- Email: [email protected]
The glomerular endothelium emerges as a key player in diabetic nephropathy
co m m e nt a r y
The transfer of pathogenic T cells by Okamoto and colleagues7 also induced increased glomerular complement deposition without clear changes in serum IgG levels, anti-dsDNA levels, or overall glomerular IgG deposition, measured semiquantitatively. Nonetheless, it is possible that the transferred T cells did enhance glomerular deposition of pathogenic autoantibodies with higher affinity or of a more damaging IgG subclass. Potentially, these autoantibodies could be responsible for the increased complement deposition found in the mice that received the pathogenic effector CD4 + T cells. The study by Okamoto et al.7 brings us a step closer to understanding how a variety of mediators, including effector CD4 + T cells, form an unholy alliance to mediate injury in lupus nephritis. The ultimate goal of new treatments, more targeted therapy with fewer side effects, may be more difficult in patients with severe lupus nephritis because of both the potential diversity of target autoantigens and the involvement of multiple arms of the effector immune response. Over the past few years there has been real progress in using newer treatments, in the form of therapies such as mycophenolate and belimumab.10 However, already it has been shown that targeting autoimmunity to a single antigen, dsDNA, is not a useful treatment, even though dsDNA was a logical candidate antigen.10 The diversity of clinical presentations and histological pictures in lupus and lupus nephritis may mean that in the future, treatment might need to be tailored in some patients to predominantly target one arm of the immune response, or even a restricted range of antigens. Further well-designed studies in relevant and well-characterized experimental models of disease based on careful observations in humans need to be performed, so that we can better understand the nature of effector responses in this at-times devastating and difficult-to-treat human disease. DISCLOSURE The authors declared no competing interests. REFERENCES 1.
2.
Murata H, Matsumura R, Koyama A et al. T cell receptor repertoire of T cells in the kidneys of patients with lupus nephritis. Arthritis Rheum 2002; 46: 2141–2147. Kitching AR, Holdsworth SR, Tipping PG. Crescentic glomerulonephritis: a manifestation of
Kidney International (2012) 82
3.
4.
5.
6.
a nephritogenic Th1 response? Histol Histopathol 2000; 15: 993–1003. Kitching AR, Holdsworth SR. The emergence of TH17 cells as effectors of renal injury. J Am Soc Nephrol 2011; 22: 235–238. Summers SA, Steinmetz OM, Li M et al. Th1 and Th17 cells induce proliferative glomerulonephritis. J Am Soc Nephrol 2009; 20: 2518–2524. Diaz Gallo C, Jevnikar AM, Brennan DC et al. Autoreactive kidney-infiltrating T-cell clones in murine lupus nephritis. Kidney Int 1992; 42: 851–859. Schwarting A, Wada T, Kinoshita K et al. IFN-gamma receptor signaling is essential for the initiation, acceleration, and destruction of autoimmune
kidney disease in MRL-Fas(lpr) mice. J Immunol 1998; 161: 494–503. 7. Okamoto A, Fujio K, Tsuno NH et al. Kidneyinfiltrating CD4+ T-cell clones promote nephritis in lupus-prone mice. Kidney Int 2012; 82: 969–979. 8. Hanrotel-Saliou C, Segalen I, Le Meur Y et al. Glomerular antibodies in lupus nephritis. Clin Rev Allergy Immunol 2011; 40: 151–158. 9. Lu L, Kaliyaperumal A, Boumpas DT et al. Major peptide autoepitopes for nucleosome-specific T cells of human lupus. J Clin Invest 1999; 104: 345–355. 10. Yildirim-Toruner C, Diamond B. Current and novel therapeutics in the treatment of systemic lupus erythematosus. J Allergy Clin Immunol 2011; 127: 303–312.
see original article on page 1010
The glomerular endothelium emerges as a key player in diabetic nephropathy Simon C. Satchell1 The effects of diabetes on glomerular structure and particularly that of the glomerular capillary wall have been extensively documented. By correlation with clinical measurements, Weil and colleagues provide important insights into the functional significance of glomerular structural changes in type 2 diabetes. Podocyte detachment correlates with albumin-to-creatinine ratio, but less strongly than does loss of endothelial fenestrations, which also correlates with reduced glomerular filtration rate. The role of the glomerular endothelium in diabetic nephropathy demands further scrutiny. Kidney International (2012) 82, 949–951. doi:10.1038/ki.2012.258
The glomerular filtration barrier is highly permeable to water and small molecules but maintains very low permeability to macromolecules.1 These characteristics are dependent on its unique three-layer structure. Glomerular endothelial cells are characterized by fenestrations, 60- to 80nm transcellular holes through the peripheral cytoplasm, essential for high water permeability.2 The glomerular basement membrane (GBM) separates glomerular endothelial cells from podocytes whose 1Academic Renal Unit, University of Bristol, Bristol, UK Correspondence: Simon C. Satchell, Academic Renal Unit, University of Bristol, Learning & Research, Southmead Hospital, Bristol BS10 5NB, UK. E-mail: [email protected]
interdigitating foot processes are spanned by the slit diaphragm (Figure 1). Until recently, the presence of fenestrations has meant that the glomerular endothelium has been considered to contribute little to the barrier to macromolecules. However, it is now appreciated that the luminal surface of the glomerular endothelium is covered with a glycocalyx, which forms a significant permeability barrier.1,3 The glycocalyx is a negatively charged hydrated mesh of cell-surfaceanchored proteoglycans and adsorbed proteins and glycosaminoglycans in dynamic equilibrium with the plasma.4 Furthermore, the glomerular filtration barrier is now understood to function as an integrated whole:1 disruption of the 949
co m m e nt a r y
Endothelial glycocalyx Fenestrated glomerular endothelium Glomerular basement membrane Podocyte foot processes and interposed slit diaphragms (green) VEGF
Podocyte detachment
VEGF
or
Reduced endothelial glycocalyx Reduced endothelial fenestrations Thickened glomerular basement membrane Increased foot process width with decreased filtration slit frequency
Figure 1 | The glomerular filtration barrier and its disruption in type 2 diabetes. (a) Schematic of the normal glomerular filtration barrier, consisting of fenestrated glomerular endothelium, glomerular basement membrane (GBM), and podocyte foot processes. The endothelium is covered by a layer of glycocalyx: a mesh of cell-surface-anchored proteoglycans (core proteins and carbohydrate side chains, predominantly heparan sulfate and chondroitin sulfate) and adsorbed plasma proteins and glycosaminoglycans. Vascular endothelial growth factor (VEGF) is produced by podocytes and acts to maintain the adjacent glomerular endothelium. (b) Disruption of the glomerular filtration barrier in type 2 diabetes as described by Weil et al.8 includes thickening of the GBM, podocyte foot process widening, and podocyte detachment.7 The overall proportion of fenestrated endothelium is reduced, but, surprisingly, this is not seen opposite areas of podocyte detachment. Disruption of the endothelial glycocalyx is not demonstrated by Weil et al.8 but is strongly implicated. Data regarding bioavailability of VEGF in the diabetic glomerulus are conflicting, but one possibility is that an initial increase is followed by a decrease as the disease progresses.
macromolecular permeability of any one layer affects overall permeability, and cell– cell communication via soluble mediators within the glomerular filtration barrier modulates its behavior. For example, vascular endothelial growth factor (VEGF) is produced by podocytes and is critical to maintenance of the adjacent glomerular endothelium, including regulation of fenestrations.2,5 In this context there has been an increasing interest in the glomerular endothelium and its contribution to glomerular filtration.6 The classical histological features of diabetic nephropathy include mesangial expansion and GBM thickening accompanied by podocyte foot process widening or ‘effacement.’ Later studies have investigated features including podocyte loss, but knowledge of glomerular endothelial behavior in diabetic nephropathy is sparse. A recent study has examined podocyte detachment and loss of endothelial fenestrations along 950
with functional parameters in type 1 diabetes,7 and Weil et al.8 (this issue) now provide detailed correlations between these ultrastructural parameters and glomerular filtration rate (GFR) and albumin-tocreatinine ratio (ACR) in type 2 diabetes. Weil et al. studied Pima Indians with type 2 diabetes, divided into normo-, micro-, and macroalbuminuric groups.8 Clinical characteristics and glomerular structural parameters derived from light and electron microscopy on renal biopsies were compared between these groups, as well as a group of nondiabetic non-Pima-Indian controls. Macroalbuminuric subjects had a longer duration of diabetes and a higher HbA1c level than normoalbuminurics. GFR was significantly higher in microalbuminuria than in the three other groups, but there were no significant blood pressure differences. As expected on the basis of previous studies, glomerular volume, fractional interstitial area, fractional mesangial area, and GBM
width were all higher in diabetes and tended to increase with increasing albuminuria. All diabetics had a reduced filtration surface area density, while the total surface area was increased in microalbuminuria. Podocyte numbers were reduced in diabetes, significantly so in micro- and macroalbuminuria. Similarly, filtration slit frequency was reduced, and foot process width increased, in microalbuminuria and more so in macroalbuminuria. Podocyte detachment was higher in all diabetic groups compared with controls (0.03%), increasing from 0.4% in normoalbuminuria to 1.48% in macroalbuminuria. In type 1 diabetes, 22 % of the GBM was not covered by intact foot processes compared with 2% of controls.7 These values are higher than those obtained by Weil et al.8 probably because the latter measured GBM completely devoid of podocytes, a more stringent measure. Nevertheless, a similar pattern of increasing detachment paralleling increasing albuminuria emerged. The proportion of fenestrated endothelium was reduced from 43.5% in controls to 27.4% and 27.2% in normo- and microalbuminuria, respectively, and further to 19.3% in macroalbuminuria. These observations are comparable to those in type 1 diabetes: 41% in controls, 32% in normo- and microalbuminuria, and 25% in macroalbuminuria.7 Fenestration loss has also been observed in animal models of diabetes.2 Comparing structural parameters within diabetic groups revealed multiple differences between macro- and both microand normoalbuminuric groups in line with the above observations. However, although there were trends, there were no significant differences between normo- and microalbuminuric groups. This suggests that the structural changes observed do not account for microalbuminuria, which may be related to structural differences not measured in this study (see below) or to hemodynamic effects: GFR was higher in micro- than in normoalbuminuria. Correlations between clinical and glomerular parameters, including the data from all subject groups, are of particular interest. GFR had few significant correlations but was correlated positively with glomerular volume and total surface area and negatively with mesangial area. GFR was positively correlated with Kidney International (2012) 82
co m m e nt a r y
percentage fenestrated endothelium but was not significantly correlated with podocyte detachment. Although models predict the importance of fenestrations for maintenance of GFR, this is only the second study to provide direct evidence, the first being a study in preeclampsia.9 ACR, on the other hand, was correlated with classical features of diabetic nephropathy including mesangial expansion and podocyte foot process widening. It was also correlated positively with podocyte detachment, but, counterintuitively, the correlation was less marked than that with foot process widening. This may be because podocyte detachment affected only a small proportion of the filtration surface compared with foot process widening. It is arguable, therefore, that measurement of podocyte detachment adds little in assessment of podocyte damage; however, it does provide an important insight into the evolution of diabetic glomerulopathy, with podocyte detachment likely to be the end result of podocyte dysfunction. ACR correlated more strongly with percentage fenestrated endothelium (negatively in this case) than did podocyte detachment. This leads to the surprising conclusion that the state of the glomerular endothelium more closely reflects both GFR and ACR than podocyte parameters. The relevance of fenestrations to GFR is easy to understand, but what of ACR? Reduction of fenestrations might be expected to lead to a reduced ACR, suggesting that this parameter is an index of disruption of another component of the glomerular filtration barrier, and the most obvious candidate is the endothelial glycocalyx. Although this study provides no direct evidence of glycocalyx disruption, there is ample evidence from other work that this occurs in diabetes.10 Although endothelial fenestrations correlated with various structural parameters, these did not include any of the podocyte parameters. This is despite the fact that comparisons between the four study groups showed that there are similar patterns of change in diabetes: podocyte detachment increased in normo- and microalbuminuria to a similar extent, while fenestrated endothelium decreased to similar extents in both groups. Both changed further in macroalbuminuria. However, when Weil Kidney International (2012) 82
et al.8 investigated their samples further by measuring endothelial fenestration directly across the GBM from detached podocytes, they found no decrease. What are the implications of this study? Firstly, do these observations threaten the paradigm of glomerular endothelial cell dependence on podocyte-produced VEGF?2,5 A number of explanations for the discrepancy present themselves: (1) VEGF levels may be abnormally high in diabetic glomeruli, so podocyte detachment in this context does not necessarily mean reduced VEGF bioavailability; (2) the proportion of GBM totally denuded of podocytes was small (⭐1.48%), and therefore the majority of a given endothelial cell will still be adjacent to an area of GBM where podocyte cytoplasm is present; (3) there may be subtle fenestration abnormalities not detected in this study; and (4) the formation of fenestrations may be at least as dependent on the ability of glomerular endothelial cells to respond to VEGF as on the ability of podocytes to produce it. Indeed, this consideration also raises the reverse question of whether podocyte dysfunction may be downstream of endothelial damage through disordered release of endothelial-cell mediators. In type 1 diabetes fenestral density did correlate with podocyte GBM coverage.7 Secondly, the glomerular endothelial contribution to glomerular filtration barrier pathophysiology is greater than generally appreciated. The importance of fenestrations is clearly demonstrated, and that of the glycocalyx is strongly implied. This study should stimulate more intensive investigation of glomerular endothelial dysfunction in diabetic nephropathy with a view to target therapies to its prevention and reversal. Investigating whether renin–angiotensin system blockade modifies these study observations, and confirming them in non-PimaIndian type 2 diabetics, are obvious next steps. The link that endothelial dysfunction provides between the glomerulus, proteinuria, and systemic vascular disease in diabetes also demands further exploration. Direct confirmation of the importance of the glomerular endothelial glycocalyx will require studies using specialized fixation and staining techniques to label the glycocalyx and investigate its relationship with ACR.4 Higher-resolution imaging will reveal fenestration changes in
more detail in terms of size, density, distribution, and correlation with glycocalyx disturbance. Thirdly, the discordance between ACR and GFR is consistent with other studies and the recognition that decline in renal function may be independent of proteinuria.11 Such observations imply that different mechanisms underlie proteinuria and loss of GFR, so that early decline in GFR may be a better predictor of progressive renal failure than proteinuria. Hence the correlations of a reduced GFR, including, in this study, loss of fenestrations, may be more informative in understanding progressive diabetic nephropathy. Finally, these findings in diabetic nephropathy, in addition to those in preeclampsia,9 suggest that fenestration loss may be important in other glomerular diseases. In the majority of cases this question has not yet been addressed, but observations in transplant glomerulopathy and in a range of animal models suggest that such investigations will be rewarding.2 DISCLOSURE The author declared no competing interests. REFERENCES 1.
Haraldsson B, Nystrom J, Deen WM. Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev 2008; 88: 451–487. 2. Satchell SC, Braet F. Glomerular endothelial cell fenestrations: an integral component of the glomerular filtration barrier. Am J Physiol Renal Physiol 2009; 296: F947–F956. 3. Curry FE, Adamson RH. Endothelial glycocalyx: permeability barrier and mechanosensor. Ann Biomed Eng 2012; 40: 828–839. 4. Reitsma S, Slaaf DW, Vink H et al. The endothelial glycocalyx: composition, functions, and visualization. Pflugers Arch 2007; 454: 345–359. 5. Eremina V, Jefferson JA, Kowalewska J et al. VEGF inhibition and renal thrombotic microangiopathy. N Engl J Med 2008; 358: 1129–1136. 6. Haraldsson B, Nystrom J. The glomerular endothelium: new insights on function and structure. Curr Opin Nephrol Hypertens 2012; 21: 258–263. 7. Toyoda M, Najafian B, Kim Y et al. Podocyte detachment and reduced glomerular capillary endothelial fenestration in human type 1 diabetic nephropathy. Diabetes 2007; 56: 2155–2160. 8. Weil EJ, Lemley KV, Mason CC et al. Podocyte detachment and reduced glomerular capillary endothelial fenestration promote kidney disease in type 2 diabetic nephropathy. Kidney Int 2012; 82: 1010–1017. 9. Lafayette RA, Druzin M, Sibley R et al. Nature of glomerular dysfunction in pre-eclampsia. Kidney Int 1998; 54: 1240–1249. 10. Salmon AH, Satchell SC. Endothelial glycocalyx dysfunction in disease: albuminuria and increased microvascular permeability. J Pathol 2012; 226: 562–574. 11. Macisaac RJ, Jerums G. Diabetic kidney disease with and without albuminuria. Curr Opin Nephrol Hypertens 2011; 20: 246–257.
951
The transfer of pathogenic T cells by Okamoto and colleagues7 also induced increased glomerular complement deposition without clear changes in serum IgG levels, anti-dsDNA levels, or overall glomerular IgG deposition, measured semiquantitatively. Nonetheless, it is possible that the transferred T cells did enhance glomerular deposition of pathogenic autoantibodies with higher affinity or of a more damaging IgG subclass. Potentially, these autoantibodies could be responsible for the increased complement deposition found in the mice that received the pathogenic effector CD4 + T cells. The study by Okamoto et al.7 brings us a step closer to understanding how a variety of mediators, including effector CD4 + T cells, form an unholy alliance to mediate injury in lupus nephritis. The ultimate goal of new treatments, more targeted therapy with fewer side effects, may be more difficult in patients with severe lupus nephritis because of both the potential diversity of target autoantigens and the involvement of multiple arms of the effector immune response. Over the past few years there has been real progress in using newer treatments, in the form of therapies such as mycophenolate and belimumab.10 However, already it has been shown that targeting autoimmunity to a single antigen, dsDNA, is not a useful treatment, even though dsDNA was a logical candidate antigen.10 The diversity of clinical presentations and histological pictures in lupus and lupus nephritis may mean that in the future, treatment might need to be tailored in some patients to predominantly target one arm of the immune response, or even a restricted range of antigens. Further well-designed studies in relevant and well-characterized experimental models of disease based on careful observations in humans need to be performed, so that we can better understand the nature of effector responses in this at-times devastating and difficult-to-treat human disease. DISCLOSURE The authors declared no competing interests. REFERENCES 1.
2.
Murata H, Matsumura R, Koyama A et al. T cell receptor repertoire of T cells in the kidneys of patients with lupus nephritis. Arthritis Rheum 2002; 46: 2141–2147. Kitching AR, Holdsworth SR, Tipping PG. Crescentic glomerulonephritis: a manifestation of
Kidney International (2012) 82
3.
4.
5.
6.
a nephritogenic Th1 response? Histol Histopathol 2000; 15: 993–1003. Kitching AR, Holdsworth SR. The emergence of TH17 cells as effectors of renal injury. J Am Soc Nephrol 2011; 22: 235–238. Summers SA, Steinmetz OM, Li M et al. Th1 and Th17 cells induce proliferative glomerulonephritis. J Am Soc Nephrol 2009; 20: 2518–2524. Diaz Gallo C, Jevnikar AM, Brennan DC et al. Autoreactive kidney-infiltrating T-cell clones in murine lupus nephritis. Kidney Int 1992; 42: 851–859. Schwarting A, Wada T, Kinoshita K et al. IFN-gamma receptor signaling is essential for the initiation, acceleration, and destruction of autoimmune
kidney disease in MRL-Fas(lpr) mice. J Immunol 1998; 161: 494–503. 7. Okamoto A, Fujio K, Tsuno NH et al. Kidneyinfiltrating CD4+ T-cell clones promote nephritis in lupus-prone mice. Kidney Int 2012; 82: 969–979. 8. Hanrotel-Saliou C, Segalen I, Le Meur Y et al. Glomerular antibodies in lupus nephritis. Clin Rev Allergy Immunol 2011; 40: 151–158. 9. Lu L, Kaliyaperumal A, Boumpas DT et al. Major peptide autoepitopes for nucleosome-specific T cells of human lupus. J Clin Invest 1999; 104: 345–355. 10. Yildirim-Toruner C, Diamond B. Current and novel therapeutics in the treatment of systemic lupus erythematosus. J Allergy Clin Immunol 2011; 127: 303–312.
see original article on page 1010
The glomerular endothelium emerges as a key player in diabetic nephropathy Simon C. Satchell1 The effects of diabetes on glomerular structure and particularly that of the glomerular capillary wall have been extensively documented. By correlation with clinical measurements, Weil and colleagues provide important insights into the functional significance of glomerular structural changes in type 2 diabetes. Podocyte detachment correlates with albumin-to-creatinine ratio, but less strongly than does loss of endothelial fenestrations, which also correlates with reduced glomerular filtration rate. The role of the glomerular endothelium in diabetic nephropathy demands further scrutiny. Kidney International (2012) 82, 949–951. doi:10.1038/ki.2012.258
The glomerular filtration barrier is highly permeable to water and small molecules but maintains very low permeability to macromolecules.1 These characteristics are dependent on its unique three-layer structure. Glomerular endothelial cells are characterized by fenestrations, 60- to 80nm transcellular holes through the peripheral cytoplasm, essential for high water permeability.2 The glomerular basement membrane (GBM) separates glomerular endothelial cells from podocytes whose 1Academic Renal Unit, University of Bristol, Bristol, UK Correspondence: Simon C. Satchell, Academic Renal Unit, University of Bristol, Learning & Research, Southmead Hospital, Bristol BS10 5NB, UK. E-mail: [email protected]
interdigitating foot processes are spanned by the slit diaphragm (Figure 1). Until recently, the presence of fenestrations has meant that the glomerular endothelium has been considered to contribute little to the barrier to macromolecules. However, it is now appreciated that the luminal surface of the glomerular endothelium is covered with a glycocalyx, which forms a significant permeability barrier.1,3 The glycocalyx is a negatively charged hydrated mesh of cell-surfaceanchored proteoglycans and adsorbed proteins and glycosaminoglycans in dynamic equilibrium with the plasma.4 Furthermore, the glomerular filtration barrier is now understood to function as an integrated whole:1 disruption of the 949
co m m e nt a r y
Endothelial glycocalyx Fenestrated glomerular endothelium Glomerular basement membrane Podocyte foot processes and interposed slit diaphragms (green) VEGF
Podocyte detachment
VEGF
or
Reduced endothelial glycocalyx Reduced endothelial fenestrations Thickened glomerular basement membrane Increased foot process width with decreased filtration slit frequency
Figure 1 | The glomerular filtration barrier and its disruption in type 2 diabetes. (a) Schematic of the normal glomerular filtration barrier, consisting of fenestrated glomerular endothelium, glomerular basement membrane (GBM), and podocyte foot processes. The endothelium is covered by a layer of glycocalyx: a mesh of cell-surface-anchored proteoglycans (core proteins and carbohydrate side chains, predominantly heparan sulfate and chondroitin sulfate) and adsorbed plasma proteins and glycosaminoglycans. Vascular endothelial growth factor (VEGF) is produced by podocytes and acts to maintain the adjacent glomerular endothelium. (b) Disruption of the glomerular filtration barrier in type 2 diabetes as described by Weil et al.8 includes thickening of the GBM, podocyte foot process widening, and podocyte detachment.7 The overall proportion of fenestrated endothelium is reduced, but, surprisingly, this is not seen opposite areas of podocyte detachment. Disruption of the endothelial glycocalyx is not demonstrated by Weil et al.8 but is strongly implicated. Data regarding bioavailability of VEGF in the diabetic glomerulus are conflicting, but one possibility is that an initial increase is followed by a decrease as the disease progresses.
macromolecular permeability of any one layer affects overall permeability, and cell– cell communication via soluble mediators within the glomerular filtration barrier modulates its behavior. For example, vascular endothelial growth factor (VEGF) is produced by podocytes and is critical to maintenance of the adjacent glomerular endothelium, including regulation of fenestrations.2,5 In this context there has been an increasing interest in the glomerular endothelium and its contribution to glomerular filtration.6 The classical histological features of diabetic nephropathy include mesangial expansion and GBM thickening accompanied by podocyte foot process widening or ‘effacement.’ Later studies have investigated features including podocyte loss, but knowledge of glomerular endothelial behavior in diabetic nephropathy is sparse. A recent study has examined podocyte detachment and loss of endothelial fenestrations along 950
with functional parameters in type 1 diabetes,7 and Weil et al.8 (this issue) now provide detailed correlations between these ultrastructural parameters and glomerular filtration rate (GFR) and albumin-tocreatinine ratio (ACR) in type 2 diabetes. Weil et al. studied Pima Indians with type 2 diabetes, divided into normo-, micro-, and macroalbuminuric groups.8 Clinical characteristics and glomerular structural parameters derived from light and electron microscopy on renal biopsies were compared between these groups, as well as a group of nondiabetic non-Pima-Indian controls. Macroalbuminuric subjects had a longer duration of diabetes and a higher HbA1c level than normoalbuminurics. GFR was significantly higher in microalbuminuria than in the three other groups, but there were no significant blood pressure differences. As expected on the basis of previous studies, glomerular volume, fractional interstitial area, fractional mesangial area, and GBM
width were all higher in diabetes and tended to increase with increasing albuminuria. All diabetics had a reduced filtration surface area density, while the total surface area was increased in microalbuminuria. Podocyte numbers were reduced in diabetes, significantly so in micro- and macroalbuminuria. Similarly, filtration slit frequency was reduced, and foot process width increased, in microalbuminuria and more so in macroalbuminuria. Podocyte detachment was higher in all diabetic groups compared with controls (0.03%), increasing from 0.4% in normoalbuminuria to 1.48% in macroalbuminuria. In type 1 diabetes, 22 % of the GBM was not covered by intact foot processes compared with 2% of controls.7 These values are higher than those obtained by Weil et al.8 probably because the latter measured GBM completely devoid of podocytes, a more stringent measure. Nevertheless, a similar pattern of increasing detachment paralleling increasing albuminuria emerged. The proportion of fenestrated endothelium was reduced from 43.5% in controls to 27.4% and 27.2% in normo- and microalbuminuria, respectively, and further to 19.3% in macroalbuminuria. These observations are comparable to those in type 1 diabetes: 41% in controls, 32% in normo- and microalbuminuria, and 25% in macroalbuminuria.7 Fenestration loss has also been observed in animal models of diabetes.2 Comparing structural parameters within diabetic groups revealed multiple differences between macro- and both microand normoalbuminuric groups in line with the above observations. However, although there were trends, there were no significant differences between normo- and microalbuminuric groups. This suggests that the structural changes observed do not account for microalbuminuria, which may be related to structural differences not measured in this study (see below) or to hemodynamic effects: GFR was higher in micro- than in normoalbuminuria. Correlations between clinical and glomerular parameters, including the data from all subject groups, are of particular interest. GFR had few significant correlations but was correlated positively with glomerular volume and total surface area and negatively with mesangial area. GFR was positively correlated with Kidney International (2012) 82
co m m e nt a r y
percentage fenestrated endothelium but was not significantly correlated with podocyte detachment. Although models predict the importance of fenestrations for maintenance of GFR, this is only the second study to provide direct evidence, the first being a study in preeclampsia.9 ACR, on the other hand, was correlated with classical features of diabetic nephropathy including mesangial expansion and podocyte foot process widening. It was also correlated positively with podocyte detachment, but, counterintuitively, the correlation was less marked than that with foot process widening. This may be because podocyte detachment affected only a small proportion of the filtration surface compared with foot process widening. It is arguable, therefore, that measurement of podocyte detachment adds little in assessment of podocyte damage; however, it does provide an important insight into the evolution of diabetic glomerulopathy, with podocyte detachment likely to be the end result of podocyte dysfunction. ACR correlated more strongly with percentage fenestrated endothelium (negatively in this case) than did podocyte detachment. This leads to the surprising conclusion that the state of the glomerular endothelium more closely reflects both GFR and ACR than podocyte parameters. The relevance of fenestrations to GFR is easy to understand, but what of ACR? Reduction of fenestrations might be expected to lead to a reduced ACR, suggesting that this parameter is an index of disruption of another component of the glomerular filtration barrier, and the most obvious candidate is the endothelial glycocalyx. Although this study provides no direct evidence of glycocalyx disruption, there is ample evidence from other work that this occurs in diabetes.10 Although endothelial fenestrations correlated with various structural parameters, these did not include any of the podocyte parameters. This is despite the fact that comparisons between the four study groups showed that there are similar patterns of change in diabetes: podocyte detachment increased in normo- and microalbuminuria to a similar extent, while fenestrated endothelium decreased to similar extents in both groups. Both changed further in macroalbuminuria. However, when Weil Kidney International (2012) 82
et al.8 investigated their samples further by measuring endothelial fenestration directly across the GBM from detached podocytes, they found no decrease. What are the implications of this study? Firstly, do these observations threaten the paradigm of glomerular endothelial cell dependence on podocyte-produced VEGF?2,5 A number of explanations for the discrepancy present themselves: (1) VEGF levels may be abnormally high in diabetic glomeruli, so podocyte detachment in this context does not necessarily mean reduced VEGF bioavailability; (2) the proportion of GBM totally denuded of podocytes was small (⭐1.48%), and therefore the majority of a given endothelial cell will still be adjacent to an area of GBM where podocyte cytoplasm is present; (3) there may be subtle fenestration abnormalities not detected in this study; and (4) the formation of fenestrations may be at least as dependent on the ability of glomerular endothelial cells to respond to VEGF as on the ability of podocytes to produce it. Indeed, this consideration also raises the reverse question of whether podocyte dysfunction may be downstream of endothelial damage through disordered release of endothelial-cell mediators. In type 1 diabetes fenestral density did correlate with podocyte GBM coverage.7 Secondly, the glomerular endothelial contribution to glomerular filtration barrier pathophysiology is greater than generally appreciated. The importance of fenestrations is clearly demonstrated, and that of the glycocalyx is strongly implied. This study should stimulate more intensive investigation of glomerular endothelial dysfunction in diabetic nephropathy with a view to target therapies to its prevention and reversal. Investigating whether renin–angiotensin system blockade modifies these study observations, and confirming them in non-PimaIndian type 2 diabetics, are obvious next steps. The link that endothelial dysfunction provides between the glomerulus, proteinuria, and systemic vascular disease in diabetes also demands further exploration. Direct confirmation of the importance of the glomerular endothelial glycocalyx will require studies using specialized fixation and staining techniques to label the glycocalyx and investigate its relationship with ACR.4 Higher-resolution imaging will reveal fenestration changes in
more detail in terms of size, density, distribution, and correlation with glycocalyx disturbance. Thirdly, the discordance between ACR and GFR is consistent with other studies and the recognition that decline in renal function may be independent of proteinuria.11 Such observations imply that different mechanisms underlie proteinuria and loss of GFR, so that early decline in GFR may be a better predictor of progressive renal failure than proteinuria. Hence the correlations of a reduced GFR, including, in this study, loss of fenestrations, may be more informative in understanding progressive diabetic nephropathy. Finally, these findings in diabetic nephropathy, in addition to those in preeclampsia,9 suggest that fenestration loss may be important in other glomerular diseases. In the majority of cases this question has not yet been addressed, but observations in transplant glomerulopathy and in a range of animal models suggest that such investigations will be rewarding.2 DISCLOSURE The author declared no competing interests. REFERENCES 1.
Haraldsson B, Nystrom J, Deen WM. Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev 2008; 88: 451–487. 2. Satchell SC, Braet F. Glomerular endothelial cell fenestrations: an integral component of the glomerular filtration barrier. Am J Physiol Renal Physiol 2009; 296: F947–F956. 3. Curry FE, Adamson RH. Endothelial glycocalyx: permeability barrier and mechanosensor. Ann Biomed Eng 2012; 40: 828–839. 4. Reitsma S, Slaaf DW, Vink H et al. The endothelial glycocalyx: composition, functions, and visualization. Pflugers Arch 2007; 454: 345–359. 5. Eremina V, Jefferson JA, Kowalewska J et al. VEGF inhibition and renal thrombotic microangiopathy. N Engl J Med 2008; 358: 1129–1136. 6. Haraldsson B, Nystrom J. The glomerular endothelium: new insights on function and structure. Curr Opin Nephrol Hypertens 2012; 21: 258–263. 7. Toyoda M, Najafian B, Kim Y et al. Podocyte detachment and reduced glomerular capillary endothelial fenestration in human type 1 diabetic nephropathy. Diabetes 2007; 56: 2155–2160. 8. Weil EJ, Lemley KV, Mason CC et al. Podocyte detachment and reduced glomerular capillary endothelial fenestration promote kidney disease in type 2 diabetic nephropathy. Kidney Int 2012; 82: 1010–1017. 9. Lafayette RA, Druzin M, Sibley R et al. Nature of glomerular dysfunction in pre-eclampsia. Kidney Int 1998; 54: 1240–1249. 10. Salmon AH, Satchell SC. Endothelial glycocalyx dysfunction in disease: albuminuria and increased microvascular permeability. J Pathol 2012; 226: 562–574. 11. Macisaac RJ, Jerums G. Diabetic kidney disease with and without albuminuria. Curr Opin Nephrol Hypertens 2011; 20: 246–257.
951