ultrastructural study of human glomerular capillary loops with iga ... · lesions of human renal...

Summary. Immunoglobulin A (IgA) nephropathy showsgreat variability regarding the histological features of thelesions of human renal glomeruli. In the present study,the quick-freezing and deep-etching (QF-DE) methodwas used to analyze the glomerular ultrastructure ofbiopsied kidney tissues from children with IgAnephropathy. Biopsied renal tissues were routinelyprepared for light microscopy, immunofluorescencemicroscopy, conventional electron microscopy, andreplica electron microscopy. The three-dimensionalultrastructure of glomeruli of the kidney was clearlyobserved by using the QF-DE method. Three layers ofglomerular basement membranes, i.e., middle, inner andouter layers, were clearly detected in the replica electronmicrographs. The middle layer was 343.0±24.2 nm(n=20) in width and formed polygonal meshworkstructures. We also observed slit diaphragms, electron-dense mesangial deposits, and increased amounts ofmesangial matrix and foot process effacement. Manydelicate filaments were found to be distributed from theapical to the bottom portions between neighboring footprocesses. The ultrastructural difference between thereplica electron micrographs and conventional electronmicrographs was found to be especially marked in theappearance of foot processes and connecting filamentsbetween the neighboring foot processes. Theexamination of extracellular matrix changes, as revealedat high resolution by the QF-DE method, gave us somemorphofunctional information relevant to themechanism of proteinuria with IgA nephropathy.

Key words: IgA nephropathy, Slit diaphragm, Basementmembrane, Quick-freezing, Replica electron micrograph

Introduction

Immunoglobulin A (IgA) nephropathy is a kidneydisease which many children suffer from. Initial reportsabout IgA nephropathy described a predominantly focaland segmental type of glomerular lesions, and what wasat first considered to be a benign kidney disease (Berger,1969; Zimmerman and Burkholder, 1975). Later, otherresearches reported a great variety of glomerular lesionsand clinical courses (Bogenschutz et al., 1990; Okada etal., 1990). Although the pathogenesis and patho-physiology of IgA nephropathy have been investigatedfor the past few decades, they remain unclear.

The quick-freezing and deep-etching (QF-DE)method for electron microscopy has often been used toexamine the three-dimensional ultrastructure of cells andextracellular matrices in situ at high resolution (Heuserand Kirschner, 1980; Ohno, 1985; Ohno et al., 1996).This QF-DE method has also been used to examinesome renal disease models in experimental animals(Naramoto et al., 1991; Nakazawa et al., 1992; Duan etal., 1993; Moriya et al., 1993, 1996), normal humankidneys (Moriya et al., 1995), and patients with diabeticnephropathy (Tanaka et al., 1996). In the present study,we applied the QF-DE method to examining the renalglomerular ultrastructure of biopsied kidney specimensfrom patients with IgA nephropathy to get somemorphofunctional insight into the kidney disease.

Materials and methods

Four children with IgA nephropathy were finallydiagnosed by kidney biopsy at the Yamanashi University

Ultrastructural study of human glomerular capillary loops with IgA nephropathy using quick-freezing and deep-etching method

E. Sawanobori1, N. Terada2, Y. Fujii2, K. Higashida1, S. Nakazawa1 and S. Ohno2

Departments of 1Pediatrics and 2Anatomy and Molecular Histology,

Interdisciplinary Graduate School of Medicine and Engineering, University of Yamanashi, Chuo-shi, Yamanashi, Japan

Histol Histopathol (2008) 23: 297-307

Offprint requests to: Shinichi Ohno, M.D., Ph.D. , Professor andChairman, Department of Anatomy and Molecular Histology,Interdisciplinary Graduate School of Medicine and Engineering,University of Yamanashi, Chuo-shi, Yamanashi 409-3898, Japan. e-mail: [email protected]

http://www.hh.um.es

Histology andHistopathology

Cellular and Molecular Biology

Hospital. The diagnosis of IgA nephropathy washistologically based on the presence of IgA as thepredominant component of immunoglobulin deposits inthe glomerular mesangium. The clinical features of thepatients at the time of the biopsy are summarized inTable 1. Patients No.1 and No.2, were evaluated afterthey had received therapy for 2 years; the others had notreceived any therapy. All four patients showed positivereactions for hematuria with moderate proteinuria (>0.5g/day) at the beginning of the clinical courses. Nopatient had petechiae or arthralgia, and serum creatinineof all patients was normal throughout the clinical course.

They were all normotensive. The urinalysis ofpatients No.1 and No.2 turned out to be negative at thetime of the second biopsy after 2 years’ treatment.Informed consent was strictly obtained from all patientsunder the guidelines for clinical experiments of theUniversity of Yamanashi.

Biopsied renal tissues were divided into fourportions that were used for routine light microscopy,immunofluorescence microscopy, conventional electronmicroscopy, or the QF-DE method. For lightmicroscopy, one part of the specimen was fixed with10% formalin, dehydrated in a graded series of ethanoland routinely embedded in paraffin wax. Thin sectionswere cut at 3-4 mm thickness and subjected to periodicacid-Schiff (PAS) and Masson trichrome (MT) staining.For immunofluorescence microscopy, the second part ofthe specimen was snap-frozen in acetone-dry ice, cut at athickness of 3-4 µm in a cryostat and immunostainedwith primary antisera against human IgG, IgA, IgM,C3c, C4c, C1q, and fibrinogen (DACO, Denmark),followed by secondary fluorescein-labeled antibody. Theimmunostained sections were observed with a Nikonfluorescence microscope.

For conventional electron microscopy, the third partof the fresh biopsied tissues was routinely prefixed with2.5% glutaraldehyde in 0.1M phosphate buffer (PB),pH7.4, for 1h and then postfixed with 1.5% osmiumtetroxide in 0.1M PB for 1h. After routine dehydrationwith a graded series of ethanol, they were embedded inQuetol-812 (Nissin EM Co., Tokyo, Japan). Ultrathinsections were cut using an ultramicrotome with adiamond knife, doubly stained with uranyl acetate andlead citrate, and examined using electron microscopes(Hitachi H-600 and H-7500; Hitachi, Tokyo, Japan).

For the QF-DE method, the last part of the fresh

biopsied tissues was processed as described previously(Hora et al., 1990; Furukawa et al., 1991; Yu et al.,1998). Briefly, small tissue blocks were slightly fixedwith 2% paraformaldehyde (PF) in 0.1M PB for 20-30min and then cut into slices with razor blades. They werewashed with 0.1M PB for 15min to remove solublematrix components from the exposed tissue surface andpostfixed with 0.25% glutaraldehyde in 0.1M PB for30min. They were immersed in 10% methanol, and thenquickly frozen by a contact freezing method with coppermetal cooled in liquid nitrogen (-196°C) using a JFD-RFA freezing apparatus (JEOL, Tokyo, Japan). Thesurface areas of the frozen renal tissues were freeze-fractured with a scalpel in liquid nitrogen as reportedbefore (Hora et al., 1990; Furukawa et al., 1991; Yu etal., 1998). The samples were deeply etched under highvacuum conditions of 2-3x10-7 Torr at -95°C in an EikoFD-3AS freeze-etching device (Eiko Co., Ibaraki, Japan)for 20-30 min. After the deep-etching, they were rotary-shadowed with platinum up to 2 nm thickness at anangle of 30° and then carbon at an angle of 90°. Theprepared replica membranes were coated with collodionin amyl acetate and treated with household bleach todissolve tissue components. They were finally washedwith distilled water and mounted on Formvar-filmedcopper grids, and immersed in amyl acetate solution todissolve the collodion. Finally, they were observed usingelectron microscopes (Hitachi H-600 and H-7500).

Results

Histopathology of glomeruli with light microscopy

The histological findings obtained from each renalbiopsy are summarized in Table 2. Two patients, No.2and No.4, were diagnosed with focal segmentalproliferative glomerulonephritis. The other patients,No.1 and No.3, were diagnosed with diffuse mesangialproliferative glomerulonephritis with segmental lesions.Mesangial matrix areas were expanded due to increasedamounts of the matrix and mesangial cell proliferation(Fig. 1a-d). Such pathological changes were found to bevariable among the glomeruli of a biopsied specimenand also within each glomerulus. By immuno-fluorescence microscopy, all specimens werepredominantly immunopositive for IgA in the mesangialareas (Fig. 1e-h), though the IgA immunoreaction was

298

Replica electron micrograph of IgA nephropathy

Table 1. Clinical data at the time of kidney biopsy.

Patient Age (yo) Sex Duration Creatinine Total Protein Albumin Proteinuria Hematuria Treatment (months) (mg/dl) (g/dl) (g/dl) (g/day) Urinalysis (sediment) before biopsy

No.1 13 F 30 0.46 6.4 4.0 (-) (-) PSL+AZT+DP+WNo.2 17 M 72 0.93 6.9 4.3 (-) (-) DP+sai-rei-touNo.3 9 F 2 0.50 5.6 3.4 1g 20-50 rbc/hpf Not doneNo.4 13 M 2 0.67 6.8 4.1 0.7g 10-20 rbc/hpf Not done

PSL: prednisolone; AZT: azathioprine; DP: dipyridamole; W: warfarin.

299


Fig. 1. Light andimmunofluorescencemicrographs showingcharacteristics of IgAnephropathy. a. There ismoderate to severe mesangialproliferation with sclerosis aswell as adhesion to Bowman’scapsule at 9 o’clock in themicrograph (patient No.1). MT-staining. x 200. b. Mesangialcells are slightly proliferated(patient No.2). MT-staining. x 200. c. Moderate mesangialcell and matrix proliferation isseen at 12 o’clock in the image(patient No.3). PAS-staining. x 200. d. Mild mesangialproliferation is seen (patientNo.4). PAS-staining. x 200. e-h. Immunofluorescentreaction products of IgA inpatients, No.1-4. They aremostly localized in mesangialareas and also partially alongglomerular capillary walls.

300


Fig. 2. Replica electron micrographs demonstrating three relatively well-ordered layers of the glomerular basement membrane (GBM) (patient No. 4). a. Foot processes (F) closely cover the GBM, and cross-linking filaments (arrows) are seen between foot processes and middle layers (asterisk),corresponding to lamina rara externa in conventional ultrathin sections. The middle layer (asterisk) has fine polygonal meshworks. Other cross-linkingfilaments are also seen between endothelial cells (E) and the middle layer (arrowhead). Bar: 500 nm. b. In slit spaces, connecting filaments (arrows)are seen between neighboring foot processes (F). Bar: 500 nm. c. Filamentous meshworks appear to be loosened at subendothelial sites at otherglomerular capillary loops (arrows). The middle layer is thinner at these areas (asterisks). F: foot processes; E: endothelial cells; L: blood capillarylumens. Bar: 500 nm.

Table 2. Histological features of kidney biopsy.

Patient Mesangial proliferation Crescents Adhesions Segmental T-I change Immunofluorescence

Cell Matrix (%) (%) sclerosis(%) (%) IgA IgG IgM C3

No.1 Diffuse ++ ++ 2 19 7 <5 +++ + ++ +Fibrocellular Mesangial

No.2 Focal + + 0 0 4 <5 ++ - + +++Mesangial

No.3 Diffuse +++ + 5 5 0 - +++ + + +++Cellular Mesangial/capillary

No.4 Focal ++ + 0 0 0 <5 +++ + ++ +++Mesangial

T-I change: tubulo-interstitial change.

also partially positive along the glomerular walls inpatient No.3 (Fig. 1g). C1q and C4c were weaklyimmunopositive in mesangial areas in patients No.1 andNo.2. Fibrinogen was immunopositive along thecapillary walls and crescent in patient No.3.

Replica electron micrographs of glomeruli in children withIgA nephropathy

Next we examined the following ultrastructural

features of the glomeruli which were obtained frombiopsied kidney samples of patients with IgAnephropathy: i) glomerular basement membranes, ii) slitdiaphragms, iii) mesangial areas, iv) endothelial cells inglomerular capillaries, and v) podocytes.

Glomerular basement membrane (GBM)

In the areas where GBM and foot processes werewell preserved in some glomeruli, the GBM consisted of

301


Fig. 3. Replica electron micrographs demonstrating structures of slit spaces and slit diaphragms between foot processes (patient No. 3). a. Slit spaces(arrows) and slit diaphragms (arrowhead), which were freeze-fractured almost in parallel to the GBM, show filamentous or sheet-like substructures,respectively. Some primary or secondary processes of podocytes (P) contain many intermediate filaments, and foot processes contain compactnetworks (asterisks). GBM: glomerular basement membrane. Bar: 200 nm. b. Glomerular capillary loops were freeze-fractured at an oblique angle tothe GBM on the left side and also in parallel to the GBM on the right side. There are narrow slit spaces and slit diaphragms (arrows) between footprocesses (F). The GBM shows tiny meshworks in the middle layer (asterisks). Some cytoplasmic sides of cell membranes are decorated withfilamentous structures (arrowheads). E: endothelial cells; L: capillary lumens. Bar: 500 nm.

middle, inner and outer layers in the replica micrograph(Fig. 2a, patient No.4), corresponding to lamina densa,lamina rara interna and lamina rara externa, respectively,as reported before (Takami et al., 1991). The middlelayer, which was 343.0±24.2 nm (n=20) in width,formed polygonal meshwork structures. In the outerlayer, the cross-linking filaments perpendicularlyconnected epithelial cell membranes with meshworks ofthe middle layer. The similar filaments of the inner layeralso connected endothelial cell membranes with themeshworks. The total width of the three layers was452.4±39.3 nm (n=20). Other connecting filaments werealso seen between neighboring foot processes (Fig.2b),and between foot processes and primary processes ofpodocytes (Fig.2c). Filamentous meshworks appeared tobe loosened at subendothelial sites, where the middlelayer became thinner (Fig.2c).

Slit diaphragms and their relationship to GBM

When slit diaphragms between foot processes werefreeze-fractured almost in parallel to the outer layersurface of GBM, they showed filamentous or sheet-likesubstructures, not zipper-like ones (Fig.3a, patient No.3).

In the areas which were freeze-fractured at an obliqueangle to the GBM, they also showed cross-linkingfilamentous substructures. The mean value of the widthof the slit diaphragms between cell membranes of thefoot processes was about 39 nm (n=20). Some primaryprocesses of podocytes contained many intermediatefilaments, and other foot processes contained thinfilament networks (Fig. 3a). In addition, at horizontallyor obliquely freeze-fractured GBM (Fig. 3b), meshworksin the middle layer of the GBM were more widely seento be localized between cell membranes of footprocesses and endothelial cells, and some cytoplasmicsides of podocytes were decorated with filamentousstructures (Fig. 3b, arrowheads).

Mesangial areas

Replica electron micrographs also showed mesangialareas of glomeruli in three-dimensional images (Fig. 4,patient No.1). The mesangial matrix had proliferated toform large areas of filamentous meshworks betweenmesangial cells. Some soluble materials were associatedwith basic filamentous structures, resulting in theformation of more compact meshworks. The

302


Fig. 4. Replica electron micrograph showing mesangial areas (patient No. 1). The mesangial matrix (MM) is proliferated to form large areas ofmeshworks between mesangial cytoplasmic processes (MC). Large amounts of materials are associated with basic filamentous structures, resulting inthe formation of compact meshworks. Bar: 1 µm. Inset; corresponding area in a conventional electron micrograph obtained from ultrathin sections. Bar: 500 nm.

303


Fig. 5. Replica electron micrographs showing luminal sides of endothelial cells (E) in glomerular capillary loops (patient No. 1). a. Endothelial fenestrae(arrows) are seen to be heterogeneously scattered along the blood capillary wall. Bar: 1 µm. Inset; higher magnification of the area. Arrows: Freeze-fractured fenestrae; Arrowheads: Invaginating pits; L: lumen; GBM: glomerular basement membrane. Bar: 500 nm. b. Slit diaphragms (arrows), whichwere freeze-fractured almost in parallel to the GBM, show sheet-like substructures between foot processes (F). The middle layer (asterisks),corresponding to the conventional lamina densa, forms fine meshworks. Some cross-linking filaments (arrowheads) connect the middle layer (asterisks)with endothelial cells (E). OL: Outer layer between horizontally freeze-fractured foot processes (F). GBM: glomerular basement membrane. Bar: 500 nm.

304


Fig. 6. Replica electron micrograph of freeze-fractured cell body of podocyte (P) and foot processes (F) or primary processes (Pp) (patient No. 2). Manythick intermediate filaments are seen to be localized in the primary processes (Pp). Other thin filaments are localized in the foot processes (F). S: cellmembrane surface of the podocyte; U: urinary space. Bars: 500 nm. Inset: higher magnification of the cell body of the podocyte. Thinner filaments(arrows) are bridging between the intermediate filaments.

conventional electron micrograph showed thecorresponding area, which was obtained in ultrathinsections (Fig. 4, inset).

Endothelial cells in glomerular capillary loops

Luminal sides of endothelial cells in glomerularcapillary loops were examined in the replica electronmicrographs (Fig. 5a, patient No.1). The number ofendothelial fenestrae decreased and they wereheterogeneously arranged. In this patient, the slitdiaphragms, which were freeze-fractured almost inparallel to the GBM (Fig. 5b), also showed sheet-likesubstructures between the foot processes. The middlelayer, corresponding to the conventional lamina densa,formed fine meshworks. Some cross-linking filamentsconnected the middle layer with cell membranes ofendothelial cells.

Podocytes and their cytoskeletons

Replica electron micrographs showed primaryprocesses of freeze-fractured podocytes (Fig. 6, patientNo.2). Many intermediate filaments were localized inthe primary processes, where they might have acted tostabilize the structures. In addition, extending secondaryprocesses contained not only intermediate filaments, butalso thin filaments. Many such filaments were localizedin the foot processes, where they were thought tofunction in dynamic movement.

Discussion

The QF-DE method is a useful electron microscopictechnique that permits three-dimensional views of cellsand tissues, including cell organelles, cytoskeletalelements and extracellular components. It has previouslybeen used to examine some renal diseases ofexperimental animal models (Naramoto et al., 1991;Nakazawa et al., 1992), but its application to humanrenal diseases is still limited because of the difficulty andcomplexity of human tissue preparation. In the presentstudy, we investigated IgA nephropathy of children byusing the QF-DE method, and observed three well-ordered layers of GBM, which were composed mainly ofmeshwork filaments (Takami et al., 1991). The totalwidth of the GBM was 452.4±39.3nm, which was widerthan that of 300-400nm obtained by conventionalelectron microscopy in normal adults (Tiebosch et al.,1989). This difference of width was probably caused bythe use of the new QF-DE method of tissue preparation,which characteristically preserves real tissuearchitecture.

The middle layer consisted of polygonal meshworkstructures. The GBM was reported to be composed oftype IV collagen, laminin, entactin/nidogen andproteoglycan (Kleppel et al., 1989; Timpl, 1989), andtype IV collagen is a major skeletal component of GBM.The meshwork structure of the GBM might contribute to

the construction of blood capillary tufts and the filtrationbarrier against large soluble materials. In human geneticdiseases of type IV collagen, such as Alport syndromeand thin basement disease, hematuria is an initial clinicalmanifestation (Kalluri et al., 1997; Ueda et al., 2002).The disruption of GBM causes pathological leakage oferythrocytes. Controversy about the major structure ofthe glomerular permeability barrier between GBM andthe slit diaphragm has continued for a long time. Earlystudies with dextran and ferritin suggested that the GBMwas a major filtration barrier against plasma proteins(Farquhar, 1961; Caulfield and Farquhar, 1974).Recently, some studies of genetic renal diseases haveindicated that the slit diaphragm is probably moreimportant than the GBM in protein sieving (Kestila etal., 1998; Shih et al., 1999; Boute et al., 2000; Donovielet al., 2001; Ciani et al., 2003). The discovery andmutation analysis of several podocyte proteins, includingnephrin (Kestila et al., 1998), Neph1 (Donoviel et al.,2001), podocin (Boute et al., 2000), Fat (Ciani et al.,2003), and CD2AP (Shih et al., 1999), have indicatedthe importance of the slit diaphragm in the glomerularpermeability barrier, and shown that defects in anyprotein components of the slit diaphragm result in thenephritic syndrome. Other mathematical studies withstructure-based models showed that the size selectivityof the glomerular capillary was largely determined bythe cellular layers, rather than the GBM (Deen et al.,2001). They also suggested that the main roles of theGBM were the maintenance of a highly regulated barrierfunction only for water, small molecular solutes/ions,and smaller plasma proteins.

In 1974 and 1975, the glomerular slit diaphragm innormal rats, mice and humans was reported to consist ofzipper-like substructures with periodic cross-bridgesextending from foot processes to a central filament(Rodewald and Karnovsky, 1974; Schneeberger et al.,1975). The spacing of the cross-bridges was defined as auniform population of rectangular pores, approximately4.0x14.0 nm in size. However, the prepared specimensin those studies were fixed with a conventional solutionof tannic acid, glutaraldehyde and osmium tetroxide(TGO). We have already reported that the so-calledzipper-like substructure was due to a technical problemduring the fixation and dehydration procedures, whichwas easily caused by shrinkage during the TGO fixationof rat glomeruli (Hora et al., 1990; Ohno et al., 1992). Inthe present study, we found slit diaphragms between twoadjacent foot processes by using the QF-DE method,partially as a sheet-like substructure. The width of theslit diaphragm had a mean value of 39 nm. A recentadvanced structural study by electron tomographysupports our contention that the zipper-like substructureis not a real one in vivo (Wartiovaara et al., 2004). It wasreported that the slit diaphragm was either a zipper typestructure or a ladder structure with more tortuous pores.However, in that study, their value of the width of humanslit diaphragms was 38±2 nm by the immuno-cryosection method, which is wider than that of 32±2 nm

305


obtained by the resin-embedding section. Moreover,their figures obtained using the high-pressure freezingand freeze-substitution method look like sheet-likesubstructures because of their smaller pores, rather thanzipper or ladder shapes (Wartiovaara et al., 2004).

As for the primary processes of podocytes, 10-nmintermediate filaments were localized in most of thecentral cytoplasm, whereas 7~9-nm actin microfilamentswere more predominant in foot processes (Andrews,1981; Vasmant et al., 1984; Pavenstädt et al., 2003). Thefoot processes were firmly attached to the slit membraneand GBM by matrix filaments. We also found manydelicate cross-linking filaments between neighboringfoot processes from the apical to the bottom portions.These cross-linking filaments, which are not visible withconventional fixation, may contribute to the podocytes`sroles in protein filtration barriers and attachmentbetween foot processes.

An increase of mesangial matrices was observed asan expansion of newly formed meshworks. Theseincreased matrices were suggested to be composed oftype IV collagen, laminin, fibronectin, perlecan andtenasin (Timpl, 1989; Miner, 1999), which exist innormal mesangium, with altered distributions accordingto mesangial proliferation and sclerosis (Oomura et al.,1989; Schnaper et al., 2003). In some reports, focal andsegmental accumulation of type I, III, or V collagen wasclosely related to the pathological development ofglomerular sclerosis (van den Born et al., 1993; Yang etal., 2001). The patients in the present study were young,and their duration of IgA nephropathy was not so long.In the light microscopic examination, only smallamounts of sclerosis and fibrosis were found in theirbiopsied kidneys. In some glomerular areas (Fig. 4),huge homogeneous filaments were widely distributed,indicating that some growth factor such as PDGF orTGF-ß might locally regulate the mesangial matrixproduction (Schocklmann et al., 1999; Yang et al.,2001).

To clarify the variable histological changes in IgAnephropathy, we examined human biopsied kidneys byelectron microscopy. The QF-DE method enabled us todetermine the three-dimensional ultrastructure ofmeshwork matrices in detail, and to detect somepathological changes and cell-cell interactions in IgAnephropathy. In addition, we are planning to performreplica-immunostaining of IgA and various slitdiapharagm-associated proteins in the replicamembranes to clarify the localization of immune-complex formation and changes of component proteinsin further studies. The detection of extracellular matrixchanges, as seen by the QF-DE method, will provideuseful information for clarifying the mechanism of IgAnephropathy.

Acknowledgements. The authors thank Miss Y. Kato in the Departmentof Anatomy, Interdisciplinary Graduate School of Medicine andEngineering, University of Yamanashi, for her technical assistance inthe present study.

References

Andrews P.M. (1981). Investigation of cytoplasmic contractile andcytoskeletal elements in the kidney glomerulus. Kidney Int. 20, 549-562.

Berger J. (1969). IgA glomerular deposits in renal disease. Transplant.Proc. 1, 939-944.

Bogenschütz O., Bohle A., Batz C., Wehrmann M., Pressler H.,Kendziorra H. and Gärtner H.V. (1990). IgA nephritis: on theimportance of morphological and clinical parameters in the long-termprognosis of 239 patients. Am. J. Nephrol. 10, 137-147.

Boute N., Gribouval O., Roselli S., Benessy F., Lee H., Fuchshuber A.,Dahan K., Gubler M.C., Niaudet P. and Antignac C. (2000). NPHS2,encoding the glomerular protein podocin, is mutated in autosomalrecessive steroid-resistant nephrotic syndrome. Nat. Genet. 24, 349-354.

Caulfield J.P. and Farquhar M.G. (1974). The permeability of glomerularcapillaries to graded dextrans. Identification of the basementmembrane as the primary filtration barrier. J. Cell Biol. 63, 883-903.

Ciani L., Patel A., Allen N.D. and Ffrench-Constant C. (2003). Micelacking the giant protocadherin mFAT1 exhibit renal slit junctionabnormalities and a partially penetrant cyclopia and anophthalmiaphenotype. Mol. Cell. Biol. 23, 3575-3582.

Deen W.M., Lazzara M.J. and Myers B.D. (2001). Structuraldeterminants of glomerular permeability. Am. J. Physiol. RenalPhysiol. 281, F579-596.

Donoviel D.B., Freed D.D., Vogel H., Potter D.G., Hawkins E., BarrishJ.P., Mathur B.N., Turner C.A., Geske R., Montgomery C.A.,Starbuck M., Brandt M., Gupta A., Ramirez-Solis R., ZambrowiczB.P. and Powell D.R. (2001). Proteinuria and perinatal lethality inmice lacking NEPH1, a novel protein with homology to NEPHRIN.Mol. Cell. Biol. 21, 4829-4836.

Duan H.J., Ohno S. and Shigematsu H. (1993). Anionic sites inglomerular basement membrane of rats with serum sicknessnephritis: quick-freezing and deep-etching study. Histochem. J. 25,367-375.

Farquhar M.G., Wissig S.L. and Palade G..E. (1961). Glomerularpermeability I. Ferritin transfer across the normal glomerularcapillary wall. J. Exp. Med. 113, 47-66.

Furukawa T., Ohno S., Oguchi H., Hora K., Tokunaga S. and Furuta S.(1991). Morphometric study of glomerular slit diaphragms fixed byrapid-freezing and freeze-substitution. Kidney Int. 40, 621-624.

Heuser J.E. and Kirschner M.W. (1980). Filament organization revealedin platinum replicas of freeze-dried cytoskeletons. J. Cell Biol. 86,212-234.

Hora K., Ohno S., Oguchi H., Furukawa T. and Furuta S. (1990). Three-dimensional study of glomerular slit diaphragm by the quick-freezingand deep-etching replica method. Eur. J. Cell Biol. 53, 402-406.

Kalluri R., Shield C.F., Todd P., Hudson B.G. and Neilson E.G. (1997).Isoform switching of type IV collagen is developmentally arrested inX-linked Alport syndrome leading to increased susceptibility of renalbasement membranes to endoproteolysis. J. Clin. Invest. 99, 2470-2478.

Kestilä M., Lenkkeri U., Männikkö M., Lamerdin J., McCready P.,Putaala H., Ruotsalainen V., Morita T., Nissinen M., Herva R.,Kashtan C.E., Peltonen L., Holmberg C., Olsen A. and TryggvasonK. (1998). Positionally cloned gene for a novel glomerular protein--nephrin--is mutated in congenital nephrotic syndrome. Mol. Cell. 1,575-582.

Kleppel M.M., Santi P.A., Cameron J.D., Wieslander J. and Michael A.F.

306


(1989). Human tissue distribution of novel basement membranecollagen. Am. J. Pathol. 134, 813-825.

Miner J.H. (1999). Renal basement membrane components. Kidney Int.56, 2016-2024.

Moriya T., Ohno S. Nakazawa K., Shigematsu H. and Yajima Y. (1993).Ultrastructural study of glomerular basement membrane in diabeticrats by quick-freezing and deep-etching method. Virchows Arch. (B)64, 107-114.

Moriya T., Ohno S. and Yajima Y. (1995). Three-dimensionalultrastructure of human glomerular basement membrane by a quick-freezing and deep-etching method. Nephron 70, 72-76.

Moriya T., Ohno S. Tanaka K., Fujii Y. and Yajima Y. (1996). Effects ofearly insulin treatment on ultrastructural changes of glomeruli indiabetic rats revealed by the quick-freezing and deep-etchingmethod. Diabetologia 39, 632-640.

Nakazawa K., Ohno S., Naramoto A., Takami H., Duan H.J., Itoh N. andShigematsu H. (1992). Immune deposits in the glomerularextracellular matrix detected by the quick-freezing and deep-etchingmethod. Nephron 62, 203-212.

Naramoto A., Ohno S., Itoh N., Shibata N., Nakazawa K., Takami H.,Duan H.J., Kasahara H. and Shigematsu H. (1991). Ultrastructure ofmatriceal changes in chronic phase of Masugi nephritis by quick-freezing and deep-etching method. Virchows Arch. (A) 418, 51-59.

Ohno S. (1985). Immunocytochemical study on the cytoplasmic side ofcell membranes infected with vesicular stomatitis virus by quick-freezing and deep-etching replica method. Histochemistry 82, 565-575.

Ohno S., Hora K., Furukawa T. and Oguchi H. (1992). Ultrastructuralstudy of the glomerular slit diaphragm in fresh unfixed kidneys by aquick-freezing method. Virchows Arch. (B) 61, 351-358.

Ohno S., Baba T., Terada N., Fujii Y. and Ueda H. (1996). Cell biologyof kidney glomerulus. Int. Rev. Cytol. 166, 181-230.

Okada K., Funai M., Kawakami K., Kagami S., Yano I. and Kuroda Y.(1990). IgA nephropathy in Japanese children and adults: acomparative study of clinicopathological features. Am. J. Nephrol.10, 191-197.

Oomura A., Nakamura T., Arakawa M., Ooshima A. and Isemura M.(1989). Alterations in the extracellular matrix components in humanglomerular diseases. Virchows Arch. (A) 415, 151-159.

Pavenstädt H., Kriz W. and Kretzler M. (2003). Cell biology of theglomerular podocyte. Physiol. Rev. 83, 253-307.

Rodewald R. and Karnovsky M.J. (1974). Porous substructure of theglomerular slit diaphragm in the rat and mouse. J. Cell Biol. 60, 423-433.

Schnaper H.M., Hayashida T., Hubchak S.C. and Poncelet A.C. (2003).TGF-ß signal transduction and mesangial cell fibrogenesis. Am. J.Physiol. Renal Physiol. 284, F243-252.

Schneeberger E.E., Levey R.H., McCluskey R.T. and Karnovsky M.J.

(1975). The isoporous substructure of the human glomerular slitdiaphragm. Kidney Int. 8, 48-52.

Schöcklmann H.O., Lang S. and Sterzel R.B. (1999). Regulation ofmesangial cell proliferation. Kidney Int. 56, 1199-1207.

Shih N.Y., Li J., Karpitskii V., Nguyen A., Dustin M.L., Kanagawa O.,Miner J.H. and Shaw A.S. (1999). Congenital nephrotic syndrome inmice lacking CD2-associated protein. Science 286, 312-315.

Takami H., Naramoto A., Shigematsu H. and Ohno S. (1991).Ultrastructure of glomerular basement membrane by quick-freezeand deep-etch methods. Kidney Int. 39, 659-664.

Tanaka K., Moriya T., Ohno S. and Yajima Y. (1996). The three-dimensional ultrastructure of diabetic glomeruli revealed by thequick-freezing and deep-etching method. Med. Electron Microsc. 29,13-22.

Tiebosch A.T., Frederik P.M., van Breda Vriesman P.J., Mooy J.M., vanRie H., van de Wiel T.W., Wolters J. and Zeppenfeldt E. (1989).Thin-basement-membrane nephropathy in adults with persistenthematuria. N. Engl. J. Med. 320, 14-18.

Timpl R. (1989). Structure and biological activity of basementmembrane proteins. Eur. J. Biochem. 180, 487-502.

Ueda T., Nakajima M., Akazawa H., Maruhashi Y., Shimoyama H.,Sakagami Y., Takagawa K., Kamitsuji H., Naito I., Sado Y. andYoshioka A. (2002). Quantitative analysis of glomerular type IVcollagen alpha3-5 chain expression in children with thin basementmembrane disease. Nephron 92, 271-278.

van den Born J., van den Heuvel L.P., Bakker M.A., Veerkamp J.H.,Assmann K.J., Weening J.J. and Berden J.H. (1993). Distribution ofGBM heparan sulfate proteoglycan core protein and side chains inhuman glomerular diseases. Kidney Int. 43, 454-463.

Vasmant D., Maurice M. and Feldmann G. (1984). Cytoskeletonultrastructure of podocytes and glomerular endothelial cells in manand in the rat. Anat. Rec. 210, 17-24.

Wartiovaara J., Öfverstedt L.G., Khoshnoodi J., Zhang J., Mäkelä E.,Sandin S., Ruotsalainen V., Cheng R.H., Jalanko H., Skoglund U.and Tryggvason K. (2004). Nephrin strands contribute to a porousslit diaphragm scaffold as revealed by electron tomography. J. Clin.Invest. 114, 1475-1483.

Yang Y., Zhang S.Y., Sich M., Béziau A., van den Heuvel L.P. andGubler M.C. (2001). Glomerular extracellular matrix and growthfactors in diffuse mesangial sclerosis. Pediatr. Nephrol. 16, 429-438.

Yu Y., Leng C.G., Kato Y. Terada N., Fujii Y. and Ohno S. (1998).Ultrastructural study of anionic sites in glomerular basementmembranes at different perfusion pressures by quick-freezing anddeep-etching method. Nephron 78, 88-95.

Zimmerman S.W. and Burkholder P.M. (1975). Immunoglobulin Anephropathy. Arch. Intern. Med. 135, 1217-1223.

Accepted September 14, 2007

307


ultrastructural study of human glomerular capillary loops with iga ... · lesions of human renal...

Documents