It has also been shown that Beclin-1, an autophagy protein, negatively regulates cGAS-mediated innate immune responses (Liang et al. 2014). On one hand, Beclin-1 directly interacts with the NTase domain of cGAS in a DNA binding dependent manner through its CCD domain, which impairs the enzymatic activity of cGAS, resulting in decreased cGAMP synthesis and subsequent impairment of type Ⅰ interferon induction by DNA virus infection. On the other hand, knockdown of cGAS decreases the formation of autophagosome upon DNA stimulation, which suggests that cGAS may be involved in the formation of autophagosome. By this way, beclin-1 enhances autophagy-mediated degradation of cytosolic pathogen DNA to prevent excessive cGAS activation. These findings showed that autophagy adopts two strategies, elimination of pathogen DNA or inhibition of cGAS enzymatic activity, to negatively regulate DNA sensing pathway to avoid excessive activation of immune response.
Interestingly, a recent study by Chen et al. (2016a) provides another piece of evidence for negative regulation of DNA sensing pathway by autophagy. They found that E3 ligase TRIM14 could stabilize cGAS by inhibiting the autophagic degradation of cGAS. The K48-linked ubiquitination of cGAS at K414 is a signal for p62, a cargo receptor of selective autophagy. Without virus infection, cGAS undergoes K48-linked ubiquitination at K414, leading to p62-dependent selective autophagic degradation of cGAS. Upon DNA virus infection, TRIM14, as an interferon-stimulated gene, is induced and recruits the deubiquitinating enzyme USP14 to cleave K48-linked ubiquitin chains of cGAS, thus inhibiting p62-cGAS interaction as well as the degradation of cGAS. These studies reveal a positive feedback regulation of cGASmediated signaling by TRIM14 and provide insights into the crosstalk between autophagy and type Ⅰ IFN signaling in innate immunity.
Viral infection triggers the production of type Ⅰ interferons and activation of inflammasomes. There are many examples of crosstalks between type Ⅰ interferons production signaling and inflammasome activation to balance these two processes (Guarda et al. 2011; Zhang et al. 2014). Wang et al. (2017c) found that cGAS is inhibited by caspases, essential components of inflammasome. Upon binding to DNA, AIM2 forms a complex with ASC and caspase-1, leading to the activation of caspase-1 (Fernandes-Alnemri et al. 2009; Hornung et al. 2009). Activated caspase-1 directly binds to and cleaves cGAS at D140/157, resulting in reduced cGAMP production and type Ⅰ IFN induction. Importantly, caspase-1 only cuts cGAS upon DNA virus infection but not RNA virus infection. Consequently, AIM2-, ASC-, and caspase-1-deficient mice display enhanced resistance to DNA virus infection. In addition, Caspase-4, 5, and 11 can cut cGAS in conditions of non-canonical inflammasome activation (Wang et al. 2017c). Their findings reveal a role for inflammasome in the regulation of cGAS-mediated induction of type Ⅰ IFNs and provided insights into the crosstalk between DNA virus induced innate immune responses and inflammasome activation.
Viruses have evolved elaborate mechanisms to antagonize the innate immune system. Plenty of studies have demonstrated that MITA, an essential adaptor downstream cGAS, is targeted by various viruses for immune evasion (Ding et al. 2013; Kalamvoki and Roizman 2014; Ma et al. 2015b; Wu et al. 2015). Recent studies supply several lines of evidence that cGAS is also antagonized by virus for immune evasion.
It is well established that Kaposi's sarcoma-associated herpesvirus (KSHV) LANA localizes to the nucleus of infected cells. However, an isoform of KSHV LANA lacking the N terminal, generated by internal translation initiation, is localized to the cytoplasm. This cytoplasmic isoform of KSHV LANA was demonstrated to recruit and antagonize cGAS (Zhang et al. 2016). Overexpression of this isoform binds to cGAS and reduces the phosphorylation of TBK1 and IRF3, resulting in impaired production of type Ⅰ IFNs and antiviral response. However, the precise mechanism has not been fully elucidated yet.
In addition to cytoplasmic isoform of LANA, ORF52, a tegument protein of KSHV, inhibits DNA sensing of cGAS (Wu et al. 2015). The enzymatic activity of cGAS is inhibited in the presence of ORF52 upon DNA stimulation or DNA virus infection. ORF52 can bind to both DNA and cGAS, and these abilities are important for cGAS inhibition. Intriguingly, its binding to cGAS is not dependent on its binding to DNA. ORF52 is a conserved protein in gammaherpes-viruses and its homologs in MHV68, RRV and EBV can also inhibit cGAS, suggesting that this is a general mechanism for immune evasion by gammaherpesviruses.
Dengue virus (DENV) is a single positive-stranded RNA virus of the family Flaviviridae, without DNA production in their life cycle. Nevertheless, it is well established that cGAS plays important roles in innate immune responses against DENV infection. It is proposed that cGAS detects mitochondrial DNA released during DENV infection (Aguirre et al. 2017; Ma and Damania 2016; West et al. 2015). A recent study showed that cGAS is degraded during DENV infection (Aguirre et al. 2017). DENV encodes a protease cofactor NS2B that promotes cGAS degradation in an autophagy–lysosome-dependent manner to avoid the detection of mitochondrial DNA, which results in the inhibition of type Ⅰ interferon production during DENV infection. It is worthy to note that DENV can also evade the innate immune response through cleavage of MITA by its NS2B3 protease (Aguirre et al. 2012; Yu et al. 2012). Therefore, DENV targets both cGAS and MITA to minimize the host innate immune antiviral response.
Since cGAS acts as an important cytosolic DNA sensor that leads to activation of innate immune antiviral responses, the activity of cGAS needs to be tightly modulated spatially and temporally by host factors to ensure the strength and duration of the immune responses. On the other hand, viruses have evolved elaborate mechanisms to antagonize the innate immune system. The cGAS-mediated signaling is targeted at every levels of signaling cascades by viruses. This review summarized the regulatory mechanisms of cGAS. The information presented in this review is summarized in Fig. 1 and Table 1. So far, our understanding of cGAS regulation is relatively limited. PQBP1 was reported to function as a coreceptor of cGAS in the sensing of retroviruses (Yoh et al. 2015), and studies also showed that IFI16 and cGAS work cooperatively during DNA virus infection in human foreskin fibroblasts (HFFs) and human keratinocytes (Almine et al. 2017; Orzalli et al. 2015). These studies indicate that the relationship between cGAS and other cytosolic DNA sensors needs to be further clarified, and also prompt us that there may be other coreceptor(s) or coactivator(s) of cGAS.
Figure 1. Schematic regulating network of cGAS. Detailed regulating proteins and virus antagonist of cGAS are depicted. Upon DNA binding to cGAS, cGAMP is synthesized by cGAS from ATP and GTP. Then binding of cGAMP to the adapter MITA induces conformational changes of MITA homodimer, leading to production of cytokines including type Ⅰ interferons. Green arrows indicate positive regulators of the corresponding process. Red arrows indicate negative regulators of the corresponding process. Upon inflammasome activation, caspase-1 inhibits the cGAS-MITA pathway by cleavage of cGAS. DENV NS2B protein promotes the autophagy–lysosome-dependent degradation of cGAS. TRIM14 and USP14 inhibit the degradation of cGAS by cleaving K48-linked ubiquitin chains of cGAS. Beclin-1 promotes the autophagic degradation of cGAS. CCP5/6, TRIM38, SENP7 and RNF185 promote the activity of cGAS by deglutamylation, SUMOylation, deSUMOylation, ubiquitination of cGAS, respectively. TTLL4, TTLL6 and Akt inhibit the activation of cGAS by monoglutamylation, polyglutamylation, phosphorylation of cGAS, respectively. LANA of KSHV interacts and antagonizes cGAS. ORF52 of KSHV interacts with cGAS and inhibits the enzymatic activity of cGAS
Table 1. Important regulation sites of cGAS
cGAS is tightly regulated by post-translational modifications, such as phosphorylation, ubiquitination, SUMOylation, and glutamylation. However, there still are some questions to be answered. Firstly, it is demonstrated that cGAS undergoes ubiquitin-dependent degradation, but the E3 ubiquitin ligases responsible for this process is still unknown. Secondly, the roles of dynamic SUMOylation of cGAS need further clarification. Thirdly, whether other PTMs, such as acetylation, glycosylation, which are also responsible for the regulation of cGAS are worthy to explore. Most importantly, how these PTMs work cooperatively to modulate the initiation, duration, and termination of cGAS-mediated signaling is pivotal. Systemic study on these dynamic processes of PTMs in vivo during virus infection will provide more exciting and convincing insights into cGAS-mediated signaling.
Finally, the roles of cGAS in autoimmune diseases and antitumor immunity have drawn more and more attention, but the regulation mechanisms of cGAS in these processes are still largely unknown. Further investigation of these questions will certainly shed new lights on our understanding of cGAS-mediated DNA sensing pathways and help to develop strategies to prevent and treat the related diseases.