Nephrol Dial Transplant (1998) 13: 99-105
Nephrology
Dialysis
Transplantation
Original Article
Heat shock protein (HSP) expression and proliferation of tubular cells
in end stage renal disease with and without haemodialysis
Amit K. Dinda', Meer Mathur", Sandeep Guleria', Sanjeev Saxena.', Suresh C. Tiwari:' and Satish C.
Dash 3
'Department of Pathology, 2Department of Surgery and 3Department of Nephrology, All India Institute of Medical Sciences,
New Delhi 29, India
Abstract
Background. Prolonged dialysis is associated with
acquired cystic kidney disease (ACKD) and also higher
incidence of renal cell carcinoma. Relationship among
dialysis, tubular cell proliferation, development of
cystic change and neoplastic transformation is not
clearly known. Whether dialysis causes additional
stress on tubular cells is also conjectural. Study of heat
shock protein (HSP) expression which are rapidly
synthesized in cells in response to a variety of stresses
may be helpful in this regard.
Methods. To evaluate dialysis induced early changes
in end stage renal disease (ESRD), kidneys from eight
adult autopsied patients were examined (group I) who
were on weekly maintenance haemodialysis for 3-12
months. The heat shock protein (HSP 72/73) expression of tubular epithelial cells and their proliferating
cell nuclear antigen (PCNA) labelling index (LI) were
studied by immunohistochemistry using monoclonal
antibodies. For comparison similar study was carried
out in 10 cases of ESRD (Group II) of similar age
and sex distribution who were not dialysed. The
atrophic tubules were sub typed morphologically into
(1) classic, (2) thyroid, (3) endocrine and (4) super
tubules.
Results. In the dialysed group (I) the percentage of
hyperplastic super tubules (10.6±4.1%) was significantly higher than in the non-dialysed group (II)
(5.2 ± 1.3%) with a higher PCNA LI (6.8 ± 2.04%)
(group II 4.9± 1.9%) (P<O.Ol to <0.001). Though
grossly not detected, but microscopic cysts and microadenoma like areas were seen in all the cases in
group I with a mean diameter of 522.66 ± 315.25 urn
and 494.85 ± 262.46!lm respectively. They were seen
in one case of group II. PCNA LI of the cells in
microadenoma (7.2±3.1%) and microcysts (6.6±
2.6%) were similar to that of super tubules in group I.
Quantitation of HSP expression by image analysis
(optical density 2.309 ± 0.155) showed a positive corCorrespondence and offprint requests to: Dr A. K. Dinda MD, PhD,
Department of Pathology, All India Institute of Medical Sciences,
New Delhi 110 029, India.
relation (r=0.7555) (P<O.OOl) with PCNA LI in super
tubules indicating a higher induction in the dialysed
group.
Conclusions. This study suggests that haemodialysis
may cause injury to tubular cells and aggravate stress
on an already compromised situation of ESRD leading
to increased cell proliferation and more hyperplastic
supertubule formation which may be the forerunner
of cyst formation as well as neoplastic transformation.
Key words: Dialysis; ESRD; HSP expression; PCNA
labelling
Introduction
The relation of acquired cystic kidney disease (ACKD)
with haemodialysis and peritoneal dialysis is well
known [1,2]. A recent comprehensive review has shown
up to 50-fold increased risk of renal cell carcinoma
(RCC) in ACKD compared with the general population [3]. It is generally believed that development of
ACKD and renal cell neoplasm is a continuous process
with evolving phenotypic expression [3]. The pathogenetic mechanism which initiates the process is not clearly
known. A large number of pathogenetic pathways have
been proposed. However, their relative importance and
consequences need further evaluation [3].
Dialysis may result in introduction of various chemicals into the circulation, such as nitrite, nitrosamine,
formaldehyde and plasticizer (from dialysis tubing)
which may induce sublethal injury to the tubular cells
and induce their proliferation [4,5]. Recent studies
have documented that the kidney in end stage renal
disease (ESRD) is not a resting organ. It shows high
proliferative activity of the tubular epithelial cells
compared with normal kidney [6].
In response to stress due to various aetiologies, heat
shock proteins (HSPs) are rapidly synthesized and
expressed in the cells [7]. These HSPs are usually
classified according to their size, HSP20 to 30 kDa,
HSP60, HSP70, HSP90 [8]. These are highly conserved
© 1998 European Renal Association-European Dialysis and Transplant Association
A. K. Dinda et al.
100
proteins throughout evolution. Although many of their
functions are still speculative, several properties have
been clearly defined. They are involved in protein
folding and unfolding which may resist denaturation
of proteins in response to various stresses such as heat
[7,8], exposure to heavy metal [9] and hypoxia [10].
HSP may have an active role in transfer of antigenic
peptides during antigen presentation [11]. In normal
rat and human kidney localization of HSP73 has been
studied [12,13]. In rat kidney HSP73 was localized in
glomerular and Bowman's epithelia as well as epithelia
of renal tubules from proximal to collecting duct in
both nuclei and cytoplasm of the cells [12]. In normal
human kidney, HSP72/73 showed a uniform fine
granular cytoplasmic staining of visceral glomerular
epithelial cells and epithelia of distal convoluted
tubules and collecting ducts without localization in
proximal tubules [13]. In cases of tubular damage such
as acute interstitial nephritis there was a significant
increase in expression of HSP in tubular epithelium
[13]. It was suggested that an increase in synthesis and
cellular expression of HSP might indicate increased
cellular stress [14]. The present study was undertaken
on HSP expression and the proliferative activity of
tubular cells in kidneys of ESRD with and without
haemodialysis. Our aim was to investigate whether
haemodialysis might cause increased expression of HSP
in tubular epithelial cells as well as induce higher
proliferative activity. Considering the heterogeneity of
the morphological types of atrophic tubules in ESRD
[6], we studied HSP and PCNA expression in each
morphological subtype of tubules separately.
Subjects and methods
From the autopsy record of the past 15 years, 18 cases were
included in this study who were suffering from ESRD. Eight
patients were managed with weekly maintenance haemodialysis for periods of 3-12 months (group I). Ten patients did
not receive any dialysis (group II). In all these cases autopsies
were performed within 2 h of death. In five cases (two in
group I and three in group II), kidney tissues were also
obtained by blind needle biopsy immediately after death and
fixed in 10% buffered formalin. Ten kidney biopsies of adults
were taken from surgical pathology record as controls which
were reported as normal.
Histology
Following gross examination and weighing of the kidneys,
sections were taken from the upper pole, lower pole and
mid-part of each kidney, fixed in 10% buffered formalin and
routinely processed for histology. Sections were taken from
these three regions in all the kidneys to avoid any sampling
bias. Paraffin sections (5 11m thick) were stained with haematoxylin and eosin (H&E), periodic acid Schiff (PAS) and
silver methanamine (SM) stains for routine histological
evaluation. The atrophic tubules were classified into four
morphological types as described earlier [6]; (1) 'classic'
atrophic tubules with simple flat epithelium and thick wrinkled basement membrane, (2) 'thyroidized' tubules with uniform round shape and casts, (3) 'endocrine' tubules with
clear epithelial cells, narrow lumen and thin basement membrane, and (4) 'super' tubules which were large dilated
tubules with hyperplastic epithelial cells. In the majority of
cases, a minimum of three paraffin blocks were examined
from three different regions of each kidney as mentioned
earlier. In all, 1000-2000 atrophic tubules were evaluated
randomly in each case to determine the relative distribution
of the four morphological types. A minimum of 20 serial
sections were examined under light microscope from each
block of tissue to identify the focus of cystic change or
microadenoma formation.
Immunohistochemistry
Serial 5 11m thick paraffin sections were taken on poly-Llysine (Sigma Chemical Co, St Louis, MO, USA) coated
slides and numbered. Two consecutive slides were used for
immunostaining with monoclonal antibody against HSP
72/73 (clone W27) and proliferating cell nuclear antigen
(PCNA) (clone PClO). Both antibodies were obtained from
Boehringer Mannheim Biochemica (Mannheim, Germany).
Immunostaining was carried out using the avidin-biotin
conjugate (ABC) immunoperoxidase technique using diaminobenzidine tetrahydrochloride (DAB) as chromogen [15].
Antigen retrieval was performed on deparaffinized sections
by microwaving in citric acid buffer twice for 5 min [10,16].
Both the primary antibodies were used in 1: 100 dilution
with overnight incubation at 4°C. The ABC kit was procured
from Vector Laboratories (Burlingame, CA, USA). Each
batch of immunostaining was accompanied by a positive
control, negative control and a control with substitution of
primary with a non-related antibody of similar immunoglobulin type.
rc»A labelling index
(LI)
PCNA LI was determined by calculating the percentage of
positively stained cell nuclei out of the total number of cell
nuclei [15]. This index was determined separately for the
four atrophic tubule subtypes as well as in proximal and
distal tubules of normal control cases. A minimum of 200
cell nuclei were examined for estimation of LI.
HSP expression and image analysis
HSP expression was studied in consecutive serial sections in
the same group of tubules to compare with PCNA positivity.
The degree of HSP expression was assessed by measuring
the optical density (OD) of cytoplasmic brown staining
following reaction with DAB. It was measured with the help
of image analysis using a standard method [17]. The
image analysis was performed using a Xilix micro imager
(Xilix Technologies Corporation, Richmond, Canada),
F-64/occulus image grabber card (Coreco Inc, Quebec,
Canada) based on a pentium IBM personal computer
(Celebris XL, Digital Corporation, NY, USA) and Optimas
5.2 image analysis software (Optimas Corporation,
Washington, USA). Briefly, a representative image was captured from an HSP-immunostained slide. The mean OD of
the cytoplasmic area of the tubular cells was determined.
A minimum of 200 cells of each tubular type were analysed.
A negative control of the consecutive serial section with
antibody substitution was similarly analysed to assess nonspecific DAB staining. The essential prerequisites for a correct
OD measurement such as identical thickness of paraffin
HSP expression and proliferative activity in ESRD with and without dialysis
sections, identical performance of immunohistochemistry and
avoidance of any counter staining were carried out [17].
101
detected. The mean weight of kidneys in group I was
62 ± 30 g (52-105) and in group II 68 ± 25 g (48-110).
Statistical analysis
Histology
The statistical analysis was performed using Microstat software package (CA, USA) [18]. Values within the same group
were compared using the paired t-test and between groups I
and II by unpaired t-test. The Wilcoxon rank sum and signed
rank test were applied when required. Correlation regression
analysis was done with determination of r value to evaluate
the significance of a correlation between PCNA index and
OD of HSP expression in each type of tubule; P<O.05 was
considered to be significant.
Results
All the cases included in this study were adult
(Table 1). The age of eight patients in the dialysed
group (I) vaired from 40 to 56 (48 ± 8.4) years with a
male: female ratio of 3 : 1. The duration of illness varied
from 3 to 10 (6.8 ± 2.2) years. Previous biopsy records
were available in five cases. Three cases presented with
nephrotic syndrome diagnosed as membranous glomerulonephritis. Two cases presented with nephritic syndrome showed membranoproliferative and idiopathic
crescentic glomerulonephritis. The remaining three
cases presented initially with features of chronic renal
failure. These patients were on maintenance haemodialysis of 4 h, twice a week for periods of 3-12 months.
The age range of the 10 patients in non-dialysed
group (II) was 42 to 62 (50 ± 9.2) years with a male
female ratio of 4: 1 (Table 1). The duration of illness
varied from 4 to 12 (7.1 ± 3.2) years. Previous biopsy
records were available in four cases. Three cases presented with nephrotic syndrome and were diagnosed
as membranous glomerulonephritis (n=2) and membranoproliferative glomerulonephritis (n = 1). One case
presented as nephritic syndrome diagnosed as having
membranoproliferative glomerulonephritis. The other
six cases presented initially with features of chronic
renal failure. None of these cases in these two groups
had any clinical or serological evidence of collagen
vascular disease or diabetes.
Gross examination revealed typical small granular
contracted kidneys in all cases. There was no evidence
of hydronephrosis. On gross serial sectioning, no cystic
change or any abnormal space occupying lesion was
Table 1. Age, sex distribution and duration of illness of ESRD
patients with haemodialysis (group I» and without dialysis (Group
II). Values are given as mean ± SD with range shown in parentheses
Group
n
8
II
10
Age (years)
M:F
Duration of illness (years)
48 ± 8.4
(40-56)
50±9.2
(42-62)
3:1
6.8±2.2
(3-10)
7.1 ±3.2
(4-12)
4:1
On light microscopic examination, all four types of
atrophic tubules were seen in both the groups. The
distribution of classic, thyroid and endocrine type of
tubules was similar in the two groups with no significant difference (Table 2). Super tubules were significantly more frequent in group I (10.6±4.1%) than in
group II (5.2± 1.326) (P<O.OOI) (Table 2). These were
tubules with a larger diameter lined by proliferating
cuboidal to columnar cells. In focal areas they showed
double or multilayering of lining cells. In some cases,
these hyperplastic epithelial cells formed papillary
luminal projections (Figure 1); such changes were
noted more frequently in group I.
The micro adenoma and microcystic areas were noted
in all cases of group I and in one case of group II.
The micro adenoma areas varied from 120 to 922 urn
in diameter with a mean of 494.85 ± 262.46 urn. The
proliferating cells showed solid nests, aggregates, as
well as tubular structures with focal areas of dilatation
(Figure 2). Mitotic figures were noted in these cells.
These lesions were randomly distributed in renal parenchyma with a pushing margin. The microcystic areas
varied in diameter from 145 to 1125 urn with a mean
of 522.66 ± 315.25 urn. These were lined by single layer
of the columnar to cuboidal cells. These cysts were
noted separately as well as in close relation to the
microadenoma-like areas (Figure 2). Both micro adenomas and microcysts were noted in close proximity to
the super tubules. They were more frequent in areas
with a high concentration of super tubules.
peNALI
The PCNA LI in normal kidneys showed lower proliferation rate in proximal tubular cells (0.24±0.14%)
than distal tubules (0.36±0. 16%) (P<0.05). In ESRD,
the proliferative potential of cells in all types of
atrophic tubules in both the groups were significantly
higher than normal (P<O.OOI) (Table 2). Super
tubules had the highest PCNA LI. It was (6.8 ± 2.04%)
significantly higher than all other types in the dialysed
group (I) (P<0.05 to <0.001) (Table 2 and Figure 3).
In the non-dialysed group (II) the PCNA LI was
4.9 ± 1.96% higher than in the thyroid and endocrine
types (P<O.OOI) whereas it was not significantly
different from
the classic atrophic tubules
(3.05±2.40%) (P<0.08). Comparison ofPCNA LI of
different types of tubules between the two groups
showed a significantly high LI of only super tubules in
group I (P<0.05), whereas the difference was not
significant for other types (Table 2). The PCNA LI of
epithelial cells of microadenoma and microcystic areas
were 7.2 ± 3.12% and 6.6 ± 2.64% respectively
(P<0.08), which were similar to those of super tubules
in group I.
A. K. Dinda et al.
102
Table 2. The distribution of morphological types of atrophic tubules , their PCNA LI and HSP expression in dialysed (group I) and
non-dialysed (group II) cases
Tubules
Classic
Thyroid
Endocrine
Super
Distribution (%)
I
II
I
II
60.2± 14.6
65.6±20.5
3.45±2.25
3.05±2.40
2.086± 0.035'
2.032±0.22
21.3 ±5.5
20.8±4.6
1.58± 0.92
1.65± 1.02
2.072± 0.016'
2.026±0.014
7.9±2.8
9.8±3.9
1.62±0.81
1.48±0.92
2.063±0.015b
2.020±0.016
PCN LI (%)
HSP expression
(optical density)
I
II
1O.6±4.l c
5.2± 1.3
6.8±2.04b
4.9± 1.96
2.309±O.l55b
2.106±0.016
Significance of difference between groups I and II; 'P<0.05, bp<O.OI and cP<O.OOI.
Fig.!. Super tubules in dialysed group (I) showing dense cytoplasmic HSP positivity (arrows) . Note double and multilayering of
lining cells (arrow heads) as well as papillary luminal projection
(immunohistochemistry x 400).
Fig. 2. Photomicrograph showing a microadenom a (AD) with an
adjacent microcyst (CS) (haematoxylin and eosin x 120).
Fig.3. Immunohistochemistry for PCNA showing higher nuclear
labelling (arrow heads) in classic (C) and super (S) tubules with
negative nuclei in endocrine (E) tubule (x 1000).
tubules showed significantly higher OD than distal
tubules of normal kidney (P<0.05 to <0.001)
(Table 2) . In the non-dialysed group (II), the epithelia
of all four morphological types of atrophic tubules
showed similar HSP expression which was significantly
lower than the corresponding types of tubules in the
dialysed group (I) (P<0 .05 to <0.001) (Table 2). In
group I, super tubules showed a higher expression (OD
2.309±0.155) than other types (P<O .OOI). High
expression was also noted in cells of micro adenoma
(2.286±0.063) and microcysts (2.272±0.056) which
was similar to the super tubules in group I. Correlation
regression analysis between the PCNA LI and OD of
HSP expression in different types of tubules showed a
significant positive correlation (r=0 .7555) (P<O.OOl)
only in the super tubules of both the groups and not
in other types. Similar positive correlation (r=0.8l25)
(P < 0.001) was also noted in the epithelium of microadenoma and microcystic areas.
HSP expression
Under light microscopic examination in normal kidney,
the proximal tubules were negative for HSP whereas
the distal tubules were positive. With image analysis
the mean OD following HSP immunostaining of the
cytoplasm of proximal tubular cell was 1.035± 0.052
and that of the distal tubule was 2.006 ± 0.16 (P < 0.01).
In both groups I and II all the different types of
Discussion
A large number of studies are available in the literature
regarding the association of dialysis with the development of ACKD and renal cell carcinoma, although the
pathogenetic mechanism remains to be clearly defined
[1-5,19] . Considering various theories and hypotheses,
HSP expression and proliferative activity in ESRD with and without dialysis
it appears to be a multifactorial mechanism [3].
Relationships among tubular epithelial cell proliferation, development of cysts as well as adenoma or
carcinoma in patients undergoing dialysis treatment
needs further evaluation. The present post-mortem
study was devised to include eight patients (group I)
under haemodialysis treatment for a period ranging
from 3 months to 1 year, to detect dialysis induced
early renal changes. In these cases, no gross cyst or
tumour was identified. For comparison, 10 nondialysed cases of ESRD were taken (group II) having
a similar age and sex distribution as well as similar
duration of illness (Table 1). In ESRD the tubulointerstitial compartment of the kidney undergoes extensive
damage with tubular atrophy and loss as well as
development of atubular glomeruli [20]. However, the
morphological appearance of these atrophic tubules is
heterogeneous and they have been classified into three
subtypes; (l) 'classic' atrophic tubules, (2) tubules
showing thyroidization and (3) endocrine tubules [6].
Along with these atrophic tubules, a fourth subtype of
large tubules with hypertrophic or hyperplastic epithelial cells has been noted which is designated as 'super'
tubules [6]. These tubules are thought to develop as a
consequence of a compensatory growth mechanism
and obstruction [21]. A previous study on ESRD using
a panel of nephron-segment specific epithelial markers
showed that the classic atrophic type of tubules were
mostly derived from the proximal tubules [6]. The
thyroid and endocrine types of tubules were mostly
positive for distal tubular epithelial markers. The super
tubules showed varying degrees of expression of both
proximal and distal tubular markers [6].
Although no segment specific epithelial cell markers
were used in the present study, the tubules were
morphologically characterized into the four groups
using histochemical stains. The relative distribution of
all these different types of atrophic tubules was similar
in dialysed (1) and non-dialysed (II) groups apart
from the super tubules (Table 2). In group I, the
percentage of super tubules (10.6±4.1%) was significantly higher (P<O.OOI) than in group II (5.2± 1.3%).
Morphologically, these tubules showed hyperplasia of
the lining epithelial cells with focal areas of double or
multiple layering and occasional mitosis. Some of them
showed papillary projection in the lumen specially in
group I (Figure 1) .
These morphological features of epithelial hyperplasia of super tubules were reflected in higher proliferation as quantified by PCNA LI (Table 2) (Figure 3).
In both the groups these tubules showed the highest
PCNA LI which was significantly different from other
types (P<0.05 to <0.001). This finding was different
from the earlier report [6] which showed classic
atrophic tubules to have highest PCNA LI. This difference may be due to the extensive sampling done in the
present study. The super tubules of group I showed
significantly higher LI (6.8 ± 2.04%) than those of
group II (4.9±1.96%) (P<O.OI) indicating a higher
proliferation rate associated with dialysis. The PCNA
LI of the cells lining the micro cystic areas (6.6±2.64%)
103
and in microadenomas (7.2 ± 3.12%) were similar to
the super tubules of dialysed group. It is possible that
the morphological transformation to super tubules is
associated with a higher proliferation rate which in
turn may be associated with transepithelial secretion
of fluid resulting in cyst formation [19,22]. We
observed some super tubules with cystic dilatation as
well as papillary projection of proliferating epithelial
cells into the lumen (Figure 1). Close proximity of
super tubules, microcysts and microadenoma was
noted in the dialysed group (Figure 2). Cyst fluid from
ACKD cases has been shown to have mitogenic effect
on proximal tubular cells and renal carcinoma cells
[23]. It also contains a significantly higher concentration of epidermal growth factor than that of simple
cysts [24]. All the cases in the dialysed group (1)
showed microcyst as well as adenoma formation,
whereas only one out of 10 cases in the non-dialysed
group (II) had these changes.
To investigate whether haemodialysis causes additional stress to tubular epithelial cells in ESRD, we
studied HSP expression using an anti-HSP 72/73 monoclonal antibody. HSPs are a group of highly conserved
proteins developed during the evolution of species
which are expressed or induced by stress [7-14]. They
promote protein assembly and folding, elimination of
malfolded proteins and stabilization of newly synthesized proteins in various cellular compartments [8,25].
Among the HSP families, the HSP70 series has been
reported as the most abundant, participating in various
cellular functions, both under normal and stress-related
conditions [26]. A recent immunohistochemical study
of HSP 72/73 in normal human kidney and kidney
biopsies from cases of minimal change disease (MCD),
focal segmental glomerulosclerosis (FSGS), membranous glomerulonephritis (MGN) and chronic interstitial
nephritis (CIN) showed 2 + to 3 + positivity of cells
in distal convoluted tubules, cortical and medullary
collecting tubules, portion of the loop of Henle and
glomerular epithelium [13]. Increased staining of these
tubules was noted in acute interstitial nephritis (AIN)
and diffuse proliferative glomerulonephritis associated
with active interstitial inflammation [13]. In an earlier
study in 28 cases of human renal disease Dodd et al.
[27] showed positive staining in proximal tubules in
ischaemic transplant nephropathy and microPAN
(poly-anteritis nodosa). In the present study we mainly
observed cytoplasmic positivity in distal tubular epithelium of the normal kidney in line with Venkatasheshan
et al. [13] and quantified the cytoplasmic expression
of HSP by measuring the OD with the help of image
analysis system.
Recently, image analysis or image cytometry has
been used by several workers for selective quantitative
analysis of the intensity of immunohistochemical reactions (optical densitometry) which appears to be proportional to the amount of antigen present at the
reaction site [28-30]. Unlike the Western blot, this
technique does not give a direct quantification of the
antigen [31], but it is very helpful for studies with
reference to morphology [29,30] like the present one,
104
where the different subtypes of tubules are intermingled
with each other in the cortex.
In ESRD without dialysis (group II) all the different
types of tubules showed significantly higher OD than
the distal tubles of normal kidney (Table 2), (P<0.05
to <0.01). Among the different types of tubules the
super tubules showed highest OD (2.106±0.016) which
was significantly higher than other types (P<O.Ol to
<0.001) (Table 2). The complex multifactorial pathogenesis of ESRD [20] had been proposed to be associated with marked alteration of tubulointerstitial
microenvironment with ischaemic [32] and metabolic
[33] stress which might increase HSP expression in all
tubular cells in comparison with the normal kidney
[13]. The cause of increased HSP expression in super
tubules was not clear in our study. It is possible that
they represent the remaining hypertrophic, hyperfunctioning nephrons which bear additional functional
stress in an altered microenvironment [20,21].
In the dialysed group (I) all the types of tubules
had significantly higher HSP expression than the nondialysed group (II) (P<0.05 to <0.001), super tubules
showing the highest (2.309 ± 0.155) (Table 2). The
increased expression of HSP with haemodialysis might
be secondary to stress caused by dialysis in addition
to alteration ofESRD. Dialysis may cause introduction
of several potentially toxic exogenous chemicals into
the circulation such as plasticizers, formaldehyde,
nitrite and nitrosamine which may cause tubular injury
as well as inducing tubular cell proliferation [3-5]. We
observed a significant positive correlation between HSP
expression and the proliferation of tubular cells
(PCNA LI) in super tubules in both dialysed and nondialysed groups as well as in cells of microcysts and
microadenomas (P<O.OOl). The present study did not
reveal the mechanism of induction of HSP secondary
to haemodialysis, although it showed a definite increase
in HSP expression in all types of atrophic tubules
(Table 2). It is also well known that cells with a higher
rate of proliferation are more susceptible to an oncogenic hit leading to neoplastic transformation [34].
This may be one of the pathogenetic mechanisms
responsible for the higher incidence of renal cell
neoplasms associated with dialysis.
Thus it may be suggested from the present study
that haemodialysis causes more stress or injury to the
tubular cells superimposed on an already compromised
situation of ESRD, leading to a higher rate of tubular
cell proliferation associated with more hyperplastic
super tubule formation which may be the forerunner
of cyst formation as well as neoplastic transformation.
Acknowledgements. The financial support for the work was provided
by Indian Council of Medical Research. The authors are grateful to
Mr Charan Singh, Mohan Kumar, Kishore and Shikander for
technical assistance as well as Mr K. K. Arora for secretarial help.
References
1. Bommer J, Ritz E, Walldherr R, Mion C, Silngeneyer A.
Acquired mu1ticystic transformation of kidneys. Contrib Nephrol
1985; 48: 189-195
A. K. Dinda et al.
2. Thomson BJ, Jenkin DAS, Allan PL, Elton RA, Winney
RJ. Acquired cystic disease of the kidney in patients with end
stage chronic renal failure: a study of prevalence and etiology.
Nephrol Dial Transplant 1986; 1: 38-43
3. Truong LD, Krishnan B, Cao JTN, Barrios R, Suki WN. Renal
neoplasms in acquired cystic kidney disease. Am J Kid Dis 1995;
26: 1-12
4. Crocker JFS, Safe SH. An animal model of hemodialysisinduced polycystic kidney disease (PKD). Kidney Int 1984;
25: 183
5. Davin T, Cosio F, Kjellstrand CM. Association of cancer with
primary renal disease and/or uremia. In: Riselbach RE, Garnick
M (eds): Cancer and the Kidney. Lea and Febinger, Philadelphia:
1982; 857-868
6. Nadasdy T, Laszik Z, Blick RE, Johnson DL, Sillva FG.
Tubular atrophy in the end stage kidney: a lectin and immunohistochemical study. Hum Patho11994; 25: 22-28
7. Schlesinger MJ. Heat shock proteins: the search for functions.
J Cell Bioi 1986; 103: 321-325
8. Burdon RH. Heat shock and heat shock proteins. Biochem
J 1986; 240: 213-324
9. Bauman JW, Liu J, Klaassen CD. Production of metallot hionein
and heat shock proteins in response to metals. Fund Appl Toxicol
1993; 21: 15-22
10. Joannidis M, Cantley LG, Spokes K et al. Induction of heatshock proteins does not prevent renal tubular injury following
ischemia. Kidney Int 1995; 47: 1752-1759
II. Li Z, Srivastava PK. A critical contemplation on the role of
heat shock proteins in transfer of antigenic peptides during
antigen presentation. Behringer Inst Mitt 1994; 94: 37-47
12. Komatsuda A, Wakui H, Imai H, Nakamoto Y, Miura AB,
Itoh H. Renal localisation of the constitutive 73-kDa heat shock
protein in normal and PAN rats. Kidney Int 1992; 41: 1204-1212
13. Venkataseshan VS, Marquet E. Heat shock protein 72/73 in
normal and diseased kidneys. Nephron 1996; 73: 442-449
14. Polla BS, Kantengwa S. Heat shock proteins and inflammation.
Curr Top Microbiol Immunol1991; 167: 93-105
15. Banerjee M, Dinda AK, Sinha S, Sarkar C, Mathur M. C-myc
oncogene expression and cell proliferation in mixed oligoastrocytoma. Int J Cancer 1996; 65: 730-733
16. Shi SR, Key ME, Kalra KL. Antigen retrieval in formalin fixed
paraffin embedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of
tissue sections. J Histochem Cytochem 1991; 39: 741-748
17. ThaI DR, Horn M, Schlole W. Selective quantitative analysis of
immunohistochemical reaction. Acta Histochem 1995; 97:
203-211
18. Dinda AK, Saxena S, Guleria S et al. Diagnosis of glomerular
haematuria: role of dysmorphic red cell, G 1 cell and bright field
microscopy. Scand J Clin Lab Invest 1997; 57: 203-208
19. Grantham n. Acquired cystic kidney. Kidney Int 1991; 40:
143-152
20. Marcussen N. Atubular glomeruli and the structural basis for
chronic renal failure. Lab Invest 1992; 66: 265-284
21. Heptinstall RH. End stage renal disease. In: Heptinstall RH
(ed) Pathology of the Kidney. 4th edn. Little Brown, Boston,
1992; 713-777
22. Grantham n. Pathogenesis of renal cyst expansion: opportunities for therapy. Am J Kidney Dis 1994; 23: 210-218
23. Torrence RJ, Elbers JD, Seline P, Clayman RV. The effect of
renal cyst fluid from patients with acquired cystic the kidneys
(ACKD) upon growth of benign and malignant human renal
cells. J Uro11988; 139: 211A [Abstr]
24. Moskowitz OW, Bonar SL, Marcus MD, Clayman RV, Avner
ED. Epidermal growth factor (EGF)) content of human and
mouse (CPK) renal cysts. Am J Kidney Dis 1989; 14: 435-436
25. Becker J, Craig EA. Heat shock proteins as molecular chaperones. Eur J Biochem 1994; 219: 11-23
26. Minowada G, Welch WJ. Clinical implications of the stress
proteins. J Clin Invest 1995; 95: 3-12
27. Dodd SM, Martin JE, Swash M, Mather K. Expression of heat
shock protein epitopes in renal disease. Clin Nephrol 1993;
39: 239-244
HSP expression and proliferative activity in ESRD with and without dialysis
28. Cohen C. Image cytometric analysis in pathology. Hum Pathol
1996; 27: 482-493
29. Charpin C, DeVictor B, Andrae L et al. p53 quantitative
immunohistochemical analysis in breast carcinomas. Hum Pathol
1995; 26: 159-166
30. Layfield LJ, Saria EA, Conlon DH, Kerns BJ. Estrogen and
progesterone receptor status determined by Ventana ES 320
automated immunohistochemical stainer and CAS 200 image
analyser in 236 early-stage breast carcinomas: prognostic significance. J Surg Oneal 1996; 61: 177-184
31. Dhillon VB, McCallum S, Latchman DS, Isenberg DA.
105
Elevation of the 90 kDa heat shock protein in specific subsets
of systemic lupus erythematosus. Q J Med 1994; 87: 215-222
32. Cohen EP, Elliott WC Jr. The role of ischaemia in acquired
cystic kidney disease. Am J Kidney Dis 1990; 15: 55-60
33. Annotation. Acquired cystic disease of the kidney. Lancet 1977;
ii: 1063
34. Masuhara M, Ogasawara H, Katyal SL, Nakamura T,
Shinozuka H. Cyclosporin stimulates hepatocyte proliferation
and accelerates development of hepatocellular carcinoma in rats.
Carcinogenesis 1993; 14: 1579-1584
Received for publication: 14.10.96
Accepted in revised form: 5.9.97