Here, we report two separate sets of analyses. First, we report a retrospective, observational statistical population analysis of COVID-19 prevalence and mortality with respect to age. Second, we perform a retrospective analysis on sera collected from 204 individuals to explore the molecular basis for similarities between poliovirus and SARS-CoV-2.

Written informed consent was obtained from all participants. Among those who had recently been inoculated with IPV, adults had been immunized with a 0.5 mL booster shot (standard dosage in the United States) and children had been vaccinated in four stages according to age-defined childhood vaccination protocols, set by the Centers for Disease Control and Prevention (CDC). IRB approval for the collection of adult sera was obtained from BioMed IRB, and pediatric samples were purchased from Discovery Life Sciences (Huntsville, AL) in accordance with the approved IRB protocol Alabama Biobank Research Trial 001 V3 (ABRT 001-V3).

In order to evaluate if there was a relationship between countries' median age and prevalence or mortality of COVID-19, data on the total number of cases and deaths from the Johns Hopkins University Coronavirus Resource Center were compiled for the 100 countries with the highest prevalence of SARS-CoV-2 (13). Prevalence and death rates were then calculated for each country based on total population size, as of February 10, 2021. Median age for each country was then included for correlation analyses, calculated using the “corrplot” package from R Studio software, version 1.4.1103 (https://cran.r-project.org/index.html), between median age and SARS-CoV-2 prevalence and death rates.

The full length nsp12 gene was cloned into an AmeriDx® E. coli over-expression vector pADX75. Similarly, the poliovirus RdRp polyoprotein region 5286bp was amplified to 6564bp (NCBI accession number ALP31139) and cloned into the pADX75 vector. These two vectors were transformed into the BL21 strain respectively. The BL21 strains harboring the vectors were selected and used for protein expression and purification, according to standard protein purification protocols.

An enzyme-linked immunosorbent assay (ELISA) was used for a quantitative analysis of the virus antibody in human sera. ELISA tests were carried out according to standard lab operating procedures. Briefly, the poliovirus vaccine (IPV)'s RdRp full length protein and the SARS-CoV-2 RdRp antigen were coated on the ELISA plate, respectively, and the human serum harboring the primary antibody was applied to each well. The primary antibody bound to the antigen coated on the well, and then the horse radish peroxidase (HRP) conjugated goat anti-human IgG/A/M secondary antibody, which recognizes the primary human antibody, formed a complex with the antigen, antigen-antibody-secondary antibody-HRP. TMB substrate was then added to the ELISA plate. Once the HRP enzyme turned the solution color to blue, the reaction was stopped by adding 2 M sulfuric acid, which converted the solution color to yellow. An ELISA reader measured and recorded the activity. We performed non-linear regression on the raw data using a four-parameter fitting algorithm using the R package “drc,” (14) which resulted in relative serum response units (reported as “arbitrary units”).

Recombinant RdRp proteins were separated by 4–20% SDS-PAGE (Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis), and then were transferred to the PVDF membrane. Mouse monoclonal antibody 4E6 was added as 1:1000 in Western blot buffer TBST, and the antibody bound to antigens on the membrane. This binding was captured using a HRP conjugated goat anti-mouse IgG/A/M secondary antibody that conjugated with HRP. The HRP's chemiluminescence substrate then displayed a change of color for the specific RdRp protein band.

RdRp activity was determined in vitro by a fluorescence method, which was modified from a technique developed by others for Zika virus RdRp (15). There was a sample vial and a control vial for each sample, where water replaced ribonucleoside-tri-phosphates (rNTPs). The enzymatic reaction consisted of 0.2 M Tris-HCl, pH 8.0, 0.125 M NaCl, 40 mM MgCl₂, 10 mM spermidine-(HCl)₃, 2.5 uM rNTPs, 2.5 mM DTT, and 40 nM Syto-82 fluorescence dye. The sample serum was diluted to 1,800 times with 1xPBS, and then 2 uL were added to both vials. Samples were incubated at 37°C for 30 min. The newly formed double stranded RNA then bound to the Syto-82 fluorescence dye, and emitted 589 nm light that could be recorded by a GeneScan^TM fluorescence reader. The fluorescence signal in each vial was quantified using the mean gray value in an ImageJ program (https://imagej.nih.gov/ij/index.html) (16). The fluorescence intensity difference between these two vials was designed to reflect the RdRp's relative activity, such that a positive value represented up-regulated enzymatic activity (normal replication, causing an individual to become ill), and a negative value represented down-regulated enzymatic activity (replication inhibited, preventing viral effects).

A cytopathic effect (CPE)-based antiviral assay was performed in vitro by infecting Vero-E6-cells in the presence or absence of immune sera to evaluate antiviral activity against SARS-CoV-2 (MEX-BC2/2020) (17). Most of the pediatric serum samples were <1 mL each, and thus not enough for a single CPE test assay, so we pooled samples from subjects of the same age. Vero E6 cells were seeded and incubated for 24 h. Sera were either pre-incubated first with target cells for 1 h at 37°C before infection with SARS-CoV-2, or cells were treated with SARS-CoV-2 for 3 h at 37°C to allow viral adsorption, and then the serum specimens were added to the target cells without removing the viral inoculum. The cells were challenged with the viral inoculum and re-suspended in DMEM with 2% FBS (DMEM2). The test sera and virus were maintained in the cell culture media for 96 h. As a surrogate marker for inhibition of viral replication, inhibition of SARS-CoV-2-induced CPE was measured using uninfected Vero E6 cells as mock control, and the cells infected with SARS-CoV-2 virus only (no serum) as vehicle control. Cell viability was monitored with the neutral red (NR) uptake assay. The average absorbance at 540 nm (A540) from each well on a 96-well plate was calculated in percentage using the following formula:

where vehicle is the O.D. from virus-infected cells without any testing serum samples, and mock is the O.D. value from the uninfected cells.

We first sought to determine whether there are age-related differences in COVID-19 prevalence and mortality by performing a retrospective literature study. Using data from the 100 countries with the highest rates of SARS-CoV-2, correlation analyses show there is a positive association between a country's median age and prevalence rate of SARS-CoV-2 (Pearson r = 0.546, p < 0.001). There is also a positive correlation between median age and the SARS-CoV-2-related death rate (Pearson r = 0.545, p < 0.001) (Figure 2). Countries with older national median ages were significantly more likely to have a higher prevalence of SARS-CoV-2, as well as a higher SARS-CoV-2-related death rate. While such an analysis has limitations, this result nevertheless suggests that there is indeed a statistically significant positive association between age and COVID-19 prevalence and mortality.

Having shown that COVID-19 prevalence and mortality is age dependent, we next sought to determine what role—if any—previous inoculation with poliovirus may play. Thus, we explored whether antibodies generated by SARS-CoV-2 RdRp protein could recognize recombinant poliovirus RdRp antigens. The tertiary and quaternary structures of both the RdRp from poliovirus and SARS-CoV-1 are right-handed palm type (11), and SARS-CoV-1 and SARS-CoV-2 have similar structures (18). As shown in Figure 3, both RdRp can be recognized by SARS-CoV-2 RdRp monoclonal antibody 4E6.

To test the hypothesis that poliovirus vaccination induces an immunity to SARS-CoV-2, we analyzed 204 serum samples from 119 individuals who had not recently been vaccinated with IPV, referred to as “pre-immunized samples,” and 85 individuals who had recently received IPV, referred to as “post-immunized individuals,” for anti-RdRp antibody titers. While the immune response to antigenic RdRp protein varies from person to person, sera from those who had been vaccinated for poliovirus showed a significant increase in antibody titers (Two-tailed t-test, p < 0.001, Figure 4A). Additionally, we compared a late-stage immunized pediatric group (children who had received at least three of the four CDC recommended polio dosages, depending on their age, N = 62, age 1–10) to a recently immunized adult group (N = 76 of the total 85 post-immunized samples, age 25–89) and observed no significant difference in antigenic specificity between pediatric and re-immunized adult sera (Two-tailed t-test, p = 0.85), indicating immunized adults display similar antibody responses as children that have received their childhood vaccinations (Figure 4B).

Having observed that the poliovirus vaccine significantly affects SARS-CoV-2 RdRp reactivity, our next goal was to understand the mechanism of action behind this phenomenon. Inhibition of SARS-CoV-2-induced cytopathic effect (CPE) in Vero-cell culture was used as a surrogate marker for the inhibition of viral replication. Polio-immune serum demonstrated an antiviral effect in Vero cells at dilutions 1:8 to 1:32, which was stronger when the antisera were pre-incubated with cells, rather than added after viral adsorption. Samples from children fully immunized with IPV (NS-1) and young adults (NS-3A and NS-3B) show the strongest inhibition of viral CPE when added to cell culture prior to virus (Figure 5). The sample labeled NS-1 was from pooled sera from children who were on average 1 years old, collected prior to the onset of the SARS-CoV-2 pandemic; these children would have been immunized against polio as part of national immunization campaigns. NS-2 samples were pooled from infants who were on average 75 days old; these infants would not have been fully immunized against polio. The remaining samples depicted in Figure 5 were randomly selected from the adult pool of paired samples to demonstrate differences in immunity, based on age and IPV immunization. Sample NS-3A is from a 27-year-old subject, who was recently given IPV. The NS-3B sample is from the same subject after obtaining a second IPV booster. Sample NS-4A is serum from a 63-year-old female subject pre-IPV vaccination; NS-4B is her post-vaccination serum sample. The NS-5 sample represents serum from a 14-year-old subject, who had not recently been vaccinated with an IPV booster.

Pediatric serum samples were pooled in order to perform the analysis for infants, as shown in Figure 5 (NS-1 and NS-2). While a lack of individual samples precludes statistical analysis, we are able to observe a qualitative trend that children who have received their childhood vaccinations by 1 year of age display stronger immunity to SARS-CoV-2, as evidenced by CPE levels, compared to children that are <4 months old and still very early in immunization protocols (NS-1 and NS-2, respectively). This result is expected if, as we propose, poliovirus immunization plays a role in the relative resistance to COVID-19 in younger populations. Samples NS-3A and NS-3B further demonstrate how polio immunization (a single inoculation, followed by a second booster 2 months later, respectively) inhibits SARS-CoV-2-induced CPE.

Protection from poliovirus or SARS-CoV-2 declines as one ages. No inhibition of SARS-CoV-2-induced CPE is evident pre-vaccination (NS-4A), though older adults' immune systems can respond to an IPV booster (NS-4B). Here, data show approximately a 35% increase of protection from viral CPE, relative to pre-vaccination levels. While the immunity is lower in the 63-year-old adult, compared to the 27-year-old adult, IPV provides some protection that the teenager (NS-5), who was not recently vaccinated, does not have.

Finally, we explored RdRp as a possible therapeutic target by evaluating the effect of polio-immune serum on SARS-CoV-2 RdRp activity in living cells. Randomly-selected sera from 17 different polio-immunized subjects (seven children; 10 adults) were added to the RdRp reaction mix at a dilution of 1:1800. RdRp activity was then measured by fluorescence signal (Figure 6), with a positive value representing regular enzymatic RdRp activity (virus will replicate normally and cause the individual to become ill) and a negative value representing inhibition of the RdRp enzyme (sera from immunized individuals reduce the possibility of viral replication and inhibit the virus from entering the cell). Of the 17 randomly-selected sera from immunized individuals, 13 (76.5%) were effective in inhibiting the enzymatic activity of RdRp.