Originally, it was thought that this protein could only be activated by cyclic GMP-AMP (cGAMP), or similar cyclic dinucleotide molecules. Now, a recent study from the University of Texas Southwestern has uncovered a previously hidden binding site on the STING protein. On one hand, the discovery of this “cryptic pocket” may be a useful target for enhancing activation. One the other hand, previous studies have also found that when some molecules bind to STING, they inhibit its activation. Considering that heightened STING activation seems to underlie some autoimmune disorders, knowing when to activate or inhibit this protein may be the key to controlling the robust inflammatory consequences of Covid-19. Here, we will discuss the discovery of the new binding site on STING as a promising therapeutic target for activating the immune system.

Before we begin, we must consider the mechanisms involved in STING-mediated immunity beginning with the detention of DNA by an enzyme called cGAMP synthase (cGAS). Upon binding to DNA, cGAS triggers a reaction that produces the second messenger cyclic GMP-AMP (cGAMP). cGAMP then binds to STING, located in the endoplasmic reticulum around the cell’s nucleus. This triggers profound changes in the STING protein that allow it to move around the cell and mount an immune defense against the invading pathogen. The cGAMP- bound STING protein then travels to the golgi apparatus, where it recruits a protein called TANK-binding kinase 1 (TBK1) and interferon regulatory factor 3 (IRF3). TBK1 phosphorylates, or “activates”, both STING and the interferon regulatory factor. Now in its final active form, STING mediates the removal of the pathogen through autophagy, while the activated interferon regulatory factor enters the nucleus to stimulate increased expression of inflammatory genes to fight the infection, including a group of cytokines known as type 1 interferons. Type 1 interferons have a dual effect that not only fights infection in the cell they were generated in but also protects nearby healthy cells. The entirety of this pathway is illustrated below.

As a transmembrane protein, STING has two main domains: a transmembrane domain and cytoplasmic ligand-binding domain. Although cGAMP and other cyclic dinucleotides bind strongly to the center of the butterfly-shaped ligand-binding domain, the challenge with treating Covid-19 with cyclic dinucleotides is that these molecules are not only involved in immune regulation. They have multiple functions throughout the body. cGMP, for example, also plays an important role in detecting light in the visual system, forming long-term memories, and the process of contracting and relaxing muscles. Administering high amounts of these molecules can significantly disrupt several of these systems.

Activating STING causes the entire structure of the protein to shift and facilitates the formation of a larger structure called an oligomer. During oligomerization, multiple individual STING proteins come together in a line to form side-by-side connections. Forming an oligomer not only allows STING to travel around the cell but this step is also important for recruiting the TBK1 enzyme needed for phosphorylation.

Despite their high affinity for STING, however, Lu et al. were surprised to find that cGAMP alone could not activate STING in a test tube. Little to no oligomers were formed and there was no evidence of phosphorylation. Since cGAMP is known to STING, researchers speculated whether adding another agonist could promote activation.

Fortunately, a previous discovery published in the European Journal of Medicinal Chemistry revealed that non-cyclic dinucleotides can also activate the immune properties of STING. This study identified a group of small compounds from the benzothiazinone family that bind to STING with varying degrees. Pryde et al. found that one molecule in particular, compound 53 (C53) shown in the figure below, robustly and consistently activated STING. However, considering how chemically and structurally different C53 is from cGAMP, where this molecule was binding to on the STING protein was a mystery. This is when researchers initially began speculating whether there may be an additional binding location somewhere else on the protein.

Using the STING agonists identified from the previous study, Lu et al. found that adding both cGAMP and C53 consistently generated high amounts of the STING oligomer. However, this effect could not be replicated when either of these compounds were mixed STING alone. Using cryo-electron microscopy, a technique that freezes proteins in place, Lu et al. confirmed that both cGAMP and C53 could simultaneously bind to the STING oligomers. Shown in the figure below, cGAMP sat in the ligand-binding domain, while C53 fit within a deep pocket within the transmembrane domain.

What role does C53 have in activating STING? To answer this question, Lu et al. further examined the structure of the STING oligomer. What they found was that these oligomers consisted of two smaller STING proteins, stabilized by interactions between the two ligand binding sites, as well as the two transmembrane domains. While the ligand-binding connections were rather weak, interactions between the transmembrane domains seemed to provide most of the structure’s stability. In fact, when investigators genetically mutated regions of the protein that facilitate interactions across transmembrane domains, the number of STING oligomers generated significantly reduced despite the fact that both C53 and cGAMP were bound to the protein. Lu et al. therefore speculated that when C53 binds to STING, it significantly changes the transmembrane domain that promotes stronger interactions. This likely explains why cGAMP alone did not produce stable STING oligomers, but adding C53 did.

Previously, it was unclear whether the transmembrane domain even had the ability to bind agonists. Now it seems like this “cryptic pocket” may be critically important for mediating innate immunity. When researchers introduced mutations in this binding pocket, for instance, the number of STING oligomers dramatically reduced, even when exposed to both C53 and cGAMP. These mutations also prevented the phosphorylation of STING later on, suggesting that this binding pocket within the transmembrane domain may mediate the effectiveness of STING after it has been activated by cGAMP.

The discovery of this hidden binding site may provide a new target for drugs and vaccines against SARS-CoV-2. Compounds, like C53, that can bind to the transmembrane domain in fact may be better therapeutic agonists for activating STING than those mimicking cGAMP. However, the specifics of how STING is innately activated in the body, not to mention how different molecules activate or inhibit its activity, remain unclear. Researchers speculate that cells have an unknown molecule similar to C53 that binds to STING’s transmembrane domain to activate the protein as needed. Identifying this molecule and understanding its mechanism of action may provide better insight into how we can capitalize on innate immunity to fight SAR-CoV-2. Next, in this series, we will take a closer look at the structure of C53 to understand which components may be the most important for STING activation.