Extensive evolutionary and functional diversity among mammalian AIM2-like receptors

Abstract

Innate immune detection of nucleic acids is important for initiation of antiviral responses. Detection of intracellular DNA activates STING-dependent type I interferons (IFNs) and the ASC-dependent inflammasome. Certain members of the AIM2-like receptor (ALR) gene family contribute to each of these pathways, but most ALRs remain uncharacterized. Here, we identify five novel murine ALRs and perform a phylogenetic analysis of mammalian ALRs, revealing a remarkable diversification of these receptors among mammals. We characterize the expression, localization, and functions of the murine and human ALRs and identify novel activators of STING-dependent IFNs and the ASC-dependent inflammasome. These findings validate ALRs as key activators of the antiviral response and provide an evolutionary and functional framework for understanding their roles in innate immunity.

Figure 1.

STING is required for sensing of natural DNA, cyclic dinucleotides and DMXAA. (A) Quantitative RT-PCR analysis of IFN-β mRNA expression in WT or Tmem173^−/− BMDMs (left) and primary, early passage MEFs (right); cells were treated with the indicated ligands and harvested for analysis every 2 h. (B) WT, Tmem173^−/−, Mavs^−/−, and Tmem173^−/− Mavs^−/− BMDMs were stimulated with the indicated ligands for 4 h. (C) BMDCs of the indicated genotypes were generated and stimulated with the indicated ligands for 4 h before harvest and quantitative RT-PCR analysis. Data are representative of two to three experiments, each with triplicate treatments for every time point (error bars represent the SEM). Statistical analysis was performed at the 6-h time point (A) and 4-h time point (B and C) with a two-way ANOVA test. ***, P < 0.001; ****, P < 0.0001; n.s. = not statistically significant (P > 0.05). Significant values in panels B and C are identical to those found in Fig. 1 A and are omitted for figure clarity.

Figure 2.

Evolutionary relationships of the ALR gene family in mammals. (A) Genomic loci containing ALRs in selected mammalian species (loci not drawn to scale). Boxes represent coding exons, with those encoding Pyrin domains in red and HIN domains in blue. Introns are omitted for clarity. ALRs from species other than mouse and human are arbitrarily named. Gray horizontal lines show which exon–exon boundaries are supported by mRNA or EST evidence; other boundaries are predicted computationally. Short diagonal lines indicate gaps in the genome assemblies. Flanking genes CADM3 (green) and SPTA1 (yellow) have numerous exons that are not all presented here. Olfactory receptor genes and pseudogenes (black boxes) are also found at one end of the locus and are labeled according to their family and subfamily (e.g., 10AA = family 10, subfamily AA) using the HORDE system. A species tree (not to scale) is shown on the left, along with the approximate divergence times. The position of horse on the tree is uncertain (Meredith et al., 2011); we have depicted it in the position that is best supported by current evidence. Chromosomal locations of the locus are in parentheses, with a “−” symbol denoting that the human locus is depicted as flipped with respect to its numbered chromosomal orientation. (B) Using the horse AIM2 as a query in blast searches, we detected fragments of AIM2 pseudogenes in the genomes of cow, sheep, and dolphin, whereas pig Aim2 appears to have only retained a recognizable exon 2. Cat and dog genomes only have a partial exon 3. (C) Three selected portions of a multi-species AIM2 alignment showing inactivating mutations shared by more than one species (red boxes). Highlighted bases are conserved in >50% of the sequences shown. These shared mutations show that the AIM2 gene was inactivated in the carnivore lineage before cat and dog diverged (∼56 mya) and in at least some species of the Cetartiodactyla lineage since before dolphins, sheep, and cattle diverged (∼58 mya). (D and E) Phylogenetic trees of the indicated mammalian ALRs based on Pyrin domains (D) and HIN domains (E). Species-specific ALR gene expansions are indicated in yellow boxes. Bootstrap values, calculated as described in the Materials and methods, are indicated at each branch point.

Figure 3.

Expression and inducibility of murine ALR mRNAs. (A) Basal expression of each ALR mRNA in untreated WT and Ifnar1^−/− BMDM, calculated relative to HPRT expression. (B) IFN-inducibility of murine ALRs: the basal expression level of each ALR was set to “1,” and the fold induction of each ALR mRNA over untreated in WT and Ifnar1^−/− BMDM is shown in response to a 6-h stimulation with 100 U/ml recombinant mouse IFN-β. (C) Ligand inducibility: CT DNA (1) or RIG-I Ligand (2). Results are representative of three experiments, with triplicate treatments in each experiment (error bars represent the SD [A] and the SEM [B and C]).

Figure 4.

Murine IFI204 is not a nonredundant sensor of DNA. (A) Knockdown efficiency in HEK293T cells transiently transfected with 500 ng HA-tagged IFI204 and 2 µg siRNA #1, siRNA #2, or control siRNA for 24 h. IFI204 protein knockdown was assessed by anti-HA immunoblot (IB). (B) Quantitative real-time PCR of IFI204 mRNA expression in BMDM (left) and MEF (right) stably expressing the indicated lentiviral siRNA constructs. (C) Quantitative RT-PCR analysis of IFN-β mRNA expression in BMDM and MEF stably expressing the indicated lentiviral siRNAs and treated as indicated for 4 h before harvest. Results are representative of two independent experiments, each with triplicate treatments for each ligand (error bars represent the SD [B] and the SEM [C]). Statistical analysis was performed with a two-way ANOVA test, and none of the stimulated knockdown cells reached a statistical significance of P < 0.05 compared with control cells.

Figure 5.

Intracellular localization of murine ALR proteins. (A) 500 ng ALR-HA expression plasmids alone (left two columns) or 500 ng ALR-HA, 500 ng ASC-eGFP, and 500 ng STING-FLAG (right columns) were transfected into 3.0 × 10⁴ HeLa cells and probed 24 h after transfection with anti-HA-ALR (red), anti-eGFP (green), anti-FLAG (blue), and DAPI (gray). (B) HA-tagged murine MNDA was immunoprecipitated from control cells and cells expressing STING-FLAG and ASC-FLAG, which can be distinguished based on their size despite the same tag. (C) HA-tagged Pyr-A was immunoprecipitated from cells expressing the indicated constructs, followed by blotting for STING-FLAG and ASC-FLAG. All data are representative of three to five independent experiments.

Figure 6.

Unbiased functional characterization of all mouse and human ALRs. (A) ISRE luciferase reporter assay in HEK293T cells transiently transfected with ISRE luciferase reporter and increasing amounts of ALR-HA expression vectors with 300 ng STING-FLAG vector, measured 24 h after transfection. Results are expressed as fold induction over ISRE-luciferase alone. (B) IL-1β ELISA of HEK293T cells stably expressing murine ASC-FLAG and murine caspase-1, transfected with increasing amounts of ALR-HA expression vectors and 100 ng pro-IL-1β vector. (C) Plot of ISRE luciferase induction versus IL-1β secretion for each ALR. (D) Plot of ISRE-luciferase induction versus IL-1β secretion for each human ALR. Results are representative of 2–5 experiments, with each treatment done in triplicate (error bars represent the SEM [A] and the SD [B]).