Highly polymerized proanthocyanidins (PAC) components from blueberry leaf and stem significantly inhibit SARS-CoV-2 infection via inhibition of ACE2 and viral 3CLpro enzymes

With the current worldwide pandemic of COVID-19, there is an urgent need to develop effective treatment and prevention methods against SARS-CoV-2 infection. We have previously reported that the proanthocyanidin (PAC) fraction in blueberry (BB) leaves has strong antiviral activity against hepatitis C virus (HCV) and human T-lymphocytic leukemia virus type 1 (HTLV-1). In this study, we used Kunisato 35 Gou (K35) derived from the rabbit eye blueberry (Vaccinium virgatum Aiton), which has a high PAC content in the leaves and stems. The mean of polymerization (mDP) of PAC in K35 was the highest of 7.88 in Fraction 8 (Fr8) from the stems and 12.28 of Fraction 7 (Fr7) in the leaves. The composition of BB-PAC in K35 is that most are B-type bonds with a small number of A-type bonds and cinchonain I as extension units. A strong antiviral effect was observed in Fr7, with a high polymerized PAC content in both the leaves and stems. Furthermore, when we examined the difference in the action of BB-PAC before and after SARS-CoV-2 infection, we found a stronger inhibitory effect in the pre-infection period. Moreover, BB-PAC Fr7 inhibited the activity of angiotensin II converting enzyme (ACE2), although no effect was observed in a neutralization test of pseudotyped SARS-CoV-2. The viral chymotrypsin-like cysteine protease (3CLpro) of SARS-CoV-2 was also inhibited by BB-PAC Fr7 in leaves and stems. These results indicate that BB-PAC has at least two different inhibitory effects, and that it is effective in suppressing SARS-CoV-2 infection regardless of the time of infection.


Introduction
The outbreak of coronavirus disease 2019 (COVID-19) is ongoing, with more than 230 million confirmed infections and 4.7 million coronavirus-related deaths. In Japan, on the other hand, the number of infected people is 1.7 million and the number of deaths is only 17,000; however, the number of COVID-19 patients is constantly increasing as a major domestic problem. Considered one of the most devastating global crises in recent years, infection with COVID-19 causes symptoms similar to those of acute respiratory distress syndrome [1]. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative virus of COVID-19, is a novel coronavirus strain [2]. To date, remdecivir has been licensed for emergency use as a therapeutic agent [3], and the efficacy of an urgently developed mRNA vaccine has been evaluated [4]. However, it will take some time to achieve herd immunity through vaccination, and although some new drugs for COVID-19 have been marketed, alternative measures are still needed to prevent infection.
We have reported that proanthocyanidins (PACs) contained in rabbit blueberry (BB) leaves have various functions, such as inhibiting hepatitis C virus replication [5], suppressing fatty liver development [6], inhibiting the progression of hepatocarcinogenesis [7], suppressing the development of adult T-cell leukemia [8], and inhibiting the proliferation of HTLV-1 virus-infected cells [9]. In recent years, we have developed Kunisato 35 gou (K35), which is suitable for use in tea leaves with high PAC content, and we are conducting functional analysis of PAC in K35 [10]. In this manuscript, we report a new function of the PAC fraction from K35 against a new type of coronavirus disease . The leaves and stems of K35 were extracted using hot water, and the extracts inhibited the RNA copy number of SARS-CoV-2 or the plaqueforming assay of SARS-CoV-2. After the extracts were fractionated using organic solvents, the fraction with high polymerization PAC content (BB-Fr7) was found to have a stronger suppressive effect against SARS-CoV-2. At the same time, BB-Fr7 with high PAC content showed potential as a prophylactic administration because it suppressed viral replication after infection by pretreatment. In addition, BB-Fr7 was found to inhibit angiotensin II converting enzyme (ACE2) and viral chymotrypsin-like cysteine protease (3CLpro), suggesting that it inhibits cell surface receptors necessary for infection and inhibits viral replication after infection. Therefore, administration of BB extracts containing highly polymerized PACs is expected to prevent new coronavirus infections.

Fractionation of blueberry leaf and stem extraction from K35
The fractionation of blueberry leaf and stem extracts has been previously described [8]. Lyophilized blueberry leaves and stem powder (300 g) were extracted for 30 min with boiling water (15 l) and then cooled in an ice bath. After centrifugation (10 min at 1700Âg), the supernatant was lyophilized and stored at À20 C until use. The lyophilized powder (20.4 g) of blueberry leaf hot water extracts (BBL-HWE) was suspended in distilled water (1.2 l) and sonicated at 50 C for 30 min. The insoluble material was then removed via filtration. The filtrate was directly subjected to Sephadex LH-20 column chromatography (3.5 cm i.d. Â 23 cm) (Sigma Aldrich, St. Louis, MO, USA), eluted with water containing increasing proportions of methanol (100% water, 20%, 40%, 60%, 80%, and 100% methanol), 60% acetone, and finally eluted with 100% acetone. The eluate was collected in an Erlenmeyer flask (1000 ml in each) to yield ten fractions (BBFr0eBBFr9), concentrated with a vacuum evaporator, and completely freeze-dried. The lyophilized powder (100 g) of blueberry stem hot water extracts (BBS-HWE) was fractionated in the same manner as above.
To quantify PAC, the fractions were measured using the butanol/ HCl method [5]. Each fraction was quantitated from 0.5 mg/ml (for Fr0eFr4) or 0.2 mg/ml (for Fr5eFr7) in methanol. Briefly, 400 ml of the solution was mixed with 3.5 ml of n-butanol/HCl (95:5, v/v) and 100 ml of 2% (w/v) NH 4 Fe(SO 4 ) 2 $12H 2 O in 2 M HCl and heated in an oil bath at 105 C for 30 min. After heating, the reaction mixture was cooled in water for 15 min and the absorbance was measured at 550 nm. PAC content was expressed in milligrams of cyanidin chloride equivalent/g dry weight.
The mean polymerization (mDP) and composition of PAC were estimated by thiolysis followed by reverse-phase high-performance liquid chromatography (HPLC). Fr4 to Fr8 (2.5 mg) were dissolved in 500 ml ethanol solution consisting of 5% (v/v) 2mercaotoethanol, 4% (v/v) 0.5 M HCl, and 32% (v/v) H 2 O, and the mixture was heated at 70 C for 7 h. The resulting thiolytic products were filtered and the filtrate was injected into the HPLC system. HPLC was carried out on a Q-Exactive mass spectrometry system (Thermo Fisher Scientific, Waltham, MA, USA) equipped with a UVevisible detector and an ODS Hypersil C18 column (4.6 mm$250 mm, 5 mm, Thermo Fisher Scientific). The separation conditions were as follows: flow rate, 0.8 ml/min; elution solvent, A (0.1% formic acid in water) and B (acetonitrile); and gradient program, 10%e28% B from 0 to 30 min, 28%e82% B from 30 to 40 min. UV detection was performed at 280 nm wavelength. The mean degree of polymerization (mDP) of PAC was calculated as follows: mDP ¼ [sum of (2-hydroxyethylthio adducts $ n) þ sum of (free flavan-3-ol $ n)]/[total free flavan-3-ol], where n represents the degree of polymerization of the detected flavan-3-ol by thiolysis. Average percentage of A-type bonds ¼ [sum of (thiolysis compounds containing A-type bonds $ n)/[total free flavan-3-ol]/ mDP Â 100. Average percentage of cinchonain I units ¼ [sum of (thiolysis compounds containing cinchonain I unit $ n)]/[total free flavan-3-ol]/mDP Â 100.
After 2 h, the inoculum was removed, washed with PBS, and replaced with fresh culture medium without BB fraction. After incubation of 24 h, the copy number of viral RNA in the culture supernatant was quantified by real-time RT-PCR, as previously described [14], with some modifications. In brief, 5 ml of culture supernatant was mixed with 5 ml of 2 Â RNA lysis buffer [2% Triton X-100, 50 mM KCl, 100 mM Tris-HCl (pH 7.4), 40% glycerol, 0.8 U/ml recombinant RNase inhibitor (Takara, cat# 2313B)] and incubated at room temperature for 10 min. The samples were diluted with RNase-free water (90 ml), and the diluted sample (2.5 ml) was used as the template for realtime RT-PCR performed according to the manufacturer's protocol using the One Step TB Green PrimeScript PLUS RT-PCR kit (Takara, cat# RR096A) and the following primers: 026NeF, 5 0 -AGCCTCTTCTCGTTCCTCATCAC-3'; and 026N-R, 5 0 -CCGCCATTGC-CAGCCATTC-3'. Fluorescent signals were acquired using a Quant-Studio 3 Real-Time PCR system (Thermo Fisher Scientific).

Cytotoxicity assay
To test biological activity, each BB fraction was dissolved in dimethyl sulfoxide (DMSO; Nacalai Tesque, Japan) at a concentration of 100 mg/ml (stock solution). Serial 4-fold dilutions of BB, ranging from 1000 mg/ml to 0.25 mg/ml, were prepared in DMEM containing 10% FBS, penicillin (100 units/mL), streptomycin (100 mg/mL), and 1 mg/ml G418 (Gibco). Diluted DMSO was used as the negative control. The cytotoxic effect of each extract was tested using the Cell Counting Kit-8 (CCK-8; Dojindo, Japan) assay following the manufacturer's instructions. The absorbance was measured at 450 nm using a microplate reader (Bio-Rad, Japan). The concentration of the BB fraction that produced a 50% reduction in cell viability (cytotoxic concentration, CC50) and the concentration that maintained >95% cell viability, the safe concentration, were calculated by regression analysis as described previously [15]. A solution below the calculated safe concentration of each BB was used in the antiviral assays.
(The remaining materials and methods are in the supplementary file.)

Purified fractionation of hot water extracts from leaves and stems of blueberry Kunisato 35 gou (K35)
We previously reported the inhibitory effects of PACs from rabbit-eye blueberry leaves on HTLV-1-infected T lymphocytes [8], ATLL, and HCV virus [5]. Recently, we developed Kunisato 35 gou (K35) from rabbit eye blueberries (Vaccinium virgatum Aiton) for leaf tea production [10] and examined its effect on the SARS virus.
To investigate the chemical structure of PAC, hot water extraction (BB-HWE) was carried out according to a previously published method (reference), and the extracts were fractionated using eluting solvents according to standard methods (Tables S1 and S2). To investigate the chemical structure of PAC, the DMACA and butanol hydrochloric acid methods were used to determine the PAC content in each fraction (Tables S3 and S4). The mean polymerization (mDP) and composition of PAC were estimated by thiolysis followed by HPLC. The results showed that Fr4eFr8 contained higher amounts of PAC. The mDP of PAC was the highest of 7.88 Fr8 from the stem and 12.28 Fr7 in the leaves, resulting in much higher degree of polymerization of the leaves than that of the stems. Regarding the structure of PAC, the content of cinchonain I as an extension unit and A-type bond was quite small in one unit of PAC, and most of them were B-type bonds (Table S5). Interestingly, the structure of PACs from the stem showed mostly B-type bonds, whereas PACs from leaves were characterized by slightly more Atype bonds and cinchonain I units. Based on these results, we next investigated the antiviral activity of Fr4 to Fr7, which had higher PAC contents in the same fractions from the leaves and stems.

Purified fractions containing highly polymerized PAC from blueberry hot water extracts significantly inhibit SARS-CoV-2 replication
To investigate the inhibitory effect of HWE and Fr4eFr7, the viral RNA expression rate in SARS-CoV-2 infected VeroE6/TMPRSS2 cells was determined. The results showed that Fr7 from BB-Stem or BB-Leaf, containing highly polymerized PAC, showed the strongest inhibition of viral RNA expression (Fig. 1A). In particular, the inhibitory activity of BB-Leaf Fr7 was over 100 times higher than that of EGC as a control, and that of the stem component (Fr6 and Fr7) was more than 10 times higher. Therefore, the fraction with the highly polymerized PAC was found to have a stronger inhibitory effect against SARS-CoV-2.
Therefore, IC 50 was determined by adding various concentrations of BB-Leaf Fr7 and BB-Stem Fr7 (Fig. 1B). The IC 50 of BB-Leaf Fr7 was 1.52 mg/ml and 1.39 mg/ml for BB-Stem Fr7. Based on these results, subsequent experiments were conducted using BB-HWE and BB-Fr7 media.

Inhibitory effect of BB-Fr7 solution on viral replication of SARS-CoV-2 by plaque-forming assay
Because the addition of high concentrations of BB-Fr7 solution to VeroE6/TMPRSS2 cells caused cytotoxicity, we decided to use a BB-Fr7 solution of less than 50 mg/ml, which does not cause cytotoxicity in the experimental system (Fig. S1). To determine the suppressive activity of viral replication on SARS-CoV-2, we conducted a viral plaque assay at a concentration of 50 mg/ml of BB-Leaf or Stem Fr7. The results showed that the number of plaques was significantly reduced in BB-Leaf Fr7-and Stem Fr7 treated cells ( Fig. 2A). We also examined the concentration-dependent inhibition of plaque formation. The results showed that the IC 50 values were 1.08 mg/ml for leaves and 1.92 mg/ml for stems (Fig. 2B).
Therefore, BB-Fr7 was found to inhibit SARS-CoV-2 replication directly or indirectly.

Blueberry-derived highly polymerized PAC components can significantly inhibit viral replication by prophylactic administration
To determine the mechanism by which BB-Fr7 inhibits SARS-CoV-2 replication, we applied BB-Fr7 to virus-infected cells at different time points. As shown in Fig. S2, we applied BB-Fr7 as a pre-infection treatment (Pre), for 2 h during infection (On), and 72 h after infection (Post). The results showed that pre-infection treatment (Pre) was the strongest inhibitor of viral replication. Furthermore, a significant suppression of viral replication was observed during and after infection (Fig. 3A), suggesting that highly polymerized PAC, the main component of BB-Fr7, has multiple inhibitory functions against SARS-CoV-2 infection before and after infection. In particular, pre-infection treatment showed stronger suppression of SARS-CoV-2 replication, indicating that prophylactic administration of BB-Fr7 was the most efficient method of administration. To understand the mode of inhibitory activity against infection, we tested whether BB-PAC could directly target SARS-CoV-2 virus particles. To this end, we mixed BB-Fr7 with the SARS-CoV-2 spike (S) protein-pseudotyped lentiviral luciferase reporter virus for 1 h and inoculated them into HeLa-ACE2-TMPRSS2 cells. All BB-PAC Fr7 from leaves, stems, and Fr4 did not suppress pseudovirus infection, even at higher concentrations (100 mg/ml) (Fig. 3B). Thus, these results suggest that BB-PAC does not directly target the viral spike protein in defense against infection.

Blueberry-derived highly polymerized PAC components protects against SARS-CoV-2 infection by inhibiting ACE2 and viral protease
Next, we examined the inhibitory effects of the BB-PAC fraction on the angiotensin-converting enzyme 2 (ACE2) receptor and transmembrane protease, serine 2 (TMPRSS2), which are essential for SARS-CoV-2 infection on the plasma membrane. The enzymatic activities of ACE2 and TMPRSS2 were examined using BB-PAC Fr7 from the leaves or stems and Fr1 as a control Fr1 had no inhibitory effect on ACE2, while Fr7 had an inhibitory effect with 24.2 mg/ml as IC 50 for leaves and 27.0 mg/ml for stems, respectively (Fig. 4A). On the other hand, Fr1 showed no inhibition for TMPRSS2, while Fr7 showed 66.5 mg/ml for leaves and 67.5 mg/ml for stems as IC 50 (Fig. 4B). The inhibitory activity of TMPRSS2 was observed at higher concentrations, resulting in weaker inhibitory activity of TMPRSS2 compared to the IC 50 of ACE2.
In addition, the effect of BB-PAC on the protease activity of SARS-CoV-2 was examined, as viral replication was suppressed by BB-PAC even after infection. In addition, since several natural compounds have been investigated using computational methods to inhibit proteases, proanthocyanidins have been identified as promising drug targets for the main protease (Mpro)/chymotrypsin-like cysteine protease (3CLpro) [16]. Therefore, we investigated the enzyme inhibition ability of our purified BB-PAC fractions (Fr1 as a control and Fr7) using a 3PLpro assay kit (Fig. 4C). Fr1, the control from leaves and stems, showed no inhibitory effect, while Fr7 showed an inhibitory effect at both, 14.5 mg/ml for leaves and 18.4 mg/ml for stems. These results indicate that BB-PAC inhibits viral replication by blocking the ACE2 receptor as a defense against viral infection and inhibiting the main viral protease in the cell.

Discussion
In this study, we report that extracts containing high levels of highly polymerized PACs from blueberry leaves or stems are effective in protecting against SARS-CoV-2 infection. We created Kunisato 35 gou (K35) from rabbit eye blueberries that contained a large amount of highly polymerized PAC in their leaves and stems for use in these experiments. The PAC structure of rabbit-eye blueberry leaves (Vaccinium virgatum Aiton) has been reported that the structure of polymerized PAC is mainly B-type bonds but there is also type A bonds and cinchonain I units [11]. In this study, we analyzed the structure of PAC from K35, and the degree of polymerization in leaves and stems was 5e12, and the structure also contained mostly B-type bonds with type A bonds and cinchonain 1 units. Moreover, PAC from leaves or stems was also found to have a strong inhibitory effect on SARS-CoV-2 replication and was found to be maximally effective when treated prior to infection, indicating that it should be used prophylactically.
As a result of our investigations using BB-PAC, we have found a wide range of antiviral effects, including those against HTLV-1 and HCV, as well as antioxidant effects, anti-inflammatory effects, and anti-obesity effects due to its effects on fat metabolism [5e10]. In the present study, we showed that BB-PAC inhibits the ACE2 enzyme as a receptor for the SARS virus [17], suggesting that BB-PAC may function in defense against infection by blocking viral entry into the cell. In addition, the BB-PAC fraction inhibited 3CL protease, which encodes an essential 3CL protease (3CLpro or Mpro) that processes polyproteins, making it an attractive target for viral inhibitors [18]. Therefore, BB-PAC shows inhibition with respect to these two major therapeutic targets, ACE2 receptor on the cell membrane and viral 3CL protease and has the potential to act as an effective inhibitor of SARS-associated coronavirus infection. Therefore, BB-PAC can be considered a safe and regular protective agent against the spread of COVID-19 as a food ingredient [19]. Of course, further validation is needed, but future development should be aimed at using it not only in COVID-19 but also in general coronavirus infections, including the common cold.

Declaration of conflicts of interest
The authors declare no conflicts of interest.