Functional cell permeable motifs within medically relevantproteins

Walter Lowa,⁎, Alison Mortlocka, Liljana Petrovskab, Tania Dottorinic, Gordon Dougand, andAndrea CrisantiaaBiological Sciences, Imperial College London, Imperial College Road, 5th floor SAF Building,London SW7 2AZ, UK.bCentre for Molecular, Microbiology and Infection, Imperial College London, London SW7 2AZ, UK.cDepartment of Experimental Medicine and Biochemical Science, University of Perugia, Perugia06122, Italy.dThe Wellcome Trust Sanger Institute, Wellcome Trust Genome Campus, Hinxton, CambridgeCB10 1SA, UK.

AbstractIncreasing experimental evidence indicates that short polybasic peptides are able to translocate acrossthe membrane of living cells. However, these peptides, often derived from viruses and insects, mayinduce unspecific effects that could mask the action of their cargoes. Here, we show that a panel oflysine and/or arginine-rich peptides, derived from human proteins involved in cell signallingpathways leading to inflammation, possess the intrinsic ability to cross intact cellular membranes.These peptides are also capable of carrying a biologically active cargo. One of these peptides,encompassing the cell permeable sequence of the Toll-receptor 4 (TLR4) adaptor protein (TIRAP)and modified to carry a dominant-negative domain of the same TIRAP protein, selectively inhibitedthe production of pro-inflammatory cytokines upon LPS challenge, in in vitro, ex vivo and in vivoexperiments. Docking studies indicated that this inhibition might be mediated by the disruption ofthe recruitment of downstream effector molecules. These results show for the first time the potentialof using for therapy cell permeable peptides derived from human proteins involved in disease.

KeywordsToll-receptor 4 (TLR4); Toll-interleukin 1 receptor domain-containing adaptor protein (TIRAP);Protein transduction domains; TAT; Antennapedia homeodomain

1 IntroductionThe identification and characterisation of cell permeable peptides, or protein transductiondomains (PTDs), has generated considerable interest, as they offer the opportunity to transport

© 2007 Elsevier B.V.This document may be redistributed and reused, subject to certain conditions.

⁎Corresponding author. Tel.: +44 207 594 1651; fax: +44 207 584 2056. w.low@imperial.ac.uk.This document was posted here by permission of the publisher. At the time of deposit, it included all changes made during peer review,copyediting, and publishing. The U.S. National Library of Medicine is responsible for all links within the document and for incorporatingany publisher-supplied amendments or retractions issued subsequently. The published journal article, guaranteed to be such by Elsevier,is available for free, on ScienceDirect.

Sponsored document fromJournal of Biotechnology

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

cargo polypeptides across cell membranes for a variety of applications. To date, PTDs havebeen used to deliver a number of molecules in vivo and in vitro, including enzymes, signallingmolecules and apoptotic proteins, with the aim to dissect intracellular signalling pathways,regulate gene transcription, manipulate inflammatory responses and inhibit tumour progression(Lindsay, 2002; Christiaens et al., 2004; Wadia and Dowdy, 2005; Gondeau et al., 2005;Bernatchez et al., 2005; Toshchakov et al., 2005).

Among the inflammatory responses, those due to activation of Toll-like receptors (TLRs) havebeen intensively studied. A number of blocking peptides have been employed in vitro todelineate signal transduction pathways downstream of TLR4 and TLR2 (Horng et al., 2001;Toshchakov et al., 2005). Exposure of TLR4 to microbial components such aslipopolysaccharide (LPS) from Gram-negative bacteria triggers an intracellular signal cascadeleading to the secretion of pro-inflammatory mediators followed by an immune response. TLRsshare homologous cytoplasmic domains that are termed the Toll-Interleukin 1 receptor (TIR)homology domains. TIR domain-containing adaptors such as MyD88 (Medzhitov et al.,1998) and the TIR domain-containing adaptor protein (TIRAP) (Horng et al., 2001), alsotermed MyD88 adaptor-like (MAL) (Fitzgerald et al., 2001), couple TLR4 receptor activationto downstream signalling. These molecules are potential targets for therapeutic interventionand therefore it is anticipated that antagonists of TLR4 signal transduction might be able tomodulate harmful pro-inflammatory responses such as those observed in septic shock or severesepsis.

TIRAP was first identified by Horng et al. (2001) and its role in TLR4 signalling wasinvestigated using a cell-permeable blocking peptide, encompassing a translocating sequenceof the antennapedia homeodomain (Joliot et al., 1991; Derossi et al., 1994) covalently linkedto a TIRAP dominant negative domain. This domain, termed the BB loop (Xu et al., 2000), ishighly conserved among TIR domains. Horng et al. (2001) speculated that this peptide mightact as a specific inhibitor of TIRAP by competing with TIRAP for interaction with TLR4.However, conflicting results obtained from studies carried out with macrophages from TIRAPknock-out mice (Yamamoto et al., 2002; Horng et al., 2002) have raised the possibility thatinhibition with the TIRAP blocking-peptides may have been caused by the unspecific bindingof the antennapedia transduction domain.

In addition to antennapedia, other functionally and structurally unrelated peptides, includingTAT from human immunodeficiency virus (Green and Loewenstein, 1988; Frankel and Pabo,1988) and VP22 from herpes simplex virus (Elliott and O’Hare, 1997), have been shown tocross intact cellular membranes. However, recent evidence suggests that, as in the case ofantennapedia mentioned above, when used as carriers, these transducing peptides can causeunwanted effects that may either counteract the action of their cargoes or unspecificallyenhance it (Fotin-Mleczek et al., 2005). Interestingly, in all cases, the ability to translocatemaps to short amino acid sequences lacking in apparent sequence homology but characterisedby a high content of basic amino acids such as arginine and lysine (Bogoyevitch et al., 2002;Vives, 2003). Substitution of arginine or lysine residues in TAT and antennapedia with alanineswas shown to result in a marked decrease in peptide internalisation (Wender et al., 2000;Gratton et al., 2003).

In order to avoid the unspecific effects produced by ‘exogenous’ carriers such as Antenapedia,VP22 and TAT, we have investigated the cell-permeability properties of lysine and arginine-rich domains of human proteins which are known to be intimately involved in cell signallingpathways inherent to inflammation. A series of synthetic peptides was tested in vitro and shownto possess the intrinsic ability to move across cellular membranes. One of these peptides,derived from the TIRAP protein, was then fused to a cargo encompassing a dominant negativedomain of the same protein. This fusion peptide, named TBX2, selectively delayed the kinetics

Low et al. Page 2

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

of p38 activation following exposure to LPS in a fashion similar to that observed in cells fromTIRAP knock-out mice. Furthermore, TBX2 selectively inhibited pro-inflammatory cytokineproduction upon LPS challenge in vivo. Docking analysis suggested that TBX2 may disruptthe heterotrimeric complex formation between TLR4, MyD88 and TIRAP by binding TLR4at different non-overlapping sites. These results provide the rationale for the development ofhighly specific peptides for interfering with signal transduction and may pave the way to thedevelopment of novel tools for the treatment of human diseases such as sepsis syndrome andchronic inflammatory diseases.

2 Materials and methods2.1 Database analysis

The EMBOSS (http://www.emboss.org) program fuzzpro was used to search SWISS-PROTfor the following motifs [K/R]-X-[K/R]-[K/R]-[K/R], [K/R]-[K/R]-X-[K/R]-[K/R] and [K/R]-[K/R]-[K/R]-X-[K/R], where K is lysine, R is arginine and X is any amino acid. The resultinglists were parsed in Perl to produce a non-redundant list of entries.

2.2 Model building and computational methodsInitially, TBX2 secondary structure prediction was obtained using the PSIPRED v2.4(Mcguffin et al., 2000) web-interfaced facilities [http://bioinf.cs.ucl.ac.uk/psipred]. The 3Dmodel structure was constructed using the protein building module of Sybyl (version 7.0), andchecked for atom and/or bond type correctness using the Sketch module of Sybyl. Hydrogenatoms were added using standard Sybyl geometries and minimized with Powell algorithm witha gradient of 1 kcal/(mol Å) and 1000 cycles to remove potential bad contacts. The peptide hasbeen refined using RAPPER (Depristo et al., 2005). The final 3D structure present in thecomplex has been minimized using the Amber force field present in the compute module ofSybyl7.0.

2.3 Homology modellingSequence homologues were retrieved using PSI-BLAST (Altschul et al., 1997). Swiss-pdb(Guex and Peitsch, 1997) CPHModels2.0 server were used to generate 3D models of eachprotein. The models of the proteins have been obtained according to the PDB crystal structurecoordinates of 1077 pdb (Tao et al., 2002). Evaluation of obtained structure values have beenperformed with the software WHAT IF http://biotech.embl-heidelberg.de:8400/) andRAMAPAGE (http://www-cryst.bioc.cam.ac.uk/rampage) and have been compared to thereference PDB. Hydrogen atoms were added to the models using standard Sybyl geometries,and minimized with the Powell algorithm with a gradient of 0.5 kcal/(mol Å) and 1000 cyclesto remove potential bad contacts. Models have been analysed using the protein preparation toolin sybyl7.0.

2.4 Protein–protein, protein–peptide docking and HINT softwareThe models of complexes between TLR4 and the adaptor proteins, Tirap and Myd88, wereconstructed using the ClusPro Algorithm (Comeau et al., 2004). The HINT software (Kellogget al., 1991) was used to predict protein–ligand, protein–protein and macrocycle–macrocycleassociations by assessing hydrophobic, hydrogen bonds and polar interactions between the twointeracting macromolecules. Protein–protein and protein–peptide complexes were minimizedusing the amber force field present is sybyl7.0.

Low et al. Page 3

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

2.5 Peptide synthesisPeptides were produced by solid phase step-wise synthesis (SPSS) using the Fmoc N-terminalprotection strategy. All peptides were made with C terminal carboxamide functionality. Chainassembly was performed on Applied Biosystems 433 or Pioneer automated synthesisers.

N terminal biotinylation was performed on the synthesiser through activation of the biotincarboxylic acid in the same way as for amino acids.

4-Chloro-7-nitrobenzofurazan (NBD) was attached to the peptide N terminal amino groupusing a manual method. This was achieved with high efficiency by incubating the N terminaldeprotected peptide-resin with a five fold excess of NBD with two fold excessdiisopropylethylamine (DIEA) in dimethylformaide (DMF) for 18 h with gentle mixing.

All peptide products were analysed by reverse phase HPLC or CZE as appropriate. Purificationwas performed by reverse phase HPLC. Confirmation of peptide molecular weight wasperformed by MALDI. All the peptides were solubilized in PBS.

Peptide sequences: TAT, GRKKRRQRRRPPQ; TATAla, GRKKAAQAAAPPQ; IRAK1,CLHRRAKRRPPMTQVYER; TLR4, GRHIFWRRLRKALLDGKSWNPE; TIRAP,GKMADWFRQTLLKKPKKRPNSPEST; TIRAPAla,GKMADWFRQTLLAAPAAAPNSPEST; BX2, LQLRDATPGGAIVS; TBX2,GKMADWFRQTLLKKPKKRPNSPESTLQLRDATPGGAIVS; Antp-TIRAP,RQIKIWFQNRRMKWKKLQLRDAAPGGAIVS; PU.l,GSKKKIRLYQFLLDLLRSGDMKDS; ICE, QLLRKKRRIFIHSVGAGT.

2.6 Peptide internalisation assaysHeLa or RAW 264.7 cells were cultured at 37 °C, 10% CO2, in Dulbecco's modified Eagle'smedium (DMEM) (Invitrogen), containing 10% foetal calf serum (FCS, Harlan), 100 U/mlpenicillin and 100 μg/ml streptomycin. Quantification of internalisation of NBD-labelledpeptides was carried out according to the procedure described by Drin et al. (2001). Briefly,cells were seeded in six-well plates (5 × 105 cells/well) and treated with 0–160 μM of peptidefor 2 h at 37 °C. Cells were washed twice in PBS, harvested using a solution of Versene (1:500)(Invitrogen) and re-suspended in 0.5 ml (final volume) of ice-cold PBS. Cells were analysedby flow cytometry using a Becton Dickinson FACS Scan instrument and the geometric meanof fluorescence intensity was considered to represent the amount of cell-associated peptide.Following flow cytometry, 5 μl of freshly prepared dithionite stock solution (1 M Na2S2O4 in1 M Tris–HCl (pH 10)) was added to cells maintained at 4 °C for 5 min and analysed againwith flow cytometry to measure residual fluorescence. Fluorescence from cells treated withdithionite only was subtracted from the fluorescence of cell-associated or internalised peptidesto infer background values.

2.7 Immunoblot analysisRAW 264.7 cells were treated with PBS or 160 μM TBX2 peptide for 2 h at 37 °C, 10%CO2, in DMEM, containing 10% foetal calf serum and antibiotics. Cells were stimulated with10 ng/ml of LPS and samples were collected after 0, 10, 20 and 45 min. Cells were lysed insample buffer and separated by 12% SDS-PAGE and transferred onto nitrocellulose membrane(Amersham). Membranes were blocked for 1 h in 5% dry milk in PBS and 0.1% Tween 20 atRT, and then probed with an anti-ACTIVE p38 (Promega) antibody for 1 h. After washingwith PBS and 0.1% Tween 20, the primary antibody was detected using a goat anti-rabbithorseradish peroxidase conjugate (1:2000, Jackson Labs). To control that different sampleswere loaded in equal amounts, the membranes were also incubated with the mouse monoclonalanti-pan phospho-tyrosine antibody 4G10 (a gift from X. Montano) followed by a goat-anti

Low et al. Page 4

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

mouse Horse Radish Peroxidase (HRPO) antibody (1:2000, Jackson Labs). Detection of HRPOantibodies was performed using an enhanced chemiluminescence detection system (ECL,Amersham).

2.8 Cytokine analysisRAW 264.7 cells were treated with different concentrations of peptides for 2 h at 37 °C, 10%CO2, in DMEM, containing 10% foetal calf serum and antibiotics and then stimulated with10 ng/ml LPS (E. coli 055:B5, Sigma), 25 μg/ml Poly I:C (InvivoGen) or 1 μg/ml R848(InvivoGen). Supernatants were collected 22 h post agonist challenge and IL-6 levelsquantified by ELISA, according to the manufacturer's instructions (R&D Systems). Primaryhuman macrophages, isolated from a single donors buffy coats, were treated with differentconcentrations of the TBX2 or the BX2 peptides for 2 h at 37 °C, 10% CO2, in DMEM,containing 10% autologous serum and antibiotics and then stimulated with 1 ng/ml LPS for18 h. The levels of TNF-α present in the supernatants were quantified by ELISA, according tothe manufacturer's instructions (Pharmingen).

2.9 Animal experimentsGroups of five female mice (strain C3H/HeN) were injected intravenously (I.V.) with a singledose of TBX2 peptide (2, 4 or 6 mg per mouse) or PBS only. After 45 min, the mice werechallenged with 10 μg of LPS injected i.v. 90 min later, the mice were anaesthetised and bloodcollected by cardiac puncture. The levels of TNF-α in the serum were quantified by ELISAaccording to the manufacturer's instructions (Pharmingen). All animal experiments wereperformed in compliance with institutionally approved protocols at Imperial College.

2.10 Cell viabilityCellTiter-Glo Luminescent Assay (Promega) was utilised to determine viability of RAW264.7 cells following treatment with peptides, according to the manufacturer's instructions.

2.11 Statistical analysisStandard deviation (S.D.) and standard error of the mean (S.E.M.) were calculated usingMicrosoft Excel. Statistical analysis was carried out using Student's t-test.

3 Results3.1 Peptides derived from human proteins and containing cationic motifs are capable ofcrossing the membrane of living cells

In addition to the well characterised polybasic PTDs (Frankel and Pabo, 1988; Green andLoewenstein, 1988; Joliot et al., 1991; Derossi et al., 1994; Elliott and O’Hare, 1997), a nuclearlocalisation signal (NLS) derived from SV40 and containing four lysines and one arginine(Peitz et al., 2002), and a peptide derived from the prion protein (PrP) containing the sequenceKKRPK (Lundberg et al., 2002), were shown to be able to translocate across cell membranes.On the basis of these observations, we formulated the hypothesis that short stretches of cationicamino acids could confer cell permeable properties to certain human peptides. We thereforescreened the human proteome databases for proteins containing motifs encompassing fivecharged residues of either arginine or lysine allowing for a single mismatch at positions 2, 3or 4. A search in the Swiss-Prot protein database (http://us.expasy.org/sprot/) identified a largenumber of non-redundant, non-hypothetical human proteins known to function as kinases,transcription factors, proto-oncogenes, receptors and DNA repair enzymes. To test ourhypothesis, focusing on cell signalling proteins linked to immune responses and inflammation,five peptides derived from proteins linked to pathological conditions and/or to cellularsignalling pathways were synthesised and their ability to function as PTDs assessed. These

Low et al. Page 5

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

included the human Toll-interleukin 1 receptor (TIR) domain-containing adapter protein(TIRAP) (Horng et al., 2001) (alternatively known as MyD88-adapter-like (Mal); Fitzgeraldet al., 2001), the Interleukin-1 receptor-associated kinase 1 (IRAK1), the toll-like receptor 4(TLR4), the transcription factor PU.1 (PU.1) and the caspase-1 inhibitor Iceberg (ICE). Thepeptides were designed to encompass the cationic motif flanked by endogenous sequences ofup to 20 amino acids, and linked to a 7-nitrobenz-2-oxo-l,3-diazol-4-yl (NBD) at the aminoterminal end (Drin et al., 2001). This fluorophore is irreversibly quenched by the addition ofcell impermeable sodium dithionite, thus enabling discrimination of peptides merely bound tothe cells from truly intracellular ones. Moreover, its use allows a direct measurement of theamount of internalised peptides in live cells, therefore avoiding artefacts due to fixation(Langeduk et al., 2004; Drin et al., 2003; Christiaens et al., 2004). Cultured epithelial HeLacells were incubated with NBD-labelled peptides at three concentrations either in the presenceor in the absence of sodium dithionite. Cultured cells were also treated with a TAT peptide anda mutated, TATAla in which five arginine residues were substituted by alanines, to serve aspositive and negative controls, respectively. Peptide internalisation was deduced byquantifying cell fluorescence by FACS analysis. All tested peptides gave fluorescence valuessimilar or higher than TAT when added to HeLa cells in the presence of sodium dithionite,whereas very little fluorescence was observed with TATAla at all concentrations tested. Thefluorescent signal increased in a peptide concentration-dependent manner (Fig. 1A), inagreement with previous reports on TAT and Antp (Drin et al., 2003; Derossi et al., 1996;Oehlke et al., 1997; Hallbrink et al., 2001). The specificity of the uptake was investigated byincubating both murine macrophage-derived RAW 264.7 and HeLa cells either with the NBD-labelled TIRAP peptide, or with a mutated version in which the basic residues forming thecationic motif were replaced with alanines (TIRAPAla) (Fig. 1B). Cells treated with NBD-TIRAPAla were substantially less fluorescent than those treated with the original TIRAPpeptide, indicating that the disruption of the cationic motif abolished the ability of the peptideto cross the cell membrane.

3.2 The human-derived cell permeable peptides can carry a biologically active cargoTo assess the biochemical and biological activity of peptides containing putative PTDs in vitroand in vivo, we combined in a synthetic peptide the functional cell permeable sequence ofTIRAP with a dominant negative domain from the same protein. This peptide, named TBX2,therefore encompassed the sequence of TIRAP shown to have cell permeable properties inHeLa and RAW 264.7 cells (Fig. 1B) (amino-acids 19–43) and a region of the BOX2 domain(Fitzgerald et al., 2001) (amino-acids 118–131), also known as the BB-loop (Xu et al., 2000).This region is highly conserved among TIR domains and forms the core of a dominant negativedomain by competing the interaction of endogenous TIRAP with the cytoplasmic tail of Tollreceptor 4 (TLR4) (Horng et al., 2001). To investigate whether the TIRAP dominant negativedomain linked to the cell permeable sequence present in TBX2 was able to inhibit TLR4activity, RAW 264.7 cells were treated with LPS and increasing concentrations of TBX2. Theeffect of TBX2 on the TIRAP-dependent activation of p38, a member of the mitogen activatedprotein kinase family, was assessed by immunoblot analysis using an antibody that selectivelyrecognises the phosphorylated, active form of p38. As shown in Fig. 2, TBX2 caused a temporalshift in the appearance of the peak of phosphorylated p38 in RAW 264.7 cells upon LPStreatment. These results are in agreement with the observed kinetics of p38 phosphorylationin isolated peritoneal macrophages (Yamamoto et al., 2002) and bone marrow-derivedmacrophages (Horng et al., 2002) from TIRAP knock-out mice in response to LPS. Notably,the treatment with the Antp-TIRAP peptide, encompassing the PTD of Antp from D.melanogaster and the BB loop sequence derived from the mouse TIRAP gene (Horng et al.,2001), resulted instead in complete inhibition of p38 phosphorylation in response to LPS (Fig.2). Previous reports had shown that the BB-loop, when fused to the PTD of Antennapedia,could inhibit the activation of MAP kinase JNK and NF-κB in RAW 264.7 cells upon LPS

Low et al. Page 6

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

challenge (Horng et al., 2001). While TBX2 reproduced the p38 activation pattern observedin TIRAP knock-out mice, Antp-TIRAP did not.

3.3 Suppression of cytokine production following transduction of the TBX2 peptide in vitro,ex vivo and in vivo

TIRAP-deficient mice show an impaired production of the inflammatory cytokine IL-6 inresponse to LPS (Yamamoto et al., 2002; Horng et al., 2002). Consistently, incubation of RAW264.7 cells with TBX2 resulted in dose-dependent inhibition of IL-6 production in responseto LPS (Fig. 3A).

To assess the specificity of TBX2-mediated inhibition on the TLR4-TIRAP signallingpathway, RAW 264.7 cells were treated with either Poly (I:C) (a ligand of TLR3) or thesynthetic compound R-848 (a ligand of TLR7) (Fig. 3A). In either case, TBX2 did not inhibitIL-6 secretion. Furthermore, no significant reduction in cellular viability was observedfollowing treatment with increasing concentrations of TBX2 over an incubation period of upto 18 h (Fig. 3B).

We also assessed the ability of TBX2 to inhibit the production of TNF-α, a key mediator ofchronic inflammatory diseases, in freshly isolated human macrophages. Our results indicatedthat at a 50 μM concentration of TBX2, TNF-α production from human macrophages wasreduced by 50% in response to LPS (Fig. 4A). A 14 amino-acid peptide, BX2, encompassingonly the TIRAP dominant negative sequence (Horng et al., 2001) and lacking the cell permeabledomain, was used as control. As expected, primary human macrophages exposed to LPSproduced similar levels of TNF-α either in the presence or absence of the BX2 peptide (Fig.4A).

To determine whether TBX2 had the ability to suppress the expression of inflammatorycytokine genes in vivo, a single dose of the peptide was administered intravenously to mice45 min prior to challenge with LPS. The levels of TNF-α, measured in ELISA on whole miceblood, were significantly reduced (over 60%) in mice treated with 2 mg of the TBX2 peptidewhen compared with control mice (Fig. 4B). A low level of IL-10 secretion was observedwhich did not differ significantly from animals treated with LPS only to animals treated withLPS and TBX2 (not shown).

3.4 TBX2 might interfer with the heterotrimeric complex formation between TLR4, MyD88and TIRAP

To obtain a better insight into the nature of TBX2 inhibition of p38 phosphorylation in responseto LPS, we have modelled the TIR domains of TLR4, MyD88 and TIRAP proteins. The TIRAPand MyD88 adaptor proteins were docked onto the TLR4 using Cluspro (Comeau et al.,2004). Protein–protein complexes (TLR4-TIRAP and TLR4-MyD88) were superimposed tovisualize the potential ternary complex associations. The models obtained were evaluated usingthe available biological information and post-processed using the HINT software (Kellogg etal., 1991). This software predicts protein–ligand, protein–protein and macrocycle–macrocycleassociations by assessing hydrophobic, hydrogen bonds and polar interactions between the twointeracting macromolecules. Our results suggest that TIRAP and MyD88 bind to TLR4 in twonon-overlapping binding sites (Fig. 5). In this model, TIRAP is predicted to associate throughits BB-loop to TLR4, whereas MyD88 is predicted to interact with the region enclosing theTLR4 BB-loop. The P714H mutation on TLR4 has been shown to disrupt MyD88/TLR4complex formation (Poltorak et al., 1998) suggesting that this residue could be involved in thebinding of Myd88. This is consistent with our model in which the BB-loop of TLR4 is directedtowards MyD88. A corresponding proline residue is also found in the BB-loops of TIRAP andMyD88. Mutation of this proline (P125H) to histidine on the TIRAP protein was proposed

Low et al. Page 7

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

(Horng et al., 2001) to disrupt TIRAP-TLR4 association. Accordingly, in our TLR4-TIRAPinteraction model, mutation of this proline (P125H) to histidine could result in hydrophobicclashes. In addition, Dunne et al. (2003) showed that the MyD88 P200 mutation has no effecton TLR4-MyD88 association. Again, this is consistent with our model in which the BB-loopof MyD88 is projected on the opposite side of TLR4.

We then considered this ternary complex model as a basis to determine whether TBX2 couldoccupy crucial sites involved in protein–protein interaction with TLR4. Thus, we performedpeptide–protein docking (TBX2-TLR4) analysis using TBX2 as ligand and TLR4, as receptor.The most likely TBX2-TLR4 binding complex was selected using the HINT software (Kellogget al., 1991), available biological information and our own results. In our model, TBX2 couldassociate through its amino and carboxyl terminus with TLR4 potentially hindering MyD88and TIRAP binding to TLR4 (Fig. 5).

4 DiscussionThe efficient and transient intracellular delivery of peptides and polypeptides, collectivelynamed PTDs, allows for reversible regulation of biological processes. As proof of principle, ithas been shown that PTDs can serve several uses in therapy. A peptide encompassing NFAT(nuclear factor of activated T cells) inhibitor linked to the cell-permeable sequence Arg10 wasused as an immunosuppressive agent, to overcome the undesired toxic effects of calcineurininhibitors in transplant therapy, enabling successful allogeneic islet transplantation in mice(Noguchi et al., 2004). More recently, a c-Jun N-terminal kinase 1 (JNK1) inhibitory peptidecovalently linked at its C terminus to a 10-amino-acid carrier peptide derived from the HIV-TAT sequence, has been shown to improve insulin resistance and glucose tolerance in diabeticmice (Kaneto et al., 2004). In this report, we show that synthetic peptides derived from humanproteins of medical interest and containing cationic motifs have the intrinsic ability totranslocate across intact extracellular membranes. Similarly to known transduction domainssuch as TAT, VP22 and Antp, the selected peptides show different extents of cell permeableproperties, combined with the ability to function as cargo carriers. In agreement with previouspublished data, such translocation properties are strongly dependent on arginine and lysineresidues (Mi et al., 2000; Park et al., 2002; Mai et al., 2002; Noguchi et al., 2005). Importantly,all peptides chosen for the analysis were capable of entering cells, indicating that although theavailable collection of five functional transduction peptides identified here is not large enoughto rigorously prove our hypothesis, the presence of stretches of arginine and lysine residues isa good marker to predict, before experimental verification, which sequence may function as aPTD. It should to be noted that the choice of the flanking sequences may also affect individualactivity.

In addition, to demonstrate intracellular delivery of biologically active polypeptides and todisclose the general therapeutic and technological applications of human peptides containingendogenous cell-permeable motifs, we used an inhibitory peptide entirely derived from theTIRAP adaptor molecule to interfere with the functionality of the TLR4 pathway, as measuredby a delay in p38 activation and by the inhibition of cytokine production in LPS challengedcells and mice. Notably, while TBX2 closely reproduced the p38 activation pattern observedin TIRAP knock-out mice, the Antp-TIRAP peptide did not, causing a total inhibition of p38activation. Although, we cannot rule out that the single amino acid substitution (T- > A) onthe TIRAP portion of TBX2 and Antp-TIRAP peptides may contribute to p38 activation, ourfindings seem to indicate that the presence of the PTD derived from the Drosophila Antphomeodomain induces unwanted biological effects, confirming what had been previouslysuggested (Toshchakov et al., 2005; Sabroe et al., 2003; Vogel and Fenton, 2003; O’Neill,2003). The toxicity and immunogenicity of several well known PTDs has been discussed indetail (Trehin R and Mrkle HP 2004). Interestingly, a recent study demonstrated that

Low et al. Page 8

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

translocating peptides such as Antp, TAT and nona-arginine (R9) can all induce effects thatantagonize or enhance the effects of the cargo being delivered (Fotin-Mleczek et al., 2005).Our results, confirming those obtained with TIRAP knock-out mice, indicate that endogenoustranslocating peptides such as TBX2 exert a specific function, avoiding unwanted effects andtoxicity.

The observation that a fraction of human proteins may contain putative PTDs raises theintriguing possibility that certain proteins may possess the ability to passively translocatethrough cell membranes, as shown for the human homeoprotein HOXB4 (Amsellem et al.,2003) and the BETA2/NeuroD Protein (Noguchi et al., 2005). Such a notion may also providea theoretical framework for understanding pathological effects associated with necrosis (asopposed to apoptosis) that commonly occur in tumours, degenerative diseases and chronicinflammatory conditions where cell permeable proteins released by damaged cells could exertan epigenetic pathological effect on the neighbouring ones.

In conclusion, the identification of cell permeable peptides originating from human proteinslinked to pathological conditions, offers therapeutical opportunities without the need to useexogenous, often deleterious, PTD sequences to facilitate cellular uptake. This would have theadvantage of minimize protein structure artefacts and unwanted biological effects due to theinsertion of foreign sequences. These proteins, either in recombinant forms or as syntheticpeptides encompassing relevant sequences, may be utilized to regulate protein–protein orprotein–DNA interactions, to transiently correct or down regulate abnormal gene function, tointerfere with signalling pathways, cell proliferation and to restore genetic defects, thereforefacilitating the exploitation of PTD technology in therapy and basic research.

AcknowledgmentsWe would like to express our gratitude to Ian Moss from the Advanced Biotechnology Centre at Imperial College forpeptide synthesis, Aaron Rae from the IC Central Services for FACS analysis and Derek Huntley from the Centre forBioinformatics at IC for his assistance. We would also like to thank Flaminia Catteruccia and Yasuhiro Ikeda forcritical reading of the manuscript and helpful suggestions.

ReferencesAltschul S.F. Madden T.L. Schaffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Gapped BLAST and

PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res.1997;25:3389–3402. [PubMed: 9254694]

Amsellem S. Pflumio F. Bardinet D. Izac B. Charneau P. Romeo P.H. Dubart-Kupperschmitt A. FichelsonS. Ex vivo expansion of human hematopoietic stem cells by direct delivery of the HOXB4homeoprotein. Nat. Med. 2003;9:1423–1427. [PubMed: 14578882]

Bernatchez P.N. Bauer P.M. Yu J. Prendergast J.S. He P. Sessa W.C. Dissecting the molecular controlof endothelial NO synthase by caveolin-1 using cell-permeable peptides. Proc. Natl. Acad. Sci. U.S.A.2005;102:761–766. [PubMed: 15637154]

Bogoyevitch M.A. Kendrick T.S. Ng D.C. Barr R.K. Taking the cell by stealth or storm? Proteintransduction domains (PTDs) as versatile vectors for delivery. DNA Cell Biol. 2002;21:879–894.[PubMed: 12573048]

Christiaens B. Grooten J. Reusens M. Joliot A. Goethals M. Vandekerckhove J. Prochiantz A. RosseneuM. Membrane interaction and cellular internalization of penetratin peptides. Eur. J. Biochem.2004;271:1187–1197. [PubMed: 15009197]

Comeau S.R. Gatchell D.W. Vajda S. Camacho C.J. ClusPro: an automated docking and discriminationmethod for the prediction of protein complexes. Bioinformatics 2004;20:45–50. [PubMed: 14693807]

Depristo M.A. De Barker P.I. Johnson R.J. Blundell T.L. Crystallographic refinement by knowledge-based exploration of complex energy landscapes. Structure (Camb) 2005;13:1311–1319. [PubMed:16154088]

Low et al. Page 9

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Derossi D. Calvet S. Trembleau A. Brunissen A. Chassaing G. Prochiantz A. Cell internalization of thethird helix of the Antennapedia homeodomain is receptor-independent. J. Biol. Chem.1996;271:18188–18193. [PubMed: 8663410]

Derossi D. Joliot A.H. Chassaing G. Prochiantz A. The third helix of the Antennapedia homeodomaintranslocates through biological membranes. J. Biol. Chem. 1994;269:10444–10450. [PubMed:8144628]

Drin G. Cottin S. Blanc E. Rees A.R. Temsamani J. Studies on the internalization mechanism of cationiccell-penetrating peptides. J. Biol. Chem. 2003;278:31192–31201. [PubMed: 12783857]

Drin G. Mazel M. Clair P. Mathieu D. Kaczorek M. Temsamani J. Physico-chemical requirements forcellular uptake of pAntp peptide. Role of lipid-binding affinity. Eur. J. Biochem. 2001;268:1304–1314. [PubMed: 11231282]

Dunne A. Ejdeback M. Ludidi P.L. O’Neill L.A. Gay N.J. Structural complementarity of Toll/interleukin-1 receptor domains in Toll-like receptors and the adaptors Mal and MyD88. J. Biol. Chem.2003;278:41443–41451. [PubMed: 12888566]

Elliott G. O’Hare P. Intercellular trafficking and protein delivery by a herpesvirus structural protein. Cell1997;88:223–233. [PubMed: 9008163]

Fitzgerald K.A. Palsson-Mcdermott E.M. Bowie A.G. Jefferies C.A. Mansell A.S. Brady G. Brint E.Dunne A. Gray P. Harte M.T. Mcmurray D. Smith D.E. Sms J.E. Bird T.A. O’Neill L.A. Mal(MyD88-adapter-like) is required for Toll-like receptor-4 signal transduction. Nature 2001;413:78–83. [PubMed: 11544529]

Fotin-Mleczek M. Welte S. Mader O. Duchardt F. Fischer R. Hufnagel H. Scheurich P. Brock R. Cationiccell-penetrating peptides interfere with TNF signaling by induction of TNF receptor internalization.J. Cell Sci. 2005;118:3339–3351. [PubMed: 16079278]

Frankel A.D. Pabo C.O. Cellular uptake of the tat protein from human immunodeficiency virus. Cell1988;55:1189–1193. [PubMed: 2849510]

Gondeau C. Gerbal-Chaloin S. Bello P. Aldrian-Herrada G. Morris M.C. Divita G. Design of a novelclass of peptide inhibitors of cyclin-dependent kinase/cyclin activation. J. Biol. Chem.2005;280:13793–13800. [PubMed: 15649889]

Gratton J.P. Yu J. Griffith J.W. Babbitt R.W. Scotland R.S. Hickey R. Giordano F.J. Sessa W.C. Cell-permeable peptides improve cellular uptake and therapeutic gene delivery of replication-deficientviruses in cells and in vivo. Nat. Med. 2003;9:357–362. [PubMed: 12598894]

Green M. Loewenstein P.M. Autonomous functional domains of chemically synthesized humanimmunodeficiency virus tat trans-activator protein. Cell 1988;55:1179–1188. [PubMed: 2849509]

Guex N. Peitsch M.C. SWISS-MODEL and the Swiss-PdbViewer: an environment for comparativeprotein modeling. Electrophoresis 1997;18:2714–2723. [PubMed: 9504803]

Hallbrink M. Floren A. Elmquist A. Pooga M. Bartfai T. Langel U. Cargo delivery kinetics of cell-penetrating peptides. Biochim. Biophys. Acta 2001;1515:101–109. [PubMed: 11718666]

Horng T. Barton G.M. Flavell R.A. Medzhitov R. The adaptor molecule TIRAP provides signallingspecificity for Toll-like receptors. Nature 2002;420:329–333. [PubMed: 12447442]

Horng T. Barton G.M. Medzhitov R. TIRAP: an adapter molecule in the Toll signaling pathway. Nat.Immunol. 2001;2:835–841. [PubMed: 11526399]

Joliot A. Pernelle C. Deagostini-Bazin H. Prochiantz A. Antennapedia homeobox peptide regulates neuralmorphogenesis. Proc. Natl. Acad. Sci. U.S.A. 1991;88:1864–1868. [PubMed: 1672046]

Kaneto H. Nakatani Y. Miyatsuka T. Kawamori D. Matsuoka T.A. Matsuhisa M. Kajimoto Y. Ichuo H.Yamasaki Y. Hori M. Possible novel therapy for diabetes with cell-permeable JNK-inhibitorypeptide. Nat. Med. 2004;10:1128–1132. [PubMed: 15448687]

Kellogg G.E. Semus S.F. Abraham D.J. HINT: a new method of empirical hydrophobic field calculationfor CoMFA. J. Comput. Aided Mol. Des. 1991;5:545–552. [PubMed: 1818090]

Langeduk J.P. Oluhoek T. Schut D. Autar R. Meloen R.H. New transport peptides broaden the horizonof applications for peptidic pharmaceuticals. Mol. Divers 2004;8:101–111. [PubMed: 15209161]

Lindsay M.A. Peptide-mediated cell delivery: application in protein target validation. Curr. Opin.Pharmacol. 2002;2:587–594. [PubMed: 12324264]

Low et al. Page 10

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Lundberg P. Magzoub M. Lindberg M. Hallbrink M. Jarvet J. Eriksson L.E. Langel U. Graslund A. Cellmembrane translocation of the N-terminal (1–28) part of the prion protein. Biochem. Biophys. Res.Commun. 2002;299:85–90. [PubMed: 12435392]

Mai J.C. Shen H. Watkins S.C. Cheng T. Robbins P.D. Efficiency of protein transduction is cell type-dependent and is enhanced by dextran sulfate. J. Biol. Chem. 2002;277:30208–30218. [PubMed:12034749]

Mcguffin L.J. Bryson K. Jones D.T. The PSIPRED protein structure prediction server. Bioinformatics2000;16:404–405. [PubMed: 10869041]

Medzhitov R. Preston-Hurlburt P. Kopp E. Stadlen A. Chen C. Ghosh S. Janeway C.A. MyD88 is anadaptor protein in the hToll/IL-1 receptor family signaling pathways. Mol. Cell 1998;2:253–258.[PubMed: 9734363]

Mi Z. Mai J. Lu X. Robbins P.D. Characterization of a class of cationic peptides able to facilitate efficientprotein transduction in vitro and in vivo. Mol. Ther. 2000;2:339–347. [PubMed: 11020349]

Noguchi H. Bonner-Weir S. Wei F.Y. Matsushita M. Matsumoto S. BETA2/NeuroD protein can betransduced into cells due to an arginine- and lysine-rich sequence. Diabetes 2005;54:2859–2866.[PubMed: 16186386]

Noguchi H. Matsushita M. Okitsu T. Moriwaki A. Tomizawa K. Kang S. Li S.T. Kobayashi N. MatsumotoS. Tanaka K. Tanaka N. Matsui H. A new cell-permeable peptide allows successful allogeneic islettransplantation in mice. Nat. Med. 2004;10:305–309. [PubMed: 14770176]

O’Neill L.A. The role of MyD88-like adapters in Toll-like receptor signal transduction. Biochem. Soc.Trans. 2003;31:643–647. [PubMed: 12773173]

Oehlke J. Beyermann M. Wiesner B. Melzig M. Berger H. Krause E. Bienert M. Evidence for extensiveand non-specific translocation of oligopeptides across plasma membranes of mammalian cells.Biochim. Biophys. Acta 1997;1330:50–60. [PubMed: 9375812]

Park J. Ryu J. Kim K.A. Lee H.J. Bahn J.H. Han K. Choi E.Y. Lee K.S. Kwon H.Y. Choi S.Y. Mutationalanalysis of a human immunodeficiency virus type 1 Tat protein transduction domain which is requiredfor delivery of an exogenous protein into mammalian cells. J. Gen. Virol. 2002;83:1173–1181.[PubMed: 11961273]

Peitz M. Pfannkuche K. Rajewsky K. Edenhofer F. Ability of the hydrophobic FGF and basic TATpeptides to promote cellular uptake of recombinant Cre recombinase: a tool for efficient geneticengineering of mammalian genomes. Proc. Natl. Acad. Sci. U.S.A. 2002;99:4489–4494. [PubMed:11904364]

Poltorak A. He X. Smirnova I. Liu M.Y. Van Huffel C. Du X. Bird Well D. Alejos E. Silva M. GalanosC. Freudenberg M. Ricciardi-Castagnoli P. Layton B. Beutler B. Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: mutations in Tlr4 gene. Science 1998;282:2085–2088. [PubMed:9851930]

Sabroe I. Read R.C. Whyte M.K. Dockrell D.H. Vogel S.N. Dower S.K. Toll-like receptors in health anddisease: complex questions remain. J. Immunol. 2003;171:1630–1635. [PubMed: 12902458]

Tao X. Xu Y. Zheng Y. Beg A.A. Tong L. An extensively associated dimer in the structure of the C713Smutant of the TIR domain of human TLR2. Biochem. Biophys. Res. Commun. 2002;299:216–221.[PubMed: 12437972]

Toshchakov V.U. Basu S. Fenton M.J. Vogel S.N. Differential involvement of BB loops of toll-IL-1resistance (TIR) domain-containing adapter proteins in TLR4- versus TLR2-mediated signaltransduction. J. Immunol. 2005;175:494–500. [PubMed: 15972684]

Vives E. Cellular uptake [correction of utake] of the Tat peptide: an endocytosis mechanism followingionic interactions. J. Mol. Recognit. 2003;16:265–271. [PubMed: 14523939]

Vogel S.N. Fenton M. Toll-like receptor 4 signalling: new perspectives on a complex signal-transductionproblem. Biochem. Soc. Trans. 2003;31:664–668. [PubMed: 12773178]

Wadia J.S. Dowdy S.F. Transmembrane delivery of protein and peptide drugs by TAT-mediatedtransduction in the treatment of cancer. Adv. Drug Deliv. Rev. 2005;57:579–596. [PubMed:15722165]

Wender P.A. Mitchell D.J. Pattabiraman K. Pelkey E.T. Steinman L. Rothbard J.B. The design, synthesis,and evaluation of molecules that enable or enhance cellular uptake: peptoid molecular transporters.Proc. Natl. Acad. Sci. U.S.A. 2000;97:13003–13008. [PubMed: 11087855]

Low et al. Page 11

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Xu Y. Tao X. Shen B. Horng T. Medzhitov R. Manley J.L. Tong L. Structural basis for signal transductionby the Toll/interleukin-1 receptor domains. Nature 2000;408:111–115. [PubMed: 11081518]

Yamamoto M. Sato S. Hemmi H. Sanjo H. Uematsu S. Kaisho T. Hoshino K. Takeuchi O. KobayashiM. Fujita T. Takeda K. Akira S. Essential role for TIRAP in activation of the signalling cascadeshared by TLR2 and TLR4. Nature 2002;420:324–329. [PubMed: 12447441]

Low et al. Page 12

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Fig. 1.Internalisation of peptides. (A) Mean fluorescence values of NBD-treated HeLa cells measuredafter the addition of sodium dithionite quencher. Background values of cell treated with sodiumdithionite only were subtracted. Cells (5 × l05 cells ml−1) were incubated with 10 (black bar),40 (grey bar) and 160 μM (white bar) of each of the NBD labelled peptides TAT(GRKKRRQRRRPPQ), TATAla (GRKKAAQAAAPPQ), TIRAP(GKMADWFRQTLLKKPKKRPNSPEST), IRAKI (CLHRRAKRRPPMTQVYER), TLR4(GRHIFWRRLRKALLDGKSWNPE), PU.1 (GSKKKIRLYQFLLDLLRSGDMKDS) andICE (QLLRKKRRIFIHSVGAGT) for 2 h. The results are the average of three independentexperiments. Error bars represent standard error of the mean (n = 3). (B) Structure to function

Low et al. Page 13

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

analysis of the cell permeable motif found in TIRAP. Increasing concentrations of the NBD-labelled peptides TIRAP (dark grey bar), TIRAPAla

(GKMADWFRQTLLAAPAAAPNSPEST) (light grey bars), TAT (black bars) and TAT Ala

(white bars) were incubated with HeLa and RAW 264.7 cells. The amount of internalisedpeptide was determined by adding sodium dithionite quencher as described. The results are theaverage of three independent experiments. Error bars represent standard error of the mean(n = 3).

Low et al. Page 14

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Fig. 2.Kinetics of p38 phosphorylation in LPS treated RAW 264.7 macrophages upon addition ofTBX2. Cells were pre-incubated with PBS, 40 μM of Antp-TIRAP(RQIKIWFQNRRMKWKKLQLRDAAPGGAIVS) or 160 μM of TBX2(GKMADWFRQTLLKKPKKRPNSPESTLQLRDATPGGAIVS) for 2 h and then stimulatedwith 10 ng/ml LPS for the indicated periods of time. Cell lysates were separated on 12% SDS-polyacrylamide gel and blotted onto nitrocellulose membrane. Activation of p38 was detectedusing a specific anti-phospho-p38 antibody. For each experiment, the amount of cell lysateloaded was assessed using an anti-pan phospo-tyrosine antibody. Immunoblot analysis isrepresentative of four independent experiments.

Low et al. Page 15

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Fig. 3.Activity of the TBX2 peptide. (A) TBX2 activity on IL-6 secretion of LPS stimulated RAW264.7 cells. RAW cells were pre-incubated with increasing concentrations of TBX2(GKMADWFRQTLLKKPKKRPNSPESTLQLRDATPGGAIVS) for 2 h and then challengedwith 10 ng/ml LPS, or 25 μg/ml Poly I:C or 1 μg/ml R848. Supernatants were collected after18 h and IL-6 levels quantified by ELISA. Results show one experiment carried out intriplicates representative of four independent experiments. Error bars represent standard errorof the mean. NT: non-treated. (B) Cell viability assessed after incubation with TBX2. RAW264.7 cells (l × 104 cells ml−1) were distributed in 96-well plates and incubated for 2 (blackbar), 4 (grey bar), 8 (white bars) and 18 h (grey dotted bars) with 10, 40 and 160 μM TBX2.

Low et al. Page 16

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Fig. 4.Activity of the TBX2 peptide ex vivo and in vivo. (A) TBX2 activity on TNF-α production ofLPS stimulated primary human macrophages. Primary human macrophages (see Section 2)were pre- treated with increasing concentrations of either TBX2(GKMADWFRQTLLKKPKKRPNSPESTLQLRDATPGGAIVS) (light grey bars) or BX2(LQLRDATPGGAIVS) (white bars) for 2 h and then challenged with 1 ng/ml LPS for 20 h.TNF-α production of non-challenged and non-treated cells is shown (NT). TNF-α levels in thesupernatants were quantified by ELISA. Results are the average of triplicates andrepresentative of two independent experiments. Error bars represent standard error of the mean.(B) TBX2 activity on TNF-α response of LPS treated mice. Groups of five mice were injected

Low et al. Page 17

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

intravenously (i.v.) with a single dose of TBX2 peptide (2, 4 or 6 mg per mouse) or PBS only.After 45 min, the mice were challenged with 10 μg of LPS injected i.v. 90 min later, the micewere anaesthetised and blood collected by cardiac puncture. TNF-α levels in the blood weredetermined by ELISA. Results are representative of two independent experiments. Levels ofTNF-α measured on non-challenged and PBS-treated mice were subtracted from TNF-α valuesobtained from peptide treated mice. Error bars represents standard error of the meann = 5. *p < 0.05, **p < 0.01.

Low et al. Page 18

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent

Fig. 5.Molecular modelling and prediction of protein–protein interaction. TLR4 (yellow), TIRAP(purple), MyD88 (red) and TBX2 (green) are shown in ribbon representation. Conservedprolines of TLR4, MyD88 and TIRAP are shown in ball and stick. Non-overlapping bindingsites for TIRAP and MyD88 on TLR4 and the hypothetical interactions between TBX2 andTLR4 are shown. The TBX2 associates via the BB loop with TLR4 on the TIRAP binding siteand via its amino terminus (α helix) projected towards the region of TLR4 that is predicted tobind MyD88.

Low et al. Page 19

Published as: J Biotechnol. 2007 May 01; 129(3-2): 555–564.

Sponsored Docum

ent Sponsored D

ocument

Sponsored Docum

ent