Epithelial Cells Through NOD1 Bacterial Peptidoglycans, Activate

Meso-Diaminopimelic Acid and Meso -Lanthionine, Amino Acids Specific to Bacterial Peptidoglycans, Activate Human Epithelial Cells through NOD1 This information is current as of September 26, 2021. Akiko Uehara, Yukari Fujimoto, Akiko Kawasaki, Shoichi Kusumoto, Koichi Fukase and Haruhiko Takada J Immunol 2006; 177:1796-1804; ; doi: 10.4049/jimmunol.177.3.1796 http://www.jimmunol.org/content/177/3/1796 Downloaded from

References This article cites 25 articles, 12 of which you can access for free at:

http://www.jimmunol.org/content/177/3/1796.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 26, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Meso-Diaminopimelic Acid and Meso-Lanthionine, Amino Acids Speciﬁc to Bacterial Peptidoglycans, Activate Human Epithelial Cells through NOD11

Akiko Uehara,* Yukari Fujimoto,† Akiko Kawasaki,† Shoichi Kusumoto,† Koichi Fukase,† and Haruhiko Takada2*

Peptidoglycans (PGNs) are ubiquitous constituents of bacterial cell walls and exhibit various immunobiological activities. Two types of minimum essential PGN structures for immunobiological activities were chemically synthesized and designated as muramyldipeptide; N-acetylmuramyl-L-alanyl-D-isoglutamine (MDP) and ␥-D-glutamyl-meso-diaminopimelic acid (iE-DAP), which are common constituents of both Gram-positive and Gram-negative bacteria, as well as most Gram-negative and some Gram- positive bacteria, respectively. Recently, intracellular receptors for MDP and iE-DAP have been demonstrated to be nucleotide- Downloaded from binding oligomerization domain (NOD)1 and NOD2, respectively. In this study, we demonstrated that chemically synthesized meso-DAP itself activated human epithelial cells from various tissues, through NOD1 to generate antibacterial factors, PGN recognition proteins and ␤-defensin 2, and cytokines in specified cases, although the activities of meso-DAP were generally weaker than those of known NOD agonists. However, stereoisomers of meso-DAP, LL-DAP, and DD-DAP were only slightly activated or remained inactive, respectively. Synthetic meso-lanthionine, which is another diamino-type amino acid specific to PGN of the specified Gram-negative bacteria, was also recognized by NOD1. In human monocytic cells, in the presence of cytochalasin D http://www.jimmunol.org/ meso-DAP induced slightly but significantly increased production of cytokines, although the cells did not respond to meso-DAP in the absent of cytochalasin D. Our findings suggest that NOD1 is a special sentinel molecule, especially in the epithelial barrier, allowing the intracellular detection of bacteria through recognizing meso-DAP or comparable moiety of PGN from specified bacteria in cooperation with NOD2, thereby playing a key role in innate immunity. The Journal of Immunology, 2006, 177: 1796–1804.

he peptidoglycans (PGNs)3 are ubiquitous constituents of that intracellular protein carrying a nucleotide-binding oligomer- the cell walls of Gram-positive and Gram-negative bac- ization domain (NOD), NOD2, was intracellular receptor for MDP by guest on September 26, 2021 T teria. Numerous studies have reported on the immunobio- (6, 7). Then, the two groups also independently revealed that DMP activities of bacterial PGN, most of which have been reproduced containing DAP was recognized by another NOD protein, NOD1 by a chemically synthesized low-molecular PGN fragment, mu- (8, 9). Recently, it has been reported that iE-DAP is the minimum ramyldipeptide (MDP) N-acetylmuramyl-L-alanyl-D-isoglutamine PGN structure capable of stimulating NOD1 on human embryonic (or MurNAc-L-Ala-D-isoGln) (Fig. 1A) (1, 2). Another type of kidney cell line cells (HEK 293T) transfected with human NOD1 PGN fragment, desmuramylpeptides (DMPs), has also been chem- (6, 10). In these reports, Chamaillard and colleagues (6) and Gi- ically synthesized to mimic PGN containing meso-diaminopimelic rardin et al. (10) described that DAP alone is completely inactive acid (meso-DAP), and the DMPs exerted similar bioactivities to in this respect by using a commercial (Sigma-Aldrich) DAP, a MDP (Fig. 1B) (3, 4). In 1982, the Fujisawa Pharmaceutical group mixture of meso-DAP, LL-DAP and DD-DAP, and meso-DAP pu- ␥ (5) reported that -D-glutamyl-meso-DAP (iE-DAP) is the mini- rified from the commercial DAP, respectively. mum structural unit capable of eliciting bioactivities induced by In various mucosal tissues, epithelial cells interact with bacteria DMPs. Recently, two research groups independently demonstrated in normal flora and have been reported to produce several proinflammatory cytokines in response to bacterial stimuli. We re- *Department of Microbiology and Immunology, Tohoku University Graduate School vealed, however, that oral epithelial cells did not show enhanced of Dentistry, Sendai, Japan; and †Department of Chemistry, Graduate School of Sci- production of proinflammatory cytokines such as IL-8 in response ence, Osaka University, Osaka, Japan to various bacterial cell surface components, whereas several co- Received for publication December 6, 2005. Accepted for publication April 26, 2006. lonic epithelial cell lines produced inflammatory cytokines upon The costs of publication of this article were defrayed in part by the payment of page stimulation with bacterial components (11). In contrast, we re- charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. cently demonstrated that chemically synthesized bacterial cell sur- 1 This work was supported in part by Grant-in-Aid 16390519 for Scientific Research face components increase the expression of PGN recognition pro- from the Japan Society for the Promotion of Science. A.U. was supported by Research teins (PGRPs) via TLRs, NOD1 and NOD2 in human oral Fellowship PD6961 from the Japan Society for the Promotion of Science. epithelial cells (12). PGRPs are a novel family of pattern recog- 2 Address correspondence and reprint requests to Dr. Haruhiko Takada, Department nition receptor molecules in innate immunity conserved from in- of Microbiology and Immunology, Tohoku University Gradate School of Dentistry, 4-1 Seiryo-machi, Aoba-ku, Sendai 980-8575, Japan. E-mail address: dent- sects to mammals. These findings suggest that epithelial cells in- [email protected] teracting constitutionally with normal flora might respond to 3 Abbreviations used in this paper: PGN, peptidoglycan; MDP, muramyldipeptide; bacterial cell surface components with the enhanced production of DMP, desmuramylpeptide; DAP, diaminopimelic acid; iE-DAP, ␥-D-glutamyl-meso- DAP; NOD, nucleotide-binding oligomerization domain; PGRP, PGN recognition antibacterial molecules, such as PGRPs, to prevent bacterial inva- protein; siRNA, short interfering RNA. sion of host cells without producing proinflammatory cytokines to

FIGURE 1. Chemically synthesized bacterial PGN fragments. Chemical structure of Lys-type (A) and DAP-type (B) PGN and the respective minimum active moiety reported to sense NOD2 and NOD1 and chemical structures of synthetic DAP isomers and reference synthetic NOD1 agonists (C). NOD2 recognizes MDP (MurNAc-L-Ala-D-isoGln) moiety from Lys-type PGN (A) and probably also MurNAc-L-Ala-D-Glu moiety from DAP-type PGN (B). NOD1 recognizes the DMP moiety containing meso-DAP (B), whose minimum essential structure was reported to be iE-DAP (␥-D-Glu-meso-DAP) until this study. iE-DAP, FK156 (D-lactyl-L-Ala-␥-D-Glu-meso-DAP-Gly), and FK565 (heptanoly-␥-D-Glu-meso-DAP-D-Ala) were used as reference NOD1 agonists. prevent possible tissue destruction through by excessive innate im- Materials and Methods mune responses. Reagents In the course of the study to examine the innate immune system in oral epithelium, we found unexpectedly that a commercial DAP spec- The synthetic MurNAc-L-Ala-D-isoGln (MDP) was purchased from Protein Research Foundation, Peptide Institute. A synthetic Escherichia imen activated human oral epithelial cells. Therefore, to further char- coli-type lipid A (LA-15-PP) was purchased from Daiichi Chemical. acterize the bacterial moiety recognized by NOD1, we have con- ␥-D-glutamyl-meso-DAP (iE-DAP) was synthesized as previously de- ducted an exhaustive study using three chemically synthesized scribed (8). Meso-lanthionine was synthesized by coupling of N-ben- ␤ stereoisomers of DAP, meso-DAP, LL-DAP, and DD-DAP, in human zyloxycarbonyl-D-cysteine with -L-chloroalanine according to the pro- epithelial and monocytic cells in culture. Furthermore, we examined cedure reported by Photaki et al. (14). The synthetic DMP, a PGN fragment containing DAP, FK156 (D-lactoyl-L-Ala-␥-D-Glu-meso-DAP- whether synthetic meso-lanthionine, which is another diamino-type Gly) and its derivative, FK565 (heptanoyl-␥-D-Glu-meso-DAP-D-Ala) (5), was amino acid speciﬁc to PGN of some Gram-negative bacteria such as supplied by Astellas Pharmaceutical (formerly Fujisawa and Yamanouchi). Fusobacterium nucleatum (13), also activated human epithelial cells. Cytochalasin D was purchased from Sigma-Aldrich. 1798 DAP ACTIVATES EPITHELIAL CELLS VIA NOD1

FIGURE 2. Induction of PGRP in oral epithelial cells stimulated with meso-DAP. A, Human oral epithelial cell line HSC-2 cells were incubated with or without (unfractionated) commercial DAP, L-Lys, or D-Gln at the indicated doses. After 8 h, the mRNA expressions of PGRP-L, PGRP-I␣, PGRP-I␤, and Downloaded from PGRP-S were analyzed by real-time PCR using specific primers. Data are expressed as mean values Ϯ SD, and significant differ- p Ͻ 0.01 vs medium alone. The results ,ء .ences are shown presented are representative of three different experiments demonstrating similar results. B, Confluent HSC-2 cells were incubated for 24 h with or without 100 ␮g/ml meso-DAP, DD-DAP, http://www.jimmunol.org/ or LL-DAP. After the incubation, cells were stained with anti- PGRP-I␤ mAb, followed by FITC-conjugated IgG. C, HSC-2 cells were treated with or without test materials for 24 h. After fixation, the cells were treated with anti-PGRP-I␤ mAb or mouse IgG1 and were incubated with Alexa Fluor 488 (green). Nuclei were visualized by staining with DAPI blue. The results presented are representative of four different experiments demonstrating similar results. Scale bar, 20 ␮m. by guest on September 26, 2021

␣ ϭϩ Preparation of the isomers of DAP CHCl3), lit. [ ]D 5.45 (c 0.5, CHCl3). After the HPLC separation of dibenzyl 2-N-(benzyloxycarbonyl) 6-NЈ-t-butoxycarbonyl diaminopime- For separation of the three isomers of DAP (2,6-diaminopimelic acid), late, the protecting groups of the obtained isomers were cleaved by reflux- Ј DL-dibenzyl 2-N-(benzyloxycarbonyl) 6-N -t-butoxycarbonyl diamin- ingfor5hin6NHCl(15␮M), and then concentrated under reduced opimelate, a racemic DAP derivative, was synthesized from DL-DAP (pur- pressure. The residues were separately washed with Et2O to give (2R,6R)- L L D D chased from Sigma-Aldrich). The four isomers ( , -(2S,6S), , -(2R,6R), DAP, meso-(2R,6S)-DAP, and (2S,6S)-DAP (DD-DAP, meso-DAP, and LL- L,D-(2S,6R), and D,L-(2R,6S)) of the derivative were separated with chiral DAP) as white crystal. Electrospray ionization-mass spectrometry (posi- ␾ ϩ HPLC column: 1 cm x 25 cm (CHIRALPAK AD-H; Daicel Chemical tive) m/z ϭ 191.10 [M ϩ H] . The 1H-nuclear magnetic resonance ϭ Industries); mobile phase: n-Hexane/2-propanol 7/3; flow rate: 4 ml/ spectroscopic data and optical rotation data were consistent with those min; detection: UV 254 nm. The retention times and the optical rotation of reported. ␣ 21 ϭϩ the compounds were: D,D-(2R,6R): 19.1 min, [ ] D 2.7 (c 1.0, ␣ 21 ϭϩ CHCl3); (2R,6S): 34.4 min, [ ] D 0.78 (c 1.0, CHCl3); (2S,6R): 47.3 Cells and cell culture ␣ 21 ϭϪ ␣ 21 ϭϪ min, [ ] D 0.80 (c 1.0, CHCl3); L,L-(2S,6S): 52.6 min, [ ] D 3.2 (c 0.6, CHCl3). The configurations of D,D-(2R,6R) and L,L-(2S,6S) forms The human oral epithelial cell lines HSC-2 (15) and HO-1-u-1 (16), the were determined with a known optical rotation after leading to the corre- human pharyngeal epithelial cell line HEp-2, the human esophageal epi- sponding DAP. The (2R,6S) and (2S,6R) forms were derived to dimethyl thelial cell line TE-1, and the human monocytic cell line THP-1 were 2-N-(benzyloxycarbonyl) 6-NЈ-t-butoxycarbonyl diaminopimelate to com- obtained from the Cancer Cell Repository, Institute of Development, Aging pare with the known optical rotations; (2R,6S)-dimethyl 2-N-(benzyloxy- and Cancer, Tohoku University (Sendai, Japan). The human colon adeno- Ј ␣ 21 ϭϩ carbonyl) 6-N -t-butoxycarbonyl diaminopimelate: [ ] D 5.0 (c 0.67, carcinoma cell line SW620 (CCL-227) was purchased from the American The Journal of Immunology 1799

weight of the products and a melting curve analysis at each data point. The primers used for PCR had the following sequences: PGRP-L, 5Ј-ACT GAGGGCTGCTGGGACCA-3Ј,5Ј-GGCCTCAGTGAATTCCTTGG-3Ј; PGRP-I␣,5Ј-GTTCCGTGTGTCCATGTGAA-3Ј,5Ј-TTGGGAAGCCA GAGAGACAC-3Ј; PGRP-I␤,5Ј-ATGTCTCCACCACGGTCTCT-3Ј,5Ј- CACCCACTGTTGTTGTGGAC-3Ј; PGRP-S, 5-CCGTGGCTGGAACT TCACG-3Ј,5Ј-GCACATCCCGGTGTCCTTTG-3Ј; and human GAPDH, 5Ј-CATCACCATCTTCCAGGAGC-3Ј,5Ј-CATGAGTCCTTCCACGAT ACC-3Ј. Cycling conditions were as follows: with PGRP-L, -I␣,-I␤, and -S, 40 cycles at 95°C for 15 s, 56°C for 15 s, and 72°C for 11 s; with GAPDH, 40 cycles at 95°C for 15 s, 56°C for 10 s and 72°C for 11 s. The results are expressed as relative mRNA accumulation corrected with reference to GAPDH mRNA as an internal standard. Measurement of cytokines The cells (104 cells per 200 ␮l) were incubated with or without stimulant in RPMI 1640 medium with 1% FCS for 24 h in 96-well flat-bottom plates (Falcon; BD Discovery Labware). In some experiments, cytochalasin D (1 ␮M) was added directly to the cell culture medium 30 min before stimulation to allow better stimulation by NOD ligands (17). After cultivation, the culture supernatants were collected and the levels of IL-8, MCP-1, IL-6 and TNF-␣ were determined with ELISA kits (BD Pharmingen). The con- Downloaded from centration of cytokines in the supernatants was determined using the SOFTMAX data analysis program (Molecular Probes). RNA interference Transfections for targeting endogenous TLR2, NOD1, and NOD2 were conducted using Lipofectamine 2000 (Invitrogen Life Technologies) and short interfering (si) RNA (final concentration, 200 nM) for 24 h at 37°C, http://www.jimmunol.org/ according to the manufacturer’s instructions. The cellular viability of the cells after transfection was Ͼ95%, as assessed by a 0.2% trypan blue ex- clusion test, and morphological character was not changed after transfection. The sequences of target mRNAs used in this study are: TLR2, 5Ј- AATCCGGAGGCTGCATATCC-3Ј; NOD1, 5Ј-AAGAGCCTCTTTGTC TTCACC-3Ј; and NOD2, 5Ј-AAGACATCTTCCAGTTACTCC-3Ј, and siRNA for TLR2, NOD1, and NOD2 were synthesized and purified by Silencer siRNA Construction kit (Ambion). Flow cytometry Flow cytometric analyses were performed using a FACSCalibur cytometer by guest on September 26, 2021 (BD Biosciences). The cells were stimulated with or without test materials for 24 h at 37°C. After incubation, cells were collected by nonenzymatic cell dissociation solution and washed in PBS. Cells were stained with anti- PGRP-I␤ mAb (mouse IgG1) (Imgenex), followed by FITC-conjugated goat anti-mouse IgG (BioSource International). Immunostaining FIGURE 3. Specific suppression of PGRP mRNA expression induced by meso-DAP using siRNA targeting NOD1. A, Oral epithelial HSC-2 cells Epithelial cells were cultured on eight-chamber glass slides (Falcon) until were transfected with buffer only or siRNA of TLR2, NOD1, or NOD2. confluent and treated with test materials for 24 h at 37°C in a 5% CO2 Lamin A/C was used as a negative control. After 24 h, total RNA was incubator and washed three times with PBS. After fixation with 4% para- extracted and RT-PCR was performed using primers for the respective formaldehyde for 15 min at room temperature, cells were treated with mouse anti-PGRP-I␤ mAb (1/100) for3hatroom temperature without genes. B, HSC-2 cells were transfected with siRNA of TLR2, NOD1, or ␤ ␮ permeabilizing. For -defensin 2 staining, intracellular staining was per- NOD2. After a 24-h incubation, cells were stimulated with 100 g/ml formed. Briefly, after fixation, the cells were treated with 0.5% Triton meso-DAP, D-Gln, or MDP or 100 ng/ml lipid A. After 8 h, the mRNA X-100 for 15 min and washed with PBS. Cells were then treated with goat p Ͻ 0.01 and #, anti-human ␤-defensin 2 (1/100) Ab for3hatroom temperature. As a ,ء .expressions of PGRPs were analyzed by real-time PCR p Ͻ 0.01 vs the respective control. Error bars indicate SD. Results are negative control, mouse IgG1 or goat Ig (DakoCytomation) was used. representative of three different experiments. Samples were then washed and incubated with Alexa Fluor 488 goat anti- mouse IgG1 (1/500) or Alexa Fluor 488 rabbit anti-goat IgG (1/500), respectively. Nuclei were visualized by staining with 4Ј,6Ј-diamino-2-phe- nylindole (DAPI) in blue (Molecular Probes). Samples were photographed Type Culture Collection. Human epitheloid carcinoma of cervix HeLa with an AxioCamMR monochrome digital camera mounted on the Zeiss (JCRB9904), HO-1-u-1, HEp-2, TE-1, SW620, and HeLa cells were grown AxioImager Z1 microscope using the application Zeiss AxioVision 4 soft- in RPMI 1640 medium (Nissui Seiyaku) with 10% heat-inactivated FCS ware (Zeiss). (Invitrogen Life Technologies). The THP-1 cells were maintained in a 5 logarithmic phase of growth by passage every 3–4 days. Cells (2 ϫ 10 /ml) Statistical method were then treated with 0.1 ␮M 22-OCT, an analog of 1␣,25-dihydroxy- vitamin D3, supplied by Chugai Pharmaceutical for 3 days. OCT treatment Statistical significances were determined using ANOVA with the Bonfer- induced the differentiation of THP-1 cells into macrophage-like cells. roni or Dunnett method. RNA isolation and real-time PCR Results Total cellular RNA was prepared from human oral epithelial cells with Up-regulation of the expressions of PGRPs upon stimulation Isogen (Nippon Gene) according to the manufacturer’s instructions. Ran- with meso-DAP in oral epithelial cells dom hexamer-primed reverse transcription was performed using 2.5 ␮lof total RNA in a 50-␮l reaction volume. Real-time PCR was performed with To examine the bacterial moiety detected by NOD1, we first used a LightCycler. The specificity of the PCR was confirmed by the molecular human oral epithelial HSC-2 cells and carefully examined the ability 1800 DAP ACTIVATES EPITHELIAL CELLS VIA NOD1

FIGURE 4. Induction of proinﬂammatory cytokines Downloaded from in epithelial cells triggered by meso-DAP. SW620 cells (A–E) and IFN-␥-treated (1000 IU/ml) HSC-2 cells (F–I) were stimulated with or without indicated dose of test materials for 24 h. J, SW620 cells were preincubated with or without cytochalasin D (1 ␮M) for 30 min before stimulation to facilitate stimulation by NOD ligands. Then, cells were stimulated with or without the http://www.jimmunol.org/ indicated dose of test materials for 24 h. At the concentration used, cytochalasin D treatment for 24 h did not affect cell viability, as assessed by measurement of lactate dehydrogenase release in the culture supernatant (data not shown). IL-8, MCP-1, IL-6, and TNF-␣ levels in the culture supernatants were determined by ELISA. Data are expressed as mean values Ϯ SD, and signiﬁ- p Ͻ 0.01 vs medium ,ء .cant differences are shown alone. The results presented are representative of three by guest on September 26, 2021 different experiments demonstrating similar results.

of PGRP induction. A commercial DAP preparation for itself up- as Clostridium perfringens and Propionibacterium acnes, whereas regulated the mRNA expressions of four PGRPs on human oral DD-DAP has not been isolated from any bacteria so far. Therefore, epithelial cells in culture in dose-dependent manners, whereas L- we chemically synthesized three isoforms of DAP (Fig. 1C). In Lys and D-Gln, both of which are also key amino acids in bacterial flow cytometry, meso-DAP induced PGRP-I␤ protein on human PGN, were inactive (Fig. 2A). The commercial DAP is a mixture oral epithelial cells, whereas DD-DAP and LL-DAP were inactive in of stereoisomers; DD-, LL-, and meso-DAP, among which meso- this respect (Fig. 2B). A similar finding was also obtained by im- DAP is a PGN component of most Gram-negative and some munostaining (Fig. 2C); meso-DAP induced PGRP-I␤ expression Gram-positive bacteria such as Mycobacterium, Corynebacterium, and LL-DAP slightly induced PGRP-I␤ expression, whereas DD- and Bacillus, and LL-DAP was isolated from specific bacteria such DAP did not induce PGRP-I␤ expression. The Journal of Immunology 1801

FIGURE 5. Induction of ␤-defensin 2 in human epithelial cells triggered by meso-DAP and meso-lanthionine. A, Human oral epithelial HSC-2 cells were incubated for 24 h in the presence or absence of 100 ␮g/ml meso-DAP, DD-DAP, LL-DAP, iE-DAP, meso- lanthionine, or MDP. After ﬁxation, cells were treated with anti-␤-defensin 2 Ab or goat IgG then visualized with Alexa Fluor 488 (green). Nuclei were visualized by staining with DAPI blue. The results presented are representative of four different experiments demonstrating similar results. B, Oral epithelial HO-1-u-1, pharyngeal epithelial HEp-2, esophageal epithelial TE-1, Downloaded from colonic epithelial SW620, and cervical epithelial HeLa cells were incubated for 24 h in the presence or absence of 100 ␮g/ml meso-DAP, DD-DAP, or LL-DAP. After ﬁxation, cells were treated with anti-␤-defensin 2 Ab or goat IgG and then visualized with Alexa Fluor 488 (green). Nuclei were visualized by staining with DAPI blue. The results presented are representative of three http://www.jimmunol.org/ different experiments demonstrating similar results. Scale bar, 20 ␮m. by guest on September 26, 2021

Specific suppression of PGRP mRNA expression induced by MCP-1, IL-6, and TNF-␣ upon stimulation with meso-DAP in meso-DAP using siRNA targeting NOD1 dose-dependent manners, although its activities were in general To clarify the signaling pathway of cellular activation by meso- less than reference NOD agonists (Fig. 4, A–E). In the assay, LL- DAP, we used RNA interference assays targeting NOD1, NOD2, DAP exhibited weak activity and DD-DAP were inactive. and TLR2 mRNA. The region of the TLR2, NOD1, and NOD2 Naive oral epithelial cells did not secrete proinflammatory mRNA targeted by the siRNA is shown in Materials and Methods. cytokines in response to bacterial cell surface components (11), ␥ We confirmed that TLR2, NOD1, and NOD2 mRNA levels were whereas IFN- -primed oral epithelial cells produced cytokines suppressed by ϳ80% using specific siRNA in oral epithelial upon stimulation with bacterial components (18). Therefore, we HSC-2 cells from 24 to 72 h of culture (Fig. 3A) (12). The up- examined whether synthetic DAP preparations activated IFN- regulated expressions of the mRNAs for PGRPs induced by meso- ␥-treated oral epithelial cells to secrete IL-8. meso-DAP in- DAP were inhibited in NOD1-silenced oral epithelial cells, but creased MCP-1, IL-6, and TNF-␣ as well as IL-8 production in neither NOD2- nor TLR2-silenced oral epithelial cells (Fig. 3). IFN-␥-treated oral epithelial HSC-2 cells (Fig. 4, F–H). LL-DAP However, siRNA for NOD1 did not inhibit responses induced by exhibited weak activity and DD-DAP was also inactive in the MDP (NOD2 agonist) nor lipid A (TLR4 agonist) in the same cells assay (Fig. 4A). However, compared with the ability of iE-DAP, (Fig. 3). These results indicated that regulation of PGRP mRNA by FK156, or FK565, the ability of meso-DAP was weakest, meso-DAP occurs through NOD1. whereas iE-DAP showed the strongest activity (Fig. 4I). In our study, very high concentrations of NOD ligands are required to Secretion of proinflammatory cytokines upon stimulation with induce cytokines. Recently, Magalhaes (17) added cytochalasin ␥ meso-DAP in intestinal epithelial cells and IFN- -primed oral D(1␮M) to the cell culture medium to allow better stimulation epithelial cells by NOD ligands. It has been well established that cytochalasin To further clarify cellular activation triggered by meso-DAP, we D inhibits actin polymerization and allow components to inter- next examined the effect on production of proinflammatory cyto- nalize into the cells (19). Therefore, we also examined IL-8- kines in human epithelial cells. As described, some human intes- inducing abilities of NOD1 ligands against colonic epithelial tinal epithelial cells such as SW620 produce inflammatory cyto- SW620 cells in the presence of cytochalasin D. The result kines upon stimulation with bacterial components (11). Human showed that stimulation with NOD1 ligands in the presence of intestinal epithelial SW620 cells exhibited the productions of IL-8, cytochalasin D (1 ␮M) exhibited a several hundred-fold higher 1802 DAP ACTIVATES EPITHELIAL CELLS VIA NOD1

FIGURE 6. Meso-DAP might not provoke cellular activation of human monocytic cells even in the presence of lipid A. OCT-treated Downloaded from THP-1 cells were stimulated with DAP isomers and reference NOD agonists (A–D)or those together with lipid A (F–H) for 24 h at indicated concentrations. IL-8, MCP-1, IL-6, and TNF-␣ level in the culture supernatants was determined by ELISA and mean values http://www.jimmunol.org/ p Ͻ 0.01 and #, p Ͻ 0.01 vs ,ء .are shown respective control. Signiﬁcant (p Ͻ 0.05) synergistic effects were detected (F–H)by ANOVA including an interaction term. THP-1 cells were stimulated with 100 ␮g/ml iE-DAP or meso-DAP test material (E) for 8 h. After incubation, mRNA expressions of PGRP-L, PGRP-I␣, PGRP-I␤, and PGRP-S were analyzed by real-time PCR. Data are expressed as mean values Ϯ SD, and signiﬁcant differences by guest on September 26, 2021 p Ͻ 0.01 vs medium alone. The ,ء .are shown results presented are representative of three different experiments demonstrating similar results.

ability to induce cytokines than stimulation with NOD ligands DAP and LL-DAP were inactive in this respect (Fig. 5A). It must alone in terms of the amount of compound required to achieve be noted here that meso-lanthionine also activated oral epithe- 50% of maximum activity (Fig. 4J). lial cells to generate ␤-defensin 2 (Figs. 5A). Oral epithelial cells were not stained with goat IgG, followed by Alexa Fluor ␤ Meso-lanthionine as well as meso-DAP induced -defensin 2 in 488 (green) (data not shown). In addition, other oral epithelial human epithelial cells cell line HO-1-u-1, the pharyngeal epithelial cell line HEp-2, In consistency with results discussed, meso-DAP induced ␤-de- the esophageal epithelial cell line TE-1, colon adenocarcinoma fensin 2 generation in oral epithelial HSC-2 cells, whereas DD- cell line SW620 and epitheloid carcinoma of cervix HeLa also The Journal of Immunology 1803

PGNs (Fig. 1). In addition, LL-DAP, which was isolated from spe- ciﬁc bacteria such as Clostridium perfringens and Propionibacte- rium acnes, also had a weak potency on cellular activation, whereas DD-DAP, which has been not isolated from any bacteria so far, was inactive in this respect. Furthermore, we demonstrated that synthetic meso-lanthionine was also recognized by NOD1. Meso- lanthionine is a counterpart to meso-DAP in some Fusobacterium and related species such as Fusobacterium nucleatum (13). In contrast, MDP (MurNAc-L-Ala-D-isoGln) or MurNAc-L-Ala-D-Glu, essential NOD2 sensing structures, are almost ubiquitously present in PGNs from both Gram-positive and Gram-negative bacteria (Fig. 1). Therefore, NOD2 can elicit the broad recognition of bacteria, whereas NOD1 mediates the host response to a speciﬁed group of bacteria. Both PGN motifs are naturally occurring deg- radation products released from bacteria during growth (20) or processed by the host cells in the lysosomal compartment. These results indicate that NOD2 and NOD1 act as cytosolic pattern recognition receptors that recognize highly conserved PGN structures

present in all and large populations of bacteria, respectively. Recent Downloaded from studies have shown that the Drosophila immune system detects Gram-positive and Gram-negative bacteria through the speciﬁc recognition of respective PGN motifs such as the meso-DAP-containing motif (21). Thus, selective host recognition of bacteria based on the PGN structure is a ubiquitous system from insects to human. It should FIGURE 7. Meso-DAP provoked cellular activation of human mono-

be emphasized that NOD1 is a likely special sentinel molecule in the http://www.jimmunol.org/ cytic cells in the presence of cytochalasin D. OCT-treated THP-1 cells were preincubated with cytochalasin D (1 ␮M) for 30 min before stimu- epithelial barrier, allowing intracellular detection of bacteria through lation to allow better stimulation by NODs ligands. Then, cells were stim- recognizing the meso-DAP motif of PGN. In fact, Inohara et al. (22) ulated with DAP isomers and reference NOD agonists for 24 h at the reported that NOD1 was expressed in multiple tissues and was indicated concentrations. IL-8, MCP-1, IL-6, and TNF-␣ level in the cul- thought to have an important role, in particularly epithelium. There- ture supernatants were determined by ELISA and the mean values are fore, NOD1 and NOD2 in cooperation might recognize DAP-type p Ͻ 0.01 vs medium alone. The results presented are represen- bacteria and induce sufficient protective responses against these ,ء .shown tative of three different experiments demonstrating similar results. bacterial invasions. In this study, MDP, iE-DAP, meso-DAP, FK156, and FK565 are used from the 1 to 100 ␮g/ml range. Girardin et al. (6, 9, 10) reported significant activity of NOD by guest on September 26, 2021 produced ␤-defensin 2 upon stimulation with meso-DAP, but agonists at the 50 pM/ml, which was a similar concentration to not DD- and LL-DAP (Fig. 5B). These cells were not stained with that of lipid A as a TLR4 agonist. It must be noted that Girardin goat IgG, followed by Alexa Fluor 488 (green) (data not et al. (6, 9, 10) microinjected test materials into cells, or cells shown). were permeabilized by digitonin, whereas we only added the test materials to cell cultures without treatment to enhance per- Human monocytic cells scarcely respond to meso-DAP meability to examine responses under physiological conditions Then, we examined whether meso-DAP activates human mono- in most of the experiments in this study. Therefore, our system cytic cells. All of three DAP isomers did not induce IL-8 produc- was not efficient for putting test materials into the cytoplasma of tion in human monocytic THP-1 cells, in which synthetic FK156, the cells, and excess amounts of materials may be required for FK565, and iE-DAP as reference NOD1 agonists and MDP as an cell activation. In this context, in the presence of cytochalasin NOD2 agonist showed marked activities to augment production of D(1␮M) to facilitate stimulation by NOD ligands according to IL-8, MCP-1, IL-6, and TNF-␣ (Fig. 6, A–D). In addition, meso- Magalhaes et al. (17), NOD1 ligands, including meso-DAP, ex- DAP did not induce the expression of PGRPs mRNA on THP-1 hibited a several hundredfold higher ability to induce IL-8 on cells (Fig. 6E). We recently demonstrated that MDP and DMPs in human epithelial SW620 cells compared with that after stimu- combination with TLR agonists synergistically induced the pro- lation with NOD1 ligands alone in terms of the amount of com- duction of IL-8 in an NOD2- and NOD1-dependent manner, re- pound required to achieve 50% of maximum activity (Fig. 4J). spectively, in human monocytic THP-1 cells. We further examined In the system, human monocytic THP-1 cells also respond to possible synergistic IL-8 secretion by meso-DAP in combination meso-DAP, although the cells scarcely respond to meso-DAP in with lipid A in THP-1 cells in culture. Meso-DAP did not enhance the absence of cytochalasin D (Fig. 7). These findings suggested cellular activation even in combination with lipid A, whereas iE- that NOD1 is capable of recognizing meso-DAP even in mono- DAP had a potency for synergistic IL-8 production (Fig. 6, F–H). cytic cells, although under the physiological conditions meso- In contrast, in the presence of cytochalasin D meso-DAP (500 DAP could not be transferred to NOD1 in monocytic cells in ng/ml) showed slightly but significantly induced the production of contrast to epithelial cells. IL-6, IL-8, MCP-1, and TNF-␣ in THP-1 cells (Fig. 7). Among human PGRPs, Wang et al. (23) reported that human PGRP-L is MurNAc-L-Ala amidase like PGRP-SC1B in Drosoph- Discussion ila, both of which digest PGN. Murine PGRP-S was suggested to In the present study, we first demonstrated that meso-DAP itself be involved in intracellular bacterial killing (24). In this and pre- activates human cells via NOD1. Meso-DAP is a specific amino vious (12) studies, PGRP-I␣ and PGRP-I␤ were more markedly acid to most Gram-negative and some Gram-positive bacteria, and up-regulated than PGRP-L or PGRP-S, although the functions of is a key moiety to make cross-link between stem peptides of these mammalian PGRP-I␣ and PGRP-I␤ are not yet clear except for 1804 DAP ACTIVATES EPITHELIAL CELLS VIA NOD1 their binding ability to PGN and Gram-positive bacteria. PGRP-I␣ can through muramyl dipeptide (MDP) detection. J. Biol. Chem. 278: and PGRP-I␤ are reported to be highly expressed in the esophagus 8869–8872. 7. Inohara, N., Y. Ogura, A. Fontalba, O. Gutierrez, F. Pons, J. Crespo, K. Fukase, (25). Both the esophagus and oral cavity are covered by squamous S. Inamura, S. Kusumoto, M. Hashimoto, et al. 2003. Host recognition of bac- epithelial cells. If PGRP-I␣ and PGRP-I␤ functioned actively in terial muramyl dipeptide mediated through NOD2: implications for Crohn’s dis- the clearance of invasive bacteria, it is of benefit for hosts that ease. J. Biol. Chem. 278: 5509–5512. 8. Chamaillard, M., M. Hashimoto, Y. Horie, J. Masumoto, S. Qiu, L. Saab, PGRP-I␣ and PGRP-I␤ expressions are highly up-regulated upon Y. Ogura, A. Kawasaki, K. Fukase, S. Kusumoto, et al. 2003. An essential role stimulation with bacterial cell surface components in oral epithe- for NOD1 in host recognition of bacterial peptidoglycan containing diaminopimelic acid. Nat. Immunol. 4: 702–707. lial cells, which are the first cells encountered by bacteria in oral 9. Girardin, S. E., I. G. Boneca, L. A. M. Carneiro, A. Antignac, M. Je´hanno, mucosa. In contrast, naive oral epithelial cells did not secrete in- J. Viala, K. Tedin, M.-K. Taha, A. Labigne, U. Za¨hringer, et al. 2003. Nod1 flammatory cytokines upon stimulation with bacterial components detects a unique muropeptide from gram-negative bacterial peptidoglycan. Sci- ␥ ence 300: 1584–1587. including NOD1 and NOD2 ligands, although IFN- -primed oral 10. Girardin, S. E., L. H. Travassos, M. Herve´, D. Blanot, I. G. Boneca, D. J. Philpott, epithelial cells produced inflammatory cytokines in the same con- P. J. Sansonetti, and D. Mengin-Lecreulx. 2003. Peptidoglycan molecular re- dition. It must be noted that some human intestinal epithelial cell quirements allowing detection by Nod1 and Nod2. J. Biol. Chem. 278: 41702–41708. lines such as SW620 and HT29 secreted IL-8 in response to LPS, 11. Uehara, A., S. Sugawara, R. Tamai, and H. Takada. 2001. Contrasting responses whereas other human intestinal epithelial cells lines such as T84 of human gingival and colonic epithelial cells to lipopolysaccharides, lipoteichoic and Caco2 cells were completely unresponsive to LPS (26, 27). acids and peptidoglycans in the presence of soluble CD14. Med. Microbiol. Im- munol. (Berl). 189: 185–192. Furthermore, the responsiveness of human intestinal epithelial 12. Uehara, A., Y. Sugawara, S. Kurata, Y. Fujimoto, K. Fukase, S. Kusumoto, SW480 cells was about one-tenth weaker than that of human in- Y. Satta, T. Sasano, S. Sugawara, and H. Takada. 2005. Chemically synthesized testinal epithelial SW620 cells, both of which were isolated from pathogen-associated molecular patterns increase the expression of peptidoglycan Downloaded from recognition proteins via Toll-like receptors, NOD1 and NOD2 in human oral the same person (11). Although proinflammatory cytokine re- epithelial cells. Cell. Microbiol. 7: 675–686. sponses of human intestinal epithelial cells to bacterial stimuli 13. Kato, K., T. Umemoto, H. Sagawa, and S. Kotani. 1979. Lantionine as an essential constitutent of cell wall peptidoglycan of Fusobacterium nucleatum. Curr. have not been examined so far, intestinal epithelial cells produce Microbiol. 3: 147–151. anti-bacterial factors upon stimulation with bacterial components 14. Photaki, I., I. Samouilidis, S. Caranikas, and L. Zervas. 1979. Lanthionine chem- including NOD ligands (28). There is a good possibility that istry. Paert 4. Synthyesis of diasteroisomeric cyclolanthionyl derivatives. J. Chem. Soc. Perkin Trans. 1: 2599–2605. naive epithelial cells in general produce antibacterial factors, http://www.jimmunol.org/ 15. Momose, F., T. Araida, A. Negishi, H. Ichijo, S. Shioda, and S. Sasaki. 1989. but not proinflammatory factors, to prevent bacterial invasion Variant sublines with different metastatic potentials selected in nude mice from without excessive inflammatory responses, which may lead to human oral squamous cell carcinomas. J. Oral. Pathol. Med. 18: 391–395. 16. Miyauchi, S., T. Moroyama, T. Sakamoto, T. Okamura, and K. Takada. 1985. tissue destruction. Establishment of human tumor cell line (Ueda-1) derived from squamous cell It is not clear whether the meso-DAP itself actually exhibits carcinoma of the floor of the mouth. Jpn. J. Oral Maxillofac. Surg. 31: bioactivities in vivo. It is interesting that NOD1-deficient mice 1347–1351. 17. Magalhaes, J. G., D. J. Philpott, M.-A. Nahori, M. Je´hanno, J. Fritz, were more susceptible to infection with Helicobacter pylori than L. L. Bourhis, J. Viala, J.-P. Hugot, M. Giovannini, J. Bertin, et al. 2005. Murine wild-type mice (29). In this context, mucosal tissues of NOD- Nod1 but not its human orthologue mediates innate immune detection of tracheal deficient mice are generally normal, but they are susceptible to cytotoxin. EMBO Rep. 6: 1201–1207.

18. Uehara, A., S. Sugawara, and H. Takada. 2002. Priming of human oral epithelial by guest on September 26, 2021 bacterial infection orally (28). Further studies using NOD1-defi- cells by interferon-␥ to secrete cytokines in response to lipopolysaccharides, li- cient mice are needed to understand the biological function of NOD1. poteichoic acids and peptidoglycans. J. Med. Microbiol. 51: 626–634. In addition, bacteria carrying DAP-containing PGN also possess mol- 19. Gruenheid, S., and B. B. Finlay. 2003. Microbial pathogenesis and cytoskeletal function. Nature 422: 775–781. ecules capable of stimulating multiple pattern recognition proteins. 20. Boneca, I. G. 2005. The role of peptidoglycan in pathogenesis. Curr. Opin. Mi- Thus, analysis of NOD1 function in the absence and presence of other crobiol. 8: 46–53. pattern recognition receptors may be required to unravel the practical 21. Takehana, A, T. Katsuyama, T. Yano, Y. Oshima, H. Takada, T. Aigaki, and S. Kurata. 2002. Overexpression of a pattern-recognition receptor, peptidoglycan- roles of NOD1 in host defense against bacteria. recognition protein-LE, activates imd/relish-mediated antibacterial defense and the prophenoloxidase cascade in Drosophila larvae. Proc. Natl. Acad. Sci. USA Acknowledgments 99: 13705–13710. 22. Inohara, N., T. Koseki, L. del Peso, Y. Hu, C. Yee, S. Chen, R. Carrio, J. Merino, We thank Dr. N. Inohara for useful advice and discussion and the late D. Liu, J. Ni, and G. Nuń˜ez. 1999. Nod1, an Apaf-1-like activator of caspase-9 Dr. S. Kotani for encouraging us to carry out this study. We thank D. and nuclear factor-␬B. J. Biol. Chem. 274: 14560–14567. Mrozek (Medical English Service, Kyoto, Japan) for reviewing this article. 23. Wang, Z.-M., X. Li, R. R. Cocklin, M. Wang, M. Wang, K. Fukase, S. Inamura, S. Kusumoto, D. Gupta, and R. Dziarski. 2003. Human peptidoglycan recognition Disclosures protein-L is an N-acetylmuramoyl-L-alanine amidase. J. Biol. Chem. 278: 49044–49052. The authors have no financial conflict of interest. 24. Liu, C., E. Gelius, G. Liu, H. Steiner, and R. Dziarski. 2000. Mammalian peptidoglycan recognition protein binds peptidoglycan with high affinity, is ex- References pressed in neutrophils, and inhibits bacterial growth. J. Biol. Chem. 275: 24490– 1. Takada, H., and S. Kotani. 1985. Immunopharmacological activities of synthetic 24499. muramylpeptides. In Immunology of the Bacterial Cell Envelope. 25. Liu, C., Z. Xu, D. Gupta, and R. Dziarski. 2001. Peptidoglycan recognition pro- D. E. S. Stewart-Tull and M. Davies, eds. Wiley, Chichester, U.K., pp. 119–152. teins: a novel family of four human innate immunity pattern recognition mole- 2. Takada, H., and S. Kotani. 1995. Muramyl dipeptide and derivatives. In The cules. J. Biol. Chem. 276: 34686–34694. Theory and Practical Application of Adjuvants. D. E. S. Stewart-Tull ed. John 26. Eckmann, L., M. F. Kagnoff, and J. Fierer. 1993. Epithelial cells secrete the Wiley & Sons, Chichester, U.K., pp. 171–202. chemokine interleukin-8 in response to bacterial entry. Infect. Immun. 61: 3. Adam, A. 1985. Muramylpeptides and analogues. Synthetic Adjuvants. Wiley, 4569–4574. Chichester, U.K., pp. 1–58. 27. Schuerer-Maly, C. C., L. Eckmann, M. F. Kagnoff, M. T. Falco, and F. E. Maly. 4. Goto, T., and H. Aoki. 1987. The immunomodulatory activities of acylpeptides. 1994. Colonic epithelial cell lines as a source of interleukin-8: stimulation by In Immunostimulants: Now and Tomorrow. I. Azuma and G. Jolle`s, eds. Spring- inflammatory cytokines and bacterial lipopolysaccharide. Immunology 81: 85–91. er-Verlag, Berlin, pp. 99–108. 28. Kobayashi, K. S., M. Chamaillard, Y. Ogura, O. Henegariu, N. Inohara, 5. Kitaura, Y., O. Nakaguchi, H. Takeno, S. Okada, S. Yonishi, K. Hemmi, J. Mori, G. Nuń˜ez, and R. A. Flavell. 2005. Nod2-dependent regulation of innate and ␥ Ј H. Senoh, Y. Mine, and M. Hashimoto. 1982. N2-( -D-glutamyl)-meso-2L,2 D- adaptive immunity in the intestinal tract. Science 307: 731–734. diaminopimelic acid as the minimal prerequisite structure of FK-156: its acyl 29. Viala, J., C. Chaput, I. G. Boneca, A. Cardona, S. E. Girardin, A. P. Moran, derivatives with potent immunostimulating activity. J. Med. Chem. 25: 335–337. R. Athman, S. Me´met, M. R. Huerre, A. J. Coyle, et al. 2004. Nod1 responds to 6. Girardin, S. E., I. G. Boneca, J. Viala, M. Chamaillard, A. Labigne, G. Thomas, peptidoglycan delivered by the Helicobacter pylori cag pathogenicity island. Nat. D. J. Philpott, and P. J. Sansonetti. 2003. Nod2 is a general sensor of peptidogly- Immunol. 5: 1166–1174.