| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Medicine (C.H., H.G., M.C.S., A.L.), University of Birmingham, Birmingham, B15 2TH, United Kingdom; Institute of Experimental Clinical Research (A.F.), Aarhus Kommunehospital, DK-8000 Aarhus C, Denmark; Lawson Research Institute (J.P., D.J.H.), St. Joseph’s Health Centre, University of Western Ontario, London, Ontario, Canada; and Department of Medicine (C.R.T.), St. Thomas’s Hospital, London, SE1 7EH, United Kingdom
Address all correspondence and requests for reprints to: Dr. A. Logan, Department of Medicine, University of Birmingham, Wolfson Research Laboratories, Queen Elizabeth Medical Centre, Edgbaston, Birmingham, B15 2TH. E-mail: a.logan{at}bham.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Due to their potential roles in inducing cellular differentiation and/or proliferation, growth factors have attracted much attention in diabetes research (5, 6). The transforming growth factors-ß (TGF-ß) are a family of homologous peptides that have clearly elucidated cellular effects, which include regulation of cell proliferation and differentiation and modulation of extracellular matrix metabolism (7). TGF-ß have been shown to increase synthesis of extracellular matrix proteins and simultaneously to block matrix degradation by decreasing the synthesis of proteases and increasing the levels of protease inhibitors (7). There are three isoforms of TGF-ß (termed TGF-ß1, TGF-ß2, and TGF-ß3) and three receptor isoforms (termed TGF-ßRI, TGF-ßRII, and TGF-ßRIII) that mediate their effects. While the renal expression of TGF-ß1 messenger RNA (mRNA) has been studied during the acute phase (up to 2 weeks) of experimental diabetes (8), little or nothing has been reported on the dynamic changes in the expression of all three isoforms of TGF-ß and their receptors that accompany the deposition of matrix molecules in acute and chronic stages of experimental diabetic nephropathy.
The streptozotocin (STZ) diabetic rat model has been widely used to study diabetic renal changes (9, 10, 11). STZ causes insulin deficiency in rats and the time course and morphological changes in the kidney closely resemble the progression of human disease (12), so that the changes up to 180 days of hyperglycemia in the STZ-treated rat represent changes seen in early incipient kidney disease. Another animal model of spontaneous insulin-dependent diabetes mellitus is the biobreeding (BB) rat, which is a genetically characterized animal model for studies investigating the pathogenesis and prevention of type I diabetes mellitus. The BB rat model was, therefore, also used in this study to determine whether the acute phase modulation of the TGF-ß axis seen in the STZ diabetic rats was the result of hyperglycemia and not just an acute toxic effect of the drug STZ. We used the techniques of immunocytochemistry, Western blotting and ribonuclease protection assay, to examine the response of the renal TGF-ß axis to hyperglycemia both in the STZ and BB rats. In both models of experimental diabetes, we found TGF-ß2 to be the most responsive isoform and TGF-ßRII the most responsive receptor within the diabetic kidney during the acute phase of disease when fibrogenesis is initiated.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Twelve female BB rats were obtained from the University of Massachusetts (Worcester, MA) at 65 days post partum, weighing approximately 140 g, and each animal was monitored daily for body weight and hyperglycemia. Control rats for this study were age matched Wistar rats that were consanguinous to the BB rat and kept under the same experimental conditions. All rats had unlimited access to chow and water. Hyperglycemia was investigated by daily measuring of urine glucose levels using dipsticks (Roche Molecular Biochemicals, Ontario, Canada) and diabetes classified as commencing when recording a value of over 14 mmol/liter. After 2 days of hyperglycemia, 6 animals were killed and a further 5 animals after 8 days. All animals remained untreated for hyperglycemia throughout this time. UAE and body weight were monitored throughout, and kidney weight was measured in each animal following sacrifice. In both studies a blood sample was taken from each rat following anaesthetic overdose (50 mg/kg, pentobarbitol ip), then the animals were rapidly dissected, the kidneys removed, weighed, snap frozen in liquid nitrogen and stored at -70 C until ready for analysis. For this part of the study, animal care was consistent with the guidelines of the Canadian Council for Animal Care, and all procedures were approved by the ethical committee of the University of Western Ontario.
Urinary albumin excretion (UAE) of STZ and BB diabetic
rats
The urinary albumin concentration in 24 h urine collections
was determined by a polyethyleneglycol RIA for the quantitation of rat
microalbuminuria using a rat albumin antibody as a standard. The method
was previously described by (9). The urine samples were stored at -20
C until the assay was performed. Rabbit antirat albumin antibody
RARa/Alb was purchased from Nordic Pharmaceuticals and Diagnostics
(Tilburg, The Netherlands). For standard and iodination, a
globulin-free rat albumin was obtained from Sigma (St.
Louis, MO). Intra and interassay coefficient of variation was less than
5% and 10%, respectively.
Antibodies
Antibodies and control peptides for the TGF-ß1, TGF-ß3,
TGF-ßRII, and TGF-ßRIII isoforms and receptors were purchased from
R & D Systems (Abingdon, Oxfordshire, UK). The
antibody against TGF-ß2 and its control peptides were a gift from
Cambridge Antibody Technology (Melbourn, Cambridgeshire, UK) and the
antibody and control peptide for TGF-ßRI was purchased from
Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). All
biotinylated secondary antibodies were purchased from Vector Laboratories, Inc. (Peterborough, Cambridgeshire, UK) and all
peroxidase conjugated secondary antibodies were purchased from
Santa Cruz Biotechnology, Inc.. The antibody against
procollagen-I C-propeptide was a gift from Zeneca Pharmaceuticals (Macclesfield, Cheshire, UK) and used for
Western blotting at a dilution of 1:10000. The antibodies against
TGF-ß1–3 all recognize the active forms of the molecules. For
immunocytochemistry, the TGF-ß1 antibody was used at a concentration
of 2 µg/ml, TGF-ß2 antibody was used at 1 µg/ml, TGF-ß3
antibody was used at 13.33 µg/ml, TGF-ßRI antibody was used at 2
µg/ml, TGF-ßRII antibody was used at 6.25 µg/ml, and TGF-ßRIII
antibody was used at 5 µg/ml. For Western blotting, the primary
antibody was used at a concentration of 2 µg/ml for TGF-ß1, 0.5
µg/ml for TGF-ß2 and 1 µg/ml for TGF-ßRII.
Probes
Specific TGF-ß1 complementary RNA probes were constructed from
a 985-kb TGF-ß1 complementary DNA (cDNA) template (gift from Dr. M.
Sporn, then at NIH, Bethesda, MD), representing the major coding
sequence of the TGF-ß1 precursor, and subcloned into pBluescript SK+
plasmid, which was cut to a 455-bp cDNA template. TGF-ß2 mouse cDNA
template (a gift from Dr .W. Vale, La Jolla, CA) was cloned from the
whole sequence of 1244 bp and cut to a 716 bp and a 502 bp cDNA, both
of which were subcloned into pBluescript KS+. The rat TGF-ßRII cDNA
fragment (a gift from Dr. W. Vale) was cut to 455 bp in length and
subcloned into psp73. It was linearized using XhoI to obtain
the antisense DNA and EcoRI for the sense DNA control. A
316-bp GAPDH (glyceraldehyde 3-phosphate-dehydrogenase) cDNA template
was used as an internal control (Ambion, Inc., Austin,
TX).
All enzymes used in probe generation and ribonuclease protection assays were obtained from Life Technologies, Inc. (Paisley, UK) or Promega Corp. (Southampton, UK). All other reagents not specified were analytical grade from Sigma, Ltd. (Poole, Dorset, UK).
Immunocytochemistry
Cryostat sections (10 µm thick) of frozen kidneys were cut on
a freezing microtome (Bright, Huntington, Cambridgeshire, UK), and
mounted on Superfrost1 Plus slides (Surgipath, St. Neots,
Cambridgeshire, UK). Immunoperoxidase staining was performed using the
ABC Vectastain Elite kit (Vector Laboratories, Inc., Peterborough, Cambridgeshire, UK) essentially using the
method previously described by Walter et al. (13). There
were three added steps: 1) the frozen sections were incubated in 4%
formaldehyde diluted in PBS for 30 min at room temperature; 2) the
re-hydrated sections were treated with 0.001% Pronase E
(Sigma) for exactly 4 min before peroxidase treatment; and
3) after the peroxidase treatment, the sections were incubated with 100
µl of avidin D and then incubated with 100 µl of biotin
(Vector Laboratories, Inc.) to inactivate endogenous
biotin before the incubation overnight with the primary antibody.
Specificity of staining was determined by either (1) replacing primary
antibody with PBS or (2) preabsorbing the primary antibodies with the
specific protein overnight at 4 C before use. These treatments all
abolished the antigen-specific staining.
Semiquantitation of immunoperoxidase staining
Immunoperoxidase staining of glomeruli was analyzed and given a
score of between 1 and 4, 1 being very little immunoreactivity and 4
being the strongest immunoreactivity noted. This analysis was performed
in a blind and randomized way for all TGF-ß isoforms and receptors on
at least 10 glomeruli per kidney section, on two kidney sections from
each animal and in four animals per group.
Protein extraction
Before snap freezing, the right kidney was split into cortex and
medulla. Cortex (100 mg) was homogenized in 1 ml of homogenization
buffer, pH 7.5, using a Teflon glass homogenizer and transferred to a
siliconized tube, and 2 mM of phenylmethylsulfonylfluoride
(PMSF, Sigma) protease inhibitor was added. Then each
sample was microfuged for 10 min at 3,000 x g at 4 C,
and the supernatant was kept as the crude cytoplasmic protein
fraction.
Protein quantification
The total amount of protein in the samples was measured. Samples
were diluted in PBS (Sigma); 800 µl of diluted sample
was placed in a cuvette (Sarstedt, Leicester, UK) and 200 µl of dye
reagent (Bio-Rad Laboratories, Inc. Hemel Hempstead, UK)
was added. Following incubation at room temperature for 10 min, the
absorbance at 595nm was read by a spectrophotometer (Lambda 2,
Perkin-Elmer Corp., Beaconsfield, UK). BSA
(Sigma) was used to prepare a standard curve, from which
total protein concentration of samples was determined.
Western blotting
Protein was extracted, and 100 µg of protein was diluted in
equal amounts of 2 x Laemelli’s buffer and heated at 95 C for 5
min. The protein samples were first run through a 4% acrylamide
stacking gel followed by either an 8% or 12% acrylamide running gel.
The gel was run on a Bio-Rad Laboratories, Inc. (Hemel
Hempstead, Hertfordshire, UK) vertical gel apparatus at 100 V (mini
gel) and 50 mA (large gel) and broad range markers (Bio-Rad Laboratories, Inc.), were also run with the samples. The
separated proteins on the SDS gel were electrophoretically transferred
onto an Immobilon-P PVDF transfer membrane (Millipore Corp., Bedford, UK).
After transfer, the membrane was washed in TBS and then incubated on a shaker in blocking solution at room temperature for 2 h to block nonspecific antibody binding sites. The membrane was then washed in TTBS and incubated with primary antibody and blocking solution overnight in a heat sealed bag at 4 C. The membrane was washed in TBS and incubated for 1 h at room temperature in a specific secondary antibody made up in TTBS to recognize the primary antibody. The membrane was then rinsed in TTBS and binding was visualized using equal volumes of luminol reagents (ECL Western Blotting Reagents, Amersham International, Little Chalfont, Buckinghamshire, UK), subsequently, the membrane was exposed to ECL film (Kodak, Amersham Pharmacia Biotech) in the dark to visualize the specific protein bands.
RNA extraction
Total RNA was extracted from the frozen diabetic and control rat
right kidneys using the RNAzol B method previously described by
(13).
RNA quantification
The concentration of RNA was quantified in 1 µl aliquots of
each sample by absorbency measurements at 260/280 nm in a UV/visual
spectrophotometer (Lambda 2, Perkin-Elmer Corp.,
Beaconsfield, Buckinghamshire, UK) as previously described by us
elsewhere (13).
Ribonuclease protection assay
Radiolabeled TGF-ß riboprobes were synthesized using the
following reaction: 1 x transcription buffer, 10 mM
DTT, 20 U RNasin, 2.5 mM ATP, GTP, UTP (Pharmacia & Upjohn, Milton Keynes, UK), 0.25 µg linearized DNA
template, 38 µCi 
32P-CTP (Amersham Pharmacia Biotech, Buckinghamshire, UK), 7.5U T3, T7 or SP6 RNA
polymerases along with markers, using the method previously described
by us elsewhere (13).
The glyceraldehyde 3-phosphate-dehydrogenase, (GAPDH), housekeeping gene is constitutively expressed in many tissues and has been reported to be a useful internal control. The level of expression of this protein does not change under normal or pathophysiologic conditions and as such the levels of GAPDH mRNA in the kidney mRNA samples should be constant. The RPA protocol described by us elsewhere (13) was followed using TGF-ß1, TGF-ß2, TGF-ßRII, and GAPDH riboprobes added to 20 µg of sample total RNA for hybridization. Controls were set up using 20 µg of yeast transfer RNA; one was digested with RNases, the other remained undigested.
The hybridized samples were loaded onto a 4%/8 M urea polyacrylamide gel and run at 500 V for approximately 1 h. Following electrophoresis, the gel was fixed using 10% gel fix solution and blotted to 3MM Whatman filter paper (BDH, Merck Ltd., Poole, UK). The gel was then dried using a vacuum gel dryer (Flowgen, Sittingbourne, Kent, UK) for 1 h at 80 C. Autoradiography was performed by placing the gel with Hyperfilm MP (Amersham Pharmacia Biotech) in an autoradiography cassette with an intensifying screen (X-ograph Ltd., Malmesbury, Wiltshire, UK) and storing for 3 days at -70 C.
Densitometric and statistical analysis
Autoradiographs were scanned into a Macintosh computer and
subjected to densitometric analysis using Deskscan II software
(Hewlett-Packard Co.). Densitometric analysis of the
TGF-ß/receptors/collagen-I C-Propeptide and GAPDH bands in each
sample in the RPA and the TGF-ß bands in the Western blots was
carried out using NIH image 1.55f software (shareware from NIH,
Bethesda, MD). The relative levels of target mRNA in each sample
by RPA analysis was regarded as the ratio between intensities of the
TGF-ß/R and the GAPDH bands on the autoradiograph. One-way ANOVA was
performed on the densitometry results obtained to find which increases
were statistically significant when compared with controls.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
.
|
.
|
).
|
but not presented
herein, are available for inspection on written request to the
corresponding author.
|
). Staining was semiquantified using
a point scoring method (Table 3
). Immunoreactivity of the TGF-ß1
isoform was localized to the cytoplasm and nuclei of most of the
glomerular cells in control rats (Fig. 2A). The cytoplasmic
immunoreactivity decreased slightly during the acute phase up to day 14
(Fig. 2C), then returned to control levels between days 30–180
(chronic phase; Fig. 2, D–F). In the surrounding kidney tubules, the
immunoreactivity of TGF-ß1 was seen in the cytoplasm and nuclei of
most cells in control rat kidneys (Fig. 2A). Upon onset of diabetes,
TGF-ß1 seemed to decrease in the cytoplasm during the acute phase, so
that residual immunoreactivity is seen in the stroma of the tubules at
day 14 (Fig. 2C); thereafter levels of immunoreactive TGF-ß1 increase
in the tubules during the chronic phase of diabetes (Fig. 2, D–F).
|
|
|
). This staining rapidly increased in
intensity in the acute phase (days 3 and 14, Fig. 3, B and C, and Table 3) and then declined back to control levels by day 90 (Fig. 3D). The
decrease in immunoreactivity continued so that immunoreactivity was
below that of controls by day 180 (Fig. 3F and Table 3). In the tubular
epithelial cells of control animals, the levels of TGF-ß2
immunoreactivity observed were low in the cytoplasm with some staining
apparent in the surrounding stroma. Occasional immunoreactive nuclei
were apparent in the tubules (Fig. 3A). Although the intensity of
staining did not change significantly, the immunoreactivity in the
cytoplasm of the tubular epithelia in the acute phase of diabetes
showed a polarized distribution with more numerous stained nuclei (day
3 and 14, Fig. 3, B and C). The pattern of tubular TGF-ß
immunoreactivity reverted to that seen in controls by day 180 (Fig. 3, D–F). Immunoreactivity of TGF-ß3 was localized to the cytoplasm and nuclei of most glomerular cells in control rat kidneys (not shown). This immunoreactivity was increased between 3 days and 14 days of hyperglycemia. By day 30, immunoreactivity had decreased back to control levels and then decreased further chronically. In the tubules of control rat kidneys, TGF-ß3 weak immunoreactivity was seen mainly in the cytoplasm and the stroma. Occasional immunopositive nuclei were seen. The immunoreactivity increased at all localizations chronically, with the TGF-ß3 immunoreactivity being polarized to the basolateral membranes and stroma where it remained until day 90 of hyperglycemia. By day 180, weak immunoreactivity returned to the cytoplasm of the tubular epithelia (not shown).
Immunoreactivity of TGF-ßRI is seen in the cytoplasm of most glomerular cells of the control kidneys. There was no obvious change in localization or intensity of staining acutely. However, chronically, TGF-ßRI immunoreactivity increased slightly to peak at day 90, but by day 180 immunoreactivity decreased again to lower levels than those seen in control kidneys (not shown). In the tubular epithelia of control kidneys, TGF-ßRI was localized to the apical cytoplasm of epithelia, with weak staining of the stroma (not shown). During the acute diabetic phase, TGF-ßRI became polarized to the apical and basolateral membranes of tubular cells. This pattern of localization was sustained throughout the chronic period of observation (not shown).
The TGF-ßRII immunoreactivity was localized to the cytoplasm and
nuclei of the glomerular cells from control rat kidneys (Fig. 4A). This immunoreactivity increased
markedly acutely (Fig. 4, B and C). However, by day 30 after the onset
of hyperglycemia, the immunoreactivity had decreased in this cellular
compartment (Fig. 4D), and by day 180 levels had returned back to those
in control kidneys (Fig. 4F). Tubular epithelia showed only weak
TGF-ßRII immunoreactivity localized to the basolateral cytoplasm and
surrounding stroma in the control kidneys (Fig. 4A). Acutely and
chronically, this immunoreactivity increased where it was localized at
the basolateral membranes and stroma of tubular epithelia, with weak
staining also in the cytoplasm (Fig. 4, B–F).
TGF-ßRIII immunoreactivity was strongly localized to the cytoplasm
and some nuclei of glomerular cells and weakly localized at the same
sites of tubular cells in control kidneys (not shown). The pattern of
staining remained relatively constant throughout the acute and chronic
time course of diabetes studied (see Table 3).
Semiquantitation using the point scoring method allowed these changes
to be summarized. Acutely, TGF-ß2 was seen to be the most responsive
ligand, immunoreactivity increasing markedly in the glomeruli at day
14, whereas TGF-ß1 immunoreactivity decreased at day 14 (Table 3).
TGF-ßRII proved to be the most responsive receptor with
immunoreactivity increasing dramatically in glomeruli at day 14 (Table 3). Groups of tubules were assessed for staining as in glomeruli.
Similar results were seen for the tubules as the glomeruli. TGF-ß2
being the most responsive ligand in the acute phase and TGF-ßRII
being the most responsive receptor (see Figs. 3 and 4).
Immunocytochemistry in BB rats
All of the immunohistochemistry data relating to the BB rat study,
summarized in Table 3 but not presented herein, are available for
inspection on written request to the corresponding author.
Very similar dynamics of the TGF-ß axis was seen in the BB rat
as observed in the STZ induced diabetic rat, taking into account the
different definitions of acute and chronic disease in the two models
(see later discussion). For example, at day 2 after onset of
hyperglycemia (acute phase), immunoreactive TGF-ß2 and TGF-ß3 were
increased and TGF-ß1 decreased in glomeruli and tubules (see Table 3). By day 8, immunoreactive TGF-ß1 and TGF-ß3 returned to control
levels and TGF-ß2 immunoreactivity decreased below control levels.
Immunoreactive TGF-ßRII also increased strongly at day 2, whereas
TGF-ßRI and RIII immunoreactivity remained at a constant levels
compared with controls. By day 8 of hyperglycemia, the immunoreactivity
of TGF-ßRI decreased whilst TGF-ßRII levels remained elevated (see
Table 3).
Western blotting in STZ diabetic rats
Immunoprecipitation and Western blotting was performed on kidney
cortex samples from three animals in each experimental group at each
time point. The immunoreactivity of TGF-ß1 showed a small decrease to
day 14, but this did not reach statistically significant levels (Fig. 5). TGF-ß1 levels then increased up to
day 90 where immunoreactivity was over three times greater
(P < 0.01) than control levels and then decreased back
to control levels by day 180 (Fig. 5). By contrast, immunoreactivity of
TGF-ß2 showed an increase by day 30 to twice that of control levels
(P < 0.05) and decreased back toward control levels by
day 180 (Fig. 6). TGF-ßRII
immunoreactivity also increased from control levels up to day 90, where
immunoreactivity was over three times that of controls
(P < 0.05 when compared with control levels; Fig. 7). The immunoreactivity of TGF-ßRII
then decreased back to control levels by day 180.
|
|
|
By ribonuclease protection assay, TGF-ß1 mRNA at day 3 was twice that
present in the control samples (P < 0.05) and levels
remained high at day 14, then decreased back to control levels by day
30 (see Fig. 8). By contrast, TGF-ß2
mRNA did not change in the acute phase of hyperglycemia (data not
shown). Similarly, no changes in the TGF-ßRII mRNA were seen by
ribonuclease protection assay within the hyperglycemia time course
examined when compared with the controls (data not shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There is now much evidence to suggest that TGF-ß plays a major role in stimulating the increase in fibrosis that occurs within the diabetic kidney. Thus, high glucose concentration in both cultured mesangial cells and proximal tubular cells leads to increased TGF-ß expression and matrix synthesis (14, 15). In addition, in vivo experiments have been reported on changes in the renal TGF-ß axis in various animal models of experimental diabetes (8, 16, 17, 18). However, previous studies have focused on changes in only one of the TGF-ß isoforms (TGF-ß1) and, in most studies, only the mRNA levels were examined (8, 16, 17). This provided us with the impetus to carry out a rigorous characterization of the dynamic changes in the whole TGF-ß system in the diabetic kidney during the onset of the disease. Hence, we examined the changes in all TGF-ß isoforms together with their receptors to fully characterize their responses in the STZ-induced diabetic rat kidney throughout a 6-month period. We also characterized the rate of fibrosis throughout this time scale to examine its co-ordination with changes in different elements of the TGF-ß system.
Like others (19, 20), we characterized fibrosis by Western blotting for the pro-collagen-I C-propeptide. The collagen propeptides are necessary for correct folding of the collagen protein. N- and C-terminal propeptides are removed by amino and carboxyl procollagen peptidases, and this occurs before collagen-I being laid down in the extracellular matrix. Hence, the presence of C-propeptide is a marker for the rate at which collagen-I is being laid down in the matrix and is, therefore, an adequate indication of when the diabetic kidney initiates fibrosis and when initiation of glomerular scarring is taking place. An antibody against procollagen-I C-propeptide allowed us to examine the rate of collagen-I accumulation during disease induction in the STZ diabetic rat kidney and to correlate this with TGF-ß isoform expression. The results showed that the rate of synthesis of procollagen-I was highest at day 3 and 14 (the acute phase) of diabetes rather than in the chronic phase, when the kidneys look to be the most fibrosed. Clearly, measuring the C-propeptide does not quantify the total amount of collagen-I in the kidney, only the periods of most turnover of procollagen-I to mature collagen-I. It is well known that matrix continues to accumulate during the chronic phase but at a lower rate, until the glomerulus is unable to function and kidney failure results (21). Our results illustrate initiation of the fibrogenic process, when the trophic orchestrators responsible are likely to be most active.
TGF-ß1, ß2, and ß3 isoforms and all three TGF-ß receptors were examined in this study by immunocytochemistry, to identify their cellular localization within the kidney during the development of diabetic kidney disease. The point scoring method allowed semiquantitation of isoform specific immunoreactivity in the kidney. We also used Western blotting to further analyze those components of the TGF-ß axis showing major changes by immunocytochemistry.
Using these techniques, of all the components of the TGF-ß axis studied the TGF-ß1 and ß2 isoforms and TGF-ßRII proved to be the most responsive elements within the kidney glomeruli. The results showed that, while levels of TGF-ß1 protein decreased in the kidney glomerulus in the acute phase of STZ-induced diabetic disease, increases in tubular TGF-ß1 protein were detected by immunohistochemistry. These increases were reflected in increased TGF-ß1 protein detected by Western blotting between days 30–90. TGF-ß1 mRNA showed a clear increase in the amount of message transcribed during the acute phase period, an observation that confirms previous reports of changes in the mRNA (8, 22, 23). This apparent paradox of protein vs. mRNA levels indicates that, in addition to the compartmentalization of the TGF-ß1 protein, there may be some posttranscriptional modifications or protein degradation occurring during disease progression in the remodeling glomerulus. Importantly, the ribonuclease protection assay does not measure the rate of transcription; it measures steady-state mRNA levels. It is evident that even though there is an apparent increase in TGF-ß1 mRNA steady-state levels during the acute phase, the actual rate of transcription could be decreased as the mRNA becomes more stable and has a longer half-life. In the kidney, many proteases are active during the remodeling process of the disease, and there could be much protein and, therefore, TGF-ß1 degradation occurring within the kidney at this time point. Any of these mechanisms of posttranscriptional regulation may be modulating apparent TGF-ß1 protein levels. Whatever the mechanism, it is of significance that the levels of TGF-ß1 protein detected in the glomeruli during the acute phase of diabetic nephropathy in the STZ-induced diabetic rat kidney do not correlate with the initiation of fibrosis, and this indicates that TGF-ß1 is not solely responsible for the promotion of glomerular fibrosis. Other factors, including other isoforms of TGF-ß, could be equally or more important fibrogenic factors in the acute diabetic kidney.
By contrast to TGF-ß1 protein, levels of immunoreactivity of the TGF-ß2 isoform were shown to increase significantly in glomeruli of the diabetic kidney during the acute phase (14–30 days) of disease progression. The spatio-temporal pattern of the TGF-ß2 increase corresponded exactly with that of pro-collagen-1C peptide. However, the rise in TGF-ß2 protein level was not reflected by changes in its mRNA level. The RNA was extracted from whole kidney cortex and not just glomeruli, so the increase in protein seen in the glomeruli at day 3 could be mirrored locally by the RNA, but the response could be diluted out in whole cortical samples. Alternatively, the stability of the mRNA may be altered. It is unlikely that the TGF-ß2 antibody is detecting the TGF-ß1 because it has no detectable cross-reactivity with this isoform (24). Importantly, TGF-ß2 protein up-regulation coincided with the initiation of fibrogenesis, as measured by the expression of procollagen-I C-propeptide. These results suggest that the TGF-ß2 isoform could be a key factor in the initiation of fibrosis that occurs within the diabetic kidney.
Levels of TGF-ßRII protein were also seen to increase markedly in the kidney during the acute phase of nephropathy, both by immunocytochemistry and Western blotting. TGF-ßRII is a necessary receptor for signaling the fibrogenic response to the nucleus, responsible for binding the ligand and recruiting TGF-ßRI in order for a signal to the nucleus to occur. It seems that TGF-ßRII is coordinately up-regulated with the TGF-ß2 isoform, suggesting that fibrogenic signaling to the nucleus is up-regulated, which could result in the increased rate of fibrosis observed in the diabetic kidney. However, like TGF-ß2, when assayed by ribonuclease protection assay, TGF-ßRII mRNA showed no significant change of expression within the diabetic kidney over the 180-day time course, a result that could reflect the sensitivity of the mRNA detection method.
The localization of TGF-ß3, TGF-ßRI, and TGF-ßRIII were also similarly examined by immunocytochemistry, and no significant changes in immunoreactivity were revealed throughout the 6-month time period in the STZ-induced diabetic rats. Therefore, they were not further analyzed quantitatively by ribonuclease protection assay or Western blotting. Notwithstanding their lack of responsiveness, these proteins could have a permissive role in kidney fibrosis.
Changes in immunoreactivity of all isoforms and receptors of the TGF-ß axis in BB rat kidney sections were seen to be similar to those of the STZ diabetic rat model. Hence, TGF-ß1 immunoreactivity decreased and TGF-ß2 immunoreactivity increased in glomeruli and tubules in the kidney at day 2 in BB rats. By day 8 in the BB rat, the TGF-ß2 immunoreactivity started to decrease, as it does in the STZ diabetic rat model, where, by day 30, the TGF-ß2 immunoreactivity had decreased back to control levels. The changes in immunoreactivity of TGF-ß2 in the tubules of BB rats also had a similar trend to those changes seen in the STZ diabetic rat model. In the BB diabetic rat model, the immunoreactivity of TGF-ßRII increased at day 2 and remained increased at day 8 in glomeruli and tubules. In summary, during the early time points of hyperglycemia, the TGF-ß receptor axis has a very similar pattern of both localization and intensity in the kidney of the BB rat model to that seen in the STZ diabetic rat model. In conclusion, the shorter term BB diabetic rat confirms those dynamic changes in expression and localization of TGF-ß isoforms and receptors seen in the STZ diabetic rat. Together with experiments performed by others that showed increased expression of TGF-ß for as long as 6 months after induction of STZ diabetes in rats, the evidence suggests that STZ does not alter the TGF-ß system via an acute toxic effect (25).
In summary, the major conclusion made from this study was that the rate of synthesis of collagen-I, one of the matrix molecules deposited in the fibrosing kidney, is highest during the acute phase of diabetes (day 3–14), which corresponds well with the increased levels of TGF-ß2 and TGF-ßRII protein but not with changes in TGF-ß1. This suggests that TGF-ß2 and TGF-ßRII may be key signaling elements stimulating kidney fibrosis, which could be potential targets for innovative therapies designed to limit fibrosis in the diabetic kidney.
|
Footnotes |
|---|
Received June 21, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J. Pannu, S. Nakerakanti, E. Smith, P. t. Dijke, and M. Trojanowska Transforming Growth Factor-beta Receptor Type I-dependent Fibrogenic Gene Program Is Mediated via Activation of Smad1 and ERK1/2 Pathways J. Biol. Chem., April 6, 2007; 282(14): 10405 - 10413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Sasser, J. C. Sullivan, J. L. Hobbs, T. Yamamoto, D. M. Pollock, P. K. Carmines, and J. S. Pollock Endothelin A Receptor Blockade Reduces Diabetic Renal Injury via an Anti-Inflammatory Mechanism J. Am. Soc. Nephrol., January 1, 2007; 18(1): 143 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Qi, X. Chen, J. Holian, E. Mreich, S. Twigg, R. E. Gilbert, and C. A. Pollock Transforming growth factor-beta1 differentially mediates fibronectin and inflammatory cytokine expression in kidney tubular cells Am J Physiol Renal Physiol, November 1, 2006; 291(5): F1070 - F1077. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Rich Genetics of Diabetes and Its Complications J. Am. Soc. Nephrol., February 1, 2006; 17(2): 353 - 360. [Full Text] [PDF]
|
|
|
|
 |
|
 B. F. Schrijvers, A. S. De Vriese, and A. Flyvbjerg From Hyperglycemia to Diabetic Kidney Disease: The Role of Metabolic, Hemodynamic, Intracellular Factors and Growth Factors/Cytokines Endocr. Rev., December 1, 2004; 25(6): 971 - 1010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Lane, J. Sun, K. Devish, and W. J. Langer Dissociation of renal TGF-{beta} and hypertrophy in female rats with diabetes mellitus Am J Physiol Renal Physiol, November 1, 2004; 287(5): F1011 - F1020. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Erman, S. Veksler, U. Gafter, G. Boner, C. Wittenberg, and D. J. van Dijk Renin-angiotensin system blockade prevents the increase in plasma transforming growth factor {beta}1, and reduces proteinuria and kidney hypertrophy in the streptozotocin-diabetic rat Journal of Renin-Angiotensin-Aldosterone System, September 1, 2004; 5(3): 146 - 151. [Abstract] [PDF]
|
|
|
|
 |
|
 M. E. Pueyo, M. Challah, D. Gauguier, L. Louedec, M. Philippe, R. Gaertner, M. Marre, J.-B. Michel, and M.-P. Jacob Transforming Growth Factor-{beta}1 Production Is Correlated With Genetically Determined ACE Expression in Congenic Rats: A Possible Link Between ACE Genotype and Diabetic Nephropathy Diabetes, April 1, 2004; 53(4): 1111 - 1118. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. N. Ziyadeh Mediators of Diabetic Renal Disease: The Case for TGF-{beta} as the Major Mediator J. Am. Soc. Nephrol., January 1, 2004; 15(90010): S55 - 57. [Abstract] [Full Text]
|
|
|
|
|
|
K. H.S. Wilson, S. E. Eckenrode, Q.-Z. Li, Q.-G. Ruan, P. Yang, J.-D. Shi, A. Davoodi-Semiromi, R. A. McIndoe, B. P. Croker, and J.-X. She Microarray Analysis of Gene Expression in the Kidneys of New- and Post-Onset Diabetic NOD Mice Diabetes, August 1, 2003; 52(8): 2151 - 2159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Mason and N. A. Wahab Extracellular Matrix Metabolism in Diabetic Nephropathy J. Am. Soc. Nephrol., May 1, 2003; 14(5): 1358 - 1373. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. T. Lee, H. Ha, M. Jung, H. Li, S. W. Hong, B. S. Cha, H. Chul Lee, and a. Y. Dong Cho Delayed Treatment with Lithospermate B Attenuates Experimental Diabetic Renal Injury J. Am. Soc. Nephrol., March 1, 2003; 14(3): 709 - 720. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Cheng and J. P. Grande Transforming Growth Factor-{beta} Signal Transduction and Progressive Renal Disease Experimental Biology and Medicine, December 1, 2002; 227(11): 943 - 956. [Abstract] [Full Text]
|
|
|
|
|
|
E. Gore-Hyer, D. Shegogue, M. Markiewicz, S. Lo, D. Hazen-Martin, E. L. Greene, G. Grotendorst, and M. Trojanowska TGF-beta and CTGF have overlapping and distinct fibrogenic effects on human renal cells Am J Physiol Renal Physiol, October 1, 2002; 283(4): F707 - F716. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||
