TLR3 Agonists as Immunotherapeutic Agents

The interaction of innate and adaptive immune responses is extremely complex and just beginning to be understood in detail. The tendency for biomedical research to establish reproducible models, cell systems and receptor pathways, and then to focus on these systems can sometimes result in opportunity that is not appreciated. The Toll-like receptor (TLR)3 pathway as a therapeutic target falls into this category. There is an immense body of data that has been developed for more than 40 years; all of which is highly relevant to the TLR3 story and yet most of which predates the elucidation of the TLR pathways. A new recognition of TLR3, its unique properties that distinguish it from other TLR pathways, the specificity of relevant and long-available TLR agonists, and new studies of this biology utilizing modern assay methodology suggest that this target may be especially valuable as an adjunct to multiple immunotherapy strategies currently in use or in development.

The TLR pathways are a primary component of the innate immune response and form a primary danger signal response to the presence of microbial pathogens detected both internally and externally in the tissues and cells of higher organisms. The TLR pathways have been reviewed elsewhere in detail [1]. TLR3 is the specific intracellular recognition system that responds to RNA virus infection. The receptor recognizes and signals in response to the intracellular presence of dsRNA intermediates associated with retroviral nucleic acid replication. The receptors are located in the endoplasm of the cells likely to be exposed and infected [2]. TLR3 can be found on epithelial cells, a broad range of antigen-processing cells, tissue dendritic cells, monocytes, mast cells, NK cells and others. TLR3 signaling results in a local cytokine burst that is microenvironment- and cell-specific in character and results in a local inflammatory response adaptively selected to provide the greatest chance of infection control for the tissue in question. Examples of such specificities can be found in the recent work of Tanaka [3] and Rozig [4], but importantly, one should appreciate that the response to TLR3 stimulation is not uniform, and is quite likely different in specific tissue microenvironments from what the general literature might teach. However, there is no question that in response to TLR3 activation, acute response molecules, such as type I interferon, TNF-α, IL-12, MCP-1 and others, are likely to be produced locally.

A large body of literature pertaining to response to specific TLR activation predates the appreciation of the TLR pathways. Notably, Coley‘s toxin [5] activates multiple TLR pathways and a large body of research with interferon inducers dating back to Field‘s report in 1967 [6] involved TLR activation. Field produced polyribosinic:polyribocytidic acid (polyIC) as a means to induce interferon, and the literature subsequent to definition of the TLR family has generally reported this double-stranded synthetic RNA to be a TLR3-specific agonist. In fact, polyIC, which has significant limiting toxicity, is bioactive via combined TLR3 and MDA-5 signaling [7,8].

Several modified versions of polyIC were developed in the 1970s to reduce toxicity and alter the biological profile of these ‘modified interferon inducers‘. Carter developed a modified polyIC dsRNA by substituting a uridylic acid at a molar ratio of 12:1 in the synthesis of the polycytidylic acid strand resulting in a double-stranded molecule with occasional mismatches and a much more rapid metabolism in vivo(polyI:polyC₁₂U; polyIC₁₂U) [9]. Levy, in contrast, modified polyIC by stabilizing the molecule with polylysine and formulating it with carboxymethylcellulose (polyICLC) [10]. The receptor specificity of the Levy polyICLC is uncertain based on a literature review; however, the polyIC₁₂U of Carter et al. appears to have exquisite specificity for TLR3, which has clearly distinguished its pharmacology from that of polyIC [8,11]. In considering the signaling pathways of the various TLR receptors, Bagchi has demonstrated distinct biological responses to TLR3 signaling amongst the TLR receptors [12], relating to TLR3 signaling via a MyD88-independent adaptor molecule in contrast to other TLRs. This distinguishes the TLR3-specific agonist pharmacological profile from the profiles of other classes of TLR agonists. Thus, the safety and bioactivity of TLR3-specific agonists can be anticipated to be unique amongst potential TLR-stimulating agents. In addition to the two historical modified polyIC compounds, an additional TLR3 agonist, IPH3102, is reportedly in preclinical development at Innate Pharma (Marseille, France). Further details regarding the specificity, structure and function of this molecule has yet to be published.

The wide-reaching biological effects of TLR3 stimulation are well documented in the literature and are consistent with a local danger signal response. Having worked extensively with monoclonal antibodies to induce antigen cross-presentation and modified cellular and humoral immune responses to tumor antigens in our laboratory, we evaluated polyIC₁₂U and confirmed the potency of the compound in inducing dendritic cell maturation, and inducing release of local cytokines in culture systems including IL-4, IL-6, IL-12p70, IFN-γ, MIP-1α and TNF-α. We also demonstrated the potential of antigen immune responses to be modified by exposure of nonsensitized mononuclear cells to the compound in co-culture with an antigen-specific T-cell stimulation. In a PSA-transgenic mouse model, we found the compound could augment both cellular and humoral immune responses to PSA and PSA–anti-PSA immune complex. This model is characterized by immune nonresponsiveness to PSA. Finally, we demonstrated that tumor targeting antibody-dependent cellular cytoxcity (ADCC) mediated by a monoclonal antibody to a tumor antigen could be enhanced in the presence of polyIC₁₂U [13]. Gowen et al. have studied the ability of polyIC₁₂U to protect mice from lethal Punta virus infection.

Interestingly, the lethality of infection in wild-type mice is substantially reduced in TLR3^-/--knockout mice. The mortality associated with infection relates to a massive inflammatory response in the lungs, less severe in the TLR3^-/--knockout animals. However, treatment of wild-type mice with polyIC₁₂U 24 h following infection with a lethal viral inoculation is fully protective, whereas dosing 4 h prior to inoculation is ineffective in preventing mortality [11,14]. In another application, Ichinohe has reported that intranasal polyIC₁₂U administered to mice in conjunction with prototype avian flu vaccines, greatly enhances the cellular and humoral immunity achievable with the vaccine and also protects mice from lethal infection [15]. A commercial formulation of polyIC₁₂U (Ampligen^® [rintatolimod], Hemispherx Biopharma, PA, USA) has been advanced to late-stage clinical development as therapy for chronic fatigue syndrome. The large body of clinical experience is important in establishing a framework of safety for this TLR3-specific agent. It is noteworthy that the recent report of Lombardi suggests that chronic fatigue syndrome is associated with chronic retroviral infection [16], although the causal association to xenotropic murine leukemia virus-related virus (XMRV) is not established. Clinical activity may be related to antiviral properties.

As TLR3 is ubiquitous in most experimental systems, the literature abounds with evidence of bioactivity, most recently for example in models of hepatitis C viral replication [17] and relating to the involvement of NK cells in yellow fever vaccination [18]; but what is evident is that lethality of infection is often associated with an overly robust immune response and that the character of an induced immune response is highly modifiable based on route of exposures cell types present at the site of pathology and the timing of exposures relative to intervention. TLR3 agonists were originally developed as interferon inducers, but with the emergence of recombinant type I interferon products, enthusiasm for the clinical development of these compounds as a source of interferon faded. The biology of the TLR pathways has now been elucidated, and TLR3 has been distinguished from other, perhaps more toxic, TLR pathways; a distinction tied to MyD88-independent signaling. Confusion in the literature regarding TLR3 may be associated with the common and incorrect assumption that the pharmacology of polyIC is TLR3 specific. Therefore, the literature must be interpreted with caution. The successful translation of combination therapies to mobilize immunity for therapeutic benefit is a challenge that can now be tackled rationally. Regulatory and commercial barriers to studying combinations of experimental agents should be revisited to facilitate the study of TLR3 agonists in conjunction with a variety of biological compounds in development or commercially available. There is potential for clinical applications in both infectious disease and oncology. With the advances in our appreciation of the complexity of induced immunity, human clinical studies addressing alternative routes of sensitization and the significant importance of timing in inducing immune stimulatory responses for therapeutic benefit can be rationally designed. It must be appreciated that previous failures in mobilizing effective therapeutic immunity do not preclude the potential of established bioactive compounds to be found clinically effective. Failed studies should only be interpreted as demonstrations of treatment strategies that did not work, and not as definitive failures of the bioactive compounds in question. The challenge is not routine steady-state receptor-based pharmacology, but rather the artful and synchronized mobilization of multifactorial natural immune pathways while avoiding excessive toxicity. Improvements to the current standards of controlling incipient and active infection, and importantly, future standards in mobilizing immunity to eradicate micro- and macro-scopic malignancy are now within reach.

Acknowledgements

Christopher F Nicodemus thanks Sara Buczynski for her assistance with the manuscript.

Financial & competing interests disclosure

Christopher F Nicodemus is a consultant to Hemispherx Biopharma, Inc. and owns shares in the company. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

Christopher F Nicodemus is a consultant to Hemispherx Biopharma, Inc. and owns shares in the company. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.