| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
MUCOSAL BIOLOGY
alters hemodynamics in newborn intestine: role of endothelin
Center for Cell and Vascular Biology, Columbus Children's Research Institute and Department of Pediatrics, The Ohio State University, Columbus, Ohio
Submitted 23 January 2006 ; accepted in final form 6 April 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
on newborn intestinal hemodynamics. IL-1 increased the release of ET-1 by primary endothelial cells in a dose-dependent manner; as well, it reduced expression of the endothelin (ET) type B (ETB) receptor on endothelial cells and increased expression of the ET type A (ETA) receptor on vascular smooth muscle cells. IL-1 increased endothelial cell endothelial nitric oxide (NO) synthase (eNOS) expression but did not enhance eNOS activity as evidenced by release of NOx into conditioned medium in response to acetylcholine or shear stress. The effects of IL-1 on flow-induced dilation were evaluated in terminal mesenteric arteries in vitro. Pretreatment with IL-1 (1 ng; 4 h) significantly attenuated vasodilation in response to flow rates of 100 and 200 µl/min. This effect was mediated, in part, by the endothelin ETA receptor; thus selective blockade of ETA receptors with BQ610 nearly restored flow-induced dilation. In contrast, exogenous ET-1 only shifted the diameter-flow curve downward without altering the percent vasodilation in response to flow. The effects of IL-1 on ileal oxygenation were then studied using in vivo gut loops. Intramesenteric artery infusion of IL-1 upstream of the gut loop caused ileal vasoconstriction and reduced the arterial-venous O2 difference across the gut loop; consequently, it reduced ileal oxygenation by 60%. This effect was significantly attenuated by pretreatment with BQ610. These data support a linkage between the proinflammatory cytokine IL-1 and vascular dysfunction within the intestinal circulation, mediated, at least in part, by the ET system.
newborn intestine; necrotizing enterocolitis; intestinal blood flow; intestinal oxygenation
, IL-6, IL-10, and IFN-
(12, 13, 59). The mechanistic basis for the ischemia relevant to NEC remains an enigma. One possibility, heretofore untested, is that proinflammatory cytokines affect the expression or function of local vascular regulatory systems within the newborn intestine in such a manner as to generate ischemia.
Precedents for the proposed interaction exist. IL-1 upregulates the vasoconstrictor peptide endothelin (ET)-1 in bovine (5) or porcine aortic (66) or human umbilical (38) endothelial cells (ECs) as well as vascular smooth muscle cells (VSMCs) harvested from the human saphenous vein or internal mammary artery (64, 65). In the latter, IL-1-induced activation of the nuclear transcription factor NF-
B underlies the enhanced transcription of prepro-ET, the precursor of ET-1 (49, 65).
IL-1 also affects the expression of ET receptors. Two subtypes of ET receptors have been identified: ETA and ETB. The former are primarily located on vascular smooth muscle, and their activation leads to smooth muscle contraction and hence vasoconstriction; the latter are primarily located on ECs, where ligand binding activates the endothelial isoform of nitric oxide (NO) synthase (NOS), leading to NO production and vasodilation (9). IL-1 increases 131I-labeled ET-1 binding in VSMC secondary to an increase in expression of the ETA receptor (44), i.e., the receptor responsible for generating ET-1-induced vasoconstriction (40). Simultaneous upregulation of ET-1 and the ETA receptor carries the potential for induction of profound and sustained ischemia capable of generating tissue hypoxia and injury, as has been demonstrated in the liver (15), intestine (48), and skeletal muscle (28).
This putative interaction between IL-1 and the ET system may be particularly applicable to the newborn intestine insofar as ET-1 is an important determinant of vascular resistance therein (41). ETA receptors are present in the newborn intestine (57); moreover, intramesenteric infusion of ET-1 at physiologically relevant levels causes profound sustained ischemia and tissue hypoxia in this tissue (42, 43). Most importantly, however, the intestine resected from human infants with NEC demonstrates an abundance of ET-1, whereas arterioles harvested from this tissue demonstrate a basal level of vasoconstriction that is significantly attenuated by ETA receptor blockade (45). The presence of both IL-1 and ET-1 in tissue removed from human infant NEC justifies further study of their potential interaction within the newborn intestine.
These experiments approached this problem by testing two hypotheses. The first hypothesis was that IL-1 upregulates the ET system in the newborn intestinal circulation and thus exerts a deleterious effect on intestinal hemodynamic regulation. The first part of this hypothesis was tested by determining the effects of IL-1 on ET-1 production and ET receptor expression at the cellular level, whereas the second was examined by noting the effects of IL-1 flow-induced dilation in terminal mesenteric arteries (TMAs) harvested from newborn swine and studied in vitro. TMAs (
150 µm) reside immediately upstream from the intramural gut microcirculation and are resistance vessels, a circumstance that renders them an excellent site to study intestinal hemodynamic regulation (53). The choice of flow-induced dilation as a benchmark of hemodynamic regulation was driven by the importance of this response in determining downstream O2 transport and hence tissue oxygenation (56), whereas an in vitro approach was selected to afford precise manipulation of environmental stimuli (41). The second hypothesis was that IL-1 compromises intestinal tissue oxygenation via upregulation of the ET system. This hypothesis was tested by measuring the effects of IL-1 on hemodynamics and oxygenation within isolated ileal gut loops studied under in vivo conditions. This approach was selected to provide data to confirm the physiological relevance of in vitro observations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The Institutional Animal Care and Use Committee of Columbus Children's Research Institute approved this work. Piglets were studied within 8 h of birth and were never allowed to initiate feeding before use to eliminate the confounding variable of nutrient absorption on newborn intestinal hemodynamics (7). Anesthesia was achieved by an intramuscular injection of 7.5 mg/kg telezol and 5.0 mg/kg xylazine and was sustained by pentobarbital sodium (5 mg/kg iv). Piglets were euthanized via an intravenous infusion of Succumb (1 ml/kg) while still anesthetized.
Pharmacological Reagents
All pharmaceuticals were obtained from Sigma (St. Louis, MO) and included ET-1, BQ610 (ETA receptor antagonist), and BQ788 (ETB receptor antagonist). All agents were dissolved in Krebs buffer of the following composition (in mM): 118 NaCl, 4.5 KCl, 2.5 CaCl2, 1.2 MgSO4, 1.2 KH2PO4, 25.0 NaHCO3, 11.0 glucose, and 0.026 EDTA; pH was 7.4 when the buffer was aerated with 21% O2-5% CO2-balance N2 at 37°C. This buffer also served as the vehicle given to generate control data as well as the suffusate and perfusate of studies of TMAs.
EC and VSMC Studies
Harvest and growth of ECs.
ECs were harvested from the mesenteric artery via collagenase digestion using the method of Kobayashi et al. (25). Cells were seeded onto gelatin coated six-well plates in media of the following composition: medium 199 with 5 mM L-glutamine, 20% fetal bovine serum, 5,000 U/ml heparin, 0.5% EC growth supplement, 105 U/ml penicillin, 10 mg/ml streptomycin, and 25 µg/ml amphoterecin. Confirmation of cell type was made by positive staining for von Willebrand factor (26). The principal contaminating cell type was vascular smooth muscle. This contamination was assessed by staining for smooth muscle 
-actin.
Experimental protocol.
Primary EC cultures at 90% confluence were placed into quiescent medium for 24 h before study (composition: DMEM with 2 µM Na pyruvate, 6 mM L-glutamine, 0.1% BSA, 105 U/ml penicillin, 10 mg/ml streptomycin, and 25 µg/ml amphoterecin). Thereafter, medium was aspirated and replaced with fresh medium containing study drugs or vehicle. If the study end point was EC release of NO or ET-1, medium was stored at –80°C and the cell count per well was determined. If the study end point required immunoblotting, cells were lysed [lysis buffer composition: HEPES buffer (pH 7.4), containing 50 mM -glycerol phosphate, 2 mM EGTA, 1 mM DTT, 10 mM NaF, 1 mM Na orthovanadate, 1% Triton X-100, 10% glycerol, 200 mM phenylmethylsulfonyl fluoride, 100 µM leupeptine, 10 nM okadaic acid, and 200 kallikrein inhibitory units of aprotinin], and the supernatant was recovered.
Generation of shear stress. Some cells were exposed to the mechanostimulus of shear stress (15 dyn/cm2 for 1 h). This perturbation was achieved using a cone-plate viscometer device that permits simultaneous application of radial shear to all wells of a six-well plate; the magnitude of shear stress generated by this instrument has been validated (60). A shear rate of 15 dyn/cm2 was selected to duplicate the wall shear stress generally noted within small arteries (60).
Harvest and growth of VSMCs.
VSMCs were harvested using the explant method and were grown in media consisting of 50% DMEM, 50% Ham's F-12 medium, 10% fetal bovine serum, 2 mM L-glutamine, 105 U/ml penicillin, 10 mg/ml streptomycin, and 25 µg/ml amphoterecin. After 10 days, VSMCs were separated from the remains of the minced aorta and replated in six-well plates (single passage). Confirmation of cell type was made by positive staining for smooth muscle -actin (23).
Experimental protocol. Single-passage VSMCs at 90% confluence were placed into serum-free medium for 48 h to generate a differentiated (spindle-shaped) phenotype. At the onset of the experiment, the quiescent medium was aspirated and replaced with fresh medium containing the pharmacological agents under study. At the completion of the protocol, the medium was discarded and cells were lysed using the buffer previously described, and the supernatant was used for immunoblotting.
Immunoblotting.
Protein separation was carried out by electrophoresis in 7% SDS-polyacrylaminde gels (10 µg protein/lane) and then transferred overnight to polyvinylidine difluoride membranes. After being blocked for 1 h in 5% dried milk in PBS-Tween buffer, the membranes were incubated with primary antibody for ETA or ETB (1:500, Alomone Lab, Jerusalem, Israel) or endothelial NOS (eNOS) or inducible NOS (iNOS; 1:1,000, BD Biosciences Pharmigen, San Jose, CA) diluted in PBS-Tween buffer for 1 h. Membranes were rinsed and incubated in horseradish peroxidase-conjugated secondary antibody (1:2,500; Calbiochem, San Diego, CA) and developed using enhanced chemiluminescene (Pierce, Rockford, IL). Membranes were then stripped and reprobed with primary antibody for -actin. Membranes were scanned, and quantity of the protein of interest was expressed as a ratio with the -actin signal.
Measurement of NO. NO is rapidly oxidized to NO2– and NO3–, hereafter designated NOx. NOx within conditioned medium was reduced to NO using 0.1 M vanadium chloride and 1.0 M HCl at 95°C. This NO was directly delivered to a Sievers NO analyzer (Sievers Instruments, Boulder, CO) by vacuum through a bubble trap collector containing 1 M NaOH. NO was reacted with analyzer-generated ozone in a chemiluminescent reaction and the product was measured. The resulting signal was fitted to a standard curve to calculate NO concentration, expressed in picograms per 105 cells (52).
Measurement of ET-1. Protein concentration of the supernatant was determined by the bicinchoninic acid method (Pierce). Peptide extraction was carried out on conditioned medium added to 1 ml of 1% trifluoroacetic acid applied to C-18 columns preequilibrated with 100% acetonitrile. After being washed three times with 1% trifluoroacetic acid, peptides were eluted from the column with 60% acetonitrile in 1% trifluoroacetic acid. Elutants were dried on a centrifugal concentrator without heat. ET-1 concentration was determined by ELISA (Peninsula Labs, San Carlos, CA).
In Vitro Studies of TMAs
Experimental apparatus. TMAs were mounted between glass cannulae set inside a Plexiglas chamber (CH/2/AS, Living Systems, Burlington, VT) that was continuously suffused with warmed (37°C) aerated Krebs buffer. Vessels were perfused with the same buffer. Pressure and flow within the vessel were established by two pressure-servo systems (PS/200/Q, Living Systems) placed at the inflow and outflow cannulae so that the axis of symmetry was set as the midpoint of the vessel. Inflow (Pi) and outflow (Po) pressures, adjusted separately, could be set to so that the vessel was pressurized in the absence of flow (i.e., Pi = Po) or so that flow moved through the vessel (i.e., Pi > Po).
Hemodynamic measurements. Pi and Po were measured with microtransducers at the inflow and outflow cannulas (PS/200/Q; Living Systems). Flow was measured with a microflowmeter (FM1; Living Systems). The chamber was mounted on the stage of an inverted microscope set in line with a video camera. Artery diameter was determined by a video dimension analyzer (V94; Living Systems).
TMA stabilization and viability determination. Pi and Po were initially set at 0 mmHg for 60 min after vessels were mounted. Pi and Po simultaneously increased to 25 mmHg and then to 45 mmHg. The second pressure was selected to duplicate pressures noted in situ (53). TMA diameter increased following both steps; however, in response to the second pressurization step, the vessel exhibited vasoconstriction after initial dilation, i.e., it displayed myogenic vasoconstriction (41). Phenylephrine (10–5 M) was added to the suffusion buffer to further contract the vessel, followed by acetylcholine (10–7 M) to confirm the presence of an intact endothelium. Vessels that failed to demonstrate a myogenic response, that failed to contract >40% to phenylephrine, or that failed to subsequently dilate >25% following acetylcholine were excluded from further study. The suffusion buffer was then changed, and the protocol was begun.
Experimental protocol.
Hemodynamic measurements were first made under zero-flow conditions with Pi and Po set at 45 mmHg. Pi was then raised, and Po was lowered by equal amounts as necessary to generate a flow rate of 100 µl/min across the vessel; the simultaneous changes in Pi and Po kept mean pressure within the vessel (Pav) at 45 mmHg [determined as (Pi + Po)/2]. The pressure gradient was then widened in a similar manner to raise the flow rate to 200 µl/min. This protocol was carried out under control conditions or following exposure of the vessel to a single pharmacological agent. Studies that involved IL-1 required a 4-h incubation with the cytokine. To account for this, all vessels were kept in buffer, with or without IL-1, for 4 h before study.
In Vivo Study of Ileal Gut Loops
Preparation of gut loops.
The piglet was anesthetized (telezol and xylazine) and ventilated. Catheters were placed into a femoral artery-vein pair for monitoring of blood gas tensions and systemic arterial pressure and for delivery of crystalloid (5% dextrose in 0.9% saline at 15 ml·kg–1·h–1). A segment of the ileum 15 cm long was isolated from the remainder of the gut so that it was perfused by a single artery-vein pair. After heparinization (500 U/kg), the vein was cannulated and the cannula was directed to a saline-filled reservoir; blood was returned to the animal via a femoral vein at a rate equal to mesenteric venous outflow. The artery leading to the gut loop as well as the mesenteric nerves were left intact so that the ileal loop was autoperfused and innervated. An electromagnetic flowmeter and pressure transducer were placed within the venous circuit to measure ileal blood flow and mesenteric venous pressure, respectively (46). The O2 concentrations of blood samples obtained from the artery-vein pair perfusing and draining the ileal gut loop were measured by a Lex-O2-Con (Chestnut Hill, MA). O2 consumption was calculated by the Fick equation, i.e., as the product of blood flow rate and the arterial-venous O2 concentration difference (a-vO2) across the gut loop. A cannula was inserted in a jejunal artery branch and advanced retrograde until its tip was immediately proximate to the main mesenteric artery trunk, just upstream from the ileal artery branch that served the isolated gut loop. Drugs were infused through this cannula. Ileal hemodynamics (pressure, flow, O2 consumption) generated under these conditions (i.e., in an anesthetized, ventilated, acutely prepared piglet) are similar to those noted in awake, spontaneously breathing, chronically cannulated piglets (47).
Experimental protocol
The ileal gut loop was allowed sufficient time (minimum 2 h) to attain steady-state conditions, defined as fluctuation of blood flow and a-vO2 of <5%. Baseline measurements were made; thereafter, a continuous infusion of IL-1 (or an equal volume of Krebs buffer) was begun into the mesenteric artery just upstream of the ileal artery serving the study gut loop. The drug infusion rate was set to generate an initial IL-1 concentration of 1 ng/ml arterial blood, based on the existing arterial flow rate at the outset of drug infusion. The infusion was continued for 1 h, and then measurements were made over the ensuing 4 h. In some studies, a single bolus infusion of BQ610 (50 nM/kg) or BQ788 (50 nM/kg) was given into the mesenteric artery just before the IL-1 infusion. These doses are sufficient to block ETA or ETB receptors, respectively, under these in vivo conditions (42).
Statistical Methods
ANOVA was used to determine the significance within each data set. One- or two-way ANOVA formats were used depending on the number of variables under consideration. If the F-statistic of the ANOVA was significant (P < .05), then post hoc Tukey's B tests were carried out to determine the sites of significance at the P < 0.05 level.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Effects on ECs and VSMCs
IL-1 increased ET-1 production by ECs in a dose-dependent manner over the range of 0.1–10 ng/ml, with the maximal effect noted at 1 ng/ml (Table 1). This effect was also time dependent over the range of 1–6 h, with a maximal effect noted at 4 h (data not shown). The mechanostimulus of wall shear stress has been demonstrated to enhance endothelial ET-1 production (18). To determine whether prior exposure to IL-1 altered the effects of shear on ET-1 production, ECs were pretreated with IL-1 and then exposed to shear stress. Under control conditions, shear stress (15 dyn/cm2 applied for 1 h) reduced EC release of ET-1 into the medium. IL-1 reversed this trend, i.e., IL-1-pretreated cells released significantly more ET-1 into the medium in response to shear; moreover, the amount released significantly exceeded that generated by IL-1 alone (Table 2).
|
|
increased EC expression of eNOS, and the maximal effect was noted at 1 ng/ml and a 4-h exposure (Fig. 1); in contrast, the cytokine had no effect on iNOS expression in these cells (data not shown). To determine whether IL-1affected eNOS function, ECs were pretreated with IL-1 and then which are exposed to acetylcholine or shear stress, established stimulants of eNOS activity (1, 6). Release of NOx into the medium was used as an indication of eNOS activity. Despite the increased expression of eNOS generated by IL-1, release of NOx in response to acetylcholine was unchanged from control; moreover, release of NOx in response to shear stress was significantly blunted by IL-1 (Table 3).
|
|
significantly reduced expression of ETB receptors on primary newborn ECs (Fig. 2). This receptor is primarily located on the endothelium in the newborn swine mesenteric artery (57), and ligand binding activates eNOS (40); accordingly, additional studies were carried out to determine whether IL-1 altered the effect of ET-1 on EC NO production. Pretreatment with IL-1 significantly attenuated the release of NOx into conditioned medium in response to ET-1 (Table 3).
|
on ETA receptor expression in VSMCs was determined. The ETA receptor mediates ET-induced vasoconstriction, and it is primarily localized to vascular smooth muscle in the swine intestine (40, 57). IL-1 significantly increased ETA expression in vascular smooth muscles in a dose-dependent manner (Fig. 3). The effect was also time dependent, with a maximal effect noted after a 4-h exposure. IL-1 had no effect on iNOS expression in these cells (data not shown).
|
on Flow-Induced Dilation in Vitro
Flow-induced dilation was clearly present in newborn TMAs. Vessel diameter at zero flow (Pi and Po at 45 mmHg) was 149 ± 7 µm (means ± SD, n = 8). Generation of a flow rate of 100 µl/min required a change in pressure (
P) of 12 ± 2 mmHg [a-v pressure (Pav) 46 ± 2 mmHg] and caused vessel diameter to increase to 217 ± 9 µm. Further widening the P to 15 ± 1 mmHg (Pav 45 ± 2 mmHg) caused diameter to increase to 249 ± 12 µm. Resistance across the vessel fell significantly (P < 0.05) from 0.11 ± 0.01 to 0.07 ± 0.02 mmHg·µl–1·min when flow was increased from 100 to 200 µl/min.
IL-1 reduced the TMA diameter noted under zero-flow conditions and also when flows of 100 and 200 µl/min were established (Fig. 4). The percent changes in TMA diameter noted in response to flows of 100 and 200 µl/min were significantly less that noted in control vessels; as well, IL-1 increased the P requisite to establish flow, increased the resistance noted at each flow rate, and eliminated the reduction in resistance noted in control vessels when flow was increased from 100 to 200 µl/min (Table 4). Interestingly, the hemodynamic changes induced by IL-1 were sustained; stated otherwise, the response to flow could not be restored by washout of IL-1. This sustained effect lasted for at least 6 h.
|
|
on ET-1 and ETA receptor expression in ECs and VSMCs, we next sought to determine whether the IL-1-induced attenuation of flow-induced dilation was mediated by ET-1. BQ610 (50 nM) had no effect on flow-induced dilation when given alone; however, when given before IL-1, it nearly eliminated IL-1-induced attenuation of the vascular response to flow (Fig. 5 and Table 5). BQ788 (50 nM) had no effect on flow-induced dilation when given alone, nor did it alter the effect of IL-1 on flow-induced dilation (Fig. 5 and Table 5).
|
|
on the TMA response to flow. ET-1 (1 nM) decreased TMA diameter under zero-flow conditions and shifted the diameter-flow curve downward but did not eliminate flow-induced dilation; thus the percent change in diameter from zero-flow baseline following application of flow was similar in ET-1-treated vessels compared with controls. Phenylephrine, which induces vasoconstriction via activation of 1-receptors, had an effect similar to ET-1 (Fig. 6 and Table 5).
|
on Intestinal Hemodynamics and Oxygenation In Vivo
Intramesenteric artery infusion of IL-1 for 1 h caused a subsequent progressive vasoconstriction within ileal gut loops and hence reduced perfusion; as well, the cytokine reduced the a-vO2 concentration gradient across the gut loop (Fig. 7). These changes conspired to reduce the calculated O2 consumption by this tissue. Administration of BQ610, either before or after 4 h of IL-1 infusion, significantly attenuated the IL-1-induced compromise in ileal perfusion and tissue oxygenation, whereas BQ788 had no effect.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
increased the release of ET-1 by ECs, upregulated ETA receptors on VSMCs, and downregulated EC ETB receptors; 2) IL-1 significantly attenuated the intrinsic vascular response of flow-induced dilation, due in part to ETA receptors; and 3) IL-1 compromised intestinal oxygenation in vivo, due in part to ETA receptors.
It is important to place these observations within the context of the IL-1 doses used in the study. Maximal effects in cell and TMA studies were noted at an applied IL-1 dose of 1 ng/ml (in medium or buffer), and this dose was used for intramesenteric artery infusion in the ileal gut loop studies. Tissue levels of IL-1 in healthy human colonic epithelium averages 40 pg/mg (21), whereas colonic explants from Crohn's disease patients demonstrate levels of 200 pg/mg (50). Cell or tissue levels of IL-1 were not measured in this study; hence, the levels achieved at the applied dose of 1 ng/ml are unknown. It remains to be determined whether the dose of 1 ng/ml generates tissue levels that are at pharmacological or pathophysiological (i.e., during acute inflammation such as occurs in NEC).
Effects of IL-1 at a Cellular Level
The maximal stimulatory effect of IL-1 on ET-1 release was observed at a cytokine concentration of 1 ng/ml, a dose lower than that reported for adult bovine (5) and porcine (29, 66) ECs and adult human vascular ECs (64, 65). Reasons for this age discrepancy cannot be deduced from the present data insofar as the mechanisms of ET-1 production, i.e., transcriptional and posttranslational phases of ET-1 production were not addressed in this study. ET-1 is initially transcribed as preproendothelin, which undergoes proteolytic cleavage, first to "big" ET-1 and then to the mature peptide, the latter cleavage via ET-converting enzyme (ECE) (49). In this context, Herman et al. (19) reported that IL-1 activated both the prepro-ET promoter and ECE activity in human umbilical vein (ECs).
The effects of shear stress and IL-1 on EC release of ET-1 into conditioned medium were disparate. By itself, the mechanostimulus of shear stress (15 dyn/cm2 applied for 1 h) reduced EC release of ET-1, a response inconsistent with most published data. These studies, carried out using aortic ECs harvested from adult animals, indicate that shear causes an early (2–4 h) increase in ET-1 production (18, 34, 67); however, this effect dissipates, so that after 6 h of sustained shear exposure, a reduction in ET-1 production occurs (31). At least two explanations for the difference in the present data and published reports exist: first, the response of newborn mesenteric artery ECs and adult aortic ECs may differ based on age- or region-specific differences in EC physiology. Second, the reduction in ET-1 production by shear may reflect the concomitant effect this stimulus had on NO production. Shear more than doubled NO release into conditioned medium, and Boulanger and Luscher (3) reported that NO reduces ET-1 production. IL-1 doubled ET-1 release by ECs under static conditions; subsequent exposure of these cells to shear caused an additional, significant rise in ET-1 release. Thus the effect of shear on EC ET-1 release was dependent on the presence of IL-1. In the absence of this cytokine, shear reduced ET-1 release; in its presence, shear increased ET-1 release. This observation is consistent with Bodin et al. (2).
IL-1 had a discrepant effect on ET receptors: downregulation of ETB on ECs vs. upregulation of ETA on VSMCs, and this difference may contribute to the effects of IL-1 on gut hemodynamics. Ligand binding to the VSMC ETA receptor increases intracellular Ca2+, induces cell contraction, and hence generates vasoconstriction of the vessel (49). In contrast, ligand binding to the EC ETB receptor activates eNOS to generate NO and hence induces vasodilation (58); moreover, NO directly interferes with ETA receptor function, thus further limiting ET-1-induced vasoconstriction (54). Normally, expression of endothelial ETB receptors exceeds expression of ETA receptors on smooth muscle in the intestinal circulation of 1-day-old swine (57), a circumstance that likely contributes to the relatively vasodilated state of this circulation (41). The physiological effects of permutation of ETB receptor expression by IL-1 was evidenced by the effects of the cytokine on endothelial NOx release. Thus a significant reduction in NOx release in response to exogenous ET-1 occurred, which likely reflected downregulation of endothelial ETB receptors. Release of NOx by ECs in response to shear stress was also compromised by IL-1, whereas the release of ET-1 was increased. Once again, these observations may reflect a relative loss of ETB receptors; hence, shear-induced release of ET-1 failed to generate a concomitant rise in NOx release insofar as ETB-induced activation of eNOS was blunted by the effect of IL-1 on ETB receptor expression.
Effects of IL-1 on Flow-Induced Dilation in TMAs In Vitro
IL-1 significantly impaired the normal hemodynamic response of TMAs to the mechanostimulus of flow, as evidenced by 1) lessened TMA diameter during flow; 2) exigency for a greater P to establish a specified flow rate; and 3) increased vascular resistance during flow. The cytokine effect was mediated, at least in part, by the changes in ETA receptor expression insofar as its selective blockade attenuated the effect; hence, BQ610 nearly restored the TMA response to flow to baseline. In contrast, the increased endothelial release of ET-1 noted following IL-1 was, by itself, insufficient to eliminate flow-induced dilation. Hence, exogenous ET-1, given at a concentration slightly greater that the Kd of the newborn swine ETA (0.36 ± 0.14 nM) and ETB (0.51 ± 0.24 nM) receptors, shifted the diameter-flow curve downward, whereas the percent changes in diameter and resistance in response to flow rates of 100 and 200 µl/min were unchanged from control (40). Phenylephrine, given at the concentration necessary to duplicate ET-1 vasoconstriction, had a similar effect. Vasoconstriction per se was not culpable. These observations are generally consistent with published reports on the effects of IL-1 on mesenteric vascular reactivity in adult rats (10, 63).
The present understanding of the basis for flow-induced dilation is that the phenomenon is mediated by several factor, including eNOS-derived NO (6, 51), oxidants, particularly H2O2 (27, 35), and K+ channels (36, 37). In this context, the effects of IL-1 were certainly more complex that singular adjustments in ET receptor expression. For example, IL-1 caused upregulation of endothelial eNOS expression, yet had no effect on NOx release under control conditions or following acetylcholine, and actually blunted NOx release in response to shear; stated otherwise, the functional activity of eNOS, as evidenced by NOx release, did not follow its expression. eNOS activity requires posttranslational modification, most importantly targeting of the enzyme to calveolae (55); as well, coupling of the reductase and oxidase domains of eNOS via tetrahydrobiopterin must be intact lest eNOS produces superoxide anion in lieu of NO (17). The increased eNOS expression in the absence of greater eNOS activity suggests compromise of eNOS function by IL-1 in TMAs. Existing reports have indicated that IL-1 enhances oxidant production (4, 33); moreover, ligand binding by the ETA receptor generates H2O2 (61). IL-1-induced oxidant production might be predicated to enhance flow-induced dilation, although oxidant stress can uncouple eNOS, reducing NO production (39). This putative interplay among eNOS, oxidants, and the ETA receptor is made more complex in the newborn intestine insofar as antioxidant systems are developmentally regulated, i.e., some systems are not fully functional at birth (8).
Effects of IL-1 on Intestinal Oxygenation In Vivo
IL-1 was infused directly into the mesenteric artery at a dose designed to approximate the in vitro levels requisite for alteration of cellular expression of the ET system and in vitro TMA hemodynamics. This infusion caused a 60 ± 5% reduction in ileal O2 consumption, as determined by the Fick equation. Reductions of both blood flow (31 ± 6%) and a-vO2 (50 ± 5%) contributed to the compromise of ileal oxygenation. The latter effect likely reflects reduction of the perfused capillary density secondary to ETA-induced closure of the precapillary sphincter as under normal conditions an inverse relationship between levels of flow and a-vO2 exist so that tissue oxygenation is controlled (16). ETA ligand binding causes vasoconstriction and reduces capillary surface area in vivo (42). On the basis of in vitro findings, interpretation of the in vivo data suggests that IL-1 enhanced ETA receptor expression and local (intestinal vascular) ET-1 production; indeed, this explanation is strengthened by the partial reversal of cytokine-induced hemodynamic effects by selective blockade with BQ610.
BQ788, a selective ETB receptor antagonist, did not alter IL-1-induced changes in ileal hemodynamics or oxygenation in the in vivo model nor did it affect flow-induced dilation in the in vitro TMA studies. At a cellular level, IL-1 significantly reduced EC ETB expression. Ligand binding by the ETB receptor activates eNOS resulting in generation of NO, a potent vasodilator (9), and, under normal circumstances, blockade of ETB receptors in the newborn intestine elicits vasoconstriction (42). If IL-1-induced downregulation of EC ETB receptors in TMAs and ileal gut loops, then blockade of the receptors might not have any significantly appreciable effect on hemodynamics.
A weakness of the in vivo portion of this study is the means used to augment IL-1 within the intestine, i.e., intramesenteric artery infusion. The bulk of IL-1 generated in vivo is derived from circulating or tissue macrophages (11). Within the intestine, IL-1 production has been localized to the mucosa (14), specifically to monocytes and tissue macrophages within the lamina propria (30, 32). Thus it is likely that the IL-1 relevant to intestinal vascular pathophysiology ascends from the mucosa and exerts its primary effect within the submucosal arteriolar circulation. Moreover, these vascular effects would likely be intertwined with those generated by other cytokines and vasoactive agents (e.g., platelet-activating factor) generated during intestinal inflammation.
Putative Relevance of These Observations to the Pathogenesis of NEC
The isolation of both cytokines (12, 13, 59) and ET-1 (45) from the human intestine resected for NEC at significantly elevated levels implies, but does not confirm, their potential role in disease pathogenesis. Data presented herein support a link between proinflammatory stimuli (IL-1) and compromise of intestinal hemodynamics via change in the ET system that favors vasoconstriction. This effect on the ET system could extend beyond generation of vasoconstriction and tissue hypoxia; thus ligand binding by ETA receptors generates an inflammatory response (22), increases oxidant production (62), and activates the phosphoinositide cycle, all processes that can generate cell injury. Of course, these data must be viewed within the caveat that observations made in newborn swine, or any newborn animal model, can only be transcribed to the human experience with caution.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
, Control; 
, IL-1
, IL-1
, IL-1
P < 0.05 for 100 vs. 0 µl/min; 
P < 0.05 for 200 vs. 100 µl/min.
