|
- 积分
- 0
- 威望
- 0
- 包包
- 0
|
作者：Timothy A. Sutton, Henry E. Mang, Silvia B. Campos, Ruben M. Sandoval, Mervin C. Yoder, Bruce A. Molitoris,作者单位：1 Division of Nephrology, Department of Medicine,Indiana Center for Biological Microscopy, and Division of Neonatology, Department of Pediatrics,Indiana University School of Medicine, and RoudebushVeterans Affairs Medical Center, Indianapolis, Indiana 46202 + W, A, F+ m$ M8 L8 G' p. r8 k
' `3 R* A: r1 k6 q: A
0 E# w& q* O4 H" k2 r& s
1 `4 j4 i9 ]) |' }- s* P
3 B3 B+ z9 [) @
, q) e/ x9 G4 y( E
! v7 n" I1 U% Y6 |/ }' N& C# v $ T: r, d8 l- a( `' W& _
8 m$ R8 x1 Y$ c* m
8 Z6 E. z' J' a c5 B( [ & |/ t7 x: \+ d. E, j, a
! z- O i4 l# M4 Y; U
1 z% z$ D. ?/ ^: d4 T' P3 {; {
【摘要】
; w3 {/ m" ?) o! R) b& j* s4 d, | The role of renal microvascular endothelial cell injury in the pathophysiology of ischemic acute renal failure (ARF) remains largely unknown.No consistent morphological alterations have been ascribed to the endotheliumof the renal microvasculature as a result of ischemia-reperfusion injury.Therefore, the purpose of this study was to examine biochemical markers of endothelial injury and morphological changes in the renal microvascular endothelium in a rodent model of ischemic ARF. Circulating von Willebrandfactor (vWF) was measured as a marker of endothelial injury. Twenty-four hoursafter ischemia, circulating vWF peaked at 124% over baseline values( P = 0.001). The FVB-TIE2/GFP mouse was utilized to localizemorphological changes in the renal microvascular endothelium. Immediatelyafter ischemia, there was a marked increase in F-actin aggregates in the basal and basolateral aspect of renal microvascular endothelial cells in thecorticomedullary junction. After 24 h of reperfusion, the pattern of F-actinstaining was more similar to that observed under physiological conditions. Inaddition, alterations in the integrity of the adherens junctions of the renalmicrovasculature, as demonstrated by loss of localization in vascularendothelial cadherin immunostaining, were observed after 24 h of reperfusion. This observation temporally correlated with the greatest extent ofpermeability defect in the renal microvasculature as identified usingfluorescent dextrans and two-photon intravital imaging. Taken together, thesefindings indicate that renal vascular endothelial injury occurs in ischemicARF and may play an important role in the pathophysiology of ischemic ARF. $ w1 t$ q1 O& E; Z8 k1 {; y
【关键词】 von Willebrand factor vascular endothelial cadherin vascular permeability' B. `* @7 L' y& V- {& f0 d8 x* ^
BOTH SUBLETHAL AND LETHAL tubular epithelial cell injury have been of central importance in explaining the decrement in glomerular filtration rate that is the hallmark of acute renal failure (ARF). However,over the last decade the paradigm of the pathophysiology of ischemic ARF hasevolved to include a complex interplay between tubular injury, inflammation,and altered renal hemodynamics. Recent studies have provided further evidencefor the role that vascular injury, in particular endothelial cell injury, plays in the pathophysiology of ischemic ARF( 5, 46 ). Endothelial cellswelling, altered endothelial cell-cell attachment, and altered endothelialcell-basement membrane attachment are some of the morphologic alterations thathave been observed in the renal microvasculature( 5, 12 ) as well as the cerebraland coronary vasculature ( 40, 42 ) after ischemic injury.Functional consequences of these morphological alterations include altered vascular reactivity, increased leukocyte adherence and extravasation, alteredcoagulation due to loss of normal endothelial function and/or barrier, andincreased interstitial edema that have been documented as a consequence ofischemic ARF in animal models( 16 ).% t; l9 a; L9 v' D
( V$ r' P+ \" ~6 {. U; X
Specialized cellular junctions similar to those in epithelial cellsmaintain endothelial cell-cell contacts. Tight junctions are more prominent in"tight" vascular beds, such as those between the endothelial cellsof the cerebral vasculature forming the blood-brain barrier, whereas they aresparse and simplified in "leaky" vascular beds, such aspostcapillary venules ( 43 ). Cadherin-containing adherens junctions are ubiquitous between endothelialcells throughout the vasculature( 43 ). Recent studieshighlighting the differences in the molecular composition of junctionalcomplexes in various vascular beds, including those within the kidney, provideinsight into the potential functional differences in the cellular junctions inthese vascular beds ( 3, 11, 25, 32 ). Disruption of endothelialadherens junctions in vivo by the use of an inhibitory antibody to vascular endothelial cadherin (VE-cadherin) has been demonstrated to induce gapsbetween endothelial cells, increase endothelial permeability, and promote theaccumulation of inflammatory cells in coronary and pulmonary vascular beds( 8 ). Furthermore, there isevidence from in vitro studies that the interaction of endothelial cell-celljunctions with the actin cytoskeleton plays an important role in regulatingendothelial paracellular transport( 34 ). Although the abovefindings underscore the importance of endothelial cell-cell junctions inmaintaining the integrity of the endothelial permeability barrier, there isessentially no in vivo information on the effect of ischemic injury on thefunction and organization of these intercellular junctions in the renalmicrovasculature due to previous technical difficulty in visualizing theendothelium of the renal microvasculature in animals and human biopsy samples.However, there are data indicating microvascular congestion and localizedinterstitial edema after renal ischemia( 15, 16 ). Therefore, wehypothesized that ischemic injury to the kidney results in alterations of theendothelial actin cytoskeleton and endothelial cell-cell junctions thatcontribute to increased vascular permeability and local interstitial edema. Inthis study, we demonstrate that ischemic renal injury results indisorganization of the actin cytoskeleton and loss of VE-cadherin localizationin the endothelium of the renal microvasculature and that these alterationsare accompanied temporally by increased permeability of the renalmicrovasculature.- ~. r: l$ V; [ ^! Y! b" h3 k2 J
* U6 b4 e. X8 ?2 [2 R; P h3 C% M8 ~MATERIALS AND METHODS
( ]4 n5 Z* ^. l* E# w0 x1 u- Q' _( t( t0 h5 s" {
Reagents and antibodies. Rhodamine-conjugated dextran (3,000 molwt), FITC-conjugated dextran (500,000 mol wt),rhodamine-conjugated-phalloidin, and Hoechst 33342 were from Molecular Probes(Eugene, OR). Rat polyclonal antibodies to mouse VE-cadherin were from BD Pharmingen (San Diego, CA). Affinity-purified Texas red-labeled sheep anti-ratIgG antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA).An Asserachrom vWF ELISA kit was from Diagnostica Stago (Parsippany, NJ), anda Quantikine M rat IL-6 immunoassay kit was from R&D Systems (Minneapolis, MN).
* I, R8 M9 M4 R2 R
/ A; ?/ k; Z1 o# T! |Animals. Male Sprague-Dawley rats (Harlan, Indianapolis, IN) andmale FVB-TIE2/GFP mice, which express green fluorescent protein (GFP) underthe direction of the endothelial-specific receptor tyrosine kinase( Tie2 ) promoter ( 33 ),were used as described below.: v6 v% `& Y* _/ v% f9 D5 ^
5 x: M3 K7 L9 w4 L: h
Surgical procedure to induce renal ischemia. All experiments wereconducted in accordance with the Guide for the Care and Use of LaboratoryAnimals (Washington, DC: National Academy Press, 1996) and approved bythe Institutional Animal Care and Use Committee. Male Sprague-Dawley ratsweighing 200-250 g were anesthetized with an intraperitoneal injectionof pentobarbital sodium (65 mg/kg) and placed on a homeothermic table tomaintain core body temperature at 37°C. For experiments involving measurement of biochemical markers of endothelial injury, a midline incisionwas made, the renal pedicles were isolated, and bilateral renal ischemia wasinduced by clamping the renal pedicles for 45 min as previously described( 31 ). Sham surgery consistedof an identical procedure with the exception of immediate release of themicroaneurysm clamps. Tail-vein blood samples from experimental, sham, andnonoperative control rats were collected, stored, and processed as perinstructions of the Asserachrom vWF ELISA kit. For experiments involving livetwo-photon microscopic imaging of rat kidneys, a flank incision was made overthe left kidney, the renal pedicle was isolated, and unilateral renal ischemiawas induced by clamping the left renal pedicle for 45 min as previouslydescribed ( 9 ). For experimentsinvolving renal ischemia in FVB-TIE2/GFP mice, 20- to 25-g mice were anesthetized utilizing 5% halothane for induction and 1.5% for maintenance andplaced on a homeothermic table to maintain core body temperature at 37°C.A midline incision was made, and the left renal pedicle was isolated. Renalischemia was induced by clamping the renal pedicle for 32 min as described byKelly et al. ( 20 ).
a' z3 v9 e$ c) h, r( s/ n8 I3 Y3 u, I7 C3 m
Fluorescence microscopy. Kidney sections from anesthetized mice were fixed ex vivo in 4% paraformaldehyde, and 50-µm vibratome sectionswere obtained. Sections were stained with rhodamine-phalloidin or ratpolyclonal antimouse VE-cadherin and polyclonal Texas red-labeled sheepanti-rat IgG secondary antibodies. Images of the microvasculature in thecorticomedullary region of the kidney were collected with an LSM-510 Zeissconfocal microscope (Heidelberg, Germany) equipped with argon and helium/neonlasers. For intravital fluorescence microscopy, 100 µl ofrhodamine-conjugated dextran (3,000 mol wt, 20 mg/ml in 0.9% saline), 500µl of FITC-conjugated dextran (500,000 mol wt, 7.5 mg/ml in 0.9% saline),and 400 µl of Hoechst 33342 (1.5 mg/ml in 0.9% saline) were injected viathe tail vein into anesthetized rats just before imaging. The left kidney ofthe anesthetized rat was imaged through a retroperitoneal window via aleft-flank incision using a Bio-Rad MRC-1024MP Laser ScanningConfocal/Multiphoton scanner (Hercules, CA) with an excitation wavelength of800 nm attached to a Nikon Diaphot inverted microscope (Fryer, Huntley, IL) asdescribed by Dunn et al. ( 9 ).Image processing was performed utilizing Metamorph software (UniversalImaging, West Chester, PA).: y7 _6 p' y. P: b' p0 _
# H& r9 l P9 mStatistics. Results are expressed as means ± SE and were analyzed for significance by paired and unpaired Student's t -tests and ANOVA.( M, X6 Q- r2 g1 ~: `" e: R
/ F: t, W6 Z; `/ Z: T" Z4 G) ?9 jRESULTS2 @. ~3 t0 A! B& N* f
1 i9 D6 \6 k5 ]
Biochemical evidence of renal endothelial injury during ischemia. A variety of studies have utilized an elevated circulating von Willebrandfactor Ag (vWF) concentration as a marker of endothelial cell injury in otherorgans in response to a variety of insults ( 1, 6, 13, 19, 28, 35, 37 ) includingischemia-reperfusion injury to the intestines( 1 ). Therefore, we quantifiedcirculating vWF in Sprageu-Dawley rats as a marker of renal endothelial injuryafter renal ischemia. Twenty-four hours after a 45-min bilateral renal arteryclamp, circulating vWF reached its maximum level and was significantlyelevated ( Table 1 ) overbaseline preclamp values ( P = 0.001) and over levels in sham-operated control animals ( P = 0.029). Circulating vWF levels in sham-operated animals were not significantly different from baseline values at 24 h.Circulating vWF decreased by 48 h after ischemia and was not significantlyelevated over baseline values or over the level in sham-operated controlanimals.
, d/ v; a7 s/ Z# s+ S) G4 H+ x' J: n
3 a( w5 _8 r8 a5 G. BTable 1. Serum von Willebrand factor increases after renal ischemia
7 `: G3 b! j$ r$ R
& o) ~6 L! O4 T, HThe actin cytoskeleton of renal microvascular endothelium was disruptedafter ischemia. To determine the effect of ischemia on the actincytoskeleton of the renal microvascular endothelium, sections of normal andischemic kidneys from FVB-TIE2/GFP mice were stained with rhodamine-phalloidinand examined by confocal microscopy. Under physiological conditions, F-actinstaining was predominantly observed along the basal aspect of renal microvascular endothelial cells ( Fig.1 A ). Immediately after 32 min of renal ischemia,alterations in the actin cytoskeleton of the renal microvascular endotheliumincluded F-actin aggregation along the lateral aspects of the cells andincreased F-actin aggregation along the basal aspects of the cells( Fig. 1 B ). After 24 hof reperfusion subsequent to ischemia, the F-actin staining pattern moreclosely resembled the pattern observed under physiological conditions( Fig. 1 C ), althoughthere remained a more aggregated appearance of the F-actin along the basal aspects of the renal microvascular endothelial cells than what was observedunder physiological conditions.
; l# J( X: u6 G( {
! P" P* x: q- @9 J) E4 JFig. 1. The actin cytoskeleton of renal endothelial cells is disrupted after renalischemia. The FVB-TIE2/GFP mouse, which expresses green fluorescent protein(GFP) driven by the endothelial TIE2 promoter, was utilized to localizechanges in the actin cytoskeleton of the microvascular endothelium (green) inthe corticomedullary region of the kidney. Rhodamine-phalloidin was used tostain F-actin (red). A : nonischemic kidney. B : kidney after32 min of renal artery clamping. Note the increase in F-actin polymerizationand aggregation in endothelial cells of the renal microvasculature (arrow). C : kidney after 24 h of reperfusion following 32 min of renal arteryclamping. Bar = 10 µm.
) Q& s( d: Z, G
+ H) y2 g9 l* I( WAdherens junctions of renal microvascular endothelium were disrupted after ischemia. To determine the effect of ischemia on the adherens junctions of the renal microvascular endothelium, the pattern of VE-cadherinimmunostaining was examined by confocal microscopy in sections of normal andischemic kidneys from FVB-TIE2/GFP mice. Under physiological conditions,VE-cadherin immunostaining was continuous along the renal microvascularendothelium ( Fig. 2 ). WhereasVE-cadherin immunostaining was not limited to intracellular endothelialcontacts, it was similar to in vivo VE-cadherin immunostaining observed inother studies ( 4, 8 ). After 32 min of ischemia,VE-cadherin staining remained in the renal microvasculature. Twenty-four hoursafter ischemia, the majority of the renal microvasculature did not stain forVE-cadherin. The loss of VE-cadherin staining after ischemia suggested a disruption of the normal junctional complex between endothelial cells of therenal microvasculature. Seventy-two hours after ischemia, VE-cadherin stainingin the renal microvasculature was similar to that observed under physiologicalconditions.
2 v$ X3 P* G. p% I9 a1 i% D$ T/ [7 b; r1 q4 V; F" v: c
Fig. 2. Adherens junctions of renal microvascular endothelium were disrupted afterischemia. The FVB-TIE2/GFP mouse was utilized to localize changes in vascularendothelial (VE)-cadherin (red) staining of the microvascular endothelium(green) in the corticomedullary region of the kidney after ischemia. WhileVE-cadherin staining was present under physiological conditions (arrow in A ) and immediately after 32 min of ischemia ( B ), loss ofVE-cadherin staining was noted after 24 h of reperfusion following 32 min ofischemia ( C ). Of note, immunostaining was observed in the lumens (*)of tubules at 24 h postischemia; however, this pattern of staining was seen inthe secondary antibody controls (data not shown), suggesting that the stainingin the tubular lumen was nonspecific secondary antibody staining and notVE-cadherin staining. Seventy-two hours after 32 min of ischemia, VE-cadherinstaining was similar to that under physiological conditions. Bar = 10µm.
5 O# `* z7 W5 Y% F$ g- H2 u+ E+ ^' P
$ D# F+ A, r* ]$ Q6 z3 k& ?3 I rAlteration in renal microvascular permeability after ischemia. Having demonstrated morphological alterations in the actin cytoskeleton andthe adherens junctions of the renal microvascular endothelium after ischemia,we undertook the next study to examine the functional correlate of theobserved morphological alterations. Technical considerations limited theeffectiveness of two-photon intravital imaging of the corticomedullary regionin rodents, as well as intravital imaging of kidneys in FVB-TIE2/GFP mice. Consequently, intravital imaging was performed in the cortical region of rats.To examine alterations in vascular permeability, a rhodamine-conjugated,low-molecular-weight (3,000) neutral dextran and a FITC-conjugated,high-molecular-weight (500,000) neutral dextran were intravenously coinjectedinto rats under physiological conditions and after ischemia. Injections of fluorescent dextrans were undertaken at selected time points to evaluate themicrovascular permeability at that particular point in time. The renalmicrocirculation was observed in vivo utilizing dual-photon confocalmicroscopy. Injection of the fluorescent-labeled dextrans under physiologicalconditions revealed that the high-molecular-weight dextran was maintained inthe vascular space and was not filtered by the glomerulus, whereas, thelow-molecular-weight dextran was rapidly filtered into the tubules,accumulated in endosomes of the proximal tubule, and concentrated in the lumenof the distal tubules as previously described( 9 )( Fig. 3 A ). Areas ofdiminished microvascular blood flow were observed immediately after renal ischemia; however, occasional areas of leakage of either the low- orhigh-molecular-weight dextran from the microvascular space into the renalinterstitium were not observed until 2 h after renal ischemia (arrows, Fig. 3 B ). Leakage ofboth dextrans appeared to reach its greatest extent 24 h after ischemia ( Fig. 3 C ). Forty-eighthours after ischemia, there were still some patchy areas of leakage but theextent of the permeability defect appeared to have improved significantly( Fig. 3 D ). Glomerular filtration of the high-molecular-weight dextran into the tubular lumen was notobserved after ischemia at any of the time points studied. As might beanticipated, leakage of the smaller dextran into the interstitium was morediffuse than that of the larger dextran( Fig. 4 ). Interestingly, areaswhere leakage of both dextrans occurred were more often observed in areas ofmarkedly diminished microvascular flow (a supplementary video of Fig. 4 C can be viewedat http://ajprenal.physiology.org/cgi/content/full/00042.2003/DC1 ).
, y3 p0 K6 G6 R1 ]+ N
! z$ X& b3 T. n/ G2 R- UFig. 3. Ischemia results in altered renal vascular permeability. A 500,000-mol wt,FITC-labeled dextran (green), 3,000-mol wt, rhodamine-labeled dextran (red),and the nuclear stain Hoechst 33342 (blue) were injected via the tail veininto sham-operated, nonischemic rats ( A ) and into rats after 2( B ), 24 ( C ), and 48 ( D ) h of reperfusion following45 min of ischemia. Intravital images of the kidneys were obtained via aretroperitoneal approach. The large FITC-labeled dextran was generallyretained in the vascular space and identifies the microvasculature. Signalvoids in the vascular space are due to the relative rapid movement of cellularelements (red and white blood cells) compared with the speed of imageacquisition. Proximal tubules are identified by the red punctate appearancefrom the internalization of the small rhodamine-labeled dextran (*). Distaltubules are identified by the luminal concentration of the smallrhodamine-labeled dextran ( ). Note the close approximation of thetubules and microvasculature in the sham-operated control rat ( A ).Scattered areas of extravasation of both the small and large dextrans (arrows)from the vascular space into the interstitium were noted as early as 2 h afterreperfusion ( B ). The extent of extravasation peaked at 24 h ofreperfusion ( C ) and returned to close to normal after 48 h ofreperfusion ( D ). Note the areas of diminished blood flow (largearrowheads in B and C ) identified by intravascular cellularelements that appear stationary during the time frame of image acquisition.Bar = 10 µm.
+ J0 B, ^$ F! F/ U) H" \- X5 \8 F9 x5 T6 f! M7 K6 b4 {5 q
Fig. 4. Extent of microvascular permeability defect is heterogeneous. A 3,000-molwt, rhodamine-labeled dextran ( A; red in C ) and a500,000-mol wt, FITC-labeled dextran ( B; green in C ) wereinjected via the tail vein into a rat after 24 h of reperfusion following 45min of ischemia. A - C are images obtained from the samefield. A : demonstration of the extent of microvascular extravasationof the low-molecular-weight dextran. B : demonstration of the extentof microvascular extravasation of the high-molecular-weight dextran. C : color overlay of A and B, with thelow-molecular-weight dextran in red, the high-molecular-weight dextran ingreen, and the arrowhead indicating an area of diminished microvascular flow.Bar = 10 µm. A supplementary video of C is a time series takenfrom the same field as shown and is composed of a 30-frame series of imagescollected at 2 frames/s, projected at 10 frames/s, demonstrating thediminished flow in the area of the greatest permeability defect.. `3 _& f) y. t; c# o: ?
$ X+ p: H/ R$ Y: W+ d: i6 U. Y3 pDISCUSSION
% q( k% P- X7 i% N; C8 p4 ?2 p% [( Q' ^% [2 W
Endothelial dysfunction and vasomotor alterations of the renal vasculaturehave played a conceptual role in the pathophysiology of ARF for a number ofyears ( 12 ). However, recentwork by Brodsky et al. ( 5 ), inparticular, has brought to light the vulnerability of the renal microvascularendothelium to ischemic injury and has highlighted the renal vascularendothelium as a potential target for injury in renal ischemia. To further anunderstanding of the role endothelial cell injury plays in ischemic ARF, wehave utilized a renal artery clamp model of ischemic ARF in rats and mice.While this model is imperfect, it has provided valuable insight into thepathophysiology of ischemic injury to the kidney( 27 ). Our study has served to extend confirmed previous findings and to begin to examine some of themechanisms by which renal microvascular endothelial injury may contribute toaltered function in one of the myriad roles to which the endothelium issubservient.
( H* R4 R! i2 L2 q! H* }5 C# p. s& O( |6 H; S. m% C. `
Our finding of an elevation in circulating vWF after renal ischemic injurycontributes to the evidence that the renal vascular endothelium is injured. Inother organ systems, release of vWF from the endothelium has been demonstratedto occur in a biphasic fashion in response to ischemia-reperfusion injury ( 10, 39 ). These studies suggestthat hypoxia, as well as reperfusion, mediate endothelial release of vWF viapotentially different mechanisms. Furthermore, inflammatory processes cancontribute to endothelial vWF release( 41 ). The peak elevation ofcirculating vWF in our study was monophasic and occurred 24 h after the initial ischemic insult; thus continued hypoxia, reperfusion injury, and/orinflammatory processes could separately or collectively play a role inendothelial release of vWF in the renal artery-clamp model of ischemic ARF.Whereas the peak increase in vWF occurred at the same time as the peak inserum creatinine in our study, given the large molecular weight of the variouscirculating forms of vWF (500-20,000), it is doubtful that reduced renalclearance of vWF significantly contributed to the elevation in circulatingvWF. While the intent of this portion of the investigation was to simplyprovide evidence for endothelial injury, the utility of such a finding mayultimately lie in the ability to characterize the nature and extent of theunderlying injury by circulating markers of endothelial injury and thus provide a useful clinical correlate for disease severity, prognosis, andtherapeutic intervention. For example, endothelial release of IL-6 has beendemonstrated to be of prognostic value in sepsis where endothelial injuryplays an important role in the underlying pathophysiology( 14 ). Although a previousstudy of intestinal ischemia-reperfusion injury did not demonstrate acorrelation between histological injury scores and the level of circulatingvWF ( 1 ), the pattern ofcirculating vWF multimers or the pattern of vWF coupled with other markers ofinjury may in the end prove to be a useful clinical tool in renal ischemia-reperfusion injury.0 a1 i- ?1 s6 R$ C3 H$ u, o: ~
. |2 ~- f, {5 H3 y) K6 n
Our examination of the morphological alterations in renal microvascular endothelial cells revealed that the normal structure of the actin cytoskeletonin renal microvascular endothelial cells is disrupted after ischemia.Alteration of the normal actin cytoskeleton of endothelial cells in vitro hasbeen demonstrated with ATP depletion as a model of ischemic injury and withH 2 O 2 as a model of oxidant-mediated reperfusion injury.ATP depletion has been demonstrated to rapidly and reversibly disrupt the normal cortical and basal F-actin structures in endothelial cells( 17, 24, 45 ), resulting in F-actinaggregation and polymerization. Oxidant-mediated endothelial cell injury also has been demonstrated to disrupt the cortical actin band in culturedendothelial cells ( 18, 29, 47 ). We observed that disruption of the actin cytoskeleton in endothelial cells of the renalmicrovasculature subjected to ischemic injury in vivo was most prominentimmediately after the ischemic insult. Consistent with the above-mentioned invitro findings, we observed an alteration of the normal cortical and basalF-actin structures of the endothelial cell with an apparent increase inF-actin polymerization and aggregation at the basal and basolateral aspects ofendothelial cells after ischemia. This F-actin polymerization and aggregationwas reminiscent of the alterations observed after ischemia in proximal tubularepithelial cells ( 30 ) and renal vascular smooth muscle cells( 26 ). Of note, the TIE2 promoter has been reported to be upregulated in vitro by hypoxia in humanumbilical vascular endothelial cells( 7 ). In our in vivo study, wedid not observe a difference in the number of vessels or a change in theintensity of GFP signal at the selected time points.! [6 o5 `- Y( k+ {
* A! ]8 `2 {) r. n" K
The alterations in the actin cytoskeleton of renal microvascular endothelial cells preceded alterations in VE-cadherin staining at endothelialcell-cell junctions. These findings are consistent with a mechanism by whichloss of integrity of the actin cytoskeleton contributes to breakdown of theactin-associated adherens junctions and contributes to the concomitantpermeability defect. Much of the present knowledge regarding the mechanismsregulating endothelial cell-cell interaction during ischemic and oxidant injury has come from in vitro models utilizing cultured endothelial cells. ATPdepletion of endothelial cell monolayers and exposure of endothelialmonolayers to oxidants such H 2 O 2 have both been demonstrated to increase endothelial permeability and intercellular gapformation ( 22, 36, 38 ). Increased endothelialpermeability in these models has been associated with internalization of VE-cadherin from endothelial adherens junctions( 2, 21 ). Although we did notobserve internalization of VE-cadherin after ischemia in vivo, othermechanisms including cleavage of VE-cadherin as described in a previous invivo study ( 4 ) may account for the diminished VE-cadherin staining observed in our study.- G- E4 W+ Q& O1 {. M- o; Q
, Z8 a0 V0 {8 J& o: b2 Q
The utilization of two-photon intravital microscopy to visualize thepermeability defect in the renal microvasculature after ischemia is aparticular strength of our study. The power of this imaging technique isdemonstrated in our ability not only to simultaneously evaluate the disparityin the permeability defect of two differently sized fluorescent probes butalso to observe a correlation, in a timed-image series, between alterations inblood flow and severity of the permeability defect. These data imply continuedreduced blood flow results in increased permeability defects betweenendothelial cells. Although these observations were limited to the corticalarea of rats due to technical considerations, presumably the permeabilitydefect in the corticomedullary area would be even more pronounced than what weobserved in the cortical microvasculature( 16 ). Implications for leakageof plasma from the vascular space and increased interstitial edema, especiallyin the corticomedullary area, include further diminishment of the compromisedmedullary blood flow by extrinsic compression of peritubular capillaries( 23 ) and by hemoconcentrationas observed in other organs( 44 ).. z; w! v- \/ S! ]+ H$ ^
u4 u$ I# @: P N2 f" J+ i
In summary, we have demonstrated that renal microvascular injury manifestedby disruption of the actin cytoskeleton and adherens junction of endothelialcells occurs after ischemic injury and that this injury is probablyfundamental to increased microvascular permeability and renal interstitialedema. Further characterization of the implications of renal microvascularinjury may provide new diagnostic and therapeutic avenues in ischemic ARF.9 d: i; f, v7 [& B
4 |: u( O6 {% k* L% RDISCLOSURES$ X( l2 G* i& m: _9 B; I
. @9 K/ C0 D2 s; pThis work was supported by National Institutes of Health (NIH) GrantDK-60621-61594, the Ralph W. and Grace M. Showalter Research Trust, andNational Kidney Foundation of Indiana grants (to T. A. Sutton), NIH GrantHL-63169 (to M. C. Yoder), and NIH Grants DK-41126, DK-53465, and DK-61594,and Veterans Affairs Medical Research Service grants (to B. A. Molitoris).4 ^1 f" `. o$ t4 f; j0 T
5 G$ o9 {# o, F, ?# ~ACKNOWLEDGMENTS
# B, Q) Z& H* _
2 n3 l2 v( n PWe acknowledge Simon Atkinson, Ken Dunn, and Katherine Kelly for valuablediscussions." \2 F' H& K) ~. w7 Q
【参考文献】; c! }: [+ y9 `# s( u& E
Abu-Zidan FM,Winterbourn CC, Bonham MJ, Simovic MO, Buss H, and Windsor JA. Small bowelischaemia-reperfusion increases plasma concentrations of oxidised proteins inrats. Eur J Surg 165:383-389, 1999.
7 g( x+ Q2 \( H' U; x( n9 q T! r/ H9 i. z
( S% Z% q7 M. [9 G, I& g
* p) E7 X/ `; S: m4 y9 K3 N: }7 A
Alexander JS,Jackson SA, Chaney E, Kevil CG, and Haselton FR. The role of cadherinendocytosis in endothelial barrier regulation: involvement of protein kinase Cand actin-cadherin interactions. Inflammation 22: 419-433,1998.; u: ?! Q$ j! T5 t
; Y- {( ^' v' b5 @& j$ c; z$ r# V: g/ w( V" d! X
; C! T9 Q2 k4 o1 |* ]+ eAurrand-Lions M, Johnson-Leger C, Wong C, Du Pasquier L, and ImhofBA. Heterogeneity of endothelial junctions is reflected by differentialexpression and specific subcellular localization of the three JAM familymembers. Blood 98:3699-3707, 2001.' F0 j' [/ \' a' V% ^1 t
+ @1 `- F9 x' v/ M# c4 u5 ~% ?$ ]" e8 m. `* g9 @! B( J
% a+ ]3 t+ z3 W6 ?% `
Bianchi C,Araujo EG, Sato K, and Sellke FW. Biochemical and structural evidence forpig myocardium adherens junction disruption by cardiopulmonary bypass. Circulation 104:I319-324, 2001.( B- y- O# U3 H2 T& q; j
) G. c0 m0 P8 W' }9 ]. d) k/ ?8 j5 Q4 d; J% C( D) X( Z( ]' a0 L
' ?$ [3 q7 V! t8 ~Brodsky SV,Yamamoto T, Tada T, Kim B, Chen J, Kajiya F, and Goligorsky MS. Endothelial dysfunction in ischemic acute renal failure: rescue bytransplanted endothelial cells. Am J Physiol RenalPhysiol 282:F1140-F1149, 2002.) H7 f" W9 B( W' @+ S+ T8 v
; b1 m- V$ u2 X2 D, r* d
/ d3 l8 a. x- {, U
4 T% u; M F6 K, d6 _Chaloner C,Blann A, and Braganza JM. Endothelial cell harbingers of adult respiratorydistress syndrome in acute pancreatitis (Abstract). Lancet 344:1167, 1994.7 _2 A& H* u0 L+ @& q
) g( N [2 [& q9 F, M+ S+ P) @; t
- M* Q8 |; b* a& S
$ s9 R% O( I8 W7 J$ o, \' jChristensen RA,Fujikawa K, Madore R, Oettgen P, and Varticovski L. NERF2, a member of theEts family of transcription factors, is increased in response to hypoxia andangiopoietin-1: a potential mechanism for Tie2 regulation during hypoxia. J Cell Biochem 85:505-515, 2002.
4 _, [8 \+ Z0 l
) ]1 ~5 Q5 O7 Z. U4 y+ i P* s0 r
5 P# q; y& c7 B% p+ k. I6 b1 i+ `" i
* G# p1 L) S. H8 Z, OCorada M,Mariotti M, Thurston G, Smith K, Kunkel R, Brockhaus M, Lampugnani MG,Martin-Padura I, Stoppacciaro A, Ruco L, McDonald DM, Ward PA, and DejanaE. Vascular endothelial-cadherin is an important determinant ofmicrovascular integrity in vivo. Proc Natl Acad SciUSA 96:9815-9820, 1999.
0 q8 @0 o1 d' u/ G) U5 |3 H3 p h& Q- o) A5 l$ L
+ w+ i2 J6 ^- y$ s8 q3 e7 U7 {0 ]( V* w9 I( X
Dunn KW,Sandoval RM, Kelly KJ, Dagher PC, Tanner GA, Atkinson SJ, Bacallao RL, andMolitoris BA. Functional studies of the kidney of living animals usingmulticolor two-photon microscopy. Am J Physiol CellPhysiol 283:C905-C916, 2002.* Y# J# B! s6 c8 Q' F' U
4 i$ Z+ g3 \. U' Z
) Y1 T/ P8 v+ P' }- g6 ]
0 S/ e6 J! B* k4 p! W4 @! g5 }- |4 LFarrell AJ,Williams RB, Stevens CR, Lawrie AS, Cox NL, and Blake DR. Exercise inducedrelease of von Willebrand factor: evidence for hypoxic reperfusionmicrovascular injury in rheumatoid arthritis. Ann RheumDis 51:1117-1122, 1992.
& L5 \# f% ^7 Z; t2 N9 ^- p; @& c0 d, p! N- J9 c0 n
/ l! }4 q0 V) `' s6 M
, q8 k" A A! dFina L,Molgaard HV, Robertson D, Bradley NJ, Monaghan P, Delia D, Sutherland DR,Baker MA, and Greaves MF. Expression of the CD34 gene in vascularendothelial cells. Blood 75:2417-2426, 1990.
' E3 T. `3 a/ v+ m
' H8 U+ ]2 E/ n' O# i( w( ]" {/ u/ Y ]! F5 x$ y
9 u0 v; w, ?# l
Flores J,DiBona DR, Beck CH, and Leaf A. The role of cell swelling in ischemicrenal damage and the protective effect of hypertonic solute. J ClinInvest 51:118-126, 1972.
# C8 A( g6 x- z) r, \, a ~1 p8 j" ^3 y
5 j- s) d0 L6 x6 a2 }8 w5 n/ x4 \3 K( U7 S+ y& `
Gralnick HR,McKeown LP, Wilson OM, Williams SB, and Elin RJ. Von Willebrand factorrelease induced by endotoxin. J Lab Clin Med 113: 118-122,1989.7 q# x: ]4 g) D# q0 L& Q
& Y& {1 v+ o2 T/ W
0 x( a; a+ x6 n" `( Y$ E
2 w8 ? X7 l5 z gHack CE andZeerleder S. The endothelium in sepsis: source of and a target forinflammation. Crit Care Med 29:S21-S27, 2001.
8 y+ F4 K7 k* G
3 z# m& p2 o# B0 t( M; q6 { B% p
9 V) {% o T: ?
Hellberg PO,Bayati A, Kallskog O, and Wolgast M. Red cell trapping after ischemia andlong-term kidney damage. Influence of hematocrit. KidneyInt 37:1240-1247, 1990.4 X* S1 W$ h5 Y' M
1 F& j6 g: j1 l) \( i7 x
5 {/ n- f3 ]( ~+ M! n2 s1 n
- y( `) ?8 W$ IHellberg PO,Kallskog OT, Ojteg G, and Wolgast M. Peritubular capillary permeabilityand intravascular RBC aggregation after ischemia: effects of neutrophils. Am J Physiol Renal Fluid Electrolyte Physiol 258: F1018-F1025,1990.
: h7 d- o- g c) k* I6 Z) u! x
/ H0 O" Z3 l/ n N
. i9 d" W8 H6 F/ D* M- x; r" a
0 g' C2 |! K P( uHinshaw DB,Armstrong BC, Beals TF, and Hyslop PA. A cellular model of endothelialcell ischemia. J Surg Res 44:527-537, 1988.
1 `, c- J8 z |) B0 z. u5 @% ?" b+ [1 B' V' d1 @6 j/ }4 M' J
7 q1 g5 |5 Q$ r! a& K8 q! L0 W; [) ]( h- {9 U7 q- X/ C
Hinshaw DB,Burger JM, Armstrong BC, and Hyslop PA. Mechanism of endothelial cellshape change in oxidant injury. J Surg Res 46: 339-349,1989.; T. c" F* I, y2 y
$ F- K' k( c2 R1 H
; _' Y$ n# ]/ t7 t. N% j
) `% l5 A( W: w0 Y. `3 ^. JJones DK, PerryEM, Grosso MA, and Voelkel NF. Release of von Willebrand factor antigen(vWF:Ag) and eicosanoids during acute injury to the isolated rat lung. Am Rev Respir Dis 145:1410-1415, 1992.: r$ ^, k% R* ]5 ]( N# z4 S
8 t& K" [9 _4 b4 G) B( l, ^1 \! n& \5 Z. j6 t
, @! X" _ j6 j6 }6 e# w
Kelly KJ,Williams WW Jr, Colvin RB, Meehan SM, Springer TA, Gutierrez-Ramos JC, andBonventre JV. Intercellular adhesion molecule-1-deficient mice areprotected against ischemic renal injury. J Clin Invest 97: 1056-1063,1996.& X' m/ c' @+ u% R5 f* m
3 W5 l1 q; ^9 G1 e: s% [ b% f/ i {
, K3 R- z! Y5 }: S% S& x
4 ?" {0 o6 q q8 c9 I* {% Y* J5 XKevil CG, OhnoN, Gute DC, Okayama N, Robinson SA, Chaney E, and Alexander JS. Role ofcadherin internalization in hydrogen peroxide-mediated endothelialpermeability. Free Radic Biol Med 24: 1015-1022,1998.- \, l8 t4 f5 }5 U
# E( s: Y! U, H! L ~
" J; C, R; d+ D) R' U! t H+ r& v& q0 g
Kevil CG,Oshima T, Alexander B, Coe LL, and Alexander JS. H 2 O 2 -mediated permeability: role of MAPK and occludin. Am J Physiol Cell Physiol 279:C21-C30, 2000.% X* N6 s! N0 g
& {; @- k+ Y' C" Y+ h" _: ?# X5 e2 E6 u! c L
, ~* ^5 d9 }3 z$ a8 Q4 i8 g
Klingebiel T,von Gise H, and Bohle A. Morphometric studies on acute renal failure inhumans during the oligoanuric and polyuric phases. ClinNephrol 20:1-10, 1983.
) p2 |0 | a& e0 h6 _2 r& ?
. D: Q& L, Y0 F0 M/ E' }0 _; _ l ?* W1 S: R2 M
: |. O- \& \0 R! \
Kuhne W,Besselmann M, Noll T, Muhs A, Watanabe H, and Piper HM. Disintegration ofcytoskeletal structure of actin filaments in energy-depleted endothelialcells. Am J Physiol Heart Circ Physiol 264: H1599-H1608,1993.
, r. V" S6 g& ^' g9 a0 O) W! [$ J5 o
/ }0 O7 Z7 G% ^- Q4 M! _7 u; D/ o8 @: Z- c. W
Kurihara H,Anderson JM, and Farquhar MG. Diversity among tight junctions in ratkidney: glomerular slit diaphragms and endothelial junctions express only oneisoform of the tight junction protein ZO-1. Proc Natl Acad SciUSA 89:7075-7079, 1992.
% u& T# h/ B9 I D. I0 u' X9 E/ B- ]8 ~
: A# f1 n9 u8 ^# a* l
8 P) r1 u6 H- d D2 F! X( v. J& @Kwon O,Phillips CL, and Molitoris BA. Ischemia induces alterations in actinfilaments in renal vascular smooth muscle cells. Am J Physiol RenalPhysiol 282:F1012-F1019, 2002.
5 k& n" g+ m" ^2 `- k
$ k, s$ f6 b1 x- [; a% V: r! E: ~: v
& ]& @7 x. a- n1 K6 L9 R, G
Lieberthal W and Nigam SK. Acute renal failure. II. Experimental models of acute renalfailure: imperfect but indispensable. Am J Physiol RenalPhysiol 278:F1-F12, 2000.+ u( _( |( y3 `$ I1 G2 k
4 ?- G4 Y3 {' j! N. g" }6 w. g: s9 D8 D
. q/ x$ l2 V9 _$ h. ^& B( I
( K+ Q- ~8 K4 ?/ S. L% ^$ `, HLip GY andBlann A. Von Willebrand factor: a marker of endothelial dysfunction invascular disorders? Cardiovasc Res 34: 255-265,1997.
( K; a6 ^% x2 G
) V( ~ j& D" K ?) Q5 q W! K: ?3 j6 t9 M
5 O* v5 Y# j9 o9 ~Lum H, Barr DA,Shaffer JR, Gordon RJ, Ezrin AM, and Malik AB. Reoxygenation ofendothelial cells increases permeability by oxidant-dependent mechanisms. Circ Res 70:991-998, 1992.$ c! C& u6 f# t
0 E; J- h* D# x6 x$ w
5 Z A2 `0 r- h5 `* T
7 h0 v3 x8 d9 a/ D1 j
Molitoris BA,Geerdes A, and McIntosh JR. Dissociation and redistribution ofNa ,K -ATPase from its surface membrane actincytoskeletal complex during cellular ATP depletion. J ClinInvest 88:462-469, 1991.' N6 j7 [7 D+ f( {: z2 a( [
4 H7 f4 {7 n8 S! ?4 J( q
# w4 A9 m/ F& Y |$ `$ A5 G: R
4 C6 r4 j1 T& d X0 Q3 `. n3 sMolitoris BA,Wilson PD, Schrier RW, and Simon FR. Ischemia induces partial loss ofsurface membrane polarity and accumulation of putative calcium ionophores. J Clin Invest 76:2097-2105, 1985.* m6 p+ }# O4 R2 J# x" m1 Z
) v# @6 A7 H( N' r7 f
. a" J) u6 E$ I" k; Y: ~% Z2 k' J/ ]& `1 [8 a
Morita K,Sasaki H, Furuse M, and Tsukita S. Endothelial claudin: claudin-5/TMVCFconstitutes tight junction strands in endothelial cells. J CellBiol 147:185-194, 1999.: l; g7 H8 k. u5 {
, j, s ~" ]1 n* D0 v% d2 F& c9 p2 m- y# d$ A \
) `; G0 f4 t6 G# X4 v6 |
Motoike T,Loughna S, Perens E, Roman BL, Liao W, Chau TC, Richardson CD, Kawate T, KunoJ, Weinstein BM, Stainier DY, and Sato TN. Universal GFP reporter for thestudy of vascular development. Genesis 28: 75-81,2000.; u* i# k1 ]# d& l' c
9 W% _" l0 o6 i1 @, U# v; E
) l9 f! ]4 X" D; }' h
. J1 ]1 M6 L2 _) m3 y) }Navarro P,Caveda L, Breviario F, Mandoteanu I, Lampugnani MG, and Dejana E. Catenin-dependent and -independent functions of vascular endothelial cadherin. J Biol Chem 270:30965-30972, 1995.: V6 D1 t# }4 F
0 q+ t% y4 e s$ I; G
* b6 J2 U2 R P& f% y4 C0 D$ s& Y& [7 R* j! x: B/ @7 _2 k
Newsholme SJ,Thudium DT, Gossett KA, Watson ES, and Schwartz LW. Evaluation of plasmavon Willebrand factor as a biomarker for acute arterial damage in rats. Toxicol Pathol 28:688-693, 2000.
, \1 S" u3 Q! e
: g }! Y: q: H* }7 r( ]5 n, q1 Q, O/ n8 S
5 G6 K& p2 J' S$ J. w; y: oNoll T, Muhs A,Besselmann M, Watanabe H, and Piper HM. Initiation of hyperpermeability inenergy-depleted coronary endothelial monolayers. Am J Physiol HeartCirc Physiol 268:H1462-H1470, 1995.
- _5 G4 F: l' Q3 j9 w1 _
. }. t, Y0 y# m, o2 @. g) ^' g5 _& e" A# Q) t1 P
3 R% E; k9 q9 W) Z! F, W' pNovotny MJ,Turrentine MA, Johnson GS, and Adams HR. Experimental endotoxemiaincreases plasma von Willebrand factor antigen concentrations in dogs with andwithout free-radical scavenger therapy. Circ Shock 23: 205-213,1987.: L) m% ^9 q1 D! \) h5 J) L
7 m3 j( ^' V' p# j7 y) O/ L" R$ @! F# A
& E8 [2 m# O& T0 o# \/ [6 TOchoa L, WaypaG, Mahoney JR Jr, Rodriguez L, and Minnear FL. Contrasting effects ofhypochlorous acid and hydrogen peroxide on endothelial permeability:prevention with cAMP drugs. Am J Respir Crit Care Med 156: 1247-1255,1997., ~. \) u! D; r, N7 ~; n) n5 V' t
7 R6 I5 D4 H: v' R3 L
: E: z8 C: j0 P' H
& C! L$ R) [& w( g( l: U0 YPinsky DJ, NakaY, Liao H, Oz MC, Wagner DD, Mayadas TN, Johnson RC, Hynes RO, Heath M, LawsonCA, and Stern DM. Hypoxia-induced exocytosis of endothelial cellWeibel-Palade bodies. A mechanism for rapid neutrophil recruitment aftercardiac preservation. J Clin Invest 97: 493-500,1996.
* L" a" e: l9 q) R
$ w' Z! y( x2 O* ~: E$ h
! q" u% b0 ~" y2 J1 }( R Q/ j8 |! K0 k5 K% \* m
Pomfy M andHuska J. The state of the microcirculatory bed after total ischaemia ofthe brain. An experimental ultrastructural study. Funct DevMorphol 2:253-258, 1992.
5 H" B# v9 ^/ ` o4 W: u+ u
% K `( e4 ^7 t3 _$ k& D1 q
9 o9 L# M5 }' S: C9 e8 B" y8 F" A7 l* v& z' Z1 s
Reinhart K,Bayer O, Brunkhorst F, and Meisner M. Markers of endothelial damage inorgan dysfunction and sepsis. Crit Care Med 30: S302-S312,2002.
7 ?- J; B/ f6 P* Q" n# z5 B5 {2 h- G9 }7 k+ V+ }0 H
9 u5 m" h7 v6 N- j
6 H3 R( ^ G3 V! b2 L3 w
VanBenthuysen KM, McMurtry IF, and Horwitz LD. Reperfusionafter acute coronary occlusion in dogs impairs endothelium-dependentrelaxation to acetylcholine and augments contractile reactivity in vitro. J Clin Invest 79:265-274, 1987.
$ I' A- i; T0 l& S0 t# h# s0 b7 \
/ r6 o% X; Z; h) T" i5 X+ d& ^& B# R4 S5 ~& M! P- Q- I
) D) v. |% n5 f+ q$ HVestweber D. Molecular mechanisms that control endothelialcell contacts. J Pathol 190:281-291, 2000.
8 v( T6 K4 C. E" i# Q6 {. v! c+ p/ A9 H9 e3 W
$ J; a- V( M" U3 n% K P! l6 l; k+ d i" O% P/ O& x: x
Vollmar B,Glasz J, Leiderer R, Post S, and Menger MD. Hepatic microcirculatoryperfusion failure is a determinant of liver dysfunction in warmischemia-reperfusion. Am J Pathol 145: 1421-1431,1994.
4 s: t+ T8 h+ T" R) F1 P$ }' {) R" }4 \
1 ~% H+ ^( c6 X1 O) T0 A1 @- v1 C
f5 U% U; n$ ]/ U" q! GWatanabe H,Kuhne W, Spahr R, Schwartz P, and Piper HM. Macromolecule permeability ofcoronary and aortic endothelial monolayers under energy depletion. Am J Physiol Heart Circ Physiol 260: H1344-H1352,1991.
5 z0 o/ i$ \+ s4 ^# ^0 G9 w
- S/ V; n: m% r" }( w( h s
) W7 x" K, Y6 Q5 u# ~# I
" k9 C3 u, ?8 O& Z! r/ GYamamoto T,Tada T, Brodsky SV, Tanaka H, Noiri E, Kajiya F, and Goligorsky MS. Intravital videomicroscopy of peritubular capillaries in renal ischemia. Am J Physiol Renal Physiol 282:F1150-F1155, 2002.+ Y8 D1 d; u+ E
, a; K) o- T1 r6 U- o6 H
" G& @/ s& g2 i7 B9 Z8 a4 h2 s: A/ h2 j% U9 Y9 i( q
Zhao Y andDavis HW. Hydrogen peroxide-induced cytoskeletal rearrangement in culturedpulmonary endothelial cells. J Cell Physiol 174: 370-379,1998. |
|
发表于 2009-4-21 13:41
