| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Communications |
a The Haartman Institute, Division of Bacteriology and Immunology, University of Helsinki, Finland;
b Medical Policlinic, University of Munich, Munich, Germany;
c Department of Pathology, University of Vienna, Vienna, Austria;
d Department of Clinical Neurosciences, University of London, London, U.K.;
e University of Frankfurt, Frankfurt, Germany; and
f Department of Anatomy, University of Heidelberg, Heidelberg, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: proteinuria • congenital nephrotic syndrome • glomerular disease • differential display • cytochrome c oxidase
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Congenital nephrotic syndromes (CNS)2 are rare pediatric kidney diseases manifesting at or soon after birth with massive proteinuria. One of the best-characterized CNS is the congenital nephrotic syndrome of the Finnish type (CNF) with treatment-resistant proteinuria and characteristic podocyte changes in the glomeruli (3 , 4 ). Prenatal diagnosis of CNF can be made on the basis of extremely high alpha-fetoprotein in maternal serum or in amniotic fluid by the second trimester of pregnancy. The energy requirements and maturation of kidney specific mitochondrial oxidative phosphorylation complexes appear by the second trimester, whereas in other fetal tissues these are observed considerably later 5-7) . CNF used to be fatal during the first months of life, but current treatment of early bilateral nephrectomy, followed by dialysis and renal transplantation, is curative and provides a unique source of tissue material for search of the basic mechanisms involved. Kestilä et al. (8) recently established the defective gene NPHS1 in CNF patients. This gene, with as yet unknown functions, encodes a transmembrane protein with similarity to cell adhesion molecules.
To search for new molecular clues for glomerular dysfunction and pathophysiology of proteinuria, we used the differential display reverse transcription polymerase chain reaction (DDRT-PCR) method (9) , which permits the detection of genes, known and unknown, that are differentially expressed in control and experimental tissues. For this purpose, we took advantage of kidneys removed from CNF patients. One DDRT-PCR product was distinctly decreased in kidneys from CNF patients and could be identified as the mRNA for subunit I of cytochrome c oxidase (COX I). As terminal component of the mitochondrial respiratory chain, cytochrome c oxidase plays a crucial role in aerobic energy production.
Here we present data with molecular, biochemical, histochemical, immunohistochemical, functional, and morphologic evidence showing consistent mitochondrial respiratory chain abnormalities in kidneys of CNF patients. Furthermore, we found enhanced lipid peroxidation (LPO) in glomeruli from CNF. Defects in mitochondrial respiratory chain functions can lead to excessive reactive oxygen species and LPO (10) . Moreover, LPO has been implicated in the pathophysiology of glomerular proteinuria (11 , 12 ). We therefore postulate that altered mitochondrial function may lead to enhanced lipid peroxidation and proteinuria in the kidney.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The kidneys at nephrectomy were perfused with Ringer's buffer solution and glomeruli were rapidly isolated as described earlier (14) , resulting in over 98% purity of the glomerular preparations. These were immediately processed for RNA isolation (15) . Samples of cortical tissue were also processed for histochemistry, immunohistochemistry measurement of respiratory chain complex activities, and electron microscopy as described previously (15) .
Kidney biopsies from patients with membranous glomerulonephritis (N=3) or adult minimal change disease (N=3; age of all patients >18 years) were obtained from routine diagnostic biopsies. The biopsies were snap-frozen and stored in liquid nitrogen until used.
For normal controls, cadaver kidneys (N=2; age of donors 12 and 18 years) unsuitable for transplantation due to vascular anatomic reasons (Department of Surgery, University of Helsinki) or the normal poles of kidneys removed because of Wilms' tumor (N=2; age 3 and 5 years) were used. Other human tissues used included samples from extrarenal arteries from CNF and control kidneys.
RNA isolation and DDRT-PCR
Total glomerular RNA was isolated from the CNF and control
kidneys as described (15
, 16
). To process the
samples for DDRT-PCR analysis (all in duplicate), a method described
earlier was used (9
, 17
), following the
manufacturers' instructions (GenHunter, Boston, Mass.). The PCR
product analysis was performed as described earlier (17)
.
Cloning and sequencing
Differentially displayed bands in the CNF glomerular samples
were cut from the dried acrylamide gels, extracted, and finally
reamplified using the same primers and conditions as for the initial
DDRT-PCR. The PCR products were then gel-purified and cloned using the
TA cloning system (In Vitrogen, San Diego, Calif.) (18)
.
For each PCR product, three to six colonies from a single
transformation were prepared (17)
. Six clones were then
sequenced and screened for homology with database sequences using the
BLAST search algorithm via Internet at the National Center for
Biotechnology, Washington, D.C.
Northern blotting
Total RNA from five CNF and four normal human cortex samples
were prepared (15
, 17
) and used for
reconfirmation of the candidate DDRT-PCR clones in Northern blotting
(17)
. To control the total RNA content and lack of
degradation in the preparations, blots were rehybridized with a human
ß-actin probe. For autoradiography, the filters were exposed on Fuji
Bas IIIS Imaging Plates and the expression was quantified using a
PhosphorImager and accompanying MacBAS software (Fuji Photo Film Co,
Japan).
To reveal transcript levels of different COX subunits, we used cloned mtDNA and cDNA probes in Northern analysis. Plasmid pCOX2 contained the complete gene for COX II and 714 bp of the gene for COX I. Plasmid pCOX4 contained a 682 bp COX IV cDNA fragment (courtesy of Dr. M. I. Lomax) and plasmid pCOX6 contained a 450 bp COX VIb cDNA fragment.
Semiquantitative RT-PCR
Because Northern blotting and histological experiments
were not sensitive enough for glomerular detection of mitochondrial
complexes, we determined the transcript levels of some respiratory
chain complex subunits by semiquantitative RT-PCR. RNA samples from
isolated glomerular fractions were used as a starting material in this
analysis (17)
. Sequence-specific oligonucleotide primers
were designed for mitochondrially encoded COX I and cytochrome
b as well as for COX VIIa, which is encoded in the nucleus.
The semiquantiation is based on the serial dilutions of control plasmid
DNA and sample cDNAs in the linear range of amplification and on the
amount of housekeeping (ß-actin) amplification product
(17)
.
Histochemistry, immunohistochemistry, and visualization of lipid
peroxidation products
Histochemical methods for respiratory chain complex II
(succinate dehydrogenase, SDH) and IV (cytochrome c oxidase,
COX) were used (19)
. For this purpose, fresh cryostat
sections were incubated with their specific substrate solutions,
producing colored insoluble granular deposits at sites of enzymatic
activity.
For immunohistochemistry the tissue sections were fixed for 3 min in acetone at -20°C and incubated with a monoclonal antibody to COX I (clone 1D6-E1-A8) or to COX IV (10G8-D12-C12) 20-22) . Lipid peroxidation products were determined immunohistochemically using antibodies specific for malonyldialdehyde, as described previously (23; kindly provided by Dr. T Montine). For visualization, fluorescein isothiocyanate (FITC) -conjugated second antibodies were used. An Olympus Ox50 microscope with a filter system for FITC fluorescence was used for microscopy.
Quantification of mitochondrial respiratory chain complexes
The steady-state level activities of the mitochondrial
respiratory complexes I to V were analyzed by solubilizing 10 mg kidney
tissue of CNF patients (N=3) and controls (N=3), and separated in the
native state by blue native polyacrylamide gel electrophoresis
(BN-PAGE). Lanes from BN-PAGE were then processed by sodium dodecyl
sulfate (SDS) -PAGE in a second dimension and quantified by
densitometry, as described previously (24
,
25
). The staining intensities of each individual complex
were normalized to the intensity of complex III in each gel.
Electron microscopy
Cortical samples from four CNF patient kidneys and four
controls were fixed and processed as previously reported
(12
, 26
).
Statistics
Mitochondrial quantitations are expressed as mean
±SD. Statistics were performed with unpaired two-tailed
t tests. Data were obtained by direct counting from the
electron micrographs at x4500 magnification. Four to six such sections
were analyzed from each sample.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
Northern blotting and in situ hybridization
Because of the high degree of identity of B36 PCR product with COX
I, another mitochondrially encoded COX subunit (COX II) as well as two
nuclear encoded subunits (COX IV and VIb) of the cytochrome
c oxidase complex were analyzed. In all Northern blotting
experiments, RNA isolated from the whole cortex was used. Similar to
DDRT-PCR product B36, both COX I and COX II showed consistent
underexpression in CNF tissue (Fig. 2)
. In contrast, the levels of the
nuclearly encoded subunits IV and VIb did not differ from the levels of
controls (see Table 1
). To localize the mRNA expression, we performed
in situ hybridization with band B36, COX I, COX II, COX IV, and COX
VIb. A decrease of B36, COX I, and COX II based on the amount of
grains/cell was observed especially in the proximal and distal tubuli
in CNF cortical kidney as compared to the controls (data not shown).
Again, in situ hybridization of the nuclearly encoded COX IV and VIb
failed to show any difference in signal intensity between CNF and
control kidney samples.
Semiquantitative RT-PCR for respiratory chain subunit
Since reactivity in immunohistochemistry and in situ hybridization
in the glomeruli was too faint, possibly due to the overwhelming number
of mitochondria in tubules, we analyzed the transcript levels of COX
subunits directly in isolated glomeruli by semiquantitative RT-PCR
(Fig. 3
). Again, a 2.5x down-regulation of the mitochondrial (mt) encoded
COX I subunits (but not the nuclearly encoded subunits) was observed in
CNF glomeruli (Table 2
). To extend the analysis to other complexes of the respiratory
chain, we also performed RT-PCR for glomerular cytochrome b
(respiratory chain complex III encoded by the mitochondrial genome).
This showed an underexpression comparable to that of COX I in CNF
glomeruli (Fig. 3
, Table 2
).
|
|
Histochemistry and immunohistochemistry
To study the involvement of mitochondria in CNF at the functional
and protein levels, we first used histochemical analysis for the
respiratory complex II (SDH; Fig. 4a, b
) and complex IV (COX; Fig. 4c, d
). Similar to
in situ hybridization, the histochemical reaction product in the
glomeruli was weak in both control and CNF kidneys, precluding reliable
comparison. The tubular reactivity of both SDH and COX were remarkably
decreased in CNF compared with control kidney cortices (see Figs. 4
a–d). Immunohistochemistry with antibodies recognizing the
mt encoded COX I showed an intense, finely granular pattern in the
proximal and distal tubular cells, whereas glomerular reactivity was
diffuse and faint. A consistent decrease in staining was seen in tubuli
of CNF kidneys as compared to controls (Fig. 5a, b
). Similar immunohistochemical analysis for the nuclear
encoded COX IV showed intense reactivity in tubular cells, but no
difference was apparent between CNF and control tissues (Fig. 5c, d
).
|
|
Lipid peroxidation
As a secondary indication of mitochondrial dysfunction, we
analyzed local lipid peroxidation products within glomeruli. The
glomerular malonyl dialdehyde reactivity appeared finely granular,
especially within podocytes (Fig. 6A, C
) and along the glomerular basement membrane, whereas
negligible reactivity was seen in the control samples (Fig. 6B
).
|
Steady-state levels of the respiratory chain complexes
Steady-state levels of the mitochondrial respiratory chain
complexes were determined with the blue native gel electrophoresis
technique (Fig. 7
). In the CNF patient tissues, the normal ratios of the complexes
excluded a specific deficiency of complexes II–V. However, CNF tissue
showed a consistent decrease of all respiratory chain complexes from
10% (complex I) to 30% (complexes II–V) of controls (Fig. 7
,
Table 3
).
|
|
Electron microscopy
To find morphological evidence of mitochondrial involvement and to
examine whether the changes observed could be due to a decreased number
of mitochondria, we studied the EM morphology of CNF and control kidney
cortex. In the CNF samples, characteristic concentric
intramitochondrial bodies consisting mostly of double membranes could
be seen within otherwise normal appearing mitochondria (Fig. 8
). In the CNF samples studied (N=4), the morphologic changes
observed were present in 13.0–44.4% of all mitochondria (mean 26.9%,
SD 3.0, standard error 1.7; P <0.001 to
controls). In controls, 0–2% of such abnormal mitochondria (mean
0.7%, SD 0.0, standard error 0.0) was found. No
statistically significant differences were found in the amount of
mitochondria in CNF as compared to controls (51.7 ±6.6 in CNF, 61.5
±2.1 in controls, P<0.1).
|
Analysis of renal vasculature from CNF patients and of biopsies of
other human proteinuric diseases
To analyze mitochondrial involvement in other CNF tissues,
histochemistry and immunohistochemistry for nuclearly and
mitochondrially encoded COX subunits were performed in renal vessels.
Although consistent reactivity was seen, no differences in the staining
intensities could be detected between CNF samples and controls (data
not shown). Furthermore, kidney biopsies from patients with heavy
proteinuria due to minimal change glomerulonephritis and membranous
glomerulonephritis were analyzed with histochemical and
immunohistochemical methods for COX. No difference could be seen in the
level of glomerular or tubular reactivity for COX in these two groups
of proteinuric diseases (not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Kestilä et al. (8) recently reported of a gene defect associated with CNF. This gene, called nephrin, encodes a transmembrane protein with similarity to immunoglobulin-like cell adhesion molecules. Although the functions of nephrin remain to be established, it does not appear to encode known mitochondrial components. However, it is interesting that a mitochondrial enzyme proline oxidase shows remarkable similarity to nephrin as determined by comparing the respective amino acid sequences. Moreover, the immediate nephrin locus in chromosome 19q13.1 contains at least two nuclearly encoded subunits of mitochondrial respiratory chain enzymes (COX VIb and COX VIIa) as well as the major protective enzyme against lipid peroxidation, the phospholipid hydroperoxide glutathione peroxidase. Whether nephrin is involved in the regulation of these mitochondrial components remains to be studied. Männikkö et al. (27) have reported an apparent mitochondrial involvement in 2 of 30 CNF patients; another 2 patients suffer from impaired hearing, a finding often observed in mitochondrial diseases (10) . Additional studies have shown involvement of mitochondria also in other human and experimental glomerular diseases, especially in the puromycin nephrosis of the rat (H. Holthöfer et al., unpublished results). Thus, our finding of altered mitochondrial functions appears to be an important pathophysiologic feature in glomerular permeability changes in general.
Cytochrome c oxidase is a 13 subunit complex of the respiratory chain crucial for cellular ATP production (6) . Three of the COX subunits are encoded by the mitochondrial genome, whereas the remaining subunits are encoded by the nucleus and are needed for the fully functional complex (28 , 29 ). Some nuclear encoded subunits are expressed in a strictly developmental stage-, tissue-, and even cell type-specific manner obviously reflecting individual cellular energy needs (30 , 31 ). It is noteworthy that the kidney shows strong COX activity in histochemical staining earlier than most other tissues at the second trimester of pregnancy (32) , coinciding with the onset of glomerular filtration. These data suggest that the glomerular filtration process may be crucially dependent on tissue- and cell type-specific mitochondrial energy production.
In addition to demonstrating identity and down-regulation of mitochondrially encoded COX I subunit in CNF kidneys, we subsequently demonstrated involvement of other mitochondrially encoded respiratory chain components, while the respective nuclearly encoded respiratory chain components remained at control levels. This suggests a defective nuclear–mitochondrial interaction rather than a primary mitochondrial defect 33-35) . The maintenance of an optimal energy balance requires finely tuned participation of tens of nuclear encoded, tissue-, cell type-, and developmental stage-specific factors, which must be correctly targeted, recognized, processed, and assembled into the mitochondria in stoichiometric amounts (36) to become fully functional.
In any tissue, both normally and abnormally functioning mitochondria (heteroplasmic) may be found simultaneously in individual cells (10) . Furthermore, tissues with a high mitotic index are able to select against heteroplasmic cells, whereas in tissues with little mitotic activity the ratio of mutant to normal mitochondria increases (37) . The threshold value (above which a mitochondrial defect leads to clinical symptoms) appears to reflect the level of dependency of energy metabolism (10 , 37 , 38 ). In our study the number of mitochondria in CNF and controls did not differ, yet steady-state levels of all the respiratory chain complexes in CNF were decreased to a level of 10–30% of controls. In the central nervous system, neuron type-specific symptoms have been observed at a similar remnant mitochondrial activity in symptomatic disease (29 , 30 ). This was explained by the fact that terminally differentiated neurons cannot compensate functional over nonfunctional mitochondria due to lack of cell (and interrelated mitochondrial) proliferation (10 , 34 , 38 ). This could also apply to glomerular podocytes, which in general do not proliferate (39) . Mandel et al. (40) showed that a decline in cellular ATP levels leads to increased epithelial permeability in MDCK cells. Moreover, Gabai and Kabakov reported (41) that such a decrease of ATP content specifically disrupts the architecture of actin and intermediate filaments, which are considered especially important for the maintenance of the complex structure of podocytes (26) . Together, these findings could explain the common morphologic changes, including flattening and retraction of podocyte foot processes, seen in proteinuria and indicate that podocytes may be ill-prepared for challenges in ATP level maintenance.
Recently, a deficiency of mitochondrial complex I has been shown to result in excessive production of superoxide radicals (42) , which in turn causes excessive formation of hydroxyl radicals and aldehydic lipid peroxidation (43) . These are intriguing findings, as glomerular lipid peroxidation has been implicated as an important factor in the generation of proteinuria (44) . Mitochondrial diseases often show changes in the kidney as a part of multiorgan involvement (10 , 45 ). Massive proteinuria and proximal tubular damage have occasionally been reported in newborns as the presenting symptoms (45 , 46 ). There is convincing evidence of excess reactive oxygen species and lipid peroxidation products in glomeruli in various experimental models, including the puromycin nephrosis of the rat 47-49) . We also found (H. Holthöfer et al, unpublished results) that in the glomerular damage induced by puromycin, mitochondrial dysfunction develops before the onset of proteinuria, suggesting a causal relationship.
Based on these considerations and on our findings of decreased renal mitochondrial function and enhanced glomerular LPO in CNF kidneys, we propose the following hypothesis: In CNF, a change in nephrin gene leads to a kidney-specific defect in the mitochondrial respiratory chain functions. This leads to excessive reactive oxygen species generation and accumulation of local LPO adducts. This, in turn, causes glomerular proteinuria. This hypothesis can now be tested in additional studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Abbreviations: BN, blue native; CNS, congenital
nephrotic syndromes; CNF, congenital nephrotic syndrome of the Finnish
type; COX, cytochrome c oxidase; DDRT-PCR, differential
display reverse transcription polymerase chain reaction; FITC,
fluorescein isothiocyanate; LPO, lipid peroxidation; mt, mitochondrial;
PAGE, polyacrylamide gel electrophoresis; SDH, succinate dehydrogenase;
SDS, sodium dodecyl sulfate.
Received for publication June 15, 1998.
Revision received October 23, 1998.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Mayrhofer, S. Krieger, N. Huttary, M. W.-F. Chang, J. Grillari, G. Allmaier, and D. Kerjaschki Alterations in Fatty Acid Utilization and an Impaired Antioxidant Defense Mechanism Are Early Events in Podocyte Injury: A Proteomic Analysis Am. J. Pathol., April 1, 2009; 174(4): 1191 - 1202. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Diomedi-Camassei, S. Di Giandomenico, F. M. Santorelli, G. Caridi, F. Piemonte, G. Montini, G. M. Ghiggeri, L. Murer, L. Barisoni, A. Pastore, et al. COQ2 Nephropathy: A Newly Described Inherited Mitochondriopathy with Primary Renal Involvement J. Am. Soc. Nephrol., October 1, 2007; 18(10): 2773 - 2780. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. McKiernan, V. C. Tuen, K. Baldwin, J. Wanagat, A. Djamali, and J. M. Aiken Adult-onset calorie restriction delays the accumulation of mitochondrial enzyme abnormalities in aging rat kidney tubular epithelial cells Am J Physiol Renal Physiol, June 1, 2007; 292(6): F1751 - F1760. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Li Campian, X. Gao, M. Qian, and J. W. Eaton Cytochrome c Oxidase Activity and Oxygen Tolerance J. Biol. Chem., April 27, 2007; 282(17): 12430 - 12438. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ahola, E. Heikkila, E. Astrom, M. Inagaki, I. Izawa, H. Pavenstadt, D. Kerjaschki, and H. Holthofer A Novel Protein, Densin, Expressed by Glomerular Podocytes J. Am. Soc. Nephrol., July 1, 2003; 14(7): 1731 - 1737. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Yamagata, K. Muro, J. Usui, M. Hagiwara, H. Kai, Y. Arakawa, Y. Shimizu, C. Tomida, K. Hirayama, M. Kobayashi, et al. Mitochondrial DNA Mutations in Focal Segmental Glomerulosclerosis Lesions J. Am. Soc. Nephrol., July 1, 2002; 13(7): 1816 - 1823. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Holthofer, H. Ahola, M.-L. Solin, S. Wang, T. Palmen, P. Luimula, A. Miettinen, and D. Kerjaschki Nephrin Localizes at the Podocyte Filtration Slit Area and Is Characteristically Spliced in the Human Kidney Am. J. Pathol., November 1, 1999; 155(5): 1681 - 1687. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ahola, S.-X. Wang, P. Luimula, M.-L. Solin, L. B. Holzman, and H. Holthofer Cloning and Expression of the Rat Nephrin Homolog Am. J. Pathol., September 1, 1999; 155(3): 907 - 913. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
