1. Introduction
Development of optimal control measures against emerging viruses, especially those with pandemic potential, is an important goal. Vaccination is considered the best option for the control of viral diseases, but it would take 3–6 months to produce pandemic-matched, effective vaccines. Therefore, antiviral drugs can be an initial control measure. One of the promising candidates for the prevention and treatment of acute respiratory virus infections is aprotinin (APR) due to its ability to inhibit proteolytic activation of some viruses, and its activity against a broad range of viruses, including IVs and SARS-CoV-2. APR (
Figure 1) was discovered in 1930 as an “inactivator” of kallikrein in bovine lymph nodes
[1] and in 1936, as an inhibitor of bovine pancreatic trypsin
[2]. It is also known as a bovine pancreatic trypsin inhibitor (BPTI) and a trypsin-kallikrein inhibitor (TKI). According to X-ray crystallography, BPTI has a three-dimensional pear-shaped molecule structure. The polypeptide chain is folded so that hydrophobic radicals are concentrated inside the molecule, while all hydrophilic radicals, with the exception of the side chain of Asp-43, are outside the molecule, exposed to the aqueous environment. This arrangement results in a very compact tertiary structure and is mainly responsible for the remarkable stability of APR against denaturation at high temperature, acids, alkalis, and organic solvents or proteolytic degradation. Other interesting features of APR are, on the one hand, its strongly dipolar character due to the concentration of negatively charged radicals at one end of the molecule, i.e., in the lower part of the pear, and on the other hand, its strong basicity molecules with an isoelectric point close to 10.5
[3][4][5]. It should be noted that, due to incompetent interpretation of several clinical trials to reduce perioperative blood loss and the need for blood transfusions in patients undergoing coronary artery bypass grafting with cardiopulmonary bypass
[6][7][8][9][10][11], the usage of APR was nearly stopped for several years worldwide
[12]. After incredible controversy in the literature
[13][14][15][16][17][18] and re-analysis
[12][15][16][17][18][19][20][21][22][23][24][25][26] of clinical trial data
[6][7][8][9][10][11], restrictions on the use of APR were lifted worldwide in 2011–2020
[26][27][28][29].
Figure 1. Tertiary structure of APR. The figure was constructed using X-ray data from RCSB PDB (PDB ID: 3LDJ). The molecule consists of a single polypeptide chain of 58 amino acid residues linked by three disulfide bridges. The molecule is about 29 Å long, 19 Å in diameter, and contains a double-stranded antiparallel β-sheet (from Ala-16 to Gly-36, orange color) twisted into a right-handed double helix with 14 residues per turn
[30][31][32].
1.1. APR Inhibitory Activity
APR is a typical “magic shotgun”
[33][34][35][36] pharmacological agent that reduces bleeding and limits the need for blood transfusion in cardiac and noncardiac surgeries
[37][38][39]. It is also a promising drug in antiviral therapy
[40][41] and especially in combination with other drugs
[41][42].
APR is a competitive pan-protease inhibitor that forms loose complexes with serine proteases and blocks their active sites. It inhibits trypsin, chymotrypsin, and plasmin at a concentration of about 125,000 cfu/mL (KIU/mL) and kallikrein at a concentration of 300,000 cfu/mL (
Table 1). Its action on kallikrein leads to inhibition of the formation of factor XIIa. As a result, both the internal coagulation pathway and fibrinolysis are inhibited. The action of APR on plasmin independently slows down fibrinolysis
[43][44][45][46]. In addition, APR also inhibits the action of nitric oxide synthase types I and II and impairs K
+ transport through Ca
2+-activated K
+ channels
[47] and interacts with other factors of the coagulation and fibrinolytic cascade, creating a hemostatic balance without increasing the risk of thrombosis.
Table 1. The inhibition constants Ki for the complexes between APR and the various enzymes
[32].
APR is an effective anti-inflammatory drug
[48][49][50][51], which is called a “broad-spectrum anti-fibrinolysin” because of its anti-inflammatory and endothelial-modulating effects
[52]. It has multiple actions that may suppress the inflammatory response, including attenuating platelet activation, maintaining platelet function, decreasing complement activation, inhibiting kallikrein production
[53], decreased release of TNF-α
[54], IL-6 and, IL-8
[55], inhibition of endogenous cytokine-induced iNOS induction
[56], decreased CPB-induced leukocyte activation
[53][57], and inhibition of up-regulation of monocyte and granulocyte adhesion molecules
[58][59]. It may reduce lung injury, reduce bronchial inflammation
[60], and attenuate reperfusion lung injury
[61].
APR is an inhibitor of host serine proteases that cleaves the hemagglutinin (HA) glycoprotein of IVs and thus reduces the virus replication. In particular, it has been shown that plasmin cleavage of HA glycoprotein of IVs can be prevented by APR. The HA glycoprotein of IVs consists either of the precursor HA (75,000 kDa) or of its subunits HA1 (50,000 kDa) and HA2 (25,000 kDa). IVs cannot initiate infection of host cells unless the HA is proteolytically cleaved
[62]. The HA1 and HA2 subunits are significantly more infectious than the HA precursor
[63]. The HA glycoprotein of IVs plays a critical role in viral binding, fusion, and entry. That is why HA is an attractive target for inhibition of the initial stage of host cell infection with IVs.
[64][65][66]. APR inhibits transmembrane protease serine S2 (TMPRSS2), which is essential in proteolytic activation of a broad range of viruses, including IVs and SARS-CoV-2
[67].
1.2. Adverse Effects of APR
The use of APR in major surgeries began in the 1960s, and the first publications addressing the anti-influenza activity of APR appeared in the early 1970s
[68]. APR was generally well tolerated in patients undergoing surgery in clinical trials
[69]. An important side effect with APR is hypersensitivity, including skin rashes, itching, dyspnea, nausea, tachycardia, and a fatal anaphylactic or anaphylactoid reaction (shock). The frequency of hypersensitivity reactions ranges from <0.1% to 5%
[70][71][72][73]. Moreover, maximum hypersensitivity up to 5% was observed if patients were repeatedly exposed to APR within six months after initial administration
[73].
Therefore, all intravenous doses of APR are administered through a central catheter. An initial (test) dose of 1 mL (10,000 KIU) APR is administered intravenously at least 10 min before the loading dose that is administered as a constant infusion dose (1M-2M KIU)
[69].
An important side effect with APR is also a clinically significant but transient increase in serum creatinine and the potential for increased renal events. An increase in serum creatinine occurs in about 8% of patients and may persist for up to nine days. The mechanism is probably reuptake in the proximal tubules
[74]. These effects are most common in patients with existing renal dysfunction
[75].
2. Antiviral Treatment of Influenza
2.1. The Structure IAV, Function of Its Proteins, and HA Cleavage of IVs
The IAV genome is a negative-sense, single-stranded, segmented RNA genome which is divided into eight segments that encode at least 11 viral proteins (
Figure 2)
[76]. The IAV is an enveloped virus consisting of an outer lipoprotein envelope and an inner ribonucleoprotein (RNP). The virus envelope contains four proteins: HA, neuraminidase (NA), the transmembrane ion channel matrix 2 (M2), and a small amount of the nuclear export protein (NEP). RNP contains RNA and four polypeptides: the main nucleocapsid protein (NP), polymerase basic 2 (PB2), polymerase basic 1 (PB1), and polymerase acidic (PA). Both modules are connected to each other by the matrix 1 (M1) protein, which maintains viral integrity
[77].
Figure 2. Schematic representation of the IAV. A lipid bilayer contains HA and NA glycoproteins and transmembrane ion channel M2 protein. M1 protein lies beneath the lipid bilayer and binds by NEP protein. Individual RNA segments are bound by a polymerase complex, consisting of the three proteins PA, PB1, and PB2, at their termini and encapsidated by the NP into a helical structure (RNP).
The HA attaches virions to sialic acid (SA) fragments of host receptors; NA is not required for viral replication but required for budding of newly formed viral particles from the surface of infected cells. It facilitates virus movement to the target cell by cleavage of SA receptors from respiratory tract mucins, and helps the release of virions from infected cells; the M1 protein is a membrane-binding and RNA-binding protein and forms a coat inside the viral envelope, determines the virion’s shape, interacts with vRNP and other cytoplasmic domains of integral membrane proteins, increases vRNPs export and decreases import, and helps assembly and budding of virions; the M2 protein is vital for viral replication, forms a proton channel in the virus envelope, lowers the pH inside the viral particle to promote uncoating of RNPs, modulates Golgi’s pH, and helps to stabilize HA’s native conformation during virus assembly; the nonstructural protein 1 (NS1) acts as a promoter of viral replication and an inhibitor of the host's immune response; NP binds nonspecifically to single-stranded RNA (ssRNA), encapsidates viral RNA, and helps recruiting RNA polymerase for synthesis of viral positive-sense RNA (cRNA); the NS2/NEP protein promotes viral RNA replication, regulates vRNP’s export from the nucleus to the cytoplasm, RNA nuclear export, and interacts with the viral M1 protein; the PA has presumably helicase-like functions and is important for viral transcription and assembly of the polymerase complex; the main PB1 responsible for elongation of the primed nascent viral mRNA is located in the nucleus of infected cells, enhances the association of three subunits of the RNA polymerase complex; the PB2, located in the nucleus of infected cells, signals the viral polymerase passage to the host’s nucleus, enhances the formation of the cap structures necessary for viral messenger RNA (mRNA) transcription, located in the mitochondria of infected cells, inhibits Interferon-β (IFN-β), and helps determine host range
[78]; the PB1-F2 protein contributes to viral pathogenicity
[79]. Virus entry into the host cell, replication, assembly, and movement of the IVs virions are detailed in the reviews
[80][81][82].
The TMPRSS2 is expressed in epithelial cells of the human respiratory tract and cleaves (activates) the HA glycoprotein of IAVs and IBVs into HA1 and HA2 subunits to allow virus fusion with host cell receptors
[83][84][85]. It was first identified in 2006 by Bottcher and colleagues
[83]. Along with TMPRSS2
[83], host proteases with trypsin-like activity, such as TMPRSS4, TMPRSS11D, ST14, KLK5, KLK12, TMPRSS11E, and TMPRSS13, have also been shown to cleave HA glycoprotein of IAVs and IBVs and support viral replication in cell cultures
[85][86][87][88][89][90]. At the same time, the proteases prostasin, hepsin, TMPRSS3, TMPRSS6, TMPRSS9, TMPRSS10, TMPRSS11B, and TMPRSS11F do not activate HA glycoprotein of IAVs and IBVs when co-expressed in mammalian cells
[88][91][92][93].
Mice deficient in TMPRSS2 expression were protected from lethal challenge with A(H7N9) or A(H1N1)pdm09 viruses but were resistant to challenge with A(H3N2) virus. This suggests that activation of HA glycoprotein of A(H3N2) virus is independent of TMPRSS2
[94][95][96][97].
TMPRSS2 was found to be crucial for proteolytic activation of the avian IAVs of H1-H11, H14, and H15 subtypes in human and mouse airway cells
[98]. Only H9 (with a R-S-S-R cleavage site) and H16 avian IAVs were proteolytically activated in the absence of TMPRSS2 activity, albeit with reduced efficiency. It was also shown that in human and murine airway cells, TMPRSS2 is the major activating protease of IAV of almost all HA subtypes having a monobasic HA cleavage site. An additional exception was HA of IBVs in human and mouse respiratory cells, which did not depend on TMPRSS2 activation
[97].
Proteolytic cleavage regulates numerous processes in human metabolism and immune responses. One key player is the ubiquitously expressed serine protease furin, which cleaves a plethora of proteins at polybasic recognition motifs. Mammalian substrates of furin include cytokines, hormones, growth factors, and receptors
[99]. Generally, HA of human and low pathogenic avian IAVs cannot be cleaved by furin as they usually only harbor a mono- or dibasic HA cleavage site. Instead, they depend on trypsin-like proteases such as TMPRSS2 or human airway trypsin-like protease for activation
[83]. Expression of such trypsin-like proteases is largely restricted to the respiratory and gastrointestinal tract. In contrast, HA of many highly pathogenic avian influenza A(H5N1) and A(H7N9) viruses can be cleaved by furin or PCSK5, which are present in many cell types
[100]. Thus, the ability to exploit furin for efficient HA cleavage and the associated increase in pathogenicity are determined by the presence of a furin consensus target site, but also by adjacent residues and the absence of masking oligosaccharide chains.
2.2. Antiviral Drugs Available for Influenza Treatment
Influenza affects about 3% to 10% of the world’s population annually. The most common complications of influenza include viral
[101] or bacterial co-infections
[102][103], which lead to the death of about half a million people each year
[104]. CDC estimates that influenza caused 29–41 million illnesses, 380,000–710,000 hospitalizations, and 22,000–38,000 deaths annually between 2010 and 2020
[105][106].
During the influenza season in the United States, mortality attributed to influenza associated with pneumonia ranges from 5.6% to 11.1%
[107]. In a cohort study including laboratory-confirmed influenza cases, those admitted with pneumonia, were more likely to require admission to an intensive care unit (ICU, 27% vs. 10%), mechanical ventilation (18% vs. 5%), and higher risk of death (9% vs. 2%)
[108]. In 2020, the CDC ranked influenza complicated with pneumonia as the ninth leading cause of death in the United States
[109].
Vaccination is considered the most effective strategy for preventing and controlling influenza in humans
[104][110]. However, current influenza vaccines have several limitations, including their limited efficacy when there is an antigenic mismatch between the vaccine composition and circulating viruses
[110]. The effectiveness of seasonal prophylaxis with influenza vaccines developed over several decades ranges from 10% to 60%
[109][111][112].
There are currently seven viral target proteins, including nine antivirals approved for the treatment of influenza
[113][114][115][116][117][118][119][120].
Historically, treatment options for influenza infections were limited to four classes of virus protein-specific drugs targeting M2, NA, PB1, or PA proteins. The first of them, inhibitors of ion channel activity of M2 protein (amantadine, rimantadine) are active only against IAV
[121]. M2 inhibitors have lost their relevance because IAVs are resistant to these compounds
[122][123][124] and IBVs are insensitive to M2 inhibitors
[113][123][125][126].
The NA inhibitors (NAI) are targeting the surface of NA glycoprotein (oseltamivir, zanamivir, peramivir, and laninamivir), and act against IAV, IBV, and ICV. The prevalence of viruses resistant to NAIs in global circulation is generally low (<2.0%)
[127]. However, with only a single oral-dosed NAI oseltamivir available on the market, the development of new and improved anti-influenza drugs is important
[127][128][129]. The drug candidate effective against oseltamivir and zanamivir resistant viruses is the NAI AV5080
[130][131].
Favipiravir (FVP) is an inhibitor of the RNA-dependent RNA polymerase (RdRp) of a broad range of RNA viruses, and it inhibits viral RNA synthesis as a chain terminator
[132]. Emerging IV, with antigenically distinct surface glycoproteins and composition of internal genes than seasonal IV, cause severe disease and high mortality rates—53.5% for A(H5N1)
[133] and 34% for A(H7N9) virus infections
[134]. FVP demonstrated antiviral activity against different subtypes of IV in animal models, including highly pathogenic A(H5N1) and oseltamivir resistant viruses
[135][136][137][138]. FVP was more efficacious than oseltamivir in inhibiting replication of the A/Puerto Rico/8/1934 (H1N1) virus in vitro and the protection of mice infected with a high dose of this virus
[139]. Following clinical trials, FVP was approved for restricted use and pandemic stockpiling in Japan in 2014
[140]. The results of two randomized, double-blind, placebo-controlled phase 3 (US316 and US317) international trials of FVP treatment of uncomplicated influenza in adults have recently been published. US316 (NCT02026349) was conducted in 14 countries in Africa, Europe, Asia, Australia and New Zealand, and the United States over three influenza seasons between January 2014 and March 2015. US317 (NCT02008344) was conducted in 10 countries and territories in the Americas between December 2013 and February 2015. In both studies, FVP demonstrated a decrease in viral titers within 1–5 days after initiation of treatment and the median time to loss of virus detection decreased by 23.2–24.0 h compared with placebo (P < 0.001). Adverse events were generally mild or moderate. The authors recommended to conduct additional studies and investigate higher FVP doses and drug combinations for the treatment of severe influenza and other RNA-containing viral infections
[141].
Baloxavir marboxil (BXM) is a prodrug of the biologically active baloxavir acid (BXA) and inhibits cap-dependent endonuclease (CEN) activity of PA proteins of IAV, IBV and ICV. BXM is the first inhibitor of this type approved in Japan (2018), the United States (2018), and Europe (2021)
[142][143][144]. A significant advantage of BXM over NAIs is its weight-adjusted single oral dose administration regimen
[145][146]. The emergence of IV with PA-I38T substitution was already detected on day three after treatment with BXM (range three–nine days), and in most cases this occurred on day five and may lead to virus rebound
[147][148][149]. The BXM analog AV5124 (prodrug of AV5116) exhibited low cytotoxicity in MDCK cells and lacked mitochondrial toxicity, resulting in favorable selective indexes. AV5116 was equipotent or more potent in vitro than BXA against wild-type viruses and viruses with reduced BXA susceptibility carrying a PA-I38T substitution
[150]. AV5124 showed promising efficacy as an anti-influenza drug candidate in a mouse animal model
[119][150][151], and treatment with 20 mg/kg or 50 mg/kg prevented death in 60% and 100% of animals, respectively
[151].
Host-targeted antiviral drugs, such as inhibitors of proteolytic activation of the HA glycoprotein of IVs (serine protease inhibitors), only recently appeared in influenza therapy. This group includes camostat, nafamostat
[152], and APR
[153]. However, in contrast to the efficacy in vitro
[154], the clinical efficacy of camostat and nafamostat is still unclear. At the same time, a hand-held metered-dose inhaler containing APR (Aerus
TM) for the treatment of influenza was developed and is used in Russia
[83][155][156].
2.3. APR for Influenza Treatment
The main advantages of APR over existing/proven antiviral drugs are its TMPRSS2 activity, thus inhibiting virus entry into host cells and virus replication
[86][92][93][157][158]. It is an excellent partner for combination therapy because it has a different mechanism of anti-influenza action than existing/approved antiviral drugs. In addition, it inhibits the processes of inflammation
[37][48][49][50][51] and vascular thrombosis
[44][45][46][47], which are very important in the treatment of complications (particularly pneumonia) caused by influenza infection.
The first publications reporting the anti-influenza activity of APR appeared in the early 1970s. It was shown that cleavage of the HA glycoprotein of IV by plasmin can be prevented by Kunitz trypsin inhibitors from bovine pancreas, i.e., by APR
[68]. APR has been actively studied as an anti-influenza drug since the early 1980s. It should be noted that a significant contribution to these studies was made by Zhirnov and colleagues
[63][153][154][159][160][161][162][163]. They demonstrated APR activity against a number of IAVs: A/Puerto Rico/8/1934 (H1N1)
[63][153], A/Aichi/2/1968 (H3N2)
[63][163], A/California/04/2009 (H1N1)pdm09
[153], A/Hamburg/05/2009 (H1N1)pdm09
[153], and oseltamivirresistant A/Brisbane/10/2007 (H3N2)
[164]; and IBVs: B/Hong Kong/1973
[162][165] and B/Lee/1940
[162][165].
More recently, Song et al.
[120] investigated antiviral APR activity in vitro among different IAV subtypes, including seasonal human IAVs [A/Puerto Rico/8/1934 (H1N1), A/California/04/2009 (H1N1)pdm09, A/Philippines/2/1982 (H3N2), A/Brisbane/10/2007 (H3N2)], avian IAVs [A/aquatic bird/Korea/CN2/2009 (H5N2), A/aquatic bird/Korea/CN5/2009 (H6N5), A/chicken /Korea/01310/2001 (H9N2)], and oseltamivirresistant IAV: [A/Brisbane/10/2007 (H3N2)] and IBV [B/Seoul/32/2011 (Yamagata-like lineage)]. The APR EC
50 values against different IAV ranged from 11 nM to 110 nM and were 39 nM for IBV
[166]. The anti-influenza activity of APR was confirmed in mice lethally challenged with A/Puerto Rico/8/1934 (H1N1) virus
[166].
In 2011, Zhirnov et al. suggested an aprotinin-based aerosol preparation for treatment of respiratory viral infections
[167]. Zhirnov et al. showed the efficacy of aerosol formulation of APR for the treatment of experimental influenza and parainfluenza bronchopneumonia in mice
[83][155][156][168][169][170][171][172][173]. In humans, APR aerosol demonstrated efficacy against natural influenza and parainfluenza infections when administered by inhalation using a manual aerosol inhaler Aerus
® [83][156]. The study was conducted in Russia during the winter–spring outbreak of influenza caused by the pandemic A(H1N1)pdm09 virus. Patients inhaled two aerosol doses of APR (160 KIU) three times a day for four–five days. In the comparison group, patients received a single oral dose of Ingavirin™ (90 mg) for five days. The authors found an approximately 10-fold reduction in viral load in patients treated with APR compared to those treated with Ingavirin™. The duration of clinical symptoms such as rhinorrhea, weakness, headache, sore throat, cough, chest pain, and fever was one–two days shorter in the APR-treated group than in the Ingavirin™-treated group. Side effects and discomfort in patients of the APR group were not detected
[167].
Acute myocarditis is a well-known complication of influenza infection and a common prelude to inflammatory dilated cardiomyopathy (DCM) that can lead to chronic heart failure
[156][174]. IAV-induced trypsin expression in the myocardium triggers acute viral myocarditis through stimulation of IAV replication, pro-MMP-9 activation, and cytokine induction. It was reported that inhibition of trypsin can prevent DCM with improvement in cardiac function after A/Puerto Rico/8/1934 (H1N1) virus infection of mice
[175][176][177]. It was shown that ectopic myocardial trypsin was involved in acute and chronic myocardial inflammation, promoting IAV infection and initiating the trypsin-MMP-9 cytokine cycle, and promoting progressive cardiac dilatation through collagen remodeling. Trypsin plays an important role in the development of DCM after IAV infection, and APR prevents the progression of myocarditis to DCM by suppressing IAV infection, interrupting the trypsin-MMP-9-cytokine cycle, and restoring collagen metabolism through inhibition of trypsin activity. Thus, pharmacological inhibition of trypsin activity may be a promising approach to the prevention of virus-induced cardiomyopathy. To date, no clinical data on the use of APR for the prevention of inflammatory DCM are available.
In conclusion, it should be noted that the efficacy of APR for the treatment of influenza and acute respiratory diseases (ARD) has strong experimental evidence in vitro and in animal models. However, there are no clinical trial data on the treatment of complications of influenza or ARD, such as pneumonia, either with APR or with combination therapy that includes APR and an anti-influenza drug with a mechanism of action other than APR. Given the pan-protease activity of APR and rehabilitation of APR, the time has come for the conclusive clinical trials using APR for the treatment of influenza pneumonia.