COVID-19: Researchers prioritize clinical trials of drugs targeting IFNAR2 and ACE2 proteins

0
798

A new study using human genetics suggests researchers should prioritize clinical trials of drugs that target two proteins to manage COVID-19 in its early stages. The findings appeared online in the journal Nature Medicine in March 2021.

Based on their analyses, the researchers are calling for prioritizing clinical trials of drugs targeting the proteins IFNAR2 and ACE2. The goal is to identify existing drugs, either FDA-approved or in clinical development for other conditions, that can be repurposed for the early management of COVID-19. Doing so, they say, will help keep people with the virus from being hospitalized.

IFNAR2 is the target for approved drugs often used by patients with relapsing forms of the central nervous system disorder multiple sclerosis.

The researchers believe the most promising ACE2 therapy against COVID-19 is a drug that was developed before the pandemic began and has been evaluated in clinical trials to reduce inflammatory response in patients with severe respiratory disorders.

Dr. Juan P. Casas, a physician epidemiologist at the Veterans Affairs Boston Healthcare System, led the study. The research included collaborators from the University of Cambridge and the European Bioinformatics Institute in England, and Istituto Italiano di Tecnologia in Italy.

“When we started this project early last summer, most COVID-19 trials were being done on hospitalized patients,” Casas explains. “Very few treatments were being tested to give to patients early in the natural history of the disease. However, as the availability of testing against coronavirus increased, an opportunity opened to identify and treat COVID-19 patients before they progress to more severe forms that require hospitalization.

“The problem we tried to overcome,” he adds, “is how to identify if existing drugs, either approved or in clinical development for other conditions, can be repurposed for the early management of COVID-19. Most commonly used strategies for drug repurposing are based on pre-clinical studies, such as experiments in cells or animal models. However, those types of studies may have problems of reproducibility or difficulties in translating their findings to humans. That usually leads to higher rates of failure in clinical trials.”

Casas and his team used genetics as the starting point to identify drugs that can be repurposed for treating COVID-19. Large-scale human genetic studies have been widely used to inform drug development programs, with some research identifying COVID-19 drug targets.

“The reason we used human genetics is as follows,” says Casas, who is also a faculty member at Harvard Medical School. “Given that more than 90% of drugs target a human protein encoded by a gene, the opportunity is there to use genetic variants within those druggable genes as instruments to anticipate the effects that drugs targeting the same protein will have. In other words, genetic studies that used variants within druggable genes can be conceived as natural randomized trials.”

To put things into perspective, he refers to a gene that encodes a protein called PCSK9. The protein is the target of a class of drugs called PCSK9 inhibitors, which are used to lower cholesterol and prevent cardiovascular disease.

Researchers discovered that class of drugs because of studies showing that people carrying a certain variant within the PCSK9 gene tend to have high levels of cholesterol and are at greater risk for cardiovascular disease.

“That kind of genetic study was pivotal to identify the PSCK9 protein as a target for drug discovery,” Casas says. “It’s known that drug targets with human genetic support have a least twice the odds of success compare to the targets without human genetic support.”

Building on these known benefits of human genetics for drug discovery, Casas and his team set out to identify all genes that encode proteins that served as targets for FDA-approved drugs or drugs in clinical development. They called this set of 1,263 genes the “actionable druggable genome.” The genes were from two large genetic datasets that totaled more than 7,500 hospitalized COVID-19 patients and more than 1 million COVID-free controls.

By comparing the genetic profiles of the hospitalized patients and the controls, and looking at which drugs target which genes, the researchers were able to pinpoint the drugs most likely to prevent severe cases of COVID-19 that require hospitalization.

The two datasets were VA’s Million Veteran Program (MVP), one of the world’s largest sources for health and genetic information, and the COVID-19 Host Genetics Initiative, a consortium of more than 1,000 scientists from over 50 countries working collaboratively to share data and ideas, recruit patients, and disseminate findings.

“This study gets to the heart of why we built MVP,” says Dr. Sumitra Muralidhar, director of the Million Veteran Program. “It demonstrates the potential of MVP to discover new treatments, in this case for COVID-19.”

ACE2 is highly relevant to COVID-19 because the coronavirus uses that protein to enter human cells. The most promising ACE2 therapy against COVID-19 is the drug APN01, which mimics the protein. The drug works by confusing the coronavirus so it attaches to the drug instead of the ACE2 protein in the human cell.

Positive evidence is emerging from small clinical trials on the effectiveness of APN01 in COVID-19 patients, especially those that are hospitalized. “Hence, if our genetic findings are correct, there’s a need to test this strategy in clinical trials in COVID-19 outpatients,” Casas says.

The IFNAR2 protein serves as the target for a drug family known as type-I interferons, one of which is interferon beta. That drug is approved for treating patients with a degenerative form of multiple sclerosis, a chronic disease that attacks the central nervous system and disrupts the flow of information within the brain and between the brain and the body.

The researchers showed that people with a certain variant of IFNAR2 had less chance of being hospitalized due to COVID-19, compared to people without the variant.

Currently, Casas is early into planning a clinical trial to test the efficiency and safety of interferon beta in COVID-19 outpatients in VA. If his genetic findings are confirmed by a trial, he says the goal would be to prescribe the drug after people are diagnosed with COVID-19 but before their conditions require hospitalization.

Casas sees a continued need for drugs to treat people in the early phase of COVID-19, despite the ongoing worldwide vaccination campaigns.

“This is largely due to two reasons,” he says. “First, it will take some time to achieve the high levels of vaccine coverage needed to create herd immunity. In addition, certain coronavirus variants are emerging that seem to lead to a reduced vaccine efficiency. We are not yet in the clear.”


Type I interferons (IFN-I) are a family of cytokines that bind the type I interferon receptor, constituted of two transmembrane subunits, IFNAR1 and IFNAR2 (Figure 1). The two receptors are constituted of an extracellular domain, which binds IFN-I, a transmembrane helix and an unstructured intracellular domain (ICD) that binds JAKs and STATs (1, 2).

JAK1 is associated with IFNAR2 and TYK2 with IFNAR1. STAT1 and STAT2 (and maybe also other STATs) were found to be constitutively bound to the ICD of IFNAR2 (3–5). Binding results in close proximity of the intracellularly associated JAKs, JAK1 and TYK2, resulting in their activation through cross phosphorylation (Figure 1) (6, 7). This also results in receptor phosphorylation, which role is still under debate (3, 8–10).

The phosphorylated STATs dissociate from the receptor and form homo and hetero dimers, which are transported to the nucleus, where they serve as transcription factors for a large number of genes. The most prominent effects are associated with STAT1/STAT2 heterodimerization, which together with IRF9 form the interferon-stimulated gene factor 3 (ISGF3), which bind a distinct group of target genes harboring the interferon-stimulated response elements (ISRE). In addition to this, IFN-I drives STAT1/STAT1 and STAT3/STAT3 homodimerization, the formation of a STAT2/IRF9 binary complex and more (6, 10–12) (Figure 2).

This leads to the transcription activation or suppression of over 1,000 genes, which drive a wide range of innate and adaptive immune functions. These, in turn respond against various pathogens, act as important regulators in tumor immunity and have a role in pathophysiology and autoimmune diseases (10, 13–18).

STAT2 knockout cells still activate a STAT1/STAT1 response mediated by IRF1, while STAT1 knockout cells activate a STAT2/IRF9-induced response (10). Surprisingly, no change in the gene induction relative to wild-type cells was observed in STAT3 knockout HeLa cells, despite the strong IFN-I–induced phosphorylation of STAT3. However, as IFN-I responses are cell-type specific, a STAT3/STAT3-induced response may still be found in other cells than HeLa.

An external file that holds a picture, illustration, etc.
Object name is fimmu-11-595739-g001.jpg
Figure 1
The interferon response is initiated by IFN-I binding to the extracellular domains of IFNAR1 and IFNAR2. Following ternary-complex formation, the associated JAK kinases cross-phosphorylate each other as well as the associated STATs and tyrosine residues on the intracellular domains of the receptors. Upon phosphorylation the STATs are released and are transported to the nucleus. The STAT1/STAT2/IRF9 complex is strongest associated with IFN-I induced gene induction, albeit other STAT complexes are activated as well (see Figure 2 for details). The STAT complexes serve as transcription factors for many IFN-I induced genes. Three main feedback mechanisms quell IFN-I activity: Receptor Ubiquitination, resulting in receptor endocytosis (which is initiated within minutes from IFN-I induction) and SOCS and USP18, which are IFN-I induced genes and thus their feedback relates to their production to high levels (which takes hours).
An external file that holds a picture, illustration, etc.
Object name is fimmu-11-595739-g002.jpg
Figure 2
Ternary, IFN-I/IFNAR1/IFNAR2 complex formation results in the activation of multiple STAT complexes that serve as transcription factors for different genes. The activated STATs and IFN-I regulated genes vary between different cells, IFN-I subtype, its concentration and duration of activation, result in a pleiotropy of responses.

Due to this wide range of physiological responses, IFN-I has provided therapeutic benefits for multiple diseases, including multiple sclerosis, some cancers and viral diseases (hepatitis B and C) (19–21). Due to the efficient activation of antiviral activities by IFN-Is, most viruses have contemplated mechanisms to avoid its actions (22–24). For example, the Ebola virus, which outbreak in central Africa killed tens of thousands of people (25, 26), avoids IFN-I activity by producing the VP24 protein that binds the karyopherin alpha nuclear transporter. Thereby, it inhibits the nuclear transport of phosphorylated STAT1, rendering cells refractory to IFN-Is.

Another example of viral mechanisms that evolved to eliminate IFN-I functions in inducing innate immunity is given by the SARS corona virus, where both the production of IFNβ and the IFN-I induced signaling are attenuated. Recently, a more infective version of SARS has emerged, SARS-CoV-2 (which causes the COVID-19 disease). COVID-19 cases have been first reported by the end of 2019 in China, and rapidly became a world-wide epidemic with unprecedented consequences (27, 28). SARS-CoV-2 seems to have originated from horseshoe bats.

Similar virus strains that circulate in bats in Hubei province in China may in the future cause further new zoonotic outbreaks (29). SARS-CoV-2 has 83% homology to the SARS-CoV virus that also spread from China in 2002 (30). SARS-CoV-2 proved to be much more infectious compared to the original SARS virus, resulting in a global epidemic.

As IFN-I drives strong antiviral activities, the mechanisms SARS-CoV and SARS-CoV-2 combat IFN-I activities has been a matter of intense research, with at least 6 proteins being identified to counteract IFN-I functions in the SARS-CoV virus (31). In addition, IFN-Is were implicated in contributing to the severity of the cytokine storm, which is a major complication of SARS-CoV and SARS-CoV-2 and can lead to respiratory distress syndrome (ARDS) and death (31, 32).

In this review I will describe our current knowledge on the involvement of IFN-Is in the development of the COVID-19 disease, and how this relates to the different activities associated with type I interferons.

Common and Unique Features of Type I Interferon Signaling
Type I interferon receptors are found on all cell types, and are a major component of the innate immune system. Human type I interferons include 13 similar IFNαs with 80% homology between them and single IFNω, κ, ϵ and β, with lower homology (30–50%). All of them bind the receptor complex, composed of IFNAR1 and IFNAR2 at the same proximal location (1, 2, 33). Despite structural similarities among the ternary IFN-I-IFNAR1-IFNAR2 complexes, IFN-Is drive a range of different activities, dependent on the cell type and the interferon subtype (34). This apparent paradox has major implications for understanding the role of IFN-I in health and disease and its varied applications as a drug against a pleiotropy of diseases.

IFN-I signaling is initiated by binding of IFN-I to its receptor. It has been suggested that cytokine receptors are pre-associated, with ligand binding activating signaling through the induction of conformational changes (35). However, more recent single-molecule receptor tracking on life cells has clearly shown that for many of the cytokines, its role is to bring the receptors into close proximity, which drives signaling (36). This seems to be the case also for IFN-I induction, as shown both using single receptor tracking and mutational analysis (Figure 1) (37, 38).

While structurally, the ternary ligand-receptor complex seems to be the same for all IFN-Is, the binding affinity differs by many orders of magnitude. The tightest binding IFN-I is IFNβ, which binds IFNAR1 with 100 nM affinity and IFNAR2 with sub-nanomolar affinity. The different IFNα subtypes bind IFNAR1 with 0.5 to 5 µM affinity and IFNAR2 with 1 to 100 nM affinity, with IFNα1 being the weakest binding IFNα (39, 40). Even weaker binding was measured for IFNϵ, with ~100-fold reduced affinity relative to IFNα proteins (15). Interestingly, IFNϵ is constitutively expressed by the reproductive tract epithelium and is regulated by hormones during the estrus cycle, reproduction, menopause and by exogenous hormones. Thus, its mode of action is different from other IFN-Is (41).

These large differences in binding affinity between IFN-I subtypes were suggested to result in major differences in biological activity. To obtain a better insight into the molecular mechanisms of their actions, IFNα2 was engineered to cover the whole range of binding affinities of natural IFN-Is to both the high affinity (IFNAR2) and low affinity (IFNAR1) receptor chains (1). These studies have shown that indeed, the binding affinity to both receptors is a major determinant of IFN-I activity (42). Using both natural and engineered IFN-Is has shown that even weak binding IFN-Is activate the cellular antiviral program at very low (pM) concentrations (39).

Moreover, the antiviral program was activated in all cell-lines tested. Despite the 50-fold higher affinity of IFNβ over IFNα2 towards binding IFNAR receptors, its potency to elicit an antiviral response is similar. For example, in WISH cells (originally thought to be of amniotic origin, but later found to be a HeLa (cervix cancer) contaminant) the EC50 for antiviral activity of IFNα2 is 0.3 pM, while the EC50 for IFNβ is 0.15 pM (43).

WISH cells have been extensively used to characterize IFN-I activity, including for definition of IFN-I unit activity. An upper limit for antiviral potency was further verified by engineering an IFNα2 variant, YNS-α8-tail, with 50-fold tighter binding to IFNAR1 and 15-fold tighter binding to IFNAR2 in comparison to IFNα2 (thereby surpassing the receptor binding affinity of natural IFNβ). Still, the EC50 for antiviral activity is only 3-fold lower in comparison to IFNα2 (44, 45).

Conversely to antiviral activity, IFNβ is much more potent in activating the antiproliferative program relative to IFNα2, a result that was also verified using the IFNα2 variant, YNS-α8-tail (45). The EC50 for antiproliferative activity on WISH cells is 2 nM for IFNα2, 50 pM for IFNβ and 20 pM for YNS-α8-tail. A similar increase in antiproliferative potency was observed also for OVCAR3 and HeLa cells. Interestingly, while antiviral activity was observed in all cell lines tested, some cell lines were not susceptible to IFN-I induced antiproliferative activity (for example T47D and K562), independent on the concentration and subtype of IFN-I (45).

To better understand the molecular basis for this finding, IFN-I induced gene expression was monitored using various IFN-I subtypes or engineered mutants on the background of different cell-lines. These experiments showed that low concentrations of weaker binding interferons activate the expression of mostly antiviral genes. Higher concentrations of interferons activate also other genes, many of them related to immune-modulation (45).

Examples for such genes are chemokines such as CXCL10 and 11, which are involved in chemotaxis of T cells and natural killer cells, induction of apoptosis, regulation of cell growth and more. We gave the term of “robust” for the common IFN-I induced program (including its antiviral activity) and “tunable” for the other programs induced by IFN-Is, which include between others antiproliferative and immunomodulatory activities (34).

Further investigations into these two programs has shown that cells with low receptor numbers activate only the robust program, and that not all cell types execute the tunable program, conversely to the robust program that is common to all cells (46). Tighter binding IFN-Is at higher concentrations are essential for the activation of the tunable program. Genes upregulated by the robust program are mostly classical antiviral genes, such as MX1 and MX2, OAS1 and 2, PKR, IFIT1, 2 and 3, ISG15, and many more. Figure 3A shows a Venn diagram of RNAseq data for 4 different cell-lines induced with IFN-I. The diagram shows that 53 genes are commonly upregulated by all 4 cell-lines.

Figure 3B shows STRING protein interaction analysis of these common genes. Clearly, these form a tightly interacting mesh of gene products. Gene Ontology analysis shows these genes to have an extremely high signature for antiviral activity and IFN-I activation. Promoter analysis of common ISGs has shown them to be driven by the classical ISRE promoter sequence (45). Conversely, for tunable genes no clear promoter sequence was identified. The exact mechanism of how tunable genes are upregulated by IFN-I is thus not yet fully understood.

An external file that holds a picture, illustration, etc.
Object name is fimmu-11-595739-g003.jpg
Figure 3
Genes which expression was upregulated by over 3-fold in the following cell lines: HeLa, T47D, K562 and OVCAR3. (A) venn diagram of the upregulated genes. (B) STRING: functional protein association network analysis of upregulated genes in all 4 cell lines (53 genes). According to STRING and GO analysis, the commonly upregulated genes have a strong antiviral signature. The top GO terms (FDR <10−25) are response to type I interferon, innate immune response, response to virus, defense response and immune system process. It is interesting to note that antiviral genes constitute most of the upregulated genes common to all 4 cell lines. Antiviral genes are also the majority of upregulated genes in K562 and T47D cells. Conversely, OVCAR3 and HeLa cells have many unique upregulated genes, many of them related to immunomodulatory functions, cell cycle, apoptosys and more.

reference link : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7561359/


More information: et al, Actionable druggable genome-wide Mendelian randomization identifies repurposing opportunities for COVID-19, Nature Medicine (2021). DOI: 10.1038/s41591-021-01310-z

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Questo sito utilizza Akismet per ridurre lo spam. Scopri come vengono elaborati i dati derivati dai commenti.