original article http://www.kidney-international.org & 2007 International Society of Nephrology

Different roles of TIR8/SIGIRR on toll-like receptor signaling in intrarenal antigen-presenting cells and tubular epithelial cells M Lech1, C Garlanda2, A Mantovani2, CJ Kirschning3, D Schlo¨ndorff1 and H-J Anders1

1Nephrological Center, Medical Policlinic, University of Munich, Munich, Germany; 2Istituto Clinico Humanitas and Fondazione Humanitas per la Ricerca, Rozzano, Italy and 3Institute of Medical Microbiology, Immunology, and Hygiene, Technical University of Munich, Munich, Germany

Toll-like receptors (TLRs) exist on both myeloid and intrinsic The toll-like receptor (TLR)/interleukin 1 receptor (IL-1R) renal cells contributing to the initiation of innate immunity superfamily has a central role for initiating innate anti- during renal with uropathogenic Escherichia coli. microbial immunity and, hence, may contribute to renal Toll–interleukin 1 receptor (IL-1R) (TIR)8/SIGIRR is an orphan pathology in infective pyelonephritis.1,2 In fact, renal tubular receptor of the TLR/IL-1R family, which suppresses TLR epithelial cells express TLR1-4 and -11 and can produce signaling of immune cells and is highly expressed in the proinflammatory and chemokines in response to . Lack of TIR8/SIGIRR is associated with enhanced renal respective TLR ligands.3–6 Furthermore, renal dendritic cells chemokine signaling upon exposure to contribute to innate immunity in the kidney as they are (LPS). This was because of TIR8/SIGIRR expression on resident activated via TLRs either directly by pathogens or by intrarenal myeloid cells rather than tubular epithelial cells endogenous molecules such as Tamm–Horsfall .7 which express it on basolateral and luminal membranes. The Uropathogenic Escherichia coli (UPEC) is the most common lack of TIR8/SIGIRR does not enhance TLR/IL-1R signaling in pathogen involved in infective pyelonephritis and UPEC tubular epithelial cells as was observed in . TIR8/ lipopolysaccharide (LPS) ligates TLR4 as part of the LPS SIGIRR is induced in monocytes treated with LPS or tumor receptor complex.8 We have recently reported experiments necrosis factor and interferon-c in a dose-dependent manner with TLR4 bone marrow chimeric mice which showed that but was downregulated in treated tubule epithelial cells. This TLR4 on intrinsic renal cells as well as on myeloid cells cell type-specific regulation and function did not relate to contributes to renal chemokine signaling and subsequent mRNA splice variants but was associated with N- and O- renal abcess formation in infective pyelonephritis with of the receptor in renal cells of myeloid and UPEC.9 nonmyeloid origin. Our studies show that resident myeloid What are the molecular mechanisms that control renal cells contribute to TLR-mediated antimicrobial immunity in TLR signaling? Excessive TLR signaling may lead to severe the kidney and that this function is controlled by TIR8/SIGIRR. inflammation and inappropriate immunity-related tissue TIR8/SIGIRR does not suppress TLR signaling in tubular damage.10 In fact, a number of negative regulators of TLRs epithelial cells, which supports their role as sensors of have been identified, which include splice variants of microbial infection in the kidney. signaling molecules, soluble TLRs, cleavage enzymes, and Kidney International (2007) 72, 182–192; doi:10.1038/sj.ki.5002293; receptors of the TLR/IL-1R family, T1/ST2 and toll–IL- published online 2 May 2007 1R(TIR)8/SIGIRR.11–13 The Tir8/Sigirr was identified by KEYWORDS: SIGIRR; toll-like receptor; lipopolysaccharide; infection searching EST databases for TIR domain-containing se- quences of yet unknown members of the TLR/IL-1R family.14,15 TIR8/SIGIRR is the only TIR domain-containing member of the TLR/IL-1R superfamily that has a single extracellular immunoglobulin (Ig) domain.14,15 The intracel- lular domain of TIR8/SIGIRR differs from the typical C terminus of TLR/IL-1Rs and is similar to that of the Drosophila toll protein.14 Neither IL-1 nor other ligands bind to TIR8/SIGIRR. TIR8/SIGIRR does not activate Correspondence: H-J Anders, Medizinische Poliklinik, Universita¨tMu¨nchen, 14,16 Pettenkoferstr. 8a, Mu¨nchen 80336, Germany. nuclear factor-kB (NF-kB), most probably because it E-mail: [email protected] does not retain two amino acids (Ser447 and Tyr536) in the 14 Received 12 October 2006; revised 2 February 2007; accepted 8 March highly conserved TIR domain. TIR8/SIGIRR was proposed 2007; published online 2 May 2007 as an endogenous inhibitor of TLR signaling, because

182 Kidney International (2007) 72, 182–192 M Lech et al.: TIR8/SIGIRR in renal TLR signaling original article

overexpression of TIR8/SIGIRR in Jurkat or HepG2 cells tubular epithelial cells. Obviously, the regulation and substantially reduced IL-1- or IL-18-induced activation of function of TIR8/Sigirr is different in immune and NF-kB.15,17,18 Tir8/Sigirr-deficient mice are more susceptible nonimmune cells in the kidney. to dextran-induced inflammatory bowel disease and a TIR8/ Sigirr-blocking was shown to aggravate Pseudomo- RESULTS 19,20 nas aeruginosa keratitis in Balb/c mice. This was referred TIR8/Sigirr is expressed in renal antigen-presenting cells and to TIR8/Sigirr-mediated suppression of TLR signaling in renal tubular epithelial cells of 6-week-old C57BL/6 mice dendritic cells, which express TIR8/Sigirr constitutively. TLR function has been reported to be age-dependent;21 Interestingly, TIR8/Sigirr mRNA is expressed at high levels hence, we analyzed TIR8/Sigirr mRNA levels by real-time in the murine and human kidney in comparison to other reverse (RT)–PCR in C57BL/6 mice, either 10 solid organs including lymphoid tissues.15,17 This may either days, 6 weeks, or 1 year of age. At 6 weeks of age, high levels relate to TIR8/Sigirr expression in intrarenal myeloid cells or of TIR8/Sigirr mRNA were found in kidneys, threefold tubular cells. We hypothesized a role for TIR8/Sigirr in compared to respective TIR8/Sigirr levels in spleen (Figure regulating TLR signaling in both cell types during exposure 1a). By contrast, brain, heart, , , small intestine, to UPEC LPS. However, we discovered that renal TLR colon, skin, and muscle expressed TIR8/Sigirr mRNA at signaling is independent of TIR8/Sigirr expression in renal lower levels as in spleen. In most organs, TIR8/Sigirr mRNA

a 0.00020 1.5 weeks * 6.0 weeks 0.00015 52 weeks

0.00010

0.00005 * * # * Tir8/Sigirr mRNA/18s rRNA # * * * * * # 0 * * * Spleen Thymus Brain Lung Liver Kidney Small Skin intestine

b c 2×10−5 Kidney Tir8/Sigirr rRNA

1×10−5

+/+ −/− Tir8/Sigirr n.d. 0 Intrarenal Tubular Mesangial CD11b+ epithelial cells cells cells d

Figure 1 | TIR8/Sigirr expression in C57BL/6 mice.(a) mRNA was extracted from organs of C57BL/6 mice of different age as indicated (n ¼ 3–6). TIR8/Sigirr mRNA expression levels were determined by real-time RT–PCR and expressed as mean of the ratio TIR8/Sigirr /18s- rRNA7s.e.m.; *Po0.05 vs 1.5 weeks, #Po0.05 vs 6 weeks. (b) TIR8/Sigirr protein expression was determined by Western blot analysis. were prepared from kidneys of 6–week-old C57BL/6 wild-type mice or TIR8/Sigirr-deficient mice as indicated. (c) Primary cells were isolated from of 6-week-old C57BL/6 mice as described in Materials and Methods. TIR8/Sigirr mRNA levels were determined by real-time RT–PCR and expressed as mean of the ratio TIR8/Sigirr/18s-rRNA7s.e.m. N.d. ¼ not detected. (d) F4/80 immunostaining for intrarenal antigen-presenting cells (black) in kidney cortex (left) and medulla (right). Original magnification  200.

Kidney International (2007) 72, 182–192 183 original article M Lech et al.: TIR8/SIGIRR in renal TLR signaling

levels declined from young to old age. Interestingly, TIR8/ originates from tubular epithelial cells and intrarenal Sigirr mRNA levels in 6-week-old C57BL/6 mice were seven- immune cells, resident antigen-presenting cells. to eightfold higher as compared to 10-day- or 1-year-old mice. The prominent expression of TIR8/Sigirr in kidneys of TIR8/Sigirr localizes to luminal and basolateral membranes of 6-week-old mice was confirmed on the protein level by tubular epithelial cells Western blot (Figure 1b). Cortex and medulla from kidneys The transmembrane molecule TIR8/Sigirr has been reported of 6-week-old C57BL/6 mice expressed equal levels of TIR8/ to suppress LPS signaling in Jurkat cells by interacting with Sigirr mRNA, suggesting a tubular or interstitial origin rather the intracellular domain of TLR4 and both extracellular Ig than glomerular cells, which predominantely locate to the domain and the intracellular TIR domain of IL-1R.18 Thus, renal cortex (data not shown). In fact, unlike primary TIR8/Sigirr should localize to outer membranes of these cells. mesangial cells tubular epithelial cells and resident CD11b/ In fact, flow cytometry of primary tubular epithelial cells F4/80-positive renal myeloid cells both expressed TIR8/Sigirr using a TIR8/Sigirr-specific antibody revealed a robust signal mRNA (Figure 1c). The latter cells localize to the interstitium on the cell surface (Figure 2a). The cellular distribution of of the renal cortex and medulla of mice (Figure 1d). These TIR8/Sigirr was confirmed by the immunostaining on data indicate that the profound renal TIR8/Sigirr expression primary tubular epithelial cells. Positive staining signals were

100 100 abTir8/Sigirr +/+ Tir8/Sigirr −/−

80 80

60 60 Counts Counts 40 40

20 20

0 0 100 101 102 103 104 100 101 102 103 104 FL2-H FL2-H cd

* ef* * * ** * *

Figure 2 | TIR8/Sigirr expression in tubular epithelial cells. (a and b) Flow cytometry for TIR8/Sigirr was performed using primary tubular epithelial cells prepared from (a) wild-type mice and (b) Tir8/Sigirr-deficient mice as indicated. Surface expression of TIR8/Sigirr (black line) is indicated by a fluorescence shift compared to the isotype control antibody (red line). (c and d) TIR8/Sigirr immunostaining with phycoerythrin- labeled secondary antibody confirms TIR8/Sigirr expression on the cellular surface in confluent growing primary tubular epithelial cells prepared from wild-type mice (c). (d) Tubular epithelial cells from Tir8/Sigirr-deficient mice lack respective positive signal. 4,6-diamidino-2- phenylindole staining cell nuclei in blue (original magnification  400). (e and f) Renal sections were prepared from 6-week-old C57BL/6 wild- type mice and Tir8/Sigirr-deficient mice and stained for TIR8/Sigirr with a phycoerythrin-labeled secondary antibody. Fluorescein isothiocyanate-phalloidin stains brush border in proximal tubular epithelial cells and cell–cell contacts in green. (e) A yellow signal is obtained at basolateral membranes (arrow heads) and brush border (*) as a result of colocalization of red TIR8/Sigirr staining and green phalloidin staining. (f) The TIR8/Sigirr signal is absent when renal sections taken from Tir8/Sigirr-deficient mice were stained accordingly, original magnification  400).

184 Kidney International (2007) 72, 182–192 M Lech et al.: TIR8/SIGIRR in renal TLR signaling original article

detected at outer membranes only (Figure 2c). The specificity Ccl2 (pg/ml) Cxcl2 (pg/ml) of positive signals in both methods was demonstrated by a Kidney slices 3000 2000 using tubular epithelial cells prepared from Tir8/Sigirr- Tir8/Sigirr +/+ # Tir8/Sigirr +/+ Tir8/Sigirr −/− * Tir8/Sigirr −/− * deficient mice (Figure 2b and d). TIR8/Sigirr immunostain- 2000 ing on kidney sections prepared from 6-week-old C57BL/6 1000 * mice revealed positive staining signals at basolateral and 1000 * luminal membranes of tubular epithelial cells (Figure 2e). In 0 0 proximal tubular epithelial cells, the colocalization with the Medium Lps Medium Lps luminal brush border (costainined with fluorescein isothio- b Intrarenal CD11b + cells cyanate–phalloidin) was apparent (Figure 2e). Glomeruli 2000 # 1000 # stained negative for TIR8/Sigirr (not shown). Negative * * controls included staining of renal sections from Tir8/ * 1000 Sigirr-deficient mice (Figure 2f), omitting the primary * 500 antibody (not shown) or preabsorbing the primary antibody * with TIR8/Sigirr-overexpressing human embryonic kidney 0 0 293 cells (not shown). Thus, TIR8/Sigirr is expressed on Medium Lps Medium Lps basolateral and luminal membranes of tubular epithelial cells c Tubular epithelial cells 1000 600 in the mouse kidney. * * * * TIR8/Sigirr suppresses renal Ccl2 and Cxcl2 production in 500 300 response to LPS in resident myeloid cells but not in tubular epithelial cells 0 0 We have previously shown that both resident intrarenal Medium Lps Medium Lps myloid cells and tubular epithelial cells contribute to renal Figure 3 | TIR8/Sigirr and LPS-induced Ccl2 and Cxcl2 production TLR4 signaling in infective pyelonephritis with uropatho- in mouse kidneys. (a) Kidney slices, freshly prepared from wild-type genic Escherichia coli (UPEC).9 Hence, we hypothesized that and Tir8/Sigirr-deficient mice, were placed in medium and exposed to 1 mg/ml ultrapure LPS. After 6 h, supernatants were harvested and lack of TIR8/Sigirr would enhance LPS-induced renal Ccl2 (left) and Cxcl2 production (right) were measured by chemokine release as a result of uncoupling TLR4 signaling enzyme-linked immunosorbent assay. Data represent means7s.e.m. in both cell types. In fact, kidney slices from Tir8/Sigirr- from three independent experiments; *Po0.05 vs medium, #Po0.05 deficient mice produced much more Ccl2 and Cxcl2 on vs TIR8/Sigirr þ / þ .(b and c) Intrarenal myeloid cells were prepared from 6-week-old wild-type and Tir8/Sigirr-deficient mice by CD11b exposure to 1 mg/ml UPEC LPS for 24 h (Figure 3a). magnetic bead isolation and cultured with granulocyte macrophage However, when we exposed primary tubular epithelial cells colony-stimulating factor as mentioned in Materials and Methods. or CD11b-positive renal myeloid cells to LPS, lack of TIR8/ Tubular epithelial cells were prepared as before. The cells were Sigirr was found to enhance LPS-induced Ccl2 or CxCL2 incubated with either 1 mg/ml ultrapure LPS in medium or medium alone for 24 h. Data represent means7s.e.m. from three independent production only in the myeloid cells but not in the tubular experiments; *Po0.05 vs medium, #Po0.05 vs TIR8/Sigirr þ / þ . epithelial cells (Figure 3b and c). These data suggest that the suppressive effect of TIR8/Sigirr on renal TLR4 signaling relates to TIR8/Sigirr in intrarenal myeloid cells and is tubular epithelial cells and spleen monocytes were isolated independent of TIR8/Sigirr in tubular epithelial cells. either from 6-week-old wild-type or Tir8/Sigirr-deficient mice. As with the TLR4 ligand LPS, Tir8/Sigirr-deficient TIR8/Sigirr does not modulate TLR signaling in tubular monocytes produced increased amounts of Ccl2 when epithelial cells exposed to ligands for TLR1/2 (Pam3Cys), TLR3 (poly I:C LPS-induced chemokine expression is mediated through NF- RNA), and TLR9 (CpG-DNA) (Figure 4b). In tubular kB. Because transfection of primary monocytes and tubular epithelial cells, lack of TIR8/Sigirr had no such effect and cells with a NF-kB-reporter gene construct was limited by no response to CpG-DNA was observed, because tubular low cell survival we used a p50 (NLS) antibody to determine epithelial cells do not express TLR9 (Figure 4c). These data nuclear translocation of NF-kB as a marker of NF-kB show that TIR8/Sigirr does not affect signaling through TLR activation. CD11b-positive monocytes and tubular epithelial tubular epithelial cells. cells were stained after 1 h of stimulation with either medium or LPS. The data obtained were consistent with those TIR8/Sigirr is a suppressor of TLR mRNA in monocytes but not obtained from LPS-induced Ccl2 and Cxcl2 production in tubular epithelial cells (Figure 4a), indicating that LPS-induced TLR signaling is Do the cell type-specific effects of TIR8/Sigirr on TLR independent of TIR8/Sigirr in tubular epithelial cells. Next, signaling relate to a modulatory effect on TLR expression we tested whether TIR8/Sigirr affects signaling of TLRs other levels? To answer this question, TLR1-9 and -11 mRNA than TLR4 in tubular epithelial cells. Spleen monocytes were expression levels were determined by real-time RT–PCR from used as a positive control for TIR8/Sigirr function. Primary monocytes and tubular epithelial cells kept under normal

Kidney International (2007) 72, 182–192 185 original article M Lech et al.: TIR8/SIGIRR in renal TLR signaling

Spleen macrophages Tubular epithelial cells a 10 10 Medium *# Medium 8 Lps 8 Lps

6 * 6

4 4 * * 2 2 Nuclear p50+ cells/hpf 0 0 +/+ −/− +/+ −/− Tir8/Sigirr Tir8/Sigirr b 1000 # Monocytes * *# Tir8/Sigirr +/+ 750 *# Tir8/Sigirr –/– *# 500

Ccl2 (pg/ml) * 250 * * * 0 Medium Pam3Cys Poly I:C Lps CpG c 10 000 Tubular epithelial cells * * 7500 * * 5000 * * Ccl2 (pg/ml) 2500

0 Medium Pam3Cys Poly I:C Lps CpG Figure 4 | TIR8/Sigirr and TLR signaling. Monocytes and tubular epithelial cells were prepared from 6-week-old Tir8/Sigirr-deficient or wild-type mice as indicated. (a) After 1 h of stimulation with LPS (1 mg/ml), the cells were stained with a p50 antibody for detection of NF-kB activation as described in Materials and Methods. Data represent means7s.e.m. of positive nuclei per high power field (hpf); *Po0.05 vs medium, #Po0.05 vs TIR8/Sigirr þ / þ .(b and c) Cells were stimulated either with medium, Pam3Cys, pI:C RNA, LPS, or CpG-DNA for 24 h. Ccl2 production was determined in supernatants by enzyme-linked immunosorbent assay. Data represent means7s.e.m. from three independent experiments; *Po0.05 vs medium, #Po0.05 vs TIR8/Sigirr þ / þ .

culture conditions. Monocytes expressed all TLRs and lack of LPS and IFN-g/TNF both increased TIR8/Sigirr expression in TIR8/Sigirr increased mRNA levels of TLR1 (13-fold), TLR5 monocytes in a dose-dependent manner with highest mRNA (23-fold), TLR6 (sevenfold), TLR7 (eightfold), TLR9 (12- levels after 18 h (Figure 6). Together, these data suggest that fold), and TLR11 (23-fold) as compared to monocytes constitutive TLR4 signaling consistently suppresses TIR8/ isolated from wild-type mice (Figure 5a). Lack of TIR8/Sigirr Sigirr mRNA in tubular epithelial cells and monocytes. By did not affect mRNA expression of TLR2, -3, -4, and -8 in contrast, LPS or exposure rather upregulates TIR8/ monocytes. By contrast, tubular epithelial cells expressed Sigirr in monocytes but not in tubular epithelial cells, TLR1-4 and -11 mRNA independent of TIR8/Sigirr (Figure indicating cell type-specific regulation of TIR8/Sigirr. 5b). Apparently, TIR8/Sigirr is a suppressor of TLR1, -5, - 6, -7, -9, and -11 mRNA expression in monocytes, but has no Monocytes and renal tubular epithelial cells express the same effect on the respective TLRs coexpressed by tubular splice variant of TIR8/Sigirr mRNA but TIR8/Sigirr protein is epithelial cells, that is TLR1 and TLR11. differently N- and O-glycosylated in these cell types Can the different functional role of TIR8/Sigirr in monocytes The regulation of TIR8/Sigirr is cell type-specific and tubular epithelial cells be explained by either the Next, we examined whether LPS or interferon-g (IFN-g)/ expression of different splice variants or by different post- tumor necrosis factor (TNF) stimulation suppress TIR8/ translational modifications? We performed real-time RT–PCR Sigirr mRNA levels in monocytes and tubular epithelial cells. using primers specific for the intracellular and extracellular In fact, both LPS and IFN-g/TNF reduced TIR8/Sigirr mRNA domain of TIR8/Sigirr (Figure 7a). Monocytes and tubular levels in tubular epithelial cells (Figure 6a and b). By contrast, epithelial cells expressed comparable levels of both the

186 Kidney International (2007) 72, 182–192 M Lech et al.: TIR8/SIGIRR in renal TLR signaling original article

a 0.0012 Monocytes Tir8/Sigirr +/+ Tir8/Sigirr −/− * 0.0006 * * TlrmRNA/18s rRNA 0 Tlr1 Tlr2 Tlr3 Tlr4 Tlr5 Tlr6 Tlr7 Tlr8 Tlr9 Tlr11

b 4×10−5 Tubular epithelial cells Tir8/Sigirr +/+ Tir8/Sigirr −/−

2×10−5 TlrmRNA/18s rRNA 0 Tlr1 Tlr2 Tlr3 Tlr4 Tlr5 Tlr6 Tlr7 Tlr8 Tlr9 Tlr11 Figure 5 | TIR8/Sigirr modulates TLR mRNA expression. (a) Monocytes and (b) tubular epithelial cells were prepared from 6-week-old Tir8/Sigirr-deficient or wild-type mice as indicated. Tlr1-9 mRNA expression levels were determined by real-time RT–PCR and expressed as mean of the ratio Tlr/18s-rRNA7s.e.m.; *Po0.05 vs TIR8/Sigirr þ / þ .

a 12×10−6 Monocytes Tubular epithelial cells

8×10−6

4×10−6

Tir/Sigirr mRNA/18s rRNA 0 Lps(g/ml) 0 0.5 1 2 0 0.5 1 2 TNF/ 0 250 500 1000 0 250 500 1000 IFN (U/ml) 0 50 100 200 0 50 100 200

15×10−6 b Monocytes Tubular epithelial cells

10×10−6 Medium Lps TNF/IFN

5×10−6

Tir8/Sigirr mRNA/18s rRNA 0 Time (h) 6 12 18 24 6 12 18 24 Figure 6 | Regulation of TIR8/Sigirr by IFN-c and TNF. Monocytes and tubular epithelial cells were prepared from 6-week-old mice of different strains as indicated. (a) Cells were stimulated with either medium, Llp, or IFN-g þ TNF. TIR8/Sigirr mRNA expression levels were determined after 18 h by real-time RT–PCR and expressed as mean of the ratio Tlr/18s-rRNA7s.e.m. (b) Cells were stimulated with either medium, LPS (1 mg/ml), or IFN-g (100U/ml) þ TNF (500 U/ml). TIR8/Sigirr mRNA expression levels were determined after 6, 12, 18, or 24 h by real-time RT–PCR and expressed as mean of the ratio Tlr/18s-rRNA7s.e.m. intracellular and extracellular domain of TIR8/Sigirr (Figure aP32-labeled full-length probe (data not shown) or a shorter 7b). This was confirmed by Northern blot analysis. Total probe matching the extracellular part of TIR8/Sigirr. A single RNA was extracted from kidneys and spleens from wild-type TIR8/Sigirr mRNA transcript of 3.5 kB was detected in both and Tir8/Sigirr-deficient mice. The RNA was hybridized with kidney and spleen and no further splice variants were

Kidney International (2007) 72, 182–192 187 original article M Lech et al.: TIR8/SIGIRR in renal TLR signaling

a 3.73 Kb

Mm00491703_gl Mm00491700_ml TIR domain extracellular domain primer 2 primer 1

−6 b 5×10 c Kidney WT Kidney KO Spleen WT Spleen KO − 2.5×10 6 28S (4.7 kb) – Tir8/Sigirr 1/2 /18s rRNA 0 18S (1.9 kb) – Primer 1 Primer 2 de f Kidney WT Kidney KO Spleen WT Spleen KO CD11b+ depleted cells WT CD11b+ depleted cells KO CD11b+ cells WT CD11b+ KO CD11b+ depleted cells WT CD11b+ depleted cells KO CD11b+ cells WT CD11b+ KO CD11b+ depleted cells WT CD11b+ depleted cells KO CD11b+ cells WT CD11b+ KO CD11b+ depleted cells WT CD11b+ depleted cells KO CD11b+ cells WT CD11b+ KO 116 kDa –

66 kDa –

45 kDa –

35 kDa – + PNGaseF + Benzyl-GalNAc Figure 7 | TIR8/Sigirr transcripts and TIR8/Sigirr glycosylation. Tubular epithelial cells and spleen monocytes were prepared from 6-week-old Tir8/Sigirr-deficient or wild-type mice. (a) An expression level of TIR8/Sigirr mRNA was determined by using primers specific for either the extracellular (primer 1) or the intracellular domain (primer 2) of TIR8/Sigirr. (b) The respective mRNA expression levels using primer 1 and primer 2 were determined by real-time RT–PCR using cDNA prepared from splenocytes (black) and tubular epithelial cells (white) and expressed as mean of the ratio to 18s-rRNA7s.e.m. (c) Northern blot was performed on RNA isolates from spleens and kidneys of wild-type (WT) and Tir8/Sigirr-deficient mice (KO) as described in Materials and Methods. Note a single TIR8/Sigirr transcript of 3.5 kB in spleen and kidney of wild-type mice. (d–f) Western blot analysis was performed on protein isolates derived from total kidney or from renal cell suspensions separated by magnetic beads isolation for CD11b-positive and -negative cells, that is renal myeloid cells and mostly tubular epithelial cells. (d and e) Note that the smaller but predominant of the two TIR8/Sigirr-specific bands seen in total kidney isolates derives exclusively from CD11b-negative renal cells. The specificity of the bands is indicated by their absence in isolates from Tir8/Sigirr-deficient mice. (e) PNGase digests the TIR8/Sigirr glycoforms from both cell types to smaller but still different proteins indicative of cell-type specific N-glycosylation of TIR8/Sigirr. (f) Inhibiting O-glycosylation with benzyl-GalNAc reduces the size of the smaller glycoform of TIR8/Sigirr in renal CD11b cells but not the larger form of TIR8/Sigirr present in both cell types.

detected (Figure 7c). However, cell type-specific functions of sites and the predicted and actual molecular weight of TIR8/ proteins may relate to post-translational modifications. In Sigirr suggest extensive glycosylation.14 We tested the fact, Western blot of total kidney protein isolates revealed two glycosylation of TIR8/Sigirr in the two types of renal cells TIR8/Sigirr-specific bands of 75 and 90 kDa (Figure 7d). by digesting protein isolates from renal myeloid and Next, we prepared renal cell suspensions and isolated CD11b- nonmyeloid cells with peptide-N-glycosidase F (PNGaseF). positive myeloid cells and CD11b-negative renal cells of This glycosidase digests N-linked but not N-glycans nonmyeloid origin. The renal myeloid cells expressed the with fucose-linked a1,3 to the Asn-bound N-acetylglucosa- larger 90 kDa form of TIR8/Sigirr, whereas the nonmyeloid mine, O-linked oligosaccharides or glycosylphosphatidylino- renal cells predominantely expressed the smaller form of sitol lipid anchors from glycoproteins, and hence, allows 75 kDa (Figure 7e). Protein size depends on protein the detection of glycosylation variants. PNGaseF treatment glycosylation. TIR8/Sigirr carries five putative glycosylation reduced the molecular mass of TIR8/Sigirr in both CD11b-

188 Kidney International (2007) 72, 182–192 M Lech et al.: TIR8/SIGIRR in renal TLR signaling original article

positive and -negative renal cells (Figure 7e). CD11b-positive cells, as we found that kidney cell suspensions prepared and -negative renal cells both carry TIR8/Sigirr glycosylated following the protocol reported by Wald et al.17 are at identical sites, digested by PNGaseF to a 60-kDa protein. commonly contaminated by myeloid cells, compromising The CD11b-negative (nonimmune) renal cells, however, conclusions on the role of TIR8/Sigirr in tubular epithelial contained additional TIR8/Sigirr glycosylation variants that cells. By using the protocol reported by Tsuboi et al.3 the are digested by PNGaseF to two other variants with a size of primary tubular epithelial cell population was 495%. approximately 45 and 50 kDa. Furthermore, we used benzyl- Furthermore, we have now backcrossed Tir8/Sigirr mutants N-acetylgalactosamine (benzyl-GalNAc, an inhibitor of for six generations to the C57BL/6 background to prepare O-glycosylation, to investigate the O-glycosylation of TIR8/ monocytes and tubular epithelial cells from mice with a Sigirr in renal immune and nonimmune cells. Benzyl-GalNAc predicted 495% identical background as compared to wild- treatment did not change the size of the larger glycoform type controls.24,25 In fact, by using a pure tubular epithelial of TIR8/Sigirr in renal myeloid and nonmyeloid cells (Figure cell population no impact of TIR8/Sigirr on TLR signaling 7f), hence, the larger TIR8/Sigirr glycoform is not O- could be detected. Obviously, intrarenal antigen-presenting glycosilated. By contrast, exposing renal nonimmune cells cells and tubular epithelial cells contribute to renal TLR4 to benzyl-GalNAc reduced the size of the smaller predomi- signaling. Although TIR8/Sigirr is expressed by both cell nant TIR8/Sigirr glycoform to about 60 kDa, suggestive of O- types, TIR8/Sigirr suppresses TLR signaling only in intrarenal glycosylation of this protein (Figure 7f). These data show that antigen-presenting cells and not in tubular epithelial cells. the cell type-specific functions of TIR8/Sigirr in the kidney Although reducing TLR signaling in renal antigen-presenting do not relate to different splicing but are associated with cells may avoid inappropriate immunity-related tissue injury, different glycosylation variants. an unrestricted activation of tubular cells may support the induction of necessary epithelial stress response elements to DISCUSSION cope with microbial infection. This study was motivated by two observations: First, our previous observation that TLR4 on intrinsic renal cells as well Identification of novel cell type-specific functions and as bone marrow-derived cells contributes to innate immunity regulation of TIR8/Sigirr in infective pyelonephritis9 and second, TIR8/Sigirr, a Intrigued by the lack of suppressive function of TIR8/Sigirr in negative regulator of TLR/IL-1R signaling was reported to tubular epithelial cells, we first excluded alternative splicing be expressed at high levels in the kidney.15,17 We hypothesized and localized the spatial expression of TIR8/Sigirr protein. that TIR8/Sigirr would control inappropriate TLR signaling Alternative splicing is common for the members of the IL- in the kidney and, if so, that TIR8/Sigirr on renal myeloid 1R/TLR superfamily and their signaling molecules.13,26,27 By and nonimmune intrinsic renal cells would contribute to this using PCR primers for the extracellular and intracellular phenomenon. Although we confirmed the first part of our domain as well as Northern blotting, we essentially excluded hypothesis, our data do not support the latter. alternative splicing of TIR8/Sigirr. TIR8/Sigirr protein localized to the basolateral and luminal membranes of TIR8/Sigirr modulates renal Ccl2 production on exposure tubular epithelial cells consistent with the spatial subcellular to LPS distribution of the other members of the IL-1R family.21 Our data clearly demonstrate that TIR8/Sigirr suppresses Post-translational modifications such as glycosylation can renal Ccl2 production on exposure to LPS. This is consistent also alter the function of immune mediators.28 Furthermore, with the exacerbation of various other inflammatory disease N-glycans have been shown to play a general role in protein models that have been induced in Tir8/Sigirr-deficient mice, folding and sorting in biosynthetic traffic of polarized renal for example endotoxic shock,17 dextran-induced colitis,19 and tubular epithelial cells.29 We confirmed glycosylation of Pseusomonas aeruginosa keratitis.20 Recognition of LPS in the TIR8/Sigirr by PNGaseF digestion or inhibition of O- kidney may involve various cell types which can broadly be glycosylation and subsequent Western blotting. Interestingly, classified into renal immune cells and nonimmune cells. PNGaseF digestion produced two glycoforms of TIR8/Sigirr Intrarenal immune cells are mainly resident antigen-present- in the kidney in addition to the single glycoform found in the ing cells which originate from the bone marrow and form a spleen and kidney. Obviously, some renal cells express an dense network in the interstitial space that grid the tubular additional glycoform of TIR8/Sigirr beyond the glycoform compartment in the healthy kidney.22,23 Consistent with the also found in immune cells. The smaller form of TIR8/Sigirr previous description of TIR8/Sigirr’s function in dendritic found in renal CD11b-negative (mostly tubular epithelial) cells and monocytes,15,17 we found that TIR8/Sigirr sup- cells is O-glycosilated via benzyl-GalNAc, but myeloid cell presses LPS-induced Ccl2 and Cxcl2 production in renal TIR8/Sigirr is not. By contrast, both the myeloid and antigen-presenting cells. Wald et al.17 reported a similar role nonmyeloid cell TIR8/Sigirr was markedly reduced in size for TIR8/Sigirr in renal tubular cells but in their study kidney after digestion with PNGaseF, which is suggestive of extensive cell suspensions were prepared from Tir8/Sigirr-deficient N-glycosylation. Thus, TIR8/Sigirr in renal tubular epithelial mice in a mixed genetic background. In this study, we applied cells is N- and O-glycosylated, whereas TIR8/Sigirr in more stringent preparation techniques for primary tubular intrarenal CD11b-positive myeloid cells lacks benzyl-Gal-

Kidney International (2007) 72, 182–192 189 original article M Lech et al.: TIR8/SIGIRR in renal TLR signaling

NAc-dependent O-glycosylation. However, the causative role complementary to 243 bp N terminus part (cut with HindIII and of such post-translational modifications and their putative BstXI) or the full-length Tir8/Sigirr. Efficacy of probe synthesis was role for the function of the TIR8/Sigirr–TLR signaling checked by polyethylenimine (PEI) chromatography. For hybridiza- complex remains to be addressed in future studies. tion, QuickHyb hybridization solution was used (Stratagene). The Notably, beyond the cell type-specific role of TIR8/Sigirr membranes were exposed on a PhosphorScreen of the Phosphor- Imager after 24 h. on TLR signaling, TIR8/Sigirr-induced suppression of TLR expression and the LPS- or IFN-g/TNF-induced regulation of TIR8/Sigirr itself are different in antigen-presenting cells and Western blot analysis Organs from mice were manually dissected, homogenized in 1 ml tubular epithelial cells. LPS challenge suppressed TIR8/Sigirr lyses buffer (50 mM Tris–HCl, pH 7.5; 150 mM NaCl; 100 mM sodium mRNA in tubular epithelial cells, which is consistent with orthovanadate, 0.5% sodium deoxycholat, 4% NP-40, 2% Triton downmodulation of renal TIR8/Sigirr mRNA after intrave- M M 15,17 X-100; 5 m ethylenediaminetetraacetic acid; 300 m sucrose; nous injection of 1 mg LPS in mice. By contrast, LPS and proteases inhibitor tablets Complete (Roche, Penzberg, Germany), TNF/IFN-g have opposite effects on TIR8/Sigirr mRNA and centrifuged for 45 min at 30 000 g. The supernatant was tested expression in spleen monocytes, a finding consistent with the for the protein concentration. Extracted proteins were incubated in recent observation of TIR8/Sigirr expression in monocytes of 2 Â loading buffer for 30 min at 651C, resolved by 12% sodium patients with sepsis.30 dodecyl sulfate–polyacrylamide gel electrophoresis, and transferred In summary, resident renal antigen-presenting cells to an Immobilon-P membrane (Millipore, Eschborn, Germany). contribute to TLR-mediated antimicrobial immunity in the After blocking with 1% Western blocking solution (Roche), the filter kidney and this function is controlled by TIR8/Sigirr, an was incubated with a goat polyclonal anti-TIR8/Sigirr antibody (1:1000; R&D Systems, Wiesbaden-Nordenstadt, Germany) over- orphan receptor of the IL-1R family. Thereby, TIR8/Sigirr night in 0.5% Western blocking solution (Roche). Immune inhibits an inappropriate innate in the complexes were visualized using a peroxidase-conjugated donkey kidney. TIR8/Sigirr is also expressed in tubular epithelial anti-goat IgG antibody (1:10 000; Dianova, Hamburg, Germany) for cells, which promote innate immune signaling through a 1 h in 0.5% Western blocking solution and processed for detection limited set of TLRs, that is TLR1–4 (and TLR11 in mice). by enhanced chemiluminescence (Amersham Pharmacia Biotech However, TIR8/Sigirr does neither inhibit TLR signaling nor Europe, Freiburg, Germany). reduce TLR expression in tubular epithelial cells supporting Before digestion with PNGaseF (NewEngland BioLabs, Frank- their role as sensors of microbial infection in the kidney. furt, Germany) protein isolates were denaturized at 1001C for 10 min in buffer containing 0.5% sodium dodecyl sulfate and 0.04 1 MATERIALS AND METHODS dithiothreitol. Digestion was performed for 1 h at 37 C in buffer Animal studies containing 0.05 M sodium phosphate pH 7.5, 1% NP-40. benzyl-N- Tir8/Sigirr-deficient mice were generated by homologous recombi- acetygalactosamine (Sigma, Taufkirchen, Germany) was used as an inhibitor for O-glycosylation at 4 mM final concentration. The cells nation of the Tir8 gene as described previously, genotyped, and with or without inhibitors were grown on 90 mm dish for 48 h. backcrossed to the C57BL/6J strain (Charles River, Sulzbach, Germany) to generation F5.15 Mice were housed in groups of five Media containing fresh inhibitor (4 mM) were changed two times during this period. mice in filter-top cages with a 12 h dark/light cycle and unlimited access to food and water. Cages, nest lets, food, and water were sterilized by autoclaving before use. All experimental procedures Flow cytometry were performed according to the German animal care and ethics Flow cytometry of primary spleen monocytes and tubular epithelial legislation and had been approved by the local government cells was performed using the goat polyclonal anti-TIR8/Sigirr authorities. antibody. A polyclonal phycoerythrin-labeled donkey anti-goat antibody was used for detection (1:1000, Dianova). A goat IgG RNA preparation and real-time quantitative (TaqMan) (BD Pharmingen) was used as isotype control. Fluorescence- RT–PCR activated cell sorting analysis was conducted using a FACScalibur Organs were snap-frozen in liquid nitrogen and stored at À801C. RT machine and CellQuest software (BD Pharmingen, Heidelberg, and real time RT–PCR from total renal RNA was prepared as Germany). described.3 Controls consisting of ddH2O were negative for target and housekeeper . Oligonucleotide primer (300 nM) and Immunostaining probes (100 nM) were from PE Biosystems (Weiterstadt, Germany) Immunostaining was performed as described previously.9 Staining (Table 1). for TIR8/Sigirr and F-actin in tubular epithelial cells was performed on 3.7% formaldehyde-fixed cell monolayer, permeabilized with Northern blot analysis 0.5% Triton X-100. Polyclonal anti-TIR8/Sigirr antibody (1:200) Total kidney and spleen RNA was isolated, analyzed for the and a phycoerythrin-labeled donkey anti-goat IgG (1:500) were used degradation and purity, and separated on a 1% formaldehyde/3- for detection. Staining for TIR8/Sigirr was performed on acetone- [N-morpholino]propanesulfonic acid agarose gel using standard fixed frozen section using the forementioned polyclonal anti- methods. The RNA was transferred on the GeneScreen membrane, SIGIRR antibody (1:1000) with a phycoerythrin-labeled donkey baked, and cross-linked with UV light. The DNA-probes syntheses anti-goat IgG antibody (1:10 000, Dianova) for detection. Costain- were performed using the prime-it random primer-labeling kit ing for tubular brush border was performed with fluorescein (Stratagene, La Jolla, CA, USA). P32-labeled DNA probes were isothiocyanate–phalloidin (1:200; Invitrogen, Paisley, UK). Staining

190 Kidney International (2007) 72, 182–192 M Lech et al.: TIR8/SIGIRR in renal TLR signaling original article

Table 1 | Probes used for real-time RT–PCR TLR Accession-number Sequence TLR1 AF316985 Forward primer: 50-GTCAAAGCTTGGAAAGAATCTGAAG-30 Reverse primer: 50-AATGAAGGAATTCCACGTTGTTTC-30 6 FAM: 50-ATCTTACCCTGAACAATG-30

TLR2 AF124741 Forward primer: 50-CACCGGTCAGAAAACAACTTACC-30 Reverse primer: 50-CAAGATCCAGAAGAGCCAAAGAG-30 6 FAM: 50-AGACAAAGCGTCAAATC-30

TLR3 AF355152 Forward primer: 50-CGAAAGTTGGACTTGTCATCAAATC-30 Reverse primer: 50-ACTTGCCAATTGTCTGGAAACAC-30 6 FAM: 50-CACTTAAAGAGTTCTCCC-30

TLR4 AF110133 Forward primer: 50-TTCAGAACTTCAGTGGCTGGATT-30 Reverse primer: 50-CCATGCCTTGTCTTCAATTGTTT-30 6 FAM: 50-ATCCAGGTGTGAAATT-30

TLR5 AF186107 Forward primer: 50-CCCAGCTTGGATGAAATATCTGTAA-30 Reverse primer: 50-CCCAGTCTTTTCTTCTTGAACACTTA-30 6 FAM: 50-CGGGCACCAGTACT-30 TLR6 AB020808 Forward primer: 50-TGAATGATGAAAACTGTCAAAGGTTAA-30 Reverse primer: 50-GGGTCACATTCAATAAGGTTGGA-30 6 FAM: 50-TGGTGAGTTCTGATAAAA-30

TLR7 AY035889 Forward primer: 50-TGCCACCTAATTTACTAGAGCTCTATCTTTAT-30 Reverse primer: 50-TAGGTCAAGAACTTGCAACTCATTG-30 6 FAM: 50-CCAAGAAAATGATTTTAATAAC-30

TLR8 AY035890 Forward primer: 50-GGATGTTAAGAGAGAAACAAACGTTTT-30 Reverse primer: 50-AAAAGCACGTCAGAATCCATGA-30 6 FAM: 50-AGAACATGGAAAACATG-30

TLR9 NM 031178 Forward primer: 50-CAATCTGACCTCCCTTCGAGTACTT-30 Reverse primer: 50-GCCACATTCTATACAGGGATTGG-30 6 FAM: 50-ATTGCCGTCGCTGCGACCATG-30

Predeveloped TaqMan assay reagents from Applied Biosystems

SIGIRR NM023059 ContextSeq AoD: 50-GAAGAGCCATGGCAGGTGTCTGTGA-30 TLR11 NM205819 ContextSeq AoD: 50-AGGAATGCTGTTGTGCAGAAGATCC-30 AY510704 18s rRNA Predeveloped TaqMan assay reagents RT–PCR, reverse transcription–polymerase chain reaction; TLR, toll-like receptor.

of NF-kB was performed on 3% paraformaldehyde-fixed cell population, 75% also stained positive for F4/80 and 45% for CD11c. monolayer. Cell membranes were permeabilized with 0.5% Triton Primary tubular cells were prepared from kidneys of 6-week-old X-100 in phosphate-buffered saline. A p50 (NLS) antibody (1:100, Tir8/Sigirr-deficient or wild-type C57BL/6 mice and grown in Santa Cruz, Heidelberg, Germany) was used for detecting nuclear culture according to a standard protocol.3 Monocytes were prepared translocation of NF-kB.31 from spleens as follows. Whole spleen was isolated, mashed, and passed through a 30-m steel-wire mesh. Cells were resuspended in Cell culture conditions for primary cells 0.83% ammonium chloride and incubated for 5 min at room Antigen-presenting cells were prepared from kidneys as follows. temperature. Monocytes were separated from lymphocytes by Kidney cell suspensions were prepared by applying the mashed adhesion to plastic. Cells were plated in RPMI 10% fetal calf serum kidney on 30 mm preseparation filters (Miltenyi Biotec GmbH, medium on six- or 12-well plates at 1 Â 106 or 5 Â 105 cells/well, Bergisch Gladbach, Germany) and CD11b magnetic bead isolation respectively. Cells were treated with medium control or IFN-g was performed (Miltenyi Biotec GmbH). After washing, columns 100 U/ml (PeproTech, Rocky Hill, NJ) þ TNF 500 U/ml (Immun- were removed from magnetic field and cells were eluted and plated Tools, Firesoythe, Germany) or TLR ligands as follows: TLR1/2: (3 Â 105 cells/ml) in 10% fetal calf serum–1% penicillin/streptavidin Pam3Cys 1 mg/ml (Alexis Biochemicals, Gru¨nberg, Germany), TLR3: (PS-RPMI) medium supplemented with 0.5 ng/ml recombinant pI:C 50 mg/ml (Sigma-Aldrich, Taufkirchen, Germany), TLR4: murine granulocyte macrophage colony-stimulating factor (Immu- ultrapure LPS 1 mg/ml (Invivogen, San Diego, CA, USA), TLR9: noTools, Friesoythe, Germany) and grown till they were 70–80% CpG-DNA 1668 1 mM (TibMolbiol, Berlin, Germany). Ligands were confluent. Flow cytometry revealed that of the CD11b-positive cell preincubated with polymyxin B sulfate (Sigma) (50 mg/ml) to

Kidney International (2007) 72, 182–192 191 original article M Lech et al.: TIR8/SIGIRR in renal TLR signaling

neutralize possible LPS contaminations. After various time intervals 11. Muzio M, Polentarutti N, Bosisio D et al. Toll-like receptors: a growing 6, 12, 18, or 24 h, culture supernatants were collected and cytokine family of immune receptors that are differentially expressed and regulated by different leukocytes. J Leukoc Biol 2000; 67: 450–456. levels were determined using a commercial enzyme-linked im- 12. Liew FY, Xu D, Brint EK, O’Neill LA. Negative regulation of toll-like munosorbent assay kit (Ccl2; R&D Systems Inc., Wiesbaden- receptor-mediated immune responses. Nat Rev Immunol 2005; 5: Nordenstadt, Germany). RNA from cells was prepared for mRNA 446–458. analysis using the Qiagen RNasy mini kit (Qiagen Sciences, Hilden, 13. Wells CA, Chalk AM, Forrest A et al. Alternate transcription of the Toll-like receptor signaling cascade. Genome Biol 2006; 7: R10. Germany). For additional experiments, 50 mg slices were cut from 14. Thomassen E, Renshaw BR, Sims JE. Identification and characterization of kidneys of Tir8/Sigirr-deficient or wild-type C57BL/6 mice and SIGIRR, a molecule representing a novel subtype of the IL-1R superfamily. placed in Dulbecco’s modified Eagle’s medium and 10% fetal calf Cytokine 1999; 11: 389–399. serum. The kidney slices were stimulated for 6 h with either medium 15. Polentarutti N, Rol GP, Muzio M et al. Unique pattern of expression and inhibition of IL-1 signaling by the IL-1 receptor family member TIR8/ or 1 mg/ml ultrapure LPS. After 6 h, supernatants were collected and SIGIRR. Eur Cytokine Netw 2003; 14: 211–218. Ccl2 and Cxcl2 levels were determined by enzyme-linked immuno- 16. Norris JL, Manley JL. Functional interactions between the pelle kinase, sorbent assay (R&D Systems Inc.). Toll receptor, and tube suggest a mechanism for activation of dorsal. Genes Dev 1996; 10: 862–872. 17. Wald D, Qin J, Zhao Z et al. SIGIRR, a negative regulator of Toll-like Statistical analysis receptor–interleukin 1 receptor signaling. Nat Immunol 2003; 4: 920–927. Data were expressed as mean7s.e.m. Comparison between groups 18. Qin J, Qian Y, Yao J et al. SIGIRR inhibits interleukin-1 receptor- and toll- were performed using Student’s t-test. A value of Po0.05 indicated like receptor 4-mediated signaling through different mechanisms. J Biol statistical significance. Chem 2005; 280: 25233–25241. 19. Garlanda C, Riva F, Polentarutti N et al. Intestinal in mice deficient in Tir8, an inhibitory member of the IL-1 receptor family. Proc ACKNOWLEDGMENTS Natl Acad Sci USA 2004; 101: 3522–3526. The work was supported by a grant from the Fritz Thyssen 20. Huang X, Hazlett LD, Du W, Barrett RP. SIGIRR promotes resistance Foundation (no. 10.04.2.140), the Deutsche Forschungsgemeinschaft against Pseudomonas aeruginosa keratitis by down-regulating type-1 (AN372/8-1 and GRK 1202), and the EU Integrated Project immunity and IL-1R1 and TLR4 signaling. J Immunol 2006; 177: ‘INNOCHEM’ (FP6-518167) to HJA and ‘MUGEN’ to CG and AM. 548–556. 21. Renshaw M, Rockwell J, Engleman C et al. Cutting edge: impaired Toll-like Parts of this work were prepared as a doctoral thesis at the Faculty receptor expression and function in aging. J Immunol 2002; 169: of Medicine, University of Munich, by ML. 4697–4701. 22. Kruger T, Benke D, Eitner F et al. Identification and functional REFERENCES characterization of dendritic cells in the healthy murine kidney and in 1. Akira S, Uematsu S, Takeuchi O. Pathogen recognition and innate experimental glomerulonephritis. J Am Soc Nephrol 2004; 15: 613–621. immunity. Cell 2006; 124: 783–801. 23. Soos TJ, Sims TN, Barisoni L et al. CX(3)CR1(+) interstitial dendritic cells 2. Anders HJ, Banas B, Schlondorff D. Signaling danger: toll-like receptors form a contiguous network throughout the entire kidney. Kidney Int 2006; and their potential roles in kidney disease. J Am Soc Nephrol 2004; 15: 70: 591–596. 854–867. 24. Mutant mice and neuroscience: recommendations concerning genetic 3. Tsuboi N, Yoshikai Y, Matsuo S et al. Roles of toll-like receptors in C-C background. Banbury conference on genetic background in mice. Neuron chemokine production by renal tubular epithelial cells. J Immunol 2002; 1997; 19: 755–759. 169: 2026–2033. 25. Sigmund CD. Viewpoint: are studies in genetically altered mice out of 4. Zhang DG, Zhang MS, Hayden MB et al. A toll-like receptor that prevents control? Arterioscler Thromb Vasc Biol 2000; 20: 1425–1429. infection by uropathogenic bacteria. Science 2004; 303: 1522–1526. 26. Bergers G, Reikerstorfer A, Braselmann S et al. Alternative usage 5. Hung CC, Chang CT, Chen KH et al. Upregulation of chemokine CXCL1/KC of the Fos-responsive gene Fit-1 generates mRNA isoforms coding for by leptospiral membrane lipoprotein preparation in renal tubule either secreted or membrane-bound proteins related to the IL-1 receptor. epithelial cells. Kidney Int 2006; 69: 1814–1822. EMBO J 1994; 13: 1176–1188. 6. Yang CW, Hung CC, Wu MS et al. Toll-like receptor 2 mediates early 27. Rossler U, Thomassen E, Hultner L et al. Secreted and membrane-bound inflammation by leptospiral outer membrane proteins in proximal tubule isoforms of T1, an orphan receptor related to IL-1-binding proteins, are cells. Kidney Int 2006; 69: 815–822. differently expressed in vivo. Dev Biol 1995; 168: 86–97. 7. Saemann MD, Weichhart T, Zeyda M et al. Tamm-Horsfall glycoprotein 28. Daniels MA, Hogquist KA, Jameson SC. Sweet ‘n’ sour: the impact of links innate immune cell activation with adaptive immunity via a Toll-like differential glycosylation on T-cell responses. Nat Immunol 2002; 3: receptor-4-dependent mechanism. J Clin Invest 2005; 115: 468–475. 903–910. 8. Poltorak A, He X, Smirnova I et al. Defective LPS signaling in C3H/HeJ and 29. Scheiffele P, Peranen J, Simons K. N-glycans as apical sorting signals in C57BL/10ScCr mice: mutations in Tlr4 gene. Science 1998; 282: epithelial cells. Nature 1995; 378: 96–98. 2085–2088. 30. Adib-Conquy M, Adrie C, Fitting C et al. Up-regulation of MyD88s and 9. Patole PS, Schubert S, Meßmer K et al. Toll-like receptor-4: intrinsic renal SIGIRR, molecules inhibiting toll-like receptor signaling, in monocytes cells and bone marrow-derived cells signal for neutrophil recruitment from septic patients. Crit Care Med 2006; 34: 2377–2385. during infective pyelonephritis. Kidney Int 2005; 68: 2582–2587. 31. Fagerlund R, Kinnunen L, Kohler M et al. NF-{kappa}B is transported into 10. Iwasaki A, Medzhitov R. Toll-like receptor control of the adaptive immune the nucleus by importin {alpha}3 and importin {alpha}4. J Biol Chem 2005; responses. Nat Immunol 2004; 5: 987–995. 280: 15942–15951.

192 Kidney International (2007) 72, 182–192