STING agonists as antiviral agents

Abstract

The cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS-STING) pathway plays a pivotal role in mounting an innate immune response against invading pathogens. Activation of this pathway by exogenous or endogenous stimuli triggers the downstream production of interferons and both pro-/anti-inflammatory cytokines. Over the past decade, hundreds of patents have been filed for the development and use of natural and synthetic STING agonists. For antivirals, synthetic STING agonists have been shown to be effective in both prophylactic and anaphylactic manners against viral infection and serve as vaccine adjuvants. This review summarizes the current application of STING agonists as antivirals to date against a variety of RNA and DNA viruses.

Introduction

Innate immunity is the first line of host defense against microbial pathogens. Upon infection of a cell, pattern recognition receptors (PRRs) respond to conserved microbial signatures called pathogen-associated molecular patterns (PAMPs) and to host damage-associated molecular patterns (DAMPs) [¹]. PRR families include Toll-like receptors (TLRs), the nucleotide binding and oligomerization, leucine-rich proteins (NLRs), retinoic acid-like receptors (RLRs), absent in melanoma 2 (AIM2)-like receptors (ALRs), and cytosolic DNA receptors [^1-6]. PRRs can detect exogenous pathogen molecular patterns such as lipopolysaccharides, lipoproteins, bacterial flagellin, and nucleic acids. Additionally, exogenous and endogenous DNA from apoptotic or damaged host cells can trigger the receptors [^7,8]. The detection of viral PAMPS and DAMPs during virus infection activates different signaling cascades depending on the nature of the stimulus. One of the cascades that is triggered by viruses is the DNA-sensing cGAS-STING pathway.

cGAS-STING pathway

Upon detection of pathogenic or host double-stranded DNA, cyclic GMP-AMP synthase (cGAS) catalyzes the formation of the cyclic dinucleotide 2′3′-cyclic GMP-AMP (2’3’-cGAMP) from ATP and GTP [^9,10]. 2’3’-cGAMP, an endogenous ligand, binds the stimulator of interferon genes (STING/TMEM173/MPYS/MITAS/ERIS) in the endoplasmic reticulum and triggers ER-Golgi trafficking of STING [^11-14]. At the Golgi, STING undergoes palmitoylation and recruits TANK-binding kinase 1 (TBK1), which triggers phosphorylation of itself and STING [^15-17]. Both interferon regulatory factor 3 (IRF3) and/or nuclear factor kappa B (NF-κB) pathways are activated, leading to the production of type I interferons and pro/anti-inflammatory cytokines [¹²]. Interferons bind their cognate receptor to activate the Janus kinase/signal transducer and activator of transcription proteins (JAK/STAT) pathway and transcription of interferon-stimulated genes (ISGs), which have many antiviral effector functions (Figure 1) [^18,19].

Figure 1. The cGAS-STING pathway.

Upon detection of pathogenic, cGAS synthesizes 2’3’-cGAMP from ATP and GTP. 2’3’-cGAMP binds STING in the endoplasmic reticulum and triggers ER-Golgi trafficking of STING. At the Golgi, STING recruits TBK1, which triggers phosphorylation of itself and STING. This activates the IRF3 and/or NF-κB, leading to the production of type I interferons and pro/anti-inflammatory cytokines. Interferons bind their cognate receptor to activate ISGs through the JAK-STAT pathway. Created in BioRender. Nelson, T. (2026) https://BioRender.com/7dwbe6o.

In addition to TBK1, STING activation also recruits IkappaB kinase complex (IKK), which phosphorylates IκBα (inhibitor of NF-κB alpha), leading to its degradation by the proteasome and allowing NF-κB to translocate to the nucleus [^20,21]. NF-κB signaling plays a role in multiple cellular processes such as proliferation, cell survival, T-cell activation, and inflammation response. Activation of NF-κB leads to the production of inflammatory cytokines (interleukins (IL)-6, IL-8, IL-12, and tumor necrosis factor-alpha (TNF-α)) and chemokines (IL-1β, RANTES, CXCL10) [^22-24]. NF-κB activation in macrophages, an essential innate immune cell for clearing infection, helps promote activation of inflammatory T-cells. The transcription of NF-κB-dependent genes NLRP3 (nod-like receptor family pyrin domain-containing 3) pro-IL-1β, and pro-IL-18 is necessary for the activation of the inflammasome [²⁵].

The NLRP3 inflammasome is a multiprotein complex composed of NLRP3, ASC (apoptosis-associated speck-like protein containing a caspase recruitment domain), and pro-caspase-1 that can be activated by the cGAS-STING pathway in different contexts [^25,26]. Both NF-κB and phosphorylated IRF3 following STING activation can translocate to the nucleus to increase NLRP3 expression and downstream ISG production further facilitates this interaction [^26-29]. STING can recruit and deubiquitinate the NLRP3 inflammasome during herpes simplex virus I (HSV-1) infection to induce an antiviral response, and mice deficient in NLRP3 were highly susceptible to herpes simplex virus 1 infection [³⁰]. On the contrary, Zika viral protein NS1 stabilizes caspase-1, promoting cleavage of cGAS and decreased interferon response to facilitate viral replication [³¹]. The impact of the cGAS/STING/NLRP3 axis needs to be examined in the context of each specific virus.

The cGAS-STING pathway has been studied for its role in controlling infection in a wide variety of viruses [summarized in ^32-36]. For its DNA-sensing ability, cGAS-STING is important for our antiviral defense against herpes simplex virus 1, Kaposi’s sarcoma-associated herpesvirus, and poxviruses, and mice deficient in the pathway are more susceptible to lethal infection [^37-39]. For RNA-sensing pathways, RLRs and specific TLRs are responsible for detecting RNA viruses and initiating response [⁴⁰]. However, STING has been shown to interact with the retinoic acid-inducible gene I/mitochondrial antiviral signaling protein (RIG-I/MAVS) complex and transmit its signals, serving as a central player in DNA-RNA-sensing crosstalk [^11,41,42].

There are multiple reports suggesting cGAS and STING play a role in the antiviral response against RNA viruses, as RNA viruses have been shown to block DNA-dependent interferon signaling [^12,31,43-48]. Like DNA viruses, deficiency in cGAS or STING supports replication of multiple RNA viruses such as vesicular stomatitis virus, Sendai virus, dengue virus, and West Nile virus [^12,44,49,50]. Mitochondrial damage by dengue virus leads to the release of mitochondrial DNA in the cytosol and subsequent activation of the cGAS-STING pathway [⁵¹]. Additionally, STING was determined to be required for interferon production against influenza A virus [⁴³]. Infection of retroviruses generates RNA:DNA hybrids and double-stranded DNA in the cytosol that can activate the pathway [^52-54]. It has also been shown that certain RNA viruses upregulate STING expression upon infection [⁴²], suggesting STING’s role in antiviral response against RNA viruses is context-dependent and its involvement is still largely unknown. However, this crosstalk suggests that STING agonists may be beneficial to both DNA and RNA viruses.

Overview of STING agonists

STING agonists can be broadly classified into cyclic dinucleotide (CDN) analogs, non-CDN agonists, and novel delivery systems. Currently, there are multiple STING agonists in clinical trials for cancer therapeutics, as STING is important for antitumor function [^55-57]. This section will discuss the STING agonists involved in antiviral studies; however, the current progress and development of STING agonists for cancer therapeutics have been reviewed in great detail by other authors [^58-64].

Synthetic CDNs were the first class to be developed and are designed to mimic the natural ligand of STING, 2’3’-cGAMP. DMXAA, known as dimethylxanthenone acetic acid or ASA404, is a mouse-specific STING agonist commonly used in the mouse viral infection models [^65-67]. DMXAA had promising results, but its failure to activate human STING prevented it from entering a clinical trial. ADU-S100, the first STING agonist in clinical trial for cancer, is an analog to 3’3’-cyclic adenosine monophosphate (c-di-AMP), the STING ligand produced by bacteria [^68,69]. ADU-S100 completed its phase I clinical trial and phase II was initiated, but efficacy concerns led to termination of the trial. More recently, MK-1454 is a potent CDN that exhibits robust interferon and cytokine response and has ongoing clinical trials, with phase I showing little safety and efficacy concerns [^70,71]. In addition to MK-1454, there are 12 more STING agonists entering clinical trials that could be explored for their antiviral potential. While these agonists have not been tested against viruses in the studies discussed below, their entry into clinical trials makes them potentially attractive antiviral therapeutics to be explored.

Non-CDNs are synthetic STING agonists that do not resemble the CDN backbone similar to 2’3’-cGAMP, although synthesis of these compounds may derive certain characteristics from 2’3’-cGAMP and its analogs. Diamidobenzimidazole (diABZI) is one of the most popular non-CDNs, derived from a class of amidobenzimidazoles using a phosphodiester linker, due to its potent induction of interferons and high solubility [⁷²]. diABZI-3 (known as diABZI STING agonist 1, Compound 3) is a tautomer and more potent form of diABZI-1 (commonly referred to as diABZI) used in viral or cancer studies, and diABZI-4 is an oral form more favorable for in vivo studies [^72,73]. Along with 2’3’-cGAMP and DMXAA, it is the most used STING agonist in antiviral studies.

STING agonists have also been discovered from natural products. α-Mangostin, a dietary xanthone isolated from mangosteen, is a novel activator of STING [⁷⁴]. Derivatives of benzofuran, an antipsychotic drug and extract of coal tar, were shown to activate STING and have broad-spectrum antiviral activity [⁷⁵]. Moreover, derivatives of benzothiophene, found in petroleum-related deposits, such as MSA-2 and its dimer have potent antitumor activity and can be delivered orally [^76,77].

Additionally, two novel delivery systems have been tested against viruses. NanoSTING is a nanoparticle formulation of 2’3’-cGAMP with stable systemic delivery, designed to serve as a mucosal vaccine adjuvant [⁷⁸]. Another study generated the exogenous STING pathway-activating complex (SPAC) is composed of membrane vesicles (MV) displaying multimeric STING on their surface (STING-MV) bound to a STING agonist [⁷⁹].

The chemical structures of STING agonists discussed in this review can be found in Table 1.

Table 1:

Chemical structures of STING agonists

Compound	Structure	PubChem ID/CAS No.	SMILES
2’3’-cGAMP		139035001	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C3NC(=NC4=O)N)OP(=O)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C6N=CN=C7N)O) OP(=O)(O1)O)O)O
3’3’-cGAMP		135471108	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C3N=C(NC4=O)N)O)O P(=O)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C(N=CN=C76)N)O)OP(=O)(O1)O)O
α-Mangostin		5281650	CC(=CCC1=C(C2=C(C=C1O)OC3=C(C2=O)C(=C(C(=C3)O)OC)CC=C(C)C)O)C
ADU-S100		135390762	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C(N=CN=C43)N)OP(=S)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C(N=CN=C76)N)O)OP(=O)(O1)[S-])[O-])O.[Na+].[Na+]
BZF-2OH (Benzofuran derivative)		n/a	n/a
BNBC		55171-63-6 (CAS No.)	n/a
C11		16272610	CC1=CC=C(C=C1)C2=NN=C(O2)SC(C3=CC=CC=C3)C(=O)NC(=O)NC
cAIMP		1507367-51-2 (CAS No)	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C3N=CNC4=O)O)OP(=O)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C(N=CN=C76)N)O)OP(=O)(O1)O)O
c-di-GMP		135440063	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C3N=C(NC4=O)N)O)OP(=O)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C6N=C(NC7=O)N)O)OP(=O)(O1)O)O
c-di-AMP		11158091	C1[C@@H]2[C@H]([C@H]([C@@H](O2)N3C=NC4=C(N=CN=C43)N)O)OP(=O)(OC[C@@H]5[C@H]([C@H]([C@@H](O5)N6C=NC7=C(N=CN=C76)N)O)OP(=O)(O1)O)O
CF501-aduvant		240872-312-4	NC(C1=CC2=C(N=C1)N(C/C=C/CN3C(NC(C4=CC(C)=NN4CC)=O)=NC5=C3C(OCCCN6CCOCC6)=CC(C(N)=O)=C5)C(NC(C7=CC(C)=NN7CC)=O)=N2)=O
ABZI		171378699	CCN1C(=CC(=N1)C)C(=O)NC2=NC3=C(N2C[C@@H](C)C4=CC=CC=C4)C=CC(=C3)C(=O)N
diABZI		168988084	CCN1C(=CC(=N1)C)C(=O)NC2=NC3=C(N2CCCCN4C5=C(C=C(C=C5)C(=O)C)N=C4NC(=O)C6=CC(=NN6CC)C)C=CC(=C3)C(=O)C
diABZI-3		131986624	CCN1C(=CC(=N1)C)C(=O)NC2=NC3=C(N2C/C=C/CN4C5=C(C=C(C=C5OCCCN6CCOCC6)C(=O)N)N=C4NC(=O)C7=CC(=NN7CC)C)C(=CC(=C3)C(=O)N)OC
diABZI-4		132000148	CCN1C(=CC(=N1)C)C(=O)NC2=NC3=C(N2C/C=C/CN4C5=C(C=C(C=C5OC CCN(C)C)C(=O)N)N=C4NC(=O)C6=CC(=NN6CC)C)C(=CC(=C3)C(=O)N)OC
DMXAA		123964	CC1=C(C2=C(C=C1)C(=O)C3=CC=CC(=C3O2)CC(=O)O)C
G10		702662-50-8 (CAS no.)	O=C1CSC2=C(C=C(C(NCC3=CC=CO3)=O)C=C2)N1CC4=C(Cl)C=CC=C4F
Glycyrrhetinic acid		10114	C[C@]12CC[C@](C[C@H]1C3=CC(=O)[C@@H]4[C@]5(CC[C@@H](C([C@@H]5CC[C@]4([C@@]3(CC2)C)C)(C)C)O)C)(C)C(=O)O
MSA-2		139547051	COC1=C(C=C2C=C(SC2=C1)C(=O)CCC(=O)O)CCCC3=C(C=C4C(=C3)C=C(S4)C(=O)CCC(=O)O)OC
SR-717		139434659	C1=CC(=NN=C1C(=O)NC2=CC(=C(C=C2C(=O)O)F)F)N3C=CN=C3

Table 1. Chemical structures of STING agonists discussed in the review with associated SMILES and PubChem ID or CAS No. n/a = not available.

STING agonists as antiviral agents

STING agonists have been tested against a multitude of virus families (Figure 2). Table 2 summarizes the different viruses and the STING agonist(s) used in the studies, organized by virus family. Table 3 covers each agonist, its viral target, and the proposed STING-dependent mechanism.

Figure 2. Viruses targeted by STING agonists.

Graphical summary of the viruses inhibited by STING agonists via the cGAS-STING pathway. Viral families are grouped by their Baltimore classification. *Denotes a non-human virus. HCoV = human coronavirus, MERS-CoV = Middle East respiratory syndrome-related coronavirus, SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2, PRRSV = porcine reproductive and respiratory syndrome virus, VEEV = Venezuelan equine encephalitis virus. Created in BioRender. Nelson, T. (2026) https://BioRender.com/u4a0drb.

Table 2.

Virus families targeted by STING agonists

	Virus Family	Virus	STING Agonist	Literature
DNA viruses	Hepadnaeviridae	Hepatitis B virus	2’3’-cGAMP DMXAA	[86-89]
	Orthoherpesviridae	Herpes simplex virus-1	2’3’-cGAMP cAIMP Compound 23 diABZI-4 DMXAA	[73,107,109,110,116]
		Herpes simplex virus-2	ADU-S100	[117]
		Pseudorabies virus	2’3’-cGAMP	[116]
		Varicella-zoster virus	2’3’- and 3’3’-cGAMP	[118]
		Cytomegalovirus	ADU-S100	[128]
		Murine cytomegalovirus*	DMXAA SPAC	[79,128]
	Poxviridae	Vaccinia virus	2’3’-cGAMP	[110]
RNA viruses	Arteriviridae	Porcine reproductive and respiratory syndrome virus*	SR-717	[154]
	Caliciviridae	Murine norovirus*	2’3’-cGAMP DMXAA	[160]
	Coronaviridae	HCoV-229E	2’3’-cGAMP Benzofuran derivatives diABZI-4	[75,164,165]
		HCoV-OC43	diABZI DMXAA	[73,167]
		Middle East respiratory syndrome-related virus (MERS)	Cyclic diguanylate monophosphate adjuvant	[199]
		SARS	CF501-adjuvant	[197]
		SARS-CoV-2	2’3’-cGAMP Benzofuran derivatives cAIMP CF501-adjuvant diABZI-1 diABZI-4 DMXAA Glycyrrhetinic acid SPAC	[73,75,78,79,164,165,187,192,193,197]
	Flaviviridae	Dengue virus	α-Mangostin BNBC	[74,214-218]
		Hepatitis C virus	α-Mangostin	[219]
		West Nile virus	2’3’-cGAMP cAIMP c-di-AMP diABZI	[192]
		Yellow fever virus	BNBC	[214]
		Zika virus	2’3’-cGAMP BNBC cAIMP c-di-AMP diABZI	[192]
	Orthomyxoviridae	Influenza A virus	2′3′-cGAMP diABZI DMXAA PS-cGAMP Adjuvant	[73,78,165,220,222]
		Influenza B virus	2′3′-cGAMP diABZI	[220]
		Avian influenza A virus (H7N9)	NanoSTING PS-cGAMP adjuvant	[221,222]
	Paramyxoviridae	Parainfluenza virus-3	ABZI diABZI diABZI-4	[164,165,225]
	Picornaviridae	Coxsackievirus	2’3’-cGAMP diABZI	[228]
		Enterovirus 68	cAIMP diABZI	[192]
		Enterovirus 71	2’3’-cGAMP	[110]
		Rhinovirus 16	2’3’-cGAMP ABZI diABZI diABZI-4	[225]
		Seneca virus A	MSA-2	[76]
	Pneumoviridae	Respiratory syncytial virus	cAIMP diABZI	[192,231]
	Rhabdoviridae	Vesicular stomatitis virus	2’3’-cGAMP	[110,116]
	Sedoreoviridae	Bovine rotavirus*	α-Mangostin	[238]
		Rhesus rotavirus*	Compounds 23/26/27	[109]
	Togaviridae	Chikungunya virus	2’3’-cGAMP 3’3’-cGAMP α-Mangostin C11 cAIMP diABZI DMXAA G10	[192,250-252]
		Sindbis virus	G10	[251]
		Mayaro virus	C11	[250]
		O’nyong’nyong virus	C11	[250]
		Ross River virus	C11	[250]
		Venezuelan equine encephalitis virus	C11 G10	[250,251]
Retroviruses	Retroviridae	Simian immunodeficiency virus*	2′3′-cGAMP c-di-AMP	[257]

Table 2: *denotes non-human virus

Table 3.

Evaluation of STING agonists against different viruses

STING Agonist	Virus Target	In vitro/ In vivo	Proposed STING-associated Mechanism	Literature
2’3’-cGAMP	Chikungunya virus	In vitro	Type I interferon	[192]
	Coxsackievirus	In vitro	Type I interferon	[228]
	HCoV-229E	In vitro	Type I interferon	[164]
	Hepatitis B virus	In vitro	Type I interferon	[87,89]
	Herpes simplex virus-1	In vitro	Type I interferon (Interferon-independent using Porcine STING)	[110,116]
	Influenza A virus	In vitro	Type I interferon	[220]
	Influenza B virus	In vitro	Type I interferon	[220]
	Murine norovirus*	In vitro	Type I interferon	[160]
	Pseudorabies	In vitro	Interferon-independent using Porcine STING	[116]
	Rhinovirus 16	In vitro	STING-dependent autophagy	[225]
	Simian immunodeficiency virus*	In vitro	n/d	[257]
	Vaccinia virus	In vitro	n/d	[110]
	Varicella-zoster virus	In vitro	Type I interferon	[118]
	Vesicular stomatitis virus	In vitro	Interferon-independent using Porcine STING	[116]
	West Nile virus	In vitro	n/d	[192]
	Zika virus	In vitro	n/d	[192]
3’3’-cGAMP	Chikungunya virus	In vitro	Type I interferon	[192]
	Varicella-zoster virus	In vitro	Type I interferon	[118]
α-Mangostin	Bovine rotavirus*	In vitro	NF-κB signaling	[238]
	Chikungunya virus	In vitro In vivo	n/d	[252,253]
	Dengue virus	In vitro	n/d
	Hepatitis C virus	In vitro	n/d	[219]
ADU-S100	Cytomegalovirus	In vitro	n/d	[128]
	Herpes simplex virus-2	In vitro In vivo	Type I interferon	[117]
Benzofuran derivatives	HCoV-229E	In vitro	IRF3-signaling	[75]
	SARS-CoV-2	In vitro	Type I interferon	[75]
BNBC	Dengue virus	In vitro	Type I interferon	[214]
	Yellow fever virus	In vitro	Type I interferon	[214]
	Zika virus	In vitro	Type I interferon	[214]
C11	Chikungunya virus	In vitro	IFNAR-dependent	[250]
	Mayaro virus	In vitro	IFNAR-dependent	[250]
	O’nyong’nyong virus	In vitro	IFNAR-dependent	[250]
	Ross River virus	In vitro	IFNAR-dependent	[250]
	Venezuelan equine encephalitis virus	In vitro	IFNAR-dependent	[250]
cAIMP	Chikungunya virus	In vitro	Type I interferon	[192]
	Respiratory syncytial virus	In vitro	n/d (ineffective)	[192]
	SARS-CoV-2	In vitro	n/d	[192]
	West Nile virus	In vitro	n/d	[192]
	Zika virus	In vitro	n/d	[192]
Cyclic diguanylate monophosphate adjuvant	Middle East respiratory syndrome-related virus	In vivo	Neutralizing antibodies	[199]
	Simian immunodeficiency virus*	In vitro	n/d	[257]
	West Nile virus	In vitro	n/d	[192]
	Zika virus	In vitro	n/d	[192]
CF501-adjuvant	SARS	In vivo	Neutralizing antibodies	[197]
	SARS-CoV-2	In vivo	Neutralizing antibodies	[197]
Compounds 23/26/27	Herpes simplex virus-1	In vitro	Type I interferon	[109]
	Rhesus rotavirus*	In vitro	n/d	[109]
diABZI-derivatives	Chikungunya virus	In vitro	Type I interferon	[192]
	Coxsackievirus	In vitro	Type I interferon	[228]
	Enterovirus 68	In vitro	Type I interferon	[192]
	HCoV-229E	In vitro	Type I interferon	[164,165]
	HCoV-OC43	In vitro Ex vivo	IRF3-dependent signaling Type I interferon (ex vivo)	[73,167]
	Herpes simplex virus-1	In vitro In vivo	Type I interferon	[73]
	Influenza A virus	In vitro In vivo	Type I interferon	[164,165,225]
	Influenza B virus	In vitro	Type I interferon	[220]
	Parainfluenza virus-3	In vitro	TBK1-dependent type I interferon (cell type-dependent)	[165,225]
	Respiratory syncytial virus	In vitro	n/d (ineffective)	[192]
	Rhinovirus 16	In vitro	STING-dependent autophagy	[225]
	SARS-CoV-2	In vitro In vivo	Type I interferon	[78,186,192]
	West Nile virus	In vitro	n/d	[192]
	Zika virus	In vitro	n/d	[192]
DMXAA	Chikungunya virus	In vivo	Type I interferon	[251]
	Murine cytomegalovirus	In vivo	Type I interferon	[79]
	HCoV-OC43	In vitro	IRF3-dependent signaling	[167]
	Hepatitis B	In vitro In vivo	Type I interferon	[86-89]
	Herpes simplex virus-1	In vitro In vivo	Type I interferon	[107]
	Influenza A virus	In vitro	Type I interferon	[220]
	Murine norovirus*	In vitro	Type I interferon	[160]
	SARS-CoV-2	In vivo	Type I interferon	[79]
G10	Chikungunya virus	In vitro	IRF3- and STAT1-dependent signaling	[251]
	Sindbis virus	In vitro	IRF3- and STAT1-dependent signaling	[251]
	Venezuelan equine encephalitis virus	In vitro	IRF3- and STAT1-dependent signaling	[251]
Glycyrrhetinic acid	SARS-CoV-2	In vitro In vivo	Type I interferon	[193]
MMAE	Herpes simplex virus-1	In vitro In vivo	n/d	[110]
	Pseudorabies virus	In vitro	n/d	[110]
	Vaccinia virus	In vitro	n/d	[110]
MSA-2	Seneca virus A	In vitro	Type I interferon	[76]
NanoSTING	Avian influenza A virus (H7N9)	In vivo	n/d	[221]
	SARS-CoV-2	In vivo	Type I interferon	[78]
	Respiratory syncytial virus	In vivo	n/d	[231]
PS-cGAMP adjuvant	Influenza A virus	In vivo	Neutralizing antibodies	[222]
	Avian influenza A virus (H7N9)	In vivo	Neutralizing antibodies	[222]
SR-717	Porcine reproductive and respiratory syndrome virus*	In vitro	Type I interferon	[154]
STING pathway-activating complexes (SPAC)	Murine cytomegalovirus	In vivo	Type I interferon	[79]
	SARS-CoV-2	In vivo	Type I interferon	[79]

Table 3: *denotes non-human virus. n/d = STING-dependency was not discussed

DNA viruses

Hepadnaviridae

The hepatitis B virus (HBV) is a causative agent of chronic liver disease worldwide, with a global burden of over 350 million people living with chronic HBV infection. HBV infection of hepatocytes fails to induce a traditional PRR-mediated immune response, which has been attributed to the inability of HBV to activate the PRRs [^80-82] or its inhibition of those pathways [⁸³]. To target cytosolic DNA-sensing pathways, the HBV polymerase protein was found to bind STING and attenuate its K63-linked polyubiquitination and function, suppressing STING signaling [⁸⁴]. Priming of an effective HBV-specific antiviral response by Kupffer cells and intrahepatic dendritic cells is important for controlling HBV infection [⁸⁵]. In the first study testing STING agonists against HBV, direct treatment of immortalized murine hepatocytes with conditioned media from STING agonists 2’3’-cGAMP- or DMXAA-treated murine macrophages was able to inhibit HBV DNA replication through JAK-dependent type I interferon response [^86,87]. However, direct treatment of immortalized murine hepatocytes with murine-specific DMXAA did not prevent HBV DNA replication. In the same study, NOD/SCID mice treated with DMXAA prior to HBV infection exhibited decreased HBV DNA and increased expression of ISGs [⁸⁶]. In a follow-up study, HBV replication did not impact cGAS and STING functionality in reconstituted immortalized murine hepatocytes, even though HBV polymerase was shown to inhibit STING signaling [⁷⁰]. In a chronic HBV rcccDNA transgenic mouse model, STING activation by DMXAA not only downregulated HBsAg and HBV DNA through type I interferon and NF-κB signaling in Kupffer cells but also reduced HBV-induced liver fibrosis [⁸⁸]. In addition to indirect anti-HBV cytokine secretion in primary human hepatocytes and nonparenchymal liver cells, 2’3’-cGAMP was able to trigger hepatic T-cell activation regardless of HBV genotype [⁸⁹].

Orthoherpesviridae

The Orthoherpesviridae family, previously known as the Herpesviridae family, encompasses a wide variety of large double-stranded DNA human viruses and is divided into three subfamilies based on their biological properties: alphaherpesvirinae, betaherpesvirinae, and gammaherpesvirinae [^90,91]. The two most known alphaherpesviruses are herpes simplex virus 1 (HSV-1) and herpes simplex virus 2 (HSV-2), etiological agents of oral and genital herpes, respectively. HSV enters the host through the mucosal epithelia, skin, or cornea and subsequently spreads to neurons to establish latency. HSV DNA can be detected by multiple PRRs, including the cGAS-STING pathway, leading to the induction of type I, II, and III interferon [^92-95].

Alphaherpesvirinae

HSV-1 encodes multiple proteins to target all components of the cGAS-STING pathway and facilitates its immune evasion. VP22, UL37, and UL41 interact with cGAS to inhibit its enzymatic activity or mediate its degradation [^96-98]. UL38, VP1-2, ICP27, UL36USP, and γ₁34.5 independently associate with STING and disrupt or prevent its interaction with TBK1 [^99-103]. HSV-1 microRNA miR-24 binds the 3’ untranslated region of STING mRNA and inhibits its translation [¹⁰⁴]. UL46 and VP24 impair the interaction between TBK1-IRF, impairing IRF3 activation and downstream signaling [^105,106]. Despite HSV-1 using multiple mechanisms to block this pathway, both CDN and non-CDN STING agonists have been explored as antivirals against HSV-1.

Mouse STING agonist DMXAA inhibited HSV-1 infection in fibroblasts isolated from wild-type C57BL/6J mice through interferon β production, while high viral titers were detected in mutant STING or IRF3-deficient mice [¹⁰⁷]. Since HSV-1 infection can result in viral encephalitis and herpes simplex keratitis, the leading cause of infectious blindness, reduction of spread and viral load to the central nervous system is important [¹⁰⁸]. Following a corneal infection, C57BL/6J mice pretreated with DMXAA had a reduction in viral load compared to vehicle control in the cornea, trigeminal ganglia, and brainstem, although this protective phenotype disappeared by five days postinfection in the trigeminal ganglia [¹⁰⁷]. Upon evaluation of neurological disease and mortality, DMXAA-treated mice regained their original weight and trended lower neurological disease scores by 9 and 21 days postinfection, respectively [¹⁰⁷]. These outcomes were only seen in pre-treatment conditions, and the in vivo mechanism was not discussed.

One of the more potent and promising STING agonists, diABZI, was tested in a mouse model of herpes simplex encephalitis (HSE). Without agonist treatment, cGAS- and STING-deficient mice exhibit weight loss, hydrocephalus, and decreased survival after HSV-1 corneal infection [⁷³]. A single dose of diABZI-4, the oral form of the drug, via retro-orbital resulted in protection from HSE in the cGAS-deficient mice, reflected in reduced HSV-1 brain tissue titers, weight loss, lethality, and hydrocephalus [⁷³]. Although serum was not isolated from the mice following HSV-1 infection to determine the mechanism, intraperitoneal injection of diABZI-4 led to interferon β production. Another study reported the antiviral effects of their novel STING agonist compound 23, a dimerized molecule based on the structure of C-170, a STING inhibitor with strong binding. Fibrosarcoma cells (HT1080) treated with compound 23 prior to HSV-1 infection had lower levels of HSV-1 protein ICP0 and ICP4 18 hours postinfection [¹⁰⁹]. Compound 23 was shown to induce interferon β and TNF-α production in mice, but poor solubility and toxicity were limiting factors [¹⁰⁹].

A third study reported that the treatment of monocytes (THP-1) and mouse fibroblast (L929) cells by 2’3’-cGAMP led to decreased HSV-1 propagation due to interferon β production and ISG expression [¹¹⁰]. This antiviral effect was further boosted by a microtubule-destabilizing agent, monomethyl auristatin E (MMAE), which altered virus trafficking within the cell. It had previously been reported that microtubule-destabilizing agents could enhance STING agonist antitumor activity in vivo, which led the authors to test these synergetic effects in an antiviral context [¹¹¹]. Combined MMAE and STING agonist treatment also inhibited HSV-1 viral propagation and prolonged mouse survival following systemic infection [¹¹⁰]. This phenotype was also seen against another alphaherpesvirus, pseudorabies virus (PRS), in the aforementioned cell lines [¹¹⁰]. PRS, like HSV-1, also encodes multiple proteins to inhibit STING-mediated antiviral signaling (UL13, UL28, US2), suggesting the potential of targeting this pathway for a therapeutic approach [^112-115].

HSV-1 has multiple proteins designed to restrict an interferon-dependent antiviral response, suggesting the need for interferon-independent mechanisms. Porcine STING can exert an interferon-independent autophagy and antiviral response through apoptosis [¹¹⁶]. In HEK293T cells, stimulation of an exogenous expression of a porcine STING by its natural agonist 2’3’-cGAMP exhibited anti-HSV-1 and anti-PRS effects in HEK293T cells, independent of interferon and autophagy [¹¹⁶]. Through multiple antiviral mechanisms and broad-spectrum strategies, STING agonists have shown potential as antivirals for HSV-1.

HSV-2 is the leading cause of genital herpes and attempts for an anti-HSV-2 vaccine have failed. To discover new HSV-2 antivirals, one study reported the ability of two STING agonists to protect from HSV-2 infection in keratinocytes, the main cells involved in HSV-2 clinical lesions, and viral infection mouse models [¹¹⁷]. In human keratinocytes (HaCaT), multiple STING agonists were able to induce phosphorylation of STAT1 and the production of ISGs viperin and ISG15, showcasing their ability to respond to STING agonists. When pretreated with ADU-S100 (2’3’-cGAM(PS)₂), HaCaT cells showed reduced levels of viral protein 5 following HSV-2 infection. Furthermore, conditioned media from ADU-S100-stimulated monocytes were able to protect wild-type but not interferon-α/β receptor (IFNAR) knockout human keratinocytes from HSV-2 infection, indicating protection is dependent on type I interferons [¹¹⁷]. In the same study, three different treatment regimens were explored in mouse studies (pre-treatment, post-infection, and two-times post-infection). When delivered intraperitoneally to C57BL/6J wild-type and cGAS-deficient mice prior to lethal HSV-2 challenge, ADU-S100 protected both mouse groups under the pre-treatment conditions and displayed decreased TCID₅₀ in vaginal wash fluid [¹¹⁷]. Only in wild-type mice did the post-treatment groups improve survival.

To determine if direct administration to epithelial surfaces could provide protection through a local immune response, topical administration of 3’3’-cAIMP, another STING agonist with a similar response to ADU-S100 and higher solubility, was applied to the vaginal mucosal surface of WT and cGAS-deficient mice 6 hours and 72 hours prior to HSV-2 infection. Both treatments led to increased interferon production and protection against HSV-2 [¹¹⁷].

Varicella-zoster virus (VZV) is the etiological agent of chickenpox and shingles, the latter caused by reactivation of the virus from latency in an infected host. Knockdown of STING prior to VZV infection led to increased viral protein expression, suggesting a link between VZV and a STING-mediated antiviral response [¹¹⁸]. There has been one report of STING agonists (2′3′- and 3′3′-cGAMPs) inhibition of VZV in vitro [¹¹⁸]. Pre-treatment of human dermal fibroblasts, a common cell type for studying VZV, with CDNs led to decreased viral titers and replication through a type I interferon response [¹¹⁸]. In the same study, without using STING agonists, it was shown that STING-mediated type III interferon production is also an important factor for controlling VZV infection.

Betaherpesvirinae

There are four known betaherpesviruses associated with disease: human cytomegalovirus (HCMV), human herpesviruses 6A and 6B, and human herpesvirus 7. Acute infection of HCMV is typically subclinical in healthy individuals but can cause life-threatening disease in the immunocompromised, elderly, and children [¹¹⁹]. While HCMV vaccines are in clinical trials, current treatments with ganciclovir and valganciclovir are encountering drug resistance, suggesting the need for new antivirals [¹²⁰].

Like many viruses, HCMV and murine CMV (MCMV) encode multiple proteins able to downregulate the endogenous cGAS-STING-mediated type I interferon response. HCMV UL31 interacts with cGAS to dissociate DNA from it and modification of viral pp71 prevents it from inhibiting STING-dependent antiviral signaling [^121,122]. HCMV UL82 and IE63 inhibit STING translocation and mediate STING degradation, respectively [^123,124]. Interestingly, murine CMV m152 inhibits STING-dependent IRF signaling but not NF-κB signaling [¹²⁵]. To overcome this inhibition, one approach is to deliver to an exogenous STING pathway-activating complex (SPAC), which is composed of membrane vesicles (MV) displaying multimeric STING on their surface (STING-MV) bound to an agonist [⁷⁹]. Two prophylactic intravenous treatments of DMXAA or SPAC (STING-MV and -DMXAA) to mice decreased MCMV load in the liver, corresponding to the decreased inflammatory cell infiltration of the liver and lungs and increased type I interferon and ISGs [⁷⁹]. When administered 24 hours after MCMV infection, SPAC-treated mice had significantly reduced the viral load and inflammatory damage to the liver and lungs compared to the control groups [⁷⁹].

In monocyte-derived dendritic cells, cGAS-STING recognition of HCMV leads to type I interferon expression through IRFs [¹²⁶]. STING can also activate NF-κB signaling, which facilitates HCMV immediate early gene expression [¹²⁷]. Pre-treatment with STING agonist ADU-S100 decreased the infection of dendritic cells, but concurrent treatment at the time of infection led to an increase in HCMV-positive cells, suggesting more studies are needed to understand STING’s antiviral role against HCMV [¹²⁸].

Gammaherpesvirinae

The gammaherpesvirus subfamily is characterized by its establishment of latency in lymphoid cells and tumorigenic abilities. Epstein-Barr virus (EBV) infects about 95% of global adults and has been associated with a wide variety of malignant diseases, including infectious mononucleosis, Burkitt’s lymphoma, nasopharyngeal carcinoma, and post-transplant lymphoproliferative disease [¹²⁹]. EBV targets both STING and TBK1 through its deubiquitinase BPLF1 [¹³⁰]. Kaposi’s sarcoma-associated herpesvirus (KSHV) is the etiological agent of Kaposi’s sarcoma, primary effusion lymphoma, and multicentric Castleman’s disease [^131,132]. It was previously established that KSHV can be detected by multiple PRRs, and specifically the cGAS-STING pathway is essential for controlling KSHV infection in vitro [^38,133-136]. Further studies determined that KSHV targets both cGAS and STING to prevent the induction of type I interferon. KSHV proteins vIRF1 block the interaction of STING with TBK1 while ORF52 and a truncated form of LANA inhibit cGAS activity [^38,137,138]. Recently, multiple KSHV microRNAs were shown to target the STING to facilitate viral lytic replication [¹³⁹].

Currently, there are no published reports on STING agonists use as EBV or KSHV antivirals. Activating STING’s dual anticancer and antiviral properties with agonists could serve as a new therapeutic approach to oncogenic viruses like EBV and KSHV.

Poxviridae

Vaccinia virus (VACV) is widely known for its use in the live-virus smallpox vaccine and its infection is very mild [¹⁴⁰]. Multiple strains of VACV have been shown to trigger the cGAS-STING pathway and inhibit downstream IRF3 activation [³⁹]. There has been only one study exploring the interplay of STING agonists against VACV. Supernatants from human monocytes (THP-1) stimulated with 2’3’-cGAMP and/or MMAE were able to protect Vero cells from VACV infection, but the mechanism was not further investigated [¹¹⁰].

RNA viruses

Arteriviridae

Porcine reproductive and respiratory syndrome virus (PRRSV) is a nonzoonotic, positive-sense single-stranded RNA virus that causes high mortality and morbidity rates in pigs, negatively impacting the farming industry [^141-143]. The virus has extensive antigen variation and continuously produces new strains, rendering current commercial inactivated and attenuated vaccines unable to provide long-lasting and effective protection [^144-146]. PRRSV infection is characterized by suppression of innate immune response in the early phase of infection, leading to poor lymphocyte proliferation and delayed neutralizing antibody responses [¹⁴⁷]. The virus has been shown to both stimulate and inhibit various TLRs [^148,149]. Due to the poor humoral response in pigs to PRRSV, there is a growing interest in targeting the innate immune system in pigs, as seen in the efficiency of TLR agonists as antivirals [^150,151].

There are currently two known PRRSV proteins that inhibit STING function, nsp2 and nsp5, by preventing their translocation to the endoplasmic reticulum [^152,153]. There has been one study examining the antiviral effects of STING agonists on PRSSV. In Marc-145 cells and porcine alveolar macrophages, both highly permissive to PRRSV, STING agonist SR-717, which is currently entering clinical trial, reduced the titers and gene and protein expression in a dose-dependent manner against multiple PRRSV strains [¹⁵⁴]. Further experiments revealed that SR-717 induced high levels of interferon β, IL-1β, IL-6, and TNF-α [¹⁵⁴]. It was determined that the agonist had no effect on virus adsorption and entry, but inhibited genome replication, assembly, and release [¹⁵⁴].

Caliciviridae

Noroviruses are the leading cause of acute gastroenteritis worldwide [^155,156]. Murine norovirus (MNV) is commonly used to study the immune mechanisms of norovirus infection and can be detected by RIG-I and MDA-5, leading to the production of type I and II interferons in a STAT1-dependent manner [^157-159]. STING has been reported to interact with RIG-I [^12,41], and a recent study explored the role of STING in MNV infection using STING agonists. The treatment of two different mouse macrophage cell lines, known to support MNV replication, with DMXAA or 2’3’-cGAMP led to decreased MNV RNA and nonstructural proteins across three different MNV strains through interferon and ISG induction [¹⁶⁰].

Coronaviridae

Coronaviruses are large enveloped, positive-sense single-stranded RNA viruses that have caused large scale epidemics over the past few decades, including the most recent outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in 2020. The Orthocoronavirinae subfamily is divided into four genera, with the alpha- and betacoronaviruses infecting mammals and containing seven human pathogenic strains. Recent research on antivirals against human coronaviruses (HCoVs), specifically SARS-CoV-2, and the emergence of crosstalk between RNA- and DNA-sensing pathways have encouraged researchers to explore the cGAS-STING pathway. To study STING agonists in HCoV, there is a robust set of air-liquid interface (ALI) lung cell culture models used across the studies discussed below, each designed to mimic certain areas of the respiratory tract [summarized in ¹⁶¹].

Alphacoronaviruses

HCoV-229E is a prototypic human strain often responsible for the common cold [¹⁶²]. It has been shown to antagonize the RLR signaling pathway, inhibiting the interferon β and NF-κB responses [¹⁶³]. While HCoV-229E has not been found to disrupt STING signaling, a related alphacoronavirus, HCoV-NL63, disrupts STING dimerization and STING-TBK1 interaction, inhibiting STING-dependent type I interferon production [⁴⁸].

The first study reporting on STING agonists against HCoV-229E found diABZI protected human fetal fibroblast cells (MRC-5), which are highly susceptible to HCoV infection, from cytopathic effects at an EC₅₀ of 3nM, a lower concentration than the current anti-HCoV drug Remdesivir [¹⁶⁴]. Interestingly, the combination treatment of diABZI and Remdesivir had a protective effect with no cytotoxicity. Differentiated small human airway epithelial cells, cultured at an ALI to stimulate a more physiological environment, treated with diABZI had a 70% reduction in viral titer and rescue of cytopathic effects in the epithelium [¹⁶⁴]. In both cell lines, protection from prophylactic diABZI treatment was dependent on TBK1-IRF3-mediated type I interferon signaling [¹⁶⁴]. Additionally, 2’3’-cGAMP exhibited antiviral effects against HCoV-229E in both cell lines, but to a lesser extent than diABZI. Another group also reported that diABZI-4 inhibited HCoV-229E in MRC-5 cells without compromising cell viability through interferon β and ISGs [¹⁶⁵]. Furthermore, a new class of STING agonists, benzofuran derivatives, was able to reduce HCoV-229E RNA accumulation in MRC-5 cells through IRF3 signaling [⁷⁵].

Betacoronaviruses

HCoV-OC43 is another HCoV associated with the common cold. Previously, STING was shown to inhibit the formation of double membrane vesicles by HCoV-OC43, which is essential for coronavirus replication [¹⁶⁶]. Mouse-specific agonist DMXAA could block HCoV-OC43 infection in human alveolar basal epithelial A549 cells, a targeted cell type by the virus, in an IRF3-dependent manner [¹⁶⁷]. Since DMXAA does not activate human STING, a human STING mutant highly responsive to DMXAA was used in this study. Treatment with diABZI-3 inhibited infection in A549 cells and cultured ex vivo human lung explants, also in an IRF3-dependent manner, reflected in the lack of HCoV-OC43 N protein and presence of phospo-IRF3 [¹⁶⁷]. This protective effect was extended to another betacoronavirus, SARS-CoV-2 (USA-WA1/2020 strain), which is discussed in more depth below. A second study confirmed diABZI antiviral efficacy against HCoV-OC43 in ACE2-expressing A549 cells through type I interferon [⁷³].

The novel SARS-CoV-2, also known as COVID-19, has caused over 7 million deaths to date and serious post-infection complications [^168-173]. Since the outbreak, numerous vaccines have been developed and distributed globally, but the rise of variants has hindered efforts to develop effective antivirals [^174-183]. SARS-CoV-2 proteins ORF3a, ORF9b, 3CL, and ORF10 antagonize RLR- and STING-mediated type I and III interferon responses [^46,184,185].

One of the most physiologically relevant culture systems for studying respiratory pathogens is the bronchial human airway epithelial cells reconstituted from bronchial biopsies in an ALI system. Treatment with diABZI-4 to the basolateral chamber prior to challenge with SARS-CoV-2 (Beta CoV/France/IDF0571/2020 strain) decreased viral RNA and maintained the epithelial layer integrity upon apical, although the dependency on interferon was not explored in this system [¹⁶⁴]. Another system, the 3D stem cell–derived induced alveolar type II (iAT2) cell cultures, recapitulated this phenotype against the USA-WA1/2020 strain when pre-treated with low doses of diABZI-4 [⁷³]. Prophylactic diABZI-4 treatment of primary nasal and bronchial epithelial tissues from two donors led to the secretion of interferon β, IL-6, and IP10 and protected from SARS-CoV-2 infection [¹⁶⁵].

The K18-hACE2 transgenic mouse is a common model for studying antivirals against SARS-CoV-2 [¹⁸⁶]. Intranasal administration, but not intraperitoneal injection, of diABZI led to decreased viral protein, RNA, and genome copy in lung tissue after 48 hours, correlating with decreased inflammatory gene expression [⁷³]. Another study found decreased viral replication and type I interferon-dependent cytokine and chemokine expression 4- and 6-days post-infection in the same mouse model against both the USA WA1/2020 strain and South African variant B.1.35 [¹⁸⁷].

SARS-CoV-2 is responsible for multiple cardiovascular complications [^188-191]. One study reported that cAIMP, an analog to 3’3’-cGAMP, and diABZI were able to protect against the USA-WA1/2020 strain in human pluripotent stem cell-derived cardiomyocytes [¹⁹²].

In addition to the common CDN and non-CDN agonists tested above, scientists have explored new types of STING agonists against SARS-CoV-2. Benzofuran derivatives, BZF-2OH and BZF-5H, were shown to prevent SARS-CoV-2 replication in human bronchial epithelial (BEAS-2B) and lung adenocarcinoma (Calu-3) cells in a type I interferon-dependent manner [⁷⁵]. Glycyrrhetinic acid, a major component of a popular traditional Chinese medicine and a non-traditional STING agonist, was determined to inhibit SARS-CoV-2 Omicron variant in Calu-3 cells with no impact on cell viability and decreased viral load in transgenic K18-ACE2 mice challenged with SARS-CoV-2 Omicron variant through type I interferon and ISG production [¹⁹³]. The novel delivery method for 2’3’-cGAMP, NanoSTING, was demonstrated to protect against the Delta variant B.1.617 in a Syrian golden hamster model [⁷⁸]. NanoSTING-treated animals had 300-fold lower viral titers in the lungs and 1000-fold lower titers in the nasal compartment, showcasing efficient delivery of 2’3’-cGAMP across the mucosa [⁷⁸]. Although not confirmed in the hamster model, NanoSTING delivery of 2’3’-cGAMP to BALB/c mice confirmed secretion of interferon β in the nasal compartment.

To study the rescue effect of STING agonists against SARS-CoV-2, the exogenous STING pathway-activating complex (SPAC), which is composed of membrane vesicles displaying multimeric STING on their surface (STING-MV) bound to an agonist (2’3’-cGAMP or DMXAA), was delivered to 80-week-old BABL/c mice [⁷⁹]. Even in older mice, lung pathology was significantly reduced as reflected in decreased levels of viral RNA load and inflammatory cytokines, parallel to an increase in IFNB1, MX1, and ISG15 transcript levels [⁷⁹].

STING agonists could also serve as potential vaccine adjuvants against SARS-CoV-2, since they are known to elicit vaccine-induced CD8+ T-cells [^194-196]. A new STING agonist, CF501, was used as an adjuvant to the SARS-CoV-2 RBD-Fc dimeric protein and produced neutralizing antibodies in mice, rabbits, and rhesus macaques [¹⁹⁷]. This vaccine could produce cross-neutralizing antibodies against SARS-CoV, SARS-like bat coronavirus Rs3367 and WIV1, which is important due to the high risk of cross-species infection, and nine pseudotyped SARS-CoV-2 variants: Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), Epsilon (B.1.427), Zeta (P.2), Eta (B.1.525), Iota (B.1.526), and Kappa (B.1.617.1) [^197,198].

Middle East respiratory syndrome-related coronavirus (MERS-CoV) spreads through contact with infected camels, with limited human-to-human transmission. While there are no published studies regarding direct treatment of MERS-CoV by STING agonists, efforts to overcome vaccine challenges led to the use of STING agonists as adjuvants. Instead of the traditional Th2-dominant adjuvants, cyclic diguanylate monophosphate, a natural STING agonist produced by bacteria, was loaded into hollow polymeric nanoparticles designed to generate an antigenic response against MERS-CoV spike antigen [¹⁹⁹]. Vaccination with the viromimetic nanoparticle produced a higher neutralizing titer compared to the control and no detection of infectious viral load following lethal MERS-CoV challenge, coinciding with antigen-specific CD8+ T-cell responses [¹⁹⁹].

Flaviviridae

The Flaviviridae family consists of enveloped positive-sense singled-stranded RNA viruses primarily spread through arthropod vectors such as ticks and mosquitoes [²⁰⁰]. This family is home to many human pathogenic viruses with high global disease burdens: dengue virus, hepatitis C virus, Japanese encephalitis virus, tick-borne encephalitis virus, West Nile virus, yellow fever virus, and Zika virus [^201-209]. Although there is no DNA stage in the flavivirus replication cycle, there are multiple reports of cGAS-STING activation and attenuation by flaviviruses, suggesting its role in controlling viral infection. The conserved flavivirus NS2B-NS3 protease targets cGAS and human STING for degradation [^{31,44,45,47,210-212}].

Antivirals for flavivirus infection can be divided into virus-targeting drugs (direct-acting antivirals) and host-directed antivirals, with STING agonists falling into the latter category [²¹³]. In a human dermal fibroblast model, 2’3’-cGAMP, cAIMP, c-di-AMP, and diABZI reduced the infection of West Nile virus and Zika virus [¹⁹²]. Separately, the STING agonist 6-bromo-N-(naphthalen-1-yl)benzo[d][1,3]dioxole-5-carboxamide (BNBC) exhibited pan-flavivirus protection against dengue virus-2, yellow fever virus, and Zika virus, also through a type I interferon response [²¹⁴].

The natural xanthone compound α-Mangostin, known to activate STING, has been tested for its antiviral and anti-inflammatory properties against multiple flaviviruses [⁷⁴]. For dengue virus, high viral titer and severity of infection correlate with excessive cytokine production. α-Mangostin pre-treatment in two human hepatoma cells and human-derived immature dendritic cells inhibited viral infection and subsequent viral production against all four dengue serotypes (DENV-1 strain Hawaii, DENV-2 strain 16681, DENV-3 strain H87, and DENV-4 strain H241), reflected in decreased production of IL-6, TNF-α, and IP-10 transcription [^215-218]. Additionally, micromolar treatment of α-Mangostin in human hepatoma cells expressing a hepatitis C replicon inhibited viral replication, although the STING-related mechanism is not explored [²¹⁹].

Orthomyxoviridae

New influenza A virus (IAV) variants are most commonly responsible for pandemics, most notably the H1N1 strain (swine flu). IAV was shown to interact with STING through its hemagglutinin fusion peptide, hindering interferon production in a cGAS-independent, STING-dependent manner [⁴³].

STING agonists 2’3’-cGAMP, diABZI, and DMXAA were able to inhibit two influenza A strains (human H1N1 and PR8, the mouse-adapted H1N1) in monocytes and macrophages [²²⁰]. In the same study, the impact of 2’3’-cGAMP and diABZI extended to multiple influenza B strains (Yamagata and Victoria) and was attributed to increased ISG15, IFNβ, and OAS1 transcript levels [²²⁰]. In a more physiologically relevant environment, diABZI-4 inhibited IAV infection in human airway epithelial cells differentiated at ALI, reflected in elevated type I interferons and elevated interferon-dependent cytokines [^73,165,220]. Treatment at the time of infection and post-infection was also able to decrease secreted IAV in human lung fibroblast cells [¹⁶⁵].

Regarding animal studies, diABZI-4 in K18-ACE2 transgenic mice, which express the human ACE2 receptor, alongside the NanoSTING agonist protected Syrian golden hamsters against lethal IAV challenge [^73,78]. NanoSTING can induce interferon-dependent and -independent responses, and it is unclear which mechanisms are involved for IAV in vivo. It was also reported that 2’3’-cGAMP can enhance the humoral, cellular, and mucosal immune response as a mucosal adjuvant against avian influenza A (H7N9) [²²¹]. 2’3’-cGAMP was designed as a vaccine adjuvant for the inactivated H1N1 vaccine, encapsulated by pulmonary surfactant (PS)-biomimetic liposomes to test against different strains of influenza. The PS-cGAMP adjuvant was able to provide protection against distant H1N1 and heterosubtypic H3N2, H5N1, and H7N9 strains for at least 6 months in mice and ferrets [²²²].

Paramyxoviridae

The Paramyxoviridae family includes Nipah virus, parainfluenza virus, measles virus, and mumps virus, which are associated with upper and lower respiratory tract infections, measles, and mumps, respectively. These RNA viruses have been shown to trigger the cGAS-STING pathway during infection [²²³]. While there is no published study on direct inhibition of cGAS or STING by paramyxoviruses, they encode viral proteins that attenuate multiple downstream components of the pathway [²²⁴].

Parainfluenza virus 3 (PIV3) is one of four subtypes that cause severe respiratory disease in young children and infants. There have been two studies focused on STING agonists against PIV3, with diABZI at the center based on its potent activity and high solubility. Pre-treatment with diABZI in HEp2 cells, which support PIV3 propagation, inhibited PIV infection and replication through TBK1-dependent type I interferon signaling [²²⁵]. In a separate study, diABZI could also inhibit PIV3 infection when added to lung fibroblast cells parallel to infection, although the mechanism is not discussed [¹⁶⁵]. This phenotype was lost when tested in PIV3-infected Vero E6 cells, indicating cell-type specific mechanisms may play a role [¹⁶⁴].

Picornaviridae

Picornaviruses cover a large range of small RNA viruses responsible for diseases such as polio, the common cold, heart disease, and hepatitis A. Vaccines and treatments are effective for certain viruses in this family, but the large tropism leaves room for more antiviral discovery. The cGAS-STING pathway can detect picornavirus infection through mitochondrial DNA sensing and is actively attenuated by the virus to facilitate its immune evasion [^226-228].

For multiple picornaviruses, there have been only one or two published studies exploring the antiviral effects of STING agonists. Enterovirus-D68, which is associated with myocarditis, could be inhibited by cAIMP and diABZI in human pluripotent stem cell-derived cardiomyocytes [^192,229]. Since EV-D68 responded to direct IFN-β treatment in the same experiment, it is suggested that the STING agonist impact is mediated through type I interferon signaling. Enterovirus-A71 had reduced infection of Vero cells, which support viral replication, when pre-treated with supernatant from 2’3’-cGAMP-stimulated donor THP-1 cells [¹¹⁰]. Coxsackievirus infection decreased in HeLa cells when pretreated with 2’3’-cGAMP or diABZI in an interferon-dependent manner [²²⁸]. Seneca virus A, known for causing vesicular disease in pigs, was highly responsive to the oral STING agonist MSA-2 in porcine cell lines through type I interferon signaling [⁷⁶].

Rhinovirus 16 is one of multiple viruses associated with the common cold and pneumonia [^165,225]. Treatment with 2’3’-cGAMP, ABZI (the monomer of diABZI), and diABZI protected HeLa cells from rhinovirus-induced cytopathic effects through a STING-dependent autophagy mechanism, although 2’3’-cGAMP required digitonin to enter the cell and exhibit antiviral activity [²²⁵]. These results were replicated in human primary bronchial epithelial cells and patient-derived nasal tissue, which represent more physiologically relevant models [^165,225].

Pneuomoviridae

Respiratory syncytial virus (RSV), also called human respiratory syncytial virus and human orthopneumovirus, causes respiratory infections, which can be severe in children and elderly populations [²³⁰]. Currently, there is a maternal RSV vaccine and three RSV vaccines for adults aged 60 and older. To further improve vaccine efficacy, NanoSTING (2’3’-cGAMP-encapsulated nanoparticles) was tested for its ability as a vaccine adjuvant by priming mucosal immunity. In both mice and cotton rats, intranasal delivery of NanoSTING and RSV prefusion protein generated neutralizing antibodies against RSV and protected the animals from virus challenge [²³¹].

Interestingly, when considering the traditional route of treatment with STING agonists, diABZI and c-AIMP could not inhibit RSV infection in human lung epithelial cells, but direct treatment with interferon β could [¹⁹²].

Rhabdoviridae

The Rhabdoviridae family consists of notable pathogens such as rabies virus and vesicular stomatitis virus (VSV). Although VSV is an RNA virus, studies have reported that mice deficient in STING are highly susceptible to VSV infection, and VSV replicated is increased in the absence of cGAS [^11,232]. VSV inhibits interferon production by modulating the alternative splicing of STING [²³³].

Two previous studies mentioned earlier in the review (see alphaherpesvirinae) also explored the antiviral effects of STING agonists against VSV. The treatment of monocytes (THP-1) and mouse fibroblast (L929) cells by 2’3’-cGAMP led to decreased VSV propagation, and this antiviral effect was further boosted by the microtubule-destabilizing agent MMAE [¹¹⁰]. Stimulation of an exogenous expression of porcine STING by its natural agonist 2’3’-cGAMP inhibited VSV infection and replication in HEK293T cells, indicated by decreased green fluorescent protein expression and plaque-forming unit counts of the reporter virus [¹¹⁶].

Sedoreoviridae

Rotavirus infections are highly contagious, non-enveloped double-stranded RNA viruses and the leading cause of severe gastroenteritis in children under 5 years of age [²³⁴]. While there are no antivirals for rotavirus infection and treatment focuses on managing symptoms, over one hundred countries have rotavirus vaccines, although the effectiveness and impact vary per country [²³⁵]. While rotavirus is recognized by RIG-I and MDA-5, it has been reported that STING could inhibit rotavirus infection [^236,237]. There have only been two studies of STING agonist-driven inhibition of rotavirus infection. In an African green monkey kidney epithelial cell line (MA104), commonly used for rotavirus studies, pre-treatment with novel STING agonist compounds 23, 26, or 27 led to decreased rhesus rotavirus NSP5 protein expression following infection through an unexplored mechanism [¹⁰⁹]. During a natural product screen for antivirals, α-Mangostin, known to activate STING, was shown to inhibit bovine rotavirus infections and induce NF-κB [²³⁸]. While the study does not discuss the mechanism, α-Mangostin is a known activator of STING [⁷⁴].

Togaviridae

Viruses belonging to this family are commonly arboviruses, transmitted via mosquito vectors and maintained in vertebrate reservoirs. Human diseases associated with these viruses include infectious arthritis, encephalitis, rashes, and fever [^239-244]. There have been reports of alphaviruses activating the cGAS-STING pathway upon infection and encoding proteins to modulate its function [^245-248]. Chikungunya virus (CHIKV) proteins nsP4, nsP2, and the capsid target cGAS for degradation, while nsP1 interacts with STING to inhibit its signaling. Downstream, nsP2 inhibits STAT1-dependent production of ISGs [^248,249].

As with flaviviruses and coronaviruses, STING agonists are tested for their pan-antiviral capability. STING agonist C11 was shown to inhibit viral titers of CHIKV, Mayaro virus, O’nyong’nyong virus, Ross River virus, and Venezuelan equine encephalitis virus (VEEV) in telomerase-transduced foreskin fibroblasts, a commonly used tool to study alphaviruses, in an IFNAR-dependent manner [²⁵⁰]. Using a similar drug screening process, another study reported that the novel STING agonist G10 could inhibit CHIKV, Sindbis virus, and VEEV in an IRF3- and STAT1-dependent fashion [²⁵¹].

Focusing only on CHIKV, a panel of STING agonists (2′3′-cGAMP, 3′3′-cGAMP, cAIMP, and diABZI) inhibits CHIKV replication in human fibroblasts through TBK1-IRF3 signaling, with the latter two molecules also able to inhibit infection when given post-infection. Since the cardiovascular system has been implicated in systemic CHIKV infection, cAIMP and diABZI were tested in cardiomyocytes and shown to decrease infection. Direct treatment with interferon β yielded similar results, suggesting the agonists were operating through type I interferon signaling [¹⁹²]. In a separate study, intraperitoneal delivery of the mouse-specific STING agonist DMXAA was shown to inhibit CHIKV viremia in mice when given prophylactically, although titers were decreased if administered after infection, but not significantly compared to the control [²⁵¹]. A second in vivo study showed that the natural compound α-Mangostin had antiviral activity against CHIKV, reflected in decreased viral load and inflammation in muscle tissue [^252,253]. Derivatives of α-Mangostin were discovered and shown to reduce CHIKV titers in Vero CCL-81 cells, but the mechanism was not discussed [²⁵³].

Retroviruses

Retroviridae

Retroviruses are a large and diverse family of RNA viruses that integrate into a host’s genome. Known retroviruses include human T-lymphotropic virus and human immunodeficiency virus type 1 and 2 (HIV-1 and HIV-2, respectively), all of which can cause severe disease in humans [^254,255]. HIV’s reverse-transcribed DNA can be recognized by cGAS and trigger STING activation [^52,256]. While there have not been any reports of STING agonists as HIV antivirals, 2’3’-cGAMP and c-di-AMP were tested against reactivated simian immunodeficiency virus in latently infected monkey and human peripheral blood mononuclear cells and found to enhance antigen-specific CD8+ T-cell responses [²⁵⁷].

Discussion

Since the discovery of STING, there has been a steady rise in the development of synthetic STING agonists. With the most recent SARS-CoV-2 pandemic, the use of STING agonists as antivirals against a variety of virus families has garnered major research interest. As summarized in this review, there have been many studies reporting on the ability of STING agonists to inhibit viral replication against fifteen different viral families, including both RNA and DNA viruses. Stimulation of cGAS-STING before, during, or after infection with STING agonists helped prime the immune response and overcome the initial pathway suppression by virally encoded proteins in a virus-specific context. The prophylactic or anaphylactic use of agonists may vary for each virus, and more research is needed to determine when dosing is best to achieve maximum efficiency.

As mentioned previously, the cGAS-STING pathway can be triggered by other pathogens, such as bacteria, parasites, and fungi [^258-261]. While bacterial CDNs trigger STING expression, the downstream secretion of interferons and cytokines can impact bacterial infection in a context-dependent manner, either harming or helping the host. This idea can be extended to parasitic functions, where activation of STING and increased interferon production can help eliminate infection, but elevated interferon is associated with more severe disease, depending on the bacteria species involved. Interestingly, STING negatively regulates the antifungal immune response [^262,263]. More studies will need to be conducted to understand the impact of STING on other infectious diseases and the therapeutic potential of STING agonists as antimicrobials.

The current research on STING agonist development has been heavily focused on cancer immunotherapies. Intrinsically, induction of type I interferons and ISGs by STING positively correlates with T-cell infiltration and cross-priming of dendritic cells in the tumor environment [^264-266]. STING activation can drive ferroptosis, autophagy, and senescence in cancer cells and is important for radiation-induced immune responses [^267-270]. Eleven STING agonists have entered clinical trials for various cancers, and dozens of others are in preclinical development for cancer therapeutics [⁵⁹]. Unfortunately, STING agonists for antiviral purposes have not reached clinical trials, and continuous efforts are needed to build a strong preclinical foundation. This is partially due to a lack of animal studies for each virus using STING agonists. Interestingly, the effectiveness of STING agonists as both anticancer and antiviral agents could provide a new approach for targeting oncogenic viruses (e.g., EBV, KSHV, HBV [^{131,132,271,272}]), as current studies have focused on either viral infection or oncogenesis but not together in the context of STING. STING agonists were shown to inhibit HBV viral replication and potentially HCV, but little was discussed regarding viral oncogenesis. Of the seven known oncogenic viruses, only HBV and human papillomavirus have effective vaccines [^273,274]. In the case of viral cancers, STING agonists could be utilized by 1) preventing initial infection of the virus and establishment of chronic infection or latency, 2) inducing cancer cell death, and 3) preventing the spread of the virus within the host.

STING agonists, particularly CDNs, are being considered for vaccine adjuvants as they generate mucosal immunity and can protect against a variety of respiratory bacterial and viral infections [^{196,197,221,275-281}]. CDNs produce balanced memory Th1, Th2, Th17, and CD8+ T-cells along with DC cross-presentation to produce a cytotoxic T-lymphocyte response, all of which are desirable for vaccine-induced immunity [^282-284].

Overall, STING agonists have shown potential as novel antiviral agents and vaccine adjuvants, contributing to our growing toolkit to combat infectious viral diseases.

Current limitations and future directions

While the field of STING agonists as antivirals and cancer therapeutics continues to grow, there are substantial hurdles to be addressed. In this review, we detailed studies across 15 different viral families, each with varying experimental conditions, models, and use of STING agonists. In many studies, DMXAA, the mouse-specific STING agonist, was used and showed effective protection against certain viruses, but these results were not repeated with a human STING agonist. Furthermore, the low solubility of certain STING agonists, such as DMXAA and 2’3’-cGAMP, requires transfection or digitonin to enter the cell. Effective delivery and high solubility are important when developing antiviral treatments. It would be beneficial for more studies to test the STING agonists currently in clinical trials, outside of ADU-S100 and SR-717, and the developed oral STING agonists such as diABZI-4 and MSA-2. Lastly, the treatment regimen compared to the time of infection is extremely important. Many studies only consider pre-treatment dosing and require further expansion to post-infection treatment. It is also not understood mechanistically how STING agonists are able to overcome the inhibition of the cGAS-STING pathway by the virally encoded antagonists or why agonist efficacy varies between viruses. All these factors impact the translational capability of STING agonists as antivirals.

Additionally, continuous or robust stimulation of STING is a concern for inflammation and autoimmunity. In the absence of foreign DNA, the cGAS-STING pathway can be activated by self-DNA, through defective clearance by the three prime repair exonuclease (TREX1) or leakage of self-DNA (mitochondrial DNA stress or nuclear DNA) [^285-287]. STING-driven inflammation caused by defective DNA sensing or STING function has been associated with multiple autoinflammatory syndromes such as STING-associated vasculopathy with onset in infancy, Aicardi–Goutières syndrome, Familial chilblain lupus, and Copa syndrome [^288-291]. STING has been associated with systemic lupus erythematosus by its regulation of NLRP3 levels and subsequent increase in IL-1β, which contributes to the onset and/or progression of the disease [²⁹²]. When considering STING’s multifaceted roles in our antiviral, anticancer, and inflammatory responses, the pros and cons of overactivation of STING must be weighed carefully.

Abbreviations

2’3’-cGAMP

2'-3'-cyclic guanosine monophosphate-adenosine monophosphate (GMP-AMP)

AIM2

absent in melanoma 2

ALI

air-liquid interphase

ALR

AIM2-like receptor

ASC

apoptosis-associated speck-like protein containing a caspase recruitment domain

CDN

cyclic dinucleotide

cGAS

cyclic GMP-AMP synthase

CHIKV

chikungunya virus

CXCL

C-X-C motif chemokine ligand

DAMPs

damage-associated molecular patterns

DENV

dengue virus

diABZI

diamidobenzimidazole

DMXAA

dimethylxanthenone acetic acid

HBV

hepatitis B virus

HCMV

human cytomegalovirus

HCOV

human coronavirus

HIV

human immunodeficiency virus

HNTV

Hantaan virus

HSE

herpes simplex encephalitis

HSV

herpes simplex virus

IAV

influenza A virus

IFNAR

interferon-α/β receptor

IKK

IkappaB kinase

IL

interleukin

IRF

interferon regulatory factor

ISG

interferon-stimulated genes

JAK

Janus kinase

KSHV

Kaposi’s sarcoma-associated herpesvirus

MAVS

mitochondrial antiviral signaling protein

MCMV

murine cytomegalovirus

MDA5

melanoma differentiation-associated gene 5

MERS-CoV

Middle East respiratory syndrome-related coronavirus

MMAE

monomethyl auristatin E

MNV

murine norovirus

MV

membrane vesicles

NF-κB

nuclear factor kappa B

NLR

nod-like receptors

NLRP3

nod-like receptor family pyrin domain-containing 3

PAMPs

pathogen-associated molecular patterns

PIV

parainfluenza virus

PRR

pattern recognition receptor

PRRSV

porcine reproductive and respiratory syndrome virus

PRS

pseudorabies virus

PS

pulmonary surfactant

RIG-I

retinoic acid-inducible gene I

SARS-CoV-2

severe acute respiratory syndrome coronavirus 2

SPAC

STING pathway-activating complex

SIV

simian immunodeficiency virus

STAT

signal transducer and activator of transcription proteins

STING

stimulator of interferon genes

TBK1

TANK-binding kinase 1

TLR

toll-like receptor

TNF-α

tumor necrosis factor-α

TREX1

three prime repair exonuclease 1

VACV

vaccinia virus

VEEV

Venezuelan equine encephalitis virus

VSV

vesicular stomatitis virus

VZV

varicella-zoster virus