Yersinia pestis TIR-domain protein forms

dimers that interact with the human adaptor

protein MyD88

Rohini R. Rana1, Peter Simpson1, Minghao Zhang1, Matthew Jennions3,

Chimaka Ukegbu1, Abigail M. Spear2, Yilmaz Alguel1, Stephen J.

Matthews1, Helen S. Atkins2 and Bernadette Byrne1*

Division of Molecular Biosciences, Imperial College London, London SW7 2AZ,

UK1, Department of Biomedical Sciences, Defence Science and Technology

Laboratory, Porton Down, Salisbury, Wiltshire, SP4 0JQ, UK2 and Membrane Protein

Laboratory, Diamond Light Source, Harwell Science and Innovation Campus,

Chilton, Didcot, Oxfordshire OX11 0DE, UK3.

*Corresponding

Dr Bernadette Byrne,

Division of Molecular Biosciences

Imperial College London

South Kensington

London

SW7 2AZ.

Fax: +44 20 7594 3022

E-mail: b.byrne@imperial.ac.uk

1

@ Crown Copyright 2010. Published with the permission of the Defence Science and

Technology Laboratory on behalf of the Controller of HMSO.

Abstract

Recent research has highlighted the presence of Toll/Interleukin 1 receptor (TIR)-

domain proteins (Tdps) in a range of bacteria, suggested to form interactions with the

human adaptor protein MyD88 and inhibit intracellular signaling from Toll-like

receptors (TLRs). A Tdp has been identified in Yersinia pestis (YpTdp), a highly

pathogenic bacterium responsible for plague. Expression of a number of YpTIR

constructs of differing lengths (YpTIR1, S130-A285; YpTIR2, I137-I273; YpTIR3,

I137-246; YpTIR4, D107-S281) as fusions with an N-terminal GB1 tag (the B1

immunoglobulin domain of Streptococcal protein G) yielded high levels of soluble

protein. Subsequent purification yielded 4-6 mg/L pure, folded protein. Thrombin

cleavage allowed separation of the GB1 tag from YpTIR4 resulting in folded protein

after cleavage. Nuclear magnetic resonance spectroscopy, size exclusion

chromatography, SDS-PAGE analysis and static light scattering all indicate that the

YpTIR forms dimers. Generation of a double Cys-less mutant resulted in an unstable

protein containing mainly monomers indicating the importance of disulphide bonds in

dimer formation. In addition, the YpTIR constructs have been shown to interact with

the human adaptor protein MyD88 using 2D NMR and GST pull down. YpTIR is an

excellent candidate for further study of the mechanism of action of pathogenic

bacterial Tdps.

2

Keywords: Innate immune evasion; TIR domain protein; Yersinia pestis; pathogenic;

MyD88

1. Introduction

Toll-like receptors (TLRs) are key components of the innate immune system, which

detect a range of pathogen associated molecular patterns (PAMPs) including

lipopolysaccharide (LPS), bacterial cell wall components and nucleic acids [1], and

initiate the first line of host defence against infection. TLRs have a conserved domain

architecture comprised of a large extracellularly located leucine rich repeat (LRR)

domain [2] and an intracellular Toll/Interleukin 1 receptor (TIR) domain [3] linked by

a single pass transmembrane region. TLRs are suggested to exist as dimers, which can

be either heterotypic, homotypic or both depending on the receptor [4,5]. Upon

interaction with a PAMP the TLR dimer is thought to undergo a molecular

rearrangement of the intracellular TIR domains to generate an active interaction

domain [4,5,6] allowing recruitment of intracellular adaptor proteins; MyD88,

MyD88 adaptor like (MAL, also called TIRAP), TIR-domain-containing adaptor

protein inducing IFN- (TRIF), TRIF-related adaptor molecule (TRAM) and sterile

- and armadillo-motif containing protein (SARM) [7]. All of the adaptors also

contain TIR domains and in some cases other protein-protein interaction domains

important in signal transduction [8].

Broadly there are two signal transduction pathways, the MyD88 dependent

pathway utilized by all TLRs except TLR3 and the MyD88 independent pathway

utilized by TLR3 and also TLR4 [7,9]. The recruitment of adaptor proteins initiates

the complex signaling cascades which upregulate gene expression of nuclear factor

3

kappa B (NF-B) and interferon response factor (IRF) controlled genes. This

ultimately initiates an innate immune response through the release of proinflammatory

cytokines [7].

Key components of the signaling cascade are the heterotypic interactions

between the TIR domains upon receptor activation and adaptor recruitment. The

importance of the TIR domains in the innate immune response has made them the

subject of intense study. The structures of the monomeric forms of the TIR domains

of TLR1 and TLR2 [10] revealed a conserved architecture comprised of a central

five-stranded parallel -sheet surrounded by five -helices. These structures

highlighted the presence of a flexible region, the BB loop connecting strand B and

helix B, which projects away from the main TIR domain. Mutational analysis

revealed that several residues within this loop have important roles in signal

transduction [10], most notably a proline residue. Mutation of the equivalent residue,

Pro712, to histidine in TLR4-TIR results in mice unable to initiate the innate immune

response upon stimulation with LPS [11]. The BB loop also plays a role in mediating

the interactions between the monomers of the dimer of TLR10-TIR [12]. The recent

NMR structure of the adaptor MyD88-TIR domain confirmed the conserved

architecture of the TIR domain family and the flexible nature of the BB loop [13].

Bioinformatics analysis has identified TIR containing proteins in a range of

both pathogenic and non-pathogenic bacteria. The first studied example of such a

protein was the TIR-like protein A (TlpA) identified in Salmonella enterica serovar

Enteridis [14]. In vitro analysis showed that this protein was able to reduce the ability

of TLR4 and MyD88 to stimulate NF-B activity. Furthermore bacteria with

disrupted TlpA genes exhibited reduced pathogenicity in mice, suggesting that TlpA is

an important virulence factor. These data indicated that the bacterial TIR-like proteins

4

were likely to have a role in evasion of the innate immune system. Studies on

equivalent proteins, from Brucella abortus (Btp1), the uropathogenic Escherichia coli

strain CFT073 (TcpC) and Brucella melintensis (TcpB) have supported a role in

immune system evasion [15,16,17].

A recent in-depth bioinformatic analysis [18] identified 922 TIR-domain

proteins in a range of fungi, archaea, viruses and pathogenic and non-pathogenic

bacteria. The high incidence of these proteins in non-pathogenic species suggests that

these proteins may have a range of functions, including evasion of the innate immune

system, depending on the organism in question. However it has been shown that the

TIR-domain protein from the non-pathogenic thermophilic bacterium, Paracoccus

dentrificans, PdTLP interacts with the human adaptor protein MyD88 in vitro [19,20].

In addition, the crystal structure of PdTIR displays a similar fold to the known human

TIR domains [20]. Given the benign nature of this organism it would seem unlikely

that this protein has a role in innate immune system evasion, therefore making it of

limited use as a model system for studying the interactions between bacterial TIR

proteins and human TIR proteins. Further work is required to reveal the precise

function of PdTIR.

Our bioinformatic analysis [18] identified a TIR domain protein in the

pathogenic bacterium, Yersina pestis (YpTdp). Y. pestis is a facultative intracellular

bacterium that causes the zoonotic diseases bubonic and pneumonic plague [21]. The

main reservoirs of the disease are rodents, birds, farm animals and their associated

fleas and the disease is transmitted to humans via flea or animal bites. Most famously

associated with the Black Death pandemics of the Middle Ages, Y. pestis continues to

cause a significant number of deaths every year mainly in Africa and Asia [22,23].

YpTdp is a 41 kDa protein containing five cysteines. Other work in our group has

5

shown that when over-expressed in vitro YpTdp is able to disrupt immune

signalling pathways but that its removal has no obvious effect on virulence and

instead affects the characteristics of Yersinia pestis growth (Spear at al,

manuscript submitted). Additionally, microarray studies have shown that the gene

encoding YpTdp is expressed in vitro in various conditions [24,25]. In order to

understand more about the role of YpTdp in immune system evasion, we have

attempted to produce the TIR domain of YpTdp (YpTIR) with a view to obtaining

high quality samples for functional and structural studies. Here, we have expressed

and purified YpTIR as a fusion with an N-terminal GB1 tag. The resulting protein is

folded and exists in solution as dimer that interacts with the TIR domain of human

adaptor protein MyD88.

2. Results

2.1 Expression and purification of GB1-tagged YpTIR constructs

The region of the YpTdp containing the TIR domain (YpTIR1; residues S130 to

A285) was estimated based on sequence alignments with TIR domains of known

structure (Figure 1). A construct based on this region and two shorter YpTIR

constructs, corresponding to residues I137-I273 (YpTIR2) and I137-N246 (YpTIR3),

were generated. YpTIR3 lacks the region of the protein corresponding to the Box 3

motif (Figure 1). All three gene fragments were cloned into the GEV2 vector as a

fusion with an N-terminal GB1 tag (the B1 immunoglobulin domain of Streptococcal

protein G) [26] and a C-terminal His tag. High-level expression was achieved for

GB1-YpTIR1 and GB1-YpTIR2 but not for GB1-YpTIR3. Affinity chromatography

followed by size exclusion chromatography (SEC) of all three constructs produced

highly pure protein although the final yield of GB1-YpTIR3 was extremely low

6

(Figure 2ab). The final yields of pure protein obtained were 5.7, 4.6 and 0.6 mg/L for

GB1-YpTIR1, GB1-YpTIR2 and GB1-YpTIR3 respectively. The size exclusion

profiles of GB1-YpTIR1 and GB1-YpTIR2 indicated that both proteins were

monodispersed (Figure 2a). However in both cases the protein had a lower retention

volume (~14 ml) than expected for a ~25 kDa protein suggesting that the protein was

present in a higher oligomeric form, possibly a dimer. The GB1-YpTIR3 eluted as a

broad peak from the SEC column suggesting that this protein was aggregating

possibly as the result of partial unfolding. The high yield, purity and quality (Figure

2ab) of GB1-YpTIR1 and GB1-YpTIR2 allowed further analysis of the proteins.

2.2 Removal of the GB1 tag

Whilst the expression and isolation of the GB1 tagged constructs is useful, it is

important to cleave the GB1 tag for certain downstream applications including some

structural studies. Attempts to cleave the GB1 tag from GB1-YpTIR1 and GB1-

YpTIR2 were unsuccessful even after extended periods of incubation with high

concentrations of thrombin protease, probably as the result of close association of the

GB1 and YpTIR domains. In order to allow efficient thrombin cleavage a further

longer YpTIR (GB1-YpTIR4, residues D107-S281) construct was designed. This was

also cloned into GEV2 and expressed and purified as described in Section 2.1 (Figure

2cd). Cleavage was followed by Co2+-IMAC to separate the His-tagged YpTIR4 from

the GB1 tag resulting in pure, cleaved protein with a yield of 2.5 mg/L.

2.3 YpTIR is a dimer in solution

In order to assess the folded state of the YpTIR constructs prior to further studies, the

GB1-YpTIR protein constructs were analysed using 1D NMR spectroscopy. All GB1-

YpTIR constructs displayed folded domains in addition to GB1 (Figure 3). YpTIR4

after thrombin cleavage also displayed folded domains. Resonance linewidths from

7

the 1D and 2D 1H-15N HSQC NMR spectra of 15N-GB1-YpTIR1 (Figure 3) strongly

suggest that the protein is dimeric, as suggested by SEC. This was further supported

by analysis using a SEC column together with static light scattering (SLS) which

indicated that the approximate molecular weights of the different YpTIR proteins

were ~40-54 kDa. The measured molecular weight of 40.9 kDa for YpTIR4

(predicted molecular weight of the monomer = 20.8 kDa) following removal of the

GB1 tag (Figure 4) strongly indicated that dimer formation is mediated through the

YpTIR domain. The YpTIR contains two Cys residues; Cys90 and Cys132. In order

to investigate the role of these residues in the formation of a stable dimeric protein we

generated both single and a double Cys-less mutant. All mutants were expressed and

isolated as GB1-tag fusion proteins and were shown to be folded by 1D NMR (data

not shown). SEC-SLS analysis of the Cys90Ser and Cys132Ser single mutants

indicated that both had the same molecular weight as the wild-type YpTIR (Fig 4) and

thus were in the dimeric form. The double Cys-less mutant resulted in a much less

stable protein. Although it proved impossible to obtain an accurate molecular weight

for this protein the SEC-SLS indicated that the protein sample contained mainly

monomers (data not shown).

2.4 Interaction of YpTIR with the human adaptor protein MyD88

The 2D NMR spectra of 15N-GB1-YpTIR1 revealed the presence of peaks

corresponding to amino acid residues in addition to GB1 in the folded regions of the

spectrum (> 9 ppm). These additional peaks were observed to be significantly broader

than expected for a monomer of YpTIR1, presumably as a result of the protein

tumbling as a dimer (Figure 5a). Upon addition of unlabelled GB1-MyD88-TIR,

specific broadening of signals corresponding to structured regions of the YpTIR

domain was observed, whilst those from GB1 were unaffected (Figure 5b). This

8

strongly suggests that MyD88-TIR binds to the YpTIR domain, causing signal

broadening presumably as a result of the increased tumbling time and/or a binding

regime that is intermediate on the NMR timescale. The interaction between the

YpTIR constructs and MyD88 was also assessed by pull down using GST-MyD88 as

bait and GB1-YpTIR2 and YpTIR4 following removal of the GB1 tag as prey. Both

proteins specifically interact with GST-MyD88-TIR as shown in Figure 5cd.

Interestingly the uncleaved GB1-YpTIR4 interacts non-specifically with the

glutathione resin (Figure 5c) while the Cys132Ser mutant does not interact at all

(Figure 5e).

3. Discussion

This study describes a biophysical and functional characterization of a TIR domain

from the highly pathogenic bacterium Y. pestis. One of the key aims was to generate

constructs that formed a compact stable structure suitable for downstream structural

and functional studies. Expression of GB1-YpTIR1 and GB1-YpTIR2 yielded high

levels of soluble protein, which could be readily isolated to high homogeneity in a

folded and functional state. Interestingly the shortest construct, GB1-YpTIR3,

expressed poorly and yielded low levels of polydispersed protein. YpTIR3 is

truncated at residue N246 removing the TIR motif sequence Box 3 (Figure 1), which

is likely to be important for maintaining the overall fold of the YpTIR.

SEC analysis of the GB1-YpTIR1 and GB1-YpTIR2 suggested a shorter

retention time than would be expected for the ~25 kDa sized monomer and indicated

that the proteins were likely to be dimeric. This was supported by NMR spectra

displaying broader than expected peaks. The dimeric status of YpTIR was confirmed

by the SEC-SLS. This is a feature unique to YpTIR as the other TIR domains are

9

present as monomers in solution [13,19]. Mutating the two Cys residues, Cys90 and

Cys132, individually to Ser had no effect on the dimeric status of YpTIR. In contrast,

while it was difficult to obtain precise data on the molecular weight of the Cys-less

mutant, it was present mainly as monomers. These results suggest that mutation of

one Cys is tolerated however the mutation of both Cys residues results in the loss of

the stable quaternary structure and indicates that disulphide bridges have a key role in

mediating dimer formation.

It is not clear what the precise molecular arrangement of the Cys residues is

within the protein. If Cys132 forms disulphide bonds with Cys90 then it could be

expected that the mutation of one or the other Cys residue would result in the

monomeric form of the protein. However since both single mutants retain the dimeric

form this suggests that Cys90 of monomer 1 forms a disulphide bond with Cys90 of

monomer 2 and Cys132 of monomer 1 forms a disulpide bond with Cys132 of

monomer 2. The mutation of Cys132 to Ser however does result in a loss of

interaction between YpTIR and MyD88. This suggests that whilst one disulphide

bridge may be sufficient to maintain the dimer, both are required to maintain the

functional conformation of the protein. A high resolution structure of YpTIR will

shed light on these issues.

Although YpTIR is the only TIR domain so far reported to form dimers in

solution [13,19] there are several crystallographic structures of TIR domains in the

form of dimers. The structure of the TIR domain of TLR10 revealed that the BB loop

is key in mediating the interaction between monomers [12]. In contrast, the structure

of the TIR domain from Paracoccus denitrificans [20] revealed that residues from the

both DD and EE loops (Figure 1) mediate the interaction between monomers. In

addition, the crystal structures of the TIR domains of TLR 1 and 2 [10] reveal the

10

formation of dimers stabilized by disulphide bridges. Initially it was suggested that

these dimers were artifacts of the crystallization process however the subsequent data

from the crystal structures of the Cys713Ser mutant of the TLR2-TIR and IL-1RAPL

also revealed the formation of disulphide linkages potentially important in dimer

formation [27,28]. The data presented here on YpTIR, together with the known

structures, indicates that the dimeric form may be a common feature for TIR proteins.

Functional characterization of the GB1-YpTIR1 was carried out by 2D NMR

analysis using 15N-GB1-YpTIR1 in the presence and absence of unlabeled GB1-

MyD88-TIR. The observed peak broadening for certain residues of the YpTIR

domain indicates an interaction between the two proteins. This is supported by GST-

pull down assays demonstrating that the GB1-YpTIR2 specifically interacts with

human GST-MyD88-TIR. However NMR titration studies were complicated by the

fact that the YpTIR exists as a dimer, precluding facile resonance assignment for

chemical shift mapping. GST pull down analysis also revealed an interaction between

GST-MyD88-TIR and YpTIR4 following thrombin cleavage indicating that the

interaction is not mediated by the GB1 tag.

In conclusion we have generated a series of highly expressing soluble, stable

and folded YpTIR constructs. We have obtained functional YpTIR, which is uniquely

dimeric in solution compared to the known TIR domains. YpTIR is an excellent

candidate for further study of the mechanism of action of pathogenic bacterial Tdps.

4 Materials and Methods

4.1 Materials

11

Oligonucleotide primers were obtained from MWG Biotech, the GEV2 expression

vector was obtained from Addgene. Co2+-IMAC resin was obtained from Clontech

Laboratories. The Superdex 200 10/300 GL column, thrombin protease and ECF

reagent were purchased from GE Life Sciences. Protease inhibitor tablets were

obtained from Roche. -mercaptoethanol (BME) and anti-His antibodies from Sigma.

Anti-GST antibodies were obtained from Merck. The Bradford Protein Assay kit for

protein concentration determination and the Imperial Protein Stain for Coomassie

staining were purchased from Pierce. SDS-PAGE gels and BL21 (DE3) Star™ cells

were purchased from Invitrogen. Molecular weight cut-off (MWCO) filters were

purchased from Millipore. The QuikChange kit was obtained from Stratagene.

4.2 Generation of expression constructs

The oligonucleotide primers used for generating the different constructs are detailed

in Table 1 (need to add in the mutagenic oligos). In each case a 5’ XhoI site and a 3’

BamHI site were incorporated to the target gene sequence. The region of the YpTdp

containing the TIR domain (YpTIR1; residues S130 to A285) was estimated based on

sequence alignments with TIR domains of known structure (Figure 1). For this work,

the putative YpTdp gene in Y. pestis strain CO92 (YPO1883) was selected. Two

shorter YpTIR constructs were also generated, corresponding to residues I137-I273

(YpTIR2) and I137-N246 (YpTIR3) and a longer construct corresponding to residues

D107-S281 (YpTIR4). YpTIR3 lacks the region of the protein corresponding to the

Box 3 motif (Figure 1). All four constructs were cloned into GEV2 [26]. The region

of the human MyD88 gene encoding for the core TIR domain ([13], residues F163-

P296) was also cloned into GEV2. The GB1-YpTIR2 Cys90Ser, Cys132Ser and

double mutants were generated using the QuikChange mutagenesis kit according to

12

manufacturers instructions. The presence of the mutations was confirmed by DNA

sequencing and the proteins expressed and purified as described above. The GST-

MyD88 expression vector was generated by cloning the region of the human MyD88

gene encoding for the core TIR domain ([13], residues F163-P296) into pGEX5.

4.3 Expression and purification of YpTIR and MyD88 constructs

The YpTIR and MyD88-TIR constructs were individually expressed in BL21 (DE3)

Star™ E. coli cells using optimized induction conditions (1 mM IPTG at 25 °C for 4

hr) in 1 L cultures of LB. The cells were lysed by sonication and following

centrifugation to removal the insoluble material, the soluble fraction was applied to

Co2+-IMAC resin (1 ml/ 1 L prep) equilibrated with Buffer A [20 mM Tris-HCl, pH

7.0, 250 mM NaCl, 2 mM BME, protease inhibitors]. The protein was batch bound

for 1 hr at 4 C. The resin was washed with 10 CVs of Buffer A and the bound

proteins eluted with Buffer A supplemented with 150 mM imidazole. The eluted

protein was exchanged into Buffer B [20mM Tris-HCl, 150 mM NaCl, pH 7.0] on a

Superdex 200 10/300 GL column. Buffer B was supplemented with 10 mM DTT for

GB1-MyD88-TIR in order keep the sample in its reduced monomeric form. The GB1-

tagged proteins were concentrated to 10 mg/ml using a 10 kDa MWCO filter. GB1-

YpTIR4 was isolated as described above and the pure protein incubated with 2.5 units

of thrombin protease per 100 g for 16 hr at 18 °C. The digested protein was then

separated from the GB1 tag by a further Co2+-IMAC step as described above. All

proteins were analysed by SDS-PAGE by separation on a 4-12% Bis-Tris NuPAGE

gels. The protein bands were either directly visualised using Imperial Protein Stain or

detected by ECF reagent following transfer to an immobilized membrane and

incubation with either anti-His or anti-GST primary antibodies.

13

4.4 Protein concentration determination

Protein concentration was determined using the Bradford Protein Assay and the UV

absorption method [29]. Bovine serum albumin (BSA) was used as a protein standard

for the Bradford Assay.

4.5 Determination of the oligomeric state of YpTIR

The oligomeric state of the GB1-YpTIR constructs at a detected protein concentration

of 1.4 mg/ml was assessed with a Malvern Viscotek TDAmax Tetra detection system,

including static light scattering, UV and refractive index detectors, connected

downstream of a Superdex-200 10/30 gel-filtration column previously equilibrated in

Buffer B. The data was analysed with the Omnisec software (Malvern) following the

manufacturers protocols.

4.5 1D/2D NMR analysis of pure protein

10% D2O was added to all the unlabeled and labeled pure protein samples prior to

recording NMR spectra for spectrometer field lock. Experiments were recorded using

either a Bruker Avance III 600 (at 600 MHz) spectrometer at 303 K. Experiments

were performed using the TopSpin software package. For 2D NMR analysis GB1-

YpTIR1 was labeled with the 15N isotope by expressing the protein in minimal (M9)

media supplemented with 15N-labeled NH4Cl (0.07%). Expression and purification of

the labeled protein was as described for the unlabeled proteins. All samples were

prepared in 20mM Tris-HCl, 150 mM NaCl, 10 mM DTT, pH 7.0 for 2D NMR. 2D

14

spectra were processed with NMRPipe [30]. Spectral manipulation and layout was

attained using NMRView (One Moon Scientific).

5. Acknowledgements

We thank Dr Tom Monie and Prof Nick Gay, University of Cambridge for helpful

discussions and advice. This work was funded by the UK Ministry of Defence.

6. References

[1] Medzhitov R. Recognition of microoganisms and activation of the immune

response. Nature 2007; 449: 819-826.

[2] Medzhitov R, Preston-Hurlbert P, Janeway Jr. CA. A human homologue of the

Drosophila Toll protein signals activation of adaptive immunity. Nature 1997; 388:

394-7.

[3] Gay NJ, Keith FJ. Drosophila Toll and IL-1 receptor. Nature 1991; 351: 355-6.

[4] Triantafilou M, Gamper FG, Haston RM, Mouratis MA, Morath M, Hartung T et

al. Membrane sorting of toll-like receptor (TLR)-2/6 and TLR1/2 heterodimers at the

cell surface determines heterotypic associations with CD36 and intracellular targeting.

J. Biol. Chem. 2006; 281: 31002-11.

[5] Latz E, Verma A, Visintin A, Gong M, Sirois CM, Klein DCG, et al. Ligand-

induced conformational changes allosterically activate Toll-like receptor 9. Nat.

Immunol. 2007; 8: 772-9.

15

[6] Gay NJ, Gangloff M. Structure and function of Toll receptors and their ligands.

Annu. Rev. Bioch. 2007; 76: 141-65.

[7] O’Neill LAJ, Bowie AG. The family of five: TIR-domain-containing adaptors in

Toll-like receptor signaling, Nat. Rev. Immunol. 2007; 7: 353-364.

[8] Monie TP, Montcrieff MC, Gay NJ. Structure and regulation of cytoplasmic

adaptor proteins involved in innate immune signaling. Immunol. Rev. 2009; 227: 161-

175.

[9] Kumar H, Kawai T, Akira S. Pathogen recognition in the innate immune response.

Biochem. J. 2009; 420: 1-16.

[10] Xu Y, Tao X, Shen B, Horng T, Medzhitov R, Manley JL, et al. Structural basis

for signal transduction by the Toll/interleukin-1 receptor domains. Nature 2000; 408:

111-115.

[11] Poltorak A, He X, Smirnova I, Liu MY, Van Huffel C, Du X, et al. Defective

LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science

1998; 282: 2085-2088.

[12] Nyman T, Stenmark P, Flodin S, Johansson I, Hammarström M, Nordlund P The

crystal structure of the human toll-like receptor 10 cytoplasmic domain reveals a

putative signaling dimer. J. Biol. Chem. 2008; 283: 11861-11865.

[13] Ohnishi H, Tochiro H, Kato Z, Orii KE, Li A, Kimura T, et al. Structural basis

for the multiple interactions of the MyD88 TIR domain in TLR4 signaling. PNAS

2009; 106: 10260-10265.

16

[14] Newman RM, Salunkhe P, Godzik A, Reed JC. Identification and

characterization of a novel bacterial virulence factor that shares homology with

mammalian Toll/Interleukin-1 receptor family proteins. Infect. Immun. 2006; 74: 594-

601.

[15] Salcedo SP, Marchesini MI, Lelouard H, Fugier E, Jolly G, Balor S, et al

Brucella control of dendritic cell maturation is dependent on the TIR-containing

protein Btp1. PLoS Pathogens 2008; 4: e21.

[16] Cirl C, Wieser A, Yadav M, Duerr S, Schubert S, Fischer H, et al. Subversion of

Toll-like receptor signaling by a unique family of Toll/Interleukin-1 receptor domain-

containing proteins. Nat. Medicine 2008; 14: 399-406.

[17] Radhakrishnan GK, Splitter GA. Biochemical and functional analysis of TIR

domain containing protein from Brucella melitensis BBRC 2010; 397: 59-63.

[18] Spear AM, Loman NJ, Atkins HS, Pallen MJ Microbial TIR domains: agents of

subversion or good housekeepers? Trends Microbiol 2009; 17: 393-8.

[19] Low LJ, Mukasa T, Reed JC, Pascual J. Characterization of a TIR-like protein

from Paracoccus denitrificans. BBRC 2007; 356: 481-486.

[20] Chan SL, Low LY, Hsu S, Li S, Liu T, Santelli E, et al. Molecular Mimicry in

innate immunity: crystal structure of a bacterial TIR domain. J. Biol. Chem. 2009;

284: 21386-92.

[21] Perry RD, Fetherston JD. Yersinia pestis-etiological agent of plague. Clin.

Microbiol. Rev. 1997; 10, 35-66.

17

[22] Pham HV, Dang DT, Tran Minh NN, Nguyen ND, Nguyen TV. Correlates of

environmental factors and human plague: an ecological study in Vietnam. Int. J.

Epidemiol. 2009; 38:1634-41.

[23] Li Y, Cui Y, Hauck Y, Platonov ME, Dai E, Song Y, et al. Genotyping and

phylogenetic analysis of Yersinia pestis by MLVA: Insights into the Worldwide

Expansion of Central Asia Plague Foci. PLoS ONE 2009; 4: e6000.

[24] Han Y, Zhou D, Pang X, Zhang L, Song Y, Tong Z, et al Comparative

transcriptome analysis of Yersinia pestis in response to hyperosmotic and high-

salinity stress. Res. Microbiol. 2005; 156: 403-415.

[25] Han Y, Qiu J, Guo Z, Gao H, Song Y, Zhou D, Yang R. Comparative

transcriptomics in Yersinia pestis: a global view of environmental modulation of gene

expression. BMC Microbiol. 2007; 7: 96.

[26] Huth JR, Bewley CA, Jackson BM, Hinnebusch AG, Clore GM, Gronenborn

AM. Design of an expression system for detecting folded protein domains and maping

macromolecular interations by NMR. Prot. Sci. 1997; 6: 2359-64.

[27] Tao X, Xu Y, Zheng Y, Beg AA, Tong L. An extensively associated dimer in

the structure of the C713S mutant of the TIR domain of human TLR2. BBRC 2002;

299: 216-21.

[28] Khan JA, Brink EJ, O’Neill LA, Tong L. Crystal structure of the

Toll/interleukin-1 receptor domain of human Il-1RAPL. J Biol Chem 2004; 279:

31664-70.

[29] A Aitken, MP Learmonth. The Protein Protocols Handbook, ed J. Walker:

Human Press. Pp 3-6, 2002

18

[30] Delaglio F, Grzesiek S, Vuister GW, Zhu G, Pfeifer J, Bax A. NMRpipe: a

multidimensional spectral processing system based on UNIX. J Biomol NMR 1995;

6:277-93.

[31] Gouy M., Guindon S. & Gascuel O. SeaView version 4 : a multiplatform

graphical user interface for sequence alignment and phylogenetic tree building. Mol

Biol Evol 2010; 27: 221-224.

.

Figure legends

Figure 1 Alignment of YpTIR1 (residues S130-A285) with other TIR domains of

known structure. The regions corresponding to the TIR motifs, Box1, 2 and 3, are

denoted by the solid rectangles above the alignment. The regions of the sequence of

TLR1-TIR corresponding to secondary structure elements as determined by X-ray

crystallographic analysis (Pdb accession code: 1FYV) are indicated below the

alignment. The alignment was generated using ClustalW and Seaview [31].

Figure 2 (a) SEC of the GB1-YpTIR1, 2 and 3 constructs and (b) SDS-PAGE

analysis of the protein following SEC. M indicates molecular weight markers and

lanes 1, 2 and 3 contain GB1-YpTIR1, GB1-YpTIR2 and GB1-YpTIR3 respectively.

SEC (c) and SDS-PAGE analysis (d) of the GB1-YpTIR4 before (lane 4) and after

(lane 5) thrombin cleavage.

Figure 3 1D NMR analysis of (a) GB1-YpTIR1 (b) GB1-YpTIR2 (c) GB1-YpTIR4

and (d) YpTIR4 (after thrombin cleavage). The arrows indicate the peaks

corresponding to folded YpTIR protein.

19

Figure 4 Representative static light scattering measurements of GB1-YpTIR1 and

cleaved YpTIR4. The data from all constructs tested; GB1-YpTIR1, GB1-YpTIR2,

GB1-YpTIR4 and cleaved YpTIR4, indicate the presence of dimers (table inset). The

measurements from the refractive index (RI) detector for GB1-YpTIR1 and cleaved

YpTIR4 are indicated in dotted and dashed lines respectively. The data for the

measured weight-average molecular weights for GB1-YpTIR1 and cleaved YpTIR4

are shown in the solid and broken lines respectively. The measured molecular weights

for GB1-YpTIR1 and cleaved YpTIR4 are 50.7 and 40.9 kDa respectively. The

predicted molecular weights of the monomeric forms of GB1-YpTIR1 and cleaved

YpTIR4 are 25.3 and 20.8 kDa respectively.

Figure 5 Analysis of the interaction between YpTIR and MyD88 using 2D (1H/15N)

HSQC NMR spectroscopy and GST pull down. The 2D (1H/15N) TROSY-type HSQC

spectra of 15N-GB1-YpTIR1 were recorded (a) in the absence and (b) presence of 1 M

equivalent of unlabeled GB1-MyD88-TIR. Peaks arising from GB1 are boxed (as

judged from the spectrum of the free domain, data not shown) and some structured

amide peaks from GB1-YpTIR1, which are broadened upon GB1-MyD88-TIR

addition are indicated with arrows. Western blot analysis of samples obtained from

GST pull down experiments revealed (c-d) a positive and specific interaction of GB1-

YpTIR2-His and cleaved GB1-YpTIR4-His with GST-MyD88-TIR and (c,e) a lack of

specific interaction of GB1-YpTIR4-His and the C132S mutant form of GB1-

YpTIR2-His with GST-MyD88-TIR. GST-MyD88-TIR in the eluate fraction from

one of the GST pull down experiments with GB1-YpTIR2-His is indicated in (f).

Samples were probed with (c-e) anti-His antibody in order to detect the wild-type and

mutant forms of GB1-YpTIR-His and (f) anti-GST antibody in order to detect GST-

MyD88-TIR.

20

Figure 1

Figure 2

Figure 3

Figure 4

21

Figure 5

22

23