*7.2. Retroviridae: HIV*

The human immunodeficiency virus-1 (HIV-1) genome consists of two identical single stranded RNA molecules and it is the causative agent of acquired immune deficiency syndrome (AIDS). HIV infection has become a clinically manageable disease since the development of combination antiretroviral therapy (ART). Globally, 1 million people died from HIV/AIDS in 2016; without ART, more than twice as many people would have died from this disease [140]. In 2020, the United Nations Programme on HIV/AIDS (UNAIDS) reported that of all people with HIV worldwide, 66% were virally suppressed [141,142]. However, emerging drug resistance and limitations in access and adherence to ART impose a threat to controlling the spread of the virus [143]. Current drugs do not eradicate the virus, making lifelong treatment necessary [143].

Several reports from in vitro and in vivo models, and HIV infected subjects, have linked HO-1 with HIV replication and its effects on HIV mediated neurodegeneration. In cultured monocytes and HIV infected mice, hemin efficiently inhibited viral replication; the effect observed in vitro was mediated by HO-1 enzymatic activity [144]. Using an in vitro model of HIV mediated neurotoxicity, HIV infection dysregulated the macrophage antioxidant response and reduced levels of HO-1. Restoration of HO-1 expression in HIV-infected

monocyte derived macrophages (MDM) reduced neurotoxin release without altering HIV replication [145]. In HIV infected subjects, HO-1 protein levels were reduced in the dorsolateral prefrontal cortex, which correlated with central nervous system (CNS) viral load markers of immune activation. In a model of human astrocytes treated with IFNγ, an HIV associated CNS immune activator, HO-1 was degraded by the immunoproteasome [146]. Additionally, the use of CoPP reduced HIV-MDM glutamate release and neurotoxicity, suggesting a role for HO-1 in HIV associated neurocognitive disorders pathogenesis [147].

Altogether, these reports propose HO-1 induction as a protective mechanism against HIV infection. In particular, the HO-1's classical functions mediate its antiviral properties against HIV.

## *7.3. Filoviridae: EBOV*

Ebola virus (EBOV) is a negative-sense RNA virus that belongs to the Filoviridae family [148,149]. From 2013 to 2015 there was an important outbreak in West Africa, which caused >25,000 infections and >10,000 deaths. The average EBOV case fatality rate is ~50% and case fatality rates has ranged from 25% to 90% in past outbreaks [15]. Hemin treatment significantly reduced EBOV replication and delayed pathogenesis in vitro, by stimulating the cellular innate response against the infection [150,151]. In MDM, hemin treatment inhibited EBOV infection in a dose dependent manner. A similar effect was observed in other cell lines, such as HeLa and human foreskin fibroblasts cells, in which hemin also reduced viral replication [150]. Furthermore, it has been reported that the ebola virus protein 35 (VP35) is a critical protein involved in the inhibition of IRF3 phosphorylation as a mechanism that might counteract the antiviral response [152]. Thus, considering that HO-1 promotes IRF3 phosphorylation and activation, its induction may represent a novel therapeutic strategy against EBOV infection.

The cited studies place HO-1 as a novel therapeutic target against EVOB infection. Notably, HO-1's noncanonical functions are involved in the present example of antiviral action.

## *7.4. Hepatic Viruses: HCV and HBV*

Hepatitis C virus (HCV), a single stranded positive sense RNA virus, is associated with chronic hepatitis, cirrhosis, steatosis and hepatocellular carcinoma [153]. HCV treatment includes a combination of pegylated IFN-α and ribavirin, which has low efficacy and important side effects. With the development of direct-acting antiviral agents, such as sofosbuvir and simeprevir, patient outcomes have greatly improved; however, the disease remains a concern [154]. Since HCV infection generates oxidative damage to hepatocytes, the modulation of HO-1 expression emerges as an attractive therapeutic approach to reduce chronic liver disease. Abdalla et al. observed lower HO-1 mRNA and protein levels in HCV infected patients' livers, while this alteration was not found in patients with other chronic liver diseases. The authors also reported HO-1 downregulation in hepatocyte cell lines expressing the HCV core protein [155]. Further, the overexpression or hemin-induction of HO-1 in the hepatoma cell line Huh7 decreased HCV replication and conferred protection against oxidative injury [156]. This effect of HO-1 on HCV replication might be explained partly by the iron dependent inactivation of the HCV RNA polymerase NS5B [157], and by the BV mediated inhibition of HCV NS3/4A protease and induction of an antiviral response by IFNα2 and IFNα17 [153,158]. Moreover, overexpression of miR-let-7c, which interferes with the production of proinflammatory cytokines in osteoarthritis and rheumatoid arthritis synovial fibroblasts [159], can reduce HCV replication by targeting HO-1 transcriptional repressor Bach1 [160].

Hepatitis B virus (HBV) is a DNA virus that causes serious liver diseases, representing the most common etiological agent for these pathologies [161]. It has been shown that pharmacological and genetic HO-1 overexpression attenuates viral replication both in vivo and in vitro in HepG2 cells [161–163], while also playing a hepatoprotective role [162]. The effect of HO-1 induction using hemin and CoPP mitigated the effects of HBV replication [6,161,162]. On the other hand, blocking HO-1 by siRNA reversed the inhibition of viral replication [6]. Interestingly, Protzer et al. evaluated the effect of HO-1 on HBV core protein by pulse-chase metabolic labeling experiments finding that HO-1 can destabilize structural proteins to prevent the formation of viral capsids, highlighting a direct HO-1 antiviral mechanism rather than limiting its effect to its anti-inflammatory properties [6].

Hence, the summarized reports established the antiviral effects of HO-1 by impairing HCV's and HBV's replication.

#### *7.5. Arbovirus: DENV and ZIKV*

Dengue virus (DENV) is a single stranded positive sense RNA virus [164] that induces oxidative stress by the activation of inflammatory regulators, such as NF-κB, and leads to the progression and pathogenesis of DENV [165]. In this pro-oxidant scenario, Tseng et al. demonstrated that HO-1 promoter activity and protein synthesis gradually increased during the early stages of DENV infection (6 to 12 h), but they were markedly decreased at later stages (24 to 72 h) [166]. Strikingly, pharmacological and genetic HO-1 induction after infection impaired viral protein synthesis and replication, and reduced DENV mortality. This effect was due to BV but not CO nor Fe2+ production [166]. The authors demonstrated that BV inhibits NS2B/NS3 DENV protease, thus promoting the antiviral IFN response and impairing its blockage by this protease [166]. Accordingly, Su et al. showed the anti-DENV activity of miR-155, which inhibits Bach1, a protein that negatively regulates the expression of many oxidative stress-response genes, including *HMOX1* [167]. This, in turn, results in the induction of HO-1, boosting the IFN responses against DENV replication by the activation of interferon induced protein kinase R (PKR), 2'-5'-oligoadenylate synthetase 1 (OAS1), OAS2, and OAS3 expression [167]. Interestingly, the summarized studies demonstrate that HO-1's antiviral effects against DENV infection involve both, its canonical, in this case mediated by BV, and noncanonical functions.

Zika virus (ZIKV), a single stranded positive sense RNA, is the causative pathogen of Zika fever in humans [168]. Using A549 and embryonic kidney (HEK-293) cell lines, El Kalamouni et al. demonstrated that ZIKV infection downregulated HO-1 expression by triggering endoplasmic-reticulum-associated protein degradation, thus halting its antiviral effects [168]. This report highlights HO-1's protective role relevance, as it demonstrates that ZIKV infection promotes the decrease in HO-1 expression levels as an evasion mechanism.

#### *7.6. Neurotropic Viruses: HSV-2 and EV71*

Herpes simplex virus (HSV) includes HSV-1 and HSV-2, two double stranded DNA viruses that belong to Herpesviridae family. HSV produces recurring lesions in skin and mucosae and can also latently infect neurons of the trigeminal or dorsal root ganglia. HSV infection can result in encephalitis and meningitis [169]. Ibañez et al. demonstrated that pharmacological induction of HO-1 by CoPP hampered HSV-2 propagation in epithelial and neuronal cells. Furthermore, by CORM-2 treatment the authors also showed that the effects of HO-1 were partly mediated by CO [170].

Enterovirus 71 (EV71) is a single stranded positive sense RNA virus that belongs to the Picornaviridae family and is the causative agent of hand foot and mouth disease in children [171]. It was demonstrated that the overexpression of HO-1, as well as CO treatment, decreased viral replication in SK-N-SH cells suggesting that the antiviral effect is mediated by the downregulation of EV71 induced ROS levels [172].

Regarding neurotropic viruses, the summarized reports showcase that HO-1 displays protective effects against HSV-2 and EV71 involving its enzymatic function.

#### *7.7. COVID-19 Causative Agent: SARS-CoV-2*

SARS-CoV-2, is the novel beta coronavirus of the Coronaviridae family whose genome is composed of a single stranded RNA molecule [173]. It has been shown that hemin, hemoglobin and protoporphyrin IX bind to SARS-CoV-2 proteins, blocking its adsorption and replication independently from HO-1 induction [174]. However, the current literature regarding HO-1's antiviral effect against SARS-CoV-2 remains unclear. Maestro et al. showed that hemin does not inhibit SARS-CoV-2 viral replication in vitro [175]. Kidney epithelial Vero-E6 and lung Calu3 cell lines were treated with hemin and results showed that, despite a strong activation of HO-1 in both cell lines, there was no effect on SARS-CoV-2 viral replication, measured by the amplification of the N viral gene by RT-qPCR [175]. However, a more recent report proposed hemin as a potential drug for treating COVID-19 via HO-1 induction [176]. Interestingly, authors observed a reduction in SARS-CoV-2 replication, both when pretreating and after SARS-CoV-2 infection treatment of Vero76 with this drug. Genetic induction or silencing of HO-1 in Vero76 cells demonstrated that the antiviral effect of hemin relies on this protein. Strikingly, this effect was mediated not only by Fe2+ and BV, but also by an HO-1 enzymatic independent mechanism. Further, they showed that hemin induced HO-1 boosted ISG15, OAS1 and MX1 protein expression in SARS-CoV-2 infected cells, highlighting the importance of stimulating the host cell's IFN response against this virus [176]. Of note, there are reports from our laboratory showcasing that *MX1* gene expression is increased in COVID-19 patients. However, *MX1* expression is lower in elderly patients, where the disease has been shown to be more severe than in younger people. Additionally, through an in depth proteomics analysis, we described MX1 as a novel HO-1 interactor in prostate cancer (PCa) cell lines [177]. Moreover, genetic and pharmacological HO-1 induction in PCa cells triggered an increase in MX1 at mRNA and protein levels, and altered HO-1 cellular localization, showcasing a clear association between both proteins. Further, *MX1* silencing with a specific siRNA significantly decreased the expression of ERS genes (*HSPA5*, *DDIT3* and *XBP1*), demonstrating the role of MX1 in pro-death events [177].

In summary, the induction of the host infected cells antiviral response appears to be critical for COVID-19 treatment, which could be partly achieved by hemin mediated HO-1 induction, also preventing viral adsorption and replication by binding SARS-CoV-2 proteins. These antiviral effects are mediated by canonical and noncanonical HO-1's functions.

#### **8. HO-1 Induction as a Strategy against COVID-19**

There are mainly two different approaches to develop antiviral therapies: (1) therapies directed against viral factors; or (2) therapies targeting the host immune system. To date, the second strategy has received increasing attention due to the fact that targeting viral factors might cause viruses to mutate, increasing the rate of resistance to antiviral drugs [178]. In contrast to the viral genome, host cells' DNA does not have a high mutational frequency. Therefore, overpowering viral infection by targeting host factors involved in the antiviral response is conceivably an effective strategy to counteract the severe consequences, while also fighting the infection [179].

During the last two years, several reports have focused on the understanding of the virus–host interaction underlying COVID-19 disease. The worrying situation of the SARS-CoV-2 pandemic and the threat of new variants, such as Omicron, which is spreading across the globe at an unprecedented rate, drive the interest of scientists to seek for new anti-SARS-CoV-2 strategies. Its enhanced transmissibility compared to the Delta variant could be explained in part by its increased rate of replication in human primary airway cultures, higher binding to ACE2, and ability to efficiently enter cells in a TMPRSS2-independent manner [180]. Fortunately, preliminary data of the Omicron variant suggest a lower virus load in both lower and upper respiratory tract, associated to less inflammatory processes in the lungs, using a mouse model of severe disease [181]. However, exceptionally high transmissibility could result in a great burden on healthcare systems across the globe. In this context, HO-1 emerges as a potential target to boost the host's response to fight the infection and prevent severe COVID-19.

Certainly, HO-1 and its reaction products possess beneficial effects for the host during viral infections: it reduces inflammation and exerts antiviral actions. The most serious COVID-19 complications are: sepsis like inflammation, coagulopathy, and cardiovascular or respiratory complications. Furthermore, respiratory failure triggers hypoxia which, in combination with neuroinflammation, generates neurological complications [182]. When

inflammation is not modulated, it turns into hyperinflammation and results in tissue damage or organ failure [183]. Enhancing HO-1 expression might help avoid the severe consequences of this disease. For example, it has been reported that HO-1 induction decreases inflammation, inhibits platelet aggregation, and increases fibrinolysis and phagocytosis, thus preventing tissue damage, thrombosis and sepsis [184]. Additionally, hemin is an activator of neuroglobin, a protein involved in oxygen transport and storage in neurons that increases oxygen's intracellular partial pressure in neurons, and is crucial to protect neurons from hypoxic injury [185–187]. In addition, as mentioned above, HO-1 has a reported antiviral activity against multiple viruses. This effect depends on its classical activity involving its reaction subproducts (BR, BV, CO and Fe2+) and the activation of the IFN pathway; interestingly, its noncanonical activity is also involved in the antiviral effect of HO-1.

HO-1 expression is essentially regulated at the transcriptional level by NRF2. It has been reported that SARS-CoV-2 infection suppresses the NRF2 antioxidant gene expression pathway, and that NRF2 agonists limit viral replication and repress the proinflammatory response of SARS-CoV-2 [188]. This evidence highlights the relevance of the NRF2 signaling pathway on the antiviral response, suggesting that the activation of NRF2 might be a useful strategy against COVID-19 [189].

As explained before, clinical complications associated with COVID-19 disease have been described in different organs, including vascular, cardiac, renal, hepatic, endocrine and neurological complications [190] (Figure 4). Interestingly, HO-1 has been reported to be associated with a reduction in tissue damage, mainly through its anti-inflammatory and antioxidative functions in different organs [4,97,147,191–207] (Figure 3). It would be interesting to address HO-1's vasoprotective and antithrombotic effects for the prevention of thromboembolic events caused by SARS-CoV-2.

**Figure 4.** HO-1's role in different sites that can be affected upon SARS-CoV-2 infection. Extra pulmonary manifestations of COVID-19 are grouped according to their site or body system. HO-1's reported functions in different experimental conditions or diseases are grouped according to the model or system in which they are studied. The image of the human body has been adapted from Uhlén et al. (Human Protein Atlas, proteinatlas.org) [208].
