3.2. Interferon suppression

Although CXCL10, as well as CCL2, are interferon (IFN)-induced genes, there is evidence for impaired or delayed Type 1 IFN signaling in SARS-CoV-2-infected cells. One ex vivo experiment with lung tissue showed that SARS-CoV-2 induced less IFNs and pro-inflammatory mediators than SARS-CoV (Chu et al., 2020). Single-cell RNA sequencing analysis of bronchoalveolar lavage samples from severe and mild COVID-19 patients revealed that SARS-CoV-2 mainly infects the epithelial and recruited inflammatory macrophage subsets (Bost et al., 2020). In the latter, a disease severity-associated downregulation of type I IFN genes was noted. Notably, IFN is known to exhibit multiple biological functions such as antiviral, antiproliferative, and immunomodulatory effects (Nile et al., 2020; Wang et al., 2019). How SARS-CoV-2 thwarts intrinsic innate immune responses in monocyte-macrophages is not defined, although in monocyte-derived dendritic cells (but not macrophages) viral antagonism of STAT1 phosphorylation was reported (Yang et al., 2020). In contrast, work in Vero cells, indicates that SARS-CoV-2-infected cells are still responsive to type I IFN treatment unlike SARS-CoV-infected cells (Lokugamage et al., 2020). Of note, ACE2 was shown to be an interferon-stimulated gene in human lung cells, which is also upregulated by smoking and viral infections (Smith et al., 2020). A discussion of possible means by which SARS-CoV-2 attenuates the interferon response can be found elsewhere (Paces et al., 2020). Recently, it was reported that the SARS-CoV-2 viral ORF6, ORF8 and N proteins were potential inhibitors of the type I interferon signaling pathway (Li et al., 2020).

In light of these observations and urgent need to identify new therapies to control COVID-19 severity, IFN approved drugs have emerged as a potential treatment for COVID-19 patients. For instance, it has been demonstrated that the administration of recombinant IFNs to SARS-CoV and SARS-CoV-2 patients decreased viral protein synthesis and replication (Falzarano et al., 2013; Li et al., 2019; Zumla et al., 2016). In agreement, a recent published study on MERS-CoV patients reported that a combination of remdisevir and IFN beta showed a superior antiviral effect when compared with lopinavir/ritonavir combination (Sheahan et al., 2020). Therefore, testing the efficacy and safety of recombinant IFNs may be a worthwhile promising approach in the setting of COVID-19. Triple antiviral therapy with lopinavir-ritonavir, ribavirin and interferon beta-1b was reported to be safe and superior to lopinavir-alone in improving symptoms and reducing viral shedding and hospitalization in those with mild to moderate COVID-19 (Hung et al., 2020). On the other hand, there is evidence that IFN might be playing an important role in COVID-19 hyper-inflammation, suggesting that timing is a consideration (Conti et al., 2020c; Lee et al., 2020). Analysis of monocytes by single-cell RNA-seq from patients with severe COVID-19 exhibited signs of a type I IFN response along with TNF/IL-1β-driven inflammation.

3.3. Possible contribution of nicotine and nicotinic acetylcholine receptors

Although multiple investigations report a detrimental impact of nicotine on COVID-19 patients through up-regulating ACE2 receptors in the lungs (Farsalinos et al., 2020; Leung et al., 2020; Russo et al., 2020), recently published epidemiological studies reveal that smokers are either asymptomatic or show less severe respiratory symptoms compared with non-smokers (Covid et al., 2020; Farsalinos et al., 2020b; Kloc et al., 2020; Miyara et al., 2020; Petrilli et al., 2020). A disruption of the cholinergic anti-inflammatory pathway in COVID-19 patients has been noted (Farsalinos et al., 2020a, 2020c). It has been reported that over-responsiveness of the immune system, otherwise known as the cytokine storm, highly correlates with enhanced severity of COVID-19 infection, substantially increasing the mortality rate (Wang et al., 2020a; Ye et al., 2020). In the human lungs, the inflammatory response is mainly mediated by lung macrophages with two main types: the alveolar and interstitial macrophages (Kloc et al., 2020). Under physiological conditions, the alveolar macrophages exhibit anti-inflammatory characteristics by dampening the adaptive immune response and suppressing pro-inflammatory cytokines release (Kloc et al., 2020). Following a viral infection such as COVID-19, the alveolar macrophages switch from the anti- to pro-inflammatory phenotype, initiating consequently an inflammatory response, then switch back during the resolution phase to the anti-inflammatory phenotype, promoting thereafter tissue repair in the site of injury (Hu and Christman, 2019; Hussell and Bell, 2014). In the context of COVID-19 infection, an accumulation of macrophages in the lungs of COVID-19 patients has been observed (Wang et al., 2020a). Besides resident macrophages, monocyte-derived and non-resident macrophages have been described in COVID-19 patients (Chua et al., 2020); however, a better understanding of their interrelationship is needed.

Of note, lung macrophages have been shown to express ACE2 receptors, facilitating therefore the entry of SARS-CoV-2 to host cells (Tsaytler et al., 2011; Verdecchia et al., 2020). Besides ACE2 receptors, lung macrophages express alpha 7 nicotinic receptors (nAChRs α7) (Abrial et al., 2012). nAChRs α7 are potentially implicated in attenuating the cytokine storm through decreasing pro-inflammatory cytokine release (Kalamida et al., 2007; Tracey, 2002). For instance, it has been indicated that activation of nAChRs α7 located on lung macrophages by acetylcholine and/or nicotine mitigates the hyper-inflammatory response mediated disease severity (Lu et al., 2014; Tindle et al., 2020). Strong evidence reveals that the cholinergic anti-inflammatory pathway mediated by nAChRs α7 inhibits the translocation of the pro-inflammatory marker NF-κB to the nucleus and activates the JAK2-STAT3 pathway, consequently suppressing the inflammatory response and decreasing the cytokine storm in the lungs (Báez-Pagán et al., 2015; Changeux et al., 2020; Lu et al., 2014). Given the observed lower number of hospitalized COVID-19 patients among smokers, the potential role of medicinal nicotine to alleviate COVID-19 progression and development should be rapidly studied and clearly distinguished from conventional smoking that has no therapeutic effects.

3.4. Chemokine profile: a possible role for CXCL10 (IP-10)

Longitudinal profiling of 71 COVID-19 patients identified early expression of inhibitory mediators IL-10 and IL-1RA, along with the chemokine CCL5 (aka RANTES), in those with mild but not severe disease (Zhao et al., 2020). CCL5 is chemotactic for T cells, as well as eosinophils and basophil. On the other hand, the majority of cytokines associated with the cytokine storm in viral infections, including IL-6 and IFN-γ, were only increased at a late stage in severe illness, with TNF and GM-CSF not showing a difference between mild and severe cases.

Multiple studies have documented the upregulation of not only inflammatory cytokines but also chemokines in COVID-19 patients. Chemokines are low molecular weight proteins that act largely as chemoattractants for immune cell recruitment during inflammation, as well as modulators of immune cell homeostasis and angiogenesis (Coperchini et al., 2020). Compared to non-ICU patients, COVID-19 patients admitted to the ICU, exhibited higher plasma levels of IL2, IL7, IL10, GSCF, CXCL10 (IP-10), CCL2 (MCP1), CCL3 (MIP1A), and TNFα, indicating activation of T-helper 1 (Th1) cell function (Huang et al., 2020), although increased circulating levels of Th2-immune related cytokines IL-4 and IL-10 implicated in inflammation suppression are noted as well (Han et al., 2020). Transcriptomic analysis of bronchoalveolar lavage fluid of COVID-19 patients revealed an upregulation of CXCL1, CXCL2, CXCL6, CXCL8 (IL8), CXCL10 (IP-10), CCL2 (MCP-1), CCL3 (MIP-1A), and CCL4 (MIP1B) (Xiong et al., 2020). CXCL10 (IP-10) is a chemoattractant for monocytes/macrophages, dendritic cells, NK cells, and T cells; CCL2 (MCP-1) is a chemoattractant for monocytes, dendritic cells, and memory T cells. CXCL2 and CXCL8, which are secreted by monocytes/macrophage, serve as potent chemoattractants for neutrophils. Single cell RNA sequencing of nasopharyngeal and bronchial samples from COVID-19 patients identified increased inflammatory macrophages that express CCL2, CCL3 (MIP-1A), CCL20, CXCL1, CXCL3, CXCL10 (IP-10), CXCL8 (IL8), IL1B and TNF-α (Chua et al., 2020). Levels correlated with disease severity. CXCL10 (IP-10) levels were previously associated with the severe acute respiratory syndrome (SARS) disease progression and resolution due to the SARS-CoV virus (Altara et al., 2016; Jiang et al., 2005), and development of ARDS in preclinical models (Coperchini et al., 2020). The elevated nasopharyngeal levels of CXCL10 with COVID-19 may permit this chemokine to be used in widespread immunoassay testing for early detection of SARS-CoV-2-infection (Cheemarla et al., 2020).

3.5. Possible contribution of GM-CSF

Mounting evidence suggests that immunomodulatory agents, including GM-CSF, could be a promising therapy for COVID-19 (Lang et al., 2020; Mehta et al., 2020a). GM-CSF is known to be implicated in the production of granulocytes, monocytes, macrophages, and dendritic cells from progenitor cells, a process known as myelopoiesis (Egea et al., 2010; Fleetwood et al., 2007). It has been demonstrated that GM-CSF is secreted by different cell types including alveolar type II epithelial cells, playing therefore a key role in the integrity of alveolar barriers and maturation of alveolar macrophages (Cakarova et al., 2009; Rösler and Herold, 2016). Multiple investigations have considered GM-CSF as a pivotal cytokine that activates both the innate and adaptive immune response. For instance, GM-CSF can polarize myeloid cells into a pro-inflammatory phenotype, releasing subsequently reactive oxygen species and pro-inflammatory cytokines such as IL-1β, IL-6, TNF-α, and chemokines including CCL17, CCL2, and IL8, which can attract lymphocytes, monocytes, and neutrophils to the site of inflammation (Hamilton, 2020). It has also been reported that GM-CSF can prime dendritic cells to activate T cells, boosting thereafter the immune response by enhancing the recruitment of myeloid cells to the site of injury (Cao et al., 2015; Komuczki et al., 2019; Zhang et al., 2013). Since the goal of enhancing lung tissues integrity and dampening hyper-active immune response may lead to a drastic decrease in morbidity and mortality rate in COVID-19 patients, administration of GM-CSF as a promising therapy is being clinically investigated (Lang et al., 2020). Pre-clinical investigations revealed that overexpression of GM-CSF decreased apoptosis in alveolar wall cells, consequently preventing hyperoxia-induced lung damage (Baleeiro et al., 2006; Paine et al., 2003). A clinical study performed by Matute-Bello et al. reported that in ARDS patients, increased GM-CSF in bronchoalveolar lavage fluid was associated with decreased mortality rate through potentially improved alveolar macrophage survival (Matute-Bello et al., 2000). This observation was further strengthened with a clinical study completed by Herold et al. showing that administration of inhaled GM-CSF to patients with pneumonia-associated ARDS enhanced oxygenation and lung compliance (Herold et al., 2014). Currently, a clinical study is assessing the potential beneficial effect of using inhaled and intravenous GM-CSF agonist in respiratory failure COVID-19 patients (NCT04400929).

The potential benefits of administrating GM-CSF agonist in the context of COVID-19 patients, however, should be carefully studied, particularly in the late stage of COVID-19 where lung injury is thought to be driven by the cytokine storm rather than viral overload (Siddiqi and Mehra, 2020). Paradoxically, considerable interest in administrating anti-GM-CSF is gaining interest in the setting of COVID-19, given that a marked increase in GM-CSF expressing natural killer, B cells, and CD⁺ 4 and CD⁺ 8 T cells was observed in COVID-19 ICU patients when compared to mild cases (Zhou et al., 2020). However, given the role of GM-CSF in boosting the immune response to remove pathogen and enhancing lung repair, it is important to consider that the observed increase could be a result of exacerbated COVID-19 severity and related comorbidities. The rational is that during COVID-19 infection, over-activation of myeloid cells could be a critical mediator of enhanced cytokine storm, consequently aggravating tissue damage. Therefore, anti-GM-CSF therapy may decrease the detrimental immune response, and thus exert beneficial effects (Barnes et al., 2020; Mehta et al., 2020a; Merad and Martin, 2020), a hypothesis that was supported by a preclinical study of SARS-CoV infection animal model, showing that GM-CSF mediated the infiltration of inflammatory monocytes/macrophages into the lungs (Channappanavar et al., 2016). Taking together, these findings suggest that GM-CSF is a key player in regulating myeloid cell induced hyper-inflammation in many tissues including the lungs. Anti-GM-CSF approach in patients with COVID-19, however, should be well monitored, given the critical contribution of GM-CSF in alveolar macrophage function and pathogen clearance.

As of the start of May 2020, there were some 49 clinical trials underway targeting the cytokine storm in COVID-19 patients (Wang et al., 2020b). The vast majority involve biologicals. Besides those involving GM-CSF, prominent among them are a number of studies involving anti-IL-6 strategies. In addition, antagonistic antibodies directed against TNF, IL-1, IL-1R, and IL-8 are being investigated for attenuating excessive immune activation and the cytokine storm (Conti et al., 2020b). The rationale behind those targeting the actions of GM-CSF latter in COVID-19 is that this cytokine constitutes an autocrine/paracrine positive feedback loop that helps drive the cytokine storm (Mehta et al., 2020b). In a preliminary study, dexamethasone showed promise in reducing mortality of hospitalized COVID-19 patients if they were receiving respiratory support (mechanical ventilation or oxygen) (Group et al., 2020), but targeting the cytokine storm via broad-spectrum immunosuppression does raise a number of concerns (Theoharides and Conti, 2020).

This work was supported by a grant to FAZ from the American University of Beirut Faculty of Medicine (MPP – 320145/320095) and by Centre National de la Recherche Scientifique (CNRS) #103507/103487/103941; Seed grant #100410; and Collaborative Research Stimulus (CRS) #103556. RA acknowledges the support of the Institute of Experimental Medical Research (IEMR, OUS). GWB acknowledges the support of the Department of Pharmacology and Toxicology (UMMC).