Advertisement
Commentary
Open Access | 
10.1172/JCI169598
Aspirin and immunotherapy: a Faustian bargain?
Eric A. Goethe,1 Amy B. Heimberger,2 and Ganesh Rao1
1Baylor College of Medicine Department of Neurosurgery, Houston, Texas, USA.
2Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA.
Address correspondence to: Ganesh Rao, Baylor College of Medicine Department of Neurosurgery, 7200 Cambridge St Ste 9A, Houston, Texas 77030, USA. Email: grao@bcm.edu.
Find articles by Goethe, E. in: JCI | PubMed | Google Scholar
1Baylor College of Medicine Department of Neurosurgery, Houston, Texas, USA.
2Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA.
Address correspondence to: Ganesh Rao, Baylor College of Medicine Department of Neurosurgery, 7200 Cambridge St Ste 9A, Houston, Texas 77030, USA. Email: grao@bcm.edu.
Find articles by
Heimberger, A.
in:
JCI
|
PubMed
|
Google Scholar
|
1Baylor College of Medicine Department of Neurosurgery, Houston, Texas, USA.
2Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA.
Address correspondence to: Ganesh Rao, Baylor College of Medicine Department of Neurosurgery, 7200 Cambridge St Ste 9A, Houston, Texas 77030, USA. Email: grao@bcm.edu.
Find articles by Rao, G. in: JCI | PubMed | Google Scholar
Published May 1, 2023 - More info
J Clin Invest. 2023;133(9):e169598. https://doi.org/10.1172/JCI169598.
© 2023 Goethe et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
Related article:
Abstract
How cancer cells evade the therapeutic effects of immune checkpoint blockade is largely unknown. Here, we report that fibrinogen-like protein 1 (FGL1), a newly identified immune checkpoint ligand, was modified by acetylation at Lys 98 in hepatocellular carcinoma (HCC), which targeted it for proteasomal degradation. Sirtuin 2 (SIRT2) deacetylated and stabilized FGL1, thus promoting immune evasion. Notably, the SIRT2 inhibitor 2-Cyano-3-[5-(2,5-dichlorophenyl)-2-furanyl]-N-5-quinolinyl-2-propenamide (AGK2) enhanced acetylation of FGL1 and reduced FGL1 protein levels in vitro. The combination of AGK2 and programmed death ligand 1 (PD-L1) blockade effectively suppressed tumor growth and improved overall survival of mice. Furthermore, aspirin, an old drug, could directly acetylate FGL1 at Lys 98 and promote its degradation in vitro. Aspirin enhanced the immunotherapeutic efficacy, induced tumor regression, and extended the lifespan of tumor-bearing mice. Furthermore, the SIRT2/FGL1 axis was expressed in HCC specimens. Collectively, these findings unveil an acetylation-mediated regulation of FGL1, identify a potential target for HCC immunotherapy, and provide therapeutic strategies for the clinical treatment of HCC.
Authors
Mingen Lin, Jing He, Xinchao Zhang, Xue Sun, Wenjing Dong, Ruonan Zhang, Yanping Xu, Lei Lv
-
The role of FGL1 in hepatocellular carcinoma
Cancer evades detection by the immune system by utilizing various modulators, such as programmed cell death protein 1 (PD-1) and lymphocyte-activating gene 3 (LAG-3) (1). LAG-3 is expressed on the surface of T cells, and its activation suppresses T cell antigen–mediated activation, signaling, and proliferation (2, 3). Fibrinogen-like protein 1 (FGL1) has been identified as a ligand of LAG-3 and is implicated in several malignancies (1, 2). Antibody-mediated knockdown of FGL1 expression has been shown to increase antigen-mediated T cell activation. However, the role of FGL1 in cancer is likely contextual and lineage dependent, since its inhibition promotes growth in lung cancer cells but exerts antitumoral effects in gastric and colon cancer (2, 3).
In this issue of the JCI, Lin and colleagues (4) evaluated the role of FGL1 in hepatocellular carcinoma (HCC). Nicotinamide (NAM) treatment resulted in the acetylation of FGL1 at Lys-98, which reduced the levels of FGL1 in HCC cell lines due to increased proteasomal degradation. Expression of sirtuin 2 (SIRT2), a NAD+-dependent deacetylase, triggered a decrease in FGL1 acetylation, while inhibition of SIRT2, by either the specific inhibitor acetylglutamate kinase 2 (AGK2) or an siRNA, resulted in increased acetylation and decreased levels of FGL1. Similarly, aspirin was shown in vitro to acetylate FGL1, leading to its proteasomal degradation (Figure 1). AGK2 was associated with increased T cell–mediated killing of HCC cells in vitro and prolonged survival, alone, and in synergy with PD-L1 inhibition, in mice. Likewise, aspirin was associated with enhanced T cell destruction of HCC cells. The authors also demonstrated improved survival in mice with HCC that were treated with aspirin and a PD-L1 inhibitor. AGK2 and aspirin were equally effective with LAG-3 inhibition at suppressing tumor growth (4).
Figure 1Acetylation of FGL1 has a regulatory role in HCC antitumor immunity. Acetylation of FGL1 at Lys-98 reduces the levels of FGL1 in HCC cells due to increased proteasomal degradation. The NAD+-dependent deacetylase SIRT2 decreases FGL1 acetylation, which facilitates interaction between LAG-3 to promote evasion from immune surveillance. In contrast, compounds that increase acetylated FGL1, including the SIRT2 inhibitor AGK2, NAM, and aspirin, increase FGL1 degradation and antitumor immunity via T cell–mediated killing.
The Lin et al. study mechanistically shows that FGL1 can be regulated by acetylation at K98 via AGK2 or aspirin treatment in vitro and excluded other acetylation sites or mechanisms of degradation. While the authors provide mechanistic support for this result, one potential weakness was the lack of assessment for FGL1 acetylation in the in vivo experiments, or in assays of T cell–mediated tumor killing, raising the possibility that antitumor immunity results from some unmeasured effect. However, the same agents shown to acetylate FGL1 in vitro were used for these experiments, and the in vivo experiments demonstrated reduced FGL1 expression by immunohistochemistry (4).
FGL family in cancer. Both FGL1 and FGL2 share the duality of modulating both tumorigenesis and immune responses. FGL1 was originally cloned from HCC, and its expression is elevated in melanoma, colorectal, prostate, lung, gastric, and breast cancers based on BioGPS tissue microarrays and The Cancer Genome Atlas (TCGA) database (5). FGL1 has been identified as a major ligand for LAG-3. FGL1 can suppress Akt signaling, but FGL1 can be suppressed by IL-6–mediated JAK2/STAT3 signal activation (6). FGL2 was previously shown by our group to be involved in the malignant transformation of low- to high-grade gliomas and is expressed in other cancers, such as HCC, prostate, B cell lymphoma, colorectal, and lung cancers (7). Knockdown or silencing of FGL2 expression has also been shown to inhibit prostate and lung cancer growth in preclinical models (8–10). FGL2 can drive tumorigenesis through ERK1/-2 and the MAPK kinase pathways. The immune-modulatory functions of FGL2 relative to FGL1 are more extensively described, in general and in the context of cancer. During viral infection, FGL2–/– mice possess a higher frequency of DCs expressing CD80, CD86, and MHC; they also show decreased PD-1 expression on T cells and increased B and T cell effector responses (11). We have shown through a series of studies of preclinical glioma that FGL2 increases the expression of PD-1 and the frequency of tumor-supportive macrophages and Tregs, suppresses the development of CD103+ DCs, and blocks the recruitment of tumor-specific, brain-resident memory T cells (7, 12, 13). In other cancers, FGL2 expression was associated with better prognostic outcomes and increased lung cancer infiltration of immune cells, such as CD8+ T cells, macrophages, B cells, and DCs (14). As such, both FGL1 and FGL2 have been widely reported to share both pro- and antitumor functions.
The role of FGL1 in tumorigenesis. The results of Lin et al. indicate that FGL1 expression is a negative prognosticator in HCC, as demonstrated by the association between FGL1 inhibition, antitumor immunity, and survival (4). Notably, these results contrast with prior findings in which downregulation of FGL1 in HCC increases the rates of tumor formation (15–17), and is a positive prognosticator (15). The discrepancy between the present study and others may be a function of analyzing the different ways in which FGL1 influences the development and progression of HCC. FGL1 loss is associated with tumor dedifferentiation and increased Akt pathway signaling in HCC and is probably an effect independent of antitumor immunity (17). Additionally, multiple other tumor suppressor loci are near FGL1, suggesting that loss of the gene (i.e., via chromosomal damage) may be linked to other tumor suppressor genes (17). Thus, the differences between studies may be a function of unaccounted coassociated tumor suppressor genes and the balance of antitumor and anti-immune responses in various models. The Lin et al. study (4) emphasizes the complexity of outcome biomarkers that play different and opposing roles in tumorigenesis. Loss of FGL1 may disinhibit HCC proliferation yet contribute to an antitumor immune response. As such, the role of FGL1 in HCC is less clear than previously presumed. Nonetheless, an alternative understanding of previously established tumor markers can drive further research and treatment development.
Copyright © 2025 American Society for Clinical Investigation
ISSN: 0021-9738 (print), 1558-8238 (online)