figure
posted on 2024-03-25, 14:20 authored by Akash R. Boda, Arthur J. Liu, Susana Castro-Pando, Benjamin T. Whitfield, Jeffrey J. Molldrem, Gheath Al-Atrash, Maria Emilia Di Francesco, Philip Jones, Casey R. Ager, Michael A. Curran Novel STING alleles fail to respond to STING agonists. HEK-Blue-ISG-KO-STING cells were reconstituted with indicated novel STING variants (n = 2; A) or STING variants plus WT STING by retroviral transduction (B), then were exposed to 100 µg/mL indicated CDN for 24 hours (n = 2). IRF3/9 activity was measured by SEAP assay. IFNα positive control indicates presence and functionality of the IRF3/9 SEAP reporter in these cells. Statistical significance in B was calculated using repeated measures one-way ANOVA test. ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. C, STING knockout 293 T cells were transduced to express human WT STING with an HA or 6xHIS tag, the STING-I3R isoform with a 6xHIS tag, or the STING-Ex4D isoform with a 6xHIS tag. Cell lysates were Western blotted with anti-6xHIS antibody (Invitrogen) followed by anti-mouse HRP (Cell Signaling Technology). Anti-human β-actin staining is shown as a loading control (Abcam).
Funding
HHS | National Institutes of Health (NIH)
History
ARTICLE ABSTRACT
Lack of robust activation of Stimulator of Interferon Genes (STING) pathway and subsequent induction of type I IFN responses is considered a barrier to antitumor immunity in acute myeloid leukemia (AML). Using common human AML cell lines as in vitro tools to evaluate the efficacy of novel STING agonists, we found most AML lines to be poor producers of IFNs upon exposure to extremely potent agonists, suggesting cell-intrinsic suppression of STING signaling may occur. We observed unexpected patterns of response that did not correlate with levels of STING pathway components or of known enzymes associated with resistance. To identify a genetic basis for these observations, we cloned and sequenced STING from the cDNA of human AML cell lines and found both frequent mutations and deviations from normal RNA splicing. We identified two novel spliced isoforms of STING in these lines and validated their expression in primary human AML samples. When transduced into reporter cells, these novel STING isoforms exhibited complete insensitivity to agonist stimulation. These observations identify alternative splicing as a mechanism of STING pathway suppression and suggest that most AML silences the STING pathway through direct modification rather than through engagement of external inhibitory factors.
We find that AML acquires resistance to innate immune activation via the STING pathway through aberrant splicing of the STING transcript including two novel forms described herein that act as dominant negatives. These data broaden understanding of how cancers evolve STING resistance, and suggest that the AML tumor microenvironment, not the cancer cell, should be the target of therapeutic interventions to activate STING.
