| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
* Medicine,
Pathology, and
Surgery, University of Cambridge, Addenbrooke’s Hospital, Cambridge; Department of Molecular Biomedical Research, Flanders’ Interuniversity Institute for Biotechnology (VIB),
Interdepartmental Program in Vascular Biology and Transplantation, Gent University, Gent, Belgium; and
|| Department of Pathology, Boyer Centre for Molecular Medicine, Yale University School of Medicine, New Haven, Connecticut, USA
1 Correspondence: Department of Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QQ, UK. E-mail: john.bradley{at}addenbrookes.nhs.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: TNF • ASK1 • Etk • apoptosis • proliferation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) in human disease has been highlighted by the efficacy of anti-TNF antibodies or soluble TNF receptors (TNFRs) in controlling disease activity in rheumatoid arthritis and other inflammatory conditions. TNF interacts with two distinct receptors, designated TNFR1 and TNFR2, and understanding the specific role of each receptor in signaling TNF is important for rational use of TNF blockade. The signaling events initiated by each TNFR vary among cell types in culture. Little is known about receptor signaling in situ in different tissues.
In many cell types, and in particular mononuclear cells, TNFR1 is thought to mediate cell death. TNFR2 may enhance TNFR1-induced cell death or promote cell activation, migration, or proliferation (1)
. In cultured endothelial cells (EC), proinflammatory and morphological changes are under the dominant control of TNFR1, although we have found that TNFR2 may contribute to these changes, particularly at low TNF concentrations (2)
These observations are consistent with the notion of "ligand passing" in which TNFR2 captures TNF and passes it to TNFR1 (3)
.
The cytoplasmic sequences of these two receptors share no homology and both are devoid of intrinsic enzyme activity. Instead, TNFR1 and TNFR2 initiate signaling by recruitment of cytosolic proteins through protein-protein interaction domains in their cytoplasmic regions. TNFR1 signals by recruitment to its death domain of TNFR-associated death domain protein (TRADD) (4)
, which serves as a supporting structure for recruitment of TNF-receptor-associated factor 2 (TRAF2) and receptor interacting protein-1 (RIP-1). The signaling complex that is formed leads to activation of transcription factors such as NF-
B and AP-1. TNFR2 does not contain a cytoplasmic death domain although it can interact directly with TRAF2 (5)
, providing a mechanism for some shared activity of TNFRs.
Identification of signaling events that are specific for either TNF receptor subtypes, could provide an important bioassay and diagnostic tool. Apoptosis signaling kinase-1 (ASK1) and endothelial/epithelial tyrosine kinase (Etk) are two distinct kinases with diverse functions that may be differentially activated by TNFRs (Fig. 1
). ASK1 is a 170 kDa protein that contains an inhibitory N-terminal domain, an internal kinase domain, and a C-terminal regulatory domain (6)
that activates multiple proapoptotic pathways in cultured cells. ASK1 activity is controlled by several mechanisms, including protein-protein interactions with thioredoxin (Trx), the dimeric phosphoserine binding molecule 14-3-3, and TRAF2. In EC, Trx binds in reduced form to the N terminus of ASK1 and induces ASK1 ubiquination and degradation (7)
. Generation of reactive oxygen species by TNF or other signals leads to Trx oxidation and dissociation from ASK1. ASK1 activation is also regulated by reversible phosphorylation in its C-terminal domain, which binds to the TRAF domain of TRAF2 (8)
. Phosphorylation of Thr845 is required for ASK1 activation (9)
whereas phosphorylation at Ser967 appears to inhibit the functions of ASK1. The scaffold protein 14-3-3 binds to ASK1 via Ser-967 in the C-terminal domain (10)
. ASK1 interacting protein-1 (AIP1), a member of the Ras GTPase-activating protein (Ras-GAP) family, promotes TNF-induced activation of ASK1 by facilitating dissociation of ASK1 from 14-3-3. The dissociation of ASK1 from 14-3-3 is accompanied by dephosphorylation of ASK1 at Ser-967 by an unknown phosphatase (11)
. Thus, ASK1 activation can be assessed by loss of phosphorylation at Ser 967 and the simultaneous acquisition of phosphorylation at Thr845.
|
Etk appears to be a TNFR2-specific kinase that has been implicated in cell adhesion, migration, proliferation, and survival (12)
, (13)
. In epithelial cells, Etk may be a novel regulator of cell junctions (14)
. In vascular EC Etk is involved in TNF-induced angiogenic events (6)
, (15)
. More specifically, Etk mediates activation of the phosphatidylinositol 3 kinase (PI3K) -Akt angiogenic pathway, which has been well documented in growth factor-stimulated cell migration (16)
. TNF activates Etk through TNFR2 in a TRAF2-independent manner. TNFR2 (but not TNFR1) associates with an inactive form of Etk in a ligand-independent fashion through the C-terminal 16-amino acid sequence of TNFR2 and multiple domains of Etk. TNF is thought to induce a conformational change in TNFR2 that triggers unfolding of the closed, inactive form of Etk. In EC, TNF induces assembly of a trimolecular complex containing TNFR2, Etk, and vascular endothelial growth factor receptor 2 (VEGFR2, also known as KDR or flk-1). Within this complex, there is a coordinate reciprocal phosphorylation of Etk and VEGFR2, resulting in PI3K activation (6)
. The appearance and phosphorylated Etk in EC is indicative of TNFR2 signaling.
We previously reported that TNFR1 and TNFR2 are expressed by different cell types in normal kidney and that TNFR expression patterns are modulated during immune mediated and ischemic renal injury (17)
. This regulated expression of TNF receptors on cells in the kidney suggests that cultured cells do not accurately replicate TNF receptor expression patterns observed in situ. In this study we have exploited TNFR-specific recruitment and activation of the cytosolic kinases ASK1 and Etk in a kidney organ culture model to determine the effects of the two TNFRs on kinase phosphorylation. Using wild-type TNF and receptor-specific muteins, we show that ASK1 and Etk are differentially activated at distinct sites in the kidney and that these changes correlate with distinct cellular responses. These results provide for the first time assays to determine the status of TNFR signaling in tissue samples, and we have applied these assays to renal allograft biopsies showing rejections and/or ischemic injury.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Kidney organ cultures
All experiments using human tissue were performed with the written informed consent of patients and the approval of the local Ethical Committee and Addenbrooke’s Hospital Tissue Bank. Renal tissue for organ culture was obtained from six kidney allograft biopsies taken immediately after reperfusion of renal transplants (time zero biopsy) or from the uninvolved pole of six kidney specimens excised because of renal tumors. Duplicate 1 mm3 fragments were placed in corning flat-bottomed 96-well tissue culture plates (Appleton Woods Limited, Selly Oak, Birmingham, UK) and immediately immersed in medium 199 (Flow, Irvine, Scotland, U.K) containing 10% heat-inactivated fetal calf serum (TCS, Botolph Claydon, Bucks, U.K) and 2.2 mM glutamine. To assess the reliability and reproducibility of these assays, multiple samples of cross sections from the medulla through to the cortex were taken from each patient to obtain randomized samples; parallel group comparison, and all samples were incubated in duplicate.
Tissue was incubated for 3 h at 37°C with either culture media alone without TNF or with 10 ng/mL of wild-type TNF (AMS Biotechnology, Europe, Ltd., Abingdon Oxon, UK) or 10 ng/mL of recombinant mutations of the wild-type TNF sequence, which enable the mutated protein ("mutein") to bind selectively to either of the TNFR subtypes (18
, 19)
. The specific double mutation of R32W, S86T (here termed R1-TNF) allows selective activation of TNFR1, whereas the D143N, A145R (termed R2-TNF) double mutation allows selective activation of the TNFR2 subtype only. Half of the harvested tissue was cryoprotected in 30% sucrose in 0.1M phosphate buffer and snap frozen in isopentane-cooled in liquid nitrogen and half was immersed in 4% paraformaldehye in 0.1M PIPES buffer pH 7.6 for 1.5 h at 4°C and processed for paraffin wax embedding and H&E staining.
Light microscopy
Single immunolabeling
Cryosections 8 µm-thick from kidney organ cultures and from normal kidney and renal allografts were permeabilized in cold methanol at –20°C for 5 min, washed in Milli-Q water, and rinsed in 0.1 M Tris-HCl buffer pH 7.5 containing 0.01% TWEEN-20 (TBS) before incubation with blocking buffer [containing 10% fetal calf serum in TBS] (Sigma-Aldrich Company Ltd., Fancy Road, Poole, Dorset, England) for 10 min. Excess fluid was removed and sections were incubated with primary antibodies at 4°C overnight, all at 1:100 dilution in blocking buffer; rabbit polyclonal raised against ASK1 phosporylated at Ser967 (anti-ASK1pSer-967; Cat #3764; Cell Signaling, New England BioLabs (UK) Ltd., Wilbury way, Hitchin, Hertfordshire, UK), rabbit polyclonal raised against ASK1 phosphorylated at Thr845 (anti-ASK1pThr845; Cat no:#3765; Cell Signaling), goat polyclonal anti-Etk (C-17; sc-8874, Bioclear UK Ltd., Mile Elm Calne, Wiltshire, UK), rabbit polyclonal anti-phospho-Etk-tyr40 (Etkp) (Cat #3211, New England Biolabs UK Ltd., Wilbury way, Hitchin, Hertfordshire, UK), mouse-anti-human CD54 (ICAM-1) (Cat # MAB2130; Chemicon International Ltd., Cardinal way, Harrow, Middlesex, UK), mouse monoclonal anti-proliferative cell nuclear antigen (PCNA) (Chemicon). After three 5 min washes, the sections were incubated at room temperature for 1 h in a secondary antibody diluted 1:100 in blocking buffer; Texas Red-conjugated goat anti-rabbit (Vector Laboratories Ltd., Bretton, Peterborough, UK) or Texas Red-conjugated rabbit anti-goat or horse anti-mouse-fluoresecent isothiocynate (FITC). Sections stained for PCNA were further incubated for 10 min with 1:1000 dilution of To-PRO-3'iodide (Molecular Probes, Eugene, Oregon, USA) for nuclei detection. Sections were mounted in Vectashield Mounting Medium (Vector) and examined with a Leica TCS-NT Confocal Laser Scanning Microscope (CLSM, Leica Microsystems, Milton Keynes, UK). For controls, the primary antibody was replaced by either non-immune serum or isotype-specific antisera and all steps were followed unchanged.
Combined immunolabeling
Sections were incubated at 4°C overnight with 1:100 dilution in blocking buffer of rabbit polyclonal anti-ASK1 (anti-pSer967 or anti-pThr845) and 1:500 dilution of mouse monoclonal anti-CD31 (Dakocytomation) or 1:20 dilution of mouse monoclonal anti-TNFR1 (IgG1/Clone: 16803.7; R&D Systems, Oxford, UK). Some sections were incubated with either goat polyclonal anti-Etk or rabbit polyclonal anti-phospho-Etk-tyr40 at 1:100 dilutions and mouse monoclonal anti-TNFR2 (Cat# MAB226; IgG2a/Clone: 22221.311 R&D systems) at 1:20 dilution or with a mouse monoclonal anti-PCNA antibody at 1:100 dilution overnight at 4°C. After 5 min (x3) washes, sections were further incubated for 1 h at room temperature with 1:100 dilutions of secondary antibody in blocking buffer; Texas Red-conjugated rabbit anti-goat or goat anti-rabbit and FITC-conjugated horse anti-mouse (Dakocytomation). Sections were mounted in Vectashield Mounting Media and imaged with CLSM as described previously. Controls included use of isotype-specific primary antibody or non-immune serum.
In situ hybridization
Nonradioactive in situ hybridization was carried out on 5 µm-thick paraffin wax sections of kidney organ cultures as described previously (17)
. Single-stranded anti-sense DNA oligonucleotide probes 5'-end labeled with digoxigenin specific for TNFR1 (gb/M60275/HUMTNFRP, 476-515) and for TNFR2 (gb/M55994/HUMTNFR2, 844-873) (MWG-Biotech AG, UK) were used. Negative controls included incubation of sections with a sense probe to either TNFR1 and TNFR2 (MW Biotech-AG, UK).
Terminal deoxynucleotidyl transferase (TdT) -mediated-digoxigenin-11-dUTP nick-end labeling (TUNEL)
Apoptotic cells were detected using TUNEL method as described (20)
. After dewaxing, paraffin wax sections from six different samples of kidney organ cultures from each of the four treatments were incubated with 50 µg/mL Proteinase-K (Roche Diagnostics, Nutley, NJ, USA), pH 7.5, for 8 min in room temperature. Sections were washed in Milli-Q water and exposed to TdT buffer [containing 200 mM potassium cacodylate, 25 mM Tris-HCI, 0.25 mg/m bovine serum albumin (BSA), 5 mM cobalt chloride, pH 6.6] for 5 min, and incubated in a moist chamber with a mixture of TdT [0.05-0.2 U/uL] and digoxigenin-11-dUTP (Roche) in TdT buffer for 30 min at 37°C. Sections were then washed in TB buffer [containing 30 mM sodium citrate, 300 mM sodium chloride] for 15 min in room temperature, rinsed with Milli-Q water, and incubated in TBS-FCS for 10 min. The sections were then incubated for 1 h with alkaline phosphatase-conjugated anti-digoxigenin-11-dUTP antibody (Roche Diagnostics). Antibody binding sites were visualized using Fast Red substrate kit (K0699, Dakocytomation) and the color developed microscopically. All sections were counterstained with 1% aqueous methyl green (Sigma, UK). Negative controls included omission of the TdT enzyme; positive controls included pretreatment of sections with 0.1 U/µL deoxynuclease-1 (DNase-1) (Roche Diagnostics) before TdT staining and staining of human tonsils.
Apoptotic and proliferative indices
The number of dead tubular cells in the cortex was counted on TUNEL stained sections from 6 different samples of kidney organ cultures from each of the 4 treatments. In a view field at a magnification of x235, the total number of dead tubular cells was counted in at least 30 tubules/field in each sample and the apoptotic index (AI) averaged for each treatment. TUNEL-positive cells within the glomeruli were not counted. In addition PCNA-positive nuclei in tubules was counted on 10 representative fields at a magnification of x235. The proliferative index (PI) was calculated as a percentage of PCNA-positive nuclei averaged for each treatment.
Statistical analysis
Statistical significance, defined as P <0.05, was determined for each TNF treatment compared with no TNF-treated cultures using paired Student’s t test. All values are given as mean ± SE.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A–C). Coexpression for TNFR1 and ASK1pThr845 was not detected on sections of normal kidney (Fig. 2D-F
). Staining for TNFR2 was confined to isolated cells in glomeruli and interstitium, with a strong signal for Etk also present in glomerular EC (Fig. 2-I
). TNFR2 and Etkp were not detected at other sites in normal kidney (Fig. 2K
J–L), and no signal was observed when the primary antibody was replaced by non-immune serum (data not shown). Similar patterns of immunostaining were seen in all 9 samples of kidney showing normal histology.
|
Expression of TNFRs, ASK1 and Etk in renal allografts with rejection or ATN
We next analyzed tissue from renal allografts, in which there is differential expression of TNFR1 and TNFR2 at different anatomical sites within the kidney during rejection or ischemic injury (17)
. In all five renal allografts with evidence of acute cellular rejection, active ASK1pThr845 was strongly demonstrated in glomerular and peritubular capillaries EC and, in some (25–50%) tubular epithelial cells (Fig. 3
A, with a strong signal for TNFR2 and Etkp observed on the majority (60–80%) of tubular epithelial cells (Fig. 3B, C
). In contrast, inactive ASK1pSer967 was not detected on these sections.
|
In all five renal allografts with ATN, there was a loss of signal for ASK1pSer967, but a strong signal for ASK1pThr845 was present on some (50–75%) tubular epithelial cells (Fig. 3D
). In ATN, there was TNFR2 up-regulation on tubular epithelial cells where it colocalized with Etkp (Fig. 3E, F
). There were a few interstitial mononuclear cells positive for Etkp but negative for TNFR2.
In all seven renal allografts showing evidence of rejection and ATN, ASK1pSer967 was largely absent and expression of ASK1pThr845 on glomerular EC and tubular epithelial cells was less marked than in rejection without ATN. Colocalization for Etkp and TNFR2 was also observed on tubular epithelial cells (data not shown).
Expression and phosphorylation of ASK1 by wild-type TNF and TNF muteins in kidney organ culture
Kidney tissue maintained in culture for 3 h without TNF showed similar pattern of expression of ASK1, Etk, and TNFRs as normal kidney. Colocalization of TNFR1 and ASK1-pSer967 (Fig. 4
A, B) but not ASK1-pThr845 (Fig. 4D, E
) was demonstrated in EC of glomerular and peritubular capillaries, and on vascular EC in a few blood vessels, which were also reactive for TNFR1. With the exception of a few vascular EC, ASK1pSer967 was absent in R1-TNF-treated cultures (Fig. 4F, G
), but coexpression of TNFR1 and ASK1pThr845 was detected in EC of glomerular and peritubular capillaries (Fig. 4H, I
), with weak signal for ASK1pThr845 also present in few tubular epithelial cells (data not shown). In sections of R2-TNF-treated cultures, colocalization of TNFR1 and ASK1pSer967 (Fig. 4J, K
) but not of ASK1pThr845 (Fig. 4L, M
) was detected on isolated cells in glomeruli and interstitium. There were a few (5%) tubular cells that coexpressed TNFR1 and ASK1pThr845. Wild-type TNF-treated cultures demonstrated a weak coexpression of TNFR1 and ASK1pSer967 (Fig. 4N, O
) and a strong coexpression of TNFR1 and ASKpThr845 (Fig. 4P, Q
) in EC of glomerular and peritubular capillaries.
|
Expression and phosphorylation of Etk by wild-type TNF and TNF muteins in kidney organ culture
Kidney tissue cultured without TNF demonstrated a moderate signal for TNFR2 on isolated cells within glomeruli and interstitium (Fig. 5
A, C), with a moderate signal for Etk (Fig. 5B
) but not Etkp (Fig. 5D
) in glomerular EC. R1-TNF-treated cultures revealed a similar pattern of TNFR2 and Etk to tissue cultured without TNF (Fig. 5E, F
), with a stronger signal and coexpression of TNFR2 and Etkp detected on isolated cells in glomeruli (Fig. 5G, H
). R2-TNF-treated cultures demonstrated new expression of TNFR2 on tubular epithelial cells, where it colocalized with Etk (Fig. 5I, J
) and Etkp (Fig. 5K, L
). Etkp was evident on 20–40% of TNFR2-expressing tubular epithelial cells. No signal for TNFR2 or Etkp was detected on glomeruli, but a weak signal was occasionally seen on vascular EC, and on interstitial mononuclear cells (data not shown). A strong coexpression for TNFR2 and Etk (Fig. 5M, N
) and Etkp (Fig. 5O, P
) was also demonstrated on tubular epithelial cells in wild-type TNF-treated cultures and in mononuclear cells in glomeruli.). Expression for ICAM-1 increased with intensity in EC of glomerular, peritubular capillaries and blood vessels, and in tubular epithelial cells in kidney organ cultures treatment with wild-type TNF and R1-TNF and R2-TNF (data not shown), which served as a useful internal control for tissue viability and for the detection technique.
|
Wild-type TNF and TNF muteins differentially regulate TNFR gene expression in kidney organ culture
To determine whether R2-TNF and wild-type TNF up-regulate TNFR2 in tubular epithelial cells through new gene expression, tissues from organ cultures were analyzed for gene expression. Expression of TNFR1 mRNA (Fig. 6
A, C) but not TNFR2 mRNA (Fig. 6B, D
) was demonstrated in glomerular EC on cultures without TNF and on R1-TNF-treated cultures. No signal for TNFR1 or TNFR2 mRNA was observed in tubular epithelial cells or in vascular EC. In contrast, R2-TNF-treated cultures showed no signal for TNFR1 mRNA (Fig. 6E
) but a strong signal for TNFR2 mRNA was detected in epithelial cells of the distal convoluted tubules and, in proximal convoluted tubules (Fig. 6F
). Glomeruli were constantly negative. Wild-type TNF-treated cultures showed a strong signal for TNFR1 mRNA in glomerular EC and in interstitial mononuclear cells (Fig. 6G
), and for TNFR2 mRNA in tubular epithelial cells (Fig. 6H
). No mRNA signal was detected after hybridization with a sense probe to either TNFR1 or TNFR2 (data not shown).
|
Wild-type TNF and TNF muteins cause different levels of cell death in kidney tissue in organ culture
The presence of cell death in kidney tissue was examined on H&E sections and by TUNEL staining. Tissue incubated for 3 h in culture media without TNF (control) showed normal histology with negative TUNEL reaction (Fig. 7
A, B). Evidence of increased cell death was observed on all TNF-treated kidney organ cultures compared with controls (R1-TNF vs. control P<0.01; R2-TNF vs. control P<0.05; wild-type TNF vs. control P<0.01); there was more cell death in cultures treated with R1-TNF (Fig. 7C, D
) than with R2-TNF (Fig. 7E, F
) (P<0.05), and more cell death in wild-type TNF-treated cultures (Fig. 7G, H
) than with R1-TNF (P<0.05) or R2-TNF (P<0.01). Apoptotic index was derived from the average number of TUNEL-positive tubular cells from six different organ culture sample from each of the 4 treatments and the results are summarized in Fig. 7I
.
|
Wild-type TNF and TNF muteins increase expression of PCNA in renal cells in kidney organ culture
Kidney organ cultures were examined for evidence of cell proliferation using antibody to PCNA, and some sections colocalized for Etkp. Kidney organ tissue cultured without TNF and with R1-TNF demonstrated occasional PCNA-positive nuclei in tubules, which were negative for Etkp (data not shown). Wild-type TNF (data not shown) and R2-TNF cultures (Fig. 8
A–C) demonstrated a strong expression for PCNA in nuclei of some tubular epithelial cells, some cells which were also reactive for Etkp (data not shown). No staining for PCNA was demonstrated in glomeruli. Proliferative index (PI) was significantly high in cultures treated with wild-type TNF (63%) (P<0.01) and R2-TNF (50%) (P<0.05) but not with R1-TNF (20%) or with no TNF treatment (10%). Calculation for the PI was based on the average percentage of PCNA-positive nuclei in tubules 6 different organ culture sample from each of the 4 treatments, and the results are summarized in Fig. 8D
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Previous experiments using bovine aortic endothelial cells had indicated that TNF induces ASK1-JNK activation through TNFR1 (21)
, whereas Etk is a TNFR2-specific kinase (15)
. Our new studies extend this work to show that ASK1 is coordinately expressed with TNFR1 and that Etk is coordinately expressed with TNFR2 in situ. In addition, experiments using TNF muteins in a kidney organ culture model establish that TNFR1 and TNFR2 cause distinct cellular responses at different sites within the kidney. Treatment of normal kidney with R1-TNF results in loss of inactive ASK1pSer967 and appearance of active ASK1pThr845 in glomerular EC and peritubular capillaries. Treatment with R1-TNF causes more cell death than R2-TNF. R2-TNF up-regulates TNFR2 and causes up-regulation and phosphorylation of Etk in tubular epithelial cells, which is associated with increased expression of PCNA. The initial cell culture experiments in large vessel-derived EC differ from the in situ setting since both receptors are expressed in the same cell. We have recently observed that cultured human dermal microvascular EC express TNFR1 and ASK-1 but not TNFR2 and Etk, similar to glomerular EC in situ. Treatment of cultured human dermal microvascular EC with R1-TNF induced Thr845 phosphorylation of ASK1 confirming that TNF activates ASK1 through TNFR1 in microvascular EC (unpublished observations, W. Min), consistent with our finding that TNFR1 activates ASK1 in microvascular EC in kidney organ culture. Up-regulation of TNFR2 by R2-TNF is associated with induction of mRNA for TNFR2, and down-regulation of TNFR1 mRNA and protein. It is puzzling that R2-TNF appears to signal before the expression of its receptor. It is possible that the R2-TNF interacts with low levels of TNFR2 expressed on tubular epithelial cells, or that R2-TNF effects are initially mediated by TNFR2-expressing mesangial and interstitial cells present in normal kidney tissue. We have not observed induction of TNFR2 in cultured human dermal microvascular EC in cell culture in response to TNF (unpublished observations, W. Min).
Although R2-TNF does not Thr phosphorylate ASK1 in kidney organ culture it does induce some cell death, albeit to a significantly lesser extent than R1-TNF or wild-type TNF. The mechanisms by which TNFR2 may contribute to ASK1-independent TNF-induced cell death are unclear. Some studies suggest that TNFR2-mediated death requires TNFR1 (22)
, but studies in T cells suggest that TNFR2 may trigger cell death independently of TNFR1 (23)
; (22)
.
We have extended our organ culture studies to characterize the expression of TNFR-related signaling molecules in human renal allograft biopsies. In acute cellular rejection TNFR1 and ASK1pSer967 are lost from glomerular EC, but staining for ASK1pThr845 is seen in these cells. TRAF2 colocalizes with TNFR1 and ASK1 in normal kidney and in rejecting allografts (unpublished observations). Based on our observations using organ culture, we believe that the change in ASK1 phosphorylation state is indicative of enzyme activation mediated through TNFR1, which occurs before the loss of TNFR1. However, we cannot rule out that some other receptor system is contributing to this change. TNFR2 is up-regulated in tubular epithelial cells during allograft rejection and ATN, where it colocalizes with phosphorylated Etk. ASK1pThr845 is also found in tubular epithelial cells of rejecting and ischemic kidney in the absence of TNFR1. However, Trx-1 and Trx-2 in tubular epithelial cells of rejecting and ischemic kidney in the presence of nitrotyrosine indicative of oxidant injury (unpublished observations) raise the possibility that ASK1 at these sites may be activated by TNF-independent signals such as oxidative stress. These changes are associated with the presence of TUNEL-positive tubular epithelial cells characterized morphologically at the electron microscopy level by condensed and fragmented nuclei (data not shown).
Our observations also provide important insights into downstream effects of ASK1 activation and renal allograft injury. In both allograft rejection and ATN, the principal target for injury is tubular epithelium. Although ASK1pThr845 is found in both glomerular endothelial and tubular epithelial cells of inflamed or ischemic kidney, expression of AIP1 occurs predominantly on tubular and interstitial cells where cell death predominantly occurs (unpublished observations). We have recently shown that AIP1 specifically interacts with the effector domain RING finger of TRAF2 to enhance TNF-induced ASK1-JNK but to inhibit IKK-NF-B signaling (21), highlighting a synergistic effect of ASK1 and AIP1 in induction of cell apoptosis.
Overall these results confirm that TNF can alter the phosphorylation state of ASK1 and Etk at distinct sites within the kidney, leading to different pathophysiological responses. Ser967 dephosphorylation and Thr845 phosphorylation of ASK1, through TNFR1, may cause tissue injury by providing proinflammatory signals and/or promote cell death. In contrast, Etk phosphorylation in ischemic injury or allograft rejection may provide an important signal for tubular cell regeneration by promoting cell proliferation. In addition, the results provide for the first time assays to determine the status of TNFR signaling in tissue samples, which will allow us to further determine the critical roles of TNF signaling molecules in pathological settings.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 25, 2005. Accepted for publication June 13, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. S. Al-Lamki, J. Wang, A. M. Tolkovsky, J. A. Bradley, J. L. Griffin, S. Thiru, E. C.Y. Wang, E. Bolton, W. Min, P. Moore, et al. TL1A Both Promotes and Protects from Renal Inflammation and Injury J. Am. Soc. Nephrol., May 1, 2008; 19(5): 953 - 960. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Luo, Y. Luo, Y. He, H. Zhang, R. Zhang, X. Li, W. L. Dobrucki, A. J. Sinusas, W. C. Sessa, and W. Min Differential Functions of Tumor Necrosis Factor Receptor 1 and 2 Signaling in Ischemia-Mediated Arteriogenesis and Angiogenesis Am. J. Pathol., November 1, 2006; 169(5): 1886 - 1898. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
