Caspase-12 Dampens the Immune Response to Malaria Independently of the Inflammasome by Targeting NF-κB Signaling

Caused by the Plasmodium protozoan parasite, malaria is a major threat to global health with an estimated 500 million new cases each year, resulting in up to 2 million deaths in Africa alone. A poor understanding of host–parasite interactions exacerbates the malaria problem. It is well documented that a Th1 response to parasitized RBCs (pRBCs) is required for pathogen clearance (1) and that an overzealous inflammatory response contributes to the onset and pathogenesis of cerebral malaria (CM) (2, 3). However, the factors involved in malaria pathogenesis are varied and largely misunderstood. Recently, the inflammasome has surfaced as a mediator of the malarial inflammatory response. Specifically, it was reported that hemozoin triggers inflammasome activation in ex vivo macrophage cultures (4, 5).

Inflammasomes are large macromolecular complexes scaffolded by the cytosolic Nod-like receptors (NLRs). They trigger inflammation by activating caspase-1, which cleaves pro–IL-1β and pro–IL-18 into their mature active forms. We have previously shown that a second member of the caspase-1 subfamily, caspase-12, which is catalytically inactive in humans and attenuated in rodents (6), acts as an inhibitor of both the inflammasome and NF-κB pathways (7–9). Expression of human caspase-12 is predominately confined to African descendents and is associated with dampened proinflammatory cytokine production and sepsis-related mortality (8). Although critical to the host response during bacterial infection (10), the role of the inflammatory caspases has yet to be well characterized in parasitic infections. Furthermore, the confinement of caspase-12 to regions where malaria is endemic renders the understanding of its function in the immune response to Plasmodium essential.

In this study, we identified a role for caspase-12 in suppressing the inflammatory response to malaria. Caspase-12 limited the immune control of parasite replication and dampened cerebral hyperinflammation in a model of CM. Unexpectedly, our experiments revealed that although the inflammasome was stimulated by Plasmodium infection, its activity was not essential for host defense. In contrast, caspase-12 deficiency led to hyperactivation of NF-κB and enhanced IFN-γ production, which mediated the phenotype of casp12^−/− mice. Mechanistically, caspase-12 competed with NF-κB essential modulator (NEMO) for association with IκB kinase (IKK)-α/β, effectively preventing the formation of the IKK complex and inhibiting downstream transcriptional activation by NF-κB.

Casp12^−/− and Casp1^−/− mice (7, 11) were back-crossed to the C57BL/6 background. Female mice were used throughout the study. Plasmodium chabaudi AS was maintained as previously described (12); mice were inoculated i.p. with 10⁶ pRBCs. Where indicated, 125 μg IFN-γ neutralizing Ab clone H22 (BD Biosciences, San Jose, CA) was administered i.p. 1 h prior to infection and again on day 6 postinfection (p.i.). The IKK inhibitor PS-1145 (Sigma-Aldrich, St. Louis, MO) was administered i.p. at 10 mg/kg daily on days 6 to 11 p.i. Plasmodium berghei ANKA-infected RBCs were passaged in a C57BL/6 mouse to ∼10% parasitemia, and mice were infected i.p. with 10² to 10³ pRBCs.

Bone marrow-derived macrophages (BMDMs) were generated as described (13), and splenocytes were prepared as in Refs. 7 and 9. RAW264.7 cells, splenocytes, or BMDMs were incubated with freshly harvested uninfected RBCs (uRBCs) or pRBCs at the indicated ratios. In some experiments, cells were prestimulated with cytochalasin D (Sigma-Aldrich) at 1 μg/ml for 20 min. For phagocytosis assays, macrophages were plated on sterile coverslips for 24 h then exposed for 2 h to uRBCs or pRBCs at an RBC:macrophage ratio of 20:1. Nonphagocytosed RBCs were lysed in H₂O, and the cells were fixed and stained with Hemacolor (EMD Chemicals, Gibbstown, NJ). For examination of signaling pathways, whole-cell lysates were extracted in Laemmli buffer at the indicated time points and probed by immunoblot with Abs for ERK, pERK, IκB-α, pIκB-α (Cell Signaling, Danvers, MA), and actin (MP Biomedical, Solon, OH). Densitometric analysis was performed using ImageJ (National Institutes of Health, Bethesda, MD) software and expressed as arbitrary units relative to actin. Plasmodium falciparum parasite cultures of the laboratory clone ITG were maintained using standard procedures (14). Human PBMCs from consenting African-American donors collected as in Ref. 8 were stimulated with uRBCs or P. falciparum pRBCs at 1:3 PBMC to pfRBC ratio for 12 and 24 h.

HEK293T cells were transfected with κB-luciferase and β-gal reporter plasmids and a plasmid encoding either rat (r)Casp12, rCasp12 (C-A), human (h)Casp12L-GFP along with TNFR associated factor-6 (TRAF-6), BCL-10, IKK-α, IKK-β, or empty vector. Where indicated, cells were treated 24 h posttransfection with 10 ng/ml TNF-α (PeproTech, Rocky Hill, NJ) for 6 h. Luciferase and β-galactosidase activity was measured in whole-cell lysates.

HEK293T cells were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Whole-cell lysates were harvested 24 h later and incubated overnight with either M2 agarose beads (Sigma-Aldrich) or the immunoprecipitating Ab along with protein G Sepharose beads (Sigma-Aldrich). Immunoprecipitates were eluted with Flag peptide (Sigma-Aldrich) or by boiling in Laemmli buffer. Eluted proteins were detected by immunoblot using Abs for rat caspase-12 (Sigma-Aldrich), NEMO (Santa Cruz, Santa Cruz, CA) or GFP, Flag, or Myc tags (Roche, Basel, Switzerland).

Cytokine production in the supernatant of splenocytes was measured by Milliplex bead-based multiplex assays (Millipore, Billerica, MA) and that of PBMCs by Cytometric Bead Arrays (BD Biosciences) according to the manufacturers’ instructions. Mouse serum cytokines were quantified by ELISA assays from R&D Systems (Minneapolis, MN) (IL-1β), MBL (Woburn, MA) (IL-18), and BD Biosciences (IFN-γ, TNF-α, and IL-10). P. chabaudi pRBC Ag was prepared as described (15), coated on assay plates, and binding of serum IgG1 or IgG2a detected with isotype/subclass-specific HRP-bound Abs (SouthernBiotech, Birmingham, AL).

Total RNA from brain or spleen was reverse-transcribed using random hexamers and MML-V Reverse-Transcriptase (Invitrogen). Quantitative PCR (qPCR) was performed using iTaq SYBR Green Supermix (Bio-Rad, Philadelphia, PA) with primers as in Ref. 9. Fold induction was calculated over uninfected levels using the ΔΔ-Ct method.

Data is represented as average ± SE. Two-tailed Student t test was used to evaluate statistical significance between groups. Log-rank test was used for survival analysis. Fisher’s exact test was used for mortality analysis.

To examine the role of caspase-12 in blood-stage malaria, wild-type and casp12^−/− mice were infected with P. chabaudi AS, and the percentage of pRBCs in the peripheral blood was monitored. Casp12^−/− mice had lower peak parasitemia on day 10 and resolved the infection significantly faster than wild-type animals (Fig. 1A). Moreover, casp12^−/− mice experienced less severe body weight loss at the peak of infection compared with wild-type mice (Supplemental Fig. 1A).

Control of parasitemia depends on an adequate Th1-polarized cytokine response and the production of parasite-specific IgG2a isotype Abs (1). To determine whether the suppressive effect of caspase-12 on parasite replication and clearance mechanisms was mediated by a dampened Th1 response, cytokine production was quantified on days 2, 5, 6, and 7 p.i., when parasitemia was comparable between the two genotypes. Fig. 1B shows that casp12^−/− animals exhibited an exaggerated cytokine induction as early as day 5 p.i. Additionally, deficiency in caspase-12 led to significantly higher levels of malaria-specific IgG2a but not IgG1 (Fig. 1C). Notably, despite their enhanced production of proinflammatory cytokines, casp12^−/− animals did not demonstrate secondary inflammatory organ damage during infection as assessed by blinded histological analysis and quantification of alanine transaminase levels in the serum (Supplemental Fig. 1B, 1C). To rule out indirect systemic effects, we examined the response of splenocytes from wild-type and casp12^−/− mice harvested on day 6 p.i. to pRBC rechallenge. It has been previously reported that whereas splenocytes extracted from naive mice react poorly to pRBCs ex vivo, those derived from Plasmodium-infected mice trigger a robust inflammatory response upon rechallenge (16), presumably due to Ag-driven cell expansion and activation. Multiplex analysis of the culture supernatants revealed that caspase-12 deficiency led to excessive IFN-γ, TNF-α, and IL-10 in response to pRBC stimulation (Fig. 1D). Most notable was the induction of IL-1β that was produced at relatively low levels by wild-type splenocytes but at 10-fold higher levels by casp12^−/− splenocytes. This was not due to increased pro–IL-1β synthesis as determined by qPCR of the pro–IL-1β transcript (Supplemental Fig. 1D). Altogether, these results indicate that caspase-12 exerts important suppressive effects on the immune response to malaria parasites.

Because caspase-12 targets both the inflammasome and NF-κB activation (7, 8, 17), we set to dissect the relative contribution of these pathways in mediating the enhanced response of casp12^−/− mice to malaria. The increased production of IL-18 and IL-1β in casp12^−/− mice suggested heightened inflammasome activation by P. chabaudi. Consistently, challenge of RAW264.7 macrophages with P. chabaudi pRBCs led to caspase-1 activation and IL-1β processing in a dose- and time-dependent manner (Fig. 2A). Pretreatment with the actin-destabilizing agent cytochalasin D abrogated this response (Fig. 2B), indicating that internalization of pRBCs or pRBC-derived products was necessary for inflammasome activation. Indeed, we observed extensive macrophage phagocytosis of pRBCs but very limited uptake of uRBCs (Supplemental Fig. 2A). To determine the effect of caspase-12 on caspase-1 activation in response to P. chabaudi pRBCs, we challenged BMDMs from wild-type, casp1^−/−, or casp12^−/− mice with either uRBCs or pRBCs. Macrophages of all genotypes had equivalent pRBC phagocytosis (Supplemental Fig. 2B), however, caspase-1 activation was highest in casp12^−/− cells (Fig. 2C), further supporting a role of caspase-12 in inhibiting caspase-1 catalysis.

To determine the role of caspase-1 activation in response to pRBCs in vivo, we infected casp1^−/− mice with P. chabaudi and followed the course of parasitemia. Surprisingly, casp1^−/− mice showed no deficiency in their ability to control blood parasitemia (Fig. 2D). Pathological features, such as anemia and weight loss, were also not different between wild-type and casp1^−/− mice (Supplemental Fig. 2C). Moreover, P. chabaudi induced equivalent systemic levels of IFN-γ, TNF-α, and IL-10 in both genotypes, despite differential induction of IL-18, which was significantly lower in casp1^−/− mice compared with that in wild-type mice (Fig. 2E). Concordantly, serum levels of parasite-specific IgG1 and IgG2a (Fig. 2F) and cytokine production by pRBC-rechallenged splenocytes (Supplemental Fig. 2D) did not differ between genotypes. These results indicated that the caspase-1 inflammasome is dispensable during Plasmodium infection. Nonetheless, hyperactivation of caspase-1 in the context of caspase-12 deficiency might contribute to host resistance. To investigate this possibility (and because casp1^−/−/casp12^−/− mice are difficult to obtain because of the genomic proximity of their corresponding genes), we resorted to inhibiting caspase-1 pharmacologically, throughout the course of infection, by treating casp12^−/− mice with the selective caspase-1 inhibitor Ac-Tyr-Val-Ala-Asp-aldehyde (Ac-YVAD-CHO). To control for the efficacy of this treatment, we measured serum IL-18 levels as a read-out of caspase-1 activation. Treatment with Ac-YVAD-CHO resulted in reduced serum IL-18 levels, consistent with caspase-1 inhibition (Supplemental Fig. 3A), but did not significantly alter the heightened parasite clearance abilities of casp12^−/− mice (Supplemental Fig. 3B), the systemic production of IFN-γ (Supplemental Fig. 3C), or their enhanced production of parasite-specific IgG2a (Supplemental Fig. 3D). These data support that, during P. chabaudi infection, caspase-1 activation occurs as a collateral response but is dispensable for parasite clearance or host resistance and that the effects of caspase-12 are not restricted to the caspase-1 pathway.

Several lines of evidence link exacerbated inflammation to the onset of CM symptoms (18). The results obtained in the P. chabaudi model prompted us to investigate whether caspase-12 expression provided protection from CM. We infected wild-type and casp12^−/− mice with P. berghei ANKA, an established model of experimental CM (19). C57BL/6 mice are highly susceptible to P. berghei ANKA and exhibit many of the neurologic symptoms elicited in human CM patients (20), usually dying within 8–10 d p.i. (Supplemental Fig. 4A). Because of this dominant genetic effect, we titrated pRBCs to attain an infectious dose (multiplicity of infection 10² to 10³) at which wild-type mice survived past day 10 p.i., though eventually succumbing to CM. Using this dose range, casp12^−/− mice were more susceptible to CM (Fig. 3A) despite generally, albeit not significantly, lower parasitemia levels in the casp12^−/− mice (Fig. 3B). The enhanced susceptibility to CM correlated with exaggerated production of proinflammatory cytokines and chemokines by casp12^−/− mice compared with that in wild-type animals. This was determined by measuring serum IL-1β levels (Fig. 3C) and induction of TNF-α, IL-6, MCP-1, and MIP-1 in the brain on day 10 p.i. (Fig. 3D). Notably, caspase-12 is induced in the brain in response to P. berghei ANKA (2) (Fig. 3E), suggesting an immunoregulatory process to control the excessive inflammatory response to the Plasmodium infection.

As for P. chabaudi, caspase-1 activity was dispensable in response to P. berghei ANKA, which is consistent with recent reports (21, 22). Casp1^−/− animals were as susceptible to CM as wild-type animals (Fig. 3A, Supplemental Fig. 4A) and showed no differences in cytokine production (Fig. 3C, 3D). Our findings in casp1^−/− animals, using the two distinct parasites P. chabaudi and P. berghei ANKA, indicated no significant role for caspase-1 in host defense to Plasmodium or in CM pathology.

The geographical distribution of the ancestral form of human caspase-12 led us to hypothesize that resistance to CM might have contributed to maintenance of its expression in regions endemic for malaria. Our results indicated that caspase-12 modulated the host response to malaria in experimental animal models. To consolidate these findings and to circumvent mouse genetics, differences in the immune response, and the use of rodent-specific parasites, we examined the role of the human caspase-12 ancestral variant in the inflammatory response to P. falciparum. PBMCs from African-American donors carrying Casp12T or Casp12C alleles were challenged ex vivo with P. falciparum pRBCs. Figure 3F shows that PBMCs from Casp12T/C heterozygote carriers had a dampened inflammatory response to P. falciparum compared with those derived from Casp12T/T individuals, independently of an effect on phagocytosis (Supplemental Fig. 4B).

Activation of NF-κB and MAPK signaling pathways has been demonstrated in response to a number of Plasmodium-derived factors including GPI (23), hemozoin (22), and pRBCs (24, 25). Consistently, we found that both NF-κB and MAPK signaling pathways were upregulated in pRBC-stimulated BMDMs (Fig. 4A). Notably, caspase-12 deficiency in these cells had no effect on ERK activation but resulted in a more robust phosphorylation and degradation of IκB-α, indicative of enhanced NF-κB signaling (Fig. 4A). Concordantly, P. chabaudi infection of wild-type and casp12^−/− mice resulted in enhanced activation of NF-κB in vivo, as shown by a more robust splenic induction of NF-κB target genes, including IκB-α, Bcl-xl, and COX-2, in casp12^−/− mice compared with that in wild-type animals (Fig. 4B).

To test whether enhanced activation of NF-κB mediated the heightened proinflammatory response of casp12^−/− mice in malaria, we treated wild-type and casp12^−/− mice with the IKK inhibitor PS-1145 daily from days 6 to 10 p.i. PS-1145 abrogated the protective effects of caspase-12 deficiency, resulting in equivalent peak parasitemia (Fig. 4C) and parasite-specific Igs at the end of the infection (Fig. 4D and Supplemental Fig. 5) between PS-1145–treated casp12^−/− and wild-type mice. During P. chabaudi infection, the reduced parasite levels observed in casp12^−/− animals were associated with enhanced cytokine synthesis, including hyperproduction of IFN-γ (Fig. 1). IFN-γ is critical for the immune control of parasitemia (26). Furthermore, IFN-γ has been shown to potentiate the NF-κB–dependent proinflammatory response to Plasmodium-derived molecules (27). To assess the physiological relevance of IFN-γ hyperproduction, we administered anti–IFN-γ neutralizing Abs to casp12^−/− mice over the course of P. chabaudi infection. Our results show that IFN-γ neutralization rendered casp12^−/− mice as susceptible to infection as wild-type mice (Fig 4E, 4F). Altogether, these data indicate that the NF-κB pathway is critical for eliciting Th1-dependent immunity to P. chabaudi and that caspase-12 specifically blunts this response.

Activation of NF-κB is central to multiple signaling pathways and occurs in response to disparate proinflammatory triggers, such as cytokines, Ags, and microbial-associated molecular patterns. To dissect the molecular mechanism of caspase-12’s inhibitory effect on NF-κB, we first set to determine whether caspase-12 is a universal inhibitor of NF-κB signaling. Using a κB-luciferase reporter system in HEK293T cells, we show that both human and rodent caspase-12 suppressed NF-κB transcriptional activity (Fig. 5A). Caspase-12 inhibited NF-κB in response to TNF-α stimulation or overexpression of TRAF-6 or BCL-10, key components of the TLR/IL-1R and TCR and BCR (TCR/BCR) pathways, respectively. These results indicated a role for caspase-12 in a common downstream event. Consistently, caspase-12 also inhibited NF-κB in response to overexpression of IKK-α or IKK-β (Fig. 5A), suggesting that it functioned at the level of the IKK complex. Notably, a catalytically inactive mutant of caspase-12 (rCasp12-Cys299Ala) was as efficient as wild-type caspase-12 in blocking NF-κB (Fig. 5A), confirming previous reports that the catalytic activity of caspase-12 is confined to autoprocessing and is dispensable for its immunomodulatory functions (7, 9). To investigate the possible involvement of caspase-12 in modulating the IKK complex, we coexpressed caspase-12 with IKK-α, IKK-β, or NEMO (IKK-γ) in HEK293T cells and examined their association by coimmunoprecipitation. Caspase-12 coimmunoprecipitated with both IKK-α and IKK-β but not with NEMO (Fig. 5B). We used RIP-2 as a positive control as we have previously shown that caspase-12 downregulates NOD signaling by binding to RIP-2 and displacing TRAF-6 from the nodosome (17). To address the consequence of caspase-12 binding to IKK-α/β, we next examined whether NEMO and caspase-12 competed for association with IKK-α/β. Fig. 4C shows that overexpression of NEMO led to diminished interaction between caspase-12 and IKK-α/β. Reciprocally, caspase-12 expression diminished the association between endogenous NEMO and the phosphorylated forms of IKK-α/β in a dose-dependent manner (Fig. 5D). Collectively, these data indicate that caspase-12 dampens NF-κB signaling by targeting the central IKK complex.

Caspase-12 is phylogenetically related to the inflammatory caspase family, which includes caspase-1, -4, and -5 in humans and caspase-1 and -11 in mice. These proteases are implicated in mediating inflammatory signaling pathways and the resulting cytokine production, but their exact role and importance in the host response to parasitic infection is still unclear. We have previously shown that caspase-12 predisposes to bacterial sepsis in both humans and animal models and prevents the proper clearance of bacterial pathogens (7). Here we show that caspase-12 also limits the elimination of parasitic infections, such as malaria. Compared with wild-type mice, casp12^−/− mice produced strikingly higher levels of cytokines, such as IFN-γ, which mediated their ability to more efficiently control parasite replication and clearance. We further demonstrate that the protective phenotype of the casp12^−/− animals is the result of a de-repression of NF-κB signaling and that caspase-12 acts as an inhibitor of the IKK complex by displacing NEMO from IKK-α/β.

We have previously reported that in the context of bacterial infections, caspase-12 is an inhibitor of the caspase-1 pathway (7). We show that the inflammasome and caspase-1 are activated by Plasmodium in vivo and in macrophages cultured in the presence of P. chabaudi-infected pRBCs. Expression of caspase-12 inhibited this response, blunting caspase-1 activation and the production of mature IL-1β and IL-18. Many studies have attempted to identify the “malaria toxin” responsible for the induction of inflammation. Two important targets have been Plasmodium GPI, a membrane glycolipid, and hemozoin, β-hematin crystals formed during heme detoxification, both capable of inducing cytokine production by isolated macrophages. Recently, purified hemozoin was reported to activate the NLRP3 inflammasome and induce inflammasome-dependent inflammation when injected in vivo (4, 5). Both our in vitro and in vivo data indicate that the inflammasome is activated during malaria infection. However, our findings in casp1^−/− animals indicate no significant importance for caspase-1 in mediating the host defense mechanisms to Plasmodium as these mice show no deficiency in either their innate or adaptive immune responses to the parasite. During Plasmodium infection, pRBCs are phagocytosed, and pRBC-derived products accumulate within tissues and macrophages (28), likely contributing to increased caspase-1 activity. Yet despite consistent evidence that parasite-derived products, such as GPI and hemozoin, are capable of inducing inflammation in isolated cell culture, there is little evidence and much debate as to their importance in vivo.

In addition to its function in the inflammasome, caspase-12 has been implicated in the downregulation of NF-κB activity (8). A critical regulator of cytokine-mediated cellular responses, NF-κB is activated downstream of a number of proinflammatory pathways and has been shown to be activated by in vitro exposure to Plasmodium-derived products (22–25). Here we show NF-κB activation by P. chabaudi both in vitro and in vivo, which, in both cases, is enhanced in the context of caspase-12 deficiency. We further demonstrate that it is the enhanced NF-κB activity in the casp12^−/− mice that is responsible for their resistant phenotype. At rest, NF-κB is sequestered in the cytoplasm by the IκB family of inhibitory proteins. Upon receptor stimulation, activation of the IKK complex (IKK-α/IKK-β/NEMO) leads to the phophorylation and ubiquitination of IκB, promoting its degradation and the release of NF-κB, which translocates to the nucleus to catalyze target gene transcription. Despite being a key event in NF-κB activation, the molecular regulation of IKK activity is still poorly understood. Here we demonstrate that caspase-12 functions as a direct inhibitor of NF-κB by interfering with the formation of the IKK-α/IKK-β/NEMO complex. Our data support a model in which caspase-12 competes with NEMO for IKK-α/β binding, effectively displacing NEMO from the complex and thus preventing the subsequent degradation of IκB and NF-κB translocation. This function is similar to that of the recently described NLRC5, a caspase activation and recruitment domain-containing member of the NLR family that competes with NEMO for IKK-α/β binding (29). Furthermore, expression of both NLRC5 and caspase-12 is induced by NF-κB–activating signals (7, 17, 29), thus forming a negative-feedback loop that directly targets IKK activity to limit inflammation.

Several lines of evidence support a role for excessive inflammation in the onset of cerebral malaria (3, 18). We demonstrate an exaggerated susceptibility to CM in casp12^−/− mice, consistent with the hyperinflammatory response mounted by these animals during Plasmodium infection. Caspase-12–deficient mice succumb to CM-related death more rapidly than their wild-type counterparts, despite generally lower parasite load. Thus, expression of caspase-12 appears protective in this context. Caspase-12 has previously been reported to be upregulated in the brains of P. berghei-infected mice (30). Consistent with our own observations, this is suggestive of a possible immune-regulatory process whereby caspase-12 production is induced in an attempt to control the excessive inflammatory response to the Plasmodium infection.

Caspase-1 activity was dispensable for CM-associated mortality as casp1^−/− mice were not protected from CM. Our findings are consistent with earlier reports showing equal susceptibility of wild-type and casp1^−/− mice to P. berghei ANKA at the conventional dose of 10⁶ pRBCs (21, 31). We further demonstrate that even at significantly lower infectious doses, caspase-1 activity does not contribute to CM mortality, parasitemia, or brain cytokine production. Recent investigations into the role of the inflammasome in malaria pathogenesis reported reduced mortality of Nlrp3^−/− mice when infected with P. chabaudi Adami (5) or a low dose of P. berghei sporozoites (4). Although a dampened inflammatory response in Nlrp3^−/− animals was proposed as the protective mechanism, there is little in vivo data to support this conclusion. Notably, Reimer et al. (21) recently demonstrated that whereas casp1^−/− mice did not differ from wild-type mice in the context of P. berghei ANKA infection (which is consistent with our results), Nlrp3^−/− mice were significantly more resistant. Therefore, findings in malaria-infected Nlrp3^−/− mice should be interpreted cautiously and not assumed to be caspase-1 dependent.

The full-length caspase-12 is expressed by the “ancestral” allele predominately in Africa and in African descendents but also in Southeast Asia and in Central and South America (32). The “derived” form of the gene has spread recently throughout the human population due to positive selection, and it is proposed that its selective advantage is resistance to sepsis (32). Using PBMCs from healthy donors carrying the ancestral Casp12C allele, we show that human caspase-12 also blunts the inflammatory response to Plasmodium challenge. We have previously reported that expression of human caspase-12 functionally substitutes for the deficiency in mouse caspase-12 in a model of Listeria monocytogenes infection (9). We have expanded on this finding to confirm that human and mouse caspase-12 are functionally equivalent in response to a parasitic infection and that despite significant genetic differences between the human pathogen, P. falciparum, and the rodent parasites, P. chabaudi and P. berghei, the results derived from mouse models translate to a setting where human cells and a human pathogen were used. Our findings are of importance in the development of novel treatments and therapies for malaria as caspase-12 expression can significantly influence the immune response to infection. This is particularly true for the African population. Recently, McCall et al. (33) attempted to investigate the association between the casp12 genotype and clinical malaria parameters in West African populations. Unfortunately, the small sample size resulted in few or no Casp12C homozygous individuals, and inconsistent methodologies with regard to age, sex, infectious status, and clinical history led to conflicting results, precluding any definitive conclusion as to the role of the human caspase-12 polymorphism in malaria. A full-scale human population genetics study would help clarify the relationship between caspase-12 and resistance to CM and would provide further support as to whether the ancestral casp12 allele was maintained in the African population by this pressure.

We thank Dr. Richard Flavell for providing casp1^−/− mice and Dr. Philippe Gros for the P. berghei ANKA parasite.

Disclosures The authors have no financial conflicts of interest.