MEK inhibitors as novel host-targeted antivirals with a dual-benefit mode of action against hyperinflammatory respiratory viral diseases

Abstract

Acute hyperinflammatory virus infections, such as influenza or coronavirus disease-19, are still a major health burden worldwide. In these diseases, a massive overproduction of pro-inflammatory cytokines and chemokines (cytokine storm syndrome) determine the severity of the disease, especially in late stages. Direct-acting antivirals against these pathogens have to be administered very early after infection to be effective and may induce viral resistance. Here, we summarize data on a host-targeted strategy using inhibitors of the cellular Raf/MEK/ERK kinase cascade that not only block replication of different RNA viruses but also suppress the hyperinflammatory cytokine response upon infection. In the first phase-II clinical trial of that approach, the MEK inhibitor Zapnometinib shows evidence of clinical benefit.

Current Opinion in Virology 2023, 59:101304

This review comes from a themed issue on Antiviral strategies

Edited by Lieve Naesens and Bruno Canard

For complete overview about the section, refer “Antiviral strategies (2023)”

https://doi.org/10.1016/j.coviro.2023.101304

1879-6257/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction

Human infections with highly pathogenic respiratory viruses, such as the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of coronavirus disease-19 (COVID-19), and influenza viruses (IV), especially of the H5-, H7-, and H9 subtypes of highly pathogenic avian influenza viruses (HPAIV), can cause life-threatening systemic inflammatory disease in humans 1, 2. Respiratory failure as well as multiple organ and vascular dysfunction, involving the lungs, heart, kidney, and the brain, are symptoms of a viral sepsis that is caused by the generation of an imbalanced immune response and hyperinflammation as a reaction to the virus infection 2, 3.

The current treatment strategies for the clinical management of patients with influenza or COVID-19 include direct-acting antiviral drugs (DAA) to inhibit replication and further dissemination of the virus. Especially, the drug class of nucleoside analogs, including Remdesivir, Molnupiravir, or Favipiravir, which target the viral polymerases and interfere with the process of viral genome replication, demonstrated high antiviral activity against both SARS-CoV-2 or HPAIV, respectively 4, 5. More recently, the coronavirus 3C-like protease inhibitor Paxlovid also showed clinical efficiency in certain risk groups [6]. However, a therapeutic benefit is only achieved when the drugs are applied within the first days post infection. These DAAs also show low efficacy in hospitalized patients with a higher Clinical Severity Score (CSS) and need of oxygen to prevent death or reduce the time to recovery [7]. Therefore, the FDA primarily recommends immunomodulatory drugs, such as dexamethasone or anticytokine (receptor) monoclonal antibodies for severe hospitalized cases, which however, exhibit a rather broad immune dampening activity and side effects 8, 9. Thus, there is still a substantial requirement for safe drugs that eventually combine antiviral activity and immunomodulatory action.

Furthermore, it is relevant to mention that DAA treatment bares the risk of provoking resistance-introducing mutations in the viral genome. For example, rapid emergence of resistant IV was observed against the M2 ion channel inhibitor Amantadine 10, 11. Reduced sensitivity and resistance to IV or SARS-CoV-2 drugs were also reported for Oseltamivir, Baloxavir, or Remdesivir, respectively 12, 13, 14. To improve the available treatments and prevent the emergence of resistance and development of a severe disease outcome including organ dysfunction, new strategies need to be developed and clinically tested.

As all viruses depend on cellular factors and mechanisms for their replication, a feasible strategy might therefore be to target host factors instead of the pathogen itself. Here, we summarize the existing literature on a host-targeted strategy aiming to inhibit virus-induced signaling processes that are misused by different viruses.

The signaling pathway in focus is the Raf/MEK/ERK kinase cascade that is also known as the classical MAPK cascade [15]. This kinase pathway converts extracellular signals detected by, for example, cell surface receptors via stepwise phosphorylation and activation of the kinases Raf and MEK, leading to the activation of the downstream kinase ERK that has many further downstream targets in the cell. Under physiological conditions, the pathway regulates cellular processes such as proliferation, differentiation, and apoptosis, depending on cell type and activation context [16]. Aberrant overactivation of the pathway is often associated with tumor development. Hence, many attempts have been and are still undertaken to pharmacologically inhibit the pathway on the level of MEK, whose only target is ERK and thereby represents the bottleneck of this kinase cascade to fight tumors [17]. MEK inhibitors are special in that they are non-ATP-competitive and thus rather specific. Accordingly, MEK inhibitors show low toxicity and little adverse side effects in humans.

Here, we review data showing that the Raf/MEK/ERK cascade is hijacked by many viruses, in particular IV and SARS-CoV-2, to boost viral replication. MEK inhibition therefore not only impairs replication of viruses but can also act immunomodulatory, thereby exhibiting a mechanism of action with dual benefit in the treatment of acute hyperinflammatory viral diseases.

The role of the Raf/MEK/ERK kinase cascade in influenza virus replication

More than 20 years ago, it was first shown that the Raf/MEK/ERK cascade is activated by IV infection [18]. With regard to its activation, the kinase cascade is somewhat different to other IV-induced signaling cascades, since it is not primarily activated by the pathogen-associated molecular pattern, 5′-triphosphate RNA. Instead, it was shown that activation occurs via accumulation of viral hemagglutinin (HA) molecules in the cell membrane [19]. This appears to induce formation of lipid rafts, which serve as signaling platforms for protein kinase C (PKC)-dependent Raf/MEK/ERK activation, presumably by bringing monomeric receptor molecules in close proximity to each other. Other findings suggest that HA triggers a switch from MEK1 SUMOylation to activation of the ERK pathway, which may further enhance activity of the cascade [20].

Strikingly, it turned out that inhibition of the pathway by inhibitors of MEK or dominant negative mutants of MEK and ERK leads to reduced influenza A- or B virus (IAV, IBV) progeny 18, 21. The Raf/MEK/ERK cascade was thus considered to be the first example of a signaling pathway activated by IV for its own benefit (reviewed in [22]). With regard to the underlying mechanism, it was shown quite early on that MEK inhibition does not affect viral RNA and protein synthesis but results in a retention of the viral genome complexes (vRNPs) in the nucleus of infected cells [18]. The exact molecular events, however, remained enigmatic for a long time. Only recently in 2020, the complete chain of events how the Raf/MEK/ERK pathway promotes IV vRNP nuclear export has been unraveled [23]. Activation of the cascade in late stages of the replication cycle leads to the phosphorylation and activation of yet another downstream kinase, named ribosomal S6 kinase 1 (RSK1), that phosphorylates the viral nucleoprotein (NP), a major constituent of vRNPs, at two distinct sites. This phosphomodification provides a signal for the association of the viral matrix protein (M1) to the NP, a step that is essential for nuclear vRNP export via the CRM1 pathway [23]. Accordingly, MEK inhibitors impede RSK activation, preventing subsequent phosphorylation of NP and M1 association [23]. This finally leads to an impaired vRNP nuclear export in the presence of the inhibitor. While this seems to represent the dominant antiviral mode of action in late stages of infection, there may also be earlier steps controlled by the pathway, for example, the vacuolar ATPase that stimulates endosomal acidification required for viral fusion during entry of the virus [24].

Ever since the virus-supportive function of the Raf/MEK/ERK signaling cascade for IV replication has been unraveled, there were attempts to use the pathway as a novel target for antiviral intervention 25, 26. This strategy is facilitated by the fact that, as mentioned above, several inhibitors of the central kinase MEK, representing the bottleneck of the signaling pathway, are developed or are even licensed for clinical use in cancer therapy 27, 28 and thus would allow a repurposing approach. Along that line, several MEK inhibitors, such as U0126 18, 21, 29, 30, CI-1040 23, 31, or the licensed drug Trametinib [32], were shown to exhibit efficient anti-influenza virus activity in vitro and/or in the mouse model without any signs of adverse events. As predicted for a host-targeting drug, MEK inhibitors display a high barrier toward emergence of resistance [21]. Interestingly, the compounds were also shown to act synergistically with licensed drugs such as Oseltamivir [33] or Baloxavir [34], which would offer the possibility of combined usage. Finally, MEK inhibitors also showed a prolonged treatment window compared with standard of care in vivo [31], which would overcome a major drawback of DAAs, such as Oseltamivir, that have to be applied very early after infection.

MEK inhibitors block replication of different RNA viruses

As there are numerous reports that several RNA viruses manipulate the Raf/MEK/ERK pathway (reviewed in [35]), it was hypothesized that also other viruses besides IV may be sensitive to MEK inhibition. Since MEK inhibitors provoke nuclear retention of IV vRNPs, a first virus to be studied additionally was Borna disease virus (BoDV), the only other RNA virus besides IV with a nuclear phase. Indeed, it could be shown that the BoDV replication is impaired by MEK inhibition, however, there was no indication of impaired nuclear transport of viral factors [36]. Respiratory syncytial virus (RSV) is another important respiratory virus with a negative strand genome and has thus been studied [37]. Activation of the Raf/MEK/ERK cascade by RSV was also biphasic as observed earlier for IV [37]. The early induction within minutes after infection appears to be associated with viral attachment. Inhibition of the MEK kinase in late stages after the virus has been internalized also resulted in a reduction of viral titers and it was revealed that the late-stage activation of ERK is required for the secretory transport of the RSV fusion protein F, leading to impaired surface accumulation of the F protein [37].

For hantavirus, it has been shown that viral assembly depends on the transport of their viral proteins to the endoplasmic-reticulum–Golgi intermediate compartment (ERGIC). This transport is facilitated by dynein [38]. Because dynein intermediate chains are a substrate of ERK, this process could explain the mode of action of MEK inhibition. Namely, treatment with MEK inhibitors interferes with the phosphorylation of ERK and thereby inhibits the phosphorylation of the dynein intermediate chains, which consequently inhibits the dynein-dependent transport of the nucleocapsid protein to the ERGIC, preventing the successful assembly of new virions. Accordingly, MEK inhibitor treatment of mice infected with hantavirus strain Puumala resulted in reduced virus titers in lung and kidney (Patent No. WO/2021/069486).

With the upcoming Covid-19 pandemic, the question arose whether SARS-CoV-2 would also exploit Raf/MEK/ERK signaling and might be susceptible to inhibition of the pathway. Earlier studies provided some evidence that coronaviruses interact with Raf/MEK/ERK signaling. For example, genomic and subgenomic RNA synthesis of the murine betacoronavirus mouse hepatitis virus could be blocked through inhibition of MEK [39]. Furthermore, it was shown that the SARS-CoV spike(S) protein affects calcium-dependent activation of PKCα and promotes COX-2 protein synthesis via Raf/MEK/ERK activation [40]. These findings prompted studies on SARS-CoV-2 and it was found that infection of cells leads to transient activation of Raf/MEK/ERK signaling in the very early phase of viral entry [41]. The virus seems to depend on signaling via this pathway since ERK knockdown limited virus replication in cell culture. Accordingly, the MEK inhibitor Zapnometinib displayed strong anti-SARS-CoV-2 activity in cell lines as well as in primary air/liquid-interphase epithelial cell cultures, with a large selective treatment window [41]. Regarding the mechanism of action, preliminary data suggest that a very early post-entry step that is common in all cells and for all SARS-CoV-2 variants is affected by the inhibitor, since Zapnometinib was equally effective against viruses using either the TMPRSS2-mediated (endosomal) or cathepsin L-mediated (surface) entry route. Follow-up studies further demonstrated that Zapnometinib synergistically potentiated the effect of DAA against SARS-CoV-2 on viral replication [42]. Treatment combinations of Zapnometinib with nucleoside inhibitors Molnupiravir and Remdesivir or 3C-like protease inhibitors Nirmatrelvir and Ritonavir, the ingredients of the drug Paxlovid, showed strong synergistic effects in all tested concentrations of Zapnometinib in a concentration range that was very well-tolerated by cells. Such a synergistic effect appears not to be trivial, since combinations of two DAAs, namely Remdesivir and Nirmatrelvir, even lead to an antagonistic effect and higher virus titers in some concentration ranges [42]. This synergism of Zapnometinib with anti-SARS-CoV-2 DAAs resembles observations from older studies done with IV, where MEK inhibitors also acted synergistic with the anti-IV drugs Oseltamivir [33] or Baloxavir [34].

MEK inhibition alleviates virus-induced proinflammatory cytokine responses

Besides an antiviral activity against several RNA viruses, there is accumulating evidence that MEK inhibitors might confer a second benefit in hyperinflammatory virus infections due to their immunomodulating activity. The Raf/MEK/ERK pathway is active in many innate and adaptive immunity processes [43]. Inhibition of MEK leads to a reduction of excessive inflammation (cytokine storm) [44], a shift from a Th2 to a Th1 response [45], a reduction of regulatory T cells (Tregs) [46], and a support of clonal expansion of T cells, all processes required for an effective antiviral immune response. Accordingly, it has been shown in vitro and in vivo that MEK inhibition results in a reduced expression of proinflammatory cytokines [29]. Interestingly, there seems to be a differential sensitivity of antiviral versus proinflammatory genes to MEK inhibition, as proinflammatory cytokines were found to be reduced, while an antiviral innate type-I IFN response seems to be unaffected 32, 41. This would be a very beneficial feature since, in contrast to broad immunosuppressive drugs such as dexamethasone, MEK inhibition in patients might rebalance overshooting proinflammatory cytokine responses while the essential antiviral type-I IFN response remains fully intact.

Zapnometinib, a MEK inhibitor with unexpected features

Zapnometinib (pINN), also known as PD-0184264 or ATR-002, is one of several active metabolites of the MEK inhibitor CI-1040 47, 48. In an attempt to find MEK-inhibiting agents with improved antiviral action, Zapnometinib was evaluated in antiviral assays in vitro, however, it had to be applied in approx. 10–15-fold higher concentrations to achieve a similar antiviral efficacy compared with the prodrug CI-1040. These results were in line with the results of cell-free kinase assays and determination of ERK activity in cells [49], where also much higher concentrations of Zapnometinib were needed to block the kinase MEK. From this result, one would conclude that Zapnometinib is not superior over CI-1040 as an antiviral drug candidate. The picture, however, completely changed when animal experiments were performed. Here, the complete opposite was observed, as 5–10-fold less Zapnometinib was needed to reduce IAV lung virus titers in infected mice [49]. This was most likely due to enhanced and sustained plasma levels of Zapnometinib versus CI-1040 48, 49, that would greatly increase bioavailability of the compound. From these data, one can conclude that cell-free enzyme assays or in vitro efficacy studies may not always be fully valid to decide about the quality of a lead compound in vivo.

Another unexpected feature of Zapnometinib was its unpredicted in vitro activity against bacterial pathogens. In in vitro coinfection experiments using IAV and subsequent infection with S. aureus, it was observed that treatment with some MEK inhibitors, including CI-1040 and U0126, led to slightly reduced bacterial titers, and this effect was shown to be greatly enhanced when using Zapnometinib [50]. Surprisingly, Zapnometinib also led to reduced bacterial growth in suspension cultures and reduced stress- and antibiotic tolerance without resistance induction. Thus, these studies identified for the first time that a particular MEK inhibitor metabolite exhibits direct antibacterial activity, which is likely due to interference with the S. aureus kinase PknB to alter the PknB/Stp phosphatase signaling system [50]. The antibacterial activity was not restricted to S. aureus, but was also found for S. pneumonia and B. subtilis strains, which express similar Ser/Thr kinases as PknB [50]. While to date it is not clear whether this in vitro activity would also translate to the animal model or to humans, one could speculate that this may be an attractive feature in regard to treatment of bacterial coinfections during an IAV infection.

Zapnometinib as an anti-infective drug in clinical development

Zapnometinib is the first MEK inhibitor that has entered the stage of clinical development as an anti-infective agent and was brought forward into phase-I clinical trials (NCT04385420). A randomized, double-blind, placebo-controlled dose escalation study, which was finalized in 2019, demonstrated the safety and tolerability of the drug in 70 healthy volunteers. The observed pharmacokinetic profile supported an intended once-daily regime in a further phase-II clinical development [51]. In March 2021, a phase-II clinical trial to evaluate the safety and efficacy of Zapnometinib in adult-hospitalized patients with COVID-19 (RESPIRE) was initiated (NCT04776044). The study was terminated in August 2022. Results from the RESPIRE study provide proof of concept for the clinical benefit of Zapnometinib. The results further indicate a clinically relevant efficacy profile for Zapnometinib, in terms of the primary endpoint, improved CSS at Day 15, with a favorable safety profile.

Conclusion

DAAs are still regarded by many as the gold standard in treatment of viral diseases. While this notion has previously already been challenged in the case of hyperinflammatory acute viral infections, such as severe influenza, it was the disease course of COVID-19 that again has taught us a different lesson. Although several DAAs are in clinical use, they only show a clinical benefit when administered early after infection, while in late and severe stages of the disease, they are not effective anymore. This is due to the fact that in these late stages, the disease is mainly driven by a derailed immune response. Thus, treatment recommendations of the FDA mainly include immunomodulatory agents for higher CSS under need of oxygen. One drug of choice is dexamethasone, which is a very broadly acting immunosuppressant, that shows good efficiency but should not be taken too early in the course of the disease, because it would prevent mounting an antiviral IFN response that is needed to fight the virus.

To solve this dilemma, it seems that a paradigm change is needed in anti-infective therapy to efficiently fight acute hyperinflammatory viral disease. The whole course of a viral disease has to be taken into account rather than only focusing on the pathogen.

While it might not be the only novel strategy, the findings summarized in this review suggest that inhibition of the cellular kinase MEK is a novel approach that fulfills several needs for an effective therapy over the course of an hyperinflammatory disease progression. Besides their antiviral activity, MEK inhibitors also dampen expression of proinflammatory cytokines that drive the cytokine storm, while leaving the antiviral type-I IFN response intact ( Fig. 1). Results of a first phase-II clinical trial employing the MEK inhibitor Zapnometinib against hospitalized COVID-19 patients indicate a clinically relevant efficacy profile in terms of the primary endpoint, improved CSS at Day 15, with a very favorable safety profile. Taken together, the reviewed work provides convincing evidence that the concept of MEK inhibition as an anti-infective strategy, which was developed more than 20 years ago (featured in [52]), has been proven to be a clinically feasible and promising approach.

Figure 1.

Infection of cells with influenza virus, SARS-CoV-2 or RSV, leads to activation of the Raf/MEK/ERK pathway via different pathogen-associated molecular patterns such as vRNA accumulation or membrane accumulation of HA (IAV/IVB). Viruses have acquired the capability to exploit cellular signaling via the core kinases Raf, MEK, and ERK to support different steps in the virus life cycle, such as vRNP export (IAV/IBV), F-protein translocation (RSV), or entry (SARS-CoV-2). Infection also leads to induction of proinflammatory cytokines (presumably by regulation of the transcription factor NF-kB) and the type-I antiviral IFN response. Inhibition of Raf/MEK/ERK signaling by MEK inhibitors such as Zapnometinib impairs virus replication and also dampens proinflammatory cytokine expression, while the activation of STAT1 and the antiviral type-I IFN response is not significantly affected (effects of MEK inhibition shown in red). MEK inhibition also affects replication of other viruses such as hantaviruses or BoDV, however, the mechanisms of pathway activation and the step in the virus life cycle that is affected are less clear.

Conflict of interest statement

Work of SL is supported by grants of the German Research Foundation (DFG) (grants SFB1009 B02 and B13, KFO342 P06, and Lu477/23–2). Work of SP is supported by the German Center for Infection Research (DZIF, TTU 01.806 (broad-spectrum antivirals) and FF 01.901 (nucleoside booster)), partner site Giessen, Germany. SL, SP, and OP are members of the German FluResearchNet, a nationwide research network on zoonotic influenza. SL, SP, and OP are cofounders and advisory board members of Atriva Therapeutics GmbH, Tübingen, Germany, a company that develops signal transduction inhibitors against viral diseases.

The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the paper.

Data availability

Data will be made available on request.