The antiviral and immunomodulatory activities of propolis: An update and future perspectives for respiratory diseases

Abstract Propolis is a complex natural product that possesses antioxidant, anti‐inflammatory, immunomodulatory, antibacterial, and antiviral properties mainly attributed to the high content in flavonoids, phenolic acids, and their derivatives. The chemical composition of propolis is multifarious, as it depends on the botanical sources from which honeybees collect resins and exudates. Nevertheless, despite this variability propolis may have a general pharmacological value, and this review systematically compiles, for the first time, the existing preclinical and clinical evidence of propolis activities as an antiviral and immunomodulatory agent, focusing on the possible application in respiratory diseases. In vitro and in vivo assays have demonstrated propolis broad‐spectrum effects on viral infectivity and replication, as well as the modulatory actions on cytokine production and immune cell activation as part of both innate and adaptive immune responses. Clinical trials confirmed propolis undeniable potential as an effective therapeutic agent; however, the lack of rigorous randomized clinical trials in the context of respiratory diseases is tangible. Since propolis is available as a dietary supplement, possible use for the prevention of respiratory diseases and their deleterious inflammatory drawbacks on the respiratory tract in humans is considered and discussed. This review opens up new perspectives on the clinical investigation of neglected propolis biological properties which, now more than ever, are particularly relevant with respect to the recent outbreaks of pandemic respiratory infections.

used in conjunction with the Baltimore classification system, which defines seven groups on the basis of the viral genome (single-or double-stranded DNA, single-stranded sense or antisense RNA, and double-stranded RNA) and the mechanism of messenger RNA (mRNA) production (with or without the aid of reverse transcriptase). 40 Historically, the first tenet of cell theory stated that living organisms consist of cells. However, viruses are sometimes considered noncellular living entities, even though they are not capable of autonomous reproduction without relying on cellular machinery to be copied. Whether a virus should be considered a living organism or a replicator at the edge of chemistry and life, it may be simply defined as "a piece of bad news wrapped in protein," to quote the British biologist and Nobel Prize winner Sir Peter Medawar. Indeed, viruses are responsible for a wide variety of human diseases. Viral infections provoke in animals an innate and an adaptive (humoral and cell-mediated) immune response. The innate immune system detects and responds to pathogens in a nonspecific manner to defend the infected host cells. Viral recognition generally occurs in two different ways through the detection of specific molecular signatures: the recognition of pathogen-associated molecular patterns (PAMPs), which are distinct molecular features of the viral particles, via pattern recognition receptors (PRRs), and the detection of cellular damage or stress induced by viral infection. As a result of PRR engagement, pro-inflammatory cytokines, mainly downstream of NF-κB activation, and type I interferons (IFNs) are induced. INF-α and IFN-β, collectively referred to as type I IFNs, are the major effector cytokines orchestrating the response of the host against viral infections. 41 Additionally, type I IFNs link innate and adaptive immune responses, enhancing dendritic cell maturation, natural killer cell cytotoxicity, and differentiation of virus-specific cytotoxic T lymphocytes. 42 A second outcome is the inflammasome-mediated activation of caspase-1, which can cleave multiple substrates including pro-interleukin (IL)−1β. Both PRR-induced pathways can also initiate apoptosis in the attempt to prevent viral replication and spread. 43,44 One of the best-characterized mechanisms for PRR activation is the recognition of viral nucleic acids, either viral genome or replication intermediates, which take place both at endosomal and cytosolic levels. Many toll-like receptors (TLRs) able to detect PAMPs from several types of pathogens are expressed in endosomes and detect viral genomes upon endocytosis, triggering type I IFN expression. 45,46 TLR3, originally identified as a sensor of dsRNA viruses, 47 activates dendritic cells after the phagocytosis of infected cells, 48 and plays a role in protecting the central nervous system against herpes simplex virus infection. 49,50 TLR7/8 and TLR 9 recognize ssRNA and dsDNA viral genomes, respectively, and stimulate pro-inflammatory cytokine and type I IFN expression. [51][52][53] On the other hand, cytosolic sensors include families of structurally related receptors: RIG-I-like receptors, which are sensors of RNA and induce the expression of type I IFNs; AIM2-like receptors, which are sensors of DNA and elicit inflammasome activation; NOD-like receptors, which are sensors of viral PAMPs and virus-induced cellular stress, and cause either IFN expression or inflammasome activation. 43,54 Cytokine production may be induced by the viral infection (primary cytokines) or be a consequence of the immune response (secondary cytokines). Although it is difficult to discriminate an intense pro-inflammatory response due to severe infection from a dysregulated cytokine response, 55 abnormally elevated levels of inflammatory cytokines are often referred to as "cytokine storm," a condition that typically accompanies certain viral infections. For example, a significant number of deaths from influenza occurs due to cytokines despite early antiviral therapy. Given the role of inflammation in the pathogenesis of these diseases, cytokines may be one of the most critical targets for an immunomodulatory approach to viral infections. 56 In this context, besides the direct role in virus clearance, also type I IFNs may be of pivotal importance exerting an anti-inflammatory activity through the induction of IL-10 production. 57 Many natural products, including propolis, contain high amounts of polyphenolic compounds characterized by antioxidant properties and able to inhibit inflammatory cytokine production as well, mainly through the impairment of the transcriptional activity of NF-κB. 58 The use of such natural products might enhance endogenous host defenses in a nonspecific manner and modulate inflammatory processes, thus qualifying propolis as a possible prophylactic or therapeutic approach to infectious diseases.
The principal targets of propolis action, represented by the pathogenic mechanisms described in this section, both related to the viral replication cycle and the innate and adaptive responses to viral signatures, are summarized in Figure 1.

| Research question
The aim of this review was to collect the existing in vitro, ex vivo, in vivo, and clinical evidence of the antiviral and immunomodulatory activities of propolis, to encourage the carrying out of rigorous clinical trials and help substantiate a rational therapeutic usage of this natural product against viral diseases, with a particular focus on the applications in respiratory affections. Whenever possible, a critical commentary of the collected evidence was offered, to unveil correlations between certain biological activities and specific propolis types or subtypes, dissecting antiviral and immunomodulatory mechanisms to clarify the possible modes of action and draw attention to the most promising products and lead compounds.

| Data sources and search criteria
Articles were searched using Web of Science, Scopus, Embase, and PubMed databases, without any language restriction, using the following search terms: "propolis" AND ("antiviral" OR "virus" OR "influenza" OR "immunomodulation" OR "immunomodulatory" OR "immune" OR "inflammation" OR "respiratory" OR "airways"). The search (title, abstract, and keywords) reported 1050 items on Web of Science, 1251 on Scopus, 844 on Embase, and 730 on PubMed, indexed until July 2021. Duplicate articles were removed. In a second step following study F I G U R E 1 Schematic representation of the pathogenic mechanisms underlying viral infections, which represent the principal targets of propolis action. dsDNA, double-stranded DNA; IFN, interferon; IL, interleukin; NK, natural killer; ssRNA, single-stranded RNA [Color figure can be viewed at wileyonlinelibrary.com] selection, the references listed in the retrieved research articles and reviews were sifted through, to identify documents that might had eluded the primary search.

| Eligibility criteria
Preclinical and clinical studies dealing with the effects of propolis produced by honeybees (Apis mellifera Linnaeus 1758) on immune mechanisms, respiratory system pathophysiology, and viruses relevant for human pathologies were considered eligible. Documents in languages different from English were included if they were published in Italian, or if an abstract in English was available and the abstract contained a sufficiently detailed description of research methods and results.

| Study selection
Full titles and abstracts of the documents retrieved in the primary search were assessed for adherence to the eligibility criteria. Subsequently, full texts (except the cases in which the full text was unavailable with the means at our disposal) of the eligible articles were carefully read and checked for inclusion.

| Data collection
Information relevant to the research question (in particular: type of propolis, study model and biological parameters, and statistically significant results) were deducted through the careful reading of abstracts and full texts of the articles. The methodological quality of the clinical trials reported in the antiviral activity section was assessed using the algorithm proposed by Jadad et al. 59 3 | ANTIVIRAL ACTIVITY Many in vitro, in vivo, and clinical studies have highlighted the broad spectrum of antiviral activities which several types of propolis share against numerous viral families responsible for relevant human diseases. Still, in many cases, the nature of these effects is yet to be fully unraveled, and the understanding of their significance is complicated by the lack of consistency between clinical and preclinical studies.
Few works considered the overall clinical efficacy of propolis extracts against respiratory tract infections, a number of diseases caused in most cases by viruses with possible bacterial and fungal superinfections. In a case-control study carried out in pre-school and school children, Crişan et al. 60 evaluated the action of a proprietary aqueous flavonoid-rich extract (NIVCRISOL) in acute inflammatory diseases of the upper respiratory tract, such as common cold, generally caused by rhinoviruses and, to a less extent, human coronaviruses, influenza viruses, or adenoviruses. The monitoring consisted of the recording of the incidence of acute rhinopharyngitis symptoms and in the periodical examination for the determination of viral burden. The results demonstrated a favorable effect of propolis local treatment in lowering the number of symptomatic cases and decreasing, and sometimes suppressing, the microbial flora of the upper airways. 60 Since the authors specified neither the composition of the microbial flora nor the reduction of specific viral titers, an antiviral effect of propolis, albeit not unlikely, may only be hypothesized, along with a probable immunomodulatory and disease-modifying activity. Cohen et al. 61 conducted a randomized, double-blind, and placebo-controlled clinical trial to evaluate the safety and effectiveness of an herbal preparation (Chizukit) containing propolis extract (50 mg/ml), an extract of Echinacea purpurea (L.) Moench aerial parts and Echinacea angustifolia DC. roots, and vitamin C, in preventing respiratory tract infections in children. The study demonstrated that the treatment significantly decreased the number of children who experienced one or more respiratory tract illness episodes during the 12 weeks of the study, the total number of episodes, and the mean number of episodes per child (primary outcome). The total number of illness days and the duration of individual episodes were also significantly lower. Moreover, the treatment reduced the number of days of fever and the use of antipyretics, the incidence of rhinitis, and daytime and nighttime cough (secondary outcome). 61 Once again, no mechanism against a specific viral species was reported, and besides, propolis was associated with other natural products, thus impeding the unequivocal attribution of a causal relationship. Szmeja et al. 62 clinically tested the therapeutic value of flavonoid-rich Canadian propolis in rhinovirus infections. The treatment shortened by 2.5-fold the duration of the disease, with the regression of symptoms starting from the first day of therapy and the complete recovery within 3 days in the treated group, in contrast with an average of 4.80 days in the placebo group. 62 Esposito et al. 63 performed a randomized, double-blind, placebo-controlled clinical trial to evaluate the efficacy of an oral spray, formulated with a poplar-type propolis proprietary extract (M.E.D. ® ) standardized through the evaluation of six markers (apigenin, chrysin, galangin, pinobanksin, pinocembrin, and quercetin) which represent the 25% of the total polyphenols indicated in the titration, on the remission of symptoms (sore throat, muffled dysphonia, and swelling and redness of the throat) associated with mild uncomplicated upper respiratory tract infections. The results demonstrated that after 3 days of treatment with propolis, 83% of subjects had remission of all symptoms, while 72% of subjects in the control group had at least one remaining symptom, indicating that propolis led to prompt resolution with an advance of 2 days. 63 This evidence may be related to a newly demonstrated physical mechanism of propolis action consisting in the generation of an extensive "exclusion zone" water layer on propolis-coated surfaces, which constitutes an effective barrier able to hinder microbial access to the potential site of infection and therefore inhibit viral entry. 64 Finally, Di Pierro et al. 65 investigated the role of another proprietary extract (Proposoma-lisclatrato ® ), a mixture of phytosome and propolis co-ground in a ratio 1:1, in an open-label, retrospective, controlled clinical analysis, conducted in children with acute otitis media or viral pharyngitis, generally caused by paramyxoviruses, rhinoviruses or adenoviruses. The oral administration of propolis reduced the severity of symptoms, avoided the use of antipyretics and anti-inflammatory drugs, and decreased the progression to tracheitis, bronchitis, and rhinosinusitis. 65 In view of this, it may be easily inferred that the pharmaceutical formulation of propolis extracts used in clinical studies seems not a negligible factor. In their study, in fact, Drago et al. 66 compared in vitro the antiviral activities of another proprietary product, Actichelated ® propolis (a blend of active substance, carrier, and biocatalyst subjected to mechano-chemical activation), and of a hydroalcoholic extract of propolis collected in Argentina and Uruguay, showing that the active antiviral concentrations of Actichelated ® propolis against different strains of adenoviruses, influenza viruses, parainfluenza viruses, and herpes simplex virus type 1 were at least ten times lower than those of the hydroalcoholic extract.
As discussed above, propolis exerts a clinically plausible antiviral activity, despite the existence of confounding factors such as the type of propolis extract, the route of administration, or the pharmaceutical formulation.
However, due to the variability in the etiology of these diseases, only a favorable role of propolis treatment could be demonstrated, without any elucidation of the effects against specific viral species. Nonetheless, a common thread among all these clinical studies is the use of flavonoid-rich propolis, but the contribution of an immunomodulatory and anti-inflammatory activity to the overall effect may not be excluded.
Many in vitro studies have tried to define the general mechanisms of action, which are not completely understood, at a cellular and molecular level. One of the most critical steps in antiviral innate immunity, for example, is the recognition of viral genome by PRRs which results, among other things, in the induction of type I IFNs and, consequently, of antiviral IFN-inducible genes. Hayakari et al. 67 have demonstrated in a model of viral-like infection that pretreatment of A549 human alveolar epithelial cells with green Brazilian propolis could inhibit poly I:C-induced IFN-β mRNA and protein expression in a concentration-dependent manner and enhance the mRNA expression of antiviral factors. Hence, propolis could increase the expression of antiviral effectors also through MAGNAVACCA ET AL. | 903 IFN-independent pathways. Moreover, propolis prevented poly I:C-induced mRNA expression of the proinflammatory chemokines IL-8 and CCL5, thus hindering the chemotaxis of polymorphonuclear leukocytes exposed to cell-conditioned medium. 67 These results once again highlight the importance of the immunomodulatory and anti-inflammatory activities, which coexist in propolis with the merely antiviral ones.
In the following paragraphs, the known direct effects of propolis on individual viral species are presented. For ease of treatment, viral species are divided according to the Baltimore classification group and taxonomic family. Herpes simplex virus type 1 (HSV-1) and 2 (HSV-2) are the etiological agents of herpes labialis and genital herpes, respectively. Varicella-zoster virus (VZV) causes varicella, a disease usually affecting children, and herpes zoster, which commonly affects adults. Acyclovir, a nucleoside analog able to arrest viral DNA synthesis, is the principal conventional drug used for the treatment of HSV-1/2 and VZV infections. However, adverse effects and the emergence of drug-resistant viral strains demand novel therapeutic agents. When HSV-1 viral particles were pretreated with propolis hydroalcoholic (70% ethanol) extract for 1-3 h before being added to HEp-2 cells, significant antiviral efficacy was observed at a concentration of 10 μg/ml, alone or in combination with acyclovir, with the greatest decrease in viral titers within the first hour. 79 In another study, the antiviral activity of Turkish propolis collected in the Hatay province was assayed against HSV-1 and HSV-2. Viral replication was significantly suppressed in the presence of propolis, as shown by the decrease in viral titers, and the effect was more massive and rapid against HSV-1. Moreover, the combination of propolis and acyclovir displayed a synergistic effect, proving to be more effective than acyclovir alone. 80 These results show the ability of propolis to impair, at least in vitro, the replication of herpes simplex virus through direct antiviral activities.
Shimizu et al. 81  Berg) D. Legrand. The oral administration of propolis extracts to BALB/c mice inoculated with HSV-1 virus delayed the development and progression of herpetic skin lesions in the early phases of infection. Propolis from B. dracunculifolia, that is, green propolis, was significantly effective in reducing viral titers in vivo, in the skin and brain of infected mice, and in vitro, in a plaque reduction assay. Propolis from M. euosma, of which moronic acid is a characteristic constituent, was active in reducing viral titers in vivo only in the brain but was inactive in the in vitro assay. Nonetheless, it enhanced delayed-type hypersensitivity to HSV-1 antigens and increased IFN-γ production from splenocytes of HSV-1-infected mice, suggesting the presence of components active in vivo after oral administration. Propolis from B. erioclada, despite being active in vitro, had no significant effect on viral titers in either skin or brain but significantly enhanced delayed-type hypersensitivity in infected mice and elevated IFN-γ production in splenocytes in vitro. 81  inflammatory and oxidative processes. 82 Altogether, these findings indicate that several Brazilian propolis subtypes seem effective in vivo against both HSV-1 and HSV-2 infection, with slight but significant differences in their modes of action which were not extensively investigated. In search of molecular mechanisms of action, Huleihel and Isanu showed that 0.5% propolis extract determined 50% inhibition of HSV-1 infection in Vero cells, providing indirect evidence of an interaction between propolis extract and cell surface; however, no evidence of a direct interaction with viral particles could be found. Based on these results, the antiviral efficacy of propolis extract against HSV-1 infection may be attributed, at least in part, to the prevention of virus absorption onto the host cells. 83 However, the authors did not state the origin of the propolis used in the study, therefore these conclusions may not apply to formation were determined at 5 µg/ml and 4 µg/ml, respectively. Both extracts showed high antiviral activity against HSV-2 in viral suspension tests, with >99% reduction of infectivity, exerting a direct concentration-and time-dependent activity when viruses were pretreated before incubation with cells. 84 These results demonstrate that flavonoids produce a synergistic, or at least additive, effect together with phenylcarboxylic acids and other phenolic compounds, achieving lower effective concentrations. Nonetheless, flavonoids appear not to be primarily responsible for the activity. Different polyphenols, flavonoids, and phenylcarboxylic acids have been identified as major constituents of the above-mentioned aqueous and GH 2002 extracts, including caffeic acid, p-coumaric acid, benzoic acid, galangin, pinocembrin, and chrysin. Therefore, the antiviral activity of the extracts and their isolated compounds was also evaluated against HSV-1 in RC-37 cells. IC 50 of aqueous and GH 2002 extracts for HSV-1 plaque formation were determined at 4 µg/ml and 3.5 µg/ml, respectively, demonstrating also in this case a higher activity for extracts richer in flavonoids. Both extracts showed high antiviral activity against HSV-1 in viral suspension tests, with >98% reduction of infectivity. Once again, the anti-herpetic activity was only observed when viruses were pretreated before infection. Among the isolated compounds, only galangin and chrysin (a flavonol and a flavone, respectively) displayed some antiviral activity, 85 not clearly explaining the activity of whole propolis extracts and the fact that propolis is almost equally effective in spite of the presence of high or low concentrations of flavonoid. Moreover, these results are partially in contrast with the work of Debiaggi et al. 86 in which also kaempferol, as well as chrysin, proved to cause a concentration-dependent reduction of intracellular replication of herpesvirus strains. However, virus infectivity was not significantly reduced. On the other hand, galangin had no effect on either the infectivity or replication, whereas quercetin was able to reduce infectivity and intracellular replication, but only at the highest concentration tested. 86 The extracts exhibited significantly higher activity and selectivity indices in any case, showing that the observed effects cannot be reduced to the effect of single components, but synergisms play a prominent role. This evidence might find confirmation in the work of Amoros et al. 87 in which, besides the fact that flavonols (e.g., galangin, quercetin, rutin, and kaempferol) resulted to be more active against HSV-1 than flavones (e.g., chrysin, apigenin, and acacetin), binary flavone-flavonol combinations proved to be more effective than individual compounds, and such a synergistic effect could explain why total propolis extracts are more active than isolated components. 87 Demir et al. 88  propolis extract a greater antiviral activity than acyclovir against both HSV-1 and HSV-2, whereas the ethanolic and soya oil extracts were found to be more active than acyclovir only against HSV-2. 88 Notably, the extracts that showed the most promising activity in this study are all cosmetic ingredients which could advantageously be exploited in the formulation of products for topical application.
The efficacy of propolis against herpes viruses has been evaluated also in humans and a consistent number of clinical studies have been conducted on formulations containing GH 2002 extract. Holcová and Hladíková 69 evaluated the efficacy and tolerability of a lip balm against herpes labialis in a double-blind, randomized, three-arm dose-finding study with three concentrations of GH 2002 (0.1%, 0.5%, and 1%). The best results were obtained with 0.5% formulation, which allowed the shortest healing time (3.4 and 5.4 days in the 50th and 90th percentile, respectively), good tolerability, and significant therapeutic results for all the secondary parameters (local pain, itching, burning, tension, and swelling). 69 Then, Arenberger et al. 70  Other studies were conducted using different proprietary extracts. A single-blind, randomized, controlled, multi-center study was undertaken to evaluate the efficacy of Canadian propolis ointment compared with acyclovir in men and women with recurrent chronic genital HSV-2 infection. Propolis was collected in a region rich in Populus spp. trees and extracted with 95% ethanol. The proprietary extract, designated ACF ® (Antiviral Complex of Flavonoids), was not standardized to a certain component. The healing process seemed to be faster in the group receiving the propolis ointment, which appeared to be more efficacious than both placebo or acyclovir in the resolution of genital herpetic lesions and of local symptoms. Moreover, the incidence of bacterial superinfection was reduced by 55%. 72 Bankova et al. 90

| Papillomaviridae
Human papilloma virus (HPV) may cause persistent infections that result in warts or pre-cancerous oral and genital lesions. The aim of the study by Iljazović et al. 74 was to clinically evaluate the efficacy of an association between interferon and a propolis herbal product in the treatment of genital HPV infection. Fifty-five HPV positive women were enrolled in the study and randomly assigned to control group (other therapeutic options, e.g., laser, cryotherapy, and podophyllin) or treatment group (interferon pessaries + pessaries containing 5% propolis extract, Aloe vera (L.) Burm.f. juice, Echinacea purpurea (L.) Moench and Calendula officinalis L.). After 3 months, HPV infection was still present in more than 90% of the subjects in the control group, while had disappeared in 71.42% of the patients in the treatment group and in 100% after 6 months. 74 In this study, however, propolis was formulated with other natural products and associated with interferon, making it impossible to dissect its individual contribution to therapeutic success. Zedan et al. 75 conducted an open-label, single-blinded, randomized, placebo-controlled clinical trial to investigate the effect of the dietary supplementation of propolis as an alternative treatment for cutaneous warts. In the case of flat and common warts, propolis treatment seemed effective in 75% and 73% of cases, respectively. 75 In consideration of the oral administration, an immunomodulatory and disease-modifying effect, rather than a direct antiviral one, cannot be excluded. To date, due to the absence of in vitro studies corroborating possible direct antiviral mechanisms, clinical evidence is insufficient to establish propolis efficacy against HPV.

| Poxviridae
Parapoxvirus is a genus of viruses, mainly zoonotic, responsible for causing orf (ecthyma contagiosum), a viral exanthem, in humans. Zeedan et al. 91 evaluated the efficacy of ethanolic and aqueous Egyptian propolis extracts, administrated via subcutaneous and intradermal injections in albino rats inoculated with parapoxvirus. Propolis determined a nearly 2-3 log decrease of infectivity titers in vitro. In vivo, noninfected animals receiving propolis showed a slight suppression of TNF-α and IFN-γ levels when compared to controls, whereas the cytokine production appeared strongly stimulated in infected rats treated with propolis. The histopathological analysis revealed acute necrotic hepatitis accompanied with disseminated intravascular coagulopathy, which is a pathognomonic process, in infected animals. Conversely, rats treated with propolis appeared normal or presented only mild lesions. 91 The in vitro evidence supports the hypothesis of a direct antiviral activity that may be relevant also in vivo, a context in which, however, it is accompanied by evident immunomodulatory effects.

| Flaviviridae
Dengue hemorrhagic fever is a mosquito-borne zoonosis caused by dengue virus strains (DENV). Soroy et al. 76 evaluated the effectiveness of a water-soluble proprietary extract (Propoelix™, a blend of poplar and Baccharis spp. propolis), rich in caffeic acid phenethyl ester (CAPE) and flavonoids, on the clinical course of patients with dengue hemorrhagic fever. The results of this double-blind, randomized, placebo-controlled trial showed a trend toward faster recovery in platelet counts of treated patients, who had a significantly shorter length of hospitalization.
Moreover, treated patients showed also a significant decline in TNF-α levels, consistent with a possible antiinflammatory and overall disease-modifying effect of propolis extract. 76 This study is supported neither by the determination of viral titers nor by any known mechanism of propolis action against DENV, and a mere immunomodulatory effect on the host should not be excluded.

| Picornaviridae
Human rhinoviruses, assigned to the genus Enterovirus and categorized in three species (A, B, and C), are the major cause of upper respiratory tract infections, accounting for more than 50% of common colds. Rhinoviruses are also associated with severe lower respiratory tract symptoms and exacerbations of chronic pulmonary diseases, as well as fatal pneumonia in elderly and immunocompromised adults. Kwon et al. 92  Silveira et al. 77 conducted an open-label, randomized, controlled trial on hospitalized adult COVID-19 patients to assess the effectiveness of a 7-day treatment with Propomax ® capsules (400 and 800 mg/die, formulated with the standardized green Brazilian propolis extract EPP-AF ® ) in conjunction with standard care. The length of hospitalization was significantly reduced in propolis-treated groups, although propolis did not significantly affect the need for oxygen supplementation. Nevertheless, patients treated with propolis tended to have a reduced need for invasive oxygen therapy. In addition, in patients treated with the higher dose (800 mg/die), a lower rate of acute kidney injury, a common complication of the disease associated with a poor prognosis, was observed. 77 The authors were not able to explicate the mechanisms behind the beneficial effects on COVID-19 patients, however, the antioxidant, immunomodulatory and anti-inflammatory properties of propolis could explain the reduction of the disease impact. reflecting the evidence that whole extract activity appears to be greater if compared to individual components.

|
Nonetheless, Serkedjieva et al. 96 evaluated in vitro the antiviral activity of six synthetic substances, esters of substituted cinnamic acids, which were identical with or analogous to propolis components found in the etheric fraction of a methanolic extract. The authors found that synthetic isopentyl ferulate, a close analog of isopent-3enyl ferulate present in propolis, significantly suppressed the replication of influenza A/Hong Kong/1/68 (H3N2) virus and the production of viral hemagglutinins in vitro and in ovo. The compound was much less active against influenza A/PR/8/34 (H1N1) virus. Effective concentration ranges of isopentyl ferulate and propolis extract etheric fraction were almost equal, and the maximal effect was observed when the substance was present in the culture medium during the whole infectious process. 96 These findings demonstrate that the antiviral activity of the fraction in a study can be unambiguously attributed to the aforementioned phenolic acid ester, arousing interest also for the research of new effective antiviral lead compounds in propolis. An aqueous propolis extract, rutin, and a rutin/ quercetin combination were assayed in mice infected with influenza A/PR/8/34 (H1N1) virus. When propolis extract was administrated intranasally before virus inoculation, a reduction in viral hemagglutinin titers in lungs was observed, but no reduction in mortality or increase in survival times could be seen. However, when propolis was administrated after infection, the reduction in hemagglutinin titers was accompanied by a slight decrease in mortality. Rutin and rutin/quercetin combination were ineffective and actually increased both hemagglutinin titers and mortality. 97 Despite the unclear significance of these findings, it is evident that propolis extract displays a biological activity not paralleled by its isolated components. Moreover, the fact that a decrease in mortality was only observed when propolis was administered after infection possibly indicates an immunomodulatory, as well as, antiviral effect.
Governa et al. 98  | 915 speculated that propolis exerted no direct effect on virus particles, nor interacted with viral components, but rather enhanced cell resistivity via the activation/inactivation of the cellular process yet to be unraveled. 100 The same research group further explored the ability of green Brazilian propolis and 3,4-dicaffeoylquinic acid to act against influenza A virus, apparently without influencing the viral components. In vivo, aqueous and ethanolic extracts, and 3,4-dicaffeoylquinic acid increased the survival times of infected BALB/c mice. In an attempt to explain the mechanism of action of 3,4-dicaffeoylquinic acid, the mRNA expression in lungs of viral hemagglutinin was found to be moderately decreased, while the mRNA of TRAIL, a proapoptotic factor with viral clearance activity, was increased. Therefore, the authors confirmed the anti-influenza activities of green Brazilian propolis extracts and hypothesized that their mode of action, at least in part, included two mechanisms: an unknown cytoprotective effect and the enhancement of viral clearance via TRAIL overexpression, both possibly induced by 3,4dicaffeoylquinic acid. Also, immunomodulatory and anti-inflammatory effects might not be excluded. Notably, such a mechanism, based on the enhancement of self-defense machineries of the host, might overcome the problems deriving from the emerging drug-resistance of influenza viruses. 101 Finally, Kai et al. 102  RSV infected BALB/c mice treated with propolis were lower than those in the control. The effect was particularly significant in the case of IFN-γ, IL-6, and IL-10. Interestingly, propolis treatment did not affect the production of anti-RSV antibodies. 103 Altogether, these results support the hypothesis of an effect on the host immune system, rather than a direct antiviral one, and demonstrate that propolis administration may be equally beneficial in attenuating the inflammatory drawbacks of viral diseases. regions. The mechanism of propolis activity in CD4 + lymphocytes involved, at least in part, the inhibition of viral entry.
Moreover, propolis displayed an additive effect with the reverse transcriptase inhibitor zidovudine but had no noticeable effect on the protease inhibitor indinavir. 104 In the study by Harish et al., the authors showed the ability of propolis to suppress HIV-1 replication in vitro in CEM cells. At high concentrations, propolis abolished syncytium formation, whereas at lower ones it was inhibited in a concentration-dependent manner. Additionally, propolis decreased p24 antigen production by 90%-100% in a concentration-dependent manner. 105 In an in vitro screening of natural products against HIV-1 in H9 T-cell line, Ito et al. 106

| IMMUNOMODULATORY ACTIVITY
The results of in vivo and clinical studies on the antiviral activity of propolis, presented in the previous section, have often highlighted other mechanisms of action involving the modulation of the host immune responses, an effect which complements direct antiviral activities but sometimes appears to be independent and even more relevant.
Some of the earliest evidence of propolis immunomodulatory activities was obtained by the group of Popov in the early 1990s. In a series of studies, the authors examined in vitro and in vivo the immunomodulatory action of a water-soluble propolis derivative on complement activity. In vitro, the extract inhibited classical and alternative complement pathways in a concentration-dependent manner. In particular, the inhibition of the classical pathway was stronger. Moreover, the extract diminished C3 protein functional activity. 108 In vivo, the extract was administered intravenously, intraperitoneally, and orally to ICR mice. An alteration of serum alternative pathway complement levels was observed. Interestingly, a significant reduction of acute inflammation in zymosan-induced paw edema was obtained after oral administration, a condition in which serum complement levels were not influenced, demonstrating that the effect was strongly dependent on the route of application. 109 Although partial, these results paved the way to the comprehension of propolis effects on the host immune response.
In the late 1990s Brätter et al. 110  without undesired side effects. 110 The most recent updates consist of two systematic reviews in which the effect of propolis supplementation on C reactive protein, TNF-α, IL-1, and IL-6 levels were investigated. The meta-analyses of randomized clinical trials highlighted a significant reduction in IL-6, C reactive protein, and TNF-α serum levels following propolis administration, whereas propolis did not exert any significant effect on IL-1. 111,112 The vast majority of evidence in this context, though, is still anchored to the preclinical level. Consequently, in the following paragraphs in vitro, ex vivo, and in vivo findings are presented, grouped according to the propolis origin and type, to highlight any peculiarity of each. This subdivision has been adopted due to the fact that propolis is a poorly standardized natural product, characterized by extreme variability and multifariousness of composition, that cannot be reduced to the mere sum of components active on specific molecular targets. Despite the similarities in composition, the overall effect is, often and however, dramatically different.
What emerges from this section is that most of the evidence is related to Brazilian propolis; however, among the thirteen known subtypes, in practice only the red and, especially, the green ones have been investigated in this context, at the expense of other varieties still poorly characterized.
The principal modes of action that emerged from preclinical studies are collectively summarized in Figure 2. propolis hydroalcoholic solution, inhibition of concanavalin A-induced splenocyte proliferation could be only seen at lower doses (2.5-5 mg/kg). Propolis alone did not induce IFN-γ release in cell culture supernatants, but concanavalin A-stimulated splenocytes from propolis-treated mice produced significantly more IFN-γ than the stimulated controls. 113 TLR-2 and TLR-4 expression, and IL-1β and IL-6 (only in splenocytes) production were upregulated in splenocytes and peritoneal macrophages, possibly indicating that propolis modulated the mechanisms of the innate immunity, activating the initial steps of the immune response. 114 In unstimulated splenic tissue, propolis did not affect T h 1 (IL-2 and IFN-γ) and T h 2 (IL-4 and IL-10) mRNA expression; IFN-γ mRNA could not be even detected in experimental groups. However, the analysis of cytokine production in ex vivo cultured splenocytes showed a reduction in IFN-γ basal production, whereas no changes were detected in IL-2, IL-4, and IL-10. 115 Pagliarone et al. 116 evaluated the same parameters in splenic cells of acutely stressed BALB/c mice treated per os with propolis extract for 3 days after being subjected to a restraint stress protocol. Higher corticosterone concentrations were characteristic markers of the stressed group, treated or not with propolis, in comparison with the control group. As far as it concerns T h 1 cytokine production, no alterations were observed regarding IL-2, while IFN-γ production was inhibited in stressed mice even after propolis treatment. As to T h 2 cytokines, IL-4 production was inhibited in stressed mice, but levels were normalized by propolis treatment. No significant differences were found in IL-10 production. These results suggest the inability of propolis extract counteract the immunosuppressive effect on IFN-γ production in acutely stressed mice. Nevertheless, propolis could increase IL-4 production, thus favoring the humoral immune response under stress. 116 The potential effect of a preventive treatment was not considered, since propolis was administered only following the stressful event. On the other hand, Missima and Sforcin considered the immunomodulatory effects on macrophages and lymphoid organs in chronically stressed BALB/c mice subjected to immobilization stress. Stressed mice showed greater H 2 O 2 production by peritoneal macrophages. Propolis treatment further increased H 2 O 2 generation but inhibited NO production by these cells.
Histopathological analysis showed no alterations in thymus, bone marrow, and adrenal glands, but propolis treatment counteracted the alterations found in the spleen, as splenic germinal centers appeared to be increased. 117 In chronically stressed C57BL/6 mice, propolis administration prevented the downregulation of splenic TLR-2 and TLR-4 mRNA expression, favoring the initial steps of the immune response, thus reinforcing the evidence obtained in in vitro and ex vivo experiments. 118 Bachiega et al. 119  demonstrated an anti-inflammatory effect of propolis in response to LPS. Finally, when monocytes were exposed to melanoma-associated antigen-1 (MAGE-1) to induce a T h 1 profile, propolis decreased CD40 expression, inhibited TNF-α and IL-6 release, and increased IL-10 production, thus favoring a deactivating status important to avoid damage to the host. The immunomodulatory activity of propolis was shown to be independent from autophagy, since LC3 expression was not induced. In addition, propolis inhibited NF-κB expression, downregulating monocyte inflammatory activity. 123 Even though the signaling pathways modulated in differentially polarized monocytes were not fully elucidated, propolis exerted an immunomodulatory activity oriented toward an anti-inflammatory profile.
These results, which appear sometimes inconsistent, show that the effects of propolis extract may be different, and even opposite, in relation to treatment concentrations and other less predictable experimental variables.
Among isolated constituents of propolis, caffeic acid downregulated TLR-2 and HLA-DR expression without exerting any effect on TLR-4 and CD80 expression. The treatment inhibited TNF-α and IL-10 production at all concentrations. In addition, while the first was affected by TLR-4, but not TLR-2 blockage, the latter was altered neither by TLR-2 nor TLR-4 blockage. 124 Cinnamic acid downregulated TLR-2 expression; on the other hand, TLR-4 expression was increased. HLA-DR expression was inhibited at all tested concentrations, while CD80 was only at the highest one (100 µg/ml). Both TNF-α and IL-10 production were inhibited by cinnamic acid treatment, but this effect was abolished by TLR-4 blockage, thus suggesting, at least in part, its involvement in cinnamic acid activity on human monocytes. 125 Altogether these results demonstrate that propolis isolated compounds may exert immunomodulatory activities on PBMCs, but nonetheless, most of the synergistic effects are still unknown, as well as the scale and significance of these effects in vivo and at a clinical level. Cardoso et al. 126 investigated the involvement of various phenolic acids (caffeic, dihydrocinnamic, and p-coumaric acids), alone or in combination, in the activity of propolis on resting and LPS-activated human PMBCs. Caffeic acid combinations enhanced TNF-α production by resting PBMCs, but not by LPS-stimulated cells. All treatments upregulated IL-10 in resting monocytes, in particular caffeic acid combinations exerted an effect actually higher than the one induced by propolis extract, whereas neither propolis nor phenolic acids affected IL-10 production by LPS-stimulated monocytes.
Propolis, dihydrocinnamic acid, p-coumaric acid, and a combination of caffeic and dihydrocinnamic acids significantly stimulated IL-6 production by resting monocytes. However, in LPS-stimulated cells, the treatments only slightly downregulated IL-6. Propolis and phenolic acids did not influence the expression of HLA-DR, CD80 and TLR-2; however, the combination of caffeic and dihydrocinnamic acids upregulated LPS-induced TLR-4 expression.
These results reaffirm the ability of Brazilian propolis extract to activate human monocytes and demonstrate that phenolic acids are differentially involved in its activity. 126 Finally, Conti et al. 127  Brazilian and Mexican propolis stimulated IL-10 production in a concentration-dependent manner, whereas Cuban propolis showed inhibitory activity. In general, propolis immunomodulatory actions depend on its origin and consequently chemical composition, and although these findings seem to be contradictory, differences in cytokine production may be due to different compositions and synergistic or antagonistic effects among the components.
Nonetheless, Brazilian, Cuban, and Mexican propolis contain different constituents that may exert mild proinflammatory effects, useful to stimulate the initial events of the immune response or anti-inflammatory activity, depending on their concentration. 127 Such considerations on propolis range of effects may well apply also to the particular case of Brazilian propolis collected in Botucatu and used in the studies presented in this section. Despite the same provenance, the variance among the responses obtained in murine or human, resting or variously activated, immune cells might be ascribed on the one hand to differences inherent in experimental models, on the other to the summation of different activities exerted by individual components and synergisms still to be unraveled. An exemplary case is that of the effect on IL-10 levels, found to be unchanged, elevated, or reduced, each time with plausible scientific explanations.
However, what seems certain is the ability of this propolis type to modulate the mechanisms of the innate immunity, possibly activating the initial steps of the immune response, and that these immunomodulatory and antiinflammatory effects are mainly the result of additive, synergistic and antagonistic interactions among individual components, which define the overall pharmacological activity. Nonetheless, more studies, both preclinical and especially clinical, are needed to definitively elucidate the direction and entity of the effects and their therapeutical exploitability.

| Brazilian propolis (other origins)
Origin, subtype, and chemical composition of propolis are crucial factors to understand the relevance of the reported findings and direct the overall interpretation of propolis biological effects and therapeutic utility. However, the indication of such information is often lacking or incomplete. In the present section, evidence has been tentatively organized by inferring missing information, trying to combine studies considering similar propolis extracts.
In a study by Takagi et al., 128 Brazilian propolis hydroalcoholic (70% ethanol) extract suppressed IgG production in C3H/HeNCrj mice, while the amount of IgM was found to be significantly higher in comparison with the control group. In C57BL/6CrSlc mice treated with propolis, the number of CD4 + cells increased, both in basal condition and MAGNAVACCA ET AL.
| 921 after total-body X-ray irradiation (2 Gy), while the number of CD8 + cells increased dramatically in basal condition, but only slightly after irradiation, without reaching the levels of nonirradiated controls. According to the authors, propolis could activate macrophages and stimulate IFN-γ production, thus resulting in a decrease in IgG and IgM production. 128 The dietary supplementation of raw green propolis improved the innate and adaptative immunity in aged Kunming mice. After 4 weeks of treatment, propolis administration promoted phagocytosis in peritoneal macrophages and significantly increased phagocytic indexes in comparison with the control. Splenic T-cell proliferation was only slightly augmented, while splenic NK cells were unaffected. Serum IgG was dramatically increased in mice receiving propolis, whereas no significant change was detected in IgM. Serum hemolysins were found to be higher in treated mice, while IL-1β, IFN-γ, and IL-4 levels were unaffected. These results indicated that green Brazilian propolis was effective in improving innate and adaptive immunity in aged mice, especially at a dose of 157.4 mg/kg. 129 These two studies present contrasting evidence, especially in the case of a preferential effect on IgG or IgM and the influence on IFN-γ production, probably owing to inherent differences in animal models.
Nevertheless, what is particularly noticeable is the ability of Brazilian propolis to exert an immunomodulatory effect, which is still to be completely unraveled.  133 Finally, Mikami et al. 134 investigated the immunomodulatory activity of green Brazilian propolis ethanolic extract (standardized to contain 8.0% artepillin C and 0.14% culifolin) on Foxp3 + regulatory T

cells (T reg ). Propolis did not affect Foxp3 expression in induced regulatory T cells (iT reg ) and natural regulatory T cells
(nT reg ), but rather supported T reg expansion and activation through the upregulation of TNFR2 expression via IRF4/ cMyc pathway, with artepillin C being a major actor of this effect. TNRF2 has in fact been recently reported as a T reg -specific receptor that contributes to the anti-inflammatory immunosuppressive feedback exerted by T reg cells. 134 Once again, it was demonstrated that the same propolis type, that is, artepillin C-rich green Brazilian propolis, may exert different but equally interesting effects on different immune cells subtypes, on the basis of the specific context taken under consideration.
In turn, Bueno-Silva et al. 36  Propolis treatment reduced NO production by 65%, without affecting cell viability, and decreased the production of IL1-β, IL-4, IL-6, IL-12, IL1-3, MCP1, and GM-CSF. 135 One of the principal constituents of red Brazilian propolis, neovestitol, was shown to specifically exert immune-modulatory effects on LPS-activated RAW 264.7 murine macrophages in vitro, reducing NO production by 60% without affecting cell viability and decreasing GM-CSF, IFNγ, IL1-β, IL-4, TNF-α, and IL-6 levels, while increasing IL-10 production, 136 configuring a pro-resolving mechanism of action. In a similar study, also vestitol inhibited NO production by 83%, without affecting cell viability, and

| Chinese propolis
The composition of Chinese propolis is generally similar to that of the European poplar type, rich in flavonoids and phenolic acids esters, and glycerides. An extract of Chinese propolis, characterized by the presence of galangin, chrysin, and pinocembrin, was able to inhibit the secretion of TNF-α, IL-1β, and chiefly IL-6, from LPS-stimulated human PBMCs, whereas it had no effect on IL-8. Additionally, propolis significantly reduced IL-10, PGE2, and NO levels and completely abrogated IFN-γ production. 140 Shi et al. 141

| Other propolis
In this section, the evidence related to propolis types of various origins, the limited number of which does not permit an individual separate discussion, is presented.

Iranian propolis ethanolic extract could inhibit the release of T h 2 cytokines, IL-13, and IL-17 induced by
Aspergillus fumigatus conidia in murine lung epithelial cells (TC-1). 144 Moreover, when the extract was administrated per os to tumor-bearing BALB/c mice with disseminated Candida albicans infection, a condition in which the mean tumor size was significantly higher when compared with control group, propolis treatment significantly reduced tumor size in both C. albicans-infected and noninfected mice. Propolis determined a significant decline in T h 2 cytokines (IL-4 and IL-10) and a slight decrease in IL-17 production, whereas a significant increment in TNF-α and IFN-γ levels was observed in comparison with control groups. 145 The reduction of IL-17 production may correlate with the inhibition of T h 17 differentiation, an effect previously reported for Brazilian propolis. T h 17 cells play a role in adaptive immunity, fungal infections, and promotion or regression of tumors in murine models, however, as mentioned earlier, might also be responsible for detrimental inflammatory effects.
Turkish propolis ethanolic extracts decreased neopterin release and tryptophan degradation mediated by indoleamine 2,3-dioxygenase activity in PBMCs, both resting and treated with mitogenic phytohemagglutinin, and reduced TNF-α and IFN-γ production in stimulated cells. Indoleamine 2,3-dioxygenase is a heme-containing enzyme important for the defense against various pathogens, produced in response to inflammation, and able to inhibit T-cell function and induce immune tolerance. Thus, the inhibition of its activity represents an immunostimulatory mechanism. 146 A dimethyilsulfoxide extract of Turkish propolis reduced phorbol myristate acetate-induced respiratory burst, and the consequent secretion of elastase, a neutrophilic marker of acute inflammatory responses, in isolated human polymorphonuclear leukocytes. 147 Medjeber et al. 148 assessed the effect of Algerian propolis hydroalcoholic (85% ethanol) extract in PBMCs from patients with celiac disease. Compared with control, PBMCs from patients showed higher NO and IFN-γ levels, which were significantly decreased by propolis treatment. In addition, propolis extract significantly increased IL-10 production, downregulated iNOS mRNA expression, and impaired NF-κB and pSTAT-3 transcription factors activity. 148 In a model of carrageenan-induced inflammation in Wistar rats, the oral pretreatment with the ethyl acetate fraction of an ethanolic extract of Algerian propolis from Tigzirt significantly reduced paw edema, counteracted white blood cell increment, and increased the enzymatic activity of superoxide dismutase, catalase, and glutathione peroxidase. Moreover, propolis reduced the levels of PGE-2 and TNF-α and the activity of myeloperoxidase in the peritoneal exudate. In particular, the authors demonstrated that propolis is able to both reduce the release of myeloperoxidase from neutrophils and directly inhibit its activity. 149 The immunomodulatory properties of Moroccan propolis, mainly characterized by pinocembrin, chrysin, quercetin, and galangin, were evaluated in vitro by Touzani et al. 150 Propolis extract alone showed no effect on the production of TNF-α and IL-6, as well as IL-10, in human PBMCs. However, when PBMCs were stimulated with LPS, propolis significantly inhibited the secretion of TNF-α and IL-6, which at a concentration of 250 μg/mL declined toward basal levels, whereas IL-10 production was strongly enhanced in a concentration-dependent manner, thus indicating a selective anti-inflammatory and pro-resolving effect on activated immune cells. 150 Shvarzbeyn and Huleihel 151  subset. Low concentrations of propolis (1-2.5 µg/ml) had a protective effect on the activity and proliferation of B cells but did not affect the percentage of CD4 + and CD8 + T-cells. No negative effects could be seen on the proliferation and vitality of NK cells. 153 Sampietro et al. 154 assessed the immunomodulatory activity of delipidated extracts of propolis from the North of Argentina (mainly derived from Salix humboldtiana Willd., Pinus spp., and Eucalyptus spp.), and purified galangin and pinocembrin. Propolis was more effective as a chemotactic agent than the isolated compounds and stimulated higher neutrophil phagocytic activity, probably due to the synergistic effects among its components. 154 Alanazi et al. 155  Mirzoeva et al. 159 considered, instead, the activity of propolis ethanolic extract and its isolated components (caffeic acid, CAPE, quercetin, and naringenin) on eicosanoid production. In vitro, low concentrations (20-200 µg/ml) of propolis prevented LTB4, LTC4, and PGE2 secretion by murine peritoneal macrophages. Among the isolated compounds, CAPE strongly suppressed LTB4 and LTC4 synthesis in a concentration-dependent manner, followed by caffeic acid and quercetin. Naringenin only slightly impaired LTC4 production. CAPE also had a concentrationdependent inhibitory effect on PGE2 secretion. In vivo, in an acute peritoneal inflammation model in C57BL/6 mice, intraperitoneal administration of propolis extract decreased LTB4, LTC4, and PGE2 production by 95%, 90%, and 70%, respectively. CAPE inhibited LTB4 and LTC4 secretion, whereas caffeic acid affected only LTC4 synthesis. None of the isolated compounds was able to inhibit PGE2 production. The dietary supplementation of propolis during acute inflammation was also effective in reducing LTB4 and LTC4, but not PGE2, production. 159 Rossi et al. 160 investigated in vitro the effect of a commercial propolis ethanolic extract, as such and deprived of CAPE, and some isolated components on cyclooxygenase (COX-1 and COX-2) activity in J774.A1 murine macrophages. Propolis extract effectively inhibited PGE2 production by COX-1 and COX-2 in a concentration-dependent manner. Among the isolated compounds, caffeic, ferulic, cinnamic, and chlorogenic acids, and pinocembrin did not affect the activity of COX isoforms, whereas CAPE and galangin showed inhibitory properties, the latter being about 10-to 20-fold less potent.
CAPE-deprived propolis extract was about 10-fold less potent than the extract as such in the inhibition of both COX-1 and COX-2, suggesting that CAPE and to a lesser extent galangin contribute to the overall activity of propolis. 160 Finally, in their work Mounieb et al. 161 examined in vivo the potential protective effect of propolis oral administration against liver injury in concanavalin A-induced hepatitis in Wistar rats, a T-cell-dependent model that causes an immune-mediated disease similar to the one induced by viral infections, in which oxidative stress, increased levels of TGF-β, and fibrosis are the hallmarks. Induction with concanavalin A caused histopathological changes, reduction in serum albumin, and significant increase in serum levels of ALT, AST, and total bilirubin. Also, lipid peroxidation in liver tissue was found to be increased, while glutathione, and superoxide dismutase, and catalase activities were markedly downregulated. Serum levels of inflammatory cytokines (TNF-α and IL-6) and TGF-β were increased. Propolis treatment was able to significantly attenuate all these deleterious effects, improving liver function. 161 Altogether, these findings primarily suggest the anti-inflammatory activity of propolis, frequently proven in activated, rather than resting, immune cells. The investigations on propolis isolated components recognized, when present, caffeic acid phenethyl ester (CAPE) as one of the main actors in propolis anti-inflammatory actions. In addition, the enhancement of pathogen clearance and of the resistance to infections may be inferred. IFN-γ, IL-1β, and IL-6 production. In vivo, subcutaneous administration of the formulation with ovalbumin to mice could effectively activate the cellular and humoral immune response, increasing levels of IgG, IL-4, and IFN-γ in serum, and the proliferation rates of splenic lymphocytes. 163 Sforcin et al. 164 investigated the effect of Brazilian (collected in UNESP campus, Botucatu, São Paulo State) and Bulgarian propolis hydroalcoholic (70% ethanol) extracts, and two isolated compounds (caffeic acid and quercetin) on the antibody production in bovine serum albumin immunized rats. Both propolis types, independently from the collection season and geographical origin, stimulated antibody production in the same magnitude after 15 days of immunization, whereas caffeic acid and quercetin were ineffective at this purpose. 164 Mojarab et al. 165  in a significant increase in the neutralizing antibody titers compared to control. 166 The same extract was evaluated in BALB/c mice associated with inactivated suid herpesvirus type 1 (SuHV-1) vaccine. The treatment with propolis extract alone did not induce significant levels of antibodies, an effect which could be obtained with the additional association of aluminum hydroxide. However, propolis was able to increase the cellular immune response, enhancing IFN-γ mRNA expression. 167 Finally, a phenolic compound-rich fraction of green Brazilian propolis methanolic extract was further investigated to determine its effectiveness in the stimulation of cellular and humoral immune response when co-administered with an inactivated vaccine against SuHV-1. The treatment significantly increased neutralizing antibody titers against SuHV-1, as well as the percentage of protected animals upon infection challenge, without requiring any other co-adjuvant. 168 Altogether, these results demonstrate that propolis extracts of different types, among which the most studied are of Chinese and Brazilian origin, possess intriguing immunostimulatory properties that could be effectively exploited to enhance humoral immunity. Therefore, the utility of propolis as an alternative and more efficient immunological adjuvant should be first confirmed in preclinical studies, propaedeutic to future clinical trials aimed at assessing the relevance of this mechanism.

| PROPOLIS APPLICATIONS IN RESPIRATORY DISEASES
An abnormal inflammation of the respiratory tract can be a life-threatening condition. Several pathologies may cause breathing difficulties, but more commonly the respiratory tract is highly reactive toward airborne toxic pollutants, cigarette smoke, irritant agents, allergens, and pathogens including viruses. A persistent inflammatory response might be due to an impairment in macrophage clearance from the inflamed site, chronic infection, oxidative stress, or local hypoxia. All these stimuli can lead to tissue remodeling of the airway walls, and definitely impair gas exchange at the level of the blood-air barrier. 169 Inflammatory responses are triggered by a complex interaction between neutrophils and epithelial cells, stimulating the recruitment of immune cells to the injured site. TNF-α and IL-8 are involved in the first phase of inflammation, increasing the expression of adhesion molecules on the endothelial cells of lung capillaries and attracting neutrophils, respectively. 170 Other inflammatory mediators generally expressed in respiratory tract inflammatory processes are MMP-9, the adhesion molecules ICAM-1 and VCAM-1, COX-2, and cytosolic phospholipase A2. 171 In particular, metalloproteinases cause the degradation of the extracellular matrix, possibly contributing to the onset of emphysema, 172 and play a significative role in the process, regulated by TGF-β, of airway tissue remodeling during asthma. 173 In the following paragraphs, the studies which demonstrate the ability of propolis, or its isolated components, to modulate the immune and inflammatory responses in the respiratory tract are presented.
Besides the clinical studies on propolis antiviral properties (Crişan et al., 60 Cohen et al., 61 Szmeja et al., 62 Esposito et al., 63 Di Pierro et al., 65 and Silveira et al. 77 ) described in the previous sections, few other works focused on the potential efficacy of propolis for the treatment of human respiratory diseases. Khayyal et al. 174 conducted a clinical study to assess the beneficial effects of a propolis-based food supplement as an adjuvant in adult patients with mild to moderate asthma treated with oral theophylline. The formulation contained an aqueous extract of propolis collected in Denmark, China, Uruguay and Brazil, and standardized to contain not less than 0.05% of aromatic acids (mainly caffeic, ferulic, isoferulic, cinnamic, and 3,4-dimethoxy-cinnamic acids), in addition to trace amounts of various flavonoids. After 2 months of treatment, patients receiving propolis showed a marked reduction in the incidence and severity of nocturnal attacks (from an average of 2.5 per week to only one) and an improvement of ventilatory functions. The serum levels of the pro-inflammatory cytokines TNF-α, ICAM-1, IL-6, and IL-8 dropped by 52%, 65%, 44%, and 30%, respectively, whereas IL-10 increased by 3-fold. PGE2, PGF2α, and LTD4 were significantly decreased to 36, 39, and 28%, respectively, of initial values. 174 This effect is consistent with the evidence concerning immunomodulatory effects presented in the previous section.
Although not properly being a respiratory disease, the consequences of the exposure to air pollutants on the human organism partially involve the respiratory tract. Sojka et al. 175  of the extract was determined at 125 µg/ml, and the effect was bacteriostatic. In addition, propolis was able to reduce biofilm formation and diminished the number of alive P. aeruginosa cells in the biofilm, suggesting that the reduction of biofilm mass was due to a diminished number of sessile bacteria. In particular, propolis partially inhibited the swimming activity of P. aeruginosa and the formation of a stable adhesion, while no effect on swarming and twitching activity could be observed. 177 The protective effect of the oral administration of a Saudi red propolis aqueous extract against lung damage was assessed in rats that were intraperitoneally injected with methicillinresistant Staphylococcus aureus. Propolis ameliorated the oxidative stress biomarker malondialdehyde, as well as the antioxidant markers glutathione peroxidase and superoxide dismutase. Moreover, propolis extract modulated the alterations in TNF-α and VEGF serum levels and ameliorated oxidative DNA damage and the apoptosis biomarker caspase-3 in the lungs. The biochemical results were supported by the histopathological observation of lung tissue. 178 Sayed et al. 179 evaluated the oral and intraperitoneal administration of Egyptian propolis hydroalcoholic (70% ethanol) and aqueous extracts in Sprague-Dawley rats for 60 days before intraperitoneal injection of S. aureus.
In control infected rats, the lung was the most affected organ, showing several focal pus nodules on the visceral surface of pleura and suppurative bronchopneumonia. The bronchial epithelium was degenerated and destructed.
In the treated group, 8.6% of the rats had morbidity manifestations; however, areas with signs of pneumonia were smaller with lower leukocytic infiltration, and bronchi were unaffected. The remaining rats (91.4%) were asymptomatic, and their lungs were microscopically normal without any lesion. Bacterial re-isolation could also not be obtained. 179 On the other hand, when the effect of the oral administration of propolis was evaluated in a rat model of Pneumocystis carinii infection, the treatment was found to be completely ineffective in reducing the number of cysts in the lungs of infected animals. 180 However, in the context of bacterial infections propolis has generally proven to exert both direct antibacterial activity, and anti-inflammatory and immunomodulatory effects, which seem particularly relevant in conditions that constitute a challenge for the immune system, such as disseminated infections and endotoxemia, and are characterized by systemic detrimental outcomes.
Barroso et al. 181 tested the hypothesis that the oral supplementation of Brazilian propolis ethanolic extract would repair lung damage in emphysema caused by cigarette smoke exposure. C57BL/6 mice were exposed to cigarette smoke for 60 days, then treated with propolis for additional 60 days. Histological analysis revealed the ability of propolis to reverse the septum and alveolar destruction, significantly improving lung histoarchitecture.
Additionally, the extract increased MMP-2 and decreased MMP-12 expression, enhancing the process of tissue repair and leukocyte recruitment. In particular, propolis promoted macrophage alternative activation, increasing IL- in ovalbumin-stimulated splenocytes from propolis-treated mice was significantly lower compared to the controls. 187 A standardized green propolis extract (EPP-AF ® ) was studied in C57BL/6 mice challenged with allergens. The extract was characterized by the presence of high quantities of artepillin C, 4,5-dicaffeoylquinic acid, drupanin, 3,5-dicaffeoylquinic acid, and p-coumaric acid. Propolis treatment decreased the total cell number and IL-5 levels in bronchoalveolar lavage fluid, and IL-13 levels in lung tissue, confirming the attenuation of T h 2 inflammation. M2 macrophage number was also reduced in the lungs of treated mice. Moreover, propolis enhanced the frequency of polymorphonuclear myeloid-derived suppressor, which can induce T reg differentiation. Accordingly, an increase in frequency and total number of CD4 + Foxp3 + T reg cells could be observed. Interestingly, this effect could be elicited directly by propolis on isolated T-cells in vitro. 188 The preclinical evidence of propolis effectiveness in models of asthma and allergy corroborates the results obtained in the above-mentioned clinical investigations.
Hu et al. 189 explored the anti-inflammatory effects of Chinese poplar propolis hydroalcoholic (80% ethanol) and aqueous extracts in a model of carrageenan-induced pleurisy and LPS-induced acute lung damage in rats. Both extracts exerted significant effects on pleural effusion, leukocyte counts, and NO and PGE2 levels in carrageenaninduced pleurisy. Propolis inhibited lung edema in the acute damage model, but the results were not significant. In addition, the hydroalcoholic extract showed a good ability to counteract the increase in neutrophils and the decrease in lymphocytes in lungs. 189 In a model of lung injury induced by oleic acid and LPS injection in Wistar rats, propolis ethanolic and aqueous extracts antagonized lung edema, decreased inflammation, and inhibited the expression and activation of NF-κB p65 subunit. 190 Borrelli et al. 191  | 931 plethora of preclinical, mainly in vivo, evidence has demonstrated the beneficial effects of this natural product in such pathological context. Propolis has proven its efficacy against bacterial infections, showing direct antibacterial activities, but also more intriguing immunomodulatory and anti-inflammatory effects that allow the preservation of lung histoarchitecture and function. Similar effects were demonstrated also in different other models of respiratory diseases, such as smoke-related emphysema, fibrosis, asthma, as well as pleurisy and generic lung injury, outlining the action profile of a wide spectrum therapeutic agent.

| Propolis and COVID-19
The severity of COVID-19, the disease caused by the recent novel coronavirus (SARS-CoV-2) outbreak and characterized by high plasma levels of inflammatory mediators (e.g., IL-1β, IL-8, IFN-γ, MCP1, and TNF-α) 195 and symptoms of acute respiratory distress, has brought to the fore again the significant role of inflammation and the so-called "cytokine storm" in the clinical outcomes of respiratory diseases. 55,56 In a retrospective study on patients with severe COVID-19, serum levels of IL-6, IL-8, and TNF-α were significantly higher compared to patients with mild disease. 196 Such an increase in levels of pro-inflammatory cytokines and chemokines is strongly associated with organ dysfunction more than the actual viral titers and leads to the development of respiratory impairment and pulmonary failure. 197,198 Based on that, the last months have seen a renewed interest in the use of propolis in the COVID-19 pandemic, also thanks to the emergence of novel potential targets in SARS-CoV-2 infection mechanism. 199 SARS-CoV-2 entry into the host cell is mediated by the interaction between angiotensin-converting enzyme 2 (ACE2) and viral spike (S) glycoproteins, primed by the host serine protease TMPRSS2. 200 Following viral entry, the pathogenesis is characterized by the upregulation of PAK1, a serine-threonine protein kinase with a nodal signaling role, involved in the suppression of the adaptive immune response, lung inflammatory processes, and fibrosis when abnormally activated. 201 Many naturally occurring compounds, including propolis components such as quercetin, myricetin, and caffeic acid, have been selected in in vitro screenings and exploiting virtual docking models as promising antiviral agents able to bind and inhibit SARS-CoV-2 key proteins. Interestingly, the majority of drug candidates are polar compounds, structurally belonging to the class of polyphenols. 202 Among the potential targets, SARS-CoV-2 M pro is a 3chymotripsin-like cysteine enzyme necessary for the processing of viral polyproteins. Hashem used an in silico molecular modeling approach to evaluate the activity of different propolis components, concluding that caffeic acid, CAPE, galangin, and chrysin possess a strong binding affinity and may inhibit the main protease of SARS-CoV-2. 203 In a similar work by Kumar et al., 204 once again CAPE was predicted to possibly interact with SARS-CoV-2 M pro . In an in silico screening of the possible activity of Egyptian propolis against SARS-CoV-2 M pro , RNA-dependent RNApolymerase (RdRp), and spike protein subunit 1 (S1), Elwakil et al. 205  Besides viral targets, propolis possesses the potentiality to interact also with the host proteins of the cell surface involved in the infection mechanism. Recent evidence showed that a TMPRSS2 inhibitor approved for clinical use could block viral entry, bringing to light an innovative treatment option. 208 Kaempferol, in fact, had been previously shown able to decrease the mRNA expression of TMPRSS2. 209 The inhibition of ACE2 may also be a relevant target against SARS-CoV-2 infection and has been suggested as a therapeutic alternative. Despite the possible drawbacks of the use of ACE inhibitors in patient with COVID-19, a recent observational study dispelled the doubts, and this class of drugs remains an important tool against potential cardiovascular complications. 210 Güler et al. 211 determined the composition of Anatolian propolis hydroalcoholic (70% ethanol) extracts in terms of phenolic acids and flavonoids and screened them as ligands for ACE2 receptors in molecular docking analysis. The results showed that rutin had the best inhibitory potential, followed by myricetin, CAPE, hesperetin, and pinocembrin. In particular, rutin could interact with zinc finger residues of the active sites of ACE2. 211 In addition, Osés et al. 212 evaluated in vitro several hydroalcoholic extracts (70% ethanol) of propolis from different origins (North-East Europe, South-West Europe, and South America) for their ACE inhibitory activity, which resulted higher than 95% for all the samples, except for one with low flavanol content. The higher values of ACE inhibition were found in samples with higher amounts of catechin and p-coumaric acid. 212 As previously stated, the pathogenesis of COVID-19 seems dependent on PAK1 abnormal activation, responsible for the suppression of the host immune system and lung fibrosis. 201 In addition PAK1 is a critical mediator of the "cytokine storm" that frequently is a cause of mortality in hospitalized patients. 213 The downregulation of PAK1 activity could restore the immune response, boosting the antibody production against SARS-CoV-2. At a molecular level, PAK1 is directly activated by Rac1, a G protein belonging to the Rho family of GTPases. Xu et al. 214 demonstrated that caffeic acid could reduce Rac1 protein and activity level, thus inhibiting downstream PAK1 activation. CAPE, green Brazilian propolis extract, and artepillin C can also selectively inhibit PAK1. 215 Finally, the inhibitory effect of Pacific propolis from Macaranga tanarius (L.) Müll. Arg., in particular the Okinawan type, on PAK-1 has recently been reviewed by Shahinozzaman and colleagues. 216 In recent times, some clinical evidence has been added to in vitro and computational findings. Fiorini et al. 217 examined the case of a 52-year-old female patient suffering from COVID-19 who self-administered a nonalcoholic preparation of green Brazilian propolis extract EPP-AF ® , at a dose of 45 drops three times a day for 14 days. After 12 days of treatment, the patient had recovered and the nasopharyngeal swab gave a negative RT-PCR test result. 217 The relevance of a single case report is debatable, but in light of the previously described clinical trial conducted by Silveira et al. 77 in the same period (June-August 2020), this first report of a possible clinical efficacy of propolis against COVID-19 contributes to supporting the therapeutic potential of green Brazilian propolis, and by extension propolis in general, against SARS-CoV-2 infection.

| CONCLUSION
Propolis is a collective term used to define a multifaceted bee product outwardly uniform in physical appearance, gross composition, and purpose, but dramatically diverse in its chemical components. From an ecological point of view, foraging for resins by honeybees is an energetically demanding activity that provides no clear individual reward.
However, it should not be forgotten that several plant species secrete resins and other exudates with antimicrobial properties in response to tissue damage, to protect buds, vegetal apices, and young leaves. Thus, honeybees have probably evolved the ability to collect antimicrobial compounds from the surrounding environment as a means of reducing the deleterious effects of parasites and pathogens by enhancing a colony-level social immunity, rather than investing only in individual immune defenses. 218 Starting from this premise, it seems clear that despite an extreme chemical variability, which depends on the botanical sources around the beehive and poses a serious issue about the product standardization, propolis possesses general pharmacological properties in terms of antimicrobial activity, which represent one of the main reasons for its clinical use against human diseases. Moreover, resins and exudates gathered by honeybees contain also a variety of plant secondary metabolites, mainly of phenolic nature, renowned for MAGNAVACCA ET AL.
their anti-inflammatory and immunomodulatory activities. Therefore, propolis is a "super-blend" of biologically active compounds evolutionarily sorted by honeybees, which pharmacological research can and should exploit to its advantage. Nevertheless, besides propolis multifariousness, there is a common thread among the chemical compositions of different samples showing similar biological activities: high concentrations of flavonoids and/or phenolic acids (mainly cinnamic and hydroxycinnamic acids, and their derivatives). As reported in this review, the most studied, and probably most promising, propolis types are the poplar and the green Brazilian ones, which are indeed characterized by flavonoids and by hydroxycinnamic acids and their prenylated derivatives, respectively.
All the collected in vitro, ex vivo, and in vivo evidence supports the hypothesis that propolis (both hydroalcoholic/ethanolic or aqueous extracts, although with slightly different mechanisms due to the different composition) may exert a direct or indirect broad-spectrum antiviral activity against several different viral families. On the one hand, in fact, propolis seems able to interfere directly, most often nonspecifically, with viral particles or steps in the viral replication cycle. On the other, it stimulates and strengthens the host innate and adaptive resistance against viral infections. Propolis and its isolated components have been shown to exert immunosuppressive activities, possibly related to the anti-inflammatory properties, on different T lymphocytes subsets, but paradoxically activate macrophagic and NK cell functions. And perhaps this is the most fascinating and promising effect. As we now know, the most devastating consequences of viral diseases often arise from a dysregulated response of the immune system and the consequent imbalance in cytokine and chemokine production. 55,56 In this context, propolis may demonstrate a dual nature as immunostimulant, to prevent the infection, and immunomodulator, to dampen the inflammatory state and counter the immune dysfunction after the onset of the disease.
As to clinical studies, almost all trials assessing propolis antiviral efficacy focused on poplar type propolis hemorrhagic fever. In this case, the most likely mechanism of action is the anti-inflammatory and immunomodulatory one, supported by the evidence of a reduction in pro-inflammatory cytokine levels.
As previously stated, propolis is a complex product with a broad spectrum of activity that goes far beyond the simple sum of its isolated components. At the current state of knowledge, due to the animal origin and the extreme variability, with all the associated standardization problems, propolis is therefore unlikely to reach the status of medicinal product recognized by regulatory agencies in the near future. Nevertheless, propolis is already available as a dietary supplement in many countries, and a plethora of preclinical studies and some clinical studies 60,62,65,76,174 considered in this review, involving the assessment of propolis systemic effects, have successfully explored the possibilities of an oral administration. Starting from these preliminary remarks, possible use for the prevention of human respiratory diseases may be inferred. Furthermore, with respect to the recent outbreaks of pandemic respiratory viral infections, propolis supplementation could be put to good use as an add-on therapy to help control viral replication and infectivity, and, above all, counteract the deleterious inflammatory drawbacks on the respiratory tract. Propolis clinical efficacy against certain viral diseases has yet been demonstrated, 60,62,65 and is supported by the preclinical evidence of a general broad-spectrum antiviral activity. Moreover, propolis possesses additional antibacterial properties, [176][177][178][179] useful against opportunistic superinfections. The modulatory effects on the balance between pro-and anti-inflammatory cytokines, together with those exerted on different immune cell subtypes, corroborate the positive effects on lung histoarchitecture in case of emphysema and fibrosis, [181][182][183]185,[219][220][221] or acute respiratory distress. 222 Finally, it should not be forgotten the possibility to exploit propolis strong antioxidant and radical scavenging abilities in case of prolonged oxygen therapy. 193 For what specifically concerns COVID-19, propolis immunomodulatory activity may be exploited in the initial phases of SARS-CoV-2 infection, characterized by the dysregulation of the immune response which facilitates viral replication, to reduce immunosuppression. Conversely, in later stages of the disease, propolis might counteract the onset of an exaggerated inflammatory response, hallmarked by a "cytokine storm" which causes multiorgan failure and makes it necessary to resort to intensive care and the use of ventilators. 223 Moreover, propolis may directly interfere with viral infectivity and replication, hindering the interaction between the virus and the host cell by targeting key proteins on both sides. Nevertheless, despite the encouraging results obtained in the first clinical trial conducted against COVID-19 with green Brazilian propolis, 77 stronger preclinical evidence is still required to corroborate such a therapeutic approach, not to mention that the effective bioavailability of propolis components at target organs in vivo is yet to be proven, being also possible that metabolites are responsible for the biological effects.
Until now, only a few clinical studies have evaluated the putative efficacy of propolis oral administration in the case of respiratory diseases, but none of them was conducted as a rigorous double-blind, controlled, randomized trial. Moreover, study backgrounds and propolis extracts differed significantly, thus providing low-quality evidence of efficacy which unfortunately prevents from formulating even a weak recommendation. Nonetheless, preclinical studies support the all-round potential of propolis in respiratory diseases and, given the current emergency caused by the COVID-19 pandemic and the limited therapeutic options, propolis should be suggested as a reasonably safe and hopefully effective therapeutic agent.
For all these reasons, it would be advisable that rigorous randomized clinical trials be conducted to ascertain propolis efficacy against respiratory tract infections and resultant diseases, opening up new perspectives for the rational use of propolis as an effective dietary supplement, or even a fully-fledged therapeutic agent in the future.

DATA AVAILABILITY STATEMENT
The data used to support the findings of this study are available from the corresponding author upon request.