Activation of liver X receptors and retinoid X receptors prevents ...

4 downloads 0 Views 514KB Size Report
Dec 21, 2004 - Nuclear receptors are ligand-dependent transcription factors ...... Akgul, C., Moulding, D. A. & Edwards, S. W. (2001) FEBS Lett 487, 318–322.
Activation of liver X receptors and retinoid X receptors prevents bacterial-induced macrophage apoptosis Annabel F. Valledor*, Li-Chung Hsu†, Sumito Ogawa*, Dominique Sawka-Verhelle*, Michael Karin†, and Christopher K. Glass*‡§ Departments of *Cellular and Molecular Medicine, †Pharmacology, and ‡Medicine, University of California at San Diego, 9500 Gilman Drive, La Jolla, CA 92093

Microbe–macrophage interactions play a central role in the pathogenesis of many infections. The ability of some bacterial pathogens to induce macrophage apoptosis has been suggested to contribute to their ability to elude innate immune responses and successfully colonize the host. Here, we provide evidence that activation of liver X receptors (LXRs) and retinoid X receptors (RXRs) inhibits apoptotic responses of macrophages to macrophage colony-stimulating factor (M-CSF) withdrawal and several inducers of apoptosis. In addition, combined activation of LXR and RXR protected macrophages from apoptosis caused by infection with Bacillus anthracis, Escherichia coli, and Salmonella typhimurium. Expression-profiling studies demonstrated that LXR and RXR agonists induced the expression of antiapoptotic regulators, including AIM兾 CT2, Bcl-XL, and Birc1a. Conversely, LXR and RXR agonists inhibited expression of proapoptotic regulators and effectors, including caspases 1, 4兾11, 7, and 12; Fas ligand; and Dnase1l3. The combination of LXR and RXR agonists was more effective than either agonist alone at inhibiting apoptosis in response to various inducers of apoptosis, and it acted synergistically to induce expression of AIM兾CT2. Inhibition of AIM兾CT2 expression in response to LXR兾RXR agonists partially reversed their antiapoptotic effects. These findings reveal unexpected roles of LXRs and RXRs in the control of macrophage survival and raise the possibility that LXR兾RXR agonists may be exploited to enhance innate immunity to bacterial pathogens that induce apoptotic programs as a strategy for evading host responses. oxysterol 兩 transcription

M

acrophages serve essential functions as regulators of immunity and homeostasis (1, 2). As participants in native immunity, macrophages phagocytose and kill invading microorganisms and elaborate signaling molecules that amplify acute inflammatory responses. Macrophages also contribute to acquired immune responses by means of specialized functions that include antigen presentation and regulation of T cell responses. Thus, regulation of macrophage differentiation and survival is critical to the overall control of the magnitude, duration, and characteristics of immune responses. Programmed cell death, or apoptosis, of lymphocyte and myeloid cells is regulated tightly through cell death receptor and mitochondrial pathways to limit amplification of immune responses and facilitate resolution of inflammation (3). Apoptosis and survival pathways are also targeted by pathogens as a means of either escaping immune surveillance or establishing residence within host cells (4). The inhibition of macrophage apoptosis may offer a strategy for augmenting innate immunity to highly virulent bacterial pathogens, such as Bacillus anthracis, Yersinia pestis, Salmonella spp., and Shigella flexneri, which have evolved various ways to kill host macrophages. The execution of all forms of programmed cell death involves the proteolytic activation of a cascade of intracellular cysteine proteases, known as caspases. Downstream effector caspases cleave specific protein targets and mediate the www.pnas.org兾cgi兾doi兾10.1073兾pnas.0407749101

deliberate disassembly of the cell into apoptotic bodies (5). A number of regulators of apoptosis function upstream and downstream of caspases by either promoting or suppressing their protease activities. For example, antiapoptotic members of the Bcl2 family act, at least in part, to preserve mitochondrial integrity and function (including its transmembrane potential, calcium-buffering capacity, and respiration efficiency), and they prevent the release of proapoptotic components. Other members of the Bcl2 family have an opposite effect, and they mediate mitochondrial dysfunction and eventual release of proapoptotic mediators (reviewed in ref. 6). One possible approach to the inhibition of macrophage apoptosis may involve a manipulation of the expression of such proteins. Nuclear receptors are ligand-dependent transcription factors that regulate diverse aspects of development and homeostasis (7). Several members of this family influence immune responses by activating or repressing cell-specific programs of gene expression in myeloid and兾or lymphoid cells (8). For example, the glucocorticoid receptor exerts potent antiinflammatory effects partly by means of its ability to inhibit the actions of proinflammatory transcription factors, such as AP-1 and NF-␬〉 and induce apoptosis of lymphocytes (9, 10). Liver X receptors (LXRs) represent a subset of the nuclear receptor superfamily that are regulated by oxidized forms of cholesterol (oxysterols) and intermediate products of the cholesterol biosynthetic pathway (11, 12). Two LXR isoforms have been identified, LXR␣ (NR1H3) and LXR␤ (NR1H2), which are encoded by distinct genes. LXRs form obligate heterodimers with retinoid X receptors (RXRs), which are members of the nuclear receptor superfamily that can be regulated by 9-cis-retinoic acid (9cRA) and long-chain polyunsaturated fatty acids (13–15). LXR兾RXR heterodimers regulate their target genes by recognizing specific LXR-response elements consisting of two direct hexanucleotide repeats separated by four nucleotides (16). Without ligands, LXR兾RXR heterodimers actively repress transcription of target genes through recruitment of the nuclear receptor corepressors NCoR (nuclear receptor corepressor) and SMRT (silencing mediator of retinoid and thyroid hormone receptors) (17, 18). Upon binding either LXR or RXR ligands, corepressors are exchanged with nuclear receptor coactivators, resulting in transcriptional activation. LXR兾RXR heterodimers induce expression of genes that mediate cholesterol efflux from cells and its ultimate excretion into bile (19). This activity has been shown to be important in the regulation of cholesterol homeostasis in macrophages, which can accumulate massive amounts of cholesterol in disease settings, such as atherosclerosis. Recent Abbreviations: M-CSF, macrophage colony-stimulating factors; LXR, liver X receptor; RXR, retinoid X receptor; siRNA, small interfering RNA; 9cRA, 9-cis-retinoic acid; TLR4, Toll-like receptor 4; BMDM, bone marrow-derived macrophage; EC, 24(S),25-epoxycholesterol; CHX, cycloheximide. §To

whom correspondence should be addressed. E-mail: [email protected].

© 2004 by The National Academy of Sciences of the USA

PNAS 兩 December 21, 2004 兩 vol. 101 兩 no. 51 兩 17813–17818

MEDICAL SCIENCES

Communicated by Michael G. Rosenfeld, University of California at San Diego, La Jolla, CA, November 18, 2004 (received for review September 7, 2004)

studies have also demonstrated that LXRs inhibit transcriptional responses to activation of Toll-like receptor 4 (TLR4) in macrophages by antagonizing the actions of NF-␬B transcription factors (20). Here, we provide evidence that LXRs and RXRs regulate macrophage survival, supporting the concept that they are important modulators of innate immunity. Materials and Methods Reagents. Anisomycin (from Streptomyces griseolus) and SB202190 were purchased from Calbiochem. Cycloheximide (CHX), 9cRA, and lipopolysaccharide were obtained from Sigma. T1317 and GW3965 were kindly donated by X-ceptor Therapeutics. We purchased 24(S),25-epoxycholesterol (EC) from Biomol (Plymouth Meeting, PA). Cells. Bone-marrow-derived macrophages (BMDM) were isolated from 8- to 10-week-old mice as described (21). LXR␣兾␤⫺/⫺ mice (22) were a gift from David Mangelsdorf (University of Texas Southwestern Medical Center, Dallas). The cells were cultured in DMEM (Cellgro, Mediatech, Herndon, VA) containing 20% FBS (HyClone) and 30% L-cell conditioned media as a source of macrophage colony-stimulating factor (M-CSF). Macrophages were obtained as a homogeneous population of adherent cells after 6–8 days of culture. Unless otherwise stated, macrophages were used at ⬍80% confluence. All experiments were performed with the approval of the University of California at San Diego Animal Subject Committee. Apoptosis Tests. DNA fragmentation was measured in triplicate samples with a photometric enzyme immunoassay (Cell-Death Detection ELISA Plus, Hoffmann–La Roche) directed toward the recognition of histone-associated DNA fragments. In some experiments, the measurement of DNA fragmentation was performed by flow cytometry. Briefly, the cells were fixed in 70% ethanol for 30 min at room temperature and then stained with propidium iodide (30 ␮g兾ml) in 0.25% tryton兾PBS containing RNase A. The DNA content of 10,000 cells was analyzed by flow cytometry using an FL-2A channel. General caspase activation was measured in triplicate samples with a quantitative fluorimetric assay (homogenous caspases assay, Hoffmann–La Roche). The exposure of phosphatidylserine in the outer leaflet of the plasma membrane was measured by annexin V staining (23, 24). Macrophages were plated in slide chambers before exposure to LXR agonists and apoptotic signals. Annexin V–Alexa 568 staining (Hoffmann–La Roche) was perfomed in situ without detaching the cells from the plate. Hoechst dye was used for nuclear staining. Several fields of at least 120 cells each were counted, and the percentage of annexin V-positive cells versus total cells was determined. Microarray Analysis. Total RNA was isolated and purified by using

TRIzol reagent (Invitrogen) and RNeasy columns (Qiagen, Valencia, CA). cRNA was generated from 10 ␮g of total RNA by using Superscript (Invitrogen) and the High-Yield RNA transcription-labeling kit (Enzo Biochem). Duplicate samples of fragmented cRNA were hybridized to U74A arrays or Codelink Uniset 1 mouse arrays according to the manufacturer’s instructions (Affymetrix, Santa Clara, CA). Data were analyzed with MICROARRAY SUITE (Affymetrix) and GENESPRING software (Silicongenetics, Redwood City, CA). Northern Blot Analysis. Total RNA was purified by using TRIzol.

RNA samples (10 ␮g per lane) were separated in 1.2% agarose gels containing formaldehyde and transferred to Genescreen nylon membranes (NEN). Hybridization to labeled probes was performed by using Quickhyb (Stratagene). 17814 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0407749101

Fig. 1. LXR and RXR agonists inhibit apoptotic responses to growth-factor withdrawal and protein synthesis inhibition. (A and B) Macrophages were prestimulated with the indicated combinations of LXR and RXR agonists for 18 h and then deprived of M-CSF for 24 h. Ligands were replaced during the deprivation phase. The percentage of fragmented DNA (subG1 population) is indicated in the graphic (PI, propidium iodide). (C) WT and LXR⫺/ ⫺ macrophages (lacking both LXR␣ and LXR␤) were plated at subconfluent densities, treated with vehicle or T1317, and then incubated with CHX (10 ␮g兾ml) for 6 h. Macrophage apoptosis was determined by DNA fragmentation. *, P ⬍ 0.05, vs. treatment with CHX alone. (D) Macrophages (40,000 cells per well) were prestimulated with vehicle, T1317 (1 ␮M), 9cRA (1 ␮M), or a combination of both for 24 h and then treated with CHX (10 ␮g兾ml) for 5 h. General caspase activity was measured by fluorimetry. Error bars represent standard deviations. *, P ⬍ 0.05, vs. treatment with CHX alone.

Small Interfering RNA (siRNA)-Mediated Knockdown of AIM. siRNA

was obtained from Ambion (Austin, TX). Two different siRNAs directed to different regions of the AIM transcript were used. The following target sequences were used: AIM-1, 5⬘A ACGGA AGACACGT TGGCTCA-3⬘; and A IM-2, 5⬘AAGATGTCGTGTTCTGGACAA-3⬘. As a control, a scrambled siRNA that is not directed to any known vertebrate gene was developed from the following target sequence: 5⬘AAGATACTCGTGATTGCACAC-3⬘. In experiments directed to study macrophage apoptosis, 8 ⫻ 104 cells were transfected by using Superfect (Qiagen) with 0.4 ␮M siRNA. The same ratio of siRNA兾cell numbers was maintained in higher-scale experiments. Results LXR Activation Inhibits Macrophage Apoptosis. While investigating

roles of LXRs in regulation of lipid homeostasis, it was noted that treatment of BMDM with LXR agonists improved their survival in the setting of growth-factor withdrawal. Therefore, we investigated potential roles of LXRs in regulation of macrophage apoptosis. Culturing BMDM for 36 h in the absence of their specific growth factor (M-CSF) resulted in increased levels of cells with subG1 DNA content (DNA content of ⬍2 N), which is an indicator of apoptosis-induced DNA fragmentation (Fig. 1 A and B). This process was attenuated when macrophages were preincubated with the synthetic LXR agonists T1317 or GW3965 or the natural agonist EC. Intriguingly, 9cRA, which is a ligand for the RXR heterodimeric partner of LXRs, had little effect on Valledor et al.

of LXR and RXR agonists significantly reduced the apoptotic responses, as measured by TUNEL staining, that were elicited by infection with B. anthracis, Escherichia coli, and the Salmonella typhimurium strain SL1344兾SipB⫺ (Fig. 2C). Although, in some cases, the antiapoptotic effects of LXR兾RXR agonists could be overcome at high multiplicities of infection (data not shown), these findings suggest that LXR and RXR promote macrophage survival in the face of bacterial infection.

Fig. 2. LXR and RXR activation protects macrophages from pathogeninduced apoptosis. (A) LXR and RXR activation protects macrophages from apoptosis induced by the combination of LPS and the p38 inhibitor SB202190, as determined by the percentage of annexin V-positive cells. (B) Representative photomicrographs of each treatment. SBL, SB202190 (5 ␮M) plus LPS (100 ng兾ml); 9cT, 9cRA (1 ␮M) plus T1317 (1 ␮M). (C) Effect of a combination of T1317 and 9cRA on apoptotic responses of macrophages exposed to the indicated multiplicity of infection (MOI) of B. anthracis, E. coli, and S. typhimurium SL1344兾SipB⫺. Error bars indicate standard deviations. *, P ⬍ 0.05; **, P ⬍ 0.01, vs. bacterial exposure in the absence of ligands.

subG1 DNA content alone, but it markedly enhanced the effects of all three LXR-specific agonists (Fig. 1 A and B). To assess whether the protective effects of LXRs are limited to the control of programmed cell death caused by growth-factor withdrawal, we extended these studies to other modes of macrophage apoptosis. As a strategy to subvert normal host-defense responses, a number of pathogens are armed with virulence factors that lead to rapid death of host macrophages (4). These virulence determinants include pore-forming toxins, protein-synthesis inhibitors, superantigens, and inhibitors of prosurvival signaling. In particular, macrophages are very sensitive to protein-synthesis inhibition (25, 26). Consistent with this observation, treatment of macrophages with CHX resulted in increased DNA fragmentation (Fig. 1C) and caspase activation (Fig. 1D). Preincubating the cells with T1317 for 24 h attenuated the apoptotic process induced by CHX in wild-type macrophages, but not in LXR-deficient macrophages (Fig. 1C). Combined treatment of macrophages with synthetic or natural LXR agonists and 9cRA resulted in an additive inhibition of caspase activation (Fig. 1D). Similar results were obtained when the macrophage apoptotic program was stimulated by anisomycin (S. griseolus) (data not shown). LXR and RXR Agonists Protect Macrophages from Pathogen-Induced Apoptosis. Recent studies have identified the p38MAPK pathway

as a target for the action of lethal factor, a virulence determinant from B. anthracis (27). Inhibition of the p38MAPK cascade sensitizes macrophages to programmed cell death in response to activation of TLR4 (26, 27). Treatment of BMDM with LXR and RXR agonists resulted in decreased levels of annexin V staining after the combined incubation with LPS and the p38 inhibitor SB202190 (Fig. 2 A and B). Therefore, we evaluated the possibility that LXR and RXR agonists could protect macrophages from apoptosis due to infection with B. anthracis and other bacterial pathogens. Indeed, preincubation with a combination Valledor et al.

tion from apoptosis, macrophages were preincubated with agonists at different time points before addition of the proapoptotic signal. Interestingly, inhibition of apoptosis in response to either anisomycin or the combination of SB202190 and LPS took place only after a 12-h preincubation of the cells with LXR and RXR ligands (Fig. 3 A and B). To identify ligand-regulated antiapoptotic genes, expression-profiling experiments were performed by using Affymetrix U74A and Codelink Uniset Mouse 1 microarrays. Because the antiapoptotic effects of LXR agonists were strongly potentiated by 9cRA, microarray experiments examined effects of the LXR agonist T1317 alone and in combination with 9cRA. Whereas T1317 alone had relatively modest effects on expression of genes with functional annotations linked to apoptosis (data not shown), the combination of T1317 and 9cRA strongly regulated several proapoptotic and antiapoptotic genes (Fig. 3C). The most significantly upregulated gene with a functional annotation linked to inhibition of apoptosis was AIM, also known as CT-2兾Api6 (28, 29). AIM was recently demonstrated to be induced by LXR agonists in liver (30). Also, the antiapototic regulators Birc1a (also known as NeuroAIP1) and Bcl-XL were up-regulated 3.3- and 2.9-fold, respectively (Fig. 3C). The combination of T1317 and 9cRA also significantly down-regulated the proapoptotic regulators兾 effectors Dnase1l3 (DNase ␥); caspases 1, 4兾11, 7, and 12; Fas ligand; Cidea; and the peptidoglycan-recognition protein Tag7. These results were confirmed in two independent microarray experiments using Codelink Mouse Uniset I microarrays, and they were also validated for the overlapping sets of genes by using Affymetrix U74A microarrays (data not shown). Together, these findings suggest that the combination of LXR and RXR agonists exert antiapoptotic effects by coordinately regulating several proapoptotic and antiapoptotic genes. An additional series of microarray experiments was performed to evaluate the influence of LXR activation on regulation of the apoptotic program induced by engagement of TLR4 (Fig. 4). Macrophages were incubated with GW3965 or vehicle for 16 h and then treated with LPS for 6 h. Of 86 genes with functional annotations linked to apoptosis and expressed in at least one condition, 23 genes were altered ⬎1.5-fold by LPS treatment. Categorizing these genes into proapoptotic and antiapoptotic functions indicated that the overall response to TLR4 engagement was primarily proapoptotic, illustrated for selected categories of genes in Fig. 4. The dominant effect of LXR activation was to counterregulate a subset of the proapoptotic program of gene expression induced by LPS. For example, the LXR agonist attenuated LPS-dependent down-regulation of the antiapoptotic proteins Bcl2, Bag3, and Birc1a. Conversely, LXR activation inhibited LPS-dependent induction of the proapoptotic factors Bax, Bak, and Bcl211, as well as caspases 1, 3, 4兾11, 7, 8, and 12. Together, these findings provided another independent line of evidence suggesting that LXRs inhibit apoptosis by coordinately regulating a network of genes that control programmed cell death. Because AIM was the gene that was the most highly induced by LXR and RXR activation, and because its antiapoptotic function is not as well established as that of Bcl-XL or Birc1a, we further characterized its regulation and function. AIM expresPNAS 兩 December 21, 2004 兩 vol. 101 兩 no. 51 兩 17815

MEDICAL SCIENCES

LXR and RXR Regulate Expression of Proapoptotic and Antiapoptotic Factors. To characterize mechanisms of LXR-mediated protec-

Fig. 3. Time requirements for effects of LXR兾RXR agonists on macrophage survival and identification of candidate genes. Macrophages were preincubated with T1317 and 9cRA for different periods of time and then stimulated with anisomycin (Anisom) (A) or SB202190 plus LPS (B) for 6 h. The levels of caspase activity (A) or the percentage of annexin V-positive cells were determined as described above. Error bars represent standard deviations. *, P ⬍ 0.05, vs. treatment with anisomycin (A) or SB plus LPS (B) alone. (C) mRNA samples from macrophages stimulated with vehicle or the combination of T1317 (1 ␮M) and 9cRA (1 ␮M) for 16 h were subjected to expression profile analysis by using Codelink Mouse Uniset 1 microarrays. The relative expression levels of genes with annotations linked to apoptosis changing by a factor of ⱖ1.5-fold are illustrated. Values are means of biological replicates. Changes in gene expression for AIM, Birc1a, Bcl-xL, Dnase1l3, and caspases 1, 7, 11, and 12 were independently confirmed by Northern blot analysis (data not shown).

sion was evaluated initially in differentiated macrophages treated with LXR agonists. AIM mRNA levels were maximally induced at 12–24 h of stimulation with T1317, which is somewhat delayed in comparison with ABCA1 and other direct LXR target genes (Fig. 5A). The combination of T1317 and 9cRA led to a much stronger induction of AIM, with maximal levels of expression occurring at 24 h, consistent with the results of microarray experiments. Both the time course of AIM induction and synergistic effects of T1317 and 9cRA correlates with the time-course requirements and combinatorial effects of both ligands on inhibition of apoptosis shown in Fig. 3. In LXRdeficient macrophages, AIM was no longer induced (Fig. 5B). AIM expression could also be induced by EC, indicating that it is subject to regulation by natural LXR ligands (Fig. 5C). Several combinations of the AIM promoter and upstream or down17816 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0407749101

Fig. 4. Activation of LXR antagonizes the proapoptotic program induced by engagement of TLR4. Macrophages were incubated with the LXR agonist GW3965 (1 ␮M) or vehicle for 16 h before treatment with LPS (100 ng兾ml) for 6 h. Total RNA was subjected to microarray analysis by using Codelink Uniset Mouse 1 microarrays. Relative expression levels for selected categories of proapoptotic and antiapoptotic genes are shown. Genes exhibiting a response to GW3965 predicted to be proapoptotic are shown in red. Genes exhibiting a response predicted to be antiapoptotic are shown in green. (A) Antiapoptotic members of the Bag and Bcl families. (B) Proapoptotic members of the Bcl family. (C) Members of the antiapoptotic baculovirus inhibitor of apoptosis protein (IAP) repeat-containing (Birc) family. (D) Members of the caspase family.

stream genomic elements were not sufficient to drive LXRdependent reporter gene expression in macrophage cell lines, raising the possibility that it is an indirect target of LXR兾RXR heterodimers (data not shown). To investigate whether AIM induction contributes to the antiapoptotic effects of LXR agonists, we inhibited its expression by using AIM-specific siRNAs. Primary macrophages were transfected with either siRNAs directed against AIM or a control siRNA designed to be unable to direct degradation of any known mouse gene. The cells were then stimulated with 9cRA and T1317 and expression of AIM was determined 24 h later by Northern blotting. As shown in Fig. 5D, transfection of macrophages with the siRNA directed against AIM reduced AIM mRNA expression by ⬇75%. Under these conditions, siRNA against AIM partially inhibited the ability of LXR and RXR agonists to protect macrophages from anisomycin-induced apoptosis (Fig. 5E). In contrast, LXR and RXR agonists were fully capable of inhibiting anisomycin-induced apoptosis in macrophages transfected with the control siRNA. These results suggest that induction of AIM expression contributes to the mechanism by which LXR and RXR agonists protect against apoptosis. Discussion Roles of LXRs in the Control of Macrophage Survival. LXRs play critical roles in the regulation of cholesterol and fatty-acid homeostasis (31). In macrophages, LXRs activate the expression Valledor et al.

tion of genes (such as ABCA1) that are required for cholesterol efflux. Unexpectedly, activation of LXR and RXR protects macrophages from apoptotic signaling pathways that are stimulated by bacterial pathogens including B. anthracis and S. typhimurium. Some pathogens, such as Listeria and Legionella, can reside intracellularly within macrophages and, thus, elude immune clearance (38). In contrast, other pathogens (exemplified by Salmonella, Shigella, and Yersinia) induce macrophage apoptosis and stimulate the release of proinf lammatory cytokines (38). The present studies suggest that use of LXR and RXR agonists may provide a useful tool for investigating the importance of apoptosis in the pathogenicity of various bacterial infections in vivo.

Fig. 5. AIM is synergistically induced by LXR and RXR agonists and contributes to their antiapoptotic effects. (A) Macrophages were stimulated with T1317 or the combination of T1317 and 9cRA for the indicated times. (B) Wild-type and LXR⫺/⫺ macrophages were incubated for 24 h with T1317, 9cRA, or a combination of both. In A and B, expression of AIM and other LXR target genes was analyzed by Northern blotting. (C) AIM is induced by EC (10 ␮M). (D) Transfection of BMDM with an siRNA against AIM significantly reduces AIM RNA levels. (E) Reduction of AIM expression reduces antiapoptotic activities of LXR and RXR agonists. Macrophages were transfected with either a control siRNA or a siRNA directed against AIM. The cells were then stimulated for 24 h with T1317, 9cRA, or a combination of both, and then treated with anisomycin for 5 h. Relative caspase activity was measured as an indicator of apoptosis. Each treatment was performed in triplicate. Error bars represent standard deviations. *, P ⫽ 0.045, vs. anisomycin treatment alone. **, P ⫽ 0.011, vs. anisomycin alone. ***, P ⫽ 0.055, vs. anisomycin alone

of a set of genes (such as the ABCA1 cholesterol transporter) that act to reduce cellular cholesterol levels (32–34). This function of LXRs has been most intensively studied in the context of atherosclerosis, which is a disease in which cholesterol-loaded macrophages accumulate within the walls of large arteries (35). Recent studies demonstrating that synthetic LXR agonists can also inhibit transcriptional events induced by TLR4 signaling suggest that LXRs have additional roles in the regulation of immune responses (20). The present studies, demonstrating that LXRs regulate macrophage survival, support this concept. A potentially important antiapoptotic role of LXRs may be to protect macrophages from cholesterol toxicity due to phagocytosis of dead cells. Programmed cell death is an important phenomenon during resolution of inflammation, and oxidative damage is a component of the apoptotic program (36). The resolution of acute inflammation requires the bulk clearance of infiltrating inflammatory cells in an ordered manner. Neutrophils participate in early phases of the inflammatory process by phagocytosing and destroying the agents that cause inflammation. Rapidly after their activation, they undergo apoptosis (37). Resident macrophages play an essential role in the clearance of apoptotic bodies, and debris generated during those conditions and the uptake of apoptotic cells results in a significant load of cellular cholesterol. Conversion of a fraction of this excess cholesterol to oxysterol ligands for LXR would result in activaValledor et al.

LXR predominantly antagonized the apoptotic program induced by engagement of TLR4 by both positively and negatively regulating gene expression. Furthermore, the combination of LXR and RXR agonists was more effective at inhibiting macrophage apoptosis than either agonist alone. The antiapoptotic factors Bcl-XL, Birc1a兾NAIP, and AIM兾CT2兾Api6 were significantly up-regulated by the combination of LXR and RXR agonists, suggesting that they are directly or indirectly regulated by RXR兾LXR heterodimers. Bcl-XL is an antiapoptotic form of Bcl-X that is related in structure and function to Bcl-2 (39). Members of the Bcl-2 family control apoptosis by several mechanisms, including alterations in cytochrome c release, which ultimately regulates caspase activation (40, 41). The balance between proapoptotic members (e.g., Bax, Bad, and Bak) and antiapoptotic members (e.g., Bcl-2, Bcl-XL, and Mcl-1) determines the fate of many types of cells. Birc1a兾NAIP is related to baculoviral inhibitor of apoptosis proteins (IAPs) (42) and directly inhibits the enzymatic activities of effector caspases 3 and 7 (43). In combination with down-regulation of caspases 1, 4兾11, 7, and 12, coordinate up-regulation of Bcl-XL and Birc1a兾 NAIP provides a likely explanation for the ability of LXR and RXR agonists to decrease caspase activities in response to exposure to apoptotic stimuli and bacterial pathogens. AIM兾 CT2兾Api6 was synergistically activated by LXR兾RXR agonists and contributed to their antiapoptotic effects. Although the mechanisms responsible for the antiapoptotic activities of AIM兾 CT2兾Api6 remain to be established, in situ hybridization studies showed high expression in specific macrophage subpopulations, including subsets of Kupffer cells in the liver, macrophages in the thymic cortex, in the marginal zone of the spleen, and in peripheral areas of granulomas (29). Nuclear receptors also play important physiological roles by negatively regulating gene expression, and microarray experiments indicated that LXR兾RXR agonists inhibited the expression of several positive regulators and effectors of apoptosis. Mechanisms of negative regulation by nuclear receptors are generally less well understood than those responsible for positive regulation, and it is possible that additive兾synergistic effects of LXR and RXR agonists results from independent activities of the two receptor subtypes. However, microarray experiments indicated that 9cRA alone had very little inhibitory activity on LPS-dependent gene expression in macrophages (data not shown). Thus, it appears to be true that the predominant role of RXR agonists as inhibitors of apoptosis is to potentiate both the positive and negative transcriptional activities of LXR agonists, most likely acting through LXR兾RXR heterodimers. Caspases 1, 4兾11, 7, and 12 were modestly down-regulated (from 1.5- to 2-fold, Figs. 3C and 4D), contributing to reduced caspase activity observed after treatment with LXR兾RXR agonists. Intriguingly, the combination of LXR and RXR agonists down-regulated several genes that contribute to apoptosis-induced DNA fragmentation. DNase ␥ and Cidea, which contribute to DNA PNAS 兩 December 21, 2004 兩 vol. 101 兩 no. 51 兩 17817

MEDICAL SCIENCES

Mechanisms of LXR兾RXR Protection from Apoptosis. Activation of

fragmentation during apoptosis (44, 45), were strongly downregulated in response to LXR兾RXR agonists. LXR兾RXR agonists also inhibited the expression of peptidoglycan recognition protein (PGLYP), which forms a cytotoxic complex with heatshock protein 70 (46). Together, these studies demonstrate that LXR and RXR coordinately regulate the network of genes that control programmed cell death, resulting in protection of macrophages from bacterial-induced apoptosis. Therefore, it will be of interest to further explore the roles of LXRs and RXRs in regulation of the responses of macrophages and other cell types to infection and test the possibility that LXR兾RXR agonists may be used to enhance innate immunity to bacterial pathogens by their ability to prevent macrophage apoptosis. We thank Drs. David Mangelsdorf and Joyce Repa (University of Texas Southwestern Medical Center) providing LXR⫺/⫺ mice; Jean Lozach and Chris Benner for assistance with microarray data analysis; Courtney

Havens and Britta Maedge for assistance with FACS analysis; Dr. Jin Mo Park for assistance with apoptosis assays; and Rene Meijer for help with fluorimetric determinations. We also thank Drs. Antonio Celada, Jordi Xaus, and Monica Comalada for helpful discussions, and Alexandra Howarth for assistance with preparation of the manuscript. This work was supported by Biostar Grant 00-10113, National Institutes of Health Grants HL56989 and ES10337, and the Stanford Reynolds Center. A.F.V. was supported by a fellowship for postdoctoral training from the Spanish Government (Beca Formacio ´n Postdoctores en el Extranjero, Ministerio de Ciencia y Tecnologı´a). D.S.-V. was supported by a fellowship from the Leukemia and Lymphoma Society. L.-C.H. was supported by work in the M.K. laboratory under National Institutes of Health Grant AI061712 and by a fellowship from the Cancer Research Institute. C.K.G. is a Stanford Reynolds Scholar. M.K. is an American Cancer Society Research Professor. Microarray analysis was performed with the assistance of the University of California at San Diego Affymetrix GeneChip and Biogem Core facilities with the support of National Institutes of Health Diabetes and Endocrinology Center Grant DK063491.

Celada, A. & Nathan, C. (1994) Immunol. Today 15, 100–102. Gordon, S. (1998) Res. Immunol. 149, 685–688. Savill, J. (1997) J. Leukocyte Biol. 61, 375–380. Weinrauch, Y. & Zychlinsky, A. (1999) Annu. Rev. Microbiol. 53, 155–187. Cohen, G. M. (1997) Biochem. J. 326, 1–16. Ranger, A. M., Malynn, B. A. & Korsmeyer, S. J. (2001) Nat. Genet. 28, 113–118. Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schu ¨tz, G., Umesono, K., Blumberg, B., Kastner, P., Mark, M., Chambon, P. & Evans, R. M. (1995) Cell 83, 835–839. Welch, J. S. & Glass, C. K. (2003) in The Macrophage as a Therapeutic Target, ed. Gordon, S. (Springer, Berlin), Vol. 158, pp. 209–226. Karin, M. (1998) Cell 93, 487–490. De Bosscher, K., Vanden Berghe, W. & Haegeman, G. (2003) Endocr. Rev. 24, 488–522. Janowski, B. A., Willy, P. J., Devi, T. R., Falck, J. R. & Mangelsdorf, D. J. (1996) Nature 383, 728–731. Janowski, B. A., Grogan, M. J., Jones, S. A., Wisely, G. B., Kliewer, S. A., Corey, E. J. & Mangelsdorf, D. J. (1999) Proc. Natl. Acad. Sci. USA 96, 266–271. Heyman, R. A., Mangelsdorf, D. J., Dyck, J. A., Stein, R. B., Eichele, G., Evans, R. M. & Thaller, C. (1992) Cell 68, 397–406. Bourguet, W., Vivat, V., Wurtz, J. M., Chambon, P., Gronemeyer, H. & Moras, D. (2000) Mol. Cell 5, 289–298. Mata de Urquiza, A. M., Liu, S., Sjoberg, M., Zetterstrom, R. H., Griffiths, W., Sjovall, J. & Perlmann, T. (2000) Science 290, 2140–2144. Willy, P. J., Umesono, K., Ong, E. S., Evans, R. M., Heyman, R. A. & Mangelsdorf, D. J. (1995) Genes Dev. 9, 1033–1045. Wagner, B. L., Valledor, A. F., Shao, G., Daige, C. L., Bischoff, E. D., Petrowski, M., Jepsen, K., Baek, S. H., Heyman, R. A., et al. (2003) Mol. Cell. Biol. 23, 5780–5789. Hu, X., Li, S., Wu, J., Xia, C. & Lala, D. S. (2003) Mol. Endocrinol. 17, 1019–1026. Repa, J. J. & Mangelsdorf, D. J. (1999) Curr. Opin. Biotechnol. 10, 557–563. Joseph, S. B., Castrillo, A., Laffitte, B. A., Mangelsdorf, D. J. & Tontonoz, P. (2003) Nat. Med. 9, 213–219. Valledor, A. F., Xaus, J., Marques, L. & Celada, A. (1999) J. Immunol. 163, 2452–2462. Repa, J. J., Liang, G., Ou, J., Bashmakov, Y., Lobaccaro, J.-M. A., Shimomura, I., Shan, B., Brown, M. S., Goldstein, J. L. & Mangelsdorf, D. J. (2000) Genes Dev. 14, 2819–2830. Koopman, G., Reutelingsperger, C. P., Kuijten, G. A., Keehnen, R. M., Pals, S. T. & van Oers, M. H. (1994) Blood 84, 1415–1420. Vermes, I., Haanen, C., Steffens-Nakken, H. & Reutelingsperger, C. (1995) J. Immunol. Methods 184, 39–51.

25. Yang, G. H., Jarvis, B. B., Chung, Y. J. & Pestka, J. J. (2000) Toxicol. Appl. Pharmacol. 164, 149–160. 26. Hsu, L. C., Park, J. M., Zhang, K., Luo, J. L., Maeda, S., Kaufman, R. J., Eckmann, L., Guiney, D. G. & Karin, M. (2004) Nature 428, 341–345. 27. Park, J. M., Greten, F. R., Li, Z. W. & Karin, M. (2002) Science 297, 2048–2051. 28. Haruta, I., Kato, Y., Hashimoto, E., Minjares, C., Kennedy, S., Uto, H., Yamauchi, K., Kobayashi, M., Yusa, S., Muller, U., et al. (2001) J. Biol. Chem. 276, 22910–22914. 29. Miyazaki, T., Hirokami, Y., Matsuhashi, N., Takatsuka, H. & Naito, M. (1999) J. Exp. Med. 189, 413–422. 30. Maxwell, K. N., Soccio, R. E., Duncan, E. M., Sehayek, E. & Breslow, J. L. (2003) J. Lipid Res. 44, 2109–2119. 31. Repa, J. J. & Mangelsdorf, D. J. (2000) Annu. Rev. Cell. Dev. Biol. 16, 459–481. 32. Venkateswaran, A., Laffitte, B. A., Joseph, S. B., Mak, P. A., Wilpitz, D. C., Edwards, P. A. & Tontonoz, P. (2000) Proc. Natl. Acad. Sci. USA 97, 12097–12102. 33. Repa, J. J., Turley, S. D., Lobaccaro, J. A., Medina, J., Li, L., Lustig, K., Shan, B., Heyman, R. A., Dietschy, J. M. & Mangelsdorf, D. J. (2000) Science 289, 1524–1529. 34. Costet, P., Luo, Y., Wang, N. & Tall, A. R. (2000) J. Biol. Chem. 275, 28240–28245. 35. Ricote, M., Valledor, A. F. & Glass, C. K. (2004) Arterioscler. Thromb. Vasc. Biol., 24, 230–239. 36. Buttke, T. M. & Sandstrom, P. A. (1994) Immunol. Today 15, 7–10. 37. Bellingan, G. J., Caldwell, H., Howie, S. E., Dransfield, I. & Haslett, C. (1996) J. Immunol. 157, 2577–2585. 38. Navarre, W. W. & Zychlinsky, A. (2000) Cell. Microbiol. 2, 265–273. 39. Chao, D. T., Linette, G. P., Boise, L. H., White, L. S., Thompson, C. B. & Korsmeyer, S. J. (1995) J. Exp. Med. 182, 821–828. 40. Akgul, C., Moulding, D. A. & Edwards, S. W. (2001) FEBS Lett 487, 318–322. 41. Kroemer, G. (1997) Nat. Med. 3, 614–620. 42. Roy, N., Mahadevan, M. S., McLean, M., Shutler, G., Yaraghi, Z., Farahani, R., Baird, S., Besner-Johnston, A., Lefebvre, C., Kang, X., et al. (1995) Cell 80, 167–178. 43. Maier, J. K., Lahoua, Z., Gendron, N. H., Fetni, R., Johnston, A., Davoodi, J., Rasper, D., Roy, S., Slack, R. S., Nicholson, D. W. & MacKenzie, A. E. (2002) J. Neurosci. 22, 2035–2043. 44. Shiokawa, D., Kobayashi, T. & Tanuma, S. (2002) J. Biol. Chem. 277, 31031–31037. 45. Inohara, N., Koseki, T., Chen, S., Wu, X. & Nunez, G. (1998) EMBO J. 17, 2526–2533. 46. Sashchenko, L. P., Dukhanina, E. A., Yashin, D. V., Shatalov, Y. V., Romanova, E. A., Korobko, E. V., Demin, A. V., Lukyanova, T. I., Kabanova, O. D., Khaidukov, S. V., et al. (2004) J. Biol. Chem. 279, 2117–2124.

1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22.

23. 24.

17818 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0407749101

Valledor et al.