Rigging Innate Immunity against the Flu

Main Text

In addition to providing great material for Hollywood movies, infectious pandemics pose a real threat to mankind. Therefore, the development of novel prophylactic and therapeutic approaches is a high priority. In a comprehensive series of studies in this issue of Molecular Therapy, Coch et al.¹ demonstrate that systemic activation of a viral RNA receptor, retinoic-acid-inducible gene I (RIG-I), not only protects mice against a lethal influenza A virus (IAV) challenge for at least a week, but also provides therapeutic benefit when administered as late as 30 hr postinfection. Further, treatment with a RIG-I agonist early after IAV infection reduces the symptoms of a bacterial superinfection without any obvious toxicity from the RIG-I therapy. Although a great deal of further work would need to be performed before such a treatment could be evaluated in humans, the results are noteworthy and likely to lead to further studies of this potential biodefense approach against IAV and studies to determine whether the protection might be extended more broadly to other potential pathogens.

The perpetual battle between pathogens and the host immune system has led to the evolution of a remarkable variety of innate immune pathways that are triggered upon detection of foreign nucleic acids in either the endosomes or cytoplasm of host cells (reviewed in Hartmann²). Some innate immune receptors, such as Toll-like receptors (TLRs), are selectively expressed in specific immune cell types, whereas other receptor pathways, such as the RNA helicase RIG-I, are expressed ubiquitously. RIG-I elegantly “recognizes” certain viral RNAs by the presence of a 5′ triphosphate on a blunt end, which is not present on host RNAs. RIG-I contains an RNA helicase domain and two CARD domains that bind the RNA duplex and trigger the downstream signaling adaptor MAVS (mitochondrial antiviral-signaling protein) (reviewed in Kohlway et al.³ and Kell and Gale⁴). RIG-I signaling via MAVS leads to the induction of type I interferon (IFN) responses via TBK1 and IRF7/8 and inflammasome activation. RIG-I is the major cytoplasmic innate immune sensor for IAV.⁵ In vitro-transcribed RNA has previously been used to activate RIG-I and has been reported to protect against flu challenge in vivo and a variety of RNA and DNA viruses in vitro, suggesting a broad potential for RIG-I activation. However, in vitro-transcribed RNAs can contain a variety of unintended structures and are subject to a range of artifacts,⁶ making the use of synthetic RNA ligands (as in the current study) essential for studies of RIG-I activation.

Coch et al.¹ administered a synthetic RIG-I ligand intravenously (i.v.), demonstrating the induction of expression of the anti-viral chemokine CXCL-10 in the lungs and complete protection from a lethal IAV challenge that persists for at least 7 days after a single treatment. With repeated administration of the RIG-I ligand, protection could be extended for several weeks; there was no desensitization or tachyphylaxis. Furthermore, the investigators demonstrated therapeutic efficacy up to 30 hr postinfection, a far greater hurdle than showing protection in the prophylactic models. One of the more intriguing results in this study derives from an experiment in which mice were first infected with flu, then, 18 hr later, subjected to postexposure RIG-I activation, and then, 4 days postflu infection, superinfected with S. pneumonia. Bacterial superinfection is a common cause of mortality in humans following flu infection, so it is of considerable interest that, under the authors’ experimental conditions, the RIG-I ligand reduced the infection severity. Although full protection was seen with RIG-I activation at 18 hr post infection, only partial protection was seen if the RIG-I agonist was administered at 30 hr. In routine clinical practice, therapeutic administration within 30 hr of infection is likely to be challenging and may be impractical considering the time required to make a diagnosis, but it is possible to envision postexposure therapy in specific settings.⁷

RIG-I does not detect viruses that lack a 5′ triphosphate or diphosphate in a blunt-ended double-stranded RNA, but triggering RIG-I with a synthetic ligand may, nevertheless, protect against pathogens that otherwise would not trigger this pathway during infection. Many pathogens inhibit the RIG-I and other innate immune defense pathways, but prophylactic therapy would avoid this issue. The authors used an RNA formulation that is already in use and appears to be safe in humans, potentially simplifying the clinical development of this approach.

Several clinical settings have been described in which the activation of innate immunity could provide a clinical benefit in defending against pulmonary infections.⁷ TLR9 agonists have been reported to protect against pulmonary and other infections by more than 20 different pathogens in species ranging from mice to sheep and primates (reviewed in Krieg⁸). However, a number of issues and unanswered questions have, so far, limited the clinical development of this and other similar approaches (Table 1). Although TLR activation has been reported to protect against IAV challenge in several prior studies,⁸ in the present mouse model, neither a TLR7/8 nor a TLR9 ligand reduced infection severity compared to the strong benefit of the RIG-I ligand. It is possible that different innate receptors may be required for protection against different pathogens and that different innate immune ligands will activate defenses that are optimal for different subsets of pathogens. For example, the TLR9 pathway has been reported to be required for resistance to gram-negative bacterial pneumonia in mice,⁹ while TLR9 deficient mice actually have increased resistance to Candida and Aspergillus infection compared to wild-type mice.¹⁰ In clinical settings, where patients are exposed to many different types of pathogens, it will be critical to understand whether activation of the RIG-I pathway to defend against a specific pathogen unintentionally increases the susceptibility to a different pathogen. Likewise, it will be important to determine the in vivo efficacy of RIG-I activation in protecting against a broader range of viral and bacterial pathogens.¹¹ Another option to explore is whether a combination of a RIG-I agonist with one or more other innate immune activators might provide superior protection.

Table 1.

Considerations in the Development of Innate Immune Activation as a Prophylaxis against Infection

Question	Comment
Is the treatment safe in healthy subjects?	large clinical trials are likely to be required to demonstrate safety
Could treatment increase the risk of other infections?	activation of innate immunity can be subverted by some pathogens to improve their replication efficiency
Is the treatment effective in the at-risk populations?	clinical studies in pediatric and elderly populations may be required
Is the route of administration practical?	subcutaneous or inhaled dosing may be acceptable; i.v. is not
Will the treatment cause the same or similar symptoms as infection?	in a real world setting of protecting against a pandemic, the ideal treatment should not induce flu-like symptoms, as this would complicate the task of distinguishing infected from uninfected subjects
Will the treatment induce a gene signature of infection?	this could complicate rapid diagnosis to distinguish infected from uninfected subjects, which would be critical in a pandemic

Unfortunately, innate immune activation generally leads to different effects in rodents and primates, so the results of RIG-I stimulation in mice may not be predictive of what will occur in humans. The first RIG-I agonist entered human clinical development in oncology earlier this year, so our understanding of the in vivo effects of RIG-I stimulation in humans is likely to expand rapidly. Another challenge for translating the present findings into humans is that the i.v. route of administration is generally considered impractical for outpatient clinical use. Further studies are required to determine whether a more convenient route of administration, such as subcutaneous delivery or inhalation, might provide a similar degree of protection.

Safety questions are likely to be paramount in advancing the clinical development of innate immune activation as an anti-infective strategy. How critical is the dose used for treatment? Considering the genetic diversity of human immunity, will dosing need to be individualized? What is the risk of RIG-I agonists triggering harmful inflammation or autoimmunity? These and other questions remain to be addressed, and the current study certainly provides a strong incentive for further investigations.