Type I IFNs are induced during microbial infections and have well-characterized antiviral activities. TRAF3 is a signaling molecule crucial for type I IFN production and, therefore, represents a potential target for disarming immune responses. Chlamydia pneumoniae is a human pathogen that primarily infects respiratory epithelial cells; the onset of symptoms takes several weeks, and the course of infection is protracted. C. pneumoniae has also been associated with a variety of chronic inflammatory conditions. Thus, typical C. pneumoniae infections of humans are consistent with an impairment in inflammatory responses to the microorganism. We demonstrate that infection of epithelial cells with C. pneumoniae does not lead to IFN-β production. Instead, infected cells are prevented from activating IFN regulatory factor 3. This effect is mediated by C. pneumoniae–dependent degradation of TRAF3, which is independent of a functional proteasome. Hence, it is likely that C. pneumoniae expresses a unique protease targeting TRAF3-dependent immune effector mechanisms.

The respiratory epithelium represents a major site for entry of pathogenic microorganisms. Therefore, the epithelial cell lining of the respiratory tract plays a pivotal role in innate immune defense against bacterial and viral pathogens. Innate immune recognition of microbial products is an essential component of this defense and is mediated by germline-encoded pattern recognition receptors (PRRs). PRRs detect pathogen-associated molecular patterns that signal the presence of a foreign microorganism to the host TLRs. These receptors, which are members of the TNFR family, represent the best-characterized class of PRRs. TLRs are localized in the cytoplasmic or endosomal membranes; of the 10 functional TLRs identified in humans, human bronchial epithelial cells express functional TLR1–6 (1). Ligation of a TLR initiates a cascade of signaling pathways that proceed in either a MyD88-dependent (TLR1, TLR2, TLR4, TLR5, and TLR6) or TRIF-dependent (TLR3 and TLR4) manner. MyD88-dependent pathways drive the induction of various inflammatory cytokines, whereas TRIF-dependent pathways are responsible for the induction of type I IFNs in addition to inflammatory cytokines (2). Another group of PRRs include RNA helicases, retinoic acid–inducible gene-I (RIG-I), and melanoma differentiation–associated gene 5 (MDA5), which recognize viral dsRNA in the cytoplasm and lead to the production of IFN-α/-β (35).

Transduction of various signals, including those initiated from all PRRs, requires participation of TNFR-associated factors (TRAFs). TRAFs are intracellular signaling molecules that serve as both adaptor proteins and E3 ubiquitin ligases. TRAF1 and TRAF2 are constitutively associated and ubiquitously expressed together with TRAF3 and TRAF6 in most cell types, whereas expression of functional TRAF5 is mainly restricted to immune cells (6). K63-linked autoubiquitination of TRAF proteins is essential for the assembly of downstream signaling effectors. The engagement of TLR3 or TLR4 leads to TRIF-dependent K63-linked ubiquitination of TRAF3. This ubiquitinated TRAF3 is crucial for downstream activation of kinases TBK1 and IKKε that catalyze the phosphorylation of IFN regulatory factor 3 (IRF3) and subsequent induction of IFN-β (79). K63-linked ubiquitination of TRAF3 also plays a role in transduction pathways initiated from RIG-I and MDA5 by binding directly to activated mitochondrial antiviral signaling protein (MAVS) (10, 11). Thus, TRAF3 represents a key signaling molecule in multiple transduction pathways, and deficiency in this adapter molecule impairs IFN-α/IFN-β induction by TLR3, TLR4, TLR7/8, TLR9 (8, 9), RIG-I, and MDA5 (10). In addition to IRF3, type I IFNs are regulated by the transcription factor IRF7, which is activated upon stimulation of endosomal TLR7/8 and TLR9 in immune cells, leading to the generation of primarily IFN-α (10). Although the importance of type I IFNs in the innate immune response to viral infections is well characterized, their role during the course of bacterial infections is less clear.

Chlamydia pneumoniae is a Gram-negative, obligate intracellular bacterium that infects mucosal surfaces of the human respiratory tract, causing pneumonia, bronchitis, pharyngitis, and sinusitis. Epidemiological data suggest that most people are infected and reinfected throughout life (12). Therefore, C. pneumoniae represents an invader commonly encountered by respiratory defenses. The pathogen has also been associated with a variety of chronic diseases, such as reactive arthritis, sarcoidosis, asthma, chronic obstructive pulmonary disease, multiple sclerosis, Alzheimer’s disease, and atherosclerosis (13). However, the role of the pathogen in the course of chronic diseases is unknown. C. pneumoniae undergoes a developmental cycle in which two functionally and morphologically distinct cell types are recognized. The infectious cell type, which is specialized for extracellular survival and transmission, is termed the “elementary body”; the intracellular, vegetative cell type is called the “reticulate body.” Upon internalization of C. pneumoniae by a host cell, the bacterium proceeds with a developmental cycle that occurs entirely within a membrane-bound vesicle termed an “inclusion.” The chlamydial inclusion does not fuse or interact with endosomes or lysosomes during productive growth of the microorganism within an epithelial cell (1416). Conversely, C. pneumoniae cannot escape the endosomal/lysosomal fusion upon invasion of an immune cell, such as a macrophage, where the pathogen is destined for degradation (17).

In this study, we demonstrate that C. pneumoniae has evolved a unique mechanism to disarm essential aspects of innate immunity. Unlike other chlamydial species, infection of epithelial cells with C. pneumoniae does not lead to IFN-β production. Moreover, the microorganism actively suppresses type I IFN induction, because infection interferes with stimulation of cells with polyinosinic–polycytidylic acid [poly(I:C)], the synthetic analog of dsRNA. Finally, we provide systematic evidence that C. pneumoniae targets TRAF3 for degradation. Therefore, C. pneumoniae infection blocks type I IFN induction by preventing phosphorylation and nuclear translocation of IRF3.

C. pneumoniae, AR-39 purchased from American Type Culture Collection (Manassas, VA), was propagated in HeLa cells and purified using MD-76R, as previously described (18). HeLa 229 and A549 cells were also obtained from American Type Culture Collection. Subconfluent monolayers of epithelial cells were infected with C. pneumoniae, AR-39 at the indicated multiplicity of infection (MOI) by incubation with chlamydial inoculum in a 37°C incubator for 2 h with occasional agitation. Chlamydia-infected and uninfected cells were cultured in RPMI 1640 (Invitrogen, Carlsbad, CA) for HeLa cells and F12K for A549 cells supplemented with 10% FBS (Sigma, St. Louis, MO) and in the presence of 1 μg/ml cycloheximide (Sigma) for the first 30 h. No antibiotics were present in the culture medium after 30 h postinfection (p.i.). In Fig. 1A, crude C. pneumoniae inoculum was prepared as described by Maass et al. (19), and HeLa cells were infected at an MOI of 3 by centrifugation at 800 × g for 30 min (20). In experiments in which human IFN-β was quantified, uninfected and chlamydiae-infected cells were incubated in the absence of antibiotics. For inhibition of chlamydial protein synthesis, C. pneumoniae–infected cells were incubated in the presence of 200 μg/ml chloramphenicol (Sigma). Growth and MD-76R purification of C. trachomatis, serovars L2 and D were performed as previously described (21).

FIGURE 1.

C. pneumoniae does not induce IFN-β. (A) HeLa cells were infected with either crude or MD-76R–purified C. pneumoniae, AR-39 at an MOI of 3, and IFN-β released into the culture media was quantified by ELISA. Uninfected and cells infected with purified C. trachomatis, L2 (MOI of 1) were included as negative and positive controls, respectively. (B) Immunoblot analysis of crude and purified C. pneumoniae inoculum probed with anti-Hsp90 and anti-TATA box binding protein (TBBP) Abs. Total amount of chlamydial protein was controlled by probing with anti-CdsJ Abs. (C) Uninfected and A549 cells infected with C. pneumoniae (MOI of 0.5) were either mock treated or transfected at 45 h p.i. with 0.5 μg/ml of poly(I:C) [p(I:C)] for an additional 20 h. IFN-β in cell culture media was measured using ELISA. *p < 0.05, Student t test. ND, None detected.

FIGURE 1.

C. pneumoniae does not induce IFN-β. (A) HeLa cells were infected with either crude or MD-76R–purified C. pneumoniae, AR-39 at an MOI of 3, and IFN-β released into the culture media was quantified by ELISA. Uninfected and cells infected with purified C. trachomatis, L2 (MOI of 1) were included as negative and positive controls, respectively. (B) Immunoblot analysis of crude and purified C. pneumoniae inoculum probed with anti-Hsp90 and anti-TATA box binding protein (TBBP) Abs. Total amount of chlamydial protein was controlled by probing with anti-CdsJ Abs. (C) Uninfected and A549 cells infected with C. pneumoniae (MOI of 0.5) were either mock treated or transfected at 45 h p.i. with 0.5 μg/ml of poly(I:C) [p(I:C)] for an additional 20 h. IFN-β in cell culture media was measured using ELISA. *p < 0.05, Student t test. ND, None detected.

Close modal

The Abs recognizing the FLAG tag (M2) and human actin were purchased from Sigma. The anti-TRAF3–specific Abs were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and LifeSpan BioSciences (Seattle, WA). The anti–phospho-IRF3 (Ser396) and anti-TATA box binding protein (TBBP) Abs were purchased from Millipore (Temecula, CA), anti-IRF3 was purchased from Santa Cruz Biotechnology, and anti-MAVS/VISA Ab was purchased from Bethyl Laboratories (Montgomery, TX). Genus-specific anti-CdsJ Abs were described previously (22). Anti-Hsp90 and anti–phospho-IκBα (Ser32 and Ser36) mAb were obtained from BD Biosciences (San Jose, CA). Poly(I:C) was purchased from Sigma. Lactacystin and MG132 were obtained from EMD Chemicals (Billerica, MA).

Uninfected and C. pneumoniae–infected HeLa cells were transfected at 45 h p.i. with either HA-TRAF3 or FLAG-TRAF2 using Lipofectamine 2000 (Invitrogen), according to the manufacturer’s instructions. The HA-TRAF3 and FLAG-TRAF2 constructs were kindly provided by Dr. E.W. Harhaj (University of Miami, Miller School of Medicine) and Dr. S.C. Sun (Department of Microbiology and Immunology, Penn State Hershey College of Medicine). The pUNO-hIPS1 vector was purchased from InvivoGen (San Diego, CA), and cotransfections of uninfected and chlamydiae-infected HeLa cells with HA-TRAF3 and pUNO-hIPS1 were performed as described above. Transfected cell cultures were incubated in RPMI 1640 supplemented with 5% FBS for an additional 19 h. For induction experiments, uninfected and C. pneumoniae– or C. trachomatis–infected HeLa cells were transfected with poly(I:C) for the indicated periods of time using Lipofectamine 2000. Mock-treated cells were incubated in the presence of Lipofectamine 2000 without poly(I:C).

Protein samples obtained from whole-cell lysates of uninfected and C. pneumoniae– or C. trachomatis–infected cultures were harvested with 8 M urea (23) and concentrated by the addition of trichloroacetic acid (Sigma) to 10% (v/v). Collected protein pellets were analyzed by SDS-PAGE electrophoresis, followed by immunoblot analysis using appropriate primary Abs and peroxidase-conjugated secondary Ab (Sigma). Nuclear fractions of uninfected and chlamydiae-infected HeLa cells were obtained through stepwise separation of membrane, cytoplasmic, and nuclear fractions using a Subcellular Protein Fractionation Kit, according to the manufacturer’s instructions (Pierce, Rockford, IL). Polypeptide bands were visualized by development with ECL Plus Western blotting Detection Reagents (GE Healthcare, Piscataway, NJ), according to the manufacturer’s instructions.

C. pneumoniae–infected and uninfected HeLa cells cotransfected with HA-TRAF3 and pUNO-hIPS1 were lysed in RIPA buffer containing 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate, and protease inhibitors (Roche, Indianapolis, IN). Precleared cell lysates were incubated with anti-HA Abs, followed by incubation with a 50% slurry of protein A–Sepharose (Sigma). The protein A–bound immunocomplexes were subjected to Western blot analysis and probed with anti-MAVS Abs, as described above.

Uninfected HeLa cells transfected with HA-TRAF3 were harvested into TBS (20 mM Tris [pH 7.4], 150 mM NaCl) (24) and gently lysed using a Par Cell Disruption Bomb (Parr Instrument Company, Moline, IL), according to the manufacturer’s instructions. Density-gradient–purified C. pneumoniae were lysed in TBS containing 1% Nonidet P-40. Different amounts of chlamydial lysate were coincubated with a constant amount of HeLa cell lysate for 1 h at 37°C, as previously described (24). Protein samples were prepared as described above.

Culture media of uninfected or C. pneumoniae– or C. trachomatis–infected HeLa cells were collected at different time points p.i. and centrifuged at 20,000 × g for 10 min, and IFN-β was measured in supernatants using an ELISA kit (PBL, Piscataway, NJ), according to the manufacturer’s instructions. Uninfected or C. pneumoniae–infected A549 cells were either mock treated or transfected with 0.5 μg/ml poly(I:C) at 48 h, as described above. Culture media were collected 20 h posttransfection, and levels of IFN-β were quantified by ELISA. The Student t test was used to analyze data.

An IFN-β response elicited during productive infection of epithelial cells with C. pneumoniae has not been demonstrated. Buss et al. (20) reported that, in human endothelial cells, C. pneumoniae induce production of IFN-β via MAVS-dependent IRF3 and IRF7 activation during the first 16 h p.i. However, in that study, endothelial cells were infected with C. pneumoniae inoculum that had not undergone purification through diatrizoate meglumine plus diatrizoate sodium, formally known as renografin, described by Caldwell et al. (21). Instead, a crude C. pneumoniae preparation was used (20). Renografin purification of chlamydiae is routinely employed for all chlamydial species, because it removes significant amounts of cellular debris from lysed eukaryotic cells, as well as reticulate bodies, which are released with elementary bodies during sonication of infected cells at the end of the chlamydial cycle. Therefore, we compared IFN-β production during infection of HeLa cells with MD-76R (diatrizoate meglumine plus diatrizoate sodium) purified and crude C. pneumoniae inoculum using ELISA assays. Although IFN-β secretion was not detected in culture media obtained from cells infected with purified C. pneumoniae at 16, 22, or 40 h p.i., the cytokine was clearly detected in media of cultures infected with crude inoculum (Fig. 1A). MD-76R–purified C. trachomatis, L2 at 40 h p.i. was included as a positive control. It is highly likely that the crude chlamydial inoculum contains ample amounts of cellular debris, including nucleic acid, which might be responsible for activation of host cells. Indeed, unlike purified C. pneumoniae, the crude inoculum contain high levels of the eukaryotic cytoplasmic protein Hsp90 and nuclear TATA box binding protein, TBBP (Fig. 1B). The lack of IFN-β production in cells infected with purified C. pneumoniae was further confirmed in A549 lung epithelial cells at 65 h p.i. (Fig. 1C). Moreover, in C. pneumoniae–infected cells transfected with poly(I:C), the production of IFN-β detected in the culture medium was inhibited by >50% compared with uninfected poly(I:C)-stimulated cells (Fig. 1C). The percentage reduction in the levels of IFN-β corresponded to the number of infected cells, because ∼50–60% of cultures contained chlamydial inclusions (data not shown). These data demonstrate that infection of epithelial cells with purified, intact C. pneumoniae does not stimulate IFN-β production at any time p.i. and suggests that the microorganism actively impairs generation of IFN-β in infected cultures.

We further confirmed that C. pneumoniae inhibits IFN-β production in infected HeLa cells in which the infection of cells at an MOI of 1 is easily and consistently achievable. Stimulation of uninfected cells with poly(I:C) for 4 h caused significant cytotoxicity, which was detected by the presence of a large number of rounding cells compared with the mock-treated cells. This cytotoxic effect of poly(I:C) appeared to be greatly reduced in cultures infected with C. pneumoniae for 64 h (Fig. 2A). In epithelial cells, the induction of IFN-β is initiated upon phosphorylation and nuclear translocation of the transcription factor IRF3. Therefore, we examined the activation status of IRF3 in Chlamydia-infected cultures. As expected, the Ser396-specific phosphorylation of IRF3 was detected in uninfected cells transfected with poly(I:C) but not in quiescent cells. Interestingly, IRF3 phosphorylation was not detected in C. pneumoniae–infected cultures, even in cells stimulated with poly(I:C) (Fig. 2B). These data were further confirmed by probing of nuclear extracts for IRF3. Nuclear translocation of IRF3 was detected in uninfected cells and cells infected with C. trachomatis, L2 at 24 h p.i. but not in C. pneumoniae–infected cells at 64 h p.i. upon induction with poly(I:C) (Fig. 2C). C. trachomatis was included as a positive control, because infection of cultures with this chlamydial species induces expression of IFN-β (25). C. trachomatis–infected cells were harvested at 24 h p.i. because of the differences in chlamydial growth rate.

FIGURE 2.

C. pneumoniae does not activate IRF3 and protects infected cells from detrimental effects of poly(I:C). (A) Phase-contrast microscopic images of uninfected and C. pneumoniae–infected HeLa cells that were either mock treated or transfected with 4 μg/ml of poly(I:C) for 4 h. Photomicrographs were acquired using an Olympus CX41 microscope. Original magnification ×400 (40× lens). (B) Immunoblot analysis of uninfected and infected HeLa cells, treated as above for 4 or 5 h, and probed with anti–phospho-IRF3–specific Ab. The amount of protein loaded per sample was controlled by probing with anti-actin Abs. (C) C. pneumoniae (64 h p.i.)– or C. trachomatis (24 h p.i.)–infected and uninfected cultures were either mock treated or transfected with 4 μg/ml of poly(I:C) for 5 h and subjected to subcellular fractionation. Immunoblot of the nuclear fraction was probed with anti-IRF3 Ab. Anti-TATA box binding protein (TBBP) was used as a loading control.

FIGURE 2.

C. pneumoniae does not activate IRF3 and protects infected cells from detrimental effects of poly(I:C). (A) Phase-contrast microscopic images of uninfected and C. pneumoniae–infected HeLa cells that were either mock treated or transfected with 4 μg/ml of poly(I:C) for 4 h. Photomicrographs were acquired using an Olympus CX41 microscope. Original magnification ×400 (40× lens). (B) Immunoblot analysis of uninfected and infected HeLa cells, treated as above for 4 or 5 h, and probed with anti–phospho-IRF3–specific Ab. The amount of protein loaded per sample was controlled by probing with anti-actin Abs. (C) C. pneumoniae (64 h p.i.)– or C. trachomatis (24 h p.i.)–infected and uninfected cultures were either mock treated or transfected with 4 μg/ml of poly(I:C) for 5 h and subjected to subcellular fractionation. Immunoblot of the nuclear fraction was probed with anti-IRF3 Ab. Anti-TATA box binding protein (TBBP) was used as a loading control.

Close modal

In epithelial cells transfected with poly(I:C), the signaling cascade leading to IRF3 activation is primarily initiated from MDA5, which recognizes this dsRNA analog (26). Ligation of MDA5 to dsRNA leads to the recruitment of MAVS, followed by activation of TRAF3 through direct association with MAVS (10, 27). Formation of TRAF3-dependent K63-linked polyubiquitin chains in induced cells is essential for activation of TBK1-IKKε kinases that phosphorylate IRF3. Conversely, in cells treated with poly(I:C) in the absence of a transfecting reagent, the dsRNA enters the cell via endocytosis and initiates the TLR3-signaling pathway, leading to the activation of IRF3 and NF-κB (reviewed in Ref. 28). Interestingly, unlike with C. trachomatis, infection of cultures with C. pneumoniae in HeLa cells treated with poly(I:C) for 1 h in the absence of a transfecting reagent prevented phosphorylation of the cytoplasmic NF-κB inhibitor, IκBα, indicating that the pathogen is also capable of inhibiting signaling initiated from TLR3 (Fig. 3A). The transduction pathways initiated from both MDA5 and TLR3 converge at one critical signaling molecule, TRAF3. Indeed, in C. pneumoniae–infected and uninfected cells cotransfected with HA-TRAF3 and pUNO-hIPS1, TRAF3 was immunoprecipitated together with MAVS from uninfected cells but not from C. pneumoniae–infected cultures. The coprecipitation of TRAF3 and MAVS was detected in Chlamydia-infected cells treated with chloramphenicol, indicating that bacterial protein synthesis is required to prevent MAVS binding to TRAF3 (Supplemental Fig. 1). Therefore, we analyzed TRAF3 levels during C. pneumoniae infection. In A549 cells, which contain detectable amounts of endogenous TRAF3, the infection with C. pneumoniae resulted in lower levels of the signaling molecule as early as 24 h p.i. compared with uninfected cells. This decrease in TRAF3 was detected throughout the course of infection (Fig. 3B). These results indicate that C. pneumoniae targets host TRAF3 during infection.

FIGURE 3.

C. pneumoniae decrease TRAF3 in infected cells. (A) Uninfected and C. pneumoniae– or C. trachomatis–infected HeLa cells were either mock treated or treated with 2 μg/ml of poly(I:C) for 1 h in the absence of a transfecting reagent. Immunoblots of cell lysates were probed with anti-S32 and S36 phospho-specific IκBα mAb. Amount of protein loaded per sample was controlled by probing with anti-actin Abs. (B) Uninfected A549 cells or cells infected with C. pneumoniae were harvested at 16, 24, 36, 48, or 64 h p.i. and probed with anti-TRAF3 or TRAF2 Abs. Probing with anti-actin mAb was used as a loading control.

FIGURE 3.

C. pneumoniae decrease TRAF3 in infected cells. (A) Uninfected and C. pneumoniae– or C. trachomatis–infected HeLa cells were either mock treated or treated with 2 μg/ml of poly(I:C) for 1 h in the absence of a transfecting reagent. Immunoblots of cell lysates were probed with anti-S32 and S36 phospho-specific IκBα mAb. Amount of protein loaded per sample was controlled by probing with anti-actin Abs. (B) Uninfected A549 cells or cells infected with C. pneumoniae were harvested at 16, 24, 36, 48, or 64 h p.i. and probed with anti-TRAF3 or TRAF2 Abs. Probing with anti-actin mAb was used as a loading control.

Close modal

The possibility that C. pneumoniae may degrade TRAF3 was investigated in uninfected and chlamydiae-infected HeLa cells ectopically expressing HA-TRAF3. It was demonstrated that overexpression of TRAF3 does not activate any known transduction pathway (29). Whole-cell lysates were harvested at 64 h p.i. for C. pneumoniae–infected cells and at 24 h p.i. for C. trachomatis, L2–infected cells and probed with TRAF3-specific Ab. In contrast with uninfected and C. trachomatis–infected cells, this signaling molecule was not readily detectable in C. pneumoniae–infected cultures (Fig. 4A), raising the possibility that TRAF3 is degraded. We performed transcriptional analysis of TRAF2, TRAF3, and TRAF5 in uninfected and C. pneumoniae–infected HeLa cells to rule out the possibility of alterations at the transcriptional level. Indeed, the mRNA message of all of the analyzed genes, including TRAF3, was not lower than in uninfected cultures (Supplemental Fig. 2). Furthermore, we examined the overexpression of TRAF3 in C. trachomatis, serovar D–infected cells; the levels of the signaling molecule remained unaltered (Supplemental Fig. 3). Therefore, the apparent TRAF3 degradation was a C. pneumoniae–specific phenomenon requiring bacterial translation (Fig. 4A). The levels of TRAF3 degradation correlated with the number of cells infected with C. pneumoniae, because more TRAF3 was detected at a lower MOI (Fig. 4B). Conversely, the ectopic expression of TRAF2 in HeLa cells was unaffected by C. pneumoniae or C. trachomatis (Fig. 4C).

FIGURE 4.

C. pneumoniae degrade host TRAF3. (A) Whole-cell lysates of uninfected and C. pneumoniae–infected at 64 h p.i., or C. trachomatis–infected at 24 h p.i. HeLa cells transfected with HA-TRAF3. Immunoblots were probed with anti-TRAF3 Ab. (B) C. pneumoniae cultures infected at an MOI of 1 or 0.5 and uninfected cells were transfected with HA-TRAF3. Resolved material was probed with anti-TRAF3 mAb. (C) Uninfected and chlamydiae-infected HeLa cells transfected with Flag-TRAF2 were probed with anti-Flag mAb. Immunoblots probed with anti-actin were used as a loading control.

FIGURE 4.

C. pneumoniae degrade host TRAF3. (A) Whole-cell lysates of uninfected and C. pneumoniae–infected at 64 h p.i., or C. trachomatis–infected at 24 h p.i. HeLa cells transfected with HA-TRAF3. Immunoblots were probed with anti-TRAF3 Ab. (B) C. pneumoniae cultures infected at an MOI of 1 or 0.5 and uninfected cells were transfected with HA-TRAF3. Resolved material was probed with anti-TRAF3 mAb. (C) Uninfected and chlamydiae-infected HeLa cells transfected with Flag-TRAF2 were probed with anti-Flag mAb. Immunoblots probed with anti-actin were used as a loading control.

Close modal

TRAF3 is an adaptor protein that also acts as an E3 ubiquitin ligase, a function essential for K63-linked autoubiquitination (30), whereas K48-linked ubiquitylation of TRAF3 marks this signaling molecule for proteasome-dependent degradation (31, 32). Interestingly, only K63-linked polyubiquitin chains are stimulated directly by TRAF3, whereas other cytoplasmic E3 ligases execute K48-linked ubiquitination (30, 33). It was recently demonstrated that TRAF3 degradation in A549 cells is mediated by the E3 ligase Triad3A and is prevented by the proteasome inhibitors lactacystin and MG132 (33). Therefore, we investigated whether the degradation of TRAF3 during infection of epithelial cells with C. pneumoniae depends on the host proteasome. Uninfected and C. pneumoniae–infected HeLa cells ectopically expressing TRAF3 were treated with lactacystin for 15 h. None of the lactacystin concentrations tested prevented TRAF3 degradation in C. pneumoniae–infected cells (Fig. 5A). We were unable to exploit the alternative proteasome inhibitor MG132 in this assay, because treatment of cultures with MG132 for 15 h resulted in toxicity and cell death (data not shown). Instead, we used a cell-free lysates approach (24) to further investigate the mechanism of TRAF3 degradation. A total of 5 × 106 uninfected HeLa cells expressing TRAF3 was lysed using a Parr Cell Disruption Bomb, and the cell lysates were coincubated for 1 h with different amounts of lysates prepared from purified C. pneumoniae. Even in the presence of MG132, chlamydial lysates were sufficient to reduce the abundance of TRAF3 (Fig. 5B). These data are consistent with changes in TRAF3 abundance not being manifested at the transcriptional level. More directly, they suggest that the protease targeting TRAF3 is chlamydial, because degradation of TRAF3 is not impaired by two commonly used inhibitors of the eukaryotic 26S proteasome, and C. pneumoniae lysates are sufficient to yield reduced detection of TRAF3.

FIGURE 5.

Degradation of TRAF3 is not inhibited by lactacystin or MG132. (A) Uninfected HeLa cells and cells infected with C. pneumoniae were transfected with HA-TRAF3 and treated with different concentrations of lactacystin for 15 h. Immunoblots were probed with anti-TRAF3 and anti-TRAF2 Abs. (B) Uninfected cells transfected with HA-TRAF3 were disrupted using a Parr Cell Disruption Bomb (1500 psi/5 min). HeLa cell lysates were coincubated with various amounts of C. pneumoniae lysates in the presence or absence of 20 μM MG132. The level of TRAF3 degradation was detected by probing immunoblots with anti-TRAF3 mAb. Probing with anti-actin was used as a loading control.

FIGURE 5.

Degradation of TRAF3 is not inhibited by lactacystin or MG132. (A) Uninfected HeLa cells and cells infected with C. pneumoniae were transfected with HA-TRAF3 and treated with different concentrations of lactacystin for 15 h. Immunoblots were probed with anti-TRAF3 and anti-TRAF2 Abs. (B) Uninfected cells transfected with HA-TRAF3 were disrupted using a Parr Cell Disruption Bomb (1500 psi/5 min). HeLa cell lysates were coincubated with various amounts of C. pneumoniae lysates in the presence or absence of 20 μM MG132. The level of TRAF3 degradation was detected by probing immunoblots with anti-TRAF3 mAb. Probing with anti-actin was used as a loading control.

Close modal

Type I IFNs are pleiotropic cytokines secreted by a variety of cells and are known to act in an autocrine, as well as paracrine, manner. Type I IFNs are primarily composed of multiple IFN-α subtypes and a single IFN-β and represent an essential component of the innate immune response to viral infections in humans. Secreted IFNs mediate antimicrobial activities by binding the IFN-α/-β receptor (IFNAR), leading to heterodimerization of STAT1 and STAT2. The heterodimer forms a complex with IRF9 and is translocated into the nucleus where it interacts with the IFN-stimulated response element. This ultimately drives the induction of transcription of type I IFN–responsive genes. The importance of type I IFNs in the innate response to numerous viruses is well established. Although type I IFNs are induced, the role of this cytokine during the course of bacterial infections is less-well defined.

TRAF3 is a key signaling molecule involved in multiple pathways that lead to generation of type I IFNs and is essential for the innate immune response to a broad range of infections. K63-linked autoubiquitination of TRAF3 is crucial for activation of IRF3 upon induction of cells with numerous stimuli. However, the signaling molecule also plays an important role as a negative regulator of various transduction pathways (31). For instance, in quiescent cells, TRAF3 binds to NF-κB inducing kinase. Upon induction of the noncanonical NF-κB pathway, TRAF3 undergoes nonself-mediated K48-linked ubiquitination and proteosomal degradation, releasing NF-κB inducing kinase, which results in phosphorylation of IKKα and p100 (NF-κB2) processing (34). Furthermore, degradative ubiquitination of TRAF3 is also required for MyD88-dependent MAPK activation (30). Although C. pneumoniae inhibits induction of IFN-β upon MDA5 and TLR3 engagement (Fig. 6), it is highly likely that by targeting TRAF3, the pathogen modulates a great number of transduction pathways within its eukaryotic host.

FIGURE 6.

C. pneumoniae inhibits multiple signaling pathways by degradation of TRAF3. Upon stimulation of either MAVS- or TRIF-dependent signaling pathways, C. pneumoniae interferes with the signaling cascade by targeting host TRAF3. This prevents transduction of the inducing signal to IRF3, thereby suppressing the IFN-β response.

FIGURE 6.

C. pneumoniae inhibits multiple signaling pathways by degradation of TRAF3. Upon stimulation of either MAVS- or TRIF-dependent signaling pathways, C. pneumoniae interferes with the signaling cascade by targeting host TRAF3. This prevents transduction of the inducing signal to IRF3, thereby suppressing the IFN-β response.

Close modal

The majority of data describing chlamydial infections and IFN production comes from in vitro and in vivo studies with Chlamydia muridarum. C. muridarum is a murine pathogen that has been established as a mouse model of genital infection because it shares many aspects of acute genital infection with C. trachomatis in women (35). The production of type I IFNs in mice infected with C. muridarum appeared to be more beneficial to the microorganism than to the host. IFNAR−/− mice were able to resolve genital infection faster and displayed reduced pathology in the oviduct compared with wild-type (WT) mice (36). In C. muridarum–infected murine macrophages, the expression of IFN-β was induced in a MyD88-dependent manner, yet the survival of chlamydiae in MyD88−/− macrophages was comparable to chlamydial survival in macrophages obtained from WT mice (37). Conversely, IFN-β production in mouse oviduct epithelial cells was primarily TLR3 dependent (i.e., MyD88 independent) (38). The production of IFN-β was also described in McCoy cells infected with C. trachomatis, serovar L2 (39), and Lad et al. (25) detected significant upregulation of IFN-β gene expression in C. trachomatis–infected HeLa cells. Unlike with C. muridarum, no difference in chlamydial burden was detected in lungs of WT, IFNAR−/−, and IRF3−/− mice infected with C. pneumoniae (40). A great number of studies described diverse effects of C. pneumoniae on various cell types in vitro, including the production of IFN-β in endothelial cells (20). However, it is absolutely essential to clearly distinguish whether the observed effects are caused directly by the pathogen or by cellular debris that may enter the host cell together with the bacterium. Furthermore, it is important to differentiate between immune responses to productive chlamydial growth and chlamydial degradation when C. pneumoniae cannot avoid fusion with the lysosomal pathway within an infected cell (17).

In this study, we demonstrated that during productive infection of human epithelial cells with purified, intact C. pneumoniae, the pathogen does not induce production of IFN-β. Furthermore, C. pneumoniae induces species-specific degradation of the host signaling molecule TRAF3, thereby preventing activation of IFR3. The impaired transduction pathway initiated from MDA5 was detected in epithelial cell lines, such as HeLa, A549, but also in HEK 293 cells (data not shown), all of which support C. pneumoniae growth, and the levels of inhibition corresponded closely to the number of infected cells. Chlamydial protein synthesis was required for all of the events caused by the bacterium. In addition, the treatment of C. pneumoniae–infected cells with the proteasomal inhibitors lactacystin and MG132 did not prevent TRAF3 degradation; coincubation of chlamydial and host cell lysates resulted in degradation of the signaling molecule, suggesting that C. pneumoniae utilizes a species-specific protease that targets TRAF3. Although chlamydiae reside within membrane-bound inclusions during their intracellular growth, insertion of chlamydial proteins into the inclusion membrane, as well as secretion of chlamydial effectors into the host cytoplasm, is essential for successful exploitation of the eukaryotic cell. Chlamydiae use sophisticated strategies to evade host immune responses. C. trachomatis and C. pneumoniae secrete a serine protease designated chlamydial protease/proteasome–like activity factor (CPAF) into the host cytosol. CPAF is responsible for degradation of several eukaryotic components, but it primarily targets the regulatory factor X5 and upstream stimulation factor 1, both of which are required for MHC Ag expression (41). Moreover, chlamydiae employ a Tail-specific protease that cleaves the p65/RelA subunit of NF-κB1 into 40- and ∼22-kDa fragments, preventing NF-κB activation during chlamydial infection (24). Due to inability of CPAF and Tail-specific protease to cleave their respective targets in 8 M urea, the target specificity of these two chlamydial proteases was recently questioned (23). We detected the degradation of TRAF3 even in the presence of 8 M urea (Figs. 3, 4), which suggests that targeting of the signaling molecule is not an artifact of postlysis degradation. Our data represent a critical shift in understanding C. pneumoniae pathogenesis. Although it is postulated that the pathogen elicits numerous proinflammatory responses by host cells, we propose that, during productive infection with purified C. pneumoniae, the bacteria actively modify and impair the host immune system, which prevents or significantly delays chlamydial recognition and efficient elimination. Indeed, C. pneumoniae seems to utilize species-specific strategies that are comparatively more efficient than those used by C. trachomatis to avoid or delay recognition by the host innate immunity. It was established that C. pneumoniae intercepts the signaling pathway from IL-17R in IL-17–stimulated epithelial cells by sequestering the adaptor molecule Act1 to the chlamydial inclusion membrane (42). The degradation of TRAF3 represents another unique tactic of C. pneumoniae that may significantly contribute to protracted or asymptomatic respiratory infections and/or chronic infections associated with this pathogen.

We thank Drs. H. Betts-Hampikian, K. Mueller, and N. Shembade for critical review of the manuscript. We also thank Dr. E. Harhaj for FLAG-TRAF2 and Dr. S.C. Sun for the HA-TRAF3 construct.

This work was supported by Public Health Service Grant AI072126 from the National Institutes of Health (to K.A.F.).

The online version of this article contains supplemental material.

Abbreviations used in this article:

CPAF

chlamydial protease/proteasome–like activity factor

IFNAR

IFN-α/-β receptor

IRF3

IFN regulatory factor 3

MAVS

mitochondrial antiviral signaling protein

MDA5

melanoma differentiation–associated gene 5

MOI

multiplicity of infection

p.i.

postinfection

poly(I:C)

polyinosinic–polycytidylic acid

PRR

pattern recognition receptor

RIG-I

retinoic acid–inducible gene-I

TRAF

TNFR-associated factor

WT

wild-type.

1
Mayer
A. K.
,
Muehmer
M.
,
Mages
J.
,
Gueinzius
K.
,
Hess
C.
,
Heeg
K.
,
Bals
R.
,
Lang
R.
,
Dalpke
A. H.
.
2007
.
Differential recognition of TLR-dependent microbial ligands in human bronchial epithelial cells.
J. Immunol.
178
:
3134
3142
.
2
Kawai
T.
,
Akira
S.
.
2010
.
The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors.
Nat. Immunol.
11
:
373
384
.
3
Andrejeva
J.
,
Childs
K. S.
,
Young
D. F.
,
Carlos
T. S.
,
Stock
N.
,
Goodbourn
S.
,
Randall
R. E.
.
2004
.
The V proteins of paramyxoviruses bind the IFN-inducible RNA helicase, mda-5, and inhibit its activation of the IFN-beta promoter.
Proc. Natl. Acad. Sci. USA
101
:
17264
17269
.
4
Yoneyama
M.
,
Kikuchi
M.
,
Natsukawa
T.
,
Shinobu
N.
,
Imaizumi
T.
,
Miyagishi
M.
,
Taira
K.
,
Akira
S.
,
Fujita
T.
.
2004
.
The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses.
Nat. Immunol.
5
:
730
737
.
5
Kato
H.
,
Sato
S.
,
Yoneyama
M.
,
Yamamoto
M.
,
Uematsu
S.
,
Matsui
K.
,
Tsujimura
T.
,
Takeda
K.
,
Fujita
T.
,
Takeuchi
O.
,
Akira
S.
.
2005
.
Cell type-specific involvement of RIG-I in antiviral response.
Immunity
23
:
19
28
.
6
Häcker
H.
,
Tseng
P. H.
,
Karin
M.
.
2011
.
Expanding TRAF function: TRAF3 as a tri-faced immune regulator.
Nat. Rev. Immunol.
11
:
457
468
.
7
Fitzgerald
K. A.
,
McWhirter
S. M.
,
Faia
K. L.
,
Rowe
D. C.
,
Latz
E.
,
Golenbock
D. T.
,
Coyle
A. J.
,
Liao
S. M.
,
Maniatis
T.
.
2003
.
IKKepsilon and TBK1 are essential components of the IRF3 signaling pathway.
Nat. Immunol.
4
:
491
496
.
8
Oganesyan
G.
,
Saha
S. K.
,
Guo
B.
,
He
J. Q.
,
Shahangian
A.
,
Zarnegar
B.
,
Perry
A.
,
Cheng
G.
.
2006
.
Critical role of TRAF3 in the Toll-like receptor-dependent and -independent antiviral response.
Nature
439
:
208
211
.
9
Häcker
H.
,
Karin
M.
.
2006
.
Regulation and function of IKK and IKK-related kinases.
Sci. STKE
2006
:
re13
.
10
Saha
S. K.
,
Pietras
E. M.
,
He
J. Q.
,
Kang
J. R.
,
Liu
S. Y.
,
Oganesyan
G.
,
Shahangian
A.
,
Zarnegar
B.
,
Shiba
T. L.
,
Wang
Y.
,
Cheng
G.
.
2006
.
Regulation of antiviral responses by a direct and specific interaction between TRAF3 and Cardif.
EMBO J.
25
:
3257
3263
.
11
Paz
S.
,
Vilasco
M.
,
Werden
S. J.
,
Arguello
M.
,
Joseph-Pillai
D.
,
Zhao
T.
,
Nguyen
T. L.
,
Sun
Q.
,
Meurs
E. F.
,
Lin
R.
,
Hiscott
J.
.
2011
.
A functional C-terminal TRAF3-binding site in MAVS participates in positive and negative regulation of the IFN antiviral response.
Cell Res.
21
:
895
910
.
12
Kuo
C. C.
,
Jackson
L. A.
,
Campbell
L. A.
,
Grayston
J. T.
.
1995
.
Chlamydia pneumoniae (TWAR).
Clin. Microbiol. Rev.
8
:
451
461
.
13
Campbell
L. A.
,
Kuo
C. C.
,
Grayston
J. T.
.
1998
.
Chlamydia pneumoniae and cardiovascular disease.
Emerg. Infect. Dis.
4
:
571
579
.
14
Schramm
N.
,
Wyrick
P. B.
.
1995
.
Cytoskeletal requirements in Chlamydia trachomatis infection of host cells.
Infect. Immun.
63
:
324
332
.
15
Heinzen
R. A.
,
Scidmore
M. A.
,
Rockey
D. D.
,
Hackstadt
T.
.
1996
.
Differential interaction with endocytic and exocytic pathways distinguish parasitophorous vacuoles of Coxiella burnetii and Chlamydia trachomatis.
Infect. Immun.
64
:
796
809
.
16
Wolf
K.
,
Hackstadt
T.
.
2001
.
Sphingomyelin trafficking in Chlamydia pneumoniae-infected cells.
Cell. Microbiol.
3
:
145
152
.
17
Wolf
K.
,
Fischer
E.
,
Hackstadt
T.
.
2005
.
Degradation of Chlamydia pneumoniae by peripheral blood monocytic cells.
Infect. Immun.
73
:
4560
4570
.
18
Wolf
K.
,
Fischer
E.
,
Hackstadt
T.
.
2000
.
Ultrastructural analysis of developmental events in Chlamydia pneumoniae-infected cells.
Infect. Immun.
68
:
2379
2385
.
19
Maass
M.
,
Bartels
C.
,
Engel
P. M.
,
Mamat
U.
,
Sievers
H. H.
.
1998
.
Endovascular presence of viable Chlamydia pneumoniae is a common phenomenon in coronary artery disease.
J. Am. Coll. Cardiol.
31
:
827
832
.
20
Buss
C.
,
Opitz
B.
,
Hocke
A. C.
,
Lippmann
J.
,
van Laak
V.
,
Hippenstiel
S.
,
Krüll
M.
,
Suttorp
N.
,
Eitel
J.
.
2010
.
Essential role of mitochondrial antiviral signaling, IFN regulatory factor (IRF)3, and IRF7 in Chlamydophila pneumoniae-mediated IFN-beta response and control of bacterial replication in human endothelial cells.
J. Immunol.
184
:
3072
3078
.
21
Caldwell
H. D.
,
Kromhout
J.
,
Schachter
J.
.
1981
.
Purification and partial characterization of the major outer membrane protein of Chlamydia trachomatis.
Infect. Immun.
31
:
1161
1176
.
22
Fields
K. A.
,
Mead
D. J.
,
Dooley
C. A.
,
Hackstadt
T.
.
2003
.
Chlamydia trachomatis type III secretion: evidence for a functional apparatus during early-cycle development.
Mol. Microbiol.
48
:
671
683
.
23
Chen
A. L.
,
Johnson
K. A.
,
Lee
J. K.
,
Sütterlin
C.
,
Tan
M.
.
2012
.
CPAF: a Chlamydial protease in search of an authentic substrate.
PLoS Pathog.
8
:
e1002842
.
24
Lad
S. P.
,
Li
J.
,
da Silva Correia
J.
,
Pan
Q.
,
Gadwal
S.
,
Ulevitch
R. J.
,
Li
E.
.
2007
.
Cleavage of p65/RelA of the NF-kappaB pathway by Chlamydia.
Proc. Natl. Acad. Sci. USA
104
:
2933
2938
.
25
Lad
S. P.
,
Fukuda
E. Y.
,
Li
J.
,
de la Maza
L. M.
,
Li
E.
.
2005
.
Up-regulation of the JAK/STAT1 signal pathway during Chlamydia trachomatis infection.
J. Immunol.
174
:
7186
7193
.
26
Kato
H.
,
Takeuchi
O.
,
Sato
S.
,
Yoneyama
M.
,
Yamamoto
M.
,
Matsui
K.
,
Uematsu
S.
,
Jung
A.
,
Kawai
T.
,
Ishii
K. J.
, et al
.
2006
.
Differential roles of MDA5 and RIG-I helicases in the recognition of RNA viruses.
Nature
441
:
101
105
.
27
Tang
E. D.
,
Wang
C. Y.
.
2009
.
MAVS self-association mediates antiviral innate immune signaling.
J. Virol.
83
:
3420
3428
.
28
Vercammen
E.
,
Staal
J.
,
Beyaert
R.
.
2008
.
Sensing of viral infection and activation of innate immunity by toll-like receptor 3.
Clin. Microbiol. Rev.
21
:
13
25
.
29
Xu
Y.
,
Cheng
G.
,
Baltimore
D.
.
1996
.
Targeted disruption of TRAF3 leads to postnatal lethality and defective T-dependent immune responses.
Immunity
5
:
407
415
.
30
Tseng
P. H.
,
Matsuzawa
A.
,
Zhang
W.
,
Mino
T.
,
Vignali
D. A.
,
Karin
M.
.
2010
.
Different modes of ubiquitination of the adaptor TRAF3 selectively activate the expression of type I interferons and proinflammatory cytokines.
Nat. Immunol.
11
:
70
75
.
31
Matsuzawa
A.
,
Tseng
P. H.
,
Vallabhapurapu
S.
,
Luo
J. L.
,
Zhang
W.
,
Wang
H.
,
Vignali
D. A.
,
Gallagher
E.
,
Karin
M.
.
2008
.
Essential cytoplasmic translocation of a cytokine receptor-assembled signaling complex.
Science
321
:
663
668
.
32
Vallabhapurapu
S.
,
Matsuzawa
A.
,
Zhang
W.
,
Tseng
P. H.
,
Keats
J. J.
,
Wang
H.
,
Vignali
D. A.
,
Bergsagel
P. L.
,
Karin
M.
.
2008
.
Nonredundant and complementary functions of TRAF2 and TRAF3 in a ubiquitination cascade that activates NIK-dependent alternative NF-kappaB signaling.
Nat. Immunol.
9
:
1364
1370
.
33
Nakhaei
P.
,
Mesplede
T.
,
Solis
M.
,
Sun
Q.
,
Zhao
T.
,
Yang
L.
,
Chuang
T. H.
,
Ware
C. F.
,
Lin
R.
,
Hiscott
J.
.
2009
.
The E3 ubiquitin ligase Triad3A negatively regulates the RIG-I/MAVS signaling pathway by targeting TRAF3 for degradation.
PLoS Pathog.
5
:
e1000650
.
34
Zarnegar
B. J.
,
Wang
Y.
,
Mahoney
D. J.
,
Dempsey
P. W.
,
Cheung
H. H.
,
He
J.
,
Shiba
T.
,
Yang
X.
,
Yeh
W. C.
,
Mak
T. W.
, et al
.
2008
.
Noncanonical NF-kappaB activation requires coordinated assembly of a regulatory complex of the adaptors cIAP1, cIAP2, TRAF2 and TRAF3 and the kinase NIK.
Nat. Immunol.
9
:
1371
1378
.
35
Farris
C. M.
,
Morrison
R. P.
.
2011
.
Vaccination against Chlamydia genital infection utilizing the murine C. muridarum model.
Infect. Immun.
79
:
986
996
.
36
Nagarajan
U. M.
,
Prantner
D.
,
Sikes
J. D.
,
Andrews
C. W.
 Jr.
,
Goodwin
A. M.
,
Nagarajan
S.
,
Darville
T.
.
2008
.
Type I interferon signaling exacerbates Chlamydia muridarum genital infection in a murine model.
Infect. Immun.
76
:
4642
4648
.
37
Nagarajan
U. M.
,
Ojcius
D. M.
,
Stahl
L.
,
Rank
R. G.
,
Darville
T.
.
2005
.
Chlamydia trachomatis induces expression of IFN-gamma-inducible protein 10 and IFN-beta independent of TLR2 and TLR4, but largely dependent on MyD88.
J. Immunol.
175
:
450
460
.
38
Derbigny
W. A.
,
Johnson
R. M.
,
Toomey
K. S.
,
Ofner
S.
,
Jayarapu
K.
.
2010
.
The Chlamydia muridarum-induced IFN-β response is TLR3-dependent in murine oviduct epithelial cells.
J. Immunol.
185
:
6689
6697
.
39
Devitt
A.
,
Lund
P. A.
,
Morris
A. G.
,
Pearce
J. H.
.
1996
.
Induction of alpha/beta interferon and dependent nitric oxide synthesis during Chlamydia trachomatis infection of McCoy cells in the absence of exogenous cytokine.
Infect. Immun.
64
:
3951
3956
.
40
Rothfuchs
A. G.
,
Trumstedt
C.
,
Mattei
F.
,
Schiavoni
G.
,
Hidmark
A.
,
Wigzell
H.
,
Rottenberg
M. E.
.
2006
.
STAT1 regulates IFN-alpha beta- and IFN-gamma-dependent control of infection with Chlamydia pneumoniae by nonhemopoietic cells.
J. Immunol.
176
:
6982
6990
.
41
Zhong
G.
,
Fan
P.
,
Ji
H.
,
Dong
F.
,
Huang
Y.
.
2001
.
Identification of a chlamydial protease-like activity factor responsible for the degradation of host transcription factors.
J. Exp. Med.
193
:
935
942
.
42
Wolf
K.
,
Plano
G. V.
,
Fields
K. A.
.
2009
.
A protein secreted by the respiratory pathogen Chlamydia pneumoniae impairs IL-17 signalling via interaction with human Act1.
Cell. Microbiol.
11
:
769
779
.

The authors have no financial conflicts of interest.