Clinical Science (1990) 78,445-450
445
Epidermal growth factor accelerates functional recovery from
ischaemic acute tubular necrosis in the rat: role of the
epidermal growth factor receptor
JILL NORMAN, YONG-KWEI TSAU, ANGELITO BACAY
AND
LEON G. FINE
Division of Nephrology, Department of Medicine, IJCLA School of Medicine, Los Angeles, California, U.S.A.
(Received 11 September/S December 1989; accepted 20 December 1989)
SUMMARY
1. Severe, ischaemic, acute tubular necrosis was
induced in rats by bilateral occlusion of the renal arteries.
The experimental group received exogenous epidermal
growth factor infused directly into the renal arterial circulation. Serum creatinine concentration was measured
daily for 1 week. Epidermal growth factor receptor binding was measured by autoradiography of whole kidney
sections. Renal cell proliferation was measured by
incorporation of [3H]thymidineinto DNA.
2. Serum creatinine concentration increased after
acute tubular necrosis with a peak at 48 h and remained
elevated above control levels after 7 days. Binding of
radiolabelled epidermal growth factor occurred in all
regions of the kidney 48 h after ischaemia. Treatment with
exogenous epidermal growth factor attenuated the rise in
serum creatinine by 4 days after acute tubular necrosis
and after 7 days serum creatinine was lower than in
animals that did not receive epidermal growth factor.
Infusion of epidermal growth factor also increased renal
DNA synthesis.
3. The increase in epidermal growth factor binding in
the kidney after acute tubular necrosis and the attenuation of the increase in serum creatinine concentration by
administration of exogenous epidermal growth factor,
suggest a role for epidermal growth factor in recovery
from ischaemic damage. The increase in DNA synthesis
in response to epidermal growth factor indicates that its
effect may be due, at least in part, to accelerated tubular
cell proliferation.
Key words: creatinine, epidermal growth factor, epidermal
growth factor receptor, necrosis, regeneration.
Abbreviations: ATN, acute tubular necrosis; EDTA,
ethylenediaminetetra-acetate; EGF, epidermal growth
factor; GFR, glomerular filtration rate; Hepes, 4-(2hydroxyethy1)-1-piperazine-ethanesulphonicacid.
INTRODUCTION
Although much attention has been paid to understanding
the pathogenesis of acute tubular necrosis (ATN) of the
kidney, in the clinical setting the syndrome is most often
encountered after it has become fully established and
when preventive measures are of no value. Early intervention to hasten the recovery process would thus be of value
in reducing morbidity or mortality. We previously
reported the potent mitogenic effect of epidermal growth
factor (EGF) on renal proximal tubular cells in primary
culture and noted that the response to EGF was
enhanced by angiotensin I1 [ 13. EGF receptors have also
been identified on a variety of cell types in the kidney,
including mesangial cells [2] and renal medullary interstitial cells [3].On the basis of these observations we
reasoned that the regenerative process in ATN may be
partially under the control of EGF and that exogenous
EGF may be efficacious in hastening the recovery phase
by promoting mitogenesis of the tubular cells. Therefore,
we examined the binding of radiolabelled EGF to the rat
kidney and the effect of exogenous EGF, infused into the
renal arterial circulation, on renal function and renal
DNA synthesis after induction of ATN.
This work was presented, in part, at the Annual Meeting of the American Society of Nephrology, 1988.
METHODS
Induction of ATN
Correspondence: Professor .I.Norman, Division of Nephrology, Department of Medicine, UCLA School of Medicine,
Los Angeles, CA 90024, U.S.A.
Male Sprague-Dawley rats (280-320 g body weight)
were anaesthetized with ether and the left femoral artery ’
was catheterized with PE-10 polyethylene tubing. The
446
J. Norman et al.
abdomen was opened via a midline incision and the tip of
the catheter was positioned in the aorta to lie cephalad to
the origin of both renal arteries. A bolus of 0.1 ml of 5%
(w/v) Methylene Blue dye was injected and the appearance of a brief blueish colouration of both kidneys confirmed the position of the catheter tip. The catheter was
then secured with silk ties around the femoral artery and
exteriorized at the back of the neck. NaCl(0.970, w/v) was
infused at a rate of 1.5 ml/h.
Both renal pedicles were exposed by blunt dissection
and a strip of aluminium foil was placed around both
renal arteries. Compression of the foil effectively
occluded the arteries and achieved complete cessation of
blood flow. Standard models of ischaemic ATN (i.e.
ischaemic periods of less than 60 min) result in loss of cell
polarity and disruption of plasma membrane function
rather than cell necrosis [4]. Pilot studies revcaled that a
60 min period of ischaemia led to an approximately fivefold increase in serum creatinine concentration within
24 h and a return to twofold above baseline by 48 h.
Since a more prolonged course of renal dysfunction was
desired, a 90 min period of renal ischaemia was induced
in all animals. After occlusion of the renal arteries, the
abdominal viscera were covered with 0.9% (w/v) NaClsoaked gauze, the midline tissues were approximated and
the animal was maintained in a heated chamber at 37°C.
After 90 min of ischaemia, the aluminium foil strips were
removed and blood flow returned to the kidneys immediately. Where the pink colouration of both kidneys was
not complete, the experiment was terminated and the
animal was killed. To confirm ATN, kidneys were fixed
and processed for routine histology of haemotoxylineosin sections.
Administration of EGF
The surgical incision was closed in two layers and the
exteriorized tip of the aortic catheter was connected to an
Alzet 2001 osmotic minipump (Alza Corp., Palo Alto,
CA, U.S.A.), implanted subcutaneously, containing either
EGF (0.1 pg/ml; tissue culture grade, Sigma) in 0.9%
(w/v) NaCl ( H =10) or 0.9% (w/v) NaCl alone (control,
12 = 10).Given a constant pumping rate of 1 pl/h, as specified by the manufacturer, the rate of the delivery of EGF
was 0.1 pg/h. The capacity of the pump allowed up to 8
days of continuous infusion.
Measurement of renal function
Animals were maintained in balance cages for 7 days.
Blood was drawn from the tip of the tail 1, 2, 3, 4
and 7 days after surgery. Creatinine concentration was
measured using a Sigma Diagnostics Creatinine Kit.
Measurement of renal DNA synthesis
Two animals from each group (EGF and control) were
killed 1 and 2 days after surgery, and DNA content and
[3H]thymidine incorporation were measured. Before
being killed, animals received an intraperitoneal injection
of [3H]thymidine (0.25 mCi per animal; sp. radioactivity
33 Ci/mmol; ICN, Costa Mesa, CA,'U.S.A.). After 2 h the
kidneys were harvested, fixed and processed for histology.
DNA was extracted from paraffin-embedded tissues as
follows. Tissue (2CO-300 mg) was pulverized to a fine
powder in a pre-cooled pestle and mortar on dry ice. The
powder was incubated in 500 mmol/l Tris-HC1 pH 9.0,
20 mmol/l ethylenediaminetetra-acetate (EDTA) pH 8.0,
10 mmol/l NaCI, 1% (w/v) sodium dodecyl sulphate and
1p g of Proteinase K/ml (incubation buffer) at 48°C for 24
h with intermittent, vigorous vortexing. The deproteinized
suspension was extracted three times with phenol-chloroform/3-methyl-l-butanol (24 :1, v/v)-incubation buffer
(3:4: 2, by vol.) and once with chloroform alone. The salt
concentration was adjusted to 300 mmol/l with sodium
acetate and 2 vol. of 100% (v/v) ethanol was added to
precipitate the DNA. The pellet was washed with 80%
(v/v) ethanol, freeze-dried and dissolved in 10 mmol/l
Tris-HCl and 1 mmol/l EDTA, pH 8.0. DNA concentration was measured by absorbance at 260 nm. [WIThymidine content was measured by scintillation counting.
Results were expressed as d.p.m./pg of DNA.
Measurement of '251-EGFbinding
Autoradiography of binding of 12SI-EGF(sp. radioactivity = 900 Ci/mmol; New England Nuclear,
Wilmington, DE, U.S.A.) was performed as previously
described [6] with minor modifications. Two animals were
killed at each time point (0, 15, 24, 48 h) after ischaemia
or sham-surgery; the kidneys were embedded in OCT
medium (Lab-Tek) and frozen at - 80°C. Frozen sections
(20 p n ) were incubated in 25 mmol/l 4-(2-hydroxyethyl)-1-piperazine-ethanesulphonic acid (Hepes), 140
mmol/l NaCI, 5 mmol/l KCI, 1 mmol/l MgCl,, 1.8 mmol/l
CaCl,, 0.1% (w/v) bovine serum albumin, 0.025% (w/v)
bacitracin, 0.0125% (w/v) ethylmaleimide, pH 7.4, and
150 pmol/l 1251-EGFwith or without
mol/l unlabelled EGF, for 20-22 h at 4°C. Sections were washed
twice in 25 mmol/l Hepes and 0.1% (w/v) bovine serum
albumin for 2 min at 4"C, rinsed in distilled water and
dried over a cold stream of air at 4°C.
Processing of the sections, autoradiography and
densitometry were performed by methods previously
described [6,7]. Mean raw density values for the concentration of EGF-binding sites were obtained from a minimum of seven separate readings on randomly selected
areas of the mid-cortical region of the kidney and were
corrected for the non-linearity of the film [6].
Since low levels of specific binding of EGF were found
in the control rat kidneys (see the Results section), we also
examined EGF binding to mouse and rabbit kidneys for
comparison and, additionally, compared kidney and liver
in the same animal.
Specific binding was calculated as the difference
between total binding and the binding measured in the
presence of lo-" mol/l unlabelled EGF.
Statistics
EGF-infused and control groups were compared using
an unpaired t-test for all components of the study.
Epidermal growth factor and tubular injury
RESULTS
Histological confirmation of ATN
Twenty-four hours after the ischaemic insult, the
kidneys were enlarged and swollen. Congestion of the
cortical and medullary interstitium was evident. By light
microscopy of haematoxylin-eosin-stained sections, the
447
glomeruli appeared normal. Patchy necrosis of proximal
tubules with loss of nuclei and nuclear fragmentation was
evident throughout the superficial and deep cortex, with
many tubules showing a vacuolated appearance and containing amorphous eosinophilic material within the
lumen. Distal tubules were not easy to define. Widening of
the interstitium with small collections of inflammatory
cells at the boundary zones of necrotic areas was evident.
No attempt was made to evaluate the effects of EGF on
renal histology.
EGF binding to rat, mouse and rabbit kidney
Autoradiograms of IZSI-EGF binding to normal
kidneys of mouse, rat and rabbit (Fig. 1) revealed a low
level of specific binding in all three species. A comparison
between EGF binding to kidney and liver of the mouse
(Fig. 1) indicated that this is not a function of the
methodology, since specific binding to the liver was much
greater than in the kidney. In mouse and rabbit kidneys,
specific binding appeared to be higher in the papillary
region and the pelvicalyceal system, whereas in the rat this
was not apparent.
Fig. 2 shows an increase in kidney size and an increase
in both specific and non-specific binding by the rat kidney
48 h after induction of ischaemic ATN. The time course
of the increased EGF binding in the kidney is shown in
Fig. 3, indicating that there is an increase (two- to six-fold)
in specific EGF binding 48 h after induction of ATN
( P < 0.0 1 vs earlier time points).
Effects of EGF on recovery from ATN
Fig. 4 illustrates the time course of serum creatinine
concentrations in EGF-treated (17 = 10) and control
( I Z = 10) rats. Values in the EGF-treated animals were
significantly lower than control values at 4 and 7 days
( P < 0.05).
Effect of EGF infusion on ['Hlthymidine incorporation
into DNA
Incorporation of ['Hlthymidine into DNA in control
and EGF-treated kidneys was measured. In controls
infused with 0.9% (w/v) NaCI, [3H]thymidine incorporation increased from 78.6 d.p.m./pg of DNA at 24 h to
160.7 d.p.m./,ug of DNA at 48 h. In EGF-treated kidneys,
[3H]thymidine incorpbration at 24 h was 152.5 d.p.m./pg
of DNA (approximately twice the control value) and
increased to 240.8 d.p.m./pg of DNA at 48 h.
Fig. 1. Autoradiograms (2-week exposure) of lzS1-EGF
(150 pmol/l) binding to frozen sections of normal rat
kidney (top), rabbit kidney (middle) and mouse kidney
and liver (bottom) in the absence (left) and presence
(right) of
mol/l unlabelled EGF. A low level of binding is seen in all three species. Increased specific binding
is apparent in the papillary region and pelvicalyceal
system in mouse and Yabbit kidney, but not in rat kidney.
The high level of specific binding [total binding (left)
minus non-specific binding (right)] in the mouse liver is
evident.
DISCUSSION
There is little information available on the factors which
promote regeneration of the renal tubular epithelium after
acute damage. Previous studies have focused mainly on
pathogenic factors and protective measures. In view of the
poor prognosis and high mortality which accompany
ischaemic ATN, insights into measures which hasten the
recovery process would be of value. In this regard, any
448
J. Norman et al.
Fig. 2. Autoradiogram (2-week exposure) of '2sI-EGF(150 pmol/l) binding to frozen sections of
normal rat kidney (top) and rat kidney 48 h after 90 min of ischaemia (bottom). Total EGF
binding is shown on the lcft and non-specific binding on right. There is an increase in specific and
non-specific binding after 48 h of ATN. An increase in kidney size is also evident.
factor which improves glomerular filtration rate (GFR)
directly would be expected to limit the extent of renal
failure. Atrial natriuretic peptide has been shown to
preserve GFR and to reduce renal tissue drainage after
ischaemic renal failure in the rat [8].In contrast, selective
renal infusion of EGF decreases GFR, mainly due to a fall
in the glomerular capillary ultrafiltration coefficient [2],so
that any effect of EGF in enhancing recovery from ATN is
unlikely to be mediated by a direct effect.
One approach to management would be to accelerate
Epidermal growth fac:tor and tubular injury
n
Control
15 h
24 h
48 h
Fig. 3. Time course of specific EGF binding to rat
kidney after 90 min of ischaemia. Increased binding
occurs after 48 h. Each bar is the m e a n f ~of ~seven
~
densitometric measurements per kidney. Readings were
made over randomly selected areas in the mid-cortical
region.
I
0
1
2
3
4
5
t
I
6
7
Time after ischaemia (days)
Fig. 4. Time course of serum creatinine concentration
after 90 min of ischaemia in the presence ( 0 ; rz = 10)and
absence ( 0 ;it = 10)of an exogenous EGF infusion. Serum
creatinine concentration was significantly lower 4 and 7
days after ischaemia (*P<0.05).
regeneration of the partially necrosed tubular epithelium
in severe cases of ATN. Based upon our observations in
vitro that EGF is a potent mitogen for renal proximal
tubular cells in primary culture [l], we investigated
whether the surviving renal cells would respond to this
mitogen after severe ATN. Although proximal tubular
cells in culture possess high-affinity receptors for EGF
(approximately 2 x 104receptors per cell irz v i m [I]), we
found that compared with the liver, the kidney has a relatively low EGF receptor density. This is, perhaps, not
surprising in view of the very low mitotic index of normal
renal tubular cells, reflecting a very slow rate of cell turnover [9]. The relatively higher level of EGF binding to the
papillary and pelvicalyceal regions of the normal kidney
of the mouse and rabbit, raises the possibility that one
role for EGF is the maintenance of the integrity of the
449
uroepithelium, which is derived from the same embryological structure (the ureteric bud) as the collecting
tubules.
Despite the relatively low control level of EGF binding,
there was a marked increase in specific EGF binding 48 h
after induction of ATN. There are two possible explanations for this finding: either the remaining tubular cells
increase the density of their EGF receptors, favouring
mitogenesis of these cells; or a new population of regenerating cells has arisen with a greater number of EGF
receptors. Our data favour the former possibility, since
the histology at 24 h did not reveal the presence of a
population of newly formed tubular epithelial cells in the
necrotic areas. The suggestion that sub-confluent cells
have a higher number of EGF receptors than confluent or
more densely packed cells is entirely consistent with the
finding irz vitro; down-regulation of EGF binding is
thought to be one factor accounting for the arrest of cell
growth at high cell density [lo].
Given that the response of the kidney is to up-regulate
EGF receptors in surviving cells, the source of EGF
which binds to such receptors remains to be determined.
Although the mammalian kidney itself produces EGF, as
suggested by the existence of messenger RNA for the
EGF precursor molecule (prepro-EGF) in the kidney
[ 111, its immunohistochemical localization to the distal
nephron and, more specifically, to the luminal surface
[12], makes it difficult to envisage that this source provides EGF for the bulk of the renal mass, i.e. the proximal
tubular cells. Indeed, the fact that EGF receptors reside
on the basolateral surface of tubular epithelial cells [13]
makes it likely that it is EGF in the renal circulation that
normally interacts with these receptors. Since EGF production by the ischaemic kidney appears to decrease
rather than to increase [5],the up-regulated EGF receptor
would have to interact predominantly with EGF reaching
the kidney via the systemic circulation to elicit a mitogenic
response.
An important finding in the present study is that
EGF was able to accelerate the rate and extent of
recovery 7 days after ischaemic ATN. It should be noted
that the EGF infusion was started after the injury, so that
its effect was clearly to alter the rate of recovery rather
than to act protectively. Apart from the protective effect
of EGF on the healing of gastric and duodenal ulcers [14,
151, this is the first demonstration that a polypeptide
growth factor can alter the natural history of a disease of a
major organ system after the disease has been established.
The application of the present finding to human ATN
requires considerably more information. It should be
noted that EGF was delivered directly into the renal
arterial (aortic) circulation in the present study, a
manoeuvre which is not without risk in humans, especially
if the catheter is left in situ for a prolonged period. It is not
clear what the effects of increased circulating EGF levels
would be, nor is it apparent whether or not EGF would be
equally effective if given for different periods of time after
the onset of ATN. It is also not clear whether EGF would
only have a therapeutic role if actual tubular cell necrosis
has occurred. Since this apparently does not occur in
450
J. Norman et al.
many instances of ATN [4], EGF may have its greatest
therapeutic value in the severest forms of ATN which are
accompanied by cell death. These studies validate a closer
look at the role of potentiating cell growth in altering the
natural history of ATN.
Note added in proof (received 16 January 1990)
Since the submission of this paper, Humes et al. [16]
have described a similar effect of subcutaneously administered EGF on renal tubular cell regeneration in postischaemic acute renal failure.
ACKNOWLEDGMENT
This work was supported by grant R01 DK34049 from
the National Institutes of Health.
REFERENCES
'
1. Norman, J., Badie-Dezfooly, B., Nord, E.P., Kurtz, I.,
Schlosser, J., Chaudhari, A. & Fine, L.G. EGF-induced
mitogenesis in proximal tubular cells: potentiation by
angiotensin 11. Am. J. Physiol. 1987; 253, F299-309.
2. Harris, R.C., Hoover, R.L., Jacobson, M.R. & Badr, K.F.
Evidence for glomerular actions of epidermal growth factor
in the rat. J. Clin. Invest. 1988; 82, 1028-39.
3. Breyer, J. & Harris, R.C. Epidermal growth factor ( E G F )
binds to specific EGF receptors and stimulates mitogenesis
in renal medullary interstitial cells [Abstract]. Kidney Int.
1988; 33,255.
4. Molitoris, B.A., Hoilien, C.A., Dahl, R., Ahnen, D.J.,
Wilson, P.D. & Kim, J. Characterization of ischemiainduced loss of epithelial polarity. J. Membr. Biol. 1988;
106,233-42.
5. Safirstein, R., Zelent, A.Z. & Price, P.M. Reduced renal
prepro-epidermal growth factor mRNA and decreased EGF
excretion in A R E Kidney Int. 1989; 36.8 10-5.
6. Bacay,A., Mantyh, C.R.,Cohen,A.J., Mantyh, P.W. 6( Fine,
L.G. Glomerular atrial natriuretic factor receptors in
primary glomerulnephropathies: studies on human renal
biopsies. Am. J. Kidney Dis. 1989; 14,386-95.
7. Rainbow, T.C., Biegon, A. & Berck, D.J.Quantitative autoradiography with tritium-labelled ligands: comparison of
biochemicals and densitometric measurements. J. Neurosci.
Methods 1984; 11,231-41.
8. Shaw, S.G., Weidmann, P., Hodler, J., Zimmerman, A. &
Peternostro, A. Atrial naturietic peptide protects againat
ischemic renal failure in the rat. J . Clin. Invest. 1987; 80,
1232-7.
9. Fine, L.G. The biology of renal hypertrophy. Kidney Int.
1986; 29,6 19-34.
10. Holley. R.W. Control of growth of kidney epithelial cells. In:
Jimenez de Asua, L., Lcvi-Montalcini, R., Shields, R. &
lacobelli, S. eds. Control mechanisms in aninial cells. New
York: Raven Press, 1989: 15-26.
11. Niall, H.D., Coghlan, J.P., Rall L.B. et al. Mouse preproepidermal growth factor synthesis by the kidney and other
tissues. Nature (London) 1985; 313,228-3 I .
12. Fisher, D.A., Salido, E.C. & Barajas, L. Epidermal growth
factor and the kidney. Annu. Rev. Physiol. 1989; 51,67-80.
13. Maratos-Flier, E., Yang Kao, G.Y., Vcrdin, E.M. & King, G.
Receptor mediated vcctoral transcytosis of epidermal
growth factor by Madin-Darby canine kidney cells. J. Cell
Biol. 1987; 105, 1595-60 1.
14. Kontwek, S.J., Radecki, T., Brzozawski, T. et al. Gastric
cytoprotection by epidermal growth factor: role of endogenous prostaglandins and DNA synthesis. Gastroenterology 1981; 81,438-43.
15. Kirkegaard, P., Olsen, P.S., Poulsen, S.S. & Nexo, E.
Epidermal growth factor inhibits cysterine-induced
duodenal ulcers. Gastroenterology 1983; 85, 1277-83.
16. Humes, D.H., Cieslinski, D.A., Coimbra, T.M., Messana,
J.M. & Galvao, C. Epidermal growth factor enhances renal
tubular cell regeneration and repair and accelerates the
recovery of renal function in postischaemic acute renal
failure. J. Clin. Invest. 1989; 84, 1757-6 I.