Báo cáo khoa học: Lipopeptide structure determines TLR2 dependent cell activation level potx

We found that the two ester bound fatty acid length threshold is beyond eight C atoms because almost no response was elicited by cellular challenge with analogues carrying shor-ter acyl

activation level

Ute Buwitt-Beckmann1, Holger Heine1, Karl-Heinz Wiesmu¨ller2, Gu¨nther Jung3, Roland Brock4 and Artur J Ulmer1

1 Department of Immunology and Cell Biology, Research Center Borstel, Borstel, Germany

2 EMC microcollections GmbH, Tu¨bingen, Germany

3 Institute of Organic Chemistry, University of Tu¨bingen, Tu¨bingen, Germany

4 Department of Molecular Biology, Institute for Cell Biology, University of Tu¨bingen, Tu¨bingen, Germany

Lipoproteins⁄ peptides (LP) are major constituents of

the cell wall of bacteria Bacterial LP are

di-O-acylated-S-(2,3-dihydroxypropyl)-cysteinyl residues N-terminally

coupled to distinct polypeptides, as found in the

macrophage-activating lipopeptide from Mycoplasma

fermentans (MALP2) [1] and in LP from the N

termi-nus of the cytochrome subunit of the photoreaction

center of Rhodopseudomonas viridis [2] The

S-(2,3-di-hydroxypropyl)-cysteine may be N-acylated with a

third fatty acid via an amide-linkage as is the case for

the LP from the cell wall of Escherichia coli [3] and numerous other Gram-negative bacteria [4] These LP activate the innate immune system and promote the formation of adaptive immunity as an adjuvant during stimulation with specific antigens

The receptor responsible for a functional recognition

of LP by cells is the Toll-like receptor 2 (TLR2) [5,6] TLRs are prominent pattern recognition receptors

of the innate immune system recognizing various invading microorganisms by conserved molecular

Keywords

TLR2; lipopeptides; ligand recognition;

structure–activity relationship

Correspondence

A J Ulmer, Research Center Borstel,

Parkallee 22, 23845 Borstel, Germany

Tel: +49 4537 188448

Fax: +49 4537 188435

E-mail: ajulmer@fz-borstel.de

(Received 19 August 2005, revised 13

October 2005, accepted 20 October 2005)

doi:10.1111/j.1742-4658.2005.05029.x

Bacterial lipoproteins⁄ peptides are composed of di-O-acylated-S-(2,3-dihyd-roxypropyl)-cysteinyl residues N-terminally coupled to distinct polypep-tides, which can be N-acylated with a third fatty acid Using a synthetic lipopeptide library we characterized the contribution of the lipid portion to the TLR2 dependent pattern recognition We found that the two ester bound fatty acid length threshold is beyond eight C atoms because almost

no response was elicited by cellular challenge with analogues carrying shor-ter acyl chains in HEK293 cells expressing recombinant human TLR2 In contrast, the amide bound fatty acid is of lesser importance While two ester-bound palmitic acids mediate a high stimulatory activity of the respective analogue, a lipopeptide carrying one amide-bound and another ester-bound palmitic acid molecule was inactive In addition, species

speci-fic LP recognition through murine and human TLR2 depended on the length of the two ester bound fatty acid chains In conclusion, our results indicate the responsibility of both ester bound acyl chains but not of the amide bound fatty acid molecule for the TLR dependent cellular recogni-tion of canonical triacylated LP, as well as a requirement for a minimal acyl chain length Thus they might support the explanation of specific immuno-stimulatory potentials of different microorganisms and provide a basis for rational design of TLR2 specific adjuvants mediating immune activation to distinct levels

Abbreviations

BbMALP2, macrophage-activating lipopeptide from Mycoplasma fermentans; LP, lipoproteins ⁄ peptides; huTLR2, human TLR2; muTLR2, murine TLR2; IL, interleukin.

structures, so called pathogen-associated molecular

patterns [7] To date 11 TLRs have been described

The receptors differ not only in ligand specificity but

also in their expression pattern on different cells of the

innate immune system In addition, distinct TLRs are

able to induce the expression of different sets of

inflammatory target genes [8] TLR2 recognizes the

most diverse set of molecular structures, including,

lipoteichoic acid, lipoarabinomanan, bacterial LP, as

well as molecules from yeast, spirochetes and fungi [8]

Unlike other TLRs, which are functionally active as

homomers [9], TLR2 forms heteromers either with

TLR1 or TLR6 [9] to attain specificity for a given

sti-mulus [10–13] The previous concept, which has stated

that di-acylated LP like MALP2 signals through

TLR2⁄ 6, whereas acylated LP, as realized in

tri-

palmitoyl-S-(2,3-dihydroxypropyl)-cysteinyl-seryl-tetra-lysine (Pam3C-SK4), signals through TLR2⁄ 1, has

recently been challenged by the finding that

Pam2C-SK4 as well as MALP2-Pam2C-SK4 is recognized by TLR2

in a TLR6-independent manner [14] The molecular

mechanism of the recognition of LP by TLR2 is not

clear yet Recently it was demonstrated that binding

of Pam3C-SK4 to the LPS-binding protein, and

soluble- or membrane-bound CD14 results in spatial

proximity of LP, CD14, TLR2 and TLR1 [15–17] and

subsequent TLR2 signalling

Among the TLR2-dependent bacterial ligands, LP

are primary candidates for analysing the structural

requirements for TLR2 activation Firstly, these

molecules are available from various bacterial species

expressing different molecular structures Secondly,

chemical synthesis provides a vast variety of

analytic-ally well defined LP analogues [18,19] Although the

structural requirements for optimal adjuvant activity

of synthetic LP have been studied in detail [20], the

structure–activity relationships for the recognition of

LP by TLR2 are still poorly defined Recently we

have systematically investigated the contribution of

all proteinogenic amino acids except cysteine in the

peptide moiety of Pam3Cys-lipohexapeptides

Ninety-five LP amide subcollections were synthesized by

combinatorial peptide chemistry None of these

lipo-hexapeptides subcollections substantially exceeded the

biological activity of Pam3C-SK4 indicating that the

biological activity is more or less independent of

peptide sequences in this LP library [19] However,

some exceptions from these general results have been

found (Wiesmu¨ller et al unpublished data)

Concern-ing the contribution of the fatty acids in

TLR2-dependent signalling evidence that the fatty acid

composition has an impact on the biological activity

has been presented [21] For this reason we have

now systematically investigated the structure-activity relationships for the lipid moiety of LP, tightly focusing on recognition by TLR2 LP libraries were synthesized having different fatty acid compositions

at the three reactive groups of the LP-scaffold, namely the ester-bound fatty acids and the amide-bound fatty acids We could demonstrate that the amide-bound fatty acids contribute, if at all, only minutely to the recognition by TLR2 In contrast, long chain ester-bound fatty acids are essential for the induction of signalling through TLR2 Interest-ingly, a longer chain-length of the ester-bound fatty acids is necessary for the activation of huTLR2 in comparison to muTLR2, revealing a species specificity

of LP recognition

Results and discussion

The structure-activity relationships in the recognition

of TLR2 by various bacterial LP have been poorly defined so far In particular, little information exists

on the relevance of the fatty acid composition and dis-tribution within the LP head group for TLR2 activa-tion In order to define the structural requirements for the lipid moiety of LP during TLR2 activation, inde-pendent from a coreceptor bias, HEK293 cells trans-fected with either the huTLR2 or muTLR2 were selected as a cellular test system These cells express both TLR1 and TLR6 At first it should be noted that all bioactive LP used in our study were found to be TLR2 dependent: There was no response to these LP

in nontransfected HEK293 cells, cells transfected with

an empty vector, nor in TLR2-deficient mice (data not shown)

In a first set of experiments a LP collection was tes-ted that was composed of synthetic LP analogues of Pam3C-SK4 in which the amide-bound as well as the two ester-bound fatty acids were modified The amide-bound fatty acids in our collection were Hex, Pe, Dec, Myr, and Pam and the ester- bound fatty acids Ac, Hex, Oct, or Pam (for abbreviations of the fatty acids see Table 1) Due to the synthesis protocol for this lipopeptide collection [22], both ester-bound fatty acids were identical HEK293 cells were transfected with huTLR2, stimulated with 1000 nm of LP and the release of interleukin (IL)-8 in the culture supernatant was determined after 24 h of culture Our results clearly indicate that the amide-bound fatty acids have,

if at all, only a minute effect on the TLR2-mediated IL-8 release: no significant differences in the response

of cells stimulated with LP containing N-bound Hex,

Pe, Dec, Myr, or Pam acyl residues were observed (Table 2)

Previously it had been postulated that the presence

or absence of an N-bound fatty acid is responsible

for converting a TLR2⁄ TLR6 dependent LP into a

TLR2⁄ TLR1 dependent LP However, our recent

results, showing that Pam2C-SK4 and also

MALP2-SK4 (the elongated MALP2 analogue

Pam2C-GNNDESNISFKEKSK4) are TLR6-independent LP,

required the rejection of this hypothesis At least all

LP with a SK4-peptid tail might be recognized by

TLR2 in a TLR6 independent manner It therefore

seems reasonable to assume that all the LP analogues

of this collection are TLR6 independent Whether

these TLR2 agonists signal in a TLR1-dependent

man-ner remains to be investigated It should be noted,

however, that even the well-studied reference

lipohexa-peptide Pam3C-SK4 exerts not only TLR1-dependent

bioactivity, but also, at least in part, a

TLR1-inde-pendent activity [12,14]

In contrast to the amide-bound fatty acids, the

iden-tity of the ester-bound fatty acids is of great relevance

for the TLR2-dependent response of the cells For

fatty acids with a short chain length (Ac or Hex) only

a minute or no response could be observed The response was slightly enhanced after stimulation with a

LP having ester-bound octanoic acids [at least with respect to nuclear factor jB (NFjB) translocation] and was maximal after stimulation with palmitoylated LP analogues The same structure-activity relationship for the stimulation of HEK293-huTLR2 cells was also observed, when the activation of NFjB, as determined

in a luciferase-reporter assay, instead of IL-8 release was estimated (Table 2) This indicates that the nature

of the ester-bound but not the amide-bound fatty acids

of LP is of great relevance throughout from TLR2-induced signal transduction up to cytokine release

To obtain more detailed information on the dose– response function of lipopeptides, for selected synthetic

LP analogues the IL-8 release by HEK293-huTLR2 cells was determined over a concentration range of 1-1000 nm As shown in Fig 2A, all four tested lipo-hexapeptide analogues with two ester-bound Pam moieties exhibited nearly the same dose–response characteristic, independently from the nature of the N-bound acyl moiety On the other hand, HEK293-huTLR2 were unresponsive to all lipohexapeptide

Table 1 Abbreviation, structure and name of fatty acids used in

this study.

Myr C14 H28 O2 Tetradecanoic acid ⁄ myristic acid

Pam C16 H32 O2 Hexadecanoic acid ⁄ palmitic acid

Table 2 Stimulation of HEK-huTLR2 by various synthetic lipohexapeptide analogues having different O- and N-acylated head groups The LP analogues were used at a concentration of 1000 n M IL-8 release was measured after a culture period of 24 h, the Luciferase-reporter assay for NFjB was performed after 6 h of stimulation The results are expressed as percentage response in relation to the response to Pam3CSK4 Each value represents the mean of triplicates.

IL-8 release

Luciferase-reporter assay for NFjB

Fig 1 Structures and denotations of some synthetic LP analogues used in this study To illustrate the structure denotation of LP logues used in this paper, the structures of three synthetic LP ana-logues are given The chemical structure of the scaffold with the four reactive groups is shown R1represents the amide-bound fatty acid, R 2 and R 3 represent the ester-bound fatty acids, and R 4 repre-sents the peptide moiety.

analogues carrying two ester bound hexanoic acids.

This lack of activity was independent from the nature

of the acyl moiety, i.e short or long chain length

(Fig 2B) In addition, for a series of LP analogues

with an increasing chain length of the two ester bound

fatty acids, no response in HEK293-huTLR2 cells was

detected for Ac, Hex, or Oct but only for Pam

moiet-ies This structure–activity relationship again was the

same for a series of LP with a short (Hex) or a long

(Myr) amide-bound fatty acid (Fig 3A and B) In

summary, all LP analogues with short O-acylated fatty

acids (Ac, Hex, or Oct) were found to be inactive for

HEK293-huTLR2 Moreover, the amide-bound fatty

acid is of low or even no relevance for the

TLR2-medi-ated stimulation of HEK-huTLR2 cells

The results, presented so far, indicate that LP

ana-logues with two ester-bound fatty acids having a short

chain length of eight carbons or less are incapable of

activating cells through TLR2, whereas a chain length

of 16 carbons, such as that present in palmitic acid is

optimal Therefore, we decided to analyse the impact

of the acyl chain length as a specific aspect of TLR2

dependent LP recognition For these investigations a

further collection of LP with a SSNASK4-peptide moi-ety was synthesized and characterized by electro-MS

It should be noted that the dose kinetic of the huTLR2 dependent cellular response, revealed upon application of lipohexapeptide Pam3C-SK4, largely resembled that obtained upon challenge with the lipo-decapeptide Pam3C-SSNASK4 (data not shown) For these analogues with a longer peptide moiety the stim-ulatory activity was a function of the chain length of the two ester-bound fatty acids as well Only a margi-nal response was found for the LP amargi-nalogues having ester-bound decanoyl (10 carbon atoms) moieties or acyl residues with less numbers of carbons Ester-bound fatty acids with at least 14 carbon atoms (Myr) were necessary to obtain an optimal response (Table 3) Activation of IL-8 release in HEK293-huT-LR2 cells again showed the same structure–activity relationships as activation of the translocation of NFjB (Table 3) These results obtained with LP at a fixed concentration of 1000 nm were again confirmed

by the acquisition of dose–response curves (Fig 4) It should be noted that a LP with ester-bound Ara moi-eties had nearly the same agonistic activity as a LP

Fig 2 IL-8 release induced in HEK-huTLR2 by various synthetic LP analogues having differently N-acylated cysteine residues HEK-huTLR2 cells were stimulated with increasing concentrations (1–1000 n M ) of synthetic LP analogues Results obtained for (A) bis-O-palmitoylated lipohexapeptide analogues and (B) bis-O-hexanoylated lipohexapeptide analogues having different N-acylated head groups After a culture period of 24 h the supernatants were harvested and the IL-8 concentration was measured by an ELISA The results are expressed as mean ± SD, n ¼ 3 The chemical structures of the LP analogues are included.

with ester-bound Pam moieties Therefore, it can be

concluded that a LP needs a minimum chain length of

the ester-bound fatty acids for optimal response

How-ever, this response cannot be increased further by an

elongation of the ester-bound fatty acids

Hitherto we could conclude that only LP with at least two long-chain ester-bound fatty acids, such as Myr or Pam, are able to signal through huTLR2 To scrutinize whether the positional distribution of the two fatty acids within the S-(2,3-dihydroxypropyl)-cysteinyl

Table 3 Stimulation of HEK-huTLR2 by various synthetic lpodecapeptide analogs having different ester-bound fatty acids The LP analogues were used at a concentration of 1000 n M IL-8 release was measured after a culture period of 24 h, the Luciferase-reporter assay for NFjB was performed after 6 h of stimulation Each value represents the mean of triplicates.

a

Chain length of O-acylated fatty acids.

Fig 3 IL-8 release induced in HEK-huTLR2 by various synthetic lipohexapeptide analogues having different O-acylated head groups HEK-huTLR2 cells were stimulated with increasing concentrations (1–1000 n M ) of synthetic LP analogues Results obtained for (A) HexPam2C-SK4 and (B) MyrPam2C-SK4 analogues having different ester-bound fatty acids After a culture period of 24 h the supernatants were harvested and the IL-8 concentration was measured in an ELISA The results are expressed as mean ± SD, n ¼ 3 The chemical struc-tures of the LP analogs are included.

headgroup is of relevance for the bioactivity, we

syn-thesized analogues of the di-acylated LP from the N

terminus of the cytochrome subunit of the

photoreac-tion centre of R viridis at our disposiphotoreac-tion, namely

Pam2C-FEPPPATTT [2] This LP has only two

ester-bound palmitoyl residues In addition we tested a

diacylated analogue of this R viridis LP with one

N-palmitoyl and one O-palmitoyl residue Only

Pam2C-FEPPPATTT having two O-palitoyl residues

stimulated the expression of IL-8 release in

HEK293-huTLR2 cells The Pam2C-FEPPPATTT analogue,

having one N- and one O-palmitoyl residue did not

show a stimulatory activity (Fig 5) This finding

indi-cates that the distribution of the fatty acids within the

LP is of great relevance for the bioactivity of

di-acylat-ed LP Only LP with two ester-bound fatty acids

pos-sess TLR2-dependent stimulatory activity

Recently it has been found that tri-lauroylated LP

(with three dodecanoic acids) are recognized by

muTLR2 but not by huTLR2 whereas both mu- and

huTLR2 are efficiently recognized by tri-palmitoylated

LP [23] This finding led us to investigate whether the

species-specific recognition of LP is a function of

the length of the ester- or amide-bound fatty acids of

the LP analogues HEK293 cells, which were

transi-ently transfected with either hu-TLR2 or mu-TLR2,

were stimulated with various members of our

Pam3C-SK4 collection at a concentration of 1000 nm and the release of IL-8 was measured after 24 h of incubation

As shown in Table 4, the response of both transfect-ants to challenge with the TLR2-independent stimulus tumour necrosis factor was largely similar LP ana-logues with short-length O-acylated fatty acids (Ac and Hex) were almost unable to stimulate either hu-TLR2- nor mu-TLR2-transfected HEK293 cells In both cases, this dependence on the chain length was independent from the type of the amide-bound fatty acid of the LP analogue However, when we incubated transfected HEK293 cells with LP analogues, having two ester-bound Oct moieties and N-acyl residues

of different length (HexOct2C-SK4, PeOct2C-SK4, DecOct2C-SK4, PamOct2C-SK4, or MyrOct2C-SK4), only HEK293-muTLR2 but not HEK293-huTLR2 were stimulated In this set of LP only bis-O-palmito-ylated LP analogues were able to stimulate both, HEK293-huTLR2 and HEK293-muTLR2 Interest-ingly, PamOct2C-SK4 exhibited a lower biological activity than the other Oct2C-SK4 derivatives in this set of LP, an observation that we cannot explain at the moment

To define the length of the fatty acids required for the stimulation through huTLR2- and muTLR2 trans-fected HEK293 cells in more detail, we made use of our second LP collection including analogues carrying

Fig 4 IL-8 release induced in huTLR2 by various synthetic lipodecapeptide analogues having different ester-bound fatty acids HEK-huTLR2 cells were stimulated with increasing concentrations (1–1000 n M ) of synthetic Pam3C-SSNASK4 analogues After a culture period of

24 h the supernatants were harvested and the IL-8 concentration was measured in an ELISA The results are expressed as mean ± SD,

n ¼ 3 In addition, an illustration of the chemical structure of the LP analogues is shown.

the longer SSNASK4 peptide chain N-Palmitoylated

lipodecapeptides with short length bis-O-acylated head

groups (PamHex2C-SSNASK4, PamBa2C-SSNASK4,

and PamPe2C-SSNASK4) did not stimulate

HEK293-huTLR2 and HEK293-muTLR2 cells (Table 5) The

stimulatory activity of PamOct2C-SSNASK4,

Pam-Pel2C-SSNASK4, and PamDec2C-SSNASK4 was low

or moderate in HEK293-muTLR2 cells but was absent

in HEK293-huTLR2 cells A strong response of HEK293-muTLR2 cells was observed upon stimula-tion with PamDod2C-SSNASK4, whereas responsive-ness of HEK293-huTLR2 to this LP was low N-palmitoylated LP carrying long length ester-bound carboxylic acids (PamMyr2C-SSNASK4, PamOle2C-SSNASK4, PamLin2C-SSNASK4, and PamAra-Hex2C-SSNASK4), all exhibited an almost similar high stimulatory activity in both, HEK293-huTLR2 and HEK293-muTLR2 cells Taken together, from these results we conclude that the species-specific recognition of LP is not a function of the length of the N-fatty acyl, but of the O-fatty acyl groups of LP analogues

Table 4 The response of HEK-huTLR2 and HEK-muTLR2 to

lipo-hexapeptide analogues with various amide- and ester-bound fatty

acids The lipopeptides were used at a concentration of 1000 n M ,

tumour necrosis factor was used at a concentration of 10 ngÆmL)1.

The results are expressed as IL-8 release (pgÆmL)1), mean ± SD,

n ¼ 3.

Table 5 The response of HEK-huTLR2 and HEK-muTLR2 to lipo-decapeptide analogues with various ester-bound fatty acids The LP analogs were used at a concentration of 1000 n M IL-8 release was measured after a culture period of 24 h Each value represents the mean ± SD of triplicates.

Structure

IL-8 release (pgÆmL)1)

Fig 5 Response of HEK-huTLR2 cells to doubly palmitoylated synthetic LP analogues from R viridis having a different positional distribution

of the two fatty acids HEK-huTLR2 cells were stimulated with rising concentrations (1–1000 n M ) of two synthetic Pam2C-FEPPPATTT ana-logues, which only differ in the positions of the two Pam There are either two ester-bound Pam or one esterfied Pam and one amidated Pam as indicated After a culture period of 24 h the supernatants were harvested and the IL-8 concentration was measured in an ELISA The results are expressed as mean ± SD, n ¼ 3 The chemical structures of the LP analogues are included.

This paper presents a systematic investigation of the

structure–activity relationships between the fatty acid

patterns in LP and the huTLR2, as well as the

muTLR2-dependent pattern recognition The analyses

revealed that the chain length of the two ester-bound

fatty acids is an essential determinant for the biological

activity of LP in huTLR2 and in

HEK293-muTLR2 cells Only fatty acids, having a chain length

of 12 or more carbons were able to substantially

sti-mulate the cells In contrast, the amide-bound fatty

acids of the LP analogues had no remarkable effect on

the recognition by the TLR2 transfected HEK293 cells

LP analogues without amide-bound fatty acid were

nearly as active as LP with this substitution (data not

shown) However, it should be kept in mind that the

amide-bound fatty acid of LP beside the peptide

moi-eties determines the coreceptor usage (TLR2⁄ TLR1 vs

TLR2⁄ TLR6 [13,24–26] Whereas the di-palmitoylated

LP from mycoplasma and the di-palmitoylated LP

form the R viridis are TLR2⁄ TRL6 dependent, their

tri-palmitoylated analogues are TLR6-independent In

addition we could demonstrate that the chain length

of the two ester-bound fatty acids is a determinant

for the species-specific recognition by huTLR2 vs

muTLR2 Whereas LP having at least two ester-bound

fatty acids with 12 carbon atoms are required for

acti-vation of huTLR2, muTLR2 is already activated by

LP having fatty acids with only eight carbon atoms

Finally, for HEK293-huTLR2 cells, the distribution of

fatty acids for diacylated LP is also of great relevance

for the bioactivity of LP In comparison to LP with

two ester-bound fatty acids, LP with one amide-bound

and one ester-bound fatty acid show a markedly

reduced bioactivity in HEK293-huTLR2 cells

Our investigations have been run in a cell model

using HEK293 cells transfected with human or murine

TLR2 This model has the advantage of being a

con-trolled in vitro system, in which the receptor, which is

responsible for the recognition of the ligand, is well

defined Preliminary results indicate that a analogous

structure–activity relationship exists also in native cells

(e.g., human and murine macrophages, unpublished

data) Furthermore, using a limited number of

Pam3C-SK4 analogues it has been shown that they induce

specific patterns of cytokines, chemokines and

costimu-latory membrane molecules in human dendritic cells,

depending on the source of amide-bound and

ester-bound fatty acids [27] Therefore, we conclude that

our results obtained in HEK293-TLR2 cells reflect the

structural relationship of LP-recognition also in native

cells and the native immune system and indeed, also

in vivo the structure of LP determines the adjuvant

activity [28] LP with a N-acyl-S-diacylglyceryl cysteine

backbone has been found in all bacteria During bac-terial infection these LP may be involved in the initi-ation of the response of the innate immune system and activation⁄ modulation of antigen-presenting cells inclu-ding dendritic cells, thereby modulating the pattern of the response of the adaptive immune system There-fore, it is reasonable to conclude that the structure of bacterial LP may play a role for the outcome of a bac-terial infection However, at present it is still poorly understood to which extent bacterial LP contribute to the extent and pattern of the immune response during bacterial infection

Taken together, our results show the substantial importance of the two ester-bound acyl residues within

LP for the induction of signalling through TLR2 Very similar to the TLR4 ligand LPS [29], the number, the chain length, and the distribution of the fatty acids is of great consequence for the stimulatory activity of LP

LP are amphiphilic molecules and have been described

to form, like LPS, supramolecular structures in aqueous solution air–water interfaces, and lipid bilayer mem-branes [30–33] As has been determined with LPS, the supramolecular conformation and the molecular shape are important for its biological activity and the recogni-tion by LPS binding molecules, e.g TLT4 [34] In con-trast to LPS, however, the supramolecular structure and molecular conformation of LP, which is optimal for the recognition by TLR2 remains to be investigated

Experimental procedures

Materials

All lipopeptide collections and single lipohexa- and lipo-decapeptides were synthesized and analysed according to published procedures [35] by EMC microcollections GmbH (Tu¨bingen, Germany, E-mail: emc@microcollections.de) The abbreviations, structures and names of the fatty acids used in this paper are given in Table 1 The abbreviations

of the amino acids relate to the single letter code The structure denotation of LP are explained in Fig 1 All SSNASK4- and SK4- LP analogues were solubilized in water at a concentration of 1 mm, resulting in a clear solu-tion Pam2C-FEPPPATTT analogues were solubilized in dimethyl sulfoxind at 1 mm and diluted in culture medium prior stimulation of the cells All synthetic LP used in this paper stimulate cells in a strictly TLR2-dependent manner

as determined by the use of TLR2-deficient mice and untransfected HEK293 cells Unless otherwise indicated, all fine chemicals were from Sigma-Aldrich (Deisenhofen, Ger-many), Serva (Heidelberg, GerGer-many), Merck Biosiences (Darmstadt, Germany) or Roche Diagnostics (Mannheim, Germany)

Stable transfected HEK293 cells

Expression plasmid containing a Flag-tagged version of

hu-TLR2 was a kind gift from P Nelson, Seattle, USA and

was subcloned into pREP9 (Invitrogen, Karlsruhe,

Germany) Stable transfection of HEK293 cells with

pREP9-Flag-huTLR2 was performed using Superfect

Transfection Reagent (Quiagen, Hilden, Germany)

accord-ing to the manufacturer’s recommendations Positive cells

were selected by fluorescence-activated cell sorting Clonal

cell lines were obtained by limiting dilution Stably

trans-fected cells were maintained in DMEM supplemented with

10% FCS, 0.5 unitsÆmL)1 penicillin, 0.5 lgÆmL)1

strepto-mycin, and 400 lgÆmL)1 G418 (Biochrom, Berlin,

Germany) Stable transfected HEK293-huTLR2 cells were

plated on 48-well dishes (Greiner, Frickenhausen,

Germany) at a concentration of 3· 105ÆmL)1 in 400 lL

complete medium without G418 The following day cells

were stimulated with the indicated LP for 24 h

Superna-tants were collected and analysed for an IL-8 content with

a commercial ELISA (Biosource, Solingen, Germany)

Transient transfected HEK293 cells

HEK293 cells were plated at a density of 1.5· 105

ÆmL)1in 96-well plates in DMEM supplemented with 10% FBS,

0.5 unitsÆmL)1penicillin, 0.5 lgÆmL)1streptomycin The

fol-lowing day, cells were transiently transfected using Polyfect

(Quiagen, Hilden, Germany) according to the manufactures’

protocol Expression plasmid containing the Flag-tagged

ver-sion of mu-TLR2 was a kind gift from E Lien, MA, USA

and was subcloned into pREP9 (Invitrogen, Karlsruhe,

Ger-many) pREP9-Flag-mu-TLR2 and pREP9-Flag-huTLR2

plasmids were used at 200 ng per transfection The efficiency

of the transient transfection was 50–70% as determined by

FACS analysis using a anti-TLR2 mAb After 6 h of

trans-fection cells were washed and stimulated for further 24 h

Interleukin-8 content in the culture supernatants was

quanti-fied using a commercial ELISA (Biosource, Camarillo, CA,

USA) The kinetics of the response are shown in Fig 6

NF-jB reporter assay

HEK293 cells were cotransfected with 150 ng per

transfec-tion NF-jB firefly luciferase reporter plasmid and 50 ng per

transfection control Renilla luciferase plasmid (both were

kindly provided by Dr D.T Golenbock, Worcester, USA)

After transfection the cells were stimulated with LP analogs

for 6 h and lysed by passive lysis buffer (Promega,

Mann-heim, Germany) Luciferase activity was measured by using

a luminometer (Berthold, Bad Wildbad, Germany) and was

calculated in relative light units as a ratio of

NF-jB-dependent firefly luciferase activity to NF-jB-inNF-jB-dependent

Renillaluciferase activity The kinetics of the response are

shown in Fig 6

Statistics

The results are given in mean of three cultures run in a given experiment Standard deviation was less than 15% or

as given in the figures and tables Each experiment was repeated at least three times One of these three experi-ments, giving representative results, is shown in each figure

Acknowledgements

The skillful technical assistance by P Prilla and

C Schneider (Research Center Borstel) and by

M Wacker and B Pa¨tzold (EMC microcollections GmbH) is acknowledged We thank A Wallisch for final reading of this manuscript This work was sup-ported by the Deutsche Forschungsgemeinschaft (Ul

68⁄ 3-1) and the Bundesministerium fu¨r Forschung und Technologie (project Biochance Nr 0312662)

References

1 Mu¨hlradt PF, Kiess M, Meyer H, Su¨ssmuth R & Jung

G (1997) Isolation, structure elucidation, and synthesis

of a macrophage stimulatory lipopeptide from Myco-plasma fermentans acting at picomolar concentration

J Exp Med 185, 1951–1958

2 Metzger JW, Beck-Sickinger AG, Loleit M, Eckert M, Bessler WG & Jung G (1995) Synthetic S-(2,3-dihydroxy-propyl)-cysteinyl peptides derived from the N-terminus of the cytochrome subunit of the photoreaction centre of Rhodopseudomonas viridisenhance murine splenocyte proliferation J Pept Sci 1, 184–190

Fig 6 Kinetics of the IL-8 release and NFjB activation after stimu-lation of HEK-huTLR2 with Pam3-CSK4 HEK-huTLR2 cells were sti-mulated with Pam3C-SK4 at a concentration of 1000 n M After various culture periods (4 h, 8 h, 16 h, 24 h, 48 h) the supernatants were harvested and the IL-8 concentration was measured in an ELISA The Luciferase-reporter assay for NFjB was performed after 2 h, 4 h, 6 h, and 8 h of stimulation The results are expressed as mean ± SD, n ¼ 3.

3 Braun V (1975) Covalent lipoprotein from the outer

membrane of Escherichia coli Biochim Biophys Acta

415, 335–377

4 Madan BM & Sankaran K (2002) DOLOP – database

of bacterial lipoproteins Bioinformatics 18, 641–643

5 Takeuchi O, Kaufmann A, Grote K, Kawai T, Hoshino

K, Morr M, Mu¨hlradt PF & Akira S (2000) Cutting

edge: preferentially the R-stereoisomer of the

mycoplas-mal lipopeptide macrophage-activating lipopeptide-2

activates immune cells through a toll-like receptor

2- and MyD88-dependent signaling pathway J Immunol

164, 554–557

6 Lien E, Sellati TJ, Yoshimura A, Flo TH, Rawadi G,

Finberg RW, Carroll JD, Espevik T, Ingalls RR, Radolf

JD & Golenbock DT (1999) Toll-like receptor 2

func-tions as a pattern recognition receptor for diverse

bac-terial products J Biol Chem 274, 33419–33425

7 Janeway CA Jr & Medzhitov R (2002) Innate immune

recognition Annu Rev Immunol 20, 197–216

8 Heine H & Lien E (2003) Toll-like receptors and their

function in innate and adaptive immunity Int Arch

Allergy Immunol 130, 180–192

9 Ozinsky A, Underhill DM, Fontenot JD, Hajjar AM,

Smith KD, Wilson CB, Schroeder L & Aderem A

(2000) The repertoire for pattern recognition of

patho-gens by the innate immune system is defined by

coop-eration between toll-like receptors Proc Natl Acad Sci

USA 97, 13766–13771

10 Alexopoulou L, Thomas V, Schnare M, Lobet Y,

Anguita J, Schoen RT, Medzhitov R, Fikrig E & Flavell

RA (2002) Hyporesponsiveness to vaccination with

Bor-relia burgdorferi OspA in humans and in TLR1- and

TLR2-deficient mice Nat Med 8, 878–884

11 Morr M, Takeuchi O, Akira S, Simon MM & Mu¨hlradt

PF (2002) Differential recognition of structural details

of bacterial lipopeptides by toll-like receptors Eur J

Immunol 32, 3337–3347

12 Takeuchi O, Sato S, Horiuchi T, Hoshino K, Takeda

K, Dong Z, Modlin RL & Akira S (2002) Cutting edge:

role of Toll-like receptor 1 in mediating immune

response to microbial lipoproteins J Immunol 169,

10–14

13 Takeuchi O, Kawai T, Mu¨hlradt PF, Morr M, Radolf

JD, Zychlinsky A, Takeda K & Akira S (2001)

Discri-mination of bacterial lipoproteins by Toll-like receptor

6 Int Immunol 13, 933–940

14 Buwitt-Beckmann U, Heine H, Wiesmu¨ller KH, Jung

G, Brock R, Akira S & Ulmer AJ (2005) Toll-like

receptor 6-independent signaling by diacylated

lipopep-tides Eur J Immunol 35, 282–289

15 Manukyan M, Triantafilou K, Triantafilou M, Mackie

A, Nilsen N, Espevik T, Wiesmu¨ller KH, Ulmer AJ &

Heine H (2005) Binding of lipopeptide to CD14 induces

physical proximity of CD14, TLR2 and TLR1 Eur J

Immunol 35, 911–921

16 Vasselon T, Detmers PA, Charron D & Haziot A (2004) TLR2 recognizes a bacterial lipopeptide through direct binding J Immunol 173, 7401–7405

17 Schro¨der NW, Heine H, Alexander C, Manukyan M, Eckert J, Hamann L, Go¨bel UB & Schumann RR (2004) Lipopolysaccharide binding protein binds to tri-acylated and ditri-acylated lipopeptides and mediates innate immune responses J Immunol 173, 2683–2691

18 Wiesmu¨ller KH, Bessler WG & Jung G (1992) Solid phase peptide synthesis of lipopeptide vaccines eliciting epitope-specific B-, T-helper and T-killer cell response Int J Pept Protein Res 40, 255–260

19 Spohn R, Buwitt-Beckmann U, Brock R, Jung G, Ulmer AJ & Wiesmu¨ller KH (2004) Synthetic lipopep-tide adjuvants and Toll-like receptor 2–structure-activity relationships Vaccine 22, 2494–2499

20 Bessler WG, Mittenbuhler K, Esche U & Huber M (2003) Lipopeptide adjuvants in combination treatment Int Immunopharmacol 3, 1217–1224

21 Mu¨ller SD, Mu¨ller MR, Huber M, Esche UU, Kirsch-ning CJ, Wagner H, Bessler WG & Mittenbuhler K (2004) Triacyl-lipopentapeptide adjuvants: TLR2-depen-dent activation of macrophages and modulation of receptor-mediated cell activation by altering acyl-moi-eties Int Immunopharmacol 4, 1287–1300

22 Reutter F, Jung G, Baier W, Treyer B, Bessler WG & Wiesmu¨ller KH (2005) Immunostimulants and Toll-like receptor ligands obtained by screening combinatorial lipopeptide collections J Pept Res 65, 375–383

23 Grabiec A, Meng G, Fichte S, Bessler W, Wagner H & Kirschning CJ (2004) Human but not murine toll-like receptor 2 discriminates between tri-palmitoylated and tri-lauroylated peptides J Biol Chem 279, 48004–48012

24 Takeuchi O, Sato S, Horiuchi T, Hoshino K, Takeda

K, Dong Z, Modlin RL & Akira S (2002) Cutting edge: role of Toll-like receptor 1 in mediating immune response to microbial lipoproteins J Immunol 169, 10–14

25 Akira S (2003) Mammalian Toll-like receptors Curr Opin Immunol 15, 5–11

26 Buwitt-Beckmann U, Heine H, Wiesmu¨ller KH, Jung

G, Brock R, Akira S & Ulmer AJ (2005) Toll-like receptor 6-independent signaling by diacylated lipopep-tides Eur J Immunol 35, 282–289

27 Reschner A, Moretta A, Landmann R, Heberer M, Spagnoli GC & Padovan E (2003) The ester-bonded palmitoyl side chains of Pam3CysSerLys4 lipopeptide account for its powerful adjuvanticity to HLA class I-restricted CD8+ T lymphocytes Eur J Immunol 33, 2044–2052

28 Bessler WG, Baier W, vd EU, Hoffmann P, Heinevetter

L, Wiesmu¨ller KH & Jung G (1997) Bacterial Lipopep-tides Constitute Efficient Novel Immunogens and Adju-vants in Parenteral and Oral Immunization Behring Institute Mitt, 390–399