Nephrol Dial Transplant (1996) 11: 635-642
Nephrology
Dialysis
Transplantation
Original Article
Synergistic renal protection by combining alkaline-diuresis with lipid
peroxidation inhibitors in rhabdomyolysis: possible interaction between
oxidant and non-oxidant mechanisms
Abdulla K. Salahudeen, Chunyou Wang, Steven A. Bigler1, Zhongyi Dai2 and Hiroyasu Tachikawa2
Departments of Medicine and 'Pathology, University of Mississippi Medical Center and 2Department of Chemistry, Jackson
State University, Jackson, USA
Abstract
Background and purpose. Heme-proteins, besides causing renal tubular obstruction, may contribute to rhabdomyolysis-induced renal injury through a heme-ironmediated lipid peroxidation process. In the present
study, we compared the combined therapy of a lipid
peroxidation inhibitor, 21-aminosteroid (21-AS) and
fluid-alkaline-mannitol (FAM) diuresis with either of
them alone to determine the efficacy of the combination
therapy and to delineate the roles of lipid peroxidation
and cast formation.
Methods and results. Employing Raman spectroscopy,
we confirmed in vitro the ability of 21-AS to inhibit
iron-induced fatty acid peroxidation. 21-AS was then
administered to rats developing renal failure from
glycerol-induced rhabdomyolysis. Although 21-AS
inhibited rhabdomyolysis-induced plasma and renal
lipid peroxidation, renal protection was incomplete.
Administration of FAM to inhibit cast formation
afforded a better renal protection. However, when
these therapies were combined to inhibit both lipid
peroxidation and cast formation, there was a synergistic renal functional protection. This was accompanied by a maximum inhibition of renal and plasma lipid
peroxidation, as well as, renal tubular necrosis and
cast formation. Compared to combination therapy,
FAM therapy alone, despite identical volume, was
accompanied by a higher tubular necrosis and cast
formation.
Conclusions. That combining a lipid peroxidation
inhibitor with fluid-alkaline diuresis in rhabdomyolysis
further lowers renal lipid peroxidation, tubular necrosis
and cast formation and synergistically limits renal
dysfunction (i) supports a role for lipid peroxidation
in the pathophysiology of rhabdomyolysis ARF,
(ii) underscores the role of intratubular heme retention,
a cause for tubular obstruction as well a source for
prodigious amount of iron, likely involved in the lipid
Correspondence and offprint requests to: Abdulla K.. Salahudeen,
MD, FRCP, Renal Division, Department of Medicine, University
of Mississippi Medical Center, 2500 North State Street, Jackson,
MS 39216-4505, USA.
peroxidation, and (iii) raises the possibility of interactions between non-oxidant and oxidant mechanisms.
Key words: acute renal failure; 21-aminosteroid; hemeproteins; lipid peroxidation; rhabdomyolysis; Raman
spectroscopy
Introduction
Following rhabdomyolysis, inordinate amounts of
myohaemoglobin are released into the systemic circulation and its prompt discharge into the renal tubules
sets the stage for the initiation of renal injury process
[1,2]. Both oxidant and non-oxidant mechanisms are
considered to be important in the complex pathophysiology of myohaemoglobin-mediated renal injury
of rhabdomyolysis [1-7]. Hypovolaemia and metabolic
acidosis facilitate tubular precipitation of myohaemoglobin and therefore, form the basis for fluid-alkalinediuresis therapy [1,2,6]. Myohaemoglobin, besides
causing tubular obstruction, may contribute to tubular
epithelial cell injury through a probable heme-ironmediated lipid peroxidation mechanism. It is conceivable that the latter process may further compromise
the dynamics of tubular fluid flow, interfering with the
complete clearance of tubular heme proteins, and thus
initiating a vicious cycle. Even in the absence of hemeprotein precipitation, which can be achieved by vigorous fluid-alkaline diuresis, tubular injury may still
occur because of the continued bathing of the tubular
cells in myohaemoglobin nitrate. The latter may facilitate tubular cell heme-loading which along with the
attendant release of catalytically active iron [7] may
potentially instigate cellular lipid peroxidation and
injury. That both oxidant and non-oxidant mechanisms
are important in rhabdomyolysis-induced renal injury
is suggested by the finding that neither diuresis nor
antioxidants therapy alone is completely protective
[3,5,7].
In this study, we attempted to define the relative
significance of the two above mentioned main median-
1996 European Dialysis and Transplant Association-European Renal Association
636
A. K. Salahudeen et at.
isms, specifically, one related to cast formation and excitation filter and a charge coupled device detector, all
the other related to oxidative injury. With regard to obtained from Spex Inc. +(Edison, NJ). A Spectra-Physics
the latter, we were particularly interested to examine (Mountain View, CA) Kr laser (647.1 nm) was used as an
the role of lipid peroxidation that often accompanies excitation source. For controlling the CCD detector and
rhabdomyolysis-induced acute renal injury [3,5]. analysing the Raman data, a Spex model DM 3000 data
Recent cell culture studies suggest a role for lipid system with an IBM PC was used.
The following samples were prepared for the Raman
peroxidation in hydrogen peroxide-induced renal epi- spectroscopic
to study the effect of
thelial cell injury [8]. Although both excessive hydro- 21-aminosteroid investigation
on iron-dependent lipid peroxidation.
gen peroxide generation and lipid peroxidation have Sample A consisted of 5 ul of 99% linolenic acid, 35 ul of
been demonstrated in kidneys sustaining injury from 100% ethanol, and 10 ul of deionized water (DIW).
rhabdomyolysis [4,5], the role of accompanying lipid Spectroscopy was carried out immediately on this vortexed
peroxidation in rhabdomyolysis-induced renal injury sample placed on a quartz plate, to obtain the spectrum of
is not well denned.
pure linolenic acid, i.e. without the possible autooxidation
21-aminosteroid (U-74006 F) the agent employed from incubation. The remaining samples (below) were subto inhibit lipid peroxidation in this study is a relatively jected to Raman spectroscopy following a 16 h incubation
new group of antioxidant that potently inhibits lipid at 37°C. Sample B (time-control) consisted of 5 ul of 99%
peroxidation, primarily by scavenging lipid peroxyl linolenic acid, 35 ul of 100% ethanol, and 10 ul of deionized
and phenoxyl radicals [9] and has been demonstrated water (DIW). Sample C (with iron) contained 5 ul of linolenic acid, 5 ul of 2 mM FeCI2 solution in DIW, 35 ul of
to potently inhibit oxidant-induced renal tubular cell ethanol,
and 5 ul of DIW. Sample D (with iron plus
lipid peroxidation and injury [8]. Furthermore, 21-aminosteroid) consisted of 5 ul of linolenic acid, 5 ul of
21-aminosteroid and its potent analog 2-methyl amin- 2mM FeCl2 in DIW, 5 ul of 1 mM 21-aminosteroid in
ochroman have been recently reported to modestly ethanol, and 30 ul of ethanol. Sample E (with iron plus
reduce renal injury in the rhabdomyolysis model BHT) consisted of 5 ul of linolenic acid, 5 ul of 2 mM FeCl2
[10,11]. Although these recent studies imply a thera- in DIW, 5 ul of 2% butylated hydroxy toluene in ethanol,
peutic potential for these compounds, the important and 30 ul of ethanol. BHT was included as a known inhibitor
question whether such therapy is better than existing of lipid peroxidation.
fluid-alkali-mannitol therapy has not been addressed
previously. Moreover, since myohaemoglobin-induced
renal injury in rhabdomyolysis is probably a com- Effect of 21-aminosteroid on glycerol-induced acute
bination of oxidant and non-oxidant mechanisms, renalfailure and lipid peroxidation parameters
another important question whether combining
21-aminosteroid with fluid-alkali-mannitol therapy Two of three groups of overnight dehydrated (18 h) male
provides additional renal protection has also not been Sprague-Dawley rats of similar weight (250-275 g), serum
creatinine, haematocrit, and serum protein were injected i.m.
addressed.
In the present study, employing Raman spectroscopy with 10 ml/kg of 50% glycerol, half each into each hind
for the first time, we confirmed the capacity of muscles under mild anaesthesia (pentobarbital sodium,
21-aminosteroid to directly inhibit iron-induced lipid 20 mg/kg, i.p.). One of these glycerol injected groups
peroxidation. We then examined in the glycerol- (GARF + 21-AS), received 21-aminosteroid; first of three
doses (10 mg/kg) was administered by tail vein 15 min after
induced rhabdomyolysis model, the effect of the
glycerol, and second and third at 4 h intervals. The other
(i) inhibiting renal lipid peroxidation by 21- group
(GRAF) injected with glycerol received vehicle of
aminosteroid, (ii) inhibiting the tubular cast forma- 21-aminosteroid, 0.15 ml/100 g body weight of 0.01 M HC1,
tion by vigorous fluid-alkali-mannitol diuresis, or 3 times in an identical manner. The third group, the sham
(iii) inhibiting both lipid peroxidation and cast forma- control, received i.m. normal saline instead of glycerol and
tion by combining these two therapies on myohaemog- the above vehicle. To minimize variability in water drinking,
lobin-induced renal injury. The data that both lipid rats were gavaged with 0.5 ml/100 g of water twice at 2 and
peroxidation and cast formation contribute to the 4 h after glycerol. Free water intake was allowed 6 h after
rhabdomyolysis-associated renal injury and the com- glycerol. Rats were reanaesthetized at 24 h after glycerol,
bination therapy affords more than an additive renal blood samples were drawn from vena caval puncture, plasma
protection suggest that such combination therapy, separated immediately stored at — 84°C. Renal cortices were
unlike either of them alone, may offer the best renal rapidly dissected out on a petri dish with ice-cold normal
outcome. Such synergistic renal protection with the saline, rinsed, blotted, weighed and frozen at -84°C.
Haematocrit (microcapillary method), plasma protein
combination therapy may also suggest the possible
existence of interactions between oxidant and non- (refractory method), and serum creatinine (Beckman autoanoxidant mechanisms, potentially compounding the alyser, Somerset, NJ) were determined.
Evidence for lipid peroxidation in the kidney samples was
severity of rhabdomyolysis-induced renal injury.
Methods
Effect of 21-aminosteroid on iron-inducedfatty acid
peroxidation: Raman spectroscopy
Raman measurements were carried out by a system consisting
of a triple-spectrometer, a micromate microscope, a tunable
sought by measuring two products of lipid peroxidation—
malondialdehyde and conjugated diene—by the assay
methods described in our previous studies [12,13]. Briefly,
renal cortex was homogenized in 100 mM potassium chloride
containing 3 mM EDTA and malondialdehyde in the supernatant was measured in the presence of antioxidants—EDTA
(3mM) and butylated hydroxytoluene (2% BHT,
lOul/ml)—to minimize procedural lipid peroxidation.
Malondialdehyde was measured as thiobarbituric acid react-
637
Lipid peroxidation in rhabdomyolysis-renal failure
ive substances (TBARS) based on the method of Ohakawa.
Malondialdehyde by the TBA reaction was also measured in
the plasma in the presence of EDTA and BHT. In a
previously reported study, we have validated this method of
measuring malondialdehyde by TBA reaction by a more
specific method employing high pressure liquid chromatography (HPLC) technique [13]. For conjugated diene determination, supernatant was extracted in chloroform-methanol
mixture (2:1), dried in nitrogen, redissolved in cyclohexane
and read at 234 nm UV. An extinction coefficient of 27 000
M " ' c m " ' was used and values were corrected for protein
by Lowry method.
In a separate study, similar groups of rats that had
undergone the above protocol were subjected to renal haemodynamic measurements at 24 h. This study was undertaken
in batches of three rats, one from each group. Renal clearance
studies, details reported previously [12,13], were performed
under anaesthesia. Fluid loss during the study was replaced
by an initial iv infusion of 6% albumin (1% of body weight)
for 30 min, followed by a maintenance hourly infusion of
0.5 ml normal saline. Blood pressure and temperature were
continuously monitored. Haematocrit and serum protein
were also monitored. To determine glomerular filtration rate
(GFR), a solution of 1 ml normal saline containing [125I]
sodium iothalamate (2 |iCi/ml) was infused as a slow single
bolus. Timed urine collections were obtained through the
bladder catheter. To measure renal blood flow, an electromagnetic flow probe (Carolina Medical Electronics, King,
NC) was hooked on to the renal artery and bloodflowwas
continuously monitored. Plasma CPK (kit no. 520-5, Sigma
Diagnostics, St Louis, MO) was measured. No lipid peroxidation parameters were determined.
Effect of combining 21-aminosteroid with fluid-alkalinemannitol therapy on glycerol-induced acute renal failure
and lipid peroxidation parameters
In this study we compared the effect of 21-aminosteroid
therapy with the conventional fluid-alkali-mannitol therapy
on rhabdomyolysis-renal injury and determined the effect of
combining these two therapies concomitantly on the kidney.
Identical model as above was employed, except that postglycerol therapy additionally included fluid-alkali-mannitol
or a combination of 21-aminosteroid and fluid-alkalimannitol. Dehydrated rats were injected i.m. with 50%
glycerol at 10 ml/kg. Fifteen minutes after glycerol, the
GARF group was injected tail-vein with the vehicle for
21-aminosteroid, 0.5 ml of 0.01 M HO, and twice more at
4h intervals. The GARF+ 21-AS group was injected similarly with 21-aminosteroid at lOmg/kg body weight in the
above vehicle. In the fluid-alkali-mannitol (GARF + FAM)
group, 1.5 ml/100 g body weight of sodium bicarbonate
solution (40 meq of NaHCO3 in 1 litre 0.45% sodium chloride
in water) was infused over 15 min using Sage minipump
(Orion Research Inc., Cambridge, MA). This was followed
by a 0.35 ml/100 g of 25% mannitol in the next 15 min.
The above NaHCO3 solution was gavaged at 0.75 ml/100 g
twice at 2 h intervals, starting 2 h after the completion
of i.v. infusion. This group also received the vehicle
for 21-aminosteroid. In the combination group
(GARF + 21-AS + FAM), the therapies, i.e. 21-aminosteroid
and fluid-alkali-mannitol were combined. Drinking water
was allowed after 6 h of glycerol. The groups of rats who
did not receive fluid-alkali-mannitol therapy were gavaged
with 0.5 ml/100 g of water twice at 2 and 4 h after glycerol.
Sample collection and measurements were carried out at 24 h
as before. Blood urea nitrogen (kit no. 535-A, Sigma
Diagnostics, St Louis, MO) levels and renal histology were
the additional measurements in this study. CPK or renal
haemodynamic measurements were not obtained. For histological evaluation of the kidney, the middle third of the right
kidney was sectioned into three coronal slices, fixed in
formalin and stained with haematoxylin and eosin.
The extent of tubular necrosis and cast formation were
blindly determined under light microscopy in the following
manner. The number of tubular sections with cast or necrosis
was counted separately in four fields of each of the three
kidney sections obtained from a rat. Eachfieldwas composed
of about 400 tubular sections. To maintain consistency and
to focus on the maximally injured proximal tubules, fields
were chosen by scanning around the cortical margin at
2.5 mm apart. Presence of cast was easily discernible
(Figure 4A) and necrosis was counted only when it was
frank as defined by the complete absence of tubular cells
(Figure 4A). The counts were assigned a score: 0-1 tubular
sections with cast or necrosis per four medium power
fields=l; 2-10 = 2; 11-20 = 3; and so on. More than 100 was
assigned a score of 11.
Materials
All chemicals were obtained from Sigma (St Louis, MO)
and radiolabels from Amersham (Arlington Heights, IL).
Statistical methods
Results are presented as mean + SEM. Data on multiple
groups were analysed by ANOVA and corrected for multiple
comparison by Tukey's HSD (Systat, Inc., Evanston, IL).
Significance was set at P<0.05 level.
Results
Effect of 21-aminosteroid on iron-inducedfatty acid
peroxidation: Raman spectroscopy
Raman bands of pure linolenic acid without incubation
(Figure 1A) displayed four bands in the 11001800 cm" 1 region: 1264 cm"1 from =CH in-plane
deformation, 1302 cm" 1 from in-phase CH2 twisting,
1441 cm" 1 from C-H bending, and 1652 cm" 1 from
C = C stretching [14]. Spectrum changes caused by the
incubation of linolenic acid with iron (Figure 1C)
included the appearance of two new bands at
1637 cm" 1 (shoulder) and 1167 cm" 1 . The intensity of
both 1264 cm" 1 and 1652 cm" 1 bands also decreased
during the incubation of linolenic acid with iron. Since
these changes are associated with desaturation of fatty
acids [14], it is concluded that the iron dependent
peroxidation reduced the number of C = C bonds. The
new band at 1167cm"1 could be associated with a
C-O bond formed by the peroxidation of linolenic
acid. The spectrum changes observed in the timecontrol sample (Figure IB) is much smaller compared
with FeCl2-incubated-linolenic acid (Figure 1C).
Inclusion of 21-aminosteroid along with the incubation
of linolenic acid and FeCl2 appreciably reduced the
FeCl2-induced changes (Figure ID). BHT, a known
inhibitor of lipid peroxidation, inhibited the FeCl2-
A. K. Salahudeen et at.
638
2.01
1.5"
Si i.
°
0.5"
0.0
10
8
6
r-
—
1200
1300 1400 1500
1600
1700 1800
2-
RAMAN SHIFT(cm-l)
Fig. 1. Incubation of linolenic acid with iron (C) resulted in the
appearance of two new bands at 1637 cm"1 (shoulder) and
1167cm"1 and reduction in the intensity of 1264 cm" 1 and
1652 cm"1. The spectrum changes observed in the time-control
sample (B) were much smaller compared with FeCl2-incubatedlinolenic acid (C). Inclusion of 21-aminosteroid and BHT along
with the incubation appreciably reduced the FeCl2-induced above
changes indicative of peroxidation (D,E). Spectrum in A is that of
linolenic acid without incubation.
SHAM
n=6
GARF
n=7
GARF+
21-AS
n=7
Fig. 2. GFR and RBF in the Sham (saline), GARF (glycerol plus
vehicle) and GARF + 21-AS (glycerol plus 21-aminosteroid) groups
of rats: *, P<0.05 versus Sham and f, P<0.05 versus GARF.
group, compared to 0.50 + 0.1 mg/dl in the sham control. Although 21-aminosteroid administration inhibited rhabdomyolysis-induced lipid peroxidation, based
induced linolenic acid spectral changes similar to on the plasma and kidney homogenate levels of malon21-aminosteroid. FeCl2-induced linolenic acid peroxi- dialdehyde and conjugated dienes, renal functional
dation in this experimental setting was further verified protection, although significant, was far from complete:
by the thiobarbituric acid reaction (data not shown). the serum creatinine in the GARF + 21-AS was
These results directly demonstrate the ability of 1.67±0.15 compared to 0.5 + 0.1 mg/dl in the sham
21-aminosteroid to inhibit iron-dependent peroxida- control (Table 1).
The above finding of modest renal protection by
tion of unsaturated fatty acid.
21-aminosteroid was confirmed in a separate study in
which GFR was measured by the 125I iothalamate
Effect of 2]-aminosteroid on glycerol-induced acute
clearance method (Figure 2). GFR at 24 h was
renal failure and lipid peroxidation parameters
1.70 + 0.12 ml/min in the sham rats injected with saline
Despite similar degree of dehydration and renal func- that was strikingly low in the glycerol injected group,
tion, i.m. administration of glycerol but not saline was at 0.059+0.33 ml/min. Confirming the serum creatinaccompanied by fairly severe acute renal failure. At ine data, 21-aminosteroid therapy, although signific24 h, serum creatinine was 2.5 + 0.3 in the GARF antly improving GFR (0.264 + 0.082 ml/min), it was
Table 1. Measurements in the sham rats, glycerol-injected rats (GARF) and glycerol-injected rats receiving 21-aminosteroid (GARF + 21-AS)
Baseline
24 h post-glycerol
Body
Weight***
(g)
Hct
(%)
Serum
protein
(g/dl)
Serum
creatinine
(mg/dl)
0.43 + 0.01
0.41 +0.01
0.40 ±0.01
6.2 + 0.04
5.9 + 0.04
5.9 + 0.06
0.5 ±0.1
2.5 + 0.3*
1.6 + 0.2**
Sham
249 + 2
GARF
250 + 3
GARF + 21-AS 249 ± 2
Kidney
Plasma
Kidney
conjugated dienes
malondialdehyde malondialdehyde
(nmol/ml)
(nmol/mg protein) (nmol/mg protein)
5.5 + 0.13
7.9 + 0.85*
4.24 + 0.70
*p<0.05 versus the rest, **versus control; ***post-dehydration; « = 4 in each group.
0.57 ± 0.07
1.40±0.13*
0.69 ±0.12
0.105 + 0.01
0.168 ±0.01*
0.095 ±0.01
Lipid peroxidation in rhabdomyolysis-renal failure
639
ineffective in maintaining the GFR in any level closer
to the normal level (Figure 2). Renal blood flow at
24 h was significantly reduced following glycerolinduced rhabdomyolysis, but was significantly better
with the administration of 21-aminosteroid (Figure 2).
Plasma CPK level, an index of muscle injury, was
elevated following glycerol-injection compared to
saline (33.3±12.5 versus 12.5±2.5 Sigma U/ml).
21-Aminosteroid therapy had no significant effect on
the glycerol-induced CPK elevation (29.5 ±1.5 Sigma
Units/ml), thus discounting any important effect of
antioxidant therapy on muscle injury in this protocol.
The striking finding of this study was the remarkable
renal protection when therapies were combined.
Compared to the 2.58 ±0.20 mg/dl of serum creatinine
in the rhabdomyolysis control (GARF), serum creatinine rose only to 0.82 ±0.10 with combination therapy
(Figure 3). The difference between the mean serum
creatinine in the GARF and GARF+ 21-AS + FAM
was 1.78, which exceeded the combined difference of
1.24 from the individual therapies. Because of the
curve-linear relation between serum creatinine and
GFR, the lower serum creatinine in the combined
therapy group may even represent a greater preservation of renal function than appreciated by serum
creatinine measurements. Findings similar to serum
Effect of combining 21-aminosteroid with fluid-alkalinecreatinine was also demonstrable with BUN values
mannitol therapy on glycerol-induced acute renal failure
(Figure 3). Thus, the renal protection from the comand lipid peroxidation parameters
bination therapy was more than additive.
Renal function with 21-aminosteroid therapy alone
Such synergistic renal functional protection with
was again significantly better than the untreated combined therapy was accompanied by an almost
glycerol injected control (Figure 3). The group treated complete inhibition of tubular cast formation as well
with fluid-alkali-mannitol alone had a slightly better as maximum reduction in tubular necrosis (Figure 4B).
renal function than the group treated with This group also had the maximum inhibition of sys21-aminosteroid alone: the serum creatinine was temic and renal lipid peroxidation (Figure 5). Plasma
1.80 + 0.2 mg/dl compared to 2.1 ±0.25 mg/dl in the malondialdehyde was 3.65±0.72 nmol/ml, the lowest
21-aminosteroid alone group (Figure 3). The respective of all the groups. The rest were 6.13±0.54 in the
BUN values were 83 + 10 and 94 + 13 mg/dl (Figure 3). GARF, 5.25±0.22 in the GARF + 21-AS (P<0.05
versus GARF) and 3.92±0.34 in the GARF + FAM
CP<0.05 versus GARF and GARF + 21-AS).
150 -i
The renal protection of 21-aminosteroid therapy
appeared to have derived from its ability to limit lipid
•
Pre-glycerol
peroxidation (Figure 5) and cell necrosis (Figure 4B)
•
Post-glycerol
and not from reducing cast formation (Figure 4B),
100since the latter score was slightly higher than the
glycerol-treated control. The fluid-alkali-mannitol
group received the same solution in an identical
manner; yet, in the absence of 21-aminosteroid admin50"
istration, there was over a three-fold increase in
myohaemoglobin cast retention and over a two-fold
increase in tubular necrosis (Figure 4B). Remarkable
also was thatfluid-alkali-mannitoltherapy, which predictably reduced the tubular heme-protein load based
on tubular cast score (Figure 4B), was also attended
by a reduction in renal lipid peroxidation (Figure 5).
11
jj!
1
Discussion
In this study, inhibition of lipid peroxidation by administering 21-aminosteroid during the evolution of hemeprotein-induced renal injury was accompanied by a
GARF +
GARF +
GARF +
GARF
modest renal improvement. 21-Aminosteroid adminisFAM
21-AS
FAM+
tration limited tubular necrosis but not the formation
21-AS
n=6 h each group
of casts. Although still incomplete, renal function was
Fig. 3. Serum creatinine and BUN were determined before and 24 h slightly better preserved with fluid-alkali-mannitol
after glycerol in GARF, GARF + 21-AS, GARF + FAM, and diuresis that appeared to have been due its effect on
GARF + FAM + 21-AS. **: Factors for conversion to SI unit is 88.4 reducing the cast formation, and not tubular necrosis.
for creatinine (umol/l) and 0.357 for BUN (mmol/l). *: /><0.05 In contrast, combination therapy aimed at inhibiting
versus GARF. t: P<0.05 versus the rest. The GARF group both lipid peroxidation and preventing cast formation
received the vehicle, GARF+ 21-AS group 21-aminosteroid, the
GARF + FAM group sodium bicarbonate in saline and mannitol was accompanied by a synergistic renal functional
protection that was also attended by a maximum
and the GARF + FAM+ 21-AS group above therapies combined.
A. K. Salahudeen et al.
640
—r—
0.4.
•t
•
•
Tubular necrosis
Tubular cast
0.2-
GARF
GARF
21-AS
GARF +
FAM
GARF +
FAM+
21-AS
Fig. 5. Lipid peroxidation products in the renal cortical homogenate
of the four groups of glycerol-injected rats receiving vehicles,
21-aminosteroid,fluid-alkali-mannitolor both 21-aminosteroid and
fluid-alkali-mannitol. *: P<0.05 versus GARF. f. P<0.05 versus
GARF + FAM.
GARF
(b)
GARF +
21-AS
GARF +
FAM
GARF-I
FAM +
21-AS
Fig. 4. (A) Two indices were scored blindly under light microscopy
on the formalin-fixed hematoxylin and eosin stained kidney sections,
one, for tubular cast indicated above by the thin arrow and the
other, for frank tubular necrosis, indicated by the thick arrow.
(B) Renal histological score of tubular necrosis and tubular cast
formation in the four groups of glycerol-injected rats receiving vehicles, 21-aminosteroid, fluid-alkali-mannitol or both
21-aminosteroid andfluid-alkali-mannitol.*: P<0.05 versus GARF.
•f: P<0.05 versus the rest.
inhibition of lipid peroxidation as well as tubular
necrosis and cast formation.
While there is little doubt that heme-proteins play a
central role in the pathogenesis of rhabdomyolysisinduced acute renal failure, accruing evidence suggests
that besides causing obstructive cast and direct tubular
toxicity (non-oxidant mechanism), heme-protein, by
being the source for an enormous amount of catalytic
iron, can contribute to renal injury through a lipid
peroxidation mechanism (oxidant mechanism) [3-5,7].
The findings in this study, besides confirming the roles
of both oxidant and non-oxidant mechanisms in the
pathogenesis of rhabdomyolysis, suggest a relative
preponderance for the role of non-oxidant mechanism.
However, therapy against the non-oxidant mechanism
alone was inadequate in preserving the renal function.
Such inadequacy is probably reflective of the inability
of fluid-alkali-mannitol to prevent heme-protein-
induced tubular necrosis. Heme-protein filtrate,
although diluted from diuresis, is still capable of generating excessive amount of catalytically active iron as
recently demonstrated [7]. Such excess iron, although
far less than in the untreated glycerol group, could still
instigate lipid peroxidation and cell necrosis. This
possibility is supported by the finding here that, compared to fluid-alkali-mannitol therapy alone, addition
of 21-aminosteroid to fluid-alkali-mannitol further
reduced the extent of tubular lipid peroxidation,
necrosis and renal dysfunction function. It should
be noted however, that, although addition of
21-aminosteroid further enhanced the renal protectiveness of alkaline-mannitol-diuresis, its therapy alone
did not offer any advantage over the conventional
fluid-alkali-mannitol diuresis. The upper end of the
recommended dose of 21-aminosteroid employed here
was adequate enough to completely suppress both the
plasma and the renal lipid peroxidation (Table 1)
and therefore, further increase in the dose of
21-aminosteroid is not expected to bring any additional
renal protection.
Fluid-alkali-mannitol alone was effective in reducing
lipid peroxidation and this may relate to the dilution
of tubular heme-protein concentration rather than
to reduced tubular damage, since tubular necrosis
was still a prominent feature in the diuresis group
(Figure 4B). Alternatively, presence of cast in tubular
Lipid peroxidation in rhabdomyolysis-renal failure
lumens may be a prerequisite to instigate the oxidative
cellular injury. The previous demonstration that direct
addition of iron or heme-protein induces cellular lipid
peroxidation and injury would however, argue against
this [7,10].
The rhabdomyolysis model in the rat, in many ways
similar to the human syndrome, is studied here (as by
many others) with prior dehydration [3-5]. Prior
dehydration allows the full expression of the renal
injury, particularly the cast formation, compared to
the non-dehydrated model [7]. Furthermore, dehydration is not an uncommon occurrence in the clinical
setting of rhabdomyolysis [1,2]. Therapies in this study
were instituted 15min after the initiation of muscle
injury; a clinically relevant question is whether further
delay in the institution of combination therapy, especially when conventional therapy is ineffective, still
affords renal protection following rhabdomyolysis.
Although lipid peroxidation is well recognized to
occur during the renal injury, particularly associated
with oxidative stress, its direct role in the causation of
renal injury is less well defined. Recent studies suggest
a role for lipid peroxidation in certain in vitro and
in vivo settings of renal injury [7,12,13,15], but data
from others indicate that its occurrence does not
necessarily mean that it is causal to the co-existing
pathophysiology [16,17]. The present data that the
inhibition of lipid peroxidation by an anti-oxidant or
by an alkaline-diuresis, the latter by reducing the
concentration of catalytically active iron, results in
renal protection further defines a role for lipid peroxidation in rhabdomyolysis-induced renal injury.
A number of studies support the view that it is the
catalytically active iron released from heme-protein
that mediates the renal injury in this model [3,5,7]. An
important mechanism of iron-induced tissue injury is
its well-known ability to induce lipid peroxidation
[18,19]. Indeed, iron-induced lipid peroxidation invitro is a model that has been extensively employed in
the study of lipid peroxidation [18,19]. It was in one
such model that the potency of 21-aminosteroid to
inhibit iron-induced lipid peroxidation wasfirstdemonstrated [20]. Subsequent studies in our laboratory and
that of others have shown that the ability of
21-aminosteroid to inhibit lipid peroxidation was not
limited for iron [7,9]. The ability of 21-aminosteroid
to inhibit lipid peroxidation resides in the amino group,
which by combining with lipid radicals limit the propagation of lipid peroxidation [9,20].
The present data from the Raman spectroscopic
study confirms the ability of 21-aminosteroid to directly inhibit iron-induced lipid peroxidation. Although,
Raman spectrum has been recorded previously for
fatty acids of varying unsaturation [14], to our knowledge this is the first time this methodology is applied
to study the spectral changes associated with the
peroxidation of a lipid [21]. The application of Raman
spectroscopy in the biomedical field is in its infancy
and currently limited to the in vivo study of atherosclerosis and the diagnosis of breast cancer [21].
We have chosen polyunsaturated linolenic acid
641
rather than arachidonic acid to complete the present
study because we found that the latter was less susceptible to iron-induced spectral changes, i.e. peroxidation,
the reason for which is not clear at present.
21-Aminosteroid, like BHT, was able to directly attenuate the iron-induced linolenic acid peroxidation. The
present finding that 21-aminosteroid inhibits fatty acid
peroxidation is consistent with our recent report that
21-aminosteroid inhibited H2O2-induced renal epithelial cell F2-isoprostane production. F2-isoprostanes are
newly identified arachidonic peroxidation products.
When infused into the kidneys, they induce intense
renal vasoconstriction [22]. In the present study,
21-aminosteroid administration was attended by preservation of renal blood flow, compared to untreated
rats with rhabdomyolysis. Increased F2-isoprostane
production may be one factor for myohaemoglobininduced renal vasoconstriction.
The fluid-alkali-mannitol protocol, adopted here
based on clinical practice, ensured that there was active
diuresis with effective alkalinization of the urine (pH
ranged 7-7.5). Although mannitol is considered to
have hydroxyl radical scavenging property, recent studies specifically addressing this issue in the setting of
rhabdomyolysis-renal injury demonstrate that the
efficacy of mannitol in rhabdomyolysis is most probably due to its diuretic effect rather than an antioxidant
effect [7]. The clinical rationale for the systemic alkalinization in rhabdomyolysis has been experimentally
substantiated [6]. The benefit includes not only reducing the heme-protein precipitation but also the formation of toxic met-hemoglobin. However, no study has
previously combined alkalinization and diuresis with
Myohemoglobin
Renal Tubules
Oxldarrt
Mechanism
Non-oxidant
Mechanism
Alkallne-dlu!re«l». /
Cast
Obstruction
Direct toxicity
\
Detferrioxamtne
y21-A mlnosterold
ureal*
Iron
Oxidative stress
Lipid peroxidation
Tubular necrosis and dysfunction
Fig. 6. This simplified scheme, based on existing data and the
findings of the present study, underscores the role of both oxidant
and non-oxidant mechanisms in the mediation of myohaemoglobininduced tubular injury. Data in this study also raises the possibility
of confounding interactions between these two mechanisms; a possibility that strengthens a case for combining therapy in rhabdomyolysis-related renal injury.
642
A. K. Salahudeen et al.
antioxidant therapy in a rhabdomyolysis model. In a References
previous experimental study on rhabdomyolysis1. Better OS. Stain JH. Early management of shock and prophyinduced acute renal failure, combining desferrioxamine
laxis of acute renal failure in traumatic rhabdomyolysis. N Engl
with mannitol-diuresis but without alkalinization was
J Med 1990; 322: 825-827
2. Odeh M. The role of reperfusion-induced injury in the pathogenaccompanied by an additive renal protection with
esis of the crush syndrome. New EngJ Med1991; 324: 1417-1422
minimal tubular cell necrosis. In that study, although
3. Paller MS. Hemoglobin and myoglobin-induced acute renal
desferrioxamine therapy alone virtually eliminated the
failure in rats: role of iron in nephrotoxicity. Am J Physiol 1988;
enormous increase in catalytically active urinary iron,
255: F539-F544
renal protection was incomplete. This finding relates 4. Guidet B, Shah SV. Enhanced in vivo hydrogen peroxide generation by rat kidney in glycerol-induced acute renal failure. Am
to our present one, i.e. 21-aminosteroid alone effecJ Physiol 1989; 257: F440-F445
tively inhibited lipid peroxidation, but it did not pro5. Shah SV, Walker PD. Evidence suggesting a role for hydroxyl
vide complete renal protection. Data on tubular casts
radical in glycerol-induced acute renal failure. Am J Physiol
1988; 255: F438-F443
in the present study suggest that the reason for incom6. Zager RA, Gamelin LM. Pathogenetic mechanisms in experiplete renal protection with antioxidants alone is likely
mental hemoglobinuric acute renal failure. Am J Physiol 1989;
due to inability to eliminate the tubular cast formation.
256: F446-F455
Thus, even if iron-mediated lipid peroxidation is inhib- 7. Zager RA. Combined mannitol and deferoxamine therapy for
myohemoglobinuric renal injury and oxidant tubular stress.
ited either by 21-aminosteroid or by desferrioxamine,
Mechanistic and therapeutic implications. J Clin Invest 1992;
tubular casts can continue to compromise renal
90: 711
function.
8. Salahudeen AK, Role of lipid peroxidation in H 2 O 2 -induced
renal epithelial (LLC-PK,) cell injury. Am J Physiol 1995;
Although kidneys end up serving a 'sacrificial role',
37: F30-F38
while trying to rid the systemic circulation off the 9. Braughler JM, Pregenzer JF. The 21-aminosteroid inhibitors of
lipid peroxidation: Reactions with lipid peroxyl and phenoxyl
flagitious myohemoglobin, the systemic nature of rhabradicals. Free Rad Biol Med 1989; 7: 125-130
domyolysis is demonstrated by the finding that lipid
10. Nath KA, Balla J, Croatt AJ, Vercellotti GM et al. Heme
peroxidation products were elevated in the systemic
protein-mediated renal injury: A protective role for
circulation and not merely confined to the kidney. This
21-aminosteroids in vitro and in vivo. Kidney Int 1995; 47:
592-602
relates to the previous observation that catalytically
Salahudeen AK. Novel lipid peroxidation inhibitor 2-methyl
active iron was markedly elevated not only in the urine 11. aminochroman
improves renal function in glycerol-induced
but also in the plasma [7]. Such systemic nature of
acute renal failure. Clin Res 1993; 41: 327A
iron release and lipid peroxidation provide further 12. Wang C, Salahudeen AK. Cyclosporine-nephrotoxicity:
Attenuation by an antioxidant-inhibitor of lipid peroxidation
indirect evidence for a link between lipid peroxidation
in-vitro and in-vivo. Transplant 1994; 58: 940-946
and heme-protein released catalytic iron.
13. Wang C, Salahudeen AK. Lipid peroxidation accompanies cyclosporine nephrotoxicity. Effects of vitamin E. Kidney Int 1995;
Both cast formation and oxidant-induced lipid per927-934
oxidation appear capable of contributing to renal 14. 49:
Tu AT. Raman spectroscopy in biology: Principles and applicainjury, independently and interdependently. That the
tions. John Wiley and Sons, New York: 1982
combination therapy employed here affords more than 15. Salahudeen A, Badr K, Morrow J, Roberts J II et al. Hydrogen
peroxide induces 21-aminosteroid-inhibitable F 2 -isoprostane
an additive renal protection along with maximum
production and cytolysis in renal tubular epithelial cells. J Am
suppression of cast formation and lipid peroxidation
Soc Nephrol 1995; 6: 1300-1303
could best be explained by the existence of confounding 16. Ramasammy LS, Christine J, Ling K et al. Effects of diphenylphenylenediamine on gentamycin-induced lipid peroxidation and
interactions between these two mechanisms (Fig. 6).
toxicity in rat renal cortex. J Pharmacol Exp Ther 1986; 238:
This notion, along with the demonstrable synergistic
83-88
efficacy of combined therapy to limit renal dysfunction 17. Tribble DL, AW TV, Jones DP et al. The pathophysiological
significance of lipid peroxidation in oxidative cell injury.
in this study, suggests that concurrent therapy aimed
Hepatology 1987; 7: 377-387
at both the oxidant and non-oxidant mechanism may
Halliwell B, Gutteridge JMC. Lipid peroxidation: A radical
yield the maximum protection against rhabdomyolysis- 18. chain-reaction.
In: Free radicals in Biology and Medicine.
induced acute renal failure.
Clarendon Press, Oxford; 1988: 188-276.
Acknowledgements. This study was supported by funding from the
American Heart (MS affiliate) grant-in-aid program and the Baxter
extramural grant program to AKS. The continued support of Dr
John O'Connell and Dr John Bower of University of Mississippi
Medical Center is appreciated, so is the gift of 21-aminosteroid
(Lazaroid) compounds by Dr John McCall of Upjohn Company.
C. Wang was a post-doctoral fellow from the Department of Surgery,
Union Hospital, Tongji Medical University, Wuham, People's
Republic of China.
19. Braughler JM, Duncan LA, Chase RL et al. The involvement
of iron in lipid peroxidation: importance of ferric to ferrous
ratios in initiation. J Biol Chem 1986; 261: 10282-10289
20. Braughler JM, Pregenzer JF, Chase RL et al. Novel
21-aminosteroid as potent inhibitors of iron-dependent lipid
peroxidation. J Biol Chem 1987; 262: 10438-10440
21. Schrader B, Keller S, Lochte T, Fendel S, Moore DS, Simon A,
Sawatzki J, Nir FT et al. Raman spectroscopy in medical
diagnosis. J Mol Struct 1995; 348: 293-296
22. Takahashi K, Nammour RM, Funaguka M et al. Glomerular
actions of a free radical-generated novel prostaglandin, 8-epiprostaglandin F 2 a, in the rat: evidence for interaction with
thromboxane A 2 receptors. J Clin Invest 1992; 90: 136-141
Received for publication: 10.10.95
Accepted in revised form: 18.12.95