS. Therefore, the release of cytokines induced by the systemic inflammatory response may be partly responsible for the apparent decrease in NO release and diminished endothelial-dependent relaxation noted in A3 and A4 vessels during chronic infection. Our data clearly show a reduced pD2value at 72 hours of infection in the precapillary microvessels, which did not occur during the earlier stages of infection.Investigatorsrecently reported an initial increase in endothelial-derived NO and vascular smooth muscle cyclic guanosine monophosphate following cecal ligation and puncture. The present data correlate with these findings in that an increased sensitivity to ACH was observed in A4 arterioles 4 hours after bacterial injection. Because ACH elicits endogenous NO release from the endothelial cell, it can be inferred that either NOS activity is up-regulated in the short term or that there is an initial increase in ACH receptors. An explanation for the increased sensitivity in A4 arterioles and not in A3 arterioles during early sepsis remains to be determined, but it is thought that paracrine microvascular control becomes dominant in the most distal microvessels.Previous studieshave shown substantial vasoconstriction or vasodilation in first-order arterioles during different phases of acute bacteremia. In our present study, the diameters of A1 vessels tended to be larger at baseline in the later stages of sepsis, but these differences were not significant (P>.05). The diameters of A1 vessels did not change significantly in any experimental group with increasing concentrations of ACH or sodium nitroprusside. The fact that neither of these agonists elicited a response indicates that these microvessels were maximally dilated from the onset of the experiment.Although this unresponsiveness may be an artifact caused by the surgical preparation, long-term stimulation can shift microvascular control in the vascular tree and relegate the larger vessels to passive conduits, as occurs with hypertension.Another possible explanation is that because these microvessels are under neural control, the use of an anesthetic may have masked or blunted any A1 or V1 microvascular response. The large A1 and A2 microvessels are the primary resistance vessels in striated muscle. The fact that they were maximally dilated throughout our studies could be a factor in the decrease in peripheral vascular resistance noted during sepsis.The time points of 4, 24, and 72 hours for investigation were chosen based on previous models to represent a progression from acute to chronic infection. Inducible NOS is thought to appear in substantial levels 6 hours after an inflammatory stimulus.Thus, the 4-hour point during early stages of bacteremia was chosen to study the short-term effects of this model on cNOS activity exclusively. The 24-hour point was chosen to represent established infection with the development and stimulation of cNOS and iNOS activity on the microvascular endothelium. The 72-hour point was selected to represent the chronic inflammatory state so that the temporal changes of the endothelial-dependent relaxation response could be studied over time.Results of this study support the use of this model in the investigation of chronic sepsis. Previous studies have used either short-term septic challenge or intraperitoneal abscess formation. The sponge implantation on the back was used to avoid a local peritoneal reaction, which could affect cremaster muscle blood flow. This model has been well documented and closely resembles the signs and symptoms of chronic clinical sepsis with intermittent aerobic and anaerobic bacteremia, increased white blood cell counts, elevated CO sustained over time, and a protein catabolic state with loss of weight and muscle mass.Repeated injections of the preformed abscess cavity achieves a sustained hyperdynamic state without death or an additional augmentation of the hyperdynamic response once the septic state has been reached.The diminished endothelial-dependent relaxation response correlates with the findings of a variety of other studies of chronic sepsis.The mean arterial pressure and heart rate remained stable (±10% from baseline) throughout the experiment. The CO did not demonstrate a significant change between experimental groups, and the baseline CO at 72 hours was 36.9 mL/min, which correlates with the definition of this model as a model of chronic hyperdynamic sepsis.The reason for the hypertensive response noted at 24 hours in the group with established infection is unexplained by this study, but it represents a transition period in activity.CONCLUSIONSOur data suggest that a differential response to septic challenge develops over time in striated muscle tissue. It is suggested that NOS activity is increased initially, but NOS activity becomes depressed in precapillary arterioles during the development of chronic sepsis. Large A1 arterioles, the primary resistance vessels in large striated muscles, are maximally dilated, which may account for the depression in peripheral vascular resistance noted during chronic hyperdynamic sepsis. Because microvessel baseline diameters were similar in the control groups and in those with sepsis, a balance between dilation and constriction is apparently maintained. Our data, however, indicate the development of a depression in the dilator response of precapillary arterioles and support the idea that this depression of endothelial-dependent relaxation is a result of a disruption in the ACH or NOS signal transduction pathways in intact endothelial cells.Statement of Clinical RelevanceThe clinical septic state is characterized by an elevation of cardiac output, a decrease in peripheral vascular resistance, and ongoing systemic acidosis. These variables hint that nutrient microvascular blood flow to vital organs is impaired, while skeletal muscle blood flow is preserved. This present study defines an alteration in dilation control mechanisms for the large skeletal muscle microvascular bed during chronic sepsis. The enhanced endothelial cell dilator capacity noted in this study contributes to the decreased peripheral vascular resistance noted during clinical sepsis.RSConnellMWHarrisonLMela-RikerQuantitative ultrastructure of skeletal and cardiac muscle in hyperdynamic sepsis.Prog Clin Biol Res.1998;264:325-332.JBoczkowskiEVicautMAubierIn vivo effects of Escherichia coliendotoxemia on diaphragmatic microcirculation in rats.J Appl Physiol.1992;72:2219-2224.ASLübbeRNGarrisonHMCryerNLAlsipPDHarrisEDRF as a possible mediator of sepsis-induced arteriolar dilation in skeletal muscle.Am J Physiol.1992;262(pt 2):H880-H887.SMoncadaRMJPalmerEAHiggsNitric oxide: physiology, pathophysiology, and pharmacology.Pharmacol Rev.1991;43:109-142.RMJPalmerThe discovery of nitric oxide in the vessel wall: a unifying concept in the pathogenesis of sepsis.Arch Surg.1993;128:396-401.JLParkerHRAdamsSelective inhibition of endothelium-dependent vasodilator capacity by Escherichia coliendotoxemia.Circ Res.1993;72:539-551.PWangZFBaIHChaudryEndothelial-dependent relaxation is depressed at the macro- and microcirculatory levels during sepsis.Am J Physiol.1995;269(pt 2):R988-R994.PRMyersJLParkerMATannerHRAdamsEffects of cytokines tumor necrosis factor alpha and interleukin 1 beta on endotoxin-mediated inhibition of endothelium-derived relaxing factor bioactivity and nitric oxide production in vascular endothelium.Shock.1994;1:73-78.PRMyersTFWrightMATannerHRAdamsEDRF and nitric oxide production in cultured endothelial cells: direct inhibition by E. coliendotoxin.Am J Physiol.1992;262(pt 2):H710-H718.TEvansACarpenterHKindermanJCohenEvidence of increased nitric oxide production in patients with the sepsis syndrome.Circ Shock.1993;41:77-81.JEParrilloPathogenic mechanisms of septic shock.N Engl J Med.1993;328:1471-1477.PWangZFBaIHChaudryNitric oxide: to block or enhance its production during sepsis.Arch Surg.1994;129:1137-1143.RAPittnerJASpitzerEndotoxin and TNF alpha directly stimulate nitric oxide formation in cultured rat hepatocytes from chronically endotoxemic rats.Biochem Biophys Res Commun.1992;185:430-435.DAGellerPDFreeswickDNguyenDifferential induction of nitric oxide synthase in hepatocytes during endotoxemia and the acute-phase response.Arch Surg.1994;129:165-171.FMRobertsonDJOffnerDPCiceriWKBeckerBAPruittDetrimental hemodynamic effects of nitric oxide synthase inhibition in septic shock.Arch Surg.1994;129:149-156.EAMinnardJShouHNaamaACechHGallagherJMDalyInhibition of nitric oxide synthase is detrimental during endotoxemia.Arch Surg.1994;129:142-148.SBaezAn open cremaster muscle preparation for the study of blood vessels by in vivo microscopy.Microvasc Res.1973;3:384-394.APetrosDBennettPVallanceEffect of nitric oxide synthase inhibitors on hypotension in patients with septic shock.Lancet.1991;338:1557-1558.SHWursterPWangREDeanIHChaudryVascular smooth muscle contractile function is impaired during early and late stages of sepsis.J Surg Res.1994;56:556-561.RFFurchgottJVZawadzkiThe obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine.Nature.1980;288:373-376.PWangZFBaIHChaudryAdministration of tumor necrosis factor-alpha in vivo depresses endothelium-dependent relaxation.Am J Physiol.1994;266(pt 2):H2535-H2541.IFlemingGJulou-SchaefferGHGrayJRParrattJCStocletEvidence that an L-arginine/nitric oxide dependent elevation of tissue cyclic GMP content is involved in depression of vascular reactivity by endotoxin.Br J Pharmacol.1991;103:1047-1052.HMCryerRNGarrisonPDHarrisRole of muscle microvasculature during hyperdynamic and hypodynamic phases of endotoxin shock in decerebrate rats.J Trauma.1988;28:312-318.MAHillGAMeiningerCalcium entry and myogenic phenomena in skeletal muscle arterioles.Am J Physiol.1994;267(pt 2):H1085-H1092.ASLübbeRNGarrisonPDHarrisHMCryerHypertension alters microvascular responses in skeletal muscle to hyperdynamic bacteremia and hypodynamic Escherichia colisepsis.J Surg Res.1989;6:108-117.AMLeferEndotoxin, cytokines, and nitric oxide in shock: an editorial comment.Shock.1994;1:79-80.CHBakerETSuttonArteriolar endothelium-dependent vasodilation occurs during endotoxin shock.Am J Physiol.1993;264(pt 2):H1118-H1123.LMela-RikerPAlexanderDBartosChronic hyperdynamic sepsis in the rat, I: characterization of the animal model.Circ Shock.1988;25:231-244.Fred A. Luchette, MD, Cincinnati, Ohio:This well-thought-out and -conducted study is a continuation of this group's research efforts in describing the alterations in the microcirculation during sepsis. An increasing number of studies indicate that the vascular endothelial cell contributes to the control of vascular smooth muscle contractile function by synthesis and release of NO under normal as well as adverse circulatory conditions. Acetylcholine, a receptor-dependent agonist, on the endothelial cell membrane leads to an increase in intracellular calcium stores from the endoplasmic reticulum. This elevation in cytoplasmic calcium concentration activates cNOS, which induces vascular relaxation. New literature contains many reports that contradict each other regarding the impact of endotoxemia or sepsis on endothelial NO production.Some reports claim that sepsis enhances production, and others demonstrate a significant reduction in production during endotoxemia. These opposing results may be because these studies were conducted at different times after the initiation of the septic insult. A major difference in the various study designs is the use of vessel rings for in vitro assessment of vascular reactivity with topical application of ACH and nitroglycerin. In contrast, the work presented here uses an in vivo model of skeletal muscle microcirculation to determine the endothelial-dependent relaxation response during the progression of infection from an acute to a chronic state. I have several questions for you, Dr Garrison.First, you demonstrate with your hemodynamic measurements that the animals do develop an increased heart rate and mean blood pressure only at 24 hours or the established infection group. At 72 hours, or the chronic state, they did not demonstrate the hemodynamic measures consistent with a hypoperfusion or hypotensive state. You also did not report any data about the acid-base status of these animals. This data would help support that the animals were in a chronic or low-CO septic state. I am not convinced that this model represents chronic sepsis seen in clinical practice. Would you comment on these 2 points?In the abstract, you conclude that maximal dilation of the A1 vessels contributes to the decrease in peripheral vascular resistance, yet you did not present any data about systemic vascular resistance. Can you really make that conclusion?My second question deals with the topical application of ACH. The production of endothelium-derived NO is the focus of your hypothesis. It is not clear to me how your muscle preparation, which uses a tissue bath for muscle incubation with ACH added to the bath medium, allows interaction of the ACH with the endothelium of any of the microcirculation. It would seem the ideal route of application would be vascular rather than topical. With the concentrations used, do you know if the animals would tolerate the same dose given intravenously? How do you know the ACH is getting to the endothelium? Rather, is the ACH interacting with autonomic neural endplates on the smooth muscle or directly on the smooth muscle rather than the endothelium? Could this explain the lack of significant vasodilation of the A4 vessels at 4 hours? Are the A1 vessels too thick to see any effect on the endothelium with application from the adventitial surface?My third comment is directed at the role of local tissue metabolites affecting the vascular reactivity in the microcirculation. Dr Hasselgren in our department has shown that skeletal muscle cell membrane Na+-K+–adenosine triphosphatase activity is increased in septic animals and is associated with an increased production of lactic acid. Could the local tissue acid-base status in the chronic sepsis (72-hour) group overshadow the role of NO and explain the depressed levels of reactivity in the chronic state? At 72 hours, could the lack of dilatory response to ACH be due to chronic local tissue acidosis causing maximal vasodilation of the microcirculation such that at the time of in vitro study, the vascular smooth muscle is fatigued from the local metabolites and not vasodilating because the endothelium is unable to produce NO?Finally, Dr Wang from Brown University reported findings on the role of adrenomedullin in producing the hyperdynamic state during early sepsis. Would you care to speculate about the interrelationship of endothelial NO (local control) and adrenomedullin (systemic regulation) during early sepsis? Which of these 2 has a more central or primary role?I look forward to your comments and thank the program committee and society for the privilege of opening the discussion.Dr Garrison:The model was chosen as a chronic infection model that simulates systemic inflammation without affecting peritonitis. You can certainly criticize our use of it, but the blood supply to the cremaster muscle must go through the lining of the peritoneum, or adjacent to it, and we tried to avoid any stimulation of those proximal vessels on its effect on the isolated skeletal muscle. We have used cremaster muscle simply because it is a thin muscle that can be spread out over our optical port and allows easy access to the striated muscle bed without having a significant amount of dissection or tissue trauma in setting up the model.You are correct that we didn't measure peripheral vascular resistance. I might add that we are studying the striated muscle bed, which accounts for about 80% to 85% of total peripheral vascular resistance with any one animal; it is the largest vascular bed. Within that vascular bed, 60% is determined by the sympathetic nervous system control of proximal A1 and A2 vessels. The A3 and A4 vessels are under paracrine control, and that represents about 40% of the total skeletal muscle vascular bed.However, in response to stress or stimulus, it is just the opposite. In this area, the A3 and A4 vessels represent the response beds, and with the exception of the acute infusion of epinephrine or direct stimulation of the sympathetic system, those vessels probably do not contribute much to the change in peripheral vascular resistance that we see. The A1 vessels are the resting vessels; the A3 and A4 vessels are the reactive vessels that locally determine blood flow.We used a topical application because that avoids the systemic effects of the drug. That is a powerful reason for using this technique where you can study local tissue reactivity without changing blood flow or pressure or whatever the perfusion is to that particular tissue. Admittedly, the vessel is exposed to ablumin, but in all of the other studies we have done, we have never seen a problem with exposure eventually to the vessel as the tissue preparation is very thin.The local tissue is maintained in an environment-controlled tissue bath, so acidosis is not an issue. It may well be in a whole animal where you have the lactic acid or other acidic compounds accumulate and secondary pH changes, but the tissue in this local bath is controlled, and so you eliminate the effects of pH. That would be a different study where you would change temperature or pH as the study variable.Finally, we have not done studies with adrenomedullin, but I suspect that that is an acute response, and therefore, the effect would be seen in the A1 and A2 vessels in the acute situation, and in the chronic situation the compound would not be of issue. I don't know this for a fact, though.Richard H. Turnage, MD, Dallas, Tex:Could you comment on this phenomenon in other vascular beds, in particular the splanchnic vascular bed? Have you made any of these observations in these vessels in that bed?Dr Garrison:As you know, we have done considerable work in the intestine. We have in our laboratory an intestinal microvascular preparation, a kidney preparation, and a liver preparation, all of which we can look at microvessels. We have done a significant amount of work in the intestinal bed, and there is, similarly, a depressed dilator response in the intestines in a 2-hit model. We have not done anything in this chronic abscess model, which is the model in the current study reported, but in a 2-hit model where the initial physiologic insult is hemorrhagic shock followed by a bacteremic incident, you can demonstrate that in the small A3 vessels, which are the premucosal blood vessels that serve the blood flow to the mucosa, there is a depressed dilator response.This study was supported in part by the Office of Research and Development, Department of Veterans Affairs Merit Review, Washington, DC.Presented at the 18th Annual Meeting of the Surgical Infection Society, New York, NY, May 1, 1998.Reprints: R. Neal Garrison, MD, Department of Surgery, University of Louisville School of Medicine, Louisville, KY 40292. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JAMA Surgery American Medical Association

Microvascular Endothelial Cell Control of Peripheral Vascular Resistance During Sepsis

Download PDF
8 pages

Loading next page...
 
/lp/american-medical-association/microvascular-endothelial-cell-control-of-peripheral-vascular-NXtO6R0MsU

References (28)

Publisher
American Medical Association
Copyright
Copyright 1998 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
ISSN
2168-6254
eISSN
2168-6262
DOI
10.1001/archsurg.133.12.1335
Publisher site
See Article on Publisher Site

Abstract

ObjectiveTo determine the endothelial-dependent control of decreased peripheral vascular resistance in skeletal muscle microvessels during evolving sepsis.Materials and InterventionsAcute (4 hours, n=7), established (24 hours, n=7), or chronic (72 hours, n=8) infection was induced in Sprague-Dawley rats (150-175 g) by injecting Escherichia coliand Bacteroides fragilis(1×109colony-forming units for both) into a subcutaneous sponge. Control animals were injected with an isotonic sodium chloride solution and analyzed at the same time points: (n=6-8 per group). Dilation in response to the topically applied endothelial-dependent agonist acetylcholine (ACH) (1×10−9to 1×10−5mol/L) was measured in inflow first-order (A1) and precapillary fourth-order (A4) arterioles in cremaster muscle in vivo with videomicroscopy. Acetylcholine dose-response curves were used to determine vascular reactivity by calculating the concentration of ACH necessary to elicit 50% of the maximal dilator response.Main Outcome MeasuresIn vivo reactivity of striated muscle microvessels to the dilation agonist ACH during acute, established, and chronic infection.ResultsA1 vessels were unresponsive to all doses of ACH at all time points. A4 vessels showed an increased dilator response during short-term treatment, which deteriorated over time to depressed dilation during chronic infection.ConclusionsPrecapillary A4 vessels have increased dilator reactivity during early sepsis, which progresses to depressed levels with chronic infection. A1 microvessels remain dilated and are not substantially influenced by endothelial dilator mechanisms initiated by ACH. Maximum dilation of the large A1 vessels appears to contribute to the decrease in peripheral vascular resistance noted during systemic infection.ACUTE SEPSIS manifests a hemodynamic state in which the cardiac output (CO) and heart rate are increased in an attempt to adequately supply peripheral perfusion. This high-output state occurs concomitantly with a decrease in systemic vascular resistance and oxygen consumption. These cardiovascular and metabolic changes reflect a failure to deliver or use oxygen at the cellular level, resulting in metabolic acidosis.Microvascular blood flow abnormalities, such as an increased number of capillaries with no flow, occur in striated muscle during sepsis.These metabolic, cardiovascular, and blood flow alterations indicate a need to assess skeletal muscle microvascular function to understand changes in peripheral resistance during the evolution of sepsis.Endothelial cells control the contractility of vascular smooth muscle cells primarily by the synthesis and release of nitric oxide (NO).Endothelial release of NO is altered during sepsis.The question remains whether endogenous NO activity in microvascular beds leads to the hemodynamic changes of sepsis. In the acute phase of sepsis, large arterioles are constricted and small terminal arterioles are dilated in rat cremaster muscle.This constriction of large-inflow arterioles reduces the overall blood volume delivered to capillaries and may ultimately contribute to metabolic acidosis.Sepsis typically begins with a hyperdynamic clinical presentation that includes tachycardia, tachypnea, acidosis, temperature elevation, leukocytosis, and early organ system dysfunction. As the infection persists, organ dysfunction, hypoperfusion, and hypotension ensue.During chronic sepsis, inducible nitric oxide synthase (iNOS) levels are increased, and constitutive NOS (cNOS) levels are decreased.These and many similar findings are the basis for iNOS inhibition as a treatment of sepsis.Recent studieshave shown, however, that a nonspecific inhibition of NOS is detrimental during the later stages of endotoxemia or septic shock. The effectiveness of iNOS inhibition may depend on the stage of sepsis, the capability of endothelial cells to release NO, and the selectivity of the antagonist. In vivo microvascular reactivity throughout the development of sepsis has not been studied. To determine the capacity of endogenous NO release by the striated muscle microvascular bed, we chose to analyze endothelial-dependent relaxation in response to acetylcholine (ACH) during the progression of infection from the acute to the chronic state. We hypothesized that the development of chronic sepsis would depress the endothelial-dependent relaxation because of a diminished capacity of the endothelial cell to release NO.MATERIALS AND METHODSANIMAL MODELMale Sprague-Dawley rats (125-135 g) were acclimated for 2 weeks in a facility approved by the American Association for Accreditation of Laboratory Animal Care. All studies were approved by the Institutional Animal Care and Use Committee and conformed to the National Institutes of Health Guidelines for Care and Use of Laboratory Animals. Animals were given an intraperitoneal injection of sodium pentobarbital (45 mg/kg). Under sterile conditions, a 2-cm2gauze sponge was placed in the subcutaneous space at the base of the tail. Animals were then housed individually and observed for 4 days to document good health and weight gain. On the fifth postoperative day, all of the animals demonstrated a weight gain (150-175 g) and normal food and water intake. On this day, the subcutaneous sponge was inoculated with either 2 mL of isotonic sodium chloride solution (control groups) or 1 mL of 1×109colony-forming units of Bacteroides fragilisplus 1 mL of 1×109colony-forming units of Escherichia coli(septic groups). This inoculation was repeated after 48 hours for animals studied at 72 hours. At 4, 24, or 72 hours, all animals were studied, respectively representing the study groups with acute, established, or chronic sepsis.The E coliwas obtained from a hospital-acquired pathogenic strain and kept frozen until used. Periodically, a colony was grown in 25 mL of tryptose-soy agar broth, streaked onto an agar plate, and incubated until individual colony growth appeared. Preceding each experiment, 1 colony was extracted from the agar plate, suspended in tryptose-soy agar broth (25 mL), and incubated at 37°C for 16 hours. The suspended bacteria were washed and suspended in isotonic sodium chloride solution. Serial dilutions were performed and measured spectrophotometrically at 620-nm wavelength to provide 1×109colony-forming units of bacteria in 1 mL of isotonic sodium chloride solution. The B fragiliswas obtained (American Type Culture Collection, Rockville, Md) in freeze-dried pellet form and suspended in Schaedler broth (40 mL) (Difco Laboratories, Detroit, Mich), which includes Oxyrase (0.8 mL) (Oxyrase Inc, Mansfield, Ohio) to maintain an anaerobic environment. Preceding each experiment, 100 µL of B fragilissuspension was added to 40 mL of fresh culture medium and incubated for 46 hours. The bacteria were then washed and calibrated in the same manner as used for E coli.EXPERIMENTAL PROTOCOLAt 4, 24, or 72 hours after initial sponge injection, each animal's cremaster muscle microvascular reactivity was analyzed in vivo by videomicroscopy. Animals were anesthetized with an intraperitoneal injection of sodium pentobarbital (45 mg/kg), and a tracheotomy tube was positioned for airway control. Transpulmonary thermodilution CO measurements were made with a small-diameter thermistor positioned in the aortic arch through the right carotid artery. Each CO measurement was made by injecting a bolus of room temperature isotonic sodium chloride (40 µL) into the superior vena cava by a jugular vein catheter. All CO measurements were made in triplicate every 10 minutes throughout the experiment. Mean arterial pressure and heart rate measurements were obtained continuously through a right femoral artery cannula.The cremaster muscle videomicroscopy preparation was originally described by Baez.Briefly, the scrotal sac was cut longitudinally, and the cremaster-covered testicle was exposed. Layers of connective tissue were gently peeled off. The cremaster muscle, with neurovascular bundle intact, was cut longitudinally and secured in a custom-made tissue bath over an optical port. Throughout the rest of the surgical procedure, the muscle tissue was kept moist with Krebs solution (sodium bicarbonate, 15.5 mmol/L; sodium chloride, 112.9 mmol/L; potassium chloride, 4.7 mmol/L; and calcium chloride diluted 1:2 parts in water, 2.55 mmol/L). The bath was filled with Krebs solution, continuously bubbled with nitrogen and carbon dioxide to maintain the pH between 7.35 and 7.45, and heated to 34°C. Pancuronium bromide, 0.1 mL of 1×10−6mol/L, was added to the bath to prevent muscle spasms.The cremaster microcirculation was directly observed by a trinocular transillumination microscope (model ACM, Carl Zeiss, Inc, Thornwood, NY), and microvascular images were transmitted by videocamera (model 5100 HS, Panasonic, Secaucus, NJ) to a monitor and videocassette recorder at ×1246 magnification. Vessel intraluminal diameters were measured on the monitor and converted to micrometers. Vascular images were read from the videocassette recording using still images to ensure accuracy.The branch order of the cremaster microvascular anatomy was determined by assigning the major inflow arteriole as the first-order arteriole (A1) and the parallel adjacent venule as the first-order venule (V1). Perpendicular branches from A1 microvessels were termed second-order arterioles(A2), and successive arteriolar branches from the second-order vessels were designated third-order(A3) and fourth-order(A4) vessels. These smaller vessels occur without adjacent venules and are terminal precapillary arterioles.Following the surgical procedure, the cremaster muscle was allowed to equilibrate in the tissue bath for 1 hour. Baseline measurements were taken of mean arterial pressure, heart rate, CO, and microvascular diameters of A1, V1, A3, and A4 microvessels. A second set of baseline measurements was repeated 10 minutes after the first set and was required to be within 10% of the initial measurements for entry into the experimental protocol. After baseline measurements were obtained, serial doses of ACH were added to the bath (final concentrations, 1×10−9to 1×10−5mol/L). Each dose of ACH remained in the bath for 10 minutes. Hemodynamic measurements were taken during the first minute of exposure, and vessel diameters were recorded at 1, 3, 5, 7, and 9 minutes for each dose of ACH. Sodium nitroprusside was added after the final ACH dose, and hemodynamic and vessel diameter measurements were taken 5 minutes later to determine maximal vessel dilation.Dose-response curves for dilation were generated by normalizing the measured vessel diameters with reference to the maximal vessel dilation. The mean baseline value was set at 0, and the maximal vessel diameters with sodium nitroprusside application were set at 100% maximal dilator capacity. The following equation was used to calculate the percentage of dilator capacity: (X−BL)/(NP−BL)×100%, where X indicates vessel diameter; BL, vessel diameter at baseline; and NP, maximal vessel diameter. From the ACH dose-response curve of each animal, vascular sensitivity was determined by the concentration of ACH necessary to elicit 50% of the maximal dilator response (EC50). The pD2value was determined as the negative log EC50, which represents the changes in microvascular sensitivity to ACH.STATISTICAL ANALYSISWithin each experimental group, the baseline values were compared with ACH doses using repeated-measures analysis of variance, followed by the Tukey-Kramer least significant difference test. Experimental groups were compared using the 1-way analysis of variance, followed by the Tukey-Kramer least significant difference test. The null hypothesis was rejected when P<.05.RESULTSHEMODYNAMICSBaseline hemodynamic values are reported in Table 1. The mean arterial pressure was similar in all experimental groups, except at 24 hours in the established infection group, which showed a 24% increase (P<.05) when compared with controls that received isotonic sodium chloride, and a 14% increase in heart rate (P<.05). Although there was a trend toward a hyperdynamic cardiovascular state in the animals with chronic infection, these changes were not significant (P>.05). All hemodynamic variables were stable during intravital microscopy and dose-response protocols.Hemodynamic Baseline Values for All Experimental Groups&ast;Infection GroupMean Arterial Pressure, mm HgHeart Rate, Beats/minCardiac Output, mL ⋅ min−1⋅ 100 g−1Acute4-h isotonic sodium chloride solution (n = 8)107.0 ± 3.5414.0 ± 14.017.5 ± 2.84-h bacteria (n = 7)106.0 ± 3.5400.0 ± 11.519.3 ± 3.8Established24-h isotonic sodium chloride solution (n = 7)106.0 ± 2.2382.0 ± 13.522.9 ± 3.024-h bacteria (n = 7)144.0 ± 7.9†457.0 ± 12.2†16.6 ± 2.9Chronic72-h isotonic sodium chloride solution (n = 6)112.0 ± 2.6371.0 ± 9.028.5 ± 3.572-h bacteria (n = 8)118.0 ± 4.5398.0 ± 8.636.9 ± 3.7&ast;Data are reported as mean ± SEM.†P<.05 vs isotonic sodium chloride control (analysis of variance followed by the Tukey-Kramer test).MICROVASCULAR RESPONSESFirst-order arterioles and venules were unresponsive to increasing concentrations of ACH and sodium nitroprusside. These first-order microvessels maintained diameters within 10% of baseline values in all experimental groups, and dose-response curves were flat. In the group showing chronic sepsis after 72 hours, the baseline diameters of A1 and V1 microvessels were slightly larger than in the groups with acute or established infection.Although the A3 microvessels in the groups with either acute or established infection showed no change in reactivity compared with those in controls, there was a shift in the EC50to the right in the group with chronic sepsis, indicating a substantial depression of dilator function (Figure 1). Precapillary A4 microvessels demonstrated a differential shift in reactivity to dilator stimulation, with an increased sensitivity and shift to the left of the EC50noted in the groups with acute sepsis (Figure 2) and a definite depression of dilator function and shift to the right of the EC50after 72 hours of infection (Figure 3). There was no change in reactivity of A4 microvessels in the groups with established infection compared with control animals. This differential shift in microvascular sensitivity to dilator stimulation is reflected in the calculated pD2values for A3 and A4 microvessels in each group. The pD2value was depressed in A3 vessels during chronic infection at 72 hours (Figure 4), whereas A4 vessels showed an early increase in the pD2value during acute infection at 4 hours, indicating a shift over time in these small precapillary microvessels.Figure 1.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A3 arterioles after 72 hours of bacterial infection. Microvascular responses are reported as the percentage (mean±SEM) of maximum dilator capacity (for the method of calculation, see the "Experimental Protocol" subsection in the "Materials and Methods" section). Open circles indicate control animals (n=6); closed squares, the group with chronic infection (n=8); BL, baseline; NP, maximal vessel diameter; and asterisk, P<.05 (by analysis of variance, followed by the Tukey-Kramer test) for isotonic sodium chloride control vs bacterial treatment.Figure 2.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A4 arterioles after 4 hours of bacterial infection. Microvascular response reporting is as given in the legend for Figure 1. No significant differences were found between the 2 experimental groups. Open circles indicate control animals (n=8); closed squares, the group with acute infection (n=7); BL, baseline; and NP, maximal vessel diameter.Figure 3.Effect of increasing concentrations of acetylcholine and sodium nitroprusside (1×10−4mol/L) on A4 arterioles after 72 hours of bacterial infection. Microvascular response reporting is as given in the legend for Figure 1. Open circles indicate control animals (n=6); closed squares, the group with chronic infection (n=8); BL, baseline; NP, maximal vessel diameter; and asterisk, P<.05 (by analysis of variance, followed by the Tukey-Kramer test) for isotonic sodium chloride control vs bacterial treatment.Figure 4.pD2values of A3 and A4 arterioles in control animals (light-shaded bars) and animals with acute, established, or chronic infection (dark-shaded bars). pD2indicates −log EC50, or dose required to elicit a 50% response of maximal dilation; asterisk, P<.05 (by analysis of variance) vs isotonic sodium chloride control.COMMENTSeveral studieshave documented that the synthesis and release of NO—the primary dilator mechanism in microvessels—is altered during sepsis. Opposing results exist,however, regarding circulating NO levels during endotoxemia or sepsis. These results are most likely caused by differences in cNOS and iNOS at various stages of sepsis. The overproduction of NO is thought to be caused by the activity of iNOS in macrophages, hepatocytes, and neutrophils, which are activated during sepsis and partially released into the systemic circulation.A decreased production of NO has been attributed to a down-regulation of cNOS, which in the initial stages of sepsis is the major endothelial-dependent control mechanism of dilator function. The quantification of microcirculatory reactivity has not been studied in the face of changing levels of NO production by the endothelial cell. We conducted the present study to determine changes in the endothelial-dependent relaxation response during the developmental phases of sepsis.Our results indicate that A3 and A4 striated muscle arterioles become less sensitive to endothelial-dependent relaxation during the development of chronic sepsis. These results correlate with the common finding of diminished constrictor and dilator responses during late-stage sepsis.These data also support the recent finding that endothelial cell–derived NO is decreased in late sepsis.Because ACH stimulates endogenous NO release primarily through cNOS, the decrease in microvascular dilator response is indicative of either a diminished or an impaired capability of endothelial cells to release NO or a decreased responsiveness of the smooth muscle cell.Parker and Adamshave reported that the process of relaxation by an endothelial-dependent, receptor-independent agonist was unaffected in guinea pig aortic and coronary vascular rings exposed to endotoxin, but was diminished for receptor-dependent agonists (ACH and adenosine diphosphate). In addition, the stimulation of a different endothelial receptor with substance P elicited a relaxation response similar to that in controls, leading to the conclusion that endotoxin inhibits endothelial cell cNOS activity and microvascular reactivity by disrupting receptor-coupled activation mechanisms shared by ACH and adenosine diphosphate.These results, coupled with our present findings, indicate that the signal transduction pathway for receptor activation of cNOS by ACH becomes disabled over time in the small precapillary microvessels when continually stimulated by sepsis. The present study documents depressed endothelial-dependent dilator function that develops over time. These endothelial cell findings thus support a diminished ACH or NOS signal transduction pathway.The possibility that physical damage to the endothelium in the small arterioles may be responsible for the diminished response was not supported in this study. According to Furchgott and Zawadzki,the presence of intact and functional endothelial cells is necessary for smooth muscle cell relaxation in response to ACH application. If endothelial cells are denuded or damaged, ACH application will result in smooth muscle cell constriction.Acetylcholine-stimulated constriction was not observed at any point during this study; thus, the endothelial lining of the cremaster microvasculature remained structurally intact. Other investigatorshave also ruled out endothelial cell damage as the cause of decreased sensitivity of the endothelial cell–dependent relaxation and constriction responses.Increasing amounts of the cytokines tumor necrosis factor α and interleukin 1β have been shown to directly inhibit endothelial-dependent relaxation by reducing NO production.Myers et alalso show that tumor necrosis factor α and interleukin 1β do not elicit the release of endothelium-derived relaxing factor from endothelial cNOS. Therefore, the release of cytokines induced by the systemic inflammatory response may be partly responsible for the apparent decrease in NO release and diminished endothelial-dependent relaxation noted in A3 and A4 vessels during chronic infection. Our data clearly show a reduced pD2value at 72 hours of infection in the precapillary microvessels, which did not occur during the earlier stages of infection.Investigatorsrecently reported an initial increase in endothelial-derived NO and vascular smooth muscle cyclic guanosine monophosphate following cecal ligation and puncture. The present data correlate with these findings in that an increased sensitivity to ACH was observed in A4 arterioles 4 hours after bacterial injection. Because ACH elicits endogenous NO release from the endothelial cell, it can be inferred that either NOS activity is up-regulated in the short term or that there is an initial increase in ACH receptors. An explanation for the increased sensitivity in A4 arterioles and not in A3 arterioles during early sepsis remains to be determined, but it is thought that paracrine microvascular control becomes dominant in the most distal microvessels.Previous studieshave shown substantial vasoconstriction or vasodilation in first-order arterioles during different phases of acute bacteremia. In our present study, the diameters of A1 vessels tended to be larger at baseline in the later stages of sepsis, but these differences were not significant (P>.05). The diameters of A1 vessels did not change significantly in any experimental group with increasing concentrations of ACH or sodium nitroprusside. The fact that neither of these agonists elicited a response indicates that these microvessels were maximally dilated from the onset of the experiment.Although this unresponsiveness may be an artifact caused by the surgical preparation, long-term stimulation can shift microvascular control in the vascular tree and relegate the larger vessels to passive conduits, as occurs with hypertension.Another possible explanation is that because these microvessels are under neural control, the use of an anesthetic may have masked or blunted any A1 or V1 microvascular response. The large A1 and A2 microvessels are the primary resistance vessels in striated muscle. The fact that they were maximally dilated throughout our studies could be a factor in the decrease in peripheral vascular resistance noted during sepsis.The time points of 4, 24, and 72 hours for investigation were chosen based on previous models to represent a progression from acute to chronic infection. Inducible NOS is thought to appear in substantial levels 6 hours after an inflammatory stimulus.Thus, the 4-hour point during early stages of bacteremia was chosen to study the short-term effects of this model on cNOS activity exclusively. The 24-hour point was chosen to represent established infection with the development and stimulation of cNOS and iNOS activity on the microvascular endothelium. The 72-hour point was selected to represent the chronic inflammatory state so that the temporal changes of the endothelial-dependent relaxation response could be studied over time.Results of this study support the use of this model in the investigation of chronic sepsis. Previous studies have used either short-term septic challenge or intraperitoneal abscess formation. The sponge implantation on the back was used to avoid a local peritoneal reaction, which could affect cremaster muscle blood flow. This model has been well documented and closely resembles the signs and symptoms of chronic clinical sepsis with intermittent aerobic and anaerobic bacteremia, increased white blood cell counts, elevated CO sustained over time, and a protein catabolic state with loss of weight and muscle mass.Repeated injections of the preformed abscess cavity achieves a sustained hyperdynamic state without death or an additional augmentation of the hyperdynamic response once the septic state has been reached.The diminished endothelial-dependent relaxation response correlates with the findings of a variety of other studies of chronic sepsis.The mean arterial pressure and heart rate remained stable (±10% from baseline) throughout the experiment. The CO did not demonstrate a significant change between experimental groups, and the baseline CO at 72 hours was 36.9 mL/min, which correlates with the definition of this model as a model of chronic hyperdynamic sepsis.The reason for the hypertensive response noted at 24 hours in the group with established infection is unexplained by this study, but it represents a transition period in activity.CONCLUSIONSOur data suggest that a differential response to septic challenge develops over time in striated muscle tissue. It is suggested that NOS activity is increased initially, but NOS activity becomes depressed in precapillary arterioles during the development of chronic sepsis. Large A1 arterioles, the primary resistance vessels in large striated muscles, are maximally dilated, which may account for the depression in peripheral vascular resistance noted during chronic hyperdynamic sepsis. Because microvessel baseline diameters were similar in the control groups and in those with sepsis, a balance between dilation and constriction is apparently maintained. Our data, however, indicate the development of a depression in the dilator response of precapillary arterioles and support the idea that this depression of endothelial-dependent relaxation is a result of a disruption in the ACH or NOS signal transduction pathways in intact endothelial cells.Statement of Clinical RelevanceThe clinical septic state is characterized by an elevation of cardiac output, a decrease in peripheral vascular resistance, and ongoing systemic acidosis. These variables hint that nutrient microvascular blood flow to vital organs is impaired, while skeletal muscle blood flow is preserved. This present study defines an alteration in dilation control mechanisms for the large skeletal muscle microvascular bed during chronic sepsis. The enhanced endothelial cell dilator capacity noted in this study contributes to the decreased peripheral vascular resistance noted during clinical sepsis.RSConnellMWHarrisonLMela-RikerQuantitative ultrastructure of skeletal and cardiac muscle in hyperdynamic sepsis.Prog Clin Biol Res.1998;264:325-332.JBoczkowskiEVicautMAubierIn vivo effects of Escherichia coliendotoxemia on diaphragmatic microcirculation in rats.J Appl Physiol.1992;72:2219-2224.ASLübbeRNGarrisonHMCryerNLAlsipPDHarrisEDRF as a possible mediator of sepsis-induced arteriolar dilation in skeletal muscle.Am J Physiol.1992;262(pt 2):H880-H887.SMoncadaRMJPalmerEAHiggsNitric oxide: physiology, pathophysiology, and pharmacology.Pharmacol Rev.1991;43:109-142.RMJPalmerThe discovery of nitric oxide in the vessel wall: a unifying concept in the pathogenesis of sepsis.Arch Surg.1993;128:396-401.JLParkerHRAdamsSelective inhibition of endothelium-dependent vasodilator capacity by Escherichia coliendotoxemia.Circ Res.1993;72:539-551.PWangZFBaIHChaudryEndothelial-dependent relaxation is depressed at the macro- and microcirculatory levels during sepsis.Am J Physiol.1995;269(pt 2):R988-R994.PRMyersJLParkerMATannerHRAdamsEffects of cytokines tumor necrosis factor alpha and interleukin 1 beta on endotoxin-mediated inhibition of endothelium-derived relaxing factor bioactivity and nitric oxide production in vascular endothelium.Shock.1994;1:73-78.PRMyersTFWrightMATannerHRAdamsEDRF and nitric oxide production in cultured endothelial cells: direct inhibition by E. coliendotoxin.Am J Physiol.1992;262(pt 2):H710-H718.TEvansACarpenterHKindermanJCohenEvidence of increased nitric oxide production in patients with the sepsis syndrome.Circ Shock.1993;41:77-81.JEParrilloPathogenic mechanisms of septic shock.N Engl J Med.1993;328:1471-1477.PWangZFBaIHChaudryNitric oxide: to block or enhance its production during sepsis.Arch Surg.1994;129:1137-1143.RAPittnerJASpitzerEndotoxin and TNF alpha directly stimulate nitric oxide formation in cultured rat hepatocytes from chronically endotoxemic rats.Biochem Biophys Res Commun.1992;185:430-435.DAGellerPDFreeswickDNguyenDifferential induction of nitric oxide synthase in hepatocytes during endotoxemia and the acute-phase response.Arch Surg.1994;129:165-171.FMRobertsonDJOffnerDPCiceriWKBeckerBAPruittDetrimental hemodynamic effects of nitric oxide synthase inhibition in septic shock.Arch Surg.1994;129:149-156.EAMinnardJShouHNaamaACechHGallagherJMDalyInhibition of nitric oxide synthase is detrimental during endotoxemia.Arch Surg.1994;129:142-148.SBaezAn open cremaster muscle preparation for the study of blood vessels by in vivo microscopy.Microvasc Res.1973;3:384-394.APetrosDBennettPVallanceEffect of nitric oxide synthase inhibitors on hypotension in patients with septic shock.Lancet.1991;338:1557-1558.SHWursterPWangREDeanIHChaudryVascular smooth muscle contractile function is impaired during early and late stages of sepsis.J Surg Res.1994;56:556-561.RFFurchgottJVZawadzkiThe obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine.Nature.1980;288:373-376.PWangZFBaIHChaudryAdministration of tumor necrosis factor-alpha in vivo depresses endothelium-dependent relaxation.Am J Physiol.1994;266(pt 2):H2535-H2541.IFlemingGJulou-SchaefferGHGrayJRParrattJCStocletEvidence that an L-arginine/nitric oxide dependent elevation of tissue cyclic GMP content is involved in depression of vascular reactivity by endotoxin.Br J Pharmacol.1991;103:1047-1052.HMCryerRNGarrisonPDHarrisRole of muscle microvasculature during hyperdynamic and hypodynamic phases of endotoxin shock in decerebrate rats.J Trauma.1988;28:312-318.MAHillGAMeiningerCalcium entry and myogenic phenomena in skeletal muscle arterioles.Am J Physiol.1994;267(pt 2):H1085-H1092.ASLübbeRNGarrisonPDHarrisHMCryerHypertension alters microvascular responses in skeletal muscle to hyperdynamic bacteremia and hypodynamic Escherichia colisepsis.J Surg Res.1989;6:108-117.AMLeferEndotoxin, cytokines, and nitric oxide in shock: an editorial comment.Shock.1994;1:79-80.CHBakerETSuttonArteriolar endothelium-dependent vasodilation occurs during endotoxin shock.Am J Physiol.1993;264(pt 2):H1118-H1123.LMela-RikerPAlexanderDBartosChronic hyperdynamic sepsis in the rat, I: characterization of the animal model.Circ Shock.1988;25:231-244.Fred A. Luchette, MD, Cincinnati, Ohio:This well-thought-out and -conducted study is a continuation of this group's research efforts in describing the alterations in the microcirculation during sepsis. An increasing number of studies indicate that the vascular endothelial cell contributes to the control of vascular smooth muscle contractile function by synthesis and release of NO under normal as well as adverse circulatory conditions. Acetylcholine, a receptor-dependent agonist, on the endothelial cell membrane leads to an increase in intracellular calcium stores from the endoplasmic reticulum. This elevation in cytoplasmic calcium concentration activates cNOS, which induces vascular relaxation. New literature contains many reports that contradict each other regarding the impact of endotoxemia or sepsis on endothelial NO production.Some reports claim that sepsis enhances production, and others demonstrate a significant reduction in production during endotoxemia. These opposing results may be because these studies were conducted at different times after the initiation of the septic insult. A major difference in the various study designs is the use of vessel rings for in vitro assessment of vascular reactivity with topical application of ACH and nitroglycerin. In contrast, the work presented here uses an in vivo model of skeletal muscle microcirculation to determine the endothelial-dependent relaxation response during the progression of infection from an acute to a chronic state. I have several questions for you, Dr Garrison.First, you demonstrate with your hemodynamic measurements that the animals do develop an increased heart rate and mean blood pressure only at 24 hours or the established infection group. At 72 hours, or the chronic state, they did not demonstrate the hemodynamic measures consistent with a hypoperfusion or hypotensive state. You also did not report any data about the acid-base status of these animals. This data would help support that the animals were in a chronic or low-CO septic state. I am not convinced that this model represents chronic sepsis seen in clinical practice. Would you comment on these 2 points?In the abstract, you conclude that maximal dilation of the A1 vessels contributes to the decrease in peripheral vascular resistance, yet you did not present any data about systemic vascular resistance. Can you really make that conclusion?My second question deals with the topical application of ACH. The production of endothelium-derived NO is the focus of your hypothesis. It is not clear to me how your muscle preparation, which uses a tissue bath for muscle incubation with ACH added to the bath medium, allows interaction of the ACH with the endothelium of any of the microcirculation. It would seem the ideal route of application would be vascular rather than topical. With the concentrations used, do you know if the animals would tolerate the same dose given intravenously? How do you know the ACH is getting to the endothelium? Rather, is the ACH interacting with autonomic neural endplates on the smooth muscle or directly on the smooth muscle rather than the endothelium? Could this explain the lack of significant vasodilation of the A4 vessels at 4 hours? Are the A1 vessels too thick to see any effect on the endothelium with application from the adventitial surface?My third comment is directed at the role of local tissue metabolites affecting the vascular reactivity in the microcirculation. Dr Hasselgren in our department has shown that skeletal muscle cell membrane Na+-K+–adenosine triphosphatase activity is increased in septic animals and is associated with an increased production of lactic acid. Could the local tissue acid-base status in the chronic sepsis (72-hour) group overshadow the role of NO and explain the depressed levels of reactivity in the chronic state? At 72 hours, could the lack of dilatory response to ACH be due to chronic local tissue acidosis causing maximal vasodilation of the microcirculation such that at the time of in vitro study, the vascular smooth muscle is fatigued from the local metabolites and not vasodilating because the endothelium is unable to produce NO?Finally, Dr Wang from Brown University reported findings on the role of adrenomedullin in producing the hyperdynamic state during early sepsis. Would you care to speculate about the interrelationship of endothelial NO (local control) and adrenomedullin (systemic regulation) during early sepsis? Which of these 2 has a more central or primary role?I look forward to your comments and thank the program committee and society for the privilege of opening the discussion.Dr Garrison:The model was chosen as a chronic infection model that simulates systemic inflammation without affecting peritonitis. You can certainly criticize our use of it, but the blood supply to the cremaster muscle must go through the lining of the peritoneum, or adjacent to it, and we tried to avoid any stimulation of those proximal vessels on its effect on the isolated skeletal muscle. We have used cremaster muscle simply because it is a thin muscle that can be spread out over our optical port and allows easy access to the striated muscle bed without having a significant amount of dissection or tissue trauma in setting up the model.You are correct that we didn't measure peripheral vascular resistance. I might add that we are studying the striated muscle bed, which accounts for about 80% to 85% of total peripheral vascular resistance with any one animal; it is the largest vascular bed. Within that vascular bed, 60% is determined by the sympathetic nervous system control of proximal A1 and A2 vessels. The A3 and A4 vessels are under paracrine control, and that represents about 40% of the total skeletal muscle vascular bed.However, in response to stress or stimulus, it is just the opposite. In this area, the A3 and A4 vessels represent the response beds, and with the exception of the acute infusion of epinephrine or direct stimulation of the sympathetic system, those vessels probably do not contribute much to the change in peripheral vascular resistance that we see. The A1 vessels are the resting vessels; the A3 and A4 vessels are the reactive vessels that locally determine blood flow.We used a topical application because that avoids the systemic effects of the drug. That is a powerful reason for using this technique where you can study local tissue reactivity without changing blood flow or pressure or whatever the perfusion is to that particular tissue. Admittedly, the vessel is exposed to ablumin, but in all of the other studies we have done, we have never seen a problem with exposure eventually to the vessel as the tissue preparation is very thin.The local tissue is maintained in an environment-controlled tissue bath, so acidosis is not an issue. It may well be in a whole animal where you have the lactic acid or other acidic compounds accumulate and secondary pH changes, but the tissue in this local bath is controlled, and so you eliminate the effects of pH. That would be a different study where you would change temperature or pH as the study variable.Finally, we have not done studies with adrenomedullin, but I suspect that that is an acute response, and therefore, the effect would be seen in the A1 and A2 vessels in the acute situation, and in the chronic situation the compound would not be of issue. I don't know this for a fact, though.Richard H. Turnage, MD, Dallas, Tex:Could you comment on this phenomenon in other vascular beds, in particular the splanchnic vascular bed? Have you made any of these observations in these vessels in that bed?Dr Garrison:As you know, we have done considerable work in the intestine. We have in our laboratory an intestinal microvascular preparation, a kidney preparation, and a liver preparation, all of which we can look at microvessels. We have done a significant amount of work in the intestinal bed, and there is, similarly, a depressed dilator response in the intestines in a 2-hit model. We have not done anything in this chronic abscess model, which is the model in the current study reported, but in a 2-hit model where the initial physiologic insult is hemorrhagic shock followed by a bacteremic incident, you can demonstrate that in the small A3 vessels, which are the premucosal blood vessels that serve the blood flow to the mucosa, there is a depressed dilator response.This study was supported in part by the Office of Research and Development, Department of Veterans Affairs Merit Review, Washington, DC.Presented at the 18th Annual Meeting of the Surgical Infection Society, New York, NY, May 1, 1998.Reprints: R. Neal Garrison, MD, Department of Surgery, University of Louisville School of Medicine, Louisville, KY 40292.

Journal

JAMA Surgery – American Medical Association

Published: Dec 1, 1998

There are no references for this article.