Glen N. Barber obtained his Ph.D in molecular virology (working on Hemorrhagic viruses) in 1989 from Porton Down and the London School of Hygiene and Tropical Medicine (LSHTM), London University. He did his post-doctoral training at the University of Washington, Seattle, USA and briefly worked on developing vaccines against human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) infection. Most of his post-doctoral work, however, focused on studying mechanisms of host defense against viral infection.
Since the mid-1950’s it was known that cells responded to virus infection by secreting potent antiviral cytokines referred to as the interferons (IFN’s). However, it was unclear how viruses caused the cell to produce IFN or how they exerted their antiviral function. Over the next 50 years, it became clear that after secretion, the IFN’s bound to cellular receptors of predominantly uninfected cells and activated the transcription of dozens of interferon-inducible anti-microbial proteins through the JAK/STAT signaling pathway. The function of many of these proteins still remains to be determined. Work at the University of Washington focused on examining the importance of an interferon-inducible protein now referred to as PKR (for protein kinase RNA-dependent). This involved demonstrating that after infection of the cell, PKR was activated by viral dsRNA species and exerted anti-viral activity by inhibiting viral translation through phosphorylation of the alpha subunit of eIF2. The lab subsequently confirmed the importance of PKR in protecting the host against infection by certain RNA viruses using animal models.
After visiting the University of Tokyo, Barber became an Assistant Professor at Emory University, Atlanta in 1996 and continued his research on host defense. The laboratory was involved with studying other dsRNA binding proteins, similar to PKR, which were called the NFARs (Nuclear Factors Associated with dsRNA). The NFARs comprised two proteins of 90 kDA and 110kDa that are produced from the same gene by alternative splicing. Research indicated that the NFARs are involved with escorting mRNAs out from the nucleus for translation. After the mRNA has been delivered to the ribosome, the NFARs normally shuttle back into the nucleus. However, if a viral RNA hijacks a ribosome, then PKR becomes activated and phosphorylates the NFARs, as well as eIF2α, an event which prevented them from re-shuttling back into the nucleus. The NFAR’s remained bound to the viral RNA, to additionally ensure that it is not translated. Indeed, controlling translation after virus infection constitutes an important component of host defense, especially since many viruses, in fact most RNA viruses, replicate exclusively in the cytoplasm of the cell.
In 1999, Barber joined the University of Miami School of Medicine as Associate Professor, becoming Professor in 2004 and Chair of the Department of Cell Biology in 2011. A major question remaining in the field that had not been answered in over half a century after their discovery was how were the IFN’s actually induced by viruses or bacteria? What was known was that it was the nucleic acid (genomes) of the viruses that were potently able to trigger IFN production, thus implying a role for cellular RNA or DNA binding receptor-like proteins in triggering this event. In the late 1990’s and early 2000’s a number of laboratories started to prove the importance of the Toll-Like Receptor (TLR) family in recognizing microbial infection and stimulating innate immune gene transcription, including IFN. Indeed, the TLR pathway constitutes an important component of host defense against viral and bacterial infection.
However, it became apparent that even in the absence of the TLR pathway, dsRNA and DNA viral species remained capable of potently triggering the production of IFN, thus suggesting the likelihood of alternate receptors and signaling pathways. At Miami the lab noted that dsRNA-mediated IFN production was ablated in cells lacking a pro-apoptotic protein referred to as FADD. However, loss of FADD or binding partner RIP1 did not appear to affect the TLR pathway. It was concluded that TLR-independent dsRNA-triggered pathways indeed existed to switch on IFN induction following RNA virus infection that involved FADD/RIP1. In the same year, the Fujita group showed that the cellular RNA binding helicases RIG-I/MDA5 were important for recognizing viral RNAs and for activating host defense gene induction, including the production of IFN. RIG-I/MDA5 downstream signaling events required FADD/RIP1 to facilitate efficient RIG-I/MDA5 signaling. The importance of FADD and RIP1 has also been shown to be significant in drosophila, where these proteins control TLR-independent innate immune gene induction in response to bacteria infection (referred to as the imd pathway). The TLR and imd pathways constitute two major ways in which insects protect themselves from microbial invasion. The TLR (TLR3, TLR7) and RIG-I/MDA5/LGP2 pathway are now known to be essential for triggering innate immune gene transcription in response to RNA virus infection.
Despite this progress, little was known relating to how IFN was produced in response to DNA pathogens (viruses, bacteria, parasites), or exogenous dsDNA species in general. In 2006, the laboratory embarked on a project to further isolate proteins that could influence the production of IFN. One cellular protein was noted to potently be able to activate IFN production. The protein was named STING (for stimulator of interferon genes). The lab went on to show that STING controlled a new cellular signaling pathway that activates innate immune gene transcription, including IFN, in response to the presence of cytosolic DNA species (the STING signaling pathway).
Generally, the cytosol is supposed to be devoid of dsDNA species. If dsDNA species are present in the cytoplasm, then it likely comes from an invading microbe, or from the nucleus of a damaged cell, or from damaged mitochondria. In these circumstances, STING triggers cytokine production, likely to attract the immune system to the region. The lab demonstrated that transient STING activity is essential for cytokine production in response to infection by DNA pathogens such as herpes simplex virus 1 (HSV1) or the bacteria Listeria.
It was apparent from the literature that patients suffering from autoimmune diseases, such as systemic lupus erythematosus (SLE) commonly produced antibodies to their own DNA (anti-nuclear antibodies- ANA). Thus, it was postulated that DNA activated innate immune pathways may play a role in manifesting such disorders. Indeed, the etiology of autoimmune/autoinflammatory diseases was largely unknown and was even speculated to involve chronic microbial infection, which constantly aggravated the immune system to produce cytokines.
A next area of focus therefore involved evaluating the role of the STING pathway in influencing inflammation. In 2001, the Nagata group demonstrated that mice lacking the exonuclease DNaseII died during embryonic development. This was due to DNaseII-deficient macrophages being unable to digest the DNA from engulfed apoptotic cells. DNA from the dead cells leaked out from the lysosomal compartments of the macrophages to produce pro-inflammatory cytokines which killed the embryos. The responsible self-DNA triggered innate immune signaling pathway proved evasive for over a decade, but was shown not to involve the TLR or RIG-/MDA5 pathways. To evaluate the role of STING in this inflammatory pathway, our lab crossed DNaseII+/- mice with STING-/- mice and found that the progeny (DNaseII-/- STING-/-) were viable, thus proving that the STING pathway was responsible for the lethal cytokine production from the macrophages.
This was the first demonstration that STING pathway was involved in the triggering of self-DNA mediated autoinflammatory disease. Of course, the severity of DNaseII deficiency means that loss of DNaseII function does not commonly occur in humans. However, cases of patients with DNaseIII (Trex1) mutations do exist and such individuals suffer from severe autoinflammatory disease and typically die shortly after birth (Aicardi-Goutieres Syndrome AGS). Mice with defects in Trex1 also succumb to lethal inflammation related disease after 8 weeks of birth. To next evaluate the role of STING signaling in Trex1 manifested disease, Trex1-/- mice were crossed with STING-/- mice and our lab again showed as a collaboration that such mice were viable.
Our data subsequently indicated that after cell division, residual DNA species can leak out of the nucleus into the cytoplasm where it is normally degraded by Trex1. In hematopoietic cells, defects in Trex1 enabled the leaked DNA to trigger STING-signaling and lethal cytokine production. The importance of hematopoietic cells in this disease was shown by transplanting normal bone marrow into irradiated Trex1-deficient mice which rescued lethality. Aside from providing new mechanistic insight into self-DNA manifested diseases, this information provided a possible treatment (transplantation) for some of these disorders as well as opened up the notion to screen for drugs that could possibly suppress STING signaling and thus treat certain inflammatory disorders.
These new observations raised the possibility that STING-signaling could conceivably be involved in a wide variety of alternate inflammatory malaise including inflammatory bowel disease (IBD) Indeed, it has now been reported that inflammation can be caused by mutations in the STING gene itself, which causes the molecule to be permanently active (STING-associated vasculopathy with onset in infancy- SAVI). Further instances of deregulated STING signaling are now being reported to be associated with inflammation.
It is now well known that inflammation can enhance the development of tumors. As an extension of the above work, our laboratory next showed that STING signaling was involved in fueling the transformation process. For example, it was observed that DNA damaging agents can induce chromatin re-arrangement and leakage into the cytosol, where it activates STING signaling and cytokine production. Data now suggests that STING signaling plays a key role in the DNA-damage response pathway and functions to alert the immune system to the damaged cell. Significantly, the lab found that STING-deficient mice were resistant to carcinogen triggered skin cancer, since the leaked DNA could not activate STING-dependent cytokine production. As a consequence, macrophages were not recruited to the damaged area an event that also eliminated additional sources of pro-inflammatory cytokines and growth factors. In summary, our laboratories work indicated that while transient STING signaling is essential for protection of the host against microbial infection and for initiation tissue repair responses, chronic STING signaling may play a key role driving certain types of inflammation driven cancers.
Previously, we had noted that transient STING activity was essential for stimulating anti-viral adaptive immunity. Through a collaboration, it was shown that STING was essential for generating anti-tumor T cell responses. Essentially, the DNA from dying tumor cells plays an important role in triggering STING signaling in engulfing macrophages. STING-dependent cytokines are essential for the efficient cross presentation of tumor antigen and the generation of anti-tumor T cell responses. STING-deficient mice do not efficiently generate anti-tumor T cell responses.
During this period, the actual mechanisms of STING signaling started to become better understood. Two reports from the Vance and Chen labs showed that cytosolic DNA species bind to a cellular synthase referred to as cGAS, which generates cyclic dinucleotides (CDN’s; such as GMP-GMP, GMP-AMP) that bind to STING. This induces an autophagy related-signaling process that leads to cytokine production. These findings have spawned an industry aimed at screening for, and using STING agonists as anti-tumor therapies due to their abilities to promote anti-tumor T cell responses. Preliminary studies have also shown that the stimulation of STING signaling using agonists can overcome resistant to checkpoint inhibitors. Finally, recent data from our lab indicates for the first time that a large percentage of colon tumors have lost STING signaling, likely to avoid the activation of DNA-damage induced cytokine production and the attention of our immunosurveillance system.
In summary, the discovery of the STING pathway shed considerable light into the mechanisms of host responses to microbial infection. These findings also provided key mechanistic insight into causes of inflammatory disease. Finally, STING signaling appears suppressed in many cancers likely since this pathway plays an important role in facilitating anti-tumor T cell responses.
Finally, during studies at Emory University, the lab noted that normal cells (for example, murine embryonic fibroblasts derived from the embryos of mice) were quite refractory to virus infection (such as vesicular stomatitis virus, VSV) compared to most tissue cultured cells such as HeLa and HEK 293 cells. It was postulated that tissue cultured cells which were transformed, may have acquired defects in their intrinsic innate immune pathways, which render them susceptible to infection. The lab characterized many types of cancer cells and found that almost all exhibited an impaired capacity to produce type I interferon. These observations lead to pioneering the use of viruses such as VSV as therapeutics for the treatment of cancer. This data enabled the initiation of Phase I trials using recombinant VSV as an oncolytic agent. The lab is presently actively involved in the development of further oncolytic agents, for evaluation in the clinic.