recovery from, postischemic acute renal failure. The goal of this study was to ascertain whether combination ...... HD, Cieslinski. DA, Coimbra. TM, Mes- sana.
An Evaluation Postischemic R.A.
Zager,2
SM.
of Antioxidant Effects on Recovery Acute Renal Failure1
Fuerstenberg,
P.H. Baehr,
D. Myerson,
and
From
B. Torok-Storb
cells, PA.
Zager,
SM.
B. Torok-Storb,
Center and Washington, (J. Am.
Soc.
Fuerstenberg, The
Fred
PH.
the Department Seattle, WA Nephrol.
Baehr,
Hutchinson
of Medicine,
1994;
D. Myerson,
Cancer
Research
University
of
4: 1588-1597)
ABSTRACT Xanthine oxidase (XO) activity and hydroxyl radical (.OH) formation are widely proposed mediators of renal reperfusion Injury, potentially altering the SeverIty of, and recovery from, postischemic acute renal failure. The goal of this study was to ascertain whether combination XO inhibitor (oxypurinol) and #{149}OH scavenger (Na benzoate) therapy, given at the time of renal ischemla, alters the extentof: ( 1) tubular necrosis and filtration failure; (2) DNA fragmentation/ apoptosis (assessed in situ by terminal deoxynucleotidyl transferase reactivity); (3) early tubular regenerative responses (proliferating cell nuclear antigen expression; (3H)thymidine incorporation); and (4) the rate and/or degree of functional and morphologic repair. The effects of XO inhibition, #{149}OH scavengers, and acatalytic iron (FeSO4) on human proximal tubular cell proliferation in vitro were also assessed with a newly established cell line (HK-2). Male Sprague-Dawley rats were subjected to 35 mm of bilateral renal arterial occlusion with or without oxypurlnol/benzoate therapy. These agents did not alter the extent of tubular necrosis or filtration failure, proIlferating cell nuclear antigen expression or thymidine incorporation, or the rate/extent of renal functional/morphologlc not
precede
repair. tubular
DNA
necrosis,
fragmentation
and
It was
did
unaffected
by antioxidant therapy. By 5 days postischemia, treatment groups demonstrated regenerating thelial fronds that protruded into the lumina. structures
contained
terminal
transferase-reactive,
but
I
R#{149}ciiv#{149}d Jun#{149} 22, 1993.
2
Corrssponsnc#{149}to R.A. Zogr,
ter,
I 124 Columbia
5trt,
August
Accepted M-178.
M.D..
Frd
10. 1993. Hutchlnson
SottI#{149}. WA 98104.
1046.6673/0408-1588503.00/0 Journal ot flu Am#{149}rtcan socl#{149}ty01 N.phroloqy Copyright 0 1994 by tM Amdcan SOCI&Y of Nphrology
1588
both epiThese
deoxynucleotidyl
morphologically
Cancer
intact,
Rsorch
suggesting the presence of apoptosis. Oxypuand #{149}OH scavengers (benzoate; dimethylthiourea) suppressed in vitro tubular cell proliferation; conversely, catalytic Fe had a growth-stimulatory effect. These results suggest that: ( 1) XO inhibition/ #{149}OH scavenger therapy has no discernible net effect on postischemic acute renal failure; (2) DNA fragmentation does not precede tubular necrosis, suggesting that It is not a primary mediator of ischemic cell death; and (3) antioxidants can be antiproliferrinol
Cn-
ativefor
human
potential Key Words: clear
tubular
beneficial
antigen,
Apoptos/s,
cells,
possibly
mitigating
their
effects. hydroxyi
thymidine,
radical,
proliferating
cell
nu-
regeneration
controversial issue concerning experimental ischemic renal injury is whether xanthine oxidase (XO) activity, with resultant free radical/hydroxyb radical (#{149}OH) generation. substantially contributes to reperfusion injury and, hence, to the severity of postischemic acute renal failure (ARF) (reviewed in references 1 -3). Although some studies have suggested that XO inhibitors (albopurinob/oxypurinol) and #{149}OH scavengers (e.g. dimethylthiourea/ benzoate) can confer renal functional and morphologic protection (e.g. references 4-8), these positive results have not been universally observed (e.g. references 9- 1 4). Because of this, the importance of renal reperfusion oxidant stress has remained in doubt (1). In studies that have documented beneficial results, it has generally been postulated that antioxidants mitigate renal failure by decreasing the extent of tubular necrosis (e.g. reference 5). However, because renal functional recovery commences within hours of ischemic damage, it is also possible that antioxidants might confer “protection” not by decreasing lethal cell injury. per Se, but rather, by enhancing tubular regeneration rates. For instance, if postischemic oxidant stress were to cause DNA damage, then prevention of it could theoretically accelerate repair. Alternatively. because selected transcription factor (e.g. nuclear factor xB; activator protein1 ) responses to tissue injury are redox sensitive (1 5- 1 8). antioxidant therapy might influence tubular regeneration independent of oxidant-induced DNA breaks. Of note, epidermal growth factor therapy improves postisehemic renri function within 24 h or less of its administration (19), indicating that enhanced tubuA
.
,
,
,
.
Volume
4
‘
Number
8
1994
#{149}
Zager
bar regeneration can, indeed, masquerade as a primary cytoprotective effect. To our knowledge, the influence of antioxidant prophylaxis on postisehemic DNA damage. tubular regeneration, and the rate and extent of organ recovcry has not been previously addressed. Thus, the goal of this study was to address these issues in an attempt to better define the robe of oxygen free radicals in postischemic ARF.
MATERIALS lschemic
AND Renal
Injury
Protocol
,
,
Influence of Antioxidant of Ischemia-lnduced
Therapy on the Tubular Necrosis
Because the extent of ischemic edly an important determinant functional recovery. the influence
of the American
protocol on the severity of the initial ischemic insult was assessed. To this end, six pairs of rats (one antioxidant/one control) were subjected to the ischemic injury protocol. and 24 h later, they were reanesthetized and the left kidneys were removed and fixed in 1 0% buffered formalin for histobogic/ histocytochemical evaluation, as described below. In addition, a terminal plasma sample was assayed for BUN and creatinine (Cr) concentrations as indices of filtration failure.
METHODS
Male Sprague-Dawbey rats (1 75 to 375 g; Charles River Laboratories, Wilmington, MA), housed under standard vivarium conditions, were used for all cxperiments. The rats were anesthetized with pentobarbital (30 to 40 mg/kg ip) and placed under heating lamps to maintain a constant 37#{176}C body temperature; a jugular venous catheter was inserted. Half of the rats received iv bolus injections of antioxidants, whereas the other half of the rats served as controls (equal iv injections of antioxidant vehicles; NaCl was used to equalize molarities). The antioxidant therapy chosen for study was oxypurinob (5 mg/kg) plus sodium benzoate (1 50 mg/kJ. The reasons for choosing this combination were as follows: (1 ) this dose of oxypurinol was demonstrated previously by this laboratory to inhibit effectively rat renal xanthine oxidase/dehydrogenase activity, as indicated by a total block in uric acid generation (1 1 1 2); (2) sodIum benzoate is a potent #{149}OH scavenger (20) and is actively transported by proximal tubular cells; (3) #{149}OH is reported to be a critical free radical for mediating proximal tubular injury(21-23); and (4) combination oxypurinol/benzoate therapy should be particularly effective in decreasing #{149}OH, both by blocking precursor (superoxide) production and by #{149}OH scavenging. Oxypurinob was administered in 0. 1 N NaOH (5 mg/ mL) (1 1 1 2); sodium benzoate was given in 5% dcxtrose/water (30 mgJmL). Ten minutes after the administration of these agents. both renal pedicbes, exposed through a midline abdominal incision, were occluded with smooth vascular clamps for 35 mm. After clamp removal, the rats were sutured and albowed to recover from anesthesia. Rats were operated on in pairs (age/weight matched; one antioxidant treated, one control) in an attempt to minimize cxperimental variability. For this reason, most statistical comparisons were performed by paired analyses (see Calculations and Statistics).
Journal
et al
Society
Extent
injury is undoubtof the rate of renal of the antioxidant
of Nephrology
Influence of Antioxidants From Postischemic ARF
on Recovery
Rates
Fourteen pairs of rats (one antioxidant/one control) were subjected to the ischemia protocol, and the rate of renal functional recovery was assessed over 5 days by the determination of daily BUN and Cr concentrations. These were determined on 200-giL tail vein plasma samples drawn on Days 1 to 4 postischemia. On the fifth day, the rats were reanesthetized and euthanized by exsanguination from the inferior vena cava. The terminal plasma was used for final BUN/ Cr determinations, and the left kidneys were fixed in 10% buffered formalin for histologic and histocytochemical evaluations, described below.
Influence of DNA
of Antioxidant Fragmentation
Therapy
on the
Extent
The following experiment assessed whether XO inhibition/’OH scavenger therapy mitigates DNA fragmentation, theoretically altering the rate of recovery from postischemic ARF. To this end, eight pairs of rats were subjected to the ischemia protocol, and either 2 h (one pair). 6 h (four pairs), 24 h (two pairs), or 5 days (one pair) later, their left kidneys were removed and fixed in 10% buffered formalin. The extent of DNA fragmentation was assessed by the newly described histocytochemical technique of Gavriell et al. (24). This method is based on the specific binding of terminal deoxynucleotidyl transferase (TdT) to the 3’-OH ends of DNA (exposed with DNA breaks), which is followed by polydeoxynucbeotide polymer synthesis. After the exposure of nuclear DNA on histologic sections (see below), the TdT is used to incorporate biotmnylated deoxyuridine at the DNA breakage sites. The biotmnybated signal is amplified by avidin-peroxidase, allowing for its identification by conventional bight microscopy. The specifics of this technique were as follows. Formalinfixed, paraffin-embedded specimens were sectioned onto three ammnopropybtriethoxysibane slides and deparaffinized at 60#{176}C overnight. Sections were rehydrated with double distilled water (DDW), with changes of fresh xybene and gradations of fresh ethanol. Nuclear proteins were degraded through incubation with proteinase K (20 igJmL for 1 5 mm at
1589
Antioxidants
and
Recovery
From
ARF
room temperature). After being washed in DDW, endogenous peroxidases were quenched in 2% hydrogen peroxide for 5 mm at room temperature. The sections were washed again in DDW and immersed in TdT buffer (30 mM Tris IPH 7.2J, 140 mM sodium cacodylate, 1 mM cobalt chloride). Sections were covered with a reaction mixture consisting of 0.3 U of TdT (Promega, Madison, WI) per microliter and 5 MM biotin-labeled UTP (Boehringer. Indianapolis. IN) in TdT buffer and were incubated in a humidified box for 60 mm at 37#{176}C.After incubation, the reaction was terminated in a solution of 300 mM sodium chloride and 30 mM sodium citrate for 1 5 mm at room temperature and was then washed in DDW. Next, the sections were covered with 2% bovine albumin solution and incubated for 10 mm at room temperature. After being rinsed, the sections were Immersed in phosphate-buffered saline for 5 mm and covered with avidin-biotin peroxidase complex (Vectastain ABC; Vector Laboratories, Burlington, CA) for 30 mm at room temperature. After being washed in phosphatebuffered saline, the sections were developed with nickel chloride-enhanced diaminobenzidine for 5 mm at room temperature, followed by counter staining with 0.25% alcian blue-0.25% methyl green. For positive controls, deparaffinized, rehydrated, proteinase K-treated normal kidney sections were incubated with DNase I (Sigma Chemical. St. Louis, MO) in a buffer consisting of 30 mM Tris (pH 7.2). 140 mM sodium cacodylate. 4 mM MgCb2, and 0.1 mM dithiothreitol for 10 mm at room temperature. After thorough washing, the sections were immersed in TdT buffer and processed as described above. Negative controls consisted of two histologic sections from two normal rat kidneys.
Influence of Antioxidant Therapy Postischemic Renal Regeneration
of
Proliferating Cell Nuclear Antigen Expression. To help gauge whether the antioxidant protocol influenced postischemic tubular regeneration, probiferating cell nuclear antigen (PCNA) (or cyclin) expression was analyzed histochemically in postischemic tissues (25-27). The term “PCNA” denotes two popubations of nuclear proteins, one of which is an auxiliary protein to DNA pobymerase #{244} (25,26). When stained with antibody in formalin-fixed tissues, cells within the late G1. S. and early G2 phases demonstrate nuclear staining, giving a semiquantitative index of cells that have left quiescent stages (Go, early G1) (25). Kidneys from 1 4 pairs of rats were studied (2 pairs at 2 h postischemia; 6 pairs at 6 h postischemia; 6 pairs at 24 h postischemia). These kidneys. plus two normal kidneys (controls), were fixed for 24 h in 10% neutral buffered formalin, dehydrated in graded ethanobs and xylene, and embedded in parapbast; 5-
1590
im sections were placed on aminopropyltriethoxysilane (APTS) slides pretreated with citrate (28), as modified by Dr. Allen Gown (Department of Pathology, University of Washington; personal communication). Slides were placed in a plastic Coplan jar with 100 mL of 10 mM citric acid (pH 6.0). The Copban jar was placed in a microwave oven (700 W) and heated for 1 to 2 mm until the solution boiled. The solution was then reheated to boiling twice more for 2 mm each, with the microwave set to 50% power. The evaporated fluid was replaced between cycles with distilled water. The Coplan jar was then placed on ice for 20 mm and washed three times in Trisbuffered saline (TBS: 0.05 M Tris-0. 1 5 M NaCl; pH 7.0). The slides were then treated with 5% normal goat serum-2% BSA in TBS for 10 mm, drained, and incubated with PC1O (M879, Dakopatts, Carpinterra, CA) antibody diluted 1 :250 for 90 mm at room temperature. After the slides were washed four times with TBS, biotinylated goat anti-mouse immunogbobulin G (6650; Tago, Burlingame, CA), diluted 1:200, was applied for 30 mm and washed. Streptavidinhorseradish peroxidase. diluted 1 :200, was applied (43-4323; Zymed. South San Francisco, CA) for 30 mm, followed by TBS washes. The chromogen nickel chboride-complexed diaminobenzidine was applied for 6 mm at room temperature. The slides were counterstained with 0.2% light green, rehydrated rapidly through graded alcohols to xylene. and covered with coverslips in Polymount (Poby Scientific, Bay Shore. NY). Controls included an unrelated antigen of the same isotype (antiamyboid. M759. Dako) that sometimes stained tubule cell cytoplasm in a granular or weak diffuse pattern; in no case, however, was nuclear staining observed). [3Hlmymidine Incorporation. As a second index of renal tubular regeneration, [3Hjthymidine incorporation into DNA was assessed. Three pairs of rats were subjected to the ischemia protocol, and 24 h later. they were lightly reanesthetized with pentobarbital. Each was given an ip injection of 200 iCi of (3Hjthymidine (in 0.5 mL of buffered salt solution). Two hours after the injection, the right and left kidneys were removed, a cross-section through the middbe of each kidney was taken (encompassing cortex/ outer medulla/inner medulla), and DNA was cxtracted (29). 13H]thymidine incorporation was cxpressed as counts per minute per milligram of DNA. To assess background thymidine uptake. three rats, subjected to sham renal ischemia. were injected with (3Hjthymidine and treated as noted above.
Influence Proximal The purinol
of Antioxidant Tubular Cell following experiment and #{149}OH scavengers
Therapy Proliferation
on Human in Culture
assessed directly
Volume
whether affect the
4
.
Number
oxyability
8
1994
#{149}
Zager
of proximal tubular cells to proliferate. To make this assessment, a newly established adult human kidney proximal tubular cell line was used (30). In brief, this cell line, designated HK-2 (human kidney-2), was established in this laboratory by the transduction of a primary culture of human proximal tubular epithehum with human papiblomavirus 16 E6/E7 genes. HK-2 cells demonstrate alkaline phosphatase, acid phosphatase, leucine aminopeptidase, and gamma glutamyltranspeptidase activities, as assessed by histochemistry; they react with antibodies directed against fibronectin, cytokeratin, a3$1 integrin, but not with antibodies against factor VIII-related antigen. 6. 19 antigen, or common acute bymphocytic beukemia antigen endopeptidase. Functionally, they demonstrate Na-dependent/phborizin-sensitive sugar transport, and their adenylate cyclase activity is stimulated by parathormone, but not by antidiuretic hormone, addition. (Thus, in sum, these cells have a proximal tubular origin and a relatively normal phenotypic expression.) To assess the effect of oxypurinob and sodium benzoate on tubular cell proliferation, HK-2 cells were plated into a 96-well, flatbottom microtiter plate at a density of 1 ,000 cells per well in a volume of 200 L of keratinocyte serumfree medium supplemented with epidermal growth factor (5 ng/mL) and bovine pituitary extract (40 ,gJ mL). After 24 h of incubation (5% CO2), the following experimental groups were established: (1 ) controls, no antioxidant addition; (2) oxypurinol addition (5 ,gJmL); (3) sodium benzoate addition (100 SM); and (4) oxypurinob plus sodium benzoate addition (N = 8 wells per treatment). The cells were incubated for 4 days and then quantified by use of the 3-14,5-dimethybthiazol-2-2,5 diphenybtetrazobium bromide (MTT) assay, as previously described (30-32). (This assay is based on the principle that MTT is actively transported by respiring mitochondria; which beads to intramitochondrial formazin crystal precipitation. which can be quantified by ELISA technology. Thus, the amount of precipitation reflects the mitochondrial mass, providing a direct quantitative index of cell number (30-321.) After the completion of the 4day incubation, the cell media were removed from the wells and replaced with 60 L of 0.83 mg of MTT/ mL of culture medium. The plate was incubated at 37#{176}C for 30 mm, the MTT solution was removed, and the precipitated formazin crystals were dissolved in 1 00 iL of dimethybsulfoxide for 60 s on a plate shaker. The absorbance was measured on an ELISA reader (Blo Tek Instruments, Burlington, VT) at 570 nm with a 630-nm reference wavelength. To further assess the influence of antioxidants on tubular cell proliferation, a second culture plate was inoculated with HK-2 cells and the following groups were established: (1 ) controls; (2) sodium benzoate addition (5 mM); (3) dimethylthiourea (DMTU) addi-
Journal
of the American
Society
of Nephrology
ef al
tion (5 mM); (4) mannitob addition (20 mM; equivalent to 5 mM benzoate or DMTU as an #{149}OH scavenger) (20); (5) catalase addition (50 JmL); and (6) FeSO4 addition (0.06 mM) (which causes #{149}OH generation) (20). After 4 days in culture, cell proliferation was assessed by the MTT assay, as described above.
Calculations
and
Statistics
All values are given as means ± 1 SE. The BUN, Cr, and thymidine incorporation data were compared by both paired and unpaired t tests. The MTT assay results were contrasted by one-way analysis of variance with after testing by unpaired t test with Benferroni correction. Histologic comparisons, where possible, are described below.
RESULTS Severity of Initial 24-h Assessments
Ischemic
Renal
Injury:
No significant differences in BUN or Cr concentrations were noted for the six pairs of rats used for histologic assessments at 24 h postischemia (antioxidant treated: BUN, 134 ± 13; Cr, 3.6 ± 0.2; Controls: BUN, 1 16 ± 9; Cr, 3.7 ± 0.2; all in milligrams per deciliter). Furthermore, antioxidant treatment did not affect the extent of tubular necrosis: at 24 h, widespread necrosis of 53 proximal tubular cells was observed in the outer medullary stripe of all kidneys, involving approximately 70% of the tubular segments in this area (Figure 1A). In addition, there was patchy necrosis seen in the medullary rays. Cortical (S/S) proximal tubular segments showed virtually no necrosis. The extent of histologic damage/necrosis appeared virtually identical among all of these 24-h postischemic kidneys, making it impossible to construct a histologic scoring system to ascertain potential statistical differences between the control and antioxidant-treated groups.
Rates
of Renal
Functional
Recovery
The sequential BUN and Cr concentrations for the 1 4 pairs of rats used to assess rates of renal functional recovery are presented in Figure 2. In agreement with the results presented above, the antioxidant protocol failed to confer functional protection, as assessed by the 24-h BUN and Cr concentrations for these rats. By 48 h postischemia. both groups of rats showed substantial improvements in renal function because minimal, or no, increments in azotemia had occurred over the 24-h values. Throughout Days 1 to 4 of follow-up, there were no significant differences in the BUN or Cr concentrations between the two groups. Furthermore, by the fifth day postischemia, essentially identical degrees of renal func-
1591
Antioxidants
and
Recovery
From
ARF
A -I
0
E
z
Day Number
Figure 2. Sequential BUN and Cr concentrations observed on Days I to 5 postischemia forthe antioxidant-treated and control rats. The severity of ARF was virtually identical for the two groups, on the basis of the Day I values, the rate of decline, and the degree of resolution (All, NS).
B
Figure
1. (A) Renal morphology 24 h after ischemic renal antioxidant treatment. Extensive tubular necrosis is apparent within the outer medullary stripe (hematoxylin and eosin stain). (B) Renal morphology 5 days after ischemb: marked tubular dilation is observed (outer medullary stripe), and intraluminal necrotic tubular debris and sloughed tubular cells (arrowheads) are apparent. in addition, a large regenerative papillary outgrowth Is shown in the center of the photograph (). injury
with
tubules were lined with cuboidal epithelial cells without well-developed brush borders; (2) interstitial widening: the interstitial region. particularly in the outer medulla, was widened because of edema and mononuclear cell accumulation; (3) intraluminal necrotic tubular debris, sloughed cells, and casts: many of the dilated tubules, particularly in the outer medulla, contained necrotic tubular cell debris and occasional casts. In addition, many desquamated cells were seen that, by hematoxylin/eosin staining, had darkly stained nuclei; and (4) papillary fronds: in occasional areas throughout the outer medulla, papillary outgrowths of regenerating epithelia were observed (Figures lB and 3). The extent of these changes did not appear to differ between the control and antioxidant-treated groups (seven kidneys cxamined in each group).
TdT Assessment tional recovery had occurred (BUN/Cr: controls, 45 ± 13/0.75 ± 0.20 mg/dL; antioxidant treated: 42 ± 5/ 0.63 ± 0.06 mg/dL; both not significant). Despite this substantial improvement in renal function, the degree of recovery remained incomplete because these BUN/Cr values were still above the normal ranges for the rat (BUN, 10 to 20 mg/dL; Cr, 0.3 to 0.5 mg/dL, respectively, as assessed in this laboratory). Of the 28 rats subjected to this 5-day time course experiment, 5 died (3 controls, 2 antioxidant treated) between 48 and 96 h postischemia, presumably of uremia. Histologic analysis of the kidneys obtained at the end of these experiments (5 days postischemia) confirmed persistent postischemic damage. Four principal abnormalities were noted as follows (Figure 1 B): (1 ) tubular dIlation: marked tubular dilation was observed, Involving large numbers of tubular segments, particularly within the outer medulla. These
1592
Day Number
of DNA
Fragmentation
Normal kidney sections in the absence of DNase treatment demonstrated no TdT-mediated nuclear staining (negative control; Figure 4A). Conversely. the treatment of these same sections with DNase (positive control) caused virtually all nuclei (e.g. of tubules, gbomeruli, vessels) to stain black with the application of the TdT-labeling technique (indicating the ability of this method to detect DNA fragmentation within formalin-fixed kidney sections) (Figure 4B). The 2-h postischemic kidneys failed to demonstrate nuclear staining (not depicted). However, by 6 h postischemia (not depicted), scattered foci of tububes within the outer medullary stripe showed strong nuclear and, in some regions, weaker cytoplasmic staining (the latter suggesting DNA fragment shedding into the cytoplasmic compartment). The involved tubules demonstrated cellular fragmentation. indicating tubular necrosis. Antioxidant therapy did not mitigate these changes observed at 6 h; ,
Volume
4
.
Number
8
‘
1994
Zager
ef al
A
Figure
3. Example
papillary
nodule
cells comprising TdT, indicating Figure 4D).
of a large, regenerating, intraluminal 5 days after ischemic injury. The these nodules were strongly reactive with that they contained fragmented DNA (see observed
rather, the extent of nuclear staining, if anything, appeared slightly greater in the antioxidant treatment group. Analysis of the 24-h postischemic kidneys revealed florid nuclear and cytoplasmic staining of cells in the outer medullary stripe, clearly corresponding to areas of proximal tubular necrosis (Figure 4C). The extent of this change appeared virtually identical in the antioxidant and control postischemic groups. precluding semiquantitative grading. Only occasional foci of nuclear positivity were observed outside of the outer medubbary stripe, being confined to necrotic proximal tubular cells within the medublary rays. TdT staining of the 5-day postischemic kidneys demonstrated only rare nuclear positivity. The involved cells were confined to the outer medullary stripe, and most of them were lying free within the proximal tubular lumina. The papillary fronds stained strongly positive by the TdT technique, mdicating that a majority of the cells attached to these fronds had fragmented DNA (Figure 4D).
PCNA
D
Staining
Normal kidneys demonstrated only occasional proximal tubular cells with PCNA-positive nuclei (Figure 5A). The kidneys examined at 2 and 6 h postischemia did not appear noticeably different from these controls, whether or not they had received antioxidant treatment. However, at 24 h postischemia, large numbers of tubular cells, particularly in the cortex, were strongly PCNA positive (Figure SB). Many outer medullary stripe tubular cells were also PCNA positive, but the extent of this finding was less impressive than that seen in the cortex, undoubtedly because of the widespread tubular cell necrosis and, hence, cell loss. To ascertain whether the PCNA staining at 24 h postischemia differed among the antioxidant versus the control group, the six pairs of matched kidneys were compared by three mndepend-
Journal
C
of the American
Society
of Nephrology
Figure 4. TdT-treated kidney sections. (A) Normal kidney section showing no TdT reactivity; (B) same kidney section, treated with DNase, now shows florid nuclear TdT reactivity, indicating the ability of this technique to demonstrate frogmented DNA; (C) marked, but focal, TdT reactivity In areas of tubular necrosis within the outer medullary stripe 24 h after ischemic injury; (D) strong TdT reactivity of cells comprising a regenerating epithelial nodule 5 days after ischemic injury.
1593
Antioxidants
and
Recovery
From
ARF
A g
250* C
*
200-
C
150
I
100 50 0 C
Oxy
B
Figure 6. HK-2 cell proliferation, as assessed by the MIT assay (reported as absorbance units x f0). Left: oxypurinol (oxy) and sodium benzoate (B; 100 ,M) each caused a highly significant depression in HK-2 proliferation 18%; P < 0.002), and when present together, their suppressive effects were additive (33%). RIght: both benzoate (B) and DMTU at 5 mM concentrations caused moderate suppresslon of HK-2 cell proliferation (P < 0.001), whereas neither catalase (cat) nor mannitol (M) exerted an effect. FeSO4 (Fe) addition (0.06 mM) had a growth-stimulatory effect (P < 0.00 1).
B
(‘.
Figure 5. PCNA staining of a normal kidney section and of a 24-h postischemic kidney (antioxidant treated). The normal kidney (A) showed only an occasional black-stained nucleus (arrowhead). However, florid nuclear staining was observed at 24 h postischemia (B) in both the cortex and outer medullary stripe (top).
ent observers and a score of “greater” or “lesser” PCNA staining was assigned to each kidney of the pair. Each of the observers graded the matched pairs the same: in three of the six pairs, the antioxidanttreated kidney was judged as having slightly less PCNA staining than the control, whereas in the remamning three pairs, the control kidney had slightly less PCNA staining than the antioxidant-treated kidney (i.e. no significant difference was observed between the two groups). ,
(3H)Thymidine
for these rats mg/dL; antioxidants,
HK-2
Cell
As either
shown in Figure 6 (left panel), oxypurinol (5 gJmL) or sodium
were:
con126 ±
Proliferation the addition benzoate
of (100
tiM) caused an 1 8% reduction in HK-2 cell proliferation (P < 0002). When oxypurinol and benzoate were present together, additive suppression resulted (33% decrease; P < 0.001 compared with nonoxidanttreated cells). As depicted in the right panel of Figure 6, 5 mM benzoate caused a 56% decrease in HK-2 cell proliferation. DMTU, a second cell-permeable #{149}OH scavenger, also decreased HK-2 cell proliferation (by 33%). However, neither mannitol (a cell membraneimpermeant #{149}OH scavenger) nor catalase exerted an effect. FeSO4 addition increased HK-2 cell proliferation by 35% (P < 0.00 1).
Incorporation
The extent of [3H]thymidine incorporation, assessed at 24 h postischemia. was virtually the same for the antioxidant-treated kidneys and the postischemic controls (43 ± 8 x 1 O versus 40 ± 9 x 1 O cpm/mg of DNA, respectively). The fact that this degree of thymidmne incorporation was 1 0-fold higher than that observed in the sham-operated rats (4 ± 1 x i0 cpm/mg of DNA; P < 0.001) indicates that increased postischemic DNA synthesis was occur-
1594
ring. BUN/Cr concentrations trols, 108 ± 3/2.8 ± 0.2 17/2.5 ± 0.2 mg/dL.
DISCUSSION In the aftermath verity of filtration ance between a ing the extent cast resolution, ances, and the or repair. Given these changes,
of ischemic renal injury, the sefailure at any time reflects a balnumber of diverse processes, includof tubular necrosis, cast formation/ intrarenal hemodynamic disturbdegree of renal tubular regeneration the number and dynamic nature of it is quite possible that a pharmaco-
Volume
4
Number
#{149}
8
‘
1994
Zager
logic intervention could have a substantial effect on any one of them, and yet. the net effect is insufficient to alter the extent of organ failure. Oxidant stress has been widely proposed as an important mediator of postischemic renal damage largely on the basis of observations that a variety of antioxidants, most notabby XO inhibitors and #{149}OH scavengers, confer functional and morphologic protection. However. that other studies have failed to reproduce these results raises the possibility that these agents may have a beneficial effect on selected injury pathways. but given the multiplicity of overlapping pathophysiologic processes consistent protection need not result. Thus, the goal of this study was to ascertain whether antioxidant (XO inhibitor/#{149}OH scavenger) therapy lessens the extent of ischemic tubular necrosis, limits the amount of DNA fragmentation, or alters rates of renal tubular regeneration or the extent of repair. By taking this approach, our goal was to evaluate more critically the effects of antioxidants on postischemic renal injury, rather than merely assessing their net effect on the severity of filtration failure. In previous experiments reported from this baboratory (9- 1 2), we were unable to demonstrate that a variety of antioxidants, including albopurinol, oxypurmnol, glutathione, superoxide dismutase, and #{149}OH scavengers (DMTU; benzoate), decreased either the severity of filtration failure or the extent of tubular necrosis. However, in each of those studies, female rats were used. Conversely, in studies that have noted antioxidant-induced protection (5-8), male rats have been used (as recently pointed out to one of us [R.A. Zager] by T.J. Burke, Denver, CO). Thus, in this study, male rats were studied to exclude the possibibity that gender explains these discrepant findings. In addition, combination antioxidant treatment (oxypurinol plus sodium benzoate) was used in an attempt to induce a sufficiently barge beneficial effect to mitigate the severity of organ failure. Overall, 23 pairs of rats had renal functional assessments at 24 h postischemia and no difference in the severity of renal failure was noted (composite BUN/Cr values: antioxidants, 1 1 7 ± 6/2.9 ± 0. 1 mg/dL; controls, 10 ± 4/3.0 ± 0.2 mg/dL). Furthermore, the extent of tubular necrosis was not affected. Thus, these findings support our previous view that X0-driven ox!dant stress is of insufficient magnitude to alter consistently the severity of postischemic ARF in the rat (10). One consequence of tissue injury is DNA fragmentation, the degree of which could affect postischemic recovery events. DNA breaks can theoretically occur either as a result of cell necrosis or as a possible pathogenetic step in the development of apoptotic cell death (2,24,33). The batter has typically been defined either by morphology (nuclear chromatin
Journal
of the American
Society
of Nephrology
et al
condensation, as assessed by electron microscopy) or by documentation of a characteristic pattern of DNA fragmentation (“baddering”) by agarose gel electrophoresis. However, the morphologic appearance of apoptosis is extremely transient (24), and “laddering” may result from DNA isolation artifacts. To circumvent these problems, Gavrieli et at. recently developed the TdT method for the in situ labeling of fragmented DNA (24). Because apoptosis may be an important mechanism for early postischemic tubular necrosis (2,33) and because oxidant stress may trigger this pathway (34), we applied the TdT technique to 0- to 24-h postischemic kidneys both to ascertain whether DNA fragmentation occurs before the onset of cell necrosis (suggesting a cause and effect relationship) and to determine whether antioxidant therapy alters its extent. These studies yielded the following insights: ( 1 ) DNA fragmentation and overt cell necrosis appeared to develop simultaneously; (2) essentially all of the fragmented DNA was confined to necrotic epithelium; and (3) antioxidant therapy did not alter its appearance. Thus, these findings suggest that these DNA breaks are a result of the dying process, rather than being a primary mediator of it and that this change results from an XO/.OH-independent mechanism. If the batter is true, then it seems unlikely that XO inhibitors and #{149}OH scavengers would alter renal tubular regeneration by the prevention of DNA breaks. Despite our findings that antioxidants did not alter the severity of filtration failure, the extent of tubular necrosis, or the amount of DNA fragmentation. it remains possible that this therapy might induce other benefits (e.g. altered transcription factor responses) that could then manifest themselves either by enhanced recovery rates or by a decrement in residual tissue damage. Although many studies have evaluated whether antioxidants alter the extent of tubular necrosis, their effect on recovery events has not been assessed. Thus, the third goal of the study was to address this issue. However, as depicted in Figure 2, virtually identical rates of decline in BUN and Cr concentrations were observed between 1 and 5 days postischemia for the control and antioxidanttreated groups. Furthermore, by 5 days postischemia, the same degree of functional and morphologic recovcry had occurred. Thus, these findings support conclusions derived from the 24-h assessments: that antioxidant therapy had no discernible effect on postischemic ARF, when assessed at the whole-organ level. Of considerable interest was that, despite the dramatic improvement in renal function by 5 days postischemia, marked morphologic abnormalities persisted. most notably tubular dilation, persistent intraluminal necrotic debris, desquamated tubular cells, and papillary outgrowths of regenerating epi.
1595
Antioxidants
and Recovery
From APF
thebia. The desquamated cells and papillary fronds were almost uniformly TdT reactive, despite appearing morphobogically intact, consistent with apoptotic cell death. Thus, these observations strongly suggest that apoptosis occurs in the late postischemic period and that it may be an important mechanism for tubular remodeling and for controlling the extent of regenerating epithelia. Because oxidant stress is increasingly recognized as a critical mediator of transcription factor responses (e.g. references 1 5 to 1 8), the provision of antioxidants could conceivably influence in vtvo tubular DNA synthetic rates. To address this possibility, PCNA expression was evaluated in postischemic formalin-fixed tissues; thymidine incorporation was also assessed. Because renal functional recovery commenced at approximately 24 h postischemic (Figure 2), these parameters were evaluated at, or before, this time. Increased PCNA staining was first apparent at 24 h postischemia, and it was widely expressed throughout the cortex and the outer medulla. The fact that the cortex was so prominently involved was surprising because one would expect new DNA synthesis to be largely confined to the outer medullary stripe. the principal area of cell necrosis. However, when sought in formalin-fixed tissues, PCNA expression can reflect late G1 S. and early G2 cell cycle events (27). Thus, the cortical PCNA positivity could have reflected cell movement from Go/early G1 to a late G1 phase. Alternatively, it may have resulted from an unmasking of constitutiveby expressed antigen, thereby allowing for its detection (35). In either event, PCNA expression was not altered by antioxidant therapy. Thymidine incorporation was virtually identical for the two groups of rats; this further suggests that this treatment did not directly alter tubular regeneration rates. Because it is impossible to assess directly the influence of antioxidants on proximal tubular cell regeneration by the use of whole-animal/whole-kidney cxperiments, the last goal of this study was to determine whether XO inhibition and #{149}OHscavenger therapy alter tubular cell proliferation in culture, with a newly established immortalized adult human proximal tubular cell line (HK-2). These experiments produced a number of noteworthy results: ( 1 ) oxypurinol and sodium benzoate (1 00 M) each caused a modest suppression (18%) of HK-2 cell proliferation. and when present together. they exerted a cumubative effect (33% suppression); (2) DMTU was also antiproliferative, suggesting that benzoate induced its influence by #{149}OH scavenging; (3) mannitob, an extracelbular #{149}OH scavenger. had no effect on HK-2 outgrowth, suggesting that benzoate and DMTU acted at an intracellular location; and (4) “catalytic” Fe enhanced HK-2 cell proliferation. Thus, these findings raise the intriguing possibility that #{149}OH exerts a .
.
1596
modest growth-stimulatory effect. If so, then antioxidants could conceivably have a negative effect on the postischemic kidney, possibly canceling out their potential beneficial influences. if so, this might help explain why consistent in vivo protection with these agents need not result. In conclusion, this study indicates that: (1 ) comb!nation XO inhibitor/#{149}OH scavenger therapy does not significantly alter the extent of postischemic tubular necrosis, the severity of filtration failure, early tubular regenerative responses (PCNA expression, thymidine incorporation), or the rate or degree of organ repair. Thus, these observations provide additional support for the view that antioxidant therapy has no consistent net effect on postischemic ARF (9- 1 2); (2) DNA fragmentation and tubular necrosis appear simultaneously in the early (24 h) postischemic kidney, suggesting that DNA breaks occur in the aftermath of ischemic tubular injury but that they are probably a consequence, not a mediator, of it; (3) during recovery from postischemic renal damage, tubular remodeling occurs (denuding of morphobogically intact cells that contain fragmented DNA), a process that may be mediated by apoptosis; and (4) antioxidants can suppress tubular cell proliferation, whereas catalytic Fe stimulates it; this raises the intriguing possibility that #{149}OH exerts a proliferative effect. Thus, antioxidants may exert detrimental, as well as beneficial, influences on the postischemic kidney, suggesting one possible reason for why consistent in vivo protection need not result.
ACKNOWLEDGMENTS The authors thank Ms. K. Burkhart, and Mr. Brian Schimpf for their Zahn for manuscript preparation. Grants
AM
38432
and
CA
15704
Ms. H. Gage, Mr. Kevin Henne, technical support and Ms. Dana This work was supported by from
the
NIH.
REFERENCES 1 . Weinberg JM: The cell biology of ischemic renal injury. Kidneylnt 1991;39:476-500. 2. Bonventre JV: Mechanisms of ischemic acute renal failure. Kidney Int 1 993;43: 1160-1178. 3. Honda N, Hishida A: Pathophysiobogy of experimental nonoliguric acute renal failure. Kidney mt 1 993;43:5 13-21. 4. Vasko KA, Dewa.l1 PA, Riley AM: Effect of albopurinol in renal ischemia. Surgery 1 972;7 1: 787-790. 5. Paller MS, Hoidal JR, Ferris TF: Oxygen free radicals in ischemic acute renal failure in the rat. J Clin Invest 1984;74: 116-1 164. 6. Baker GL, Cony RJ, Autor AP: Oxygen free radical induced damage in kidneys subjected to warm ischemia and reperfusion. Ann Surg 198S; 202:628-641. 7, Dillon JJ, Grossman SH, Finn WF: Effects of oxypurinob on renal reperfusion injury in the rat. Renal Failure 1993;15:37-45.
Volume
4
‘
Number
8
1994
#{149}
Zager
8.
9.
10.
11. 12. 13.
1 4.
1 5.
1 6.
1 7.
18.
1 9.
20.
2 1.
Journal
Hansson RB, Gustafsson B, Jonsson D, et al.: Effect of xanthine oxidase inhibition on renal circulation after ischemia. Transplant Proc 1982; 14:51-58. Zager BA: Hypoperfusion induced acute renal failure in the rat: An evaluation of oxidant tissue injury. Circ Res 1988;62:430-435. Gamelin LM, Zager RA: Evidence against oxidant injury as a critical mediator of post-ischemic acute renal failure. Am J Physiob 1988; 255:F450-F460. Zager RA, Gmur DJ: Effects of xanthine oxidase inhibition on ischemic acute renal failure in the rat. Am J Physiob 1989;257:F953-F958. Zager PA: Hyperthermia: Effects on renal ischemic reperfusion injury in the rat. Lab Invest 1 990;63:360-369. BIrd JE, Milhoan K, Wilson CB, et al.: Ischemic acute renal failure and antioxidant therapy in the rat. The relation between glomerubar and tubular dysfunction. J Clin Invest 1988;81: 1630-1638. Doctor RB, Mandel U: Minimal role of xanthine oxidase and oxygen free radicals in rat renal tubular reoxygenation injury. J Am Soc Nephrol 1991 ; 1:9S9-969. Staal FJT, Roederer M, Herzenberg LA, Herzenberg LA: Intracellular thiols regulate activation of nuclear factor KB and transcription of human immunodeficiency virus. Proc Natl Acad Sc! USA 1990;87:9943-9947. Molitor JA, Ballard DM1, Greene WC: KB-specific DNA binding proteins are differentially inhibited by enhancer mutations and biological oxidation. The New Biologist 1 99 1 ;3:987-996. Schreck R, Rieber P, Baeuerle PA: Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-KB transcription factor and HW- 1 . EMBO J 1 99 1 ; 10: 2247-2258. Meyer M, Schreck R, Baeuerle PA: H202 and antioxidants have opposite effects on activation of NF-KB and AP-1 in intact cells: AP-1 as a secondary antioxidant responsive factor. EMBO J 1993;12:2005-2015. Humes HD, Cieslinski DA, Coimbra TM, Messana JM, Galvao C: Epidermal growth factor enhances renal tubule cell regeneration and repair and accelerates the recovery of renal function in post-ischemic acute renal failure. J Cbin Invest 1989;84:1757-1761. Zager RA, Foerder CA: Effects of inorganic iron and myogbobin on in vitro proximal tubular lipid peroxidation and cytotoxicity. J Clin Invest 1 992;89:989-995. Paller MS, Hedlund BE: Role of iron in postischemic renal injury in the rat. Kidney i 988; 34:474-480.
mt
of the
American
Society
of Nephrology
22.
23.
24.
25.
26.
27. 28.
29.
30.
3 1.
32.
33.
34.
35.
et al
Paller MS: Hemoglobin-and myogbobininduced acute renal Taibure in rats: Robe of iron in nephrotoxicity. Am J Physiol 1988;255: F539-F544. Shah SV, Walker PD: Evidence suggesting a robe for hydroxyl radical in gbyceror-induced acute renal failure. Am J Physiol 1988;255: F438-F443. Gavrleli Y, Sherman Y, Ben-Sasson SA: Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J Cell Biol 1992;1 19:493-501. Connolly KM, Bogdanffy MS: Evaluation of proliferating cell nuclear antigen (PCNA) as an endogenous marker of cell proliferation in rat liver: A dual-stain comparison with 5-bromo-2’ deoxyuridine. J Histochem Cytochem 1993;41 :1-6. Bravo R, Macdonald-Bravo H: Existence of two populations of cycbin/probiferating cell nuclear antigen during the cell cycle: Association with DNA replication sites. J Cell Biob 1 987; 10: 1549-1554. Bravo R, Rainier F, Blundell PA, MacdonaldBravo H: Cyclin/PCNA is the auxiliary protein of DNA polymerase-#{244}. Nature 1987;326:515-517. Cattorettl G, Becker MH, Key G, et at.: Monocbonal antibodies against recombinant parts of the Ki-67 antigen (MIB 1 and MIB 3) detect probiferating celfs in microwave-processed formalin-fixed paraffin sections. J Pathol 1992; 1 68:357-363. Sambrook J, Frltsch EF, Manlatis T. Molecular Cloning. A Laboratory Manual. Vol. 2. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press; 1989:9.16-9.20. Ryan MJ, Johnson G, Fuerstenberg SM, Zager RA, Torok-Storb B: HK-2: An immortalized proximal tubule epithebial cell line from normal adult human kidney. Kidney Int 1994;45:48-57. Mossman T: Rapid coborimetric assay for cebbubar growth and survival: Application to proliferation and cytotoxicity assays. J Immunol Methods 1983;65:55-63. Slater TF, Sawyer B, Strauli UD: Studies on succinate-tetrazolium reductase systems. III. Points of coupling of four different tetrazobium salts. Biochim Biophys Acta 1963;77:383-393. Schumer M, Columbel MC, Sawczuk IS, et al.: Morphologic, biochemical, and molecular cvidence of apoptosis during the reperfusion phase after brief periods of renal ischemia. Am J Pathol 1992; 140:831-838. Ueda N, Shah SV: Endonuclease-induced DNA damage and cell death in oxidant injury to renal tubular epithelial cells. J Cbin Invest 1992;90: 2593-257. Schwartlng R: Little missed markers and Ki-67. Lab Invest 1993;68:597-599.
1597
