Cathepsins are required for Toll-like receptor 9 responses

Available online at www.sciencedirect.com

Biochemical and Biophysical Research Communications 367 (2008) 693–699 www.elsevier.com/locate/ybbrc

Cathepsins are required for Toll-like receptor 9 responses Fumi Matsumoto a,1, Shin-ichiroh Saitoh a,1, Ryutaroh Fukui a, Toshihiko Kobayashi a,b, Natsuko Tanimura a,b, Kazunori Konno a, Yutaka Kusumoto a, Sachiko Akashi-Takamura a, Kensuke Miyake a,* a

Division of Infectious Genetics, The Institute of Medical Science, Department of Microbiology and Immunology, The University of Tokyo, 4-6-1 Shirokanedai, Minatoku, Tokyo 108-8639, Japan b Japan Society for the Promotion of Science, Tokyo 102-8472, Japan Received 7 December 2007 Available online 31 December 2007

Abstract Toll-like receptors (TLR) recognize a variety of microbial products and activate defense responses. Pathogen sensing by TLR2/4 requires accessory molecules, whereas little is known about a molecule required for DNA recognition by TLR9. After endocytosis of microbes, microbial DNA is exposed and recognized by TLR9 in lysosomes. We here show that cathepsins, lysosomal cysteine proteases, are required for TLR9 responses. A cell line Ba/F3 was found to be defective in TLR9 responses despite enforced TLR9 expression. Functional cloning with Ba/F3 identified cathepsin B/L as a molecule required for TLR9 responses. The protease activity was essential for the complementing effect. TLR9 responses were also conferred by cathepsin S or F, but not by cathepsin H. TLR9-dependent B cell proliferation and CD86 upregulation were apparently downregulated by cathepsin B/L inhibitors. Cathepsin B inhibitor downregulated interaction of CpG-B with TLR9 in 293T cells. These results suggest roles for cathepsins in DNA recognition by TLR9. Ó 2007 Elsevier Inc. All rights reserved. Keywords: Toll-like receptor; Innate immunity; Protease

Innate immunity is the first line of defense against microbial infection [1]. The Toll family of receptors plays an essential role in innate recognition of microbial products and activation of defense responses [2]. Toll-like receptors (TLRs) are type I transmembrane proteins that contain a large, leucine-rich repeat in an extracellular region and a Toll/IL-1 receptor homology (TIR) domain in a cytoplasmic region [3]. TLRs require additional molecules for microbial recognition. Lipopolysaccharide (LPS) is recognized principally by the coreceptor MD-2 and signaled by TLR4 [4,5]. CD14 and LPS-binding protein (LBP) have a role in facilitating LPS interaction with the LPS sensor TLR4/MD-2. Tri-acylated lipopeptides in microbial membrane are directly recognized by TLR1/TLR2 heterodimer *

1

Corresponding author. Fax: +81 3 5449 5410. E-mail address: [email protected] (K. Miyake). These authors contributed equally to this study.

0006-291X/$ - see front matter Ó 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.12.130

[6], but CD14 has a role in enhancing responses to tri-acylated lipopeptides. Similarly, CD36 was shown to be required for TLR2/TLR6-dependent response to di-acylated lipopeptides [7]. TLRs can be divided into two groups based on their subcellular location. TLR1, 2, 4, 5, and 6 reside on the plasma membrane, whereas TLR3, 7, 8, and 9 are intracellular and signal from endosome/lysosome. Cell surface and intracellular TLRs recognize microbial membrane components and nucleic acids, respectively. Whereas, coreceptors and accessory molecules for cell surface TLRs have been identified and well-documented, those for intracellular, nucleic acid-sensing TLRs are largely unknown. Only a few molecules have been shown to modulate TLR9 responses to DNA. TLR9 responses to self DNA were enhanced by additional DNA-binding molecules such as anti-DNA Ab or an anti-microbial peptide LL37 [8,9], leading to autoimmune diseases like systemic lupus

694

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

erythematosus or psoriasis, respectively. HMGB-1, a nuclear DNA-binding protein released from necrotic cells, was reported to bind to DNA and augment TLR9 responses [10,11]. Regulatory roles for these DNA-associated molecules suggest a mechanism modulating TLR9 responses. It is possible that TLR9 requires additional molecules for TLR9 recognition as TLR4 requires MD-2, CD14, and LBP. Proteases mediate the hydrolysis of peptide bonds and can be classified, depending on the mechanism by which they elicit proteolysis, as aspartic, cysteine, serine or metalloproteases. Most mammalian lysosomal cysteine proteases are known as cathepsins, although not all cathepsins are cysteine proteases [12]. In human, there are 11 known cysteine cathepsins, cathepsin B, C, F, H, K, L, O, S, V, X, and W, all of which belong to the papain family of the clan CA cysteine proteases [13]. We have here identified cathepsins as molecules required for TLR3/7/9 responses. Possible roles for cathepsins in DNA recognition by TLR9 are also addressed in the present study. Materials and methods Antibodies and reagents. Anti-Flag-agarose, streptavidin–agarose, antiactin Ab, and lipid A were purchased from Sigma–Aldrich. Anti-TLR9 and anti-CD86 were purchased from eBioscience. Abs to IRAK-1, Grb2, cathepsin L, B, S, F, and H were purchased from Santa Cruz Biotechnology. Anti-IjBa was purchased from Cell Signaling TECHNOLOGY. Anti-GFP and LysoTracker were purchased from Invitrogen. CpG-B (50 TCCATGACGTTCCTGATGCT-30 ), or its derivatives were synthesized by Hokkaido System Science. FSL-1, MALP-2, and Pam3CSK4 were purchased from EMC microcollections. Loxoribine (7-allyl-7, 8-dihydro8-oxo-guanosine) and Poly(I:C) were purchased from InvivoGen. Cathepsin L inhibitor I, cathepsin B inhibitor CA-074 Me, leupeptin, and pepstatin were purchased from CALBIOCHEM. Cells and expression constructs. Murine IL-3 dependent cell line Ba/F3, murine B cell line M12, and macrophage cell line RAW264.7 were cultured as reported previously [14]. Mouse TLR9-flag or TLR9-GFP were cloned into pMXpuro retrovirus vector (supplied from Dr. Kitamura, Tokyo, Japan). The mouse cDNAs encoding TLR4, MD-2, and CD14 were cloned into pcDNA3 expression vector. The mouse cDNAs encoding cathepsin L, B, S, F, or H were cloned into pMX retrovirus vector. Cathepsin L mutants (H163A or G149R) and cathepsin B mutant (H199A) were generated by QuikChange Site-Directed Mutagenesis Kit (Stratagene). TLR9-GFP was also cloned into pCAGGS expression vector to transfect transiently into 293T cells by lipofection (Lipofectamine2000, Invitrogen). Functional cloning. cDNA was synthesized from RAW264.7 cells or murine spleens, and cloned into pMX retrovirus vector. About 2 million independent colonies were obtained in each library. These cDNA libraries were packaged and transduced into Ba/F3 cells expressing TLR4/MD-2, CD14, pNF-jB-hrGFP and TLR9-flag. These cells were stimulated with 100 nM CpG-B for 24 h and activated cells were selected for the expression of GFP by sorting with flow cytometry. cDNAs derived from the library were recovered from genomic DNA by PCR. CpG-DNA uptake assays. M12 cells were treated with various inhibitors and then were stimulated with CpG-B-FITC. After incubation, cells were washed with a staining buffer (2.5% FCS and 0.01% NaN3 in PBS) twice and cell surface CpG-B-FITC was quenched with 50% Trypan blue. The fluorescence intensity inside the cells was measured by flow cytometry. Ligand-binding studies. CpG-B interaction with TLR9 was conducted as described previously [15]. 293T cells transiently expressing TLR9-GFP were treated with various inhibitors and then stimulated with biotinylated

CpG-B. After washing twice, cells were lysed and subjected to immunoprecipitation with streptavidin–agarose and coprecipitated TLR9-GFP was immunoprobed with anti-GFP Ab.

Results and discussion The lack of CpG-B-induced NF-jB activation in Ba/F3 cells To search for a molecule required for TLR9 responses, we first looked for a cell line that does not respond to TLR9 ligand CpG-B. IL-3-dependent cell line Ba/F3 was unable to activate NF-jB in response to CpG-B, whereas B cell line M12 and macrophage line RAW264.7 activated NF-jB as revealed by IjBa degradation (Fig. 1A). We also examined degradation of IL-1R-associated kinase-1 (IRAK-1), which is tightly coupled with IRAK-1 phosphorylation [16]. CpG-B-induced IRAK-1 degradation was also severely impaired in Ba/F3 cells. When stimulated with TLR2/TLR6 ligand FSL-1, NF-jB activation and IRAK-1 degradation were seen in all the three cell lines (Fig. 1A), demonstrating that Ba/F3 is not defective in the common TLR signaling pathway. Another possibility was next addressed that Ba/F3 is specifically impaired in TLR9 responses. To make sure the expression of TLR9 protein, we established and studied Ba/F3 cells expressing TLR9-flag, which was confirmed by immunoprecipitation with the anti-flag Ab and immunoprobing with the antiTLR9 Ab (Fig. 1B, left). Despite expression of TLR9 protein, CpG-B-dependent IjBa degradation remained undetectable (Fig. 1B, right). To further confirm defective NFjB activation in CpG responses, we established Ba/F3 cells expressing TLR9-GFP and a NF-jB reporter gene, NFjB-luciferase. TLR9-GFP expression was confirmed by immunoprecipitation and immunoprobing with anti-GFP Ab (Fig. 1C, top). NF-jB activation was observed only when Ba/F3 cells were stimulated with a TLR2/TLR6 ligand MALP-2 (Fig. 1C, bottom). TLR9-GFP expression did not make any difference from Ba/F3 cells. By contrast, NF-jB activation in response to CpG as well as TLR2/ TLR6 ligand MALP-2 was clearly detected in B cell line M12 expressing NF-jB-luciferase (Fig. 1C, bottom). Functional cloning of cathepsin B and L as a molecule required for CpG-B-induced NF-jB activation in Ba/F3 cells To identify a molecule required for CpG-dependent NFjB activation in Ba/F3 cells, we conducted functional cDNA cloning. To detect NF-jB activation with flow cytometry, we established Ba/F3 cells stably expressing NF-jB-GFP and TLR9. TLR4/MD-2 and CD14 were also expressed to confirm GFP induction with lipid A. We chose a clone in which GFP expression is almost undetectable in the resting state and upregulated about two orders of magnitude 24 h after lipid A stimulation (Fig. 2A, compare histograms with solid and dotted lines). Retroviral vectorbased cDNA library was prepared from CpG-responsive RAW264.7 cells or spleen cells, and transduced into the

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

Ba/F3

A 0

10

30

60 120

0

10

30

60 120 min

IRAK-1 I κB α Grb2

M12 IRAK-1 I κB α Grb2 RAW264.7 IRAK-1 I κB α Grb2 CpG-B

B

FSL-1

Ba/F3 Ba/F3 /TLR9-flag

0

10

30 60 min

I κB α Grb2

IP: flag WB: TLR9

C

CpG-B

Ba/F3 Ba/F3/κ B / κB /TLR9-GFP IP & WB: GFP

**

100000 90000 80000 70000

*: p<0.05 **: p<0.01 **

**

**

60000 50000 40000 30000 20000 10000 0

(-) CpG M Ba/F3/κ B

(-) CpG M Ba/F3/κ B /TLR9-GFP

(-) CpG M M12/ κ B

Fig. 1. The lack of TLR9-dependent NF-jB activation in Ba/F3 cells. (A) Ba/F3, M12, and RAW264.7 cells were stimulated with TLR9 ligand CpG-B at 1 lM (left) or TLR2/6 ligand FSL-1 at 1 lg/ml (right). Cells were lysed and subjected to immunoprobing with the indicated antibodies. (B) Ba/F3 cells or those expressing TLR9-flag were subjected to immunoprecipitation with anti-flag Ab and immunoprobing with antiTLR9 Ab (left). Ba/F3 cells expressing TLR9-flag were stimulated with CpG-B (1 lM). Cells were lysed and subjected to SDS–PAGE and immunoprobed with the indicated antibodies (right). (C) Ba/F3 cells expressing the NF-jB-luciferase reporter gene with or without TLR9-GFP were subjected to lysate extraction, SDS-PAGE and immunoprobing with anti-GFP Ab (top). These cells were stimulated with CpG-B (1 lM) or TLR2/6 ligand MALP-2 (5 lg/ml) for 5 h. NF-jB activation was assessed by luciferase activity (bottom). Data are represented with the mean values with standard deviation from triplicate samples. Similar results were obtained from three independent experiments.

695

Ba/F3 cells expressing TLR4/MD-2, CD14, NF-jB-GFP and TLR9-flag. GFP-positive cells after CpG-B stimulation for 24 h were sorted, and allowed to grow until GFP expression returned to the basal level. After three round of sorting, cDNAs were recovered and retransduced to identify a cDNA conferring CpG-responsiveness on Ba/ F3 cells. Functional cloning finally identified cathepsin B and cathepsin L as a molecule required for TLR9 responses in Ba/F3 cells. When cathepsin B/L was expressed, CpGdependent GFP upregulation was observed, although CpG-B-induced GFP-upregulation was significantly lower than lipid A-induced upregulation (Fig. 2A). We examined expression of mRNA encoding cathepsin B/L in Ba/F3, M12, and RAW264.7 cells. All the cell lines including Ba/F3 expressed cathepsin B/L mRNA. Cathepsin L mRNA was apparently lower in Ba/F3 cells than the other two cell lines whereas no difference in expression of cathepsin B (data not shown). Although cathepsin B/L mRNA is expressed in Ba/F3 cells, expression of the cathepsin proteins might be lower in Ba/F3 cells than in M12 or RAW296.7 cells, and could not be sufficient for TLR9 responses. Indeed, expression of cathepsin B/L protein was hardly detectable with immunoprobing of whole cell lysates (Fig. 2A). Apparently higher expression of cathepsin B/L in transfected cells upregulated GFP induction in response to CpG-B. We next asked requirement for the protease activity of cathepsin B/L in TLR9 activation. We used cathepsin mutants substituting histidine in the active site for alanine (H163A for cathepsin L and H199A for cathepsin B) [13]. We also used a spontaneous cathepsin L mutation (G149R) found in the furless (fs) mice [17]. The apparent molecular mass of these mutant cathepsins seemed to be distinct from wild type cathepsins probably due to impaired processing from pro-enzyme to mature enzyme (Fig. 2A). All the mutants lacking protease activity were expressed as much as wild type cathepsin B/L, but failed to confer CpG-B responsiveness on Ba/F3 cells (Fig. 2A). Cathepsin B and L belong to the papain family of the clan CA cysteine proteases. There are 11 known cysteine cathepsins in human. In the case of rodents, there are 10 known cathepsins including cathepsins B, C, F, H, K, L, O, S, X, and W. Among them, cathepsin B, F, H, K, L, and S are endopeptidases. We next examined if other endopeptidase cathepsins (cathepsin S, F, and H) are able to influence CpG-B responses. Cathepin K was not included, because its expression is restricted to osteoclasts or ovary [18]. Interestingly, cathepsin S and F were able to confer CpG-B responsiveness on Ba/F3 cells, whereas cathepsin H was not. Further study is required to reveal differences between cathepsin H and the rest of the cathepsins. Cathepsin inhibitors suppress TLR3/7/9 responses in splenic B cells To study roles for cathepsins in endogenously expressed TLR9, we studied the effect of cathepsin

696

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

A

mock None CpG-B (100nM) LipidA (1 μg/ml)

Cathepsin L mock Cts

L (H163A)

Cathepsin B

B (H199A)

L (G149R)

procathepsin cathepsin

B

Cathepsin S

Cathepsin F

Cathepsin H

Fig. 2. Cathepsin augments CpG-B-induced NF-jB activation in Ba/F3 cells. Ba/F3 cells expressing TLR4/MD-2, CD14, TLR9-flag, and NF-jB-GFP reporter gene (left, mock) were further transfected with: cathepsin L, cathepsin L mutants (H163A or G149R), cathepsin B, cathepsin B mutant (H199A), cathepsin S, cathepsin F, or cathepsin H as indicated in (A and B). Expression of cathepsin was shown by immunoprobing of whole cell lysates with the corresponding anti-cathepsin Abs. NF-jB-dependent GFP-induction in these cells stimulated with CpG-B at 100 nM (closed histograms) or lipid A at 1 lg/ml (histograms with dotted lines) for 24 h was shown. Open histograms with solid lines show GFP expression in unstimulated cells. Data are representative of three independent experiments.

inhibitors in CpG-B-induced NF-jB activation in B cell line M12. We used inhibitors for cysteine protease (leupeptin), cathepsin L (Li-I), cathepsin B (CA-074 Me), or aspartic protease inhibitor (pepstatin) as a negative control. M12 cells were stimulated with CpG-B or TLR2/ TLR6 ligand FSL-1, and NF-jB activation was detected by luciferase activity (Fig. 3A). All the cathepsin inhibitors, but not an aspartic inhibitor pepstatin, significantly suppressed CpG-B-induced NF-jB activation, whereas no significant effect was seen in FSL-1 responses. To confirm these results with primary B cells in the spleen, cathepsin inhibitors were added in B cell stimulation with CpG-B. We also used other TLR ligands including TLR7 ligand Loxoribine, TLR3 ligand polyI:C, and TLR1/TLR2 ligand Pam3CSK4 (Fig. 3B). After 24 h culture, cell surface CD86 was stained. Upregulation of costimulatory molecule CD86 was apparently downregulated by cathepsin inhibitors but not by pepstatin. Similar inhibition was observed in B cell response to Loxoribine or polyI:C, but not to TLR1/TLR2 ligand Pam3CSK4 (Fig. 3B). TLR ligandinduced B cell proliferation was also studied. All the protease inhibitors except pepstatin significantly inhibited B cell proliferation induced by CpG-B or Loxoribine, but not by Pam3CSK4 (Fig. 3C). These results suggest that the cysteine protease activity of cathepsins was required for B cell responses to TLR9 and probably to TLR7 and TLR3.

Cathepsin L and B inhibitors suppress the CpG-B recognition by TLR9 To address a mechanism how cathepsin is required for TLR9 responses, we first studied IRAK-1 degradation, which is coupled with IRAK-1 phosphorylation. CpG-Bdependent IRAK-1 degradation was observed in M12 and RAW264.7 cells but not in Ba/F3 cells (Fig. 1A). Given that cathepsin is required for TLR9 responses, cathepsin inhibitors are likely to inhibit CpG-B-dependent IRAK-1 degradation in M12 or RAW264.7 cells. TLR9GFP-expressing M12 or RAW264.7 cells were treated with cathepsin inhibitors, and then stimulated with CpG-B. IRAK-1 degradation upon CpG-B stimulation in M12 cells was significantly suppressed by the inhibitors for cathepsin B or L, whereas that IRAK-1 degradation upon stimulation with TLR2/TLR6 ligand FSL-1 was not altered (Fig. 4A). Similar results were obtained with RAW264.7 cells (Fig. 4B). The effect of cathepsin inhibitors is specific for TLR9 but not for TLR2/TLR6. Moreover, cathepsins reside in endosome/lysosome but not in the cytoplasm. Given that TLR9 recognizes CpG-B in endosome/lysosome, cathepsin is likely to have a role in CpG-B recognition by TLR9. To reveal a mechanism by which cathepsins influence CpG-B recognition, we first addressed a possibility that cathepsin is required for CpG-B uptake. B cell line M12 was incubated with CpG-B conjugated with FITC (CpG-B-FITC)

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

A

*: p<0.05 **: p<0.01 B

35000

** ** *

30000 25000

(-)

CpG-B

Loxoribine

697

polyI:C

Pam3CSK4

DMSO

20000

leupeptin

15000

Li- Ι

10000

Stim:

(-)

DM leu S O pe p ti n CA L -07 i- Ι p e p 4 Me s ta ti n

Inhibitor:

DM leu S O pe p ti n CA L -07 i- Ι 4 pep Me s ta ti n

0

DM leu S O pe p ti n CA L -07 i- Ι p e p 4 Me s ta ti n

5000

CpG-B

FSL-1

CA-074 Me pepstatin CD86

C

*: p<0.05 **: p<0.01

70000

** ** **

60000 50000

**** **

40000 30000 20000

(-)

CpG-B

Loxoribine

D l e u MS O pe p ti n CA -07 Li- Ι pep 4 Me s ta ti n

D leu MSO pe p ti n CA -07 Li- Ι p e p 4 Me s ta ti n

D leu MSO pe p ti n CA -07 Li- Ι 4 pep Me s ta ti n

0

D leu MSO pe p ti n CA -07 Li- Ι 4 pep Me s ta ti n

10000

Pam3CSK4

Fig. 3. TLR3/7/9 responses in splenic B cells were suppressed by cathepsin inhibitors. (A) M12 cells expressing the NF-jB-luciferase reporter gene were treated with vehicle DMSO, cysteine protease inhibitor leupeptin (250 lM), cathepsin L inhibitor I (Li-I) (10 lM), cathepsin B inhibitor CA-074 Me (10 lM), aspartic protease inhibitor pepstatin (25 lM) for 2 h, and then stimulated with CpG-B (10 nM) or FLS-1 (100 ng/ml) for 5 h. NF-jB activation was assessed by luciferase activity. (B) Enriched splenic B cells were treated with DMSO, leupeptin (250 lM), Li-I (3 lM), CA-074 Me (3 lM), pepstatin (25 lM) for 2 h, and then stimulated with CpG-B (10 nM), Loxoribine (1 lM), poly(I:C) (30 lg/ml), or Pam3CSK4 (100 ng/ml) for 24 h. Cells were stained with anti-CD86 Ab. Shaded histograms show CD86 expression on B cells, whereas open histograms show control staining with the second reagent alone. (C) Enriched splenic B cells were treated with DMSO, leupeptin (25 lM), Li-I (1 lM), CA-074 Me (1 lM), pepstatin (2.5 lM) for 2 h, and then stimulated with CpG-B (10 nM), Loxoribine (1 lM), or Pam3CSK4 (100 ng/ml) for 2 d. 1 lCi [3H] thymidine was pulsed for the last 6 h, and incorporated thymidine was counted by a scintillation counter. Data are represented by the mean values with standard deviations from triplicate wells. Similar results were obtained from three independent experiments. *p < 0.05, **p < 0.01.

at 37 or 4 °C with sodium azide for 2 h. After quenching cell surface CpG-B-FITC, cells were analyzed on flow cytometry, and the amount of CpG-B-FITC inside the M12 cells was determined by comparing M12 cells incubated with 37 °C and those kept 4 °C with sodium azide. Significant CpG-B-FITC uptake was seen in 30 and 120 min after addition of CpG-B-FITC (Fig. 4C). Pretreatment with inhibitors for cathepsin B/L had no effect on CpG-B-FITC uptake. Cathepsin might have a role in CpG-B trafficking to lysosome, where TLR9 recognizes CpG-B. The subcellular distribution of CpG-B-rhodamine was studied with confocal microscopy. M12 was stimulated with CpG-B-rhodamine, and incubated for 2 h. CpG-B-rhodamine was visualized with LysoTracker, a marker for lysosome

(Fig. 4D). CpG-B-rhodamine colocalized with LysoTracker in M12 cells and pretreatment with cathepsin inhibitors had no effect on the subcellular distribution of CpG-B-rhodamine (Fig. 4D). Finally, we addressed a possibility that cathepsins have a role in the interaction of CpG-B with TLR9. As reported previously [15], the interaction of CpG-B with TLR9 can be detected with 293T cells transiently expressing TLR9. 293T cells expressing TLR9GFP were stimulated with biotinylated CpG-B, and CpG-B was precipitated with streptavidin–agarose. Coprecipitated TLR9-GFP was probed with anti-GFP Ab. Cathepsin B inhibitor CA-074 Me but not cathepsin inhibitor Li-I was able to downregulate coprecipitation of TLR9-GFP (Fig. 4E), suggesting a possible role for cathepsin B in the interaction of CpG-B with TLR9.

698

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

A

Cell: M12/TLR9-GFP Inhibitor: DMSO -+ IRAK-1

Li-Ι -+-+

CA-074 Me

DMSO Li-Ι -+ -+

D

CA-074 Me -+

(NC) DMSO Green: lysosome

DIC

Red: CpG-rho

merge

actin CpG-B (3 μM)

B

FSL-1 (1 μg/ml)

Cell: RAW/TLR9-GFP Inhibitor: DMSO Li-Ι -+ -+ IRAK-1

CA-074 Me -+

DMSO Li-Ι -+ -+

CA-074 Me -+

CA-074 Me

Li-Ι

actin CpG-B (1 μM) 30

C

FSL-1 (1 μg/ml) 120

min

DMSO

E

Cell: 293T/TLR9-GFP

Li-Ι

CA-074 Me

Stim (-) CpG-B-FITC 4 OC with 0.1%NaN3 CpG-B-FITC 37 OC

Inhibitor : CpG-B bio :

-

pepstatin Li-Ι +++

CA-074 Me -+

StAv IP WCL WB: GFP

Fig. 4. Effect of cathepsin inhibitors on CpG-B binding to TLR9. (A,B) M12 cells expressing TLR9-GFP (A) or RAW264.7 cells expressing TLR9-GFP (B) were treated with DMSO, Li-I (30 lM), or CA-074 Me (30 lM) for 2 h, and then stimulated with CpG-B or FSL-1. After 2 h, cells were lysed and subjected to SDS–PAGE and immunoprobing of IRAK-1. (C) M12 cells were treated with DMSO, Li-I (30 lM), or CA-074 Me (30 lM) for 2 h, and then incubated with CpG-B-FITC (1 lM). After 30 or 120 min incubation, CpG-B-FITC on the cell surface was quenched with 50% Trypan blue and CpG-BFITC inside the cells were analyzed by flow cytometry. Shaded histograms show cells without any treatment, whereas histograms with solid line or dashed line show cells incubated with CpG-B-FITC (1 lM) at 37 °C or at 4 °C in the presence of 0.1% sodium azide, respectively. (D) M12 cells were treated as in C for 2 h, and then incubated with CpG-B-rhodamine (3 lM) for 2 h. Cells were stained with a lysosome marker LysoTracker for the last 10 min, and CpG-B-rhodamine (red) and lysosome (green) were visualized by confocal microscopy. D, 293T cells transiently expressing TLR9-GFP were treated with DMSO, pepstatin (75 lM), Li-I (30 lM), or CA-074 Me (30 lM) for 2 h at 37 °C. Cells were then incubated with biotinylated CpG-B (3 lM) for 2 h at 37 °C. Biotinylated CpG-B was precipitated from cell lysates with streptavidin–agarose and coprecipitated TLR9-GFP was immunoprobed with anti-GFP Ab (upper). TLR9-GFP expression was shown by immunoprobing of whole cell lysates (lower). Similar results were obtained in three independent experiments.

The present study identified cathepsins as molecules required for TLR9 recognition. Cathepsin inhibitors were not able to affect CpG-B uptake or subsequent trafficking to lysosome (Fig. 4C and D), whereas Cathepsin B inhibitor suppressed the interaction of CpG-B with TLR9 in 293T cells (Fig. 4E), suggesting a role for cathepsins in CpG-B binding to TLR9. Given that the protease activity is required for their complementing effect (Fig. 2A), a substrate might have a role in CpG-B binding to TLR9. In this regard, it is of note that TLR9 responses are influenced by a variety of DNA-binding molecules such as HMGB-1, anti-microbial peptide LL37, and HSP90 [8,10,11,19]. Cathepsins could regulate CpG-B-TLR9 interaction by acting on these DNA-binding molecules. CpG-B binding to TLR9 might not be a simple interaction but a complex reaction requiring cathepsins. Further study has to address this possibility. Cathepsin L inhibitor inhibited CpG-B responses and CpG-B-dependent IRAK-1 degradation (Figs. 3 and 4),

but not CpG-B binding to TLR9 (Fig. 4E), suggesting that cathepsin L is required after CpG-B binding to TLR9. Cathepsin L might have a role in CpG-B-dependent conformation change of TLR9. Cathepsins seem to contribute to at least two distinct steps of CpG-B recognition, and Ba/ F3 cells are likely to be impaired in both steps. Two more cathepsins, cathepsin S and F, complemented TLR9 responses in Ba/F3 cells (Fig. 2B). As is the case with antigen presentation [12], cathepsins may have a redundant role in CpG-B recognition by TLR9. Further study has to reveal the relationship among cathepsins in their effects on TLR9 responses. Acknowledgments The authors thank Mr. Yuji Motoi for his technical assistance. This work was supported (in part) by: a contract research fund from the Ministry of Education, Culture, Sports, Science and Technology for Program of

F. Matsumoto et al. / Biochemical and Biophysical Research Communications 367 (2008) 693–699

Founding Research Centers for Emerging and Reemerging Infectious Diseases; and Strategic cooperation to control emerging and reemerging infections funded by The Special Coordination Funds for Promoting Science and Technology of Ministry of Education, Culture, Sports, Science and Technology (MEXT). References [1] C. Janeway, R. Medzhitov, Innate immune recognition, Annual Review of Immunology 20 (2002) 197–216. [2] K. Takeda, T. Kaisho, S. Akira, Toll-like receptors, Annual Review of Immunology 21 (2003) 335–376. [3] R. Medzhitov, P. Preston-Hurlburt, C.A. Janeway Jr., A human homologue of the Drosophila Toll protein signals activation of adaptive immunity, Nature 388 (1997) 394–397. [4] H.M. Kim, B.S. Park, J.I. Kim, S.E. Kim, J. Lee, S.C. Oh, P. Enkhbayar, N. Matsushima, H. Lee, O.J. Yoo, J.O. Lee, Crystal structure of the TLR4-MD-2 complex with bound endotoxin antagonist Eritoran, Cell 130 (2007) 906–917. [5] U. Ohto, K. Fukase, K. Miyake, Y. Satow, Crystal structures of human MD-2 and its complex with antiendotoxic lipid IVa, Science 316 (2007) 1632–1634. [6] M.S. Jin, S.E. Kim, J.Y. Heo, M.E. Lee, H.M. Kim, S.G. Paik, H. Lee, J.O. Lee, Crystal structure of the TLR1–TLR2 heterodimer induced by binding of a tri-acylated lipopeptide, Cell 130 (2007) 1071–1082. [7] K. Hoebe, P. Georgel, S. Rutschmann, X. Du, S. Mudd, K. Crozat, S. Sovath, L. Shamel, T. Hartung, U. Zahringer, B. Beutler, CD36 is a sensor of diacylglycerides, Nature 433 (2005) 523–527. [8] R. Lande, J. Gregorio, V. Facchinetti, B. Chatterjee, Y.H. Wang, B. Homey, W. Cao, Y.H. Wang, B. Su, F.O. Nestle, T. Zal, I. Mellman, J.M. Schroder, Y.J. Liu, M. Gilliet, Plasmacytoid dendritic cells sense selfDNA coupled with antimicrobial peptide, Nature 449 (2007) 564–569. [9] E.A. Leadbetter, I.R. Rifkin, A.M. Hohlbaum, B.C. Beaudette, M.J. Shlomchik, A. Marshak-Rothstein, Chromatin-IgG complexes activate B cells by dual engagement of IgM and Toll-like receptors, Nature 416 (2002) 603–607.

699

[10] J. Tian, A.M. Avalos, S.Y. Mao, B. Chen, K. Senthil, H. Wu, P. Parroche, S. Drabic, D. Golenbock, C. Sirois, J. Hua, L.L. An, L. Audoly, G. La Rosa, A. Bierhaus, P. Naworth, A. MarshakRothstein, M.K. Crow, K.A. Fitzgerald, E. Latz, P.A. Kiener, A.J. Coyle, Toll-like receptor 9-dependent activation by DNA-containing immune complexes is mediated by HMGB1 and RAGE, Nature Immunology 8 (2007) 487–496. [11] S. Ivanov, A.M. Dragoi, X. Wang, C. Dallacosta, J. Louten, G. Musco, G. Sitia, G.S. Yap, Y. Wan, C.A. Biron, M.E. Bianchi, H. Wang, W.M. Chu, A novel role for HMGB1 in TLR9-mediated inflammatory responses to CpG-DNA, Blood 110 (2007) 1970–1981. [12] K. Honey, A.Y. Rudensky, Lysosomal cysteine proteases regulate antigen presentation, Nature Reviews Immunology 3 (2003) 472–482. [13] D. Turk, G. Guncar, Lysosomal cysteine proteases (cathepsins): promising drug targets, Acta Crystallographica Section D Biological Crystallography 59 (2003) 203–213. [14] M. Kobayashi, S. Saitoh, N. Tanimura, K. Takahashi, K. Kawasaki, M. Nishijima, Y. Fujimoto, K. Fukase, S. Akashi-Takamura, K. Miyake, Regulatory roles for MD-2 and TLR4 in ligand-induced receptor clustering, Journal of Immunology 176 (2006) 6211–6218. [15] E. Latz, A. Schoenemeyer, A. Visintin, K.A. Fitzgerald, B.G. Monks, C.F. Knetter, E. Lien, N.J. Nilsen, T. Espevik, D.T. Golenbock, TLR9 signals after translocating from the ER to CpG DNA in the lysosome, Nature Immunology 5 (2004) 190–198. [16] M. Yamamoto, S. Sato, H. Hemmi, H. Sanjo, S. Uematsu, T. Kaisho, K. Hoshino, O. Takeuchi, M. Kobayashi, T. Fujita, K. Takeda, S. Akira, Essential role for TIRAP in activation of the signalling cascade shared by TLR2 and TLR4, Nature 420 (2002) 324–329. [17] W. Roth, J. Deussing, V.A. Botchkarev, M. Pauly-Evers, P. Saftig, A. Hafner, P. Schmidt, W. Schmahl, J. Scherer, I. Anton-Lamprecht, K. Von Figura, R. Paus, C. Peters, Cathepsin L deficiency as molecular defect of furless: hyperproliferation of keratinocytes and pertubation of hair follicle cycling, FASEB Journal 14 (2000) 2075–2086. [18] T. Zavasnik-Bergant, B. Turk, Cysteine cathepsins in the immune response, Tissue Antigens 67 (2006) 349–355. [19] F.G. Zhu, D.S. Pisetsky, Role of the heat shock protein 90 in immune response stimulation by bacterial DNA and synthetic oligonucleotides, Infection and Immunity 69 (2001) 5546–5552.