Cell Activation and Apoptosis by Bacterial Lipoproteins Through Toll-like Receptor-2

See allHide authors and affiliations

Science  30 Jul 1999:
Vol. 285, Issue 5428, pp. 736-739
DOI: 10.1126/science.285.5428.736


Apoptosis is implicated in the generation and resolution of inflammation in response to bacterial pathogens. All bacterial pathogens produce lipoproteins (BLPs), which trigger the innate immune response. BLPs were found to induce apoptosis in THP-1 monocytic cells through human Toll-like receptor–2 (hTLR2). BLPs also initiated apoptosis in an epithelial cell line transfected with hTLR2. In addition, BLPs stimulated nuclear factor–κB, a transcriptional activator of multiple host defense genes, and activated the respiratory burst through hTLR2. Thus, hTLR2 is a molecular link between microbial products, apoptosis, and host defense mechanisms.

The innate immune system coordinates the inflammatory response to pathogens. To do so, cells of the innate immune system discriminate between self and nonself by receptors that identify molecules synthesized exclusively by microbes (1). These include lipopolysaccharide (LPS), peptidoglycans, lipotechoic acids, and BLPs (2). BLPs are characterized by a unique, NH2-terminal lipo-amino acid,N-acyl-S-diacylglyceryl cysteine, and are ideal targets for innate immune surveillance because they are produced by all bacteria. Although BLPs are known to activate nuclear factor–κb (NF-κB) (3), cytokine production (4), and B cell expansion (5), it is unclear how the BLP signal is transduced into an intracellular message. Candidate BLP signal transducers are Toll receptors, which are characterized by an extracellular leucine-rich repeat domain and an interleukin-1 (IL-1) receptor type 1–like intracellular signaling domain (6). In Drosophila, Toll receptors are important for resistance to microbial pathogens (7). Toll and TLRs activate homologous signal transduction pathways leading to nuclear localization of NF-κB/Rel–type transcription factors (8). Both hTLR2 (9, 10) and murine TLR4 (11) (mTLR4) are implicated in the innate response to LPS. mTLR4 does not appear to function analogously in LPS and BLP signaling. A mutation in mTLR4, which renders cells insensitive to LPS, does not abrogate BLP-induced responses in mice (11,12). Therefore, we investigated the role of hTLR2 in BLP signaling.

A population of human embryonic kidney 293 cells, which do not express hTLR2 (10, 13), were stably transfected with an epitope-tagged hTLR2 (293hTLR2) and tested for the ability to respond to BLP in an NF-κB luciferase reporter gene assay (14). Synthetic lipoprotein analogs consisting of a palmitylated version of N-acyl-S-diacylglyceryl cysteine (Pam3Cys) and a few COOH-terminal amino acids mimic the immunomodulatory effects of BLP (15). The synthetic bacterial lipopeptide, Pam3CysSerLys4(sBLP), induced expression of the luciferase reporter gene in 293hTLR2 cells, but not in the parental line (Fig. 1A). Concentrations of sBLP orEscherichia coli murein lipoprotein (MLP) as low as 200 pM activated the reporter gene (Fig. 1B). The lipo–amino acid Pam3Cys and a monoacylated derivative of sBLP (msBLP) generated by base hydrolysis (16) did not activate the luciferase reporter gene in 293 or 293hTLR2 cells (Fig. 1, A and B). This is consistent with previous observations that the acyl groups and peptide moieties of sBLP are critical for cell activation (5, 17).

Figure 1

hTLR2 mediates BLP-induced signaling. (A) An NF-κB reporter gene is activated in 293hTLR2 cells treated with sBLP. 293hTLR2 cells and parental 293 cells were transiently transfected with an NF-κB–regulated luciferase reporter construct and stimulated with sBLP (1000 ng/ml) (black bar), Pam3Cys (1000 ng/ml) (gray bar), or an equivalent dilution of sBLP diluent (H2O–0.05% HSA) (white bar). (B) Dose-response curve for NF-κB activation by BLP. 293hTLR2 cells were transiently transfected with an NF-κB–regulated luciferase reporter construct and incubated with sBLP (◊), MLP (□), or msBLP (▵). (C) hTLR2-specific mAb 2392 blocks sBLP-mediated ROS production in PBLs. PBLs were preincubated with medium alone (○), mAb 2392 (25 μg/ml of) (⋆), or an isotype control antibody (25 μg/ml) (▿) and exposed to sBLP (280 pg/ml). ROS production was measured by lucigenin-enhanced chemiluminescence every 10 min on a luminometer (33). Results are the average of two independent samples.

The respiratory burst of phagocytes generates reactive oxygen species (ROS) that contribute to antimicrobial defense (18). We isolated a monoclonal antibody specific for hTLR2 (mAb 2392) (19). This antibody, but not an isotype control, completely abrogated the generation of ROS by peripheral blood leukocytes (PBLs) after exposure to sBLP (Fig. 1C), indicating that mAb 2392 is a blocking antibody. Although a role for other TLRs cannot be excluded, these data indicate that hTLR2 is critical for BLP responses and link Toll receptors to a transcription-independent, antimicrobial defense mechanism.

After incubation with sBLP, 293hTLR2 cells began to round up, bleb, and vacuolate (Fig. 2, E and F) (20), morphologic changes that are characteristic of apoptosis (21). The induction of cell death by sBLP was more pronounced in 293 cells coexpressing hTLR2 and human CD14 (293hTLR2/hCD14) (Fig. 2, G and H). CD14 is a membrane protein that potentiates LPS and BLP signaling (22). sBLP did not affect the morphology of parental 293 cells (Fig. 2, A and B) or cells expressing only hCD14 (293hCD14) (Fig. 2, C and D). After incubation with sBLP, 293hTLR2 cells and 293hTLR2/hCD14 cells showed a significant increase in apoptosis above background (Fig. 2I). As little as 10 ng/ml of either sBLP or MLP induced apoptosis in 293hTLR2 cells (Fig. 2L). Moreover, coexpression of hCD14 and hTLR2 augmented the total number of apoptotic cells, as well as sensitizing the cells to 10-fold lower doses of sBLP or MLP (Fig. 2M). Neither compound induced apoptosis in 293 or 293hCD14 cells (Fig. 2, I to K). Consistent with the NF-κB activation assays, neither msBLP nor Pam3Cys induced apoptosis (Fig. 2, J to M). In comparison to sBLP, 103-fold higher concentrations of LPS were required to obtain similar numbers of apoptotic cells (Fig. 2, L and M). These assays were performed at serum concentrations (5%) sufficient to promote the interaction of LPS with CD14 (23). The LPS preparation was active because both sBLP and LPS induced a respiratory burst in PBLs at concentrations less than 1 ng/ml (24).

Figure 2

BLPs induce apoptosis in 293 cells expressing hTLR2. (A to H) Morphology of 293 cells and stable transfectants exposed to sBLP. 293 (A and B), 293hCD14 (C and D), 293hTLR2 (E and F), and 293hTLR2/hCD14 (G and H) cells were incubated in medium alone (A, C, E, and G) or with sBLP (100 ng/ ml) (B, D, F, and H). Cells with apoptotic morphology including rounding up (arrows), vacuolation (dashed arrow), and membrane blebbing (arrowheads) are indicated. Bar, 10 μm (A to H). (I) Time course of BLP-induced, hTLR2-mediated apoptosis. 293 (○), 293hCD14 (▵), 293hTLR2 (□), and 293hTLR2/hCD14 (◊) cells were incubated in media only or media with sBLP (100 ng/ml), TUNEL stained, and analyzed by flow cytometry. (J to M) Dose-response curves for the induction of apoptosis by BLP. 293 (J), 293hCD14 (K), 293hTLR2 (L), and 293hTLR2/hCD14 (M) cells were incubated with sBLP (◊), MLP (□), msBLP (▵), Pam3Cys (○), orShigella flexneri type 1A LPS (⋆) for 60 hours, TUNEL stained, and analyzed by flow cytometry. (N) sBLP directly induces apoptosis in cells expressing hTLR2. 293 cells were transiently transfected with GFP and the indicated plasmid, incubated with sBLP (100 ng/ml), collected, stained for cell death with annexin V, and analyzed by flow cytometry. Results are presented as the fold induction of cell death in the GFP-negative (white bars) and GFP-positive (black bars) populations relative to cells transfected with control vector (pKRN) without the addition of sBLP. Data are the mean ± SD of two independent samples. Some SDs are within the limits of the data points.

We investigated whether the induction of apoptosis by BLPs through hTLR2 was direct or indirect (25). 293 cells were transiently cotransfected with expression vectors encoding hTLR2 and green fluorescent protein (GFP). After exposure to sBLP, an induction of apoptosis was detected only in the GFP-positive population (Fig. 2N). We did not detect cell death above background without the addition of sBLP or after transfection with the control vector. Thus, BLPs induce apoptosis directly through hTLR2 and not by stimulating the secretion of a soluble mediator of apoptosis.

The transcription of hTLR2 in THP-1 cells, a monocytic cell line that expresses hTLR2 endogenously (10), was confirmed by reverse transcriptase–polymerase chain reaction (13), and its surface expression by flow cytometric analysis with mAb 2392 (24). THP-1 cells exposed to sBLP or MLP (10 ng/ml) showed a modest increase in cytotoxicity (15%) (Fig. 3A) (26). However, initial treatment of the cells with the phorbol ester PMA (phorbol 12-myristate 13-acetate) or cotreatment with the translation inhibitor cycloheximide markedly increased cytotoxicity (Fig. 3, B and C). The cells displayed morphologic features of apoptosis, including cell shrinkage and blebbing. BLP-induced cell death occurred much more rapidly in THP-1 cells (6 hours) than in 293hTLR2 or 293hTLR2/hCD14 cells (>30 hours), suggesting that monocytes express a more appropriate pathway to execute hTLR2-mediated cell death. Neither msBLP nor Pam3Cys were active in this assay (Fig. 3, A to C). One thousand–fold higher concentrations of LPS than sBLP were required to observe a cytotoxic effect in PMA- or cycloheximide-treated THP-1 cells (Fig. 3, B and C). Similar results were obtained in the presence of recombinant LPS binding protein (24). THP-1 cells were preincubated with mAb 2392 and exposed to sBLP. This antibody to hTLR2, but not an isotype control, attenuated the cell death response to sBLP (Fig. 3D). Therefore, the induction of apoptosis by sBLP requires hTLR2.

Figure 3

BLPs are cytotoxic to the human monocytic cell line THP-1. (A to C) Dose-response curves for the induction of cell death by BLPs in THP-1 cells. THP-1 cells (A) or THP-1 cells initially treated with PMA (B) or cotreated with cycloheximide (C) were incubated with sBLP (◊), MLP (□), msBLP (▵), Pam3Cys (○), or S. flexneri type 1A LPS (⋆). (D) hTLR2-specific mAb blocks sBLP-mediated cell death in THP-1 cells. THP-1 cells were initially incubated in medium (1), isotype control antibody (25 μg/ml) (2), mAb 2392 (10 μg/ml) (3), or mAb 2392 (25 μg/ml) (4) and treated with sBLP (625 pg/ml) and cycloheximide (50 μg/ml). Data are the mean ± SD of three samples. Some SDs are within the limits of the data points.

Brightbill et al. report that BLPs activate NF-κB–dependent transcription through hTLR2 (27). We describe similar findings. However, we find that BLPs can also mediate apoptosis through hTLR2. Thus, hTLR2 signals for both cell activation and apoptosis. This dual signaling capacity is precedented by tumor necrosis factor receptor–1 (TNFR1), which stimulates parallel pathways to apoptosis and NF-κB activation (28).

Although many bacterial pathogens induce apoptosis in host cells (29), the implications of this phenomenon remain elusive. BLP-induced apoptosis could be important for (i) the initiation of inflammation (30), (ii) the resolution of inflammation (31), and (iii) generating the proper signals necessary for adaptive immune responses. The observation that sBLPs are excellent adjuvants supports the third hypothesis (32). The BLP-hTLR2 apoptotic pathway emerges as a mechanism potentially fulfilling multiple roles in the genesis and progression of innate and adaptive immune responses to bacteria.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed. E-mail: zychlins{at}


View Abstract

Navigate This Article