Suppression of STING Associated with LKB1 Loss in KRAS-Driven Lung Cancer

Abstract

KRAS-driven lung cancers frequently inactivate TP53 and/or STK11/LKB1, defining tumor subclasses with emerging clinical relevance. Specifically, KRAS-LKB1 (KL)-mutant lung cancers are particularly aggressive, lack PD-L1, and respond poorly to immune checkpoint blockade (ICB). The mechanistic basis for this impaired immunogenicity, despite the overall high mutational load of KRAS-mutant lung cancers, remains obscure. Here, we report that LKB1 loss results in marked silencing of stimulator of interferon genes (STING) expression and insensitivity to cytoplasmic double-strand DNA (dsDNA) sensing. This effect is mediated at least in part by hyperactivation of DNMT1 and EZH2 activity related to elevated S-adenylmethionine levels and reinforced by DNMT1 upregulation. Ectopic expression of STING in KL cells engages IRF3 and STAT1 signaling downstream of TBK1 and impairs cellular fitness, due to the pathologic accumulation of cytoplasmic mitochondrial dsDNA associated with mitochondrial dysfunction. Thus, silencing of STING avoids these negative consequences of LKB1 inactivation, while facilitating immune escape. SIGNIFICANCE: Oncogenic KRAS-mutant lung cancers remain treatment-refractory and are resistant to ICB in the setting of LKB1 loss. These results begin to uncover the key underlying mechanism and identify strategies to restore STING expression, with important therapeutic implications because mitochondrial dysfunction is an obligate component of this tumor subtype.See related commentary by Corte and Byers, p. 16.This article is highlighted in the In This Issue feature, p. 1.

Figure 1.

LKB1 modulates STING expression in KRAS mutant lung cancer. A, Venn diagram showing differentially expressed genes between KL and KP in The Cancer Genome Atlas (TCGA) and Cancer Cell Line Encyclopedia (CCLE). Top-ranked gene signatures derived from differentially expressed genes are represented. B, Schematic of dsRNA and dsDNA sensing pathways that induce type 1 IFN. C, Relative RPKM values of genes related to these pathways in KL and KP cells from CCLE. D, Representative STING IHC images of primary KL and KP non-small cell lung cancer (NSCLC) samples (upper panel). Insets highlight tumor cell STING expression. STING intensity in cancer cells was scored in a blinded manner (lower panel) on a scale of 0-3. IHC0: No staining, IHC1: low staining, IHC2: moderate staining, IHC3: high staining. E, Immunoblot (IB) of the indicated proteins in KL (red) or KP (blue) cells (left). KL cell lines with an asterisk contain p53 mutation. STING expression was quantified by ImageJ and normalized to β-actin (right). F, qRT-PCR of STING in KL or KP cells. Each point represents one cell line. G, IB of the indicated proteins in KP cells transduced with scramble or LKB1 sgRNA. H,I, IB of the indicated proteins in KL cells transduced with the indicated vectors. p-values were calculated by unpaired two-tailed Student’s t test (C,E,F), or Fisher’s exact test (D), *p<0.05, **p<0.01.

Figure 2.

Hyperactivation of DNMT1 and EZH2 suppress STING expression in KL cells. A, qRT-PCR of STING in KL cells transduced with GFP or LKB1. B, IB of the indicated proteins in KL cells. C, Dot plot heatmaps showing DNA methylation levels in the STING promoter between KL and KP cells in CCLE (red = hypermethylation). The location of each position is as follows, P1: 5:138861649, P2: 5:138861807, P3: 5:138862442, P4: 5:138862470. D, E, Levels of DNA methylation (D) or DNMT1 binding (E) within the 5’ UTR of STING (see methods) normalized to input in KL cells. F, qRT-PCR of STING in KL cells treated with 100 nM DAC, 5 μM GSK126, 5 μM KDM5-C70 or 500 nM UNC0638 for 7 d. G, H, IB of the indicated proteins (G) or qRT-PCR of STING (H) in KL cells transduced with GFP or LKB1, and treated ± 100 nM DAC for 7 d. I, J, H3K27me3 levels at 5’ UTR of STING normalized to the input in KL cells transduced with GFP or LKB1 (I) or treated with 5 μM GSK126 for 7 d (J) (n=4 replicates from two independent experiments). K, IB of the indicated proteins in KL cells treated with 100 nM DAC and/or 5 μM GSK126 for 7 d. L, Measurement of SAM in A549 cells treated with 100 nM DAC and/or 5 μM GSK126 for 3 d. (n=4 replicates, representative of two independent experiments). p-values were calculated by one-way (E,F,H,L), two-way (A) ANOVA followed by Tukey’s post-hoc test, or unpaired two-tailed Student’s t test (D,I,J) *p<0.05, **p<0.01.

Figure 3.

Defective sDNA sensing and impaired T cell chemotaxis due to LKB1 inactivation. A, IB of the indicated proteins in KL cells transduced with GFP or LKB1, and treated ± 1 μg/ml poly (dA:dT) for 4 hr. B, ELISA of human IFN-β, CXCL10 or CCL5 levels in conditioned medium (CM) derived from KL cells transduced with GFP or LKB1, and treated ± 1 μg/ml poly (dA:dT) for 24 hr. C, IB of the indicated proteins in A549 cells transduced with GFP or LKB1, pre-treated ± 100 nM DAC for 7 d, and treated ± 1 μg/ml poly (dA:dT) for 4 hr. D, ELISA of human CXCL10 levels in CM derived from A549 cells transduced with GFP or LKB1, pre-treated ± 100 nM DAC for 7 d, and treated ± 1 μg/ml poly (dA:dT) for 24h. E, Quantification of Jurkat CXCR3 infiltration into H1355 tumor spheroids (See supplementary Fig S3I, J). Values were normalized to each control. F, G, PD-L1 expression in H1944 cells transduced with NanoLuc or LKB1 (F), or A549 cells transduced with NanoLuc or LKB1, pre-treated ± 100 nM DAC for 7 d (G), and treated ± 125 ng/ml poly (dA:dT) for 12 hr. MFI: Mean fluorescence intensity. ΔMFI: (dAdT-Ctr)/Ctr. Data are representative of three independent experiments. H,I,J, IHC images and analysis from primary LKB1 negative; STING IHC0/1 (n=12) and LKB1 positive; STING IHC2/3 NSCLC (n=22) samples (See Figure 1D). Red arrows highlight stained CD8+ T cells in both tumor epithelium (red) and stroma (green) (H). PathAI (see Methods) was used to quantify CD3+/CD4+/CD8+ T-cell infiltration (I) and tumor PD-L1 expression (J). p-values were calculated by two-way (B,D,E) ANOVA followed by Tukey’s post-hoc test, or unpaired two-tailed Student’s t test (F,G,I,J). *p<0.05, **p<0.01.

Figure 4.

STING is poorly tolerated by KRAS-LKB1 mutant cells due to the pathologic accumulation of cytoplasmic dsDNA. A,B, IB of the indicated proteins (left) and quantification (right) of pTBK1 induction (A) or qRT-PCR of CXCL10 (B) in KL (red) or KP (blue) cells 7d following STING over-expression. (n=4 replicates from two independent experiments). C, CXCL10 ELISA in CM derived from KL cells 7 d following STING over-expression ± 1 μM TBK1 inhibitor compound 1 for 72 hr. D,E, Relative cell number 9 d following STING over-expression ± 1 μM Ruxo or 1 μM TBK1 inhibitor compound 1 (n>3, biological replicates) following STING over-expression (n=4, biological replicates). F, IB of the indicated proteins 7 d following STING re-expression ± 1 μM Ruxo treatment. G, H2009, H1944, or H2122 cells stained with PicoGreen, Mitotracker, and Hoechst (left). Scale bars: 3 μm. Signal intensity at each region of interest (ROI) was quantified by ImageJ (right). H, qPCR of mtDNA in cytoplasmic fraction (n=6, biological replicates). I, Parental H2122, H2122 Rho 0, or H2122 cells transduced with scramble or POLG sgRNA stained with PicoGreen, Mitotracker, and Hoechst (left). Scale bars: 3 μm. Signal intensity at each ROI was quantified by ImageJ (right). J, Heat map of cytokine profiles in CM 7 d following STING re-expression. CMs were collected 72 hr after medium change. Scores = ratio of log₂ fold change following STING re-expression relative to parental H2122 (left) or H2122 transduced with scramble sgRNA (right). K, IB of the indicated proteins 7 d following STING re-expression. L, Relative cell number 9 d following STING over-expression (n=3, biological replicates). Values of STING-overexpressing cells were normalized to each control. p-values were calculated by one-way (A,G,H,I) or two-way (B,C,D,E,L) ANOVA followed by Tukey’s post-hoc test. *p<0.05, **p<0.01.