Activation of Tubular Epithelial Cells in Diabetic Nephropathy

  1. Michael Morcos1,
  2. Ahmed A.R. Sayed1,
  3. Angelika Bierhaus1,
  4. Benito Yard2,
  5. Rüdiger Waldherr3,
  6. Wolfgang Merz4,
  7. Ingrid Kloeting5,
  8. Erwin Schleicher6,
  9. Stefani Mentz1,
  10. Randa F. Abd el Baki1,
  11. Hans Tritschler7,
  12. Michael Kasper8,
  13. Vedat Schwenger1,
  14. Andreas Hamann1,
  15. Klaus A. Dugi1,
  16. Anne-Marie Schmidt9,
  17. David Stern9,
  18. Reinhard Ziegler1,
  19. Hans U. Haering6,
  20. Martin Andrassy1,
  21. Fokko van der Woude2 and
  22. Peter P. Nawroth1
  1. 1Department of Internal Medicine 1 and Department of Nephrology, University of Heidelberg, Heidelberg, Germany
  2. 2Department of Nephrologie, University Hospital of Mannheim, Mannheim, Germany
  3. 3Gemeinschaftspraxis für Pathologie, Heidelberg, Germany
  4. 4Biochemiezentrum, University of Heidelberg, Germany
  5. 5Department of Laboratory Animal Science, Institute of Pathophysiology, Faculty of Medicine, University Greifswald, Greifswald, Germany
  6. 6Department of Medicine, University of Tuebingen, Tuebingen, Germany
  7. 7Asta Medica, Frankfurt, Germany
  8. 8Department of Anatomy, Department of Pathology and Institute of Food Chemistry, Technical University Dresden, Dresden, Germany
  9. 9College of Surgeons, Columbia University, New York, New York


    Previous studies have shown that renal function in type 2 diabetes correlates better with tubular changes than with glomerular pathology. Since advanced glycation end products (AGEs; AGE-albumin) and in particular carboxymethyllysine (CML) are known to play a central role in diabetic nephropathy, we studied the activation of nuclear factor κB (NF-κB) in tubular epithelial cells in vivo and in vitro by AGE-albumin and CML. Urine samples from healthy control subjects (n = 50) and type 2 diabetic patients (n = 100) were collected and tested for excretion of CML and the presence of proximal tubular epithelial cells (pTECs). CML excretion was significantly higher in diabetic patients than in healthy control subjects (P < 0.0001) and correlated with the degree of albuminuria (r = 0.7, P < 0.0001), while there was no correlation between CML excretion and HbA1c (r = 0.03, P = 0.76). Urine sediments from 20 of 100 patients contained pTECs, evidenced by cytokeratin 18 positivity, while healthy control subjects (n = 50) showed none (P < 0.0001). Activated NF-κB could be detected in the nuclear region of excreted pTECs in 8 of 20 patients with pTECs in the urine sediment (40%). Five of eight NF-κBp65 antigen-positive cells stained positive for interleukin-6 (IL-6) antigen (62%), while only one of the NF-κB-negative cells showed IL-6 positivity. pTECs in the urine sediment correlated positively with albuminuria (r = 0.57, P < 0.0001) and CML excretion (r = 0.55, P < 0.0001). Immunohistochemistry in diabetic rat kidneys and a human diabetic kidney confirmed strong expression of NF-κB in tubular cells. To further prove an AGE/CML-induced NF-κB activation in pTECs, NF-κB activation was studied in cultured human pTECs by electrophoretic mobility shift assays (EMSAs) and Western blot. Stimulation of NF-κB binding activity was dose dependent and was one-half maximal at 250 nmol/l AGE-albumin or CML and time dependent at a maximum of activation after 4 days. Functional relevance of the observed NF-κB activation was demonstrated in pTECs transfected with a NF-κB-driven luciferase reporter plasmid and was associated with an increased release of IL-6 into the supernatant. The AGE- and CML-dependent activation of NF-κBp65 and NF-κB-dependent IL-6 expression could be inhibited using the soluble form of the receptor for AGEs (RAGE) (soluble RAGE [sRAGE]), RAGE-specific antibody, or the antioxidant thioctic acid. In addition transcriptional activity and IL-6 release from transfected cells could be inhibited by overexpression of the NF-κB-specific inhibitor κBα. The findings that excreted pTECs demonstrate activated NF-κB and IL-6 antigen and that AGE-albumin and CML lead to a perpetuated activation of NF-κB in vitro infer that a perpetuated increase in proinflammtory gene products, such as IL-6, plays a role in damaging the renal tubule.


    • Address correspondence and reprint requests to Michael Morcos, MD, Department of Internal Medicine 1, University of Heidelberg, Bergheimerstr 58, 69115 Heidelberg, Germany. E-mail: michael_morcos{at}

      Received for publication 25 February 2002 and accepted in revised form 13 September 2002.

      M.M. and A.A.R.S. contributed equally to this study.

      AGE, advanced glycation end product; CC, contingency coefficient; CML, carboxymethyllysine; ELISA, enzyme-linked immunosorbent assay; EMSA, electrophoretic mobility shift assay; IκB, inhibitory κB; LPS, lipopolysaccharide; NF-κB, nuclear factor κB; pTEC, proximal tubular epithelial cell; RAGE, receptor for AGEs; sRAGE, soluble RAGE; TA, thioctic acid.

    | Table of Contents