Porcine Toll-like receptors: Recognition of Salmonella enterica serovar Choleraesuis and influence of polymorphisms

Molecular Immunology 48 (2011) 1114–1120

Contents lists available at ScienceDirect

Molecular Immunology journal homepage: www.elsevier.com/locate/molimm

Porcine Toll-like receptors: Recognition of Salmonella enterica serovar Choleraesuis and influence of polymorphisms Hiroki Shinkai a,b , Rintaro Suzuki c , Masato Akiba d , Naohiko Okumura b,e , Hirohide Uenishi a,b,∗ a

Division of Animal Sciences, National Institute of Agrobiological Sciences (NIAS), 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan Animal Genome Research Program, NIAS/STAFF, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan Division of Plant Sciences, National Institute of Agrobiological Sciences, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan d Safety Research Team, National Institute of Animal Health, 3-1-5 Kannondai, Tsukuba, Ibaraki 305-0856, Japan e Second Research Division, Institute of Society for Techno-innovation of Agriculture, Forestry and Fisheries (STAFF-Institute), 446-1 Ippaizuka, Kamiyokoba, Tsukuba, Ibaraki 305-0854, Japan b c

a r t i c l e

i n f o

Article history: Received 27 December 2010 Received in revised form 10 February 2011 Accepted 12 February 2011 Available online 8 March 2011 Keywords: Salmonella Choleraesuis Toll-like receptor Single nucleotide polymorphism Molecular modeling Pig

a b s t r a c t Salmonella enterica serovar Choleraesuis (SC) is a highly invasive pathogen that causes enteric and septicemic diseases in pigs. Although there have been some reports on gene expression profiles in the course of infection with SC in pigs, little is known about the genes involved in the infection. By measuring activation, as represented by nuclear factor-␬B activity, after stimulation by the pathogen, we showed the involvement of Toll-like receptor (TLR) 5 and the TLR2–TLR1 heterodimer in the recognition of SC. We previously found single nucleotide polymorphisms (SNPs) in the TLRs of various pig populations. Here we demonstrated that the polymorphisms resulting in amino acid changes TLR5R148L , TLR5P402L , and TLR2V703M attenuated the responses to SC by the cells transfected with the TLR genes. Each of these three SNPs was differently restricted in distribution among breeds. These results suggest that there are differences in resistance to salmonellosis among breeds; these differences may be of great importance for the pig industry in terms of breeding and vaccine development. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction Toll-like receptors (TLRs) play crucial roles in innate immunity by recognizing conserved microbial components, and they have profound effects on adaptive immunity. TLRs are type I transmembrane glycoproteins characterized by extracellular domains containing variable numbers of leucine-rich-repeat (LRR) motifs, a single transmembrane domain, and a cytoplasmic signaling region composed mainly of a Toll/interleukin-1 receptor (TIR) domain (Akira and Takeda, 2004; Matsushima et al., 2007). TLRs are divided into two groups: those that recognize molecules such as proteins and lipids from bacteria and are expressed on the surfaces of immune cells (TLR1, TLR2, TLR4, TLR5, and TLR6); and those that sense nucleic acids from pathogens and are localized in intracellular compartments (TLR3, TLR7, TLR8, and TLR9) (Akira and Takeda,

Abbreviations: ELAM-1, endothelial leukocyte adhesion molecule-1; LRR, leucine-rich repeat; PRR, pattern recognition receptor; SC, Salmonella enterica serovar Choleraesuis; ST, Salmonella enterica serovar Typhimurium; TIR, Toll/interleukin-1 receptor; TLR, Toll-like receptor. ∗ Corresponding author at: Division of Animal Sciences, National Institute of Agrobiological Sciences (NIAS), 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8602, Japan. Tel.: +81 29 838 8664; fax: +81 29 838 8674. E-mail address: [email protected] (H. Uenishi). 0161-5890/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.molimm.2011.02.004

2004). Polymorphisms in TLRs are associated with resistance and susceptibility to various diseases (Lazarus et al., 2002; Schröder and Schumann, 2005). In human TLR4, the cosegregating polymorphisms D299G and T399I have been found at a higher frequency among people hyporesponsive to inhaled lipopolysaccharide (LPS) than in a control population (Arbour et al., 2000). D299G polymorphism is also associated with increased risk of Crohn’s disease, ulcerative colitis, or severe sepsis following burn injury (Barber et al., 2004; Franchimont et al., 2004). The R677W and R753Q polymorphisms in human TLR2 are associated with diseases resulting from infection with mycobacteria; R677W is associated with lepromatous leprosy caused by Mycobacterium leprae and R753Q is associated with tuberculosis caused by Mycobacterium tuberculosis (Bochud et al., 2003; Kang and Chae, 2001; Ogus et al., 2004). In human TLR5, the presence of a stop codon (R392stop) abolishes the ability to recognize flagellin and is associated with pneumonia caused by Legionella pneumophila (Hawn et al., 2003). In our previous studies of single nucleotide polymorphisms (SNPs) in porcine TLR genes, we found that most of the nonsynonymous SNPs in the coding sequences of genes encoding TLRs expressed on the cell surface were present in the extracellular region involved in pathogen recognition, rather than the intracellular region (Morozumi and Uenishi, 2009; Shinkai et al., 2006;

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

Uenishi and Shinkai, 2009). This discovery prompted us to examine the influences of the resulting amino acid alterations on recognition of the pathogens that cause serious damage in the pig industry. Salmonella spp. are ubiquitously present in nature and have been recovered from nearly all vertebrates (Edwards et al., 2002). Among more than 2500 different serovars, Salmonella enterica serovar Choleraesuis (SC) has a narrow host range and infects predominantly pigs and occasionally humans, whereas other serovars, such as Salmonella enterica serovar Typhimurium (ST) have broad host ranges (Chiu et al., 2004). SC can cause enterocolitis, pneumonia, septicemia, and hepatitis in pigs, and it is more serious than ST infection, which usually causes only enterocolitis. Human systemic infections caused by SC are considered to be acquired from pigs (Chiu et al., 2006). Here, we examined porcine TLRs recognizing SC and demonstrated the involvement of TLR5 and the TLR2–TLR1 heterodimer. Furthermore, we identified some SNPs in TLR5 and TLR2 genes that affected induction of the response of the cells tranfected with the TLR genes after stimulation by the ligands and showed that each of these important SNPs existed in different porcine breeds. These results suggest the possibility of genetic improvement of disease resistance to salmonellosis in pigs. 2. Materials and methods 2.1. Cells and reagents Human embryonic kidney (HEK) 293 cells (American Type Culture Collection [ATCC] CRL-1573) were maintained at 37 ◦ C in Dulbecco’s modified Eagle medium (Gibco/Invitrogen, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (Gibco/Invitrogen) in a 5% CO2 incubator. Synthetic bacterial lipoprotein Pam3 CSK4 and flagellin purified from ST were purchased from InvivoGen (San Diego, CA, USA). SC (ATCC 7001) cultured in LB medium to a concentration of 1 × 109 colony forming units (cfu)/ml were killed in 70% ethanol and suspended in phosphate-buffered saline. 2.2. Expression vectors The coding regions of the porcine TLR1, TLR2, TLR5, and TLR6 genes were amplified by PCR of the genomic DNA from Berkshire (TLR1 and TLR5) and Large White (TLR2 and TLR6) breed pigs with primers comprised of the most common nucleotide sequences in the European pigs (TLR1, GenBank ID: AB208695; TLR2, GenBank ID: AB208696; TLR5, GenBank ID: AB208697; TLR6, GenBank ID: AB208698) (Shinkai et al., 2006). PCR amplification was conducted by using PfuTurbo DNA Polymerase (Stratagene, La Jolla, CA, USA). Reference expression vectors were constructed by inserting the coding regions into pEF6/V5-His TOPO (Invitrogen, Carlsbad, CA, USA), as previously described (Bochud et al., 2003). Expression vectors encoding mutant proteins were produced by introducing each of the 20 SNPs in TLR1, 10 SNPs in TLR2, and 12 SNPs in TLR5 into the above reference vectors by using a QuikChange Site-Directed Mutagenesis Kit (Stratagene) in accordance with the manufacturer’s instructions. A nuclear factor␬B (NF-␬B)-dependent endothelial leukocyte adhesion molecule-1 (ELAM-1) firefly luciferase construct was generated by cloning five NF-␬B binding sites (GGGACTTTCC × 5) (Schindler and Baichwal, 1994), followed by a fragment (−130 to +26 bp) of the human ELAM-1 promoter, into the luciferase reporter vector GL4.10 (Promega, Madison, WI, USA). Thymidine kinase Renilla luciferase vector (pGL4.74) for use as a control was purchased from Promega.

1115

2.3. Luciferase reporter assay HEK293 cells were transiently cotransfected, using FuGENE 6 Transfection Reagent (Roche Diagnostics, Mannheim, Germany), at a density of 6 × 104 cells per well in a 96-well plate format, with 100 ng of one of the TLR expression vectors, 100 ng NF-␬Bdependent ELAM-1 firefly luciferase construct, and 20 ng thymidine kinase Renilla luciferase vector. In the case of the cotransfection of TLR2 and its heterodimer partner, TLR1 or TLR6, 50 ng of each of the vectors was used for transfection. After 24 h, the cells were stimulated by the ligands for 5 h. Luciferase activity was measured by using a Dual-Glo Luciferase Assay System (Promega). After the firefly luciferase units had been divided by the Renilla units, the values of stimulated cells were divided by those of unstimulated cells to show the degree of induction of NF-␬B. All experiments were performed in triplicate wells three times to confirm reproducibility. 2.4. RT-PCR Total RNA from HEK293 cells was extracted by using Isogen (Nippon Gene, Tokyo, Japan), and then treated with DNase I to eliminate DNA contamination. Human spleen total RNA was purchased from Ambion (Austin, TX, USA). cDNA was synthesized by using PowerScript Reverse Transcriptase (Clontech, Palo Alto, CA, USA). PCR was performed by using Advantage 2 Polymerase Mix (Clontech) and primers for amplification of human TLR genes (Kadowaki et al., 2001) and the human glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH), as follows: forward, 5 -GGAAGGTGAAGGTCGGAGTCAACGG-3 ; and reverse, 5 -CCTGGAAGATGGTGATGGGATTTCC-3 . The PCR was conducted with 30 cycles consisting of 95 ◦ C for 30 s and 68 ◦ C for 1 min after incubation at 95 ◦ C for 1 min. 2.5. Genotyping Pig genomic DNA was extracted by using blood or tissue samples. We collected samples from 144 unrelated (i.e., at least two generations of kinship apart from each other) European commercial pigs (36 each of the Berkshire, Duroc, Landrace, and Large White breeds). We also used 24 Chinese pigs (12 each of the Jinhua and Meishan breeds) and 12 each of Clawn miniature pigs and Japanese wild boars. Genomic DNA from these samples was purified by a standard protocol based on phenol–chloroform extraction (Sambrook and Russell, 2001). SNPs were genotyped by sequencing the PCR products, including the corresponding nucleotides, as previously described (Shinkai et al., 2006). 2.6. Sequence alignment Comparison of amino acid sequences of TLR5 of vertebrates was conducted by using ClustalX2 (Larkin et al., 2007). Amino acid alignment for prediction of 3D models of porcine TLR5 was done manually so that the typically conserved residues and topologically equivalent residues in LRRs were aligned. 2.7. Prediction of 3D models of porcine TLR5 The 3D structure of the ectodomain of porcine TLR5 was conˇ and Blundell, 1993). structed by using MODELLER Release 9v8 (Sali The crystal structures of a human apo-TLR3-ectodomain (PDB ID: 1ZIW, 2A0Z), a mouse apo-TLR3-ectodomain (PDB ID: 3CIG), two mouse TLR3-ectodomains complexed with double-stranded RNA (PDB ID: 3CIY), a mouse TLR4-ectodomain complexed with MD-2 (PDB ID: 2Z64), and a human (TLR4–MD-2–LPS)2 duplexheterodimer (PDB ID: 3FXI) were used as templates (Bell et al., 2005; Choe et al., 2005; Kim et al., 2007; Liu et al., 2008; Park

1116

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

(Fig. 1A). These results indicate that TLR5 and the TLR2–TLR1 heterodimer are involved in the recognition of SC. HEK293 expressed endogenous TLR1 (Fig. 1B), which might be involved in the recognition of SC by forming a heterodimer with the transfected pig TLR2, although we cannot exclude the possibility that pig TLR2 alone can recognize SC. 3.2. Influence of amino acid polymorphisms in porcine TLRs on the recognition of SC

Fig. 1. TLRs involved in the recognition of Salmonella enterica serovar Choleraesuis (SC). (A) HEK293 cells were transiently cotransfected with porcine TLR1, TLR2, TLR5, or TLR6 expression vector, or combinations of these vectors, or empty vector, together with an NF-␬B-dependent ELAM-1 firefly luciferase reporter and a thymidine kinase Renilla luciferase control. Cells were stimulated with increasing concentrations of killed SC for 5 h, and luciferase activity was then measured. The results are means ± SD of triplicate wells and are representative of three independent experiments. Significant differences (P < 0.01, two-tailed Welch’s t-test) in responses of cells transfected with TLR1 and TLR2 from that with TLR2 alone were observed at every concentration thus conducted. (B) Endogenous expression of TLR mRNA in HEK293 cells. Availability of PCR primers was confirmed by using total RNA from human spleen. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

et al., 2009). The structures were then visualized by using a combination of the PyMOL Molecular Graphics System (Version 1.3; Schrödinger, Manheim, Germany) and POV-Ray (Version 3.6; Persistence of Vision Pty. Ltd., Victoria, Australia). 3. Results 3.1. Porcine TLRs involved in recognition of SC SC is a facultative anaerobic, gram-negative and peritrichously flagellated rod; we expected that it would be sensed by TLRs expressed on the surfaces of immune system cells. To identify porcine TLRs involved in the recognition of SC, we conducted a luciferase reporter assay by using HEK293 cells transiently transfected with TLR genes either alone or in combination. Single transfection with TLR2 or TLR5 increased NF-␬B induction in a dose-dependent manner after stimulation with SC (Fig. 1A). TLR2 recognizes bacterial lipoproteins that have tri- or di-acylated Nterminal cysteine residues, respectively, by dimerizing with TLR1 or TLR6 (Takeuchi et al., 2001, 2002). Cotransfection of TLR1 with TLR2 achieved nearly twice the response to SC as with TLR2 alone (P < 0.01 by Welch’s t-test), but addition of TLR6 to TLR2 did not increase the degree of NF-␬B induction beyond that achieved with TLR2 alone

We previously reported 20, 10, and 12 nonsynonymous SNPs in porcine TLR1, TLR2 and TLR5, respectively, and we found that most of them were present in the extracellular region involved in the recognition of pathogens (Shinkai et al., 2006). To examine the influence of these amino acid polymorphisms on the response to SC, we performed luciferase reporter assays with expression vectors carrying each of these SNPs. In the case of TLR5, a transfectant with R148L (a vector with an amino acid substitution located at the 5th LRR unit; Supplementary Fig. 1) showed about half the NF-␬B induction shown by the reference vector after stimulation with SC; a transfectant with P402L (a vector with an amino acid substitution at the 14th LRR unit) showed about one-third the NF-␬B induction shown by the reference vector (Fig. 2). Similar results were obtained in stimulation with flagellin, a standard ligand of TLR5 that was purified from ST (Hayashi et al., 2001). In the case of TLR2, a V703M mutant vector decreased responses to SC and Pam3 CSK4 , a standard ligand of the TLR2–TLR1 heterodimer (Takeuchi et al., 2002), by nearly half (Fig. 3A). TLR5R148L and TLR5P402L likely have direct impacts on interaction with the pathogen because of their locations in LRR domains of the extracellular region (Fig. 2B), whereas TLR2V703M likely affects signal transduction after interaction with the pathogen because it is located in the TIR domain of the intracellular region. In the case of TLR1, no amino acid polymorphisms dramatically changed the responses to either SC or Pam3 CSK4 (Fig. 3B). 3.3. Restricted distribution of SNPs in porcine TLR genes affecting response to SC In our previous study, the amino acid changes TLR2V703M , TLR5R148L , and TLR5P402L , which resulted from the SNPs TLR2G2107A , TLR5G443T , and TLR5C1205T , respectively, each existed in only a single breed among the 11 studied (Shinkai et al., 2006). Here we examined the genotypes of these SNPs in an additional 192 pigs of eight breeds to confirm the restricted distributions among breeds (Supplementary Table 1). As in the previous study, we found the SNPs TLR2G2107A , TLR5G443T , and TLR5C1205T only in Berkshire, Jinhua, and Landrace, respectively. There were obviously fewer pigs with these alleles than with the other alleles (Figs. 2A and 3A; Supplementary Table 1). There were no pigs that were homozygous for the TLR5G443T polymorphism, which results in a leucine residue at the 148th amino acid of TLR5 (Supplementary Table 1). 4. Discussion We demonstrated here that porcine TLR1, TLR2, and TLR5 involved in the recognition of SC in pigs. Infection with Salmonella spp. markedly alters the transcriptional profile in pig tissues; previous studies have reported on microarray analyses designed to clarify the molecular pathways of the porcine transcriptional response to SC infection. Expression of genes involved in apoptosis and the T helper 1 (Th1)-type immune response is induced during SC infection in the porcine lung (Uthe et al., 2007; Zhao et al., 2006). Microarray analyses have also indicated that many genes associated with the proinflammatory process, many of which are induced by the NF-␬B pathway, are involved with the response

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

A

S. Choleraesuis

100 75

NF-kB luciferrase on) (fold inductio

50 25 0 flagellin

20 15 10 5

B Consensus Swine i LRR5 5 Bovine LRR5 Human LRR5 Mouse LRR5 Chicken LRR5

Consensus Swine LRR14 Bovine LRR14 Human LRR14 Mouse LRR14 Chicken LRR14

D720N

S640F

R604Q

S618N

R522S

T416S

P402L

H278Q

G46D

R148L

M20I

S22F

Reference

EF6

0

TLR5-mutant

LxxLxLxxNxLxxxxxxxxxxxxxx LTRLDLSKNQIQSLHLHPSFQELNS LTHLDLSKNKIQSLYLHPSFRELNS LTRLDLSKNQIRSLYLHPSFGKLNS LARLDLSGNQIHSLRLHSSFRELNS LEELDLSGNQITKLHPHPLFYNLTI * .**** *:* .* *. * :*. LxxLxLxxNxLxxxxxxxx LNTLDLRDNALKTIQFIPS LNTLDLRDNALKTIQFI S LNTLDLRINALKTIYFLPS LQTLDLRDNALTTIHFIPS LQTLDLRDNALKAIGFIPS LKIIDLRDNAIKKLPSFPH *: :*** **:. : :*

Fig. 2. Influence of amino acid polymorphisms in porcine TLR5 on the recognition of Salmonella enterica serovar Choleraesuis (SC). (A) HEK293 cells were transiently cotransfected with one of the TLR5-mutant vectors, together with an NF-␬B reporter and Renilla control. Cells were stimulated with 1 × 107 cfu/ml SC or 10 ng/ml flagellin for 5 h, and luciferase activity was then measured. The results are means ± SD of triplicate wells and are representative of three independent experiments. Significant differences in response of TLR5R148L and TLRP402L to SC (open triangles, P < 0.001) and flagellin (closed triangles, P < 0.05) from those of the reference were observed by two-tailed Welch’s t-test. These results were reproducibly obtained in three independent experiments. (B) Comparative amino acid sequences of leucine-rich repeat (LRR) motifs containing the R148L or P402L polymorphisms of porcine TLR5 in vertebrates. Asterisks, colons, and periods under the aligned sequences indicate complete match, strong conservation, and weaker conservation of amino acids, respectively, in accordance with the standard of the ClustalX2 alignment software (Larkin et al., 2007). The LRR motif consists of the highly conserved segment, LxxLxLxxNxL, and the subsequent variable segment, in which “L” is leucine, isoleucine, valine, or phenylalanine, “N” is asparagine, threonine, serine, or cysteine and “x” is any amino acid. Polymorphic sites in pigs are underlined.

to SC in mesenteric lymph nodes (Wang et al., 2008), suggesting the importance of recognition of this pathogen by pattern recognition receptors (PRRs), particularly TLRs. Therefore, polymorphisms that occur in such TLRs and affect ligand recognition may influence the response to Salmonella infection and the onset of diarrhea of pigs. Many cases of involvement of TLRs in mucosal immunity and infections of mucosal tissues have been reported. Such observations prompted us to explore the associations between polymorphisms in TLRs and pig diseases such as pneumonia and diarrhea, which

1117

threaten the productivity of pig livestock. Our previous report (Shinkai et al., 2006), as well as the additional genotyping in this study, demonstrated that polymorphisms affecting the molecular function of the TLRs had a characteristically biased distribution among pig breeds. In the present study, alleles that poorly respond to the ligand were less observed than the other alleles. Furthermore, TLR2V703M was found only in Berkshire pigs and TLR5P402L only in Landrace pigs, all of which were sampled as unrelated individuals (i.e., lacking kinship). This suggests that the presence of these alleles endows a disadvantage in the resistance of pigs to pathogens in vivo. Landrace breed pigs frequently possess a defective allele type of MX1, which loses activity to suppress influenza A virus replication (Morozumi et al., 2001; Nakajima et al., 2007), suggesting that Landrace breed has not been well improved in terms of disease susceptibility. Alternatively, the biased distribution may have originated in the founding stock of these pig breeds, although the presence of these poorly responsive alleles may have some benefit in a feeding environment and therefore have remained to date. Experimental challenge with infectious agents in pigs possessing different alleles of TLRs will shed light on the role of TLRs and their polymorphisms in disease susceptibility. Recognition of microbes in mucosal tissues by PRRs is highly orchestrated, and there are complicated connections among the signals from PRRs (Clavel and Haller, 2007). Sole signaling by TLR2 without stimulation through nucleotide binding oligomerization domain 2 (NOD2) is not adequately controlled in the bowel tissues, although these two PRRs recognize components from peptidoglycan (Watanabe et al., 2006). Association studies in humans have indicated that a defect in the recognition of bowel microflora by NOD2 causes Crohn’s disease, probably because of failure to attenuate activated PRR signals by feedback mechanisms, such as secretion of IL-10 (Hugot et al., 2001; Moreira et al., 2008). We previously showed the existence of augmentation and abrogation of the response of porcine NOD2 to muramyl dipeptide by polymorphisms in the coding sequence of the NOD2 gene (Jozaki et al., 2009). The way in which PRRs combine to recognize Salmonella and the influence of polymorphisms in PRRs on pig diarrhea need to be comprehensively investigated in future. TLR2 has been regarded as a receptor that works with TLR1 or TLR6 as a heterodimer for lipopeptides derived from Grampositive bacteria (Takeuchi et al., 1999). However, a recent study found that biofilms generated by Gram-negative bacteria such as Salmonella spp. are recognized by the TLR2–TLR1 heterodimer (Tükel et al., 2010). This suggests that TLR2 has a role as a sensor for Gram-negative bacteria, including Salmonella spp. Here, we demonstrated that porcine TLR2 and TLR1 are involved in the response to the cell bodies of SC, indicating that the TLR2–TLR1 heterodimer is also important for recognition of Salmonella in this species. Crystal structure analysis of human TLR2–TLR1 heterodimer has elucidated the involvement of the 9th to 12th LRR units in TLR1 in ligand binding, and of the 11th to 14th LRRs in dimerization with TLR2 (Jin et al., 2007). In porcine TLR1, there are no nonsynonymous SNPs in a large region between the amino acid polymorphisms of A217T and L434M, corresponding to the 8th to 16th LRR units, despite the existence of a number of polymorphisms in the other LRRs (Shinkai et al., 2006). This suggests the importance and rigidity of this region in porcine TLR1 and the flexibility of the rest. The TIR domains of TLRs have similar structures to each other, and structural information on the TIRs derived from several TLRs has been reported (Gautam et al., 2006; Khan et al., 2004; Xu et al., 2000). Although the heterodimer interface between the TIRs of TLR2 and TLR1 or TLR6 is unknown, crystallography of human TLR10 has shown the homodimeric structure of its TIR domains, which is in good agreement with the biochemical data accumulated to date (Nyman et al., 2008). Alignment of the TIR domain of

1118

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

A

S. Choleraesuis

60 45

NF-kB luciferase (fold induction)

30 15 0 Pam3CSK4

80 60 40 20

V703M

K517Q

V504L

T432A

A338T

N239R

K216E

R210P

P136A

T126A

Reference

EF6

0

TLR2-mutant

B S. Choleraesuis

300

NF-kB luciferase (fold induction)

200 100 0 Pam3CSK4

200 150 100 50

EF6 Reference I18T S113L M117T D131N I135V G169S S178A A217T L434M T440I I451V N458S I500T N527D S553G S557T E559K A596V G603S H769Y

0

TLR1-mutant

Fig. 3. Influence of amino acid polymorphisms in porcine TLR2 and TLR1 on recognition of Salmonella enterica serovar Choleraesuis (SC). HEK293 cells were transiently cotransfected with one of the TLR2-mutant vectors (A) or with the same amount of TLR2-reference vector and one of the TLR1-mutant vectors (B), together with NF-␬B reporter and Renilla control. Cells were stimulated with 5 × 107 cfu/ml SC or 100 ng/ml Pam3 CSK4 (A), or with 1 × 107 cfu/ml SC or 100 ng/ml Pam3 CSK4 (B) for 5 h, and luciferase activity was then measured. Results are means ± SD of triplicate wells and are representative of three independent experiments. Significant differences of the responses of TLR2V703M to SC (closed triangle, P < 0.05) and Pam3 CSK4 (open triangle, P < 0.001) from those of the reference were observed by two-tailed Welch’s t-test. These results were reproducibly obtained in three independent experiments (A). There were no significant differences (P < 0.05, two-tailed Welch’s t-test) observed reproducibly in three independent experiments among responses of the TLR1-mutant vectors to SC or Pam3 CSK4 from those of the reference (B).

porcine TLR2 with that of human TLR10 demonstrated that V703 of porcine TLR2 is located inside the structure of the TIR domain (data not shown). This indicates that V703 does not interact directly with the TIR domain of TLR1 or TLR6 or with signaling molecules in the downstream of TLR2; instead, it alters the structure of the surface of the TIR domain, resulting in downregulation of the stimulatory signal from the ligand. Both of the observed SNPs in porcine TLR5 that affected ligand recognition were mapped within LRR units. Because the crystal structure of the TLR5 molecule has not yet been reported, we modeled the 3D structure of the ectodomain of porcine TLR5 on the basis of the crystal structures of TLR3 and TLR4 in human and mouse (Bell et al., 2005; Choe et al., 2005; Kim et al., 2007; Liu et al., 2008; Park et al., 2009). TLR3 contains 24 LRRs, and TLR4 contains 22 LRRs. The number of LRRs in TLR5 is uncertain and could be 21 or 22. Although a value of 22 LRRs has been adopted in several recent reports (Gong

et al., 2010; Matsushima et al., 2007; Wei et al., 2011), we present a model with 21 LRRs, because the 22-LRR model contains repeats that are too short (Fig. 4A and B). Superposition of the modeled TLR5 ectodomain onto the TLR4–MD-2 complex shows that R148 in porcine TLR5, located in the 5th LRR unit, corresponds to the region of TLR4 proximal to the MD-2 protein (Fig. 4C and Supplementary Fig. 1). This implies that amino acid change at R148 may confer a structural change on the site involved in ligand recognition by porcine TLR5. Furthermore, the positively charged residue is conserved among mammals at the location corresponding to R148 of porcine TLR5 (Fig. 2B). Interestingly, chicken TLR5 possesses a negatively charged glutamic acid at the residue, suggesting that there is a difference between chickens and mammals in the reactivity of TLR5 to flagellated bacteria, although further investigation is required to verify this possibility. In contrast, superposition of TLR5 onto the TLR3–RNA complex shows that R148 is 18 A˚ away from the

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

1119

Fig. 4. Predicted structure of porcine TLR5. The models of TLR5 (yellow and turquoise) were aligned to the structures of TLR4–MD-2 complex (A) or TLR3–RNA complex (B) and is shown with MD-2 (pink) or RNA (purple) molecules from the original complexes. R148 and P402 residues in TLR5 are also shown, and those on one chain (yellow) are labeled. (C) Close-up view of TLR5 aligned to TLR4. R148, D293, and D365 in one chain (yellow) and P402 in the other chain (turquoise) are shown. (D) R148 in TLR5 aligned to TLR3 is shown from another perspective.

superposed RNA molecule (Fig. 4B and D), suggesting that ligand recognition by TLR5 is quite dissimilar from that by TLR3. In our alignment, P402 was located on an insertion in the 14th LRR of TLR5 (Supplementary Fig. 1). This insertion protrudes toward the superposed MD-2 molecule (Fig. 4C). Although the actual conformation of this insertion is unknown, it may be in contact with flagellin. Often proline is a key residue for maintaining the structure of the polypeptide, and substitution of proline to another residue in receptor molecules often abrogates the ligand recognition ability of the receptors (MacArthur and Thornton, 1991; Omueti et al., 2007). An in vitro binding assay using expression vectors for truncated human TLR5 has shown that amino acid sequence 386–407 of TLR5 is a possible binding site for flagellin (Mizel et al., 2003). In contrast, molecular modeling and mutagenesis by another group suggest that different portions (around the 295th and 367th aspartic acids) form concavities on the inner surface of the horseshoe-shaped ectodomain and are responsible for the interaction of human TLR5 with flagellin (Andersen-Nissen et al., 2007). Our data support the importance of the structure of the 14th LRR domain, which corresponds to the region suggested by Mizel et al. (2003), in the recognition of flagellin by TLR5. This is also endorsed by the observation that proline corresponding to P402 of porcine TLR5 is conserved in the TLR5 molecules of other vertebrates (Fig. 2B). Nevertheless, it seems possible that residues R148 and P402 in porcine TLR5 form an interaction surface for flagellin together with D293 and D365, which correspond to D295 and D367, respectively, in human TLR5 (Fig. 4C). In conclusion, we showed the recognition of SC by porcine TLRs, and demonstrated that some of the known polymorphisms in porcine TLRs affected recognition of the bacteria. Investigation of effects of these polymorphisms in vivo has to be carried out; however, these particular polymorphisms may be good indices for pig breeding aimed at disease resistance. They may give us clues to the design of synthetic vaccines that will prevent SC-

related pig diarrhea and zoonotic infections of humans from pigs. Conflict of interest The authors declare that there is no conflict of interest in this study. Acknowledgements We thank Dr. Yasuko Hanafusa (National Institute of Animal Health) for arrangement of the collaboration in this study. This work was supported by the Animal Genome Research Project of the Ministry of Agriculture, Forestry, and Fisheries of Japan and by a Grant-in-Aid from the Japan Racing Association. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2011.02.004. References Akira, S., Takeda, K., 2004. Toll-like receptor signalling. Nat. Rev. Immunol. 4, 499–511. Andersen-Nissen, E., Smith, K.D., Bonneau, R., Strong, R.K., Aderem, A., 2007. A conserved surface on Toll-like receptor 5 recognizes bacterial flagellin. J. Exp. Med. 204, 393–403. Arbour, N.C., Lorenz, E., Schutte, B.C., Zabner, J., Kline, J.N., Jones, M., Frees, K., Watt, J.L., Schwartz, D.A., 2000. TLR4 mutations are associated with endotoxin hyporesponsiveness in humans. Nat. Genet. 25, 187–191. Barber, R.C., Aragaki, C.C., Rivera-Chavez, F.A., Purdue, G.F., Hunt, J.L., Horton, J.W., 2004. TLR4 and TNF-␣ polymorphisms are associated with an increased risk for severe sepsis following burn injury. J. Med. Genet. 41, 808–813. Bell, J.K., Botos, I., Hall, P.R., Askins, J., Shiloach, J., Segal, D.M., Davies, D.R., 2005. The molecular structure of the Toll-like receptor 3 ligand-binding domain. Proc. Natl. Acad. Sci. USA 102, 10976–10980.

1120

H. Shinkai et al. / Molecular Immunology 48 (2011) 1114–1120

Bochud, P.Y., Hawn, T.R., Aderem, A., 2003. Cutting edge: a Toll-like receptor 2 polymorphism that is associated with lepromatous leprosy is unable to mediate mycobacterial signaling. J. Immunol. 170, 3451–3454. Chiu, C.H., Chuang, C.H., Chiu, S., Su, L.H., Lin, T.Y., 2006. Salmonella enterica serotype Choleraesuis infections in pediatric patients. Pediatrics 117, e1193–e1196. Chiu, C.H., Su, L.H., Chu, C., 2004. Salmonella enterica serotype Choleraesuis: epidemiology, pathogenesis, clinical disease, and treatment. Clin. Microbiol. Rev. 17, 311–322. Choe, J., Kelker, M.S., Wilson, I.A., 2005. Crystal structure of human Toll-like receptor 3 (TLR3) ectodomain. Science 309, 581–585. Clavel, T., Haller, D., 2007. Bacteria- and host-derived mechanisms to control intestinal epithelial cell homeostasis: implications for chronic inflammation. Inflamm. Bowel Dis. 13, 1153–1164. Edwards, R.A., Olsen, G.J., Maloy, S.R., 2002. Comparative genomics of closely related salmonellae. Trends Microbiol. 10, 94–99. Franchimont, D., Vermeire, S., El Housni, H., Pierik, M., Van Steen, K., Gustot, T., Quertinmont, E., Abramowicz, M., Van Gossum, A., Devière, J., Rutgeerts, P., 2004. Deficient host–bacteria interactions in inflammatory bowel disease? The toll-like receptor (TLR)-4 Asp299gly polymorphism is associated with Crohn’s disease and ulcerative colitis. Gut 53, 987–992. Gautam, J.K., Ashish, Comeau, L.D., Krueger, J.K., Smith Jr., M.F., 2006. Structural and functional evidence for the role of the TLR2 DD loop in TLR1/TLR2 heterodimerization and signaling. J. Biol. Chem. 281, 30132–30142. Gong, J., Wei, T., Zhang, N., Jamitzky, F., Heckl, W.M., Rössle, S.C., Stark, R.W., 2010. TollML: a database of toll-like receptor structural motifs. J. Mol. Model. 16, 1283–1289. Hawn, T.R., Verbon, A., Lettinga, K.D., Zhao, L.P., Li, S.S., Laws, R.J., Skerrett, S.J., Beutler, B., Schroeder, L., Nachman, A., Ozinsky, A., Smith, K.D., Aderem, A., 2003. A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to Legionnaires’ disease. J. Exp. Med. 198, 1563–1572. Hayashi, F., Smith, K.D., Ozinsky, A., Hawn, T.R., Yi, E.C., Goodlett, D.R., Eng, J.K., Akira, S., Underhill, D.M., Aderem, A., 2001. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410, 1099–1103. Hugot, J.P., Chamaillard, M., Zouali, H., Lesage, S., Cézard, J.P., Belaiche, J., Almer, S., Tysk, C., O’Morain, C.A., Gassull, M., Binder, V., Finkel, Y., Cortot, A., Modigliani, R., Laurent-Puig, P., Gower-Rousseau, C., Macry, J., Colombel, J.F., Sahbatou, M., Thomas, G., 2001. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature 411, 599–603. Jin, M.S., Kim, S.E., Heo, J.Y., Lee, M.E., Kim, H.M., Paik, S.G., Lee, H., Lee, J.O., 2007. Crystal structure of the TLR1–TLR2 heterodimer induced by binding of a triacylated lipopeptide. Cell 130, 1071–1082. Jozaki, K., Shinkai, H., Tanaka-Matsuda, M., Morozumi, T., Matsumoto, T., Toki, D., Okumura, N., Eguchi-Ogawa, T., Kojima-Shibata, C., Kadowaki, H., Suzuki, E., Wada, Y., Uenishi, H., 2009. Influence of polymorphisms in porcine NOD2 on ligand recognition. Mol. Immunol. 47, 247–252. Kadowaki, N., Ho, S., Antonenko, S., Malefyt, R.W., Kastelein, R.A., Bazan, F., Liu, Y.J., 2001. Subsets of human dendritic cell precursors express different Toll-like receptors and respond to different microbial antigens. J. Exp. Med. 194, 863–869. Kang, T.J., Chae, G.T., 2001. Detection of Toll-like receptor 2 (TLR2) mutation in the lepromatous leprosy patients. FEMS Immunol. Med. Microbiol. 31, 53–58. Khan, J.A., Brint, E.K., O’Neill, L.A., Tong, L., 2004. Crystal structure of the Toll/interleukin-1 receptor domain of human IL-1RAPL. J. Biol. Chem. 279, 31664–31670. Kim, H.M., Park, B.S., Kim, J.I., Kim, S.E., Lee, J., Oh, S.C., Enkhbayar, P., Matsushima, N., Lee, H., Yoo, O.J., Lee, J.O., 2007. Crystal structure of the TLR4–MD-2 complex with bound endotoxin antagonist Eritoran. Cell 130, 906–917. Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J., Higgins, D.G., 2007. Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947–2948. Lazarus, R., Vercelli, D., Palmer, L.J., Klimecki, W.J., Silverman, E.K., Richter, B., Riva, A., Ramoni, M., Martinez, F.D., Weiss, S.T., Kwiatkowski, D.J., 2002. Single nucleotide polymorphisms in innate immunity genes: abundant variation and potential role in complex human disease. Immunol. Rev. 190, 9–25. Liu, L., Botos, I., Wang, Y., Leonard, J.N., Shiloach, J., Segal, D.M., Davies, D.R., 2008. Structural basis of Toll-like receptor 3 signaling with double-stranded RNA. Science 320, 379–381. MacArthur, M.W., Thornton, J.M., 1991. Influence of proline residues on protein conformation. J. Mol. Biol. 218, 397–412. Matsushima, N., Tanaka, T., Enkhbayar, P., Mikami, T., Taga, M., Yamada, K., Kuroki, Y., 2007. Comparative sequence analysis of leucine-rich repeats (LRRs) within vertebrate toll-like receptors. BMC Genomics 8, 124. Mizel, S.B., West, A.P., Hantgan, R.R., 2003. Identification of a sequence in human toll-like receptor 5 required for the binding of Gram-negative flagellin. J. Biol. Chem. 278, 23624–23629.

˜ Moreira, L.O., El Kasmi, K.C., Smith, A.M., Finkelstein, D., Fillon, S., Kim, Y.G., Núnez, G., Tuomanen, E., Murray, P.J., 2008. The TLR2–MyD88–NOD2–RIPK2 signalling axis regulates a balanced pro-inflammatory and IL-10-mediated anti-inflammatory cytokine response to Gram-positive cell walls. Cell. Microbiol. 10, 2067–2077. Morozumi, T., Sumantri, C., Nakajima, E., Kobayashi, E., Asano, A., Oishi, T., Mitsuhashi, T., Watanabe, T., Hamasima, N., 2001. Three types of polymorphisms in exon 14 in porcine Mx1 gene. Biochem. Genet. 39, 251–260. Morozumi, T., Uenishi, H., 2009. Polymorphism distribution and structural conservation in RNA-sensing Toll-like receptors 3, 7, and 8 in pigs. Biochim. Biophys. Acta 1790, 267–274. Nakajima, E., Morozumi, T., Tsukamoto, K., Watanabe, T., Plastow, G., Mitsuhashi, T., 2007. A naturally occurring variant of porcine Mx1 associated with increased susceptibility to influenza virus in vitro. Biochem. Genet. 45, 11–24. Nyman, T., Stenmark, P., Flodin, S., Johansson, I., Hammarström, M., Nordlund, P., 2008. The crystal structure of the human Toll-like receptor 10 cytoplasmic domain reveals a putative signaling dimer. J. Biol. Chem. 283, 11861–11865. Ogus, A.C., Yoldas, B., Ozdemir, T., Uguz, A., Olcen, S., Keser, I., Coskun, M., Cilli, A., Yegin, O., 2004. The Arg753Gln polymorphism of the human Toll-like receptor 2 gene in tuberculosis disease. Eur. Respir. J. 23, 219–223. Omueti, K.O., Mazur, D.J., Thompson, K.S., Lyle, E.A., Tapping, R.I., 2007. The polymorphism P315L of human Toll-like receptor 1 impairs innate immune sensing of microbial cell wall components. J. Immunol. 178, 6387–6394. Park, B.S., Song, D.H., Kim, H.M., Choi, B.S., Lee, H., Lee, J.O., 2009. The structural basis of lipopolysaccharide recognition by the TLR4–MD-2 complex. Nature 458, 1191–1195. ˇ Sali, A., Blundell, T.L., 1993. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234, 779–815. Sambrook, J., Russell, D.W., 2001. Molecular Cloning, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Schindler, U., Baichwal, V.R., 1994. Three NF-␬B binding sites in the human E-selectin gene required for maximal tumor necrosis factor alpha-induced expression. Mol. Cell. Biol. 14, 5820–5831. Schröder, N.W., Schumann, R.R., 2005. Single nucleotide polymorphisms of Toll-like receptors and susceptibility to infectious disease. Lancet Infect. Dis. 5, 156–164. Shinkai, H., Tanaka, M., Morozumi, T., Eguchi-Ogawa, T., Okumura, N., Muneta, Y., Awata, T., Uenishi, H., 2006. Biased distribution of single nucleotide polymorphisms (SNPs) in porcine Toll-like receptor 1 (TLR1), TLR2, TLR4, TLR5, and TLR6 genes. Immunogenetics 58, 324–330. Takeuchi, O., Hoshino, K., Kawai, T., Sanjo, H., Takada, H., Ogawa, T., Takeda, K., Akira, S., 1999. Differential roles of TLR2 and TLR4 in recognition of Gram-negative and Gram-positive bacterial cell wall components. Immunity 11, 443–451. Takeuchi, O., Kawai, T., Muhlradt, P.F., Morr, M., Radolf, J.D., Zychlinsky, A., Takeda, K., Akira, S., 2001. Discrimination of bacterial lipoproteins by Toll-like receptor 6. Int. Immunol. 13, 933–940. Takeuchi, O., Sato, S., Horiuchi, T., Hoshino, K., Takeda, K., Dong, Z., Modlin, R.L., Akira, S., 2002. Cutting edge: role of Toll-like receptor 1 in mediating immune response to microbial lipoproteins. J. Immunol. 169, 10–14. Tükel, C¸., Nishimori, J.H., Wilson, R.P., Winter, M.G., Keestra, A.M., van Putten, J.P., Bäumler, A.J., 2010. Toll-like receptors 1 and 2 cooperatively mediate immune responses to curli, a common amyloid from enterobacterial biofilms. Cell. Microbiol. 12, 1495–1505. Uenishi, H., Shinkai, H., 2009. Porcine Toll-like receptors: the front line of pathogen monitoring and possible implications for disease resistance. Dev. Comp. Immunol. 33, 353–361. Uthe, J.J., Royaee, A., Lunney, J.K., Stabel, T.J., Zhao, S.H., Tuggle, C.K., Bearson, S.M., 2007. Porcine differential gene expression in response to Salmonella enterica serovars Choleraesuis and Typhimurium. Mol. Immunol. 44, 2900–2914. Wang, Y., Couture, O.P., Qu, L., Uthe, J.J., Bearson, S.M., Kuhar, D., Lunney, J.K., Nettleton, D., Dekkers, J.C., Tuggle, C.K., 2008. Analysis of porcine transcriptional response to Salmonella enterica serovar Choleraesuis suggests novel targets of NFkappaB are activated in the mesenteric lymph node. BMC Genomics 9, 437. Watanabe, T., Kitani, A., Murray, P.J., Wakatsuki, Y., Fuss, I.J., Strober, W., 2006. Nucleotide binding oligomerization domain 2 deficiency leads to dysregulated TLR2 signaling and induction of antigen-specific colitis. Immunity 25, 473–485. Wei, T., Gong, J., Rössle, S.C., Jamitzky, F., Heckl, W.M., Stark, R.W., 2011. A leucinerich repeat assembly approach for homology modeling of the human TLR5-10 and mouse TLR11-13 ectodomains. J. Mol. Model. 17, 27–37. Xu, Y., Tao, X., Shen, B., Horng, T., Medzhitov, R., Manley, J.L., Tong, L., 2000. Structural basis for signal transduction by the Toll/interleukin-1 receptor domains. Nature 408, 111–115. Zhao, S.H., Kuhar, D., Lunney, J.K., Dawson, H., Guidry, C., Uthe, J.J., Bearson, S.M., Recknor, J., Nettleton, D., Tuggle, C.K., 2006. Gene expression profiling in Salmonella Choleraesuis-infected porcine lung using a long oligonucleotide microarray. Mamm. Genome 17, 777–789.