Abstract

Background. The molecular basis for the alteration of glomerular podocyte phenotype in nephrotic syndrome probably involves adaptive changes of the actin cytoskeleton. α-Actinin-4 is an actin cross-linking protein that also interacts with intra- and intercellular adhesion molecules and elements of the transmembrane signal transduction pathway and is implicated in nephrotic syndrome by animal models and human genetic studies.

Methods. We have performed the first quantitative immunoelectron microscopy study of α-actinin-4 expression in humans, analysing 12 cases of minimal change nephrosis (MCNS) and 16 cases of idiopathic membranous nephropathy (MGN), and comparing this with expression in normal tissue from seven nephrectomies (Nx). α-Actinin-4 was visualized by immunogold labelling of plastic-embedded whole glomerular cross-sections, and analysed using LUCIA software.

Results. Despite podocyte effacement, α-actinin-4 expression (group mean±1 SD) in MCNS was similar to that seen in normal Nx podocytes. In contrast, α-actinin-4 expression in MGN was significantly higher than in MCNS or Nx (P<0.001). Furthermore, in MGN cases showing a segmental deposition, expression of α-actinin-4 was significantly higher only in those capillary loop segments containing deposits, whereas in those segments without deposits expression was unchanged compared with that seen in Nx (P<0.001). α-Actinin-4 expression was higher in MGN cases where subepithelial deposits abutted podocyte cytoplasm, and lower in disease stages where deposits were contained within the glomerular basement membrane or were being reabsorbed.

Conclusions. Elevated α-actinin-4 expression is observed only in MGN, and only in areas of subepithelial deposits. Further investigation into the cause of this may reveal insights into the pathogenesis of acquired nephrosis.

Introduction

Nephrotic syndrome, the hallmark of many acquired and inherited nephropathies, is characterized by the clinical triad of heavy proteinuria, hypoalbuminuria and oedema. Regardless of the underlying diagnosis, the condition demonstrates stereotypical ultrastructural changes in the visceral glomerular epithelial cells (podocytes) with retraction and effacement of the highly specialized interdigitating glomerular podocyte foot processes that normally overlie the surface of the glomerular basement membrane (GBM). The molecular basis for this alteration of the podocyte phenotype, involving cytoskeletal reorganization with the formation of a flattened, simple epithelium is the subject of intense ongoing research [1,2]. Studies of familial nephrotic syndrome [3] and animal models [4,5], have implicated several molecules, both within and without the podocytes as candidates regulating the modulation of pathogenetic changes in the barrier to protein filtration and regulation of the podocyte actin cytoskeleton.

One such candidate, α-actinin-4, is a member of the actin-binding and cross-linking family of proteins that are encoded by four highly homologous α-actinin genes which are highly conserved among different species [3]. Expression of two isoforms, ACTN2 and ACTN3, is limited to the skeletal muscle sarcomere, whereas the distribution of non-muscle isoforms, ACTN1 and ACTN4, is widespread. The α-actinin molecule is an elongated, symmetrical, anti-parallel dimeric rod, with actin-binding sites at either end that enable cross-linkage of F-actin filaments into contractile bundles [3,6], and the formation of an anchoring complex for the actin cytoskeleton at focal contacts on the plasma membrane. The key proteins with which α-actinin-4 is known to interact include intercellular adhesion molecules ICAM-1 and -2 [7], β1-integrins [8], and intracellular focal contact-associated proteins [9]. Binding of α-actinin-4 to the cytoplasmic domain of the β1-integrin subunit of the α31 complex is thought to be important in anchoring actin microfilaments of podocyte foot processes to the GBM [8]. In addition to these structural and transmembrane functions, the biological versatility of α-actinin-4 is further demonstrated by its interactions with components of signal transduction pathways [10]. These associations strongly suggest that α-actinin-4 is a key component of podocytes which through altered expression, or redistribution, is able to modulate reorganization of the actin cytoskeleton in response to changes in the microenvironment of the cell. In normal human kidney, α-actinin-4 is the only isoform expressed [3]. It is expressed in blood vessel walls and in glomeruli, where it is localized predominantly within the dense bodies of podocyte foot processes, and to a minor extent in mesangial cytoplasmic processes [11]. Recently, familial focal and segmental glomerulosclerosis has been associated with a gain of function mutation of the ACTN4 gene [3].

In this report, we have performed the first analysis of α-actinin-4 and actin expression in human acquired nephrotic syndrome. By studying minimal change nephrotic syndrome (MCNS) we can draw inferences on the influence of proteinuria alone, and by study of idiopathic membranous glomerulonephritis (MGN) we can examine the additional effect of subepithelial deposition on the podocyte expression of α-actinin-4 and the actin cytoskeleton. Our immunogold electron microscopy study enabled precise quantitation of the expression and distribution of α-actinin-4 at the ultrastructural level in these two pathologies and indicates significant differences between them that may be important in understanding disease pathogenesis and provide novel therapeutic targets in the future.

Materials and methods

Archived renal biopsy and nephrectomy tissue blocks (1 mm3), fixed in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, and embedded in LR Gold resin (London Resin Company, Basingstoke, Hants, UK) [12] were used under local ethical guidelines and approval (Project Nos. 02/068 and 02/072). Tissue that showed ‘normal’ histology, dissected from the unaffected poles of kidneys that had been surgically removed for the treatment of renal carcinoma, was used as normal control (Nx, n = 7; age range 8–70 years; male:female ratio 5:2). The study included renal biopsy tissue from patients with MCNS, and patients with MGN and nephrotic range proteinuria. The original pathology diagnosis in all of the biopsies and the control tissue included in the study was independently reviewed. Sampling of glomeruli was limited by both the technique and the small amount of biopsy tissue available. The study was based upon analysis of single glomeruli from each biopsy that were chosen according to their superior orientation and preservation.

Reagents

Monoclonal mouse anti-human α-actinin (isotype IgG1) was purchased from Insight Biotechnology, Wembley, Middlesex, UK. This antibody was not specific for isoform-4. The specificity of staining of this antibody in the disease and control groups studied was subsequently confirmed in neutralization control experiments using a newly available polyclonal rabbit anti-human α-actinin-4 isoform specific antibody purchased from ImmunoGlobe GmbH, Himmelstadt, Germany. Cytoskeletal actin was demonstrated using a rabbit IgG poyclonal antibody directed towards highly conserved amino acid residues (20–33) of the N-terminal region (Sigma-Aldrich, Poole, Dorset, UK). Electron microscopy grade goat anti-mouse IgG, and goat anti-rabbit IgG gold conjugates (10 nm) were purchased from British Biocell International, Cardiff, UK.

Immunoelectron microscopy

Ultrathin full glomerular cross-sections nominally cut at a thickness of 90 nm that showed silver/gold interference colours were placed on 300 µm single-hole gold grids (Gilder, Grantham, Lincolnshire, UK), on which LR Gold blank sections of similar thickness had been previously mounted to provide mechanical support for the subsequent demanding handling conditions. Four duplicate sections from three or four biopsies, including a mixture of Nx, MCNS and MGN cases were included in each staining run.

Demonstration of actin did not require antigen retrieval, whilst this was necessary for the detection of α-actinin and was achieved by immersion of section grids in 0.1 M sodium citrate buffer, pH 6.0 at 95°C for 10 min. Sections were left to cool for 20 min in the citrate buffer, jet-washed in distilled water and immersed in drops of 0.5 M NH4Cl in 0.1 M phosphate buffer, pH 7.3 for 20 min. After blocking non-specific staining in 20% (w/v) normal goat serum (NGS) for 10 min, sections were incubated overnight at 4°C in mouse anti-human α-actinin, diluted 1:20 in phosphate-buffered saline, pH 7.3 containing 1% bovine serum albumin (w/v) and 0.1% Tween-20 (w/v) and 5% NGS (PBS). After washing in this PBS buffer system, sections were incubated in 20% NGS for 10 min followed by incubation for 2 h at 20°C in goat anti-mouse IgG conjugated to 10 nm gold particles, diluted 1:50 in PBS. Sections were jet-washed in distilled water, silver enhanced for 12 min [13], counterstained in saturated aqueous uranyl acetate for 3 min, rinsed briefly in distilled water and allowed to dry. Actin was demonstrated by the same procedure using the rabbit anti-actin antibody and the goat anti-rabbit IgG gold conjugate at a dilution of 1:200 and 1:100, respectively.

Appropriate negative controls included replacement of the primary antibody with non-immune serum. Inclusion of control sections that were pre-incubated overnight at 4°C with 1:200 rabbit anti-human α-actinin-4 (5× optimal concentration) prior to incubation with the monoclonal antibody was used to confirm the isoform specificity of α-actinin immunolabelling in all groups. Sections were examined at 40 kV using a Jeol JEM 100CX electron microscope. The use of single-hole grids enabled whole glomerular cross-sections to be studied from each case. A minimum of six representative areas from each glomerulus were photographed at a standard magnification of ×3000, and entire negatives were scanned into a computer using an Epson Perfection 1200S scanner and Epson Twain version 5.0 scanning software (Seiko Epson Corporation).

Image analysis

Scanned images were downloaded on to CD-ROM and analysed using Lucia L/G imaging software (Nikon UK, Kingston upon Thames, Surrey, UK). In all cases, peripheral glomerular capillary loop segments that had been cut perpendicular were selected for analysis. Tangential sections, paramesangial segments, and loops that were sharply reflected, or those which were contiguous with shared epithelium, were excluded.

Some early membranous glomerulopathies (stage I and II), showed a segmental, discontinuous distribution of immune complex deposition. In these cases, capillary loop segments with deposits were analysed separately from those segments without clearly discernible deposits.

Statistical analysis

Each measurement was based upon the density of α-actinin labelling, observed as the number of gold particles detected along a measured length of GBM. The minimum length of GBM that was analysed was 250 µm. In preliminary studies this represented a length that gave a constant mean density of α-actinin labelling in ‘normal’ control glomeruli. Calculation of this length of GBM was based on a summation average graph, as detailed in our previous studies [14]. For each measurement in individuals in each of the three study groups, the number of gold particles per µm length of GBM was calculated. Individual measurements for all cases in each of the study groups were combined and the mean density of labelling and standard deviation calculated and compared. Differences in α-actinin labelling in the study groups were identified by analysis of individual measurements in each population using a one-way ANOVA test. A P-value <0.005 was regarded as statistically significant.

Results

Patient characteristics

The patient characteristics of all biopsies included in the study are summarized in Tables 1 and 2. The MGN group included an age spread and sex ratio that is typical of idiopathic membranous nephropathy. The MCNS group included eight paediatric and four adult patients since there was no significant difference in α-actinin expression in these two groups.

Table 1.

Minimal change glomerulonephritis

Code no. Age Sex GBM length (µm) Mean particles/µm 
81.91 70 302 2.76±1.11 
32.92 10 496 2.14±0.49 
46.92 09 383 3.88±0.60 
86.92 09 263 1.44±0.46 
05.93 36 517 2.50±1.02 
26.93 10 418 2.43±0.51 
33.93 08 361 1.82±0.92 
40.93 13 492 1.37±0.34 
45.93 11 458 1.17±0.34 
48.93 79 302 3.45±0.78 
66.94 14 503 1.81±0.53 
84.94 59 411 2.07±0.62 
Code no. Age Sex GBM length (µm) Mean particles/µm 
81.91 70 302 2.76±1.11 
32.92 10 496 2.14±0.49 
46.92 09 383 3.88±0.60 
86.92 09 263 1.44±0.46 
05.93 36 517 2.50±1.02 
26.93 10 418 2.43±0.51 
33.93 08 361 1.82±0.92 
40.93 13 492 1.37±0.34 
45.93 11 458 1.17±0.34 
48.93 79 302 3.45±0.78 
66.94 14 503 1.81±0.53 
84.94 59 411 2.07±0.62 

Mean age 27.3 years, range 8–79; male:female 8:4; n = 12.

Table 2.

Membranous glomerulonephritis

Code no. Stage Age Sex GBM length (µm) Mean particles/µm (no deposits) 
47.93 76 258 (314) 8.12±3.61 (1.93±0.88) 
01.95 22 314 (287) 5.61±1.47 (2.01±0.97) 
23.92 I/II 21 537 6.84±1.98 
29.92 I/II 36 357 5.12±1.99 
39.92 I/II 26 270 4.74±1.49 
86.94 I/II 47 514 6.22±1.96 
40.92 II 40 413 7.14±4.55 
16.93 II 59 395 5.12±1.87 
106.93 II 57 423 4.75±1.42 
29.94 II 55 302 4.49±1.08 
44.94 II 75 411 3.57±1.10 
16.95 II 53 268 (311) 4.80±1.90 (2.04±1.29) 
17.95 II 39 372 3.30±1.54 
03.94 II/III 55 392 2.97±0.86 
62.93 IV 37 326 3.60±2.25 
57.92 IV 33 324 7.68±2.06 
Code no. Stage Age Sex GBM length (µm) Mean particles/µm (no deposits) 
47.93 76 258 (314) 8.12±3.61 (1.93±0.88) 
01.95 22 314 (287) 5.61±1.47 (2.01±0.97) 
23.92 I/II 21 537 6.84±1.98 
29.92 I/II 36 357 5.12±1.99 
39.92 I/II 26 270 4.74±1.49 
86.94 I/II 47 514 6.22±1.96 
40.92 II 40 413 7.14±4.55 
16.93 II 59 395 5.12±1.87 
106.93 II 57 423 4.75±1.42 
29.94 II 55 302 4.49±1.08 
44.94 II 75 411 3.57±1.10 
16.95 II 53 268 (311) 4.80±1.90 (2.04±1.29) 
17.95 II 39 372 3.30±1.54 
03.94 II/III 55 392 2.97±0.86 
62.93 IV 37 326 3.60±2.25 
57.92 IV 33 324 7.68±2.06 

Mean age 45.7 years, range 21–76; male:female 12/4; n = 16.

Figures in parentheses show labelling in GBM lengths with no deposits.

α-Actinin-4 immunogold electron microscopy

The ability to block >97% of staining by the pre-incubation of sections with the isoform-specific rabbit polyclonal antibody confirmed that α-actinin-4 was almost exclusively expressed in all groups.

Immunogold electron microscopy precisely localized α-actinin-4 in podocytes. Expression in normal glomeruli was restricted to the podocyte foot processes, with only minor localization of gold particles in the cell body (Figure 1). Where localization in the podocyte body did occur, this was in the vicinity of the plasma membrane.

Fig. 1.

Localization of α-actinin-4 in glomerular podocytes of normal kidney. The distribution of α-actinin is restricted to foot processes, with minimal staining of the cell body. Labelling is associated with more electron-dense areas of cell cytoplasm (inset, area within brackets). US, urinary space; cap, glomerular capillary; pod, podocyte. Magnification ×10 250; inset ×28 500.

Fig. 1.

Localization of α-actinin-4 in glomerular podocytes of normal kidney. The distribution of α-actinin is restricted to foot processes, with minimal staining of the cell body. Labelling is associated with more electron-dense areas of cell cytoplasm (inset, area within brackets). US, urinary space; cap, glomerular capillary; pod, podocyte. Magnification ×10 250; inset ×28 500.

Careful examination at higher magnification revealed that immunogold labelling was associated with electron-dense patches in the foot process cytoplasm (Figure 1, inset). These areas of densification, which probably represent condensations of cytoskeletal F-actin bundles [5], were more prominent in the effaced podocytes of MCNS and were a striking feature of MGN. However, in MCNS, labelling of α-actinin-4 resembled that seen in normal glomeruli (Figure 2). The cytoplasmic densifications in MGN were particularly prominent where immune deposits abutted the overlying epithelium. α-Actinin-4 immunolabelling was greatest in those cases of MGN where subepithelial deposits abutted the overlying epithelium, being predominantly associated with the areas of cytoplasmic densification (Figure 3).

Fig. 2.

Localization of α-actinin-4 in an effaced podocyte of a case of MCNS. The podocyte shows stereotypical nephrotic changes with swelling of the cell body which includes lipid droplets, and fusion of foot processes. α-Actinin expression appears similar to that seen in normal kidney. US, urinary space; glomerular capillary; pod, podocyte; *, lipid droplet. Magnification ×10 250.

Fig. 2.

Localization of α-actinin-4 in an effaced podocyte of a case of MCNS. The podocyte shows stereotypical nephrotic changes with swelling of the cell body which includes lipid droplets, and fusion of foot processes. α-Actinin expression appears similar to that seen in normal kidney. US, urinary space; glomerular capillary; pod, podocyte; *, lipid droplet. Magnification ×10 250.

Fig. 3.

Localization of α-actinin-4 in stage II MGN. Increased expression is restricted to a prominent layer of electron-dense cytoplasm at the interface where the conspicuous subepithelial deposits abut the basal surface of the flattened podocyte. US, urinary space; GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×13 500.

Fig. 3.

Localization of α-actinin-4 in stage II MGN. Increased expression is restricted to a prominent layer of electron-dense cytoplasm at the interface where the conspicuous subepithelial deposits abut the basal surface of the flattened podocyte. US, urinary space; GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×13 500.

In the later stages of MGN, when even large deposits were localized deep in the GBM and enclosed by GBM-like material (stage III) or showed evidence of reabsorption, seen as electron-lucent areas (stage IV), labelling was reduced (Figure 4). In these later stages, those segments containing deposits that still abutted the overlying podocyte continued to demonstrate higher levels of immunogold localization. In MGN cases with a segmental rather than a uniform deposition, heavy labelling was observed in regions with deposits, but this was less along segments where deposition was not discernible (Figure 5).

Fig. 4.

Localization of α-actinin-4 in a late (stage IV) MGN case. Glomerular capillary loop segments show typical areas of deposit reabsorption, seen as electron lucent holes, and deposits that have been overgrown by new GBM-like material (arrowheads). Expression of α-actinin in these areas is less than that seen in an adjacent loop segment (above) where deposits are still seen to abut the now highly disorganized podocyte cytoplasm. GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×10 250.

Fig. 4.

Localization of α-actinin-4 in a late (stage IV) MGN case. Glomerular capillary loop segments show typical areas of deposit reabsorption, seen as electron lucent holes, and deposits that have been overgrown by new GBM-like material (arrowheads). Expression of α-actinin in these areas is less than that seen in an adjacent loop segment (above) where deposits are still seen to abut the now highly disorganized podocyte cytoplasm. GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×10 250.

Fig. 5.

Localization of α-actinin-4 in an MGN case showing segmental deposition. High expression of α-actinin in the glomerular capillary loop segment that contains membranous deposits (above) contrasts with that seen in the segment which contains no discernible deposits (between arrows). US, urinary space; cap, glomerular capillary; pod, podocyte; E, endothelial cell. Magnification ×13 500.

Fig. 5.

Localization of α-actinin-4 in an MGN case showing segmental deposition. High expression of α-actinin in the glomerular capillary loop segment that contains membranous deposits (above) contrasts with that seen in the segment which contains no discernible deposits (between arrows). US, urinary space; cap, glomerular capillary; pod, podocyte; E, endothelial cell. Magnification ×13 500.

Actin gold immunoelectron microscopy

Actin was demonstrated in normal glomeruli, in the cytoplasmic processes of mesangial > epithelial podocytes > endothelial cells. Overall, the labelling density was low, and was predominantly localized in the region of the plasma membrane of these cells. The foot processes of epithelial podocytes were preferentially labelled (Figure 6). In MCNS, the distribution and level of expression of actin labelling resembled that seen in normal glomeruli. Red blood cells were also labelled with the anti-actin antibody. This may represent elements of the spectrin-actin network that constitutes the erythrocyte membrane skeleton.

Fig. 6.

Localization of actin in the peripheral glomerular capillary wall of normal kidney. Podocytes and endothelial cells are labelled. In podocytes labelling is restricted to foot processes, with minimal staining of the cell body. A circulating red blood cell is also labelled. US, urinary space; cap, glomerular capillary; pod, podocyte; E, endothelial cell. Magnification ×11 000.

Fig. 6.

Localization of actin in the peripheral glomerular capillary wall of normal kidney. Podocytes and endothelial cells are labelled. In podocytes labelling is restricted to foot processes, with minimal staining of the cell body. A circulating red blood cell is also labelled. US, urinary space; cap, glomerular capillary; pod, podocyte; E, endothelial cell. Magnification ×11 000.

In contrast to normal and MCNS groups, actin expression was markedly elevated in MGN. However, increased labelling of actin in different compartments of the expanded glomerular capillary wall in stage II/III and late stage (IV) MGN precluded precise quantitation. Figure 7 shows typical actin labelling in stage II MGN. Increased labelling was most prominent in the electron-dense ribbons of basal cytoplasm overlying the deposits, confirming the presence of actin within this condensed layer. Interestingly, the antibody predominantly labelled the peripheral margins of these areas. Elevated actin expression also extended into the podocyte cell body with labelling of reticular structures and electron-dense margins adjacent to the plasma membrane (Figure 7). Surprisingly, labelled sites were also demonstrated around the peripheral margins of immune deposits within the GBM, and occasionally within deposits themselves (Figure 7). Some non-specific labelling of condensed plasma was also evident in MGN cases.

Fig. 7.

Localization of actin in stage II MGN. Increased expression is present in the electron-dense cytoplasm (arrowed) at the interface where subepithelial deposits abut the basal surface of the flattened podocyte. Actin is also present in the body of the podocyte with labelling of reticular structures, and the electron-dense margins of the plasma membrane (arrowheads). US, urinary space; GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×13 000.

Fig. 7.

Localization of actin in stage II MGN. Increased expression is present in the electron-dense cytoplasm (arrowed) at the interface where subepithelial deposits abut the basal surface of the flattened podocyte. Actin is also present in the body of the podocyte with labelling of reticular structures, and the electron-dense margins of the plasma membrane (arrowheads). US, urinary space; GBM, glomerular basement membrane; cap, glomerular capillary; pod, podocyte. Magnification ×13 000.

Quantitative analysis of immunogold labelling of α-actinin-4

Quantitative analysis of the number of particles detected per unit length of GBM in the three study groups confirmed the qualitative observations (Figure 8). In Nx controls, the mean number of gold particles per µm length of GBM was similar to that seen in the MCNS group [2.46±1.0 (±1 SD), based upon 108 measurements in seven Nx cases, compared with 2.25±1.0 (±1 SD), based upon 204 measurements in 12 MCNS cases (P>0.05)]. In contrast with these groups, the mean number of gold particles per µm length of GBM in the MGN group, based upon 226 measurements in 16 MGN cases was significantly higher [5.13±2.57 (±1 SD)] than Nx (P<0.001) and MCNS (P<0.001). α-Actinin-4 expression in MGN was analysed further. We identified three MGN cases in our sample that showed a segmental rather than a uniform distribution of deposits. In these cases, quantitation of the number of gold particles per µm length of GBM in segments with deposits (MGN dep) was significantly higher than in those without discernible deposits (MGN 0): 5.90±2.57 (±1 SD), based upon 25 measurements in MGN dep cases, compared with 1.99±1.03 (±1 SD), based upon 39 measurements in MGN 0 cases (P<0.001) (Figure 9).

Fig. 8.

α-Actinin-4 expression, shown as the group mean number of gold particles per µm length of GBM±1 SD. In the MGN group α-actinin-4 expression is significantly higher than that seen in either the MCNS group (*P<0.001) or the Nx group (+P<0.001).

Fig. 8.

α-Actinin-4 expression, shown as the group mean number of gold particles per µm length of GBM±1 SD. In the MGN group α-actinin-4 expression is significantly higher than that seen in either the MCNS group (*P<0.001) or the Nx group (+P<0.001).

Fig. 9.

α-Actinin-4 expression, shown as the group mean number of gold particles per µm length of GBM±1 SD. The α-actinin expression in areas with deposits (MGN dep) is significantly higher than in those areas of GBM without deposits (MGN 0) (*P<0.001). Expression of α-actinin-4 in areas without deposit was not significantly different from normal Nx controls (+P>0.05).

Fig. 9.

α-Actinin-4 expression, shown as the group mean number of gold particles per µm length of GBM±1 SD. The α-actinin expression in areas with deposits (MGN dep) is significantly higher than in those areas of GBM without deposits (MGN 0) (*P<0.001). Expression of α-actinin-4 in areas without deposit was not significantly different from normal Nx controls (+P>0.05).

Discussion

Our present study provides the first quantitative analysis of α-actinin-4 expression in glomerular epithelial podocytes in acquired human nephrotic syndrome. Although sampling was restricted to the analysis of single glomeruli per case in each of the study groups, the diffuse nature of the pathologies enables valid conclusions to be reached. The application of immunogold at the ultrastructural level enables a much more precise analysis of protein expression than could be reached using optical microscopy methods which at best are only semi-quantitative.

Encoded by one of four highly homologous genes, α-actinin-4 is widely expressed throughout the body. In the normal kidney, α-actinins-1, -2 and -3 are not expressed [3], but α-actinin-4 is localized in the foot processes of podocytes and in blood vessel walls [35,11]. In addition to α-actinin-4, the α-actinin-1 isotype has also been reported to be expressed by cultured mouse podocytes [15]. Therefore, it is highly likely that the staining pattern in podocytes in this study represents α-actinin-4 expression. In our study, the ability to block immunogold staining using the isoform-specific polyclonal α-actinin-4 antibody confirmed that this isoform was almost exclusively expressed by all groups. In the absence of a specific α-actinin-1 antibody, we were unable to confirm whether the few residual (<3%) unblocked sites were due to expression of this isoform.

The biological diversity of α-actinin-4 enables it to interact not only with transmembrane receptors (integrins, intercellular adhesion molecules) [7,8], that are susceptible to influence by factors outside the cell, but also components of the actin cytoskeleton [3,6]. Furthermore, α-actinin-4 also interacts with intracellular contact-associated proteins [9] and molecules of the signal transduction pathways [10]. Therefore, it represents a strong candidate molecule to modulate changes in the podocyte cytoskeleton, either directly, or by intermediate signalling pathways in response to external factors. Knockout (abstract; Kos et al., Am Soc Nephrol [F-FC079] 2002) and transgenic studies (abstract; Michaud et al., Am Soc Nephrol [SU-P0292] 2002) have emphasized the critical role for α-actinin-4 in maintaining podocyte integrity in animals, as has the association of familial focal and segmental glomerulosclerosis with a gain of function mutation of the ACTN4 gene [3] in human disease.

In this study, we found that α-actinin-4 expression is augmented in MGN, but not in nephrotic syndrome secondary to MCNS. Furthermore, the increased α-actinin expression was only associated with areas of subepithelial deposition, but remained unchanged in glomerular segments where there was no such deposition. In addition, in later stages of MGN, when deposits are being reabsorbed or where GBM material encloses the deposits and the glomerular epithelial podocyte is once again in contact with intact basement membrane, α-actinin expression was diminished. In view of the functional diversity of α-actinin-4 in podocytes the potential stimuli leading to changes in its expression in MGN may be complex and varied.

Nephrotic range proteinuria and podocyte foot process effacement and detachment are common to both MCNS and MGN. However, these pathologies differ in the deposition of immune deposits and a marked irregularity of the GBM in MGN. Our data support the view that the increased α-actinin-4 expression follows displacement of the podocyte from its attachment to GBM by immune deposits that provide a microenvironmental stretch stimulus on the podocyte basal membrane. Certainly, subjecting these cells to mechanical stresses in vitro leads to reorganization of the F-actin cytoskeleton, indicating that these cells are mechanosensitive [16]. There is further evidence that interruption of the podocyte/GBM interaction is important. In Masugi nephritis, an animal model of anti-GBM disease, Shirato et al. [5] demonstrated augmented α-actinin-4 expression associated with an increased prominence of cytoskeletal microfilaments and dense bodies, coincidental with marked foot process effacement. These authors suggested that increased α-actinin-4 expression may be an adaptive response of the podocyte to the observed development of an undulating contour and irregular contact between the basal surface of effaced foot processes and the underlying GBM, this being caused by the disruption of integrin/GBM interactions by antibodies to β1-integrin that are present in the anti-rat GBM sera used to induce the disease [5]. In contrast, in studies of mice with defective GBM, all of whom developed significant proteinuria and alteration of podocyte architecture, but in which no lifting of epithelial cells was observed, there was no alteration in α-actinin-4 expression [17].

The apparent increase in α-actinin-4 expression may be due to a change in the distribution of actin filaments and associated molecules adjacent to the GBM in patients with MGN. Elevated α-actinin-4 was restricted to areas of condensed cytoplasm, whereas actin expression was increased in these areas, but was also more widespread, suggesting a different molecular rearrangement in different regions of the effaced and expanded podocyte in MGN. Augmented expression of these proteins was not seen in MCNS patients, although Lachapelle and Bendayan [18] have confirmed a co-distribution of α-actinin and actin in condensed areas of podocyte cytoplasm in nephrotic rat kidneys. The rat model is, however, complicated by a binding affinity of α-actinin-4 to the puromycin aminonucleoside used to induce the nephrosis [19]. There is also a possibility that factors additional to physical disruption of the podocyte and its attachment to GBM play a role in augmented α-actinin-4 expression. It is also possible that the specific components found within immune deposits may be responsible for the increased expression/re-arrangement of α-actinin. It is well established that in Heymann nephritis, development of disease is critically linked to the insertion of terminal complement components into the glomerular epithelial cell membrane [20], and more recently, such membrane injury has been demonstrated to induce specific intracellular signalling cascades with which α-actinin-4 is known to interact [21].

Clearly, increased α-actinin-4 expression is not a direct consequence of nephrosis. Whether augmented α-actinin-4 expression in MGN is a primary or a secondary response to deposit formation requires further investigation. From our results, and the known actin-bundling properties of α-actinin-4 [6] it seems likely that a primary, augmented expression of α-actinin-4 leads to cross-linkage of actin filaments resulting in densification of the basal lamina in MGN. However, the increased concentration of both α-actinin-4 and actin in the basal lamina could also be an epi-phenomenon related to the condensation of cytoskeletal components in this region by unidentified factors. The limitation of phenotypic changes in inherited mutations of α-actinin-4 to the glomerulus suggests that it may in time become an important target for therapeutic intervention.

The authors thank Doreen Crellin for expert technical assistance in the preparation of tissue for immunoelectron microscopy. The Academic Unit of Surgery at St James's University Hospital kindly provided use of the LUCIATM imaging system. The project was supported by the Yorkshire Kidney Research Fund.

Conflict of interest statement. None declared.

References

1
Smoyer WE, Mundel P. Regulation of podocyte structure during the development of nephrotic syndrome.
J Mol Med
 
1998
;
76
:
172
–183
2
Kerjaschki D. Caught flat-footed: podocyte damage and the molecular bases of focal glomerulosclerosis.
J Clin Invest
 
2001
;
108
:
1583
–1587
3
Kaplan JM, Kim SH, North KN
et al
 . Mutations in ACTN4, encoding alpha-actinin-4, cause familial focal segmental glomerulosclerosis.
Nature Genet
 
2000
;
24
:
251
–256
4
Smoyer WE, Mundel P, Gupta A, Welsh MJ. Podocyte alpha-actinin induction precedes foot process effacement in experimental nephrotic syndrome.
Am J Physiol
 
1997
;
273
:
t-7
5
Shirato I, Sakai T, Kimura K, Tomino Y, Kriz W. Cytoskeletal changes in podocytes associated with foot process effacement in Masugi nephritis.
Am J Pathol
 
1996
;
148
:
1283
–1296
6
Djinovic-Carugo K, Young P, Gautel M, Saraste M. Structure of the alpha-actinin rod: molecular basis for cross-linking of actin filaments.
Cell
 
1999
;
98
:
537
–546
7
Heiska L, Kantor C, Parr T et al. Binding of the cytoplasmic domain of intercellular adhesion molecule-2 (ICAM-2) to alpha-actinin.
J Biol Chem
 
1996
;
271
:
26214
–26219
8
Otey CA, Vasquez GB, Burridge K, Erickson BW. Mapping of the alpha-actinin binding site within the beta 1 integrin cytoplasmic domain.
J Biol Chem
 
1993
;
268
:
21193
–21197
9
Reinhard M, Zumbrunn J, Jaquemar D et al. An alpha-actinin binding site of zyxin is essential for subcellular zyxin localization and alpha-actinin recruitment.
J Biol Chem
 
1999
;
274
:
13410
–13418
10
Mukai H, Toshimori M, Shibata H et al. Interaction of PKN with alpha-actinin.
J Biol Chem
 
1997
;
272
:
4740
–4746
11
Drenckhahn D, Franke RP. Ultrastructural organization of contractile and cytoskeletal proteins in glomerular podocytes of chicken, rat, and man.
Lab Invest
 
1988
;
59
:
673
–682
12
Goode NP, Shires M, Aparicio SR, Davison AM. Cationic colloidal gold—a novel marker for the demonstration of glomerular polyanion status in routine renal biopsies.
Nephrol Dial Transplant
 
1991
;
6
:
923
–930
13
Holgate CS, Jackson P, Pollard K, Lunny D, Bird CC. Effect fixation on T and B lymphocyte surface membrane antigen demonstration in paraffin processed tissue.
J Pathol
 
1986
;
149
:
293
–300
14
Sinniah R, Khan TN. Renal tubular basement membrane changes in tubulointerstitial damage in patients with glomerular diseases.
Ultrastruct Pathol
 
1999
;
23
:
359
–368
15
Welsch T, Endlich N, Kriz W, Endlich K. CD2AP and p130Cas localize to different F-actin structures in podocytes.
Am J Physiol Renal Fluid Electrolyte Physiol
 
2001
;
281
:
F769
–F777
16
Endlich N, Kress KR, Reiser J et al. Podocytes respond to mechanical stress
in vitro
 .
J Am Soc Nephrol
 
2001
;
12
:
413
–422
17
Hamano Y, Grunkemeyer JA, Sudhakar A et al. Determinants of vascular permeability in the kidney glomerulus.
J Biol Chem
 
2002
;
277
:
31154
–31162
18
Lachapelle M, Bendayan M. Contractile proteins in podocytes: immunocytochemical localization of actin and alpha-actinin in normal and nephrotic rat kidneys.
Virchows Arch B Cell Pathol
 
1991
;
60
:
105
–111
19
Goto H, Wakui H, Komatsuda A et al. Renal alpha-actinin-4: purification and puromycin aminonucleoside-binding property.
Nephron Exp Nephrol
 
2003
;
93
:
e27
–e35
20
Topham PS, Haydar SA, Kuphal R, Lightfoot JD, Salant DJ. Complement-mediated injury reversibly disrupts glomerular epithelial cell actin microfilaments and focal adhesions.
Kidney Int
 
1999
;
55
:
1763
–1775
21
Cybulsky AV, Papillon J, McTavish AJ. Complement activates phospholipases and protein kinases in glomerular epithelial cells.
Kidney Int
 
1998
;
54
:
360
–372

Author notes

1Renal Research Unit, 2Department of Histopathology and 3Department of Renal Medicine, St James's University Hospital, The Leeds Teaching Hospitals NHS Trust, Leeds, UK

Comments

0 Comments