The Scaffolding Protein IQGAP1 Interacts with NLRC3 and Inhibits Type I IFN Production

Abstract

Sensing of cytosolic nucleotides is a critical initial step in the elaboration of type I IFN. One of several upstream receptors, cyclic GMP-AMP synthase, binds to cytosolic DNA and generates dicyclic nucleotides that act as secondary messengers. These secondary messengers bind directly to stimulator of IFN genes (STING). STING recruits TNFR-associated NF-κB kinase-binding kinase 1 which acts as a critical node that allows for efficient activation of IFN regulatory factors to drive the antiviral transcriptome. NLRC3 is a recently characterized nucleotide-binding domain, leucine-rich repeat containing protein (NLR) that negatively regulates the type I IFN pathway by inhibiting subcellular redistribution and effective signaling of STING, thus blunting the transcription of type I IFNs. NLRC3 is predominantly expressed in lymphoid and myeloid cells. IQGAP1 was identified as a putative interacting partner of NLRC3 through yeast two-hybrid screening. In this article, we show that IQGAP1 associates with NLRC3 and can disrupt the NLRC3-STING interaction in the cytosol of human epithelial cells. Furthermore, knockdown of IQGAP1 in THP1 and HeLa cells causes significantly more IFN-β production in response to cytosolic nucleic acids. This result phenocopies NLRC3-deficient macrophages and fibroblasts and short hairpin RNA knockdown of NLRC3 in THP1 cells. Our findings suggest that IQGAP1 is a novel regulator of type I IFN production, possibly via interacting with NLRC3 in human monocytic and epithelial cells.

Figure 1

NLRC3 yeast two-hybrid constructs. Schematic representation of both bait (NLRC3) and prey (IQGAP1) constructs used to screen with a human leukocyte and activated mononuclear cell or human universal cDNA libraries. Abreviations used for NLRC3 are CARD= caspase recruitment and activation domain, NOD= nucleotide oligomerization domain, NBD= nucleotide binding domain, HD1= helical domain 1, WHD= winged helix domain, HD2= helical domain 2, LRR= leucine rich repeats; the asterisk and star represent the approximate locations of the Walker A and Walker B motifs. Abreviations used for IQGAP1 are CHD= calponin homology domain, GRD= rasGAP related domain, RGCT= rasGAP carboxy terminus. Amino acid alignment represents the minimal binding region of IQGAP1, amino acids 1438–1657.

Figure 2

IQGAP1 and NLRC3 are preferentially expressed in the immune system. RT-PCR was used to determine the tissue expression of human IQGAP1 and NLRC3. GAPDH was used as a positive control. BioGPS was queried for human IQGAP1 and NLRC3 expression in specific immune cells. The top overlapping cells and tissues for IQGAP1 and NLRC3 were graphed. B) Microarray results show relative IQGAP1 expression is highest in myeloid cells with moderate levels in lymphocytes. C) Relative NLRC3 expression is highest in lymphocytes with moderate expression in myeloid cells. Arrows have been used to illustrate the difference expression levels in primary monocytic cells. Abbreviations in panel A, B and C: NTC= no template control, BM=bone marrow, PBL=peripheral blood leukocyte, PB=peripheral blood, NK= natural killer cell. D and E) Across tissue expression for IQGAP1 (probe 200791_s_at) and NLRC3 (probe 236295_s_at), using immune cells and tissues, are shown. Z-scores ± median absolute deviation is plotted. A dashed line represents a z-score of 5 which signifies that a given transcript is expressed in a tissue. Arrows depict cells of monocyte/macrophage lineage. Abbreviations used in panel C and D: BM=bone marrow, DC=dendritic cell, Treg= regulatory T cell, CB=cord blood, DC=dendritic cell, GC=germinal center, NK=natural killer, PBMC=peripheral blood mononuclear cell. F) QPCR analysis of NLRC3 expression in human transformed cells lines. Values are normalized to HPRT levels using ΔC_t method. Asterisk= p<.05 when compared to NTC

Figure 3

IQGAP1 specifically interacts with NLRC3 in human cells. HEK293T cells were transiently transfected with different epitope tagged full length and truncation versions of IQGAP1 and NLRC3. Lysates were immunoprecipitated for IQGAP1 and immunoblots were probed to detect NLRC3 or vice versa. A) The RGCT domain (amino acids 1438–1657) of IQGAP1 interacts with full length NLRC3. B) The NBD (amino acids 60–646) of NLRC3 interacts with full length IQGAP1. C and D) Full length IQGAP1 (either amino terminal FLAG or GFP tagged) interacts with HA-tagged NLRC3 but not with HA tagged IL-1β, CIITA, NLRX1 nor NOD2. E) Reciprocal co-immunoprecipitations confirm the interaction between IQGAP1 and NLRC3 but not an irrelevant protein such as IL-1β. F) Mutations in the Walker A and Walker B motifs (WAB) do not alter the ability of NLRC3 to interact with IQGAP1. Images are representative of triplicate experiments.

Figure 4

Confocal microscopy of HeLa cells transfected with epitope tagged IQGAP1-FLAG or NLRC3-HA. Cells were fixed, permeabilized and stained with antibodies specific to the epitope tag followed by species-specific Alexa Fluor 488 and Alexa 568 secondary antibodies. Cells are representative of more than three independent transfections. An endoplasmic reticulum specific fusion protein (Addgene Plasmid #56310) was used to highlight the endoplasmic reticulum. The Golgi Apparatus was stained with an anti-Golgin-97 antibody; whereas the mitochondria was identified with an anti-Tomm20 antibody. The white arrow represents NLRC3-HA aggregates seen in a percentage of cells

Figure 5

Confocal microscopy of HeLa cells co-transfected with FLAG-tagged IQGAP1 and HA-tagged NLRC3 (A) and stained for endogenous proteins: B) phalloidin conjugated to Alexa 647 C) tubulin D) Endogenous IQGAP1 boxed inset shows robust cortical colocalization with NLRC3. E) shRNA scramble F) shIQGAP1 HeLa cells were transfected with NLRC3-HA and stained for endogenous IQGAP1 and epitope tagged NLRC3 and phalloidin. White arrows represent cortical actin staining of NLRC3 and IQGAP1 in panel E. G) SW480 cells were transfected with epitope tagged NLRC3, epitope tagged IQGAP1 and cotransfected. Cells were fractionated using subcellular fractionation kit. Immunoblots were probed with antibodies to specific proteins in each fraction. Images are representative of at least three individual replicates.

Figure 6

IQGAP1 and NLRC3 show partial colocalization with recombinant STING, TBK1 and not MAVS. HeLa cells were transiently transfected different expression constructs. Panels A–D) Confocal microscopy of HeLa cells co-transfected with NLRC3 and associated expression constructs. Panels A and B show two distinct colocalization patterns of STING: (A) puncta and (B) endoplasmic reticulum.

Figure 7

shRNA mediated knockdown of IQGAP1 in different cell lines induces hyperactivation of the IFN-β pathway. In panels A–C) Stable integrants (panel A and D represent THP-1 cells, panel B and E represent THP-1 dual cells and panel C represents HeLa cells) were selected and stimulated as indicated. Interferon or cytokine levels were measured by ELISA and secreted Lucia^™. Values represent means ± standard deviation of triplicate samples. Data are representative of at least three biological replicates. * represents a p<.05 when compared to Scramble shRNA cells. Panels D and E show no significant differences relative to the scramble shRNA.

Figure 8

IQGAP1 sequesters NLRC3 to the actin cytoskeleton. A) HEK293T cells were transiently transfected with epitope tagged IQGAP1, NLRC3 and STING. Lysates were immunoprecipitated for NLRC3-HA and blotted for interacting proteins. Data are representative of three replicates. B) IQGAP1 and STING were cotransfected in the absence of NLRC3-HA and immunoprecipitated for IQGAP1 and blotted for STING. Cells were lysed in stringent (PBS with 1% Triton X-100 + 0.5% Sodium Deoxycholate shown in lane 1) and less stringent (PBS with 1% Triton X-100 shown in lane 2) conditions. Data are representative of three replicates.