Inhibitor of Apoptosis Protein cIAP2 Is Essential for Lipopolysaccharide-Induced Macrophage Survival

Generation of germ line chimeras and homozygous mice.

129/sv genomic clones (13) spanning the mouse ciap2 gene were used to construct a replacement-type targeting vector in which an internal ribosome entry site (IRES)-lacZ and phosphoglycerate kinase (PGK)-neomycin (neo) cassette (SA-IRES-βgeo; (16) replaced exons 2 to 5 in the plasmid pKO (M. Holcik and R. G. Korneluk, unpublished). The resultant targeting vector (pKO.hiap1) was comprised of a 4.1-kb 5′ arm and a 5.5-kb 3′ arm bracketing the IRES-lacZ/PGK-neo insertion. RW4 embryonic stem (ES) cells were electroporated as described previously (28), and DNA from neomycin-resistant clones was extracted and analyzed. Disruption of the ciap2 allele was confirmed by Southern blot analysis of EcoRV-digested genomic DNA after hybridization with a probe corresponding to exon 1 of the ciap2 gene. Chimeric mice were produced by morula aggregation (27) with targeted RW-4 cells. Chimeric male progeny were mated with 129/SvJ females and heterozygous progenies were backcrossed to C57BL/6 mice for at least 10 generations. Heterozygous mice were then crossed to produce homozygous cIAP^−/− mice. Both the electroporation of ES cells and the generation of chimeric animals were performed at the Genome Systems, Inc., facility (St. Louis, MO). Mice were housed in a specific-pathogen-free environment, and all experiments were performed in accordance with the guidelines of the Canadian Council on Animal Care and protocols approved by the University of Ottawa Animal Care Committee.

Southern blot analysis.

Genomic DNA was isolated by standard methods and digested with EcoRV, separated on agarose gels, and transferred to Biodyne Nylon Paper (Life Technologies, Rockville, MD). Full-length ³²P-labeled cIAP2 cDNA probes were prepared by using Rediprime (Amersham Pharmacia) and [³²P]dCTP (Amersham Pharmacia) according to the manufacturer's directions. Membranes were washed with 0.1× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate)-0.1% sodium dodecyl sulfate at 65°C for 10 min and exposed to X-ray film (Kodak).

Western blot analysis.

Mouse tissue was weighed, lysed in 5 volumes (wt/vol) of lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1 mM EDTA, 1 mM vanadate, 1% [vol/vol] Nonidet P-40, 0.25% [vol/vol] sodium deoxycholate, 1 μg of leupeptin/ml, 1 μg of aprotinin/ml, and 1 μM phenylmethylsulfonyl fluoride), and then crushed and well mixed. The samples were then rotated for 45 min at 4°C. The samples were then centrifuged for 15 min at ∼14,000 rpm in a microcentrifuge. The supernatant was collected and assayed by using a BCA kit (Pierce); then, equal amounts of protein samples were loaded per lane, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and analyzed by Western blotting with rabbit polyclonal α-cIAP2, α-cIAP1, or α-XIAP antibodies (1:2,500 dilution; Ægera), followed by anti-rabbit horseradish peroxidase-conjugated secondary antibody (Amersham), and the immune complexes were visualized by using an enhanced chemiluminescence kit (Roche).

mRNA isolation and quantitative mRNA analysis.

mRNA was extracted from mouse embryonic fibroblasts or peritoneal macrophages by using QIAGEN RNeasy 96-well extraction kit (QIAGEN, Mississauga, Ontario, Canada) and tested with a TaqMan instrument (Perkin-Elmer, Foster City, CA) with specific DNA probes for murine xiap, ciap1, and ciap2 and the TaqMan EZ RT-PCR kit (QIAGEN).

Animal LPS models.

Adult 4- to 6-week-old mice (n = 6 to 10) were injected intraperitoneally (i.p.) with a range of LPS doses (10 to 200 mg of LPS/kg) from Escherichia coli (Sigma) in a total volume of 0.2 ml of nonpyrogenic saline. It should be noted that the LPS used (L4516; Sigma) might contain <1% protein and RNA content, and therefore some of the effects seen may be attributed to alternative signaling TLR pathways other than TLR4.

Adult 4- to 6-week-old mice (n = 3), cIAP2^−/− and littermate control mice, were injected i.p. with a 100% lethal dose (LD₁₀₀) of LPS (35 mg of LPS/kg). At the appropriate times mice were anesthetized with pentobarbital and killed by cervical dislocation. The plasma was then collected from these animals and used to determine concentrations of IL-1β, TNF-α, and IL-12 in serum with an ELISA kit (R&D Systems).

Adult 4- to 6-week-old mice (n = 6) were injected i.p. with LPS (35 mg of LPS/kg), and at time zero and 5 h the mice were euthanized and the cells from the peritoneum were collected and stained to determine macrophage, T-cell, and B-cell percentages. The macrophages were also stained with fluorescein isothiocyanate (FITC)-labeled annexin V (Immunotech, Marseille, France) to determine apoptotic status.

Flow cytometry.

T and B cells were isolated from mouse lymphoid tissue (spleen) by first mincing and then pressing the tissue through a 10-μm-pore-size metal mesh and were then counted by the trypan blue exclusion method. Macrophages were harvested from mice by using repeated i.p. lavage (three times with 10 ml of media: 5% fetal calf serum [FCS], 50 μM β-mercaptoethanol, and 125 mM l-glutamine, penicillin, and streptomycin) at 4°C and washed once (centrifuged at 800 × g for 15 min) and resuspended at ∼2 × 10⁷ cells/ml for peritoneal macrophages. Subsequently, peritoneal macrophages were layered on 5 ml of room temperature Lympholyte-M (CedarLane, Canada) and centrifuged at 1,500 × g for 20 min at room temperature. For splenic macrophages, 3 ml of the minced and pressed tissue (∼2 × 10⁷ splenocytes/ml) was layered on 5 ml of room temperature Lympholyte-M (CedarLane, Canada) and centrifuged at 1,500 × g for 20 min at room temperature. Subsequently, either peritoneum- or spleen-derived macrophages were collected from the interface layer, washed twice with complete Dulbecco modified Eagle medium (DMEM), and resuspended in 1 ml of complete DMEM. The resulting cells were then counted by using trypan blue exclusion and a DiffQuik stain kit (IMEB, Inc., San Marcos, CA). Cells (10⁵ to 10⁶) were incubated with the following conjugated monoclonal antibodies: α-CD3ɛ-FITC, CD4-phycoerythrin (PE), CD8a-Cy-Chrome, CD69-PE, B220-FITC, B220-PE, CD11b-Cy-Chrome (Pharmingen) F4/80-FITC and F4/80-PE (Cedarlane Laboratories, Hornby, Ontario, Canada). Flow cytometric analyses were performed on a Coulter XL cytometer (Coulter, Canada). Concentrations of TNF-α and IL-1β in primary tissue culture supernatants were determined by using an enzyme-linked immunosorbent assay kit (R&D Systems).

Primary tissue culture and death assays.

Primary cultures were maintained in DMEM supplemented with 10% FCS and 10 ng of granulocyte-macrophage colony-stimulating factor (R&D Systems)/ml for macrophages and with 5% FCS, 50 μM β-mercaptoethanol, and 125 mM l-glutamine, penicillin, and streptomycin (∼85% confirmed via a FITC-stained α-CD3 antibody using flow cytometry) for T cells, B cells, and thymocytes.

Confluent primary cultures of peritoneum-derived macrophages were either pretreated with LPS (10 μg/ml, 4 h) or not pretreated prior to exposure to α-Fas antibody (20 μg/ml, clone Jo2) and then TUNEL stained (Roche) to assess cell viability. Primary cultures of spleen-derived T cells were preincubated with a range of IL-7 concentrations (0, 5, and 10 ng/ml) and then exposed to dexamethasone (100 nM) and T-cell survival was monitored over a 12-h period.

Establishment of cIAP2^−/− mice.

To study the function of cIAP2, we disrupted the murine ciap2 gene by homologous recombination in ES cells (Fig. 1A and B). cIAP2^−/− mice showed no overt phenotype, were fertile, followed Mendelian frequency of inheritance, and appeared healthy up to 52 weeks of age. Whole-mouse and organ weights, as well as primary and secondary lymphoid compartment cell count numbers, for cIAP2^−/− and control littermates, up to the age of 36 weeks, showed no significant difference (data not shown). Western blot analysis and quantitative reverse transcription-PCR showed no compensatory increase of IAP family members in cIAP2-ablated tissue (Fig. 1C and D).

cIAP2 is highly upregulated in macrophages treated with LPS.

LPS-induced activation of macrophages causes the upregulation of a multitude of genes, including the production or release of inflammatory mediators and the upregulation of cell surface receptors and of cell survival proteins. Conflicting previous reports suggested that in macrophages cIAP2 is upregulated or remains unchanged in response to LPS treatment. Also, it has been suggested that either XIAP or cIAP1 levels would rise to compensate for the loss of cIAP2. Therefore, we assayed the iap mRNA levels of LPS-treated macrophages, derived from the peritoneal cavity, relative to untreated controls. Peritoneal macrophages were cultured at 10⁵ cell/well (96 well, flat-bottom plate) and exposed to various concentrations of LPS for 18 h. xiap and ciap1 message levels remained relatively unchanged over the range of LPS doses (0.1 ng/ml to 10 mg/ml) (Fig. 2A). In contrast, ciap2 levels increased dramatically, up to 30-fold above untreated controls (P < 0.01 for all LPS doses) (Fig. 2A). The innate immune system provides protection within the first minutes to hours after a pathogenic challenge. Therefore, a LPS-ciap2 mRNA time course assay was performed (100 ng of LPS/ml). ciap2 mRNA was substantially upregulated, >20-fold above untreated controls, within 1 h, and this increased level was maintained for 24 h (Fig. 2B).

cIAP2^−/− mice are resistant to LPS-induced endotoxic shock.

Schoemaker et al. proposed an important role of cIAP2 in LPS-induced apoptotic death of liver cells sensitized to endotoxic shock by d-galactosamine (DGLN) (22). This work also predicted that cIAP2-null mice would therefore be highly susceptible to LPS-induced endotoxic shock. Thus, to investigate the role of cIAP2 in an innate immune response the cIAP2^−/− mice were treated with LPS. Surprisingly, contrary to the predicted outcome, an i.p.-injected administration of LPS (0 to 40 mg/gm) proved fatal to both wild-type (Fig. 3A) and cIAP2^−/+ (data not shown) mice but not to cIAP2-null mice (Fig. 3A). All cIAP2^−/− mice treated with 40 mg of LPS/kg or less survived, whereas littermate controls succumbed in a dose-dependent manner. In fact, the LD₁₀₀ dose for cIAP2^−/− mice was found to be approximately threefold greater (100 mg of LPS/kg) compared to that of control littermates (35 mg of LPS/kg). Moreover, cIAP2^−/− mice survived 2 to 7 days even at an LPS dose of 200 mg of LPS/kg, in contrast to control littermates that all died within 24 h, even at the lower dose of 40 mg of LPS/kg (Fig. 3B).

cIAP2^−/− mice are susceptible to Fas-, platelet-activating factor (PAF)-, and DGLN/LPS-induced death.

In order to determine the sensitivity of cIAP2^−/− mice to other lethal insults and inflammatory mediators, additional triggers were tested. The response of these animals to an i.p. injection of α-Fas antibodies (100 μg/mouse) (Table 1) and the effect of treatment with PAF, an inflammatory mediator that acts downstream of the LPS activation of macrophages (24) (Table 2), were examined. In addition, we exposed cIAP2^−/− mice to a second mode of LPS-induced toxicity, where treatment with DGLN sensitizes mice to endotoxic shock. In contrast to LPS alone, LPS with DGLN caused a rapid demise of the mice (<3 h, Table 3). In all cases, cIAP2^−/− and control littermates demonstrated similar sensitivity and died at identical rates.

cIAP2^−/− mice display an attenuated inflammatory response.

LPS directly activates macrophages to produce large amounts of IL-1β and TNF-α and to mediate a cascade of events leading to endotoxic shock. We therefore assayed the levels of these proinflammatory cytokines in serum from cIAP2^−/− mice treated with LPS (35 mg of LPS/kg). In cIAP2^−/− mice, IL-1β levels in serum peaked at 4 h and then markedly dropped off (Table 4). This was in contrast to littermate controls, where IL-1β levels decreased much later. Likewise, comparable initial TNF-α serum levels were observed in both groups; however, the TNF-α levels dropped off to approximately 10 pg/ml in cIAP2^−/− mice by 10 h, whereas in littermate controls the TNF-α serum levels stabilized to approximately 400 pg/ml and were maintained until death (Table 4). As a further study we also investigated the concentrations of IL-12 (Table 4) in serum after an injection of LPS (35 g/kg). Yet again, we observed an attenuation of a macrophage cytokine, IL-12, by the 6-h time point. Nevertheless, cIAP2^−/− mice did display early outward signs of sepsis, such as eye exudates and ruffled fur; however, their condition quickly ameliorated, corresponding to the observed waning of the LPS-induced inflammatory cytokines seen within the cIAP2^−/− mice.

Macrophage cytokine production and cell counts are not impaired in cIAP2^−/− mice.

The inability of the cIAP2^−/− mice to sustain IL-1β and TNF-α levels in serum in response to LPS suggests a dysfunction of the macrophages. This dysfunction may be correlated to a reduced initial number of macrophages, to a block in the LPS-induced signaling pathway, or to an increased apoptotic susceptibility of the cIAP2^−/− macrophages. However, cIAP2^−/− mice have comparable initial cell count numbers of peritoneum- and spleen-derived macrophages relative to control littermates, as assessed by trypan blue exclusion and the Diff Quik stain kit and flow cytometry (Fig. 4A). The cIAP2^−/− and wild-type macrophages also stained similarly for the LPS-binding receptor, CD14 (Fig. 4B). Moreover, isolated splenic B cells derived from cIAP2^−/− mice proliferated normally in response to various concentrations of LPS (0.1 to 100 μg/ml) (Fig. 4C). In addition, primary cultures of cIAP2^−/− peritoneal macrophages generated comparable levels of TNF-α and IL-1β as littermate controls when exposed to various doses of LPS (0.1 to 1,000 ng/ml) at 10 and 24 h in vitro (Fig. 4D to G).

cIAP2^−/− macrophages and T cells are unresponsive to antiapoptotic LPS and IL-7 signals.

Despite the demonstrated antiapoptotic properties of cIAP2, purified B cells, T cells, and mouse embryonic fibroblasts from cIAP2^−/− mice displayed no significant differences in susceptibility to a variety of apoptotic triggers (α-Fas antibody, C2-ceramide, and dexamethasone) in vitro compared to wild-type cells (data not shown). However, since cIAP2 is strongly inducible by NF-κB activation, it could be argued that the antiapoptotic properties of cIAP2 could be observed only under appropriate conditions. Macrophages exposed to LPS normally show an increased vigor and resistance toward various apoptotic triggers (3). Indeed, macrophages derived from wild-type mice and pretreated with LPS showed the expected resistance to Fas-induced death compared to wild-type-derived macrophages that were not preexposed to LPS (Fig. 5A) as observed by TUNEL (terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling) staining. In sharp contrast, macrophages derived from cIAP2^−/−mice displayed no difference in their ability to resist Fas-mediated killing with or without preexposure to LPS (Fig. 5A). This suggests that LPS-induced upregulation of cIAP2, within a macrophage, was the essential protective component against Fas mediated death and is in agreement with the findings of Cui et al. (3), who demonstrated that cIAP2 upregulation in vitro was at least in part responsible for the general antiapoptotic resistance displayed by an LPS-induced macrophage. A similar dependency on cIAP2 was also observed with another key regulatory immune cell, the T cell (20). T cells derived from wild-type littermates and preexposed to IL-7 (5 or 10 ng/ml) showed a dose-dependent resistance against dexamethasone-induced death (Fig. 5B). In contrast, T cells derived from cIAP2^−/− mice with or without IL-7 pretreatment displayed no difference in the resistance to apoptosis (Fig. 5B).

cIAP2^−/− macrophages display an increased sensitivity to apoptosis.

We have demonstrated that macrophages derived from cIAP2^−/− mice (i) had typical initial cell count numbers compared to control littermates, (ii) produced normal levels of proinflammatory cytokines in response to LPS in vitro, and (iii) were highly susceptible to apoptotic triggers relative to control littermates when activated by LPS. It was also found that both cIAP2^−/− and control littermates were equally sensitive to PAF, an inflammatory mediator that acts downstream of the LPS-induced activation of macrophages. Given these observations, we predicted that cIAP2^−/− mice were resistant to endotoxic shock due to the inability of the cIAP2^−/−-mouse-derived macrophages to upregulate cIAP2, thus leading to a loss of viability and hence a loss of the ability of the cIAP2^−/− mice to produce a lethal inflammatory response. Therefore, in cIAP2^−/− mice injected with a normal lethal dose of LPS (35 mg/kg) the expected results would be either a rapid loss of the macrophage populations and/or an increased apoptotic state of macrophages from cIAP2^−/− mice relative to control littermates.

To determine the apoptotic sensitivity of cIAP2^−/− macrophages within an LPS-induced proinflammatory environment, the peritoneal macrophage cell numbers before and during an endotoxin-elicited response were assessed by trypan blue exclusion, Diff Quik staining, and flow cytometry (using the macrophage marker PE-conjugated α-F4/80 antibody). In addition, the apoptotic status of the peritoneal and splenic macrophages derived from LPS-injected cIAP2^−/− and littermate control mice was also assessed via flow cytometry (using the macrophage marker PE conjugated α-F4/80 antibody and FITC-conjugated annexin V). cIAP2^−/− mice demonstrated a markedly reduced number of peritoneum-derived macrophages at 5 h after LPS injection relative to littermate controls (Fig. 5C). However, total lymphocyte cell count numbers (B and T cells) were comparatively unaffected in both animal types (Fig. 5D). In addition, at 5 h after LPS injection both the peritoneal and the splenic macrophages from cIAP2^−/− mice stained ∼100% positive for annexin V (Table 5). Therefore, peritoneal and splenic macrophages from cIAP2^−/− mice undergoing LPS-induced endotoxic shock are highly sensitive to apoptotic stimuli compared to control littermates in vivo.

Liver-derived macrophages are key mediators of LPS-induced endotoxic shock (7, 12). Therefore, to further investigate the survival status of macrophages within LPS-treated cIAP2^−/− mice, we examined cell count levels of liver-derived macrophages within liver sections. Macrophage cell count numbers within both cIAP2^−/− and control mice were similar at 2 h after i.p. LPS treatment (Fig. 6A): 235 ± 35 and 271 ± 29, respectively (n = 5, average of three sections [average of five fields per section] per mouse; P < 0.01). However, at the 6-h time point (Fig. 6B) there was a precipitous drop of liver-derived macrophages of the cIAP2^−/− mice (21 ± 13) versus that of the control littermates (148 ± 19; n = 5, average of three sections [average of five fields per section] per mouse; P < 0.01]). Therefore, like peritoneal and splenic macrophages, the liver-derived macrophages from cIAP2^−/− mice are highly sensitive to LPS in vivo.

LPS confers apoptotic resistance to macrophages via induction of cIAP2 protein.

The observed rapid induction of cIAP2 in macrophages in response to LPS activation suggested that, at least in part, cIAP2 might be a key resistance component for maintaining macrophage viability under apoptotic conditions. Indeed, the in vivo findings presented here clearly demonstrate that cIAP2 has a critical antiapoptotic role in sustaining macrophage viability. Peritoneal macrophages derived from either cIAP2^−/− or control littermates displayed similar sensitivity to Fas-induced death; however, when pretreated with LPS, only macrophages from cIAP2^+/+ mice displayed an increase resistance to Fas-induced apoptosis.

Ablation of cIAP2 renders macrophages susceptible to apoptosis during endotoxic shock in vivo.

We have shown that LPS activation of macrophages induces a prompt and robust induction of cIAP2. Moreover, loss of cIAP2 protects mice from acute endotoxic shock, and this is associated with the effects of cIAP2 loss on the deaths of macrophages that normally produce large amounts of proinflammatory cytokines. In support of this notion, peritoneal macrophage numbers from cIAP2^−/− mice are reduced considerably by the 5-h time point after LPS injection relative to cIAP2^−/− mice at time zero hours. More importantly, peritoneal macrophage numbers from cIAP2^−/− mice are lower by the 5-h time point after LPS injection relative to wild-type mice at the same time point time. In addition, the peritoneal and splenic macrophages from cIAP2^−/− mice 5 h after LPS injection were ∼100% apoptotic for both macrophage populations.

cIAP2^−/− mouse-derived macrophage cytokine levels are normal in response to LPS treatment in vitro.

The loss of cytokine production in cIAP2^−/− mice treated with a lethal dose of LPS may be due to a signaling dysfunction of the macrophages. Similar to the results obtained for cIAP2^−/− mice, TLR4-null (8) and MyD88-deficient mice (9) have also been found to be resistant to LPS-induced endotoxic shock. However, B cells isolated from these animals failed to proliferate in response to LPS, whereas B cells from cIAP2^−/− mice responded normally to LPS. Furthermore, cultured macrophages derived from either TLR4- or MyD88-null mice were unable to produce proinflammatory cytokines. Clearly, in the case of TLR4 and MyD88 deficiency, the observed resistance to endotoxic shock is due to a block of the LPS-induced activation pathway of the macrophage. In contrast, cIAP2^−/− mouse macrophages exposed to LPS generate normal levels of TNF-α and IL-1β, suggesting that the classical LPS-induced NF-κB pathway is intact in macrophages lacking cIAP2.

Proposed mechanism of action: cIAP2^−/− mice resist endotoxic shock.

LPS challenge of mice causes the activation of multiple types of genes, including the upregulation of the survival genes. The proinflammatory response generates an inherently hostile environment that can be lethal to both pathogen and host immune cell. Therefore, expression of prosurvival genes is probably vital to maintain macrophage viability during an immune response. The inability to upregulate cIAP2 renders LPS-activated macrophages highly susceptible to apoptotic triggers, thereby quickly eliminating the resident macrophage population soon after the initiation of a systemic inflammatory response. This leads to the loss of the principal source of proinflammatory cytokines and subsequently to the attenuation of the immune response, preventing the development of multiple organ failure.

Unlike cIAP1 and XIAP, the results presented here indicate that cIAP2 regulation is dependent upon signal transduction pathways. An upregulation of cIAP2 mRNA was elicited in T cells upon exposure to IL-7, and a dramatic increase was observed in macrophages treated with LPS. There is a possibility that the cIAPs and XIAP may be able to functionally “stand in” for one another. Although the caspases inhibited by the IAPs coincide, cIAP1 and -2 bind caspases with significantly lower affinities than XIAP. In addition, each IAP has unique properties and cellular localizations. XIAP is involved in the TAK1/JNK1 signaling cascade (21), whereas the cIAPs associate with TRAFs (19). In addition, and more importantly, the observed differences in IAP regulation, demonstrated here, serve to underscore the nonredundant physiological functions of the IAPs and indicate that these proteins cannot completely functionally substitute for each other.

Possible therapeutic applications.

The results presented here suggest that antagonizing cIAP2 expression and/or function may have therapeutic benefit in patients with sepsis. Neither IL-1β- nor TNF-α-deficient mice alone are resistant to LPS-induced endotoxic shock (4, 15, 18). The ablation of cIAP2 expression results not only in a loss of sustained IL-1β production but also in a loss of TNF-α. Therefore, a pharmacological ablation of cIAP2 will potentially limit the severity of inflammatory diseases by transiently abolishing IL-1β- and TNF-α-producing macrophages. These findings may be extended to other macrophage-dependent inflammatory disorders such as colitis.

Our results suggest that cIAP2 is a highly regulated protein whereby its apoptotic inhibitory properties can be observed only under a suitable situation. In addition, the cIAP2-inducing agents, LPS and IL-7, imbue their target cells, macrophages and T cells, respectively, with an increased apoptotic resistance. More importantly, cIAP2-null macrophages and T cells are unable to respond to these protective signals, indicating that cIAP2 is the crucial protective component.