Detection of viral antigens leads to antigen presentation to natural killer cells and CD8
+ T-cells, which activate the immune response by production of proinflammatory cytokines and chemokines
[161]. Normally, viruses are first faced by the innate immune response, but efficient counteraction may also require mobilization of the adaptive immune response (antigen targeting by immune cells). SARS-CoV-2 infection is linked to both responses
[162]. However, BETs, and especially BRD4, are mainly involved in the innate immune response (reviewed in
[48]). Production of proinflammatory cytokines is highly dependent on the Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-ĸB) signaling pathway
[163]. Cytokines are essential for the immune response, but their aberrant dysregulation leads to hyperinflammation and may cause severe damage to tissues resulting in organ failure and death. This uncontrolled systemic inflammatory response is known as cytokine storm (CS)
[164] and may account for up to 5% of COVID-19 patients
[165]. Indeed, over-stimulation of the immune response can be more damaging than the virus infection itself. ACE2 is involved in cleaving angiotensin II into anti-inflammatory angiotensins 1–7. Thus, it has been suggested that ACE2 blocking by SARS-CoV-2 binding should result in accumulation of pro-inflammatory angiotensin II, contributing to the exacerbated immune response
[166]. However, uncontrolled inflammation might also rely on pre-inflammatory states of certain organs and/or tissues.
3.1. SARS-CoV-2
Coronaviruses belong to the subfamily
Orthocoronavirinae of the family
Coronaviridae in the order
Nidovirales. They are highly diverse, enveloped, positive-sense and single-stranded RNA viruses
[167]. According to their genome structure and phylogenetic relationships, there are four genera of coronaviruses: alpha-, beta-, gamma- and delta-coronaviruses. While gamma- and delta-coronaviruses infect birds and some mammals, alpha- and beta-coronaviruses are responsible for the infection of various types of mammals
[168], causing respiratory disorders in humans and gastroenteritis in other animals
[169][170]. SARS-CoV-2 belongs to the beta-coronaviruses and is the seventh member of the coronavirus family to cause infections in humans
[171]. Among coronaviruses infecting humans we find several common cold viruses like HCoV-OC43, HCoV-HKU1 and HCoV-229E
[172], but coronaviruses with a high pathogenic capacity in humans have emerged in the last two decades, including SARS-CoV in 2002 and 2003 which caused 8000 confirmed cases with a 10% fatality rate, and MERS-CoV in 2012 with 2500 cases and a death rate of 36%
[173]. Analysis of SARS-CoV-2 genome revealed 79.5% genomic identity with SARS-CoV
[170].
The SARS-CoV-2 genome of approximately 30 kb encodes 14 open reading frames (ORFs) (
Figure 2A). The ORF1ab is located in the 5′ region and encodes the overlapping polyproteins 1ab and 1a, which undergo auto-proteolysis to give rise to 16 non-structural proteins (nsps), mostly contributing to the formation of the replicase/transcriptase complex (RTC). At the 3′ end up to 13 ORFs are present, which include four structural proteins (
Figure 2B): Spike (S), Envelope (E), Membrane (M) and Nucleocapsid (N) and nine putative accessory factors
[174][175].
Figure 2. SARS-CoV-2 genomic structure and proteins. (A) Schematic representation of the SARS-CoV-2 genome, indicating the different derived proteins. ORF 1ab give rise to two polyproteins for non-structural proteins: pp1a and pp1ab, this last through a programmed ribosomal frameshift. Structural proteins comprise: Spike (S), Envelope (E), Membrane (M) and Nucleocapsid (N) proteins. (B) Schematic representation of the viral particle showing the viral envelope, the single strand RNA (ssRNA) genome and structural proteins (proteins are not represented to scale). Interaction of S protein with ACE2 (under the control of BET proteins) and direct interaction of E protein with BETs (putatively involving the bromodomains) is also indicated.
The N protein binds to the viral genome and is involved in RNA replication, virion formation and immune evasion, and also interacts with the M protein
[176]. The M protein promotes viral particle assembly and budding through interaction with the N protein and accessory proteins 3a and 7a
[161][177]. The most relevant structural proteins for BET inhibition are E and S. E protein has been demonstrated to interact with BRD2 and BRD4
[6]. It is the smallest structural protein and facilitates the production, maturation and release of virions
[178]. S protein is a transmembrane protein that facilitates the binding of the viral envelope to the host receptor. The main receptor of SARS-CoV-2 in host cells is ACE2
[170][179], which is expressed in the colon, gallbladder, heart, kidney, epididymis, breast, ovary, lung, prostate, esophagus, tongue, liver, pancreas, and cerebellum
[180]. Similar to other coronaviruses, SARS-CoV-2 needs to proteolyze S protein to activate the endocytic pathway. Host proteases, including transmembrane protease serine 2 (TMPRSS2), cathepsin L and furine have been shown to be involved in the process
[179][181][182]. TMPRSS2 is highly expressed in certain tissues and co-expressed with ACE2 in bronchial branches, lungs and nasal epithelial cells, which explains part of the tissue tropism of SARS-CoV-2
[183][184]. Notably, expression of ACE2, but also of TMPRSS2, is under the control of BET proteins, what has been exploited to fight SARS-CoV-2 infection
[7][8][9][10] (see below).
Of the 4 structural proteins, SARS-CoV-2 shares more than 90% amino acid identity with SARS-CoV, except for the S protein, which presents greater divergence
[170][185]. SARS-CoV uses the same receptor as SARS-CoV-2 to infect cells, suggesting that both viruses may have similar life cycles
[170]. However, the binding affinity of the S protein of SARS-CoV-2 to the human receptor is much higher than that of SARS-CoV
[186]. This may be due to sequence differences, which makes the binding affinity of SARS-CoV-2 to ACE2 10 to 20 times higher than that of SARS-CoV, thus, enhancing the spread of SARS-CoV-2
[187][188]. The uncontrolled expansion of the virus is giving rise to new variants, with enhanced infectivity in some cases, that can challenge the control of the pandemic and compromise the efficiency of recently developed vaccines. Both, mutations in S protein leading to enhanced affinity for ACE2 receptor, and mutations reducing neutralizing activity of antibodies (immune escape), may be related to higher infectivity of new variants
[189]. Since all SARS-CoV-2 variants use ACE2 for entry into the host cell, it is predicted that strategies targeting ACE2, should be effective in reducing infection by new variants. These strategies include BET inhibition, as explained below.
Once SARS-CoV-2 has entered the cell, the genetic material is released into the cytoplasm and starts translation. The first and only region directly translated from the genome is that coding for nsps (ORF1ab)
[190]. Polyproteins 1ab and 1a are processed by the action of chymotrypsin-like protease (3CLpro, or main protease (Mpro)) and one or two papain-like proteases (PLpro) that are encoded by the virus
[191]. The RTC locates in double membrane vesicles creating a protective microenvironment for replication of genomic RNA and transcription of subgenomic mRNAs. The subgenomic mRNAs are translated into accessory and structural proteins M, S and E that are isolated in the endoplasmic reticulum and then translocated to the endoplasmic reticulum-Golgi intermediate compartment. Subsequently they interact with the newly produced genomic RNA encapsidated by N protein, resulting in the formation of vesicles that are exported out of the cell through exocytosis
[190][192][193].
3.2. SARS-CoV-2 Induced Immune Response and BET Proteins
The immune response is activated through recognition of both pathogen-associated and damage-associated molecular patterns by cell surface and intracellular pattern recognition receptors. In this scenario, toll-like family of pattern recognition receptors (TLR) play an important function. Expression of both TLR3 and TLR4 is upregulated by SARS-CoV-2
[194][195]. In addition, S protein interacts with and activates TLR4
[196][197]. BETs have been shown to positively regulate TLR4 expression in pancreatic ductal adenocarcinoma and in acute myocardial infarction rodent models
[198][199]. In turn, TLR3 is activated by foreign RNAs molecules (particularly dsRNAs derived from virus replication), and it has been established that TLR3-induced acute airway inflammation and remodeling is efficiently neutralized by BET inhibitors, as it depends on BRD4
[200].
TLR3 and TLR4 stimulation results in NF-ĸB activation
[161]. Of note, SARS-CoV-2 infection results in higher NF-ĸB pathway activation
[201]. Notably, BRD4 control of innate immunity largely relies on BRD4 regulation of canonical NF-ĸB pathway, associated with transcription factor RELA. Under normal conditions, IĸBα blocks RELA in the cytoplasm. Inflammation-associated activation of IĸB kinases results in IĸBα phosphorylation, which is ubiquitinated and then targeted for degradation. Liberated RELA, undergoes translocation to the nucleus for regulation of inflammatory and immunomodulatory genes. RELA-mediated transcription activation depends on BRD4. Transcription activation by RELA requires its acetylation at K310, which is recognized by BRD4
[202]. BRD4 binding seems to stabilize RELA, since BET inhibition or BRD4 depletion leads to RELA ubiquitination and degradation
[202][203]. In turn, RELA acetylation requires phosphorylation at S276
[204], which also depends on BRD4
[205]. Phosphorylation-coupled acetylation of RELA has been shown to facilitate BRD4 binding and recruitment of P-TEFb for transcriptional elongation of inflammatory cytokine genes upon RSV viral infection
[204].
Activated NF-ĸB cooperates with Interferon Regulatory Factor (IRF) 3 to induce proinflammatory cytokines like type I interferon molecules (IFNs) and Tumor Necrosis Factor (TNF)
[161]. Notably, upregulation of both types of molecules has been observed in COVID-19 patients
[195]. IRFs are the main transcription factors involved in production of IFNs and are key regulators of antiviral immunity. It has been described that following RSV viral infection, the BRD4/RELA complex recruits the P-TEFb component CDK9 to
IRF1 and
IRF7 promoters for enhanced expression, and BRD4 inhibition has proven to alleviate viral-associated inflammation in this system
[200]. Besides, it has been shown that virus infection in macrophages downregulates BRD3 expression and that BRD3 depletion impairs virus-mediated production of IFN-ß
[206].
The tumorigenesis-associated JAK-STAT pathway also cooperates with NF-ĸB in triggering the immune response
[207]. Through extracellular stimuli (mainly interleukin-6), the membrane receptor-associated Janus kinase (JAK) activates Signal Transducer and Activator of Transcription (STAT) factors to regulate the expression of cytokine-responsive genes
[208]. JAK-STAT activation results in phosphorylated STAT, which enters the nucleus to activate transcription of IFN-stimulated genes. In human pluripotent stem cell-derived cardiac organoids (hPSC-COs), Mills et al. have reported that simulation of the COVID-19-associated CS leads to phosphorylation of STAT1 at S727 site
[8]. On the other hand, BET inhibition has shown to efficiently inhibit the phosphorylation of STAT3
[209]. Moreover, combined inhibition of BET and JAK proteins has been shown to efficiently reverse inflammation linked to bone marrow fibrosis
[210].
Brd4 siRNA delivery through liposome nanoparticles efficiently suppresses RELA and STAT3 activation in LPS-induced mouse models of inflammation
[211].
Another important determinant of the immune response triggered by SARS-CoV-2 is NLRP3, which is well expressed in various cell types such as lung epithelial, kidney, cardiac, endothelial, hematopoietic and innate immune cells
[212]. NLRP3 is the most studied component of the inflammosomes, which are multiprotein oligomers of the innate immune system responsible for the activation of the inflammatory response. Inflammosomes strikingly participate in Caspase 1 activation, which leads to the induction of pyroptosis, a newly introduced type of programmed cell death associated with inflammation
[213]. It has been shown that BRD4 inhibition prevents proliferation and epithelial to mesenchymal transition in renal cell carcinoma by increasing NLRP3 levels, which results in activated Caspase 1 and pyroptosis
[214].
Besides triggering inflammation, SARS-CoV-2 infection, like other respiratory viral infections, is linked to oxidative stress of the epithelium, whose cells activate the transcription factor Nuclear factor erythroid-derived 2-Related Factor 2 (NRF2) for protection against oxidation and inflammation
[215][216]. Notably, it has been described that BRD4 downregulation or BET inhibitors lead to NRF2 stabilization, which results in decreased reactive oxygen species production
[217].
TNF is a well-known mediator of inflammation-associated heart failure, which induces systolic dysfunction
[218]. Indeed, simulated SARS-CoV-2 infection by TNF treatment of hPSC-COs also leads to systolic dysfunction
[8]. Although quite different in origin, atherosclerosis and heart failure have in common the involvement of TNF in mediating the inflammatory response. It has been shown that TNF-mediated inflammation in endothelial cells directs the formation of RELA- and BRD4-dependent SEs, being BET inhibition able to abrogate SEs-derived transcription and atherosclerosis
[219]. These types of SEs are regulated in a highly dynamic way. They are tightly associated with disease
[36] and have been denominated “latent enhancers”, being usually formed in response to noxious stimuli in terminally differentiated cells
[220]. In these conditions they are flooded with transcription-associated proteins, sharing many features with SEs. Once stimulus ceases, most of them remain in a latent state of memory, which enables faster and greater induction by next stimulation. Cardiac hypertrophy promotes great changes in methylation and acetylation of chromatin leading to activation of a great number of enhancers
[221]. In heart failure models, BRD4 occupies the majority of activated enhancers and BET inhibition disturbs associated transcription, suppressing cardiomyocyte hypertrophy
[222].
3.3. BET Inhibition for COVID-19 Treatment
Since organ damage associated with SARS-CoV-2 infection is tightly linked to overactivation of the immune response, a way to avoid damage is to neutralize hyperinflammation. On top of that, blocking virus recognition and entry into host cells will prevent triggering of the immune response, which will also result in aborted inflammation. BET inhibition has a lot to do with both approaches. We have highlighted the fundamental role BET proteins play in the control of the immune response, and we have indicated that expression of the main receptor in host cells for SARS-CoV-2 entry, the ACE2 protein, but also that of the associated TMPRSS2 protease, is under the control of BETs. Thus, BET inhibition, by interfering at different levels of the virus infection, may result in beneficial outcomes when used for treating COVID-19. It has been indicated that ACE2 expression is stimulated by activation of the immune response
[223], and early in SARS-CoV-2 research became clear that antagonizing ACE2 and/or TMPRSS2 could be of interest for COVID-19 therapies
[179][224][225][226][227].
From an unbiased CRISPRi screen to uncover druggable pathways controlling SARS-CoV-2 S protein binding to human cells, Tian et al. have determined that BRD2 is a key player of the cellular response to SARS-CoV-2 infection
[10]. In this screen, they used Calu-3 cells, a lung epithelial cancer cell line that endogenously expresses ACE2. As expected,
ACE2 downregulation was the major cause of impaired S protein binding. However, they also found
BRD2 among downregulated genes leading to decreased S protein binding. In fact, they showed that downregulation of
BRD2 correlated with lower levels of
ACE2 transcript and thereby of protein. Overexpression of the full BRD2 protein recovered
ACE2 transcriptional levels, demonstrating that BRD2 is required for ACE2 expression. Furthermore, it was shown that downregulation of
BRD2 produced a complete inhibition of viral replication in these cells, showing levels similar to those observed when
ACE2 was downregulated. The use of BET inhibitors produced effects similar to those observed upon
BRD2 downregulation, with reduced
ACE2 mRNA levels and S protein binding. Decreased
ACE2 mRNA levels was observed in both primary human bronchial epithelial cells and cardiomyocytes. Viral replication was also affected by BET inhibitors, and in a similar way to that observed when downregulating
BRD2 or
ACE2. In addition, BET inhibition led to marked downregulation of genes that are involved in the response to type I IFN, whose expression is induced by SARS-CoV-2 both in patients and in cell cultures
[10]. Therefore, these results suggest that BRD2 could be used as a therapeutic target for the treatment of COVID-19.
Moreover, SARS-CoV-2 infection is known to cause cardiac damage and dysfunction in 20–30% of hospitalized patients
[228] and in the absence of infection well known inflammatory mediators such as TNF are associated with heart failure
[218]. Mills et al. used hPSC-COs models, phosphoproteomic studies and single nuclei RNA sequencing to identify therapeutic targets and treatments for cardiac dysfunction. They studied the effects of several proinflammatory cytokines that are increased in COVID-19 patients and observed that they produced cardiac dysfunction, being TNF associated with systolic dysfunction and combination of IFN-γ, IL-1β and poly(I:C) with diastolic dysfunction, which is one of the most common dysfunction observed in COVID-19 patients
[229]. The cardiac CS produced by IFN-γ, IL-1β and poly(I:C) induced 91 phosphosites, including one site on STAT1 and two sites on BRD4. Specific inhibitors exist for both proteins. However, while the different treatments used to inhibit STAT1 phosphorylation did not prevent CS-induced diastolic dysfunction, of several BET inhibitors tested (INCB054329, JQ1, RXV-2157, apabetalone and ABBV-744), the first four showed protection. This study demonstrated that CS-mediated diastolic dysfunction is mediated by BRD4-dependent mechanisms that can be blocked using BET inhibitors. Also in mice, they showed that response in the heart triggered by SARS-CoV-2 infection was partially blocked by treatment with INCB054329. Pre-incubation of hPSC-COs with INCB054329 prior to infection reduced ACE2 expression and decreased intracellular viral RNA, demonstrating the potential of BET protein inhibitors to block SARS-CoV-2 infection and prevent dysfunction
[8]. BET inhibitors with dual BDI and BDII activities may display side effects
[43], making necessary to determine the selectivity of the inhibition. Importantly, molecules specifically inhibiting BDII, like RXV-2157 and apabetalone, efficiently blocked SARS-CoV-2 infection by decreasing ACE2 expression and thereby SARS-CoV-2 S protein binding, showing that selective inhibitors against BDII are potential candidates to prevent heart damage caused by COVID-19
[8].
In line with this, a more recent study carried out by Gilham et al., shows that apabetalone, as JQ1, produces downregulation of
ACE2 expression in different cell types: Calu-3 cells, Vero E6 cell (monkey kidney epithelial cells), hepatocarcinoma cells HepG2 and Huh-7, and primary human hepatocytes
[7]. Decrease in mRNA levels is accompanied by a decrease in protein levels in Calu-3 and Vero E6 cells. Apabetalone treatment also decreased
DPP4 expression in Calu-3 cells.
DDP4 encodes for Dipeptidyl peptidase 4 (CD26), a potential cofactor for SARS-CoV-2 entry into the host cell, as its presence on the cell surface facilitates viral binding
[230][231]. Apabetalone and other BET protein inhibitors attenuate SARS-CoV-2 S protein binding and abrogate SARS-CoV-2 infection. These results and its well-established safety profile, together with the dual mechanism of action simultaneously combating hyperinflammation
[232][233][234][235] and ACE2-mediated viral entry, make apabetalone a good candidate for treating SARS-CoV-2 infection
[7]. Indeed, a clinical trial with apabetalone for COVID-19 treatment has been approved: NCT04894266 identifier (
https://clinicaltrials.gov/, accessed on 22 July 2021).
Additional studies have reinforced the interest on BET targeting as an effective tool against SARS-CoV-2 infection. Expression of the key host proteins ACE2 and TMPRSS2 mediating virus entry in the cell is regulated by androgens
[9][236] and transcriptional repression of the androgen receptor (AR) enhanceosome with AR or BET inhibitors suppressing SARS-CoV-2 infection in vitro
[9]. Studies have shown that
AR, ACE2 and
TMPRSS2 are co-expressed in various types of human and murine lung epithelial cells, including alveolar and bronchial cells. Adult immune-competent C57BL/6 male mice were castrated to create an androgen-deprived condition and compared with non-castrated mice and with castrated mice treated for 5 days with testosterone. The experiments showed that
Tmprss2 and
Ace2 are positively regulated by androgens
[9]. In the AR-positive LNCaP prostate cancer cells, which can be infected by SARS-CoV-2, it has been shown that blocking AR signaling with the use of different AR antagonists approved for prostate cancer treatment affects the infective capacity of SARS-CoV-2, showing a dose-dependent decrease in the expression of
TMPRSS2 and
ACE2. Likewise, the use of different BET protein inhibitors showed decreased expression of
TMPRSS2 and
ACE2, suggesting once again that BET proteins play a role in regulating the expression of the SARS-CoV-2 entry factor ACE2 and that this is independent of AR regulation, supporting the idea that inhibition of BET proteins may be useful to mitigate SARS-CoV-2 infection
[9].
SUPT16H protein together with SSRP1 constitute FACT (FAcilitates Chromatin Transcription), a heterodimeric histone chaperone associated with chromatin remodeling during gene transcription
[237]. BRD4 stabilizes SUPT16H by recognizing SUPT16H acetylation at K647
[238]. Targeting SUPT16H by RNAi-based approaches or pharmacological inhibition leads to the induction of interferons and interferon-stimulated genes, efficiently inhibiting SARS-CoV-2 infection, but also infection by other viruses like Zika and influenza
[238]. This raises the question whether the effects of BET inhibitors on SARS-CoV-2 infection are mediated, at least in part, by altered SUPT16H stability. Besides BET proteins, another critical druggable target identified in SARS-CoV-2 interactome is mTOR
[6]. It has been shown that the dual BET/mTOR-PI3K-α SF2523 inhibitor effectively blocks SARS-CoV-2 replication in lung bronchial epithelial cells in vitro
[239]. Moreover, synergistic effects are observed when SF2523 is combined with remdesivir. Additional molecules like flavonoids and allergen fragrance molecules have been proposed as compounds of interest to interfere with SARS-CoV-2 infection through BET inhibition
[240][241].
In sum, BET inhibition for beneficial effects on SARS-CoV-2 infection seems to operate at different levels (
Figure 3). By targeting BETs, different processes/pathways relevant for virus infection can be simultaneously interfered. As explained, BET inhibition should result in attenuated inflammation but also in decreased ACE2 expression, which will result in hampered infection, collectively leading to reduced tissue damage. Besides, direct targeting of host proteins by viral proteins may also have consequences but it is an additional piece of the landscape at best. In the case of E protein interacting with BETs we do not know at present whether interaction leads to significant impaired BET function. This seems not to be the case, as Tian et al. have reported that E protein overexpression has mild and non-overlapping effects on transcriptome in comparison with the effect of
BRD2 knockdown or BET inhibition
[10]. In the case that the bromodomains mediate the interaction with E protein we can anticipate that the use of BET inhibitors may be detrimental for the cell, as inhibitors can enhance E protein-mediated interference at bromodomain level. However, we can speculate about the use of well-defined concentrations of BET inhibitors resulting in E protein dissociation without grossly affecting chromatin attachment of BET proteins, thus resulting in a benefit for the cell. Nevertheless, as demonstrated, the use of BET inhibitors has proven to be an efficient tool in fighting SARS-CoV-2 infection effects for many other reasons.
Figure 3. Effects of BET inhibition on SARS-CoV-2 infection. On one hand, expression of the SARS-CoV-2 receptor ACE2 depends on BET proteins. On the other hand, virus infection triggers the immune response, leading to BET-dependent activation of inflammation, which in turn may also activate ACE2 expression. Uncontrolled inflammation may cause severe tissue damage. Besides, SARS-CoV-2 E protein interacts with BETs, but no associated effects have been reported to date. The use of BET inhibitors (BET-i) attenuates ACE2 expression and counteracts inflammation, thus, reducing infection and tissue damage.