Atmospheric PM2.5 before and after Lockdown in relation to COVID-19 Evolution and daily Viral Counts: Could Viral Natural Selection have occurred due to changes in the Airborne Pollutant PM2.5 acting as a Vector for SARS-CoV-2?

Genes coding for SARS-CoV-2 have been detected on the microscopic airborne pollutant particulate matter, which has been suggested as a vector for COVID-19 transmission. Lockdown in China has been shown to be associated with significant reduction in pollution including the particulate matter component which coincided with the appearance of a viral mutant (Clade G) which steadily displaced the original Clade D after lockdown. The reason why Clade G developed a fitness advantage is as yet unknown. This paper examines the possible role of airborne particulate matter PM2.5 as selective pressure determining viral Clade predominance and further shedding light on the mode of SARS-CoV-2 transmission.

Methods

The average levels of PM2.5 of a number of cities were obtained from the Air Quality Index (AQI), a real-time assessment of atmospheric pollution. The daily average PM2.5 levels were assessed between January 23rd and April 29 ^th 2020 determined by the timeline when viral counts in Beijing and other cities were available. Daily viral counts of Clades D and G were available starting from the 12 ^th February as determined by the scientific literature published in August 2020. The cities chosen were Beijing, Sheffield, Nottingham, Sydney and Cambridge because of their substantially elevated viral counts compared to other cities. Cities as opposed to vaster areas/nations were chosen as PM2.5 levels vary across regions and countries.

Results

For the time period assessed, the Beijing PM2.5 pattern initiated with highly elevated mean PM2.5 levels of 155.8µg/m3 (SD+/-73.6) during high viral counts, followed by 82.1µg/m3 (SD+/-44.9) (p<0.04) when the viral counts decreased. In all the other cities assessed, the pattern differed whereby the PM2.5 levels increased significantly over the preceding baseline contemporaneously with the viral count rise. The changes in these cities’ PM2.5 levels were on average 31.5µg/m3 before viral counts rose and 56.35µg/m3 contemporaneous with viral count rise. The average levels of PM2.5 in these cities started to decrease one week after lockdown to 46µg/m3 when measured over 2 weeks post-lockdown.

As regards the viral counts from data retrieved from Beijing, the latter part of the bell-shaped curve and a subsequent smaller curve of the viral count was available for evaluation. The average viral count for Clade D in Beijing was 11.1(SD+/-13.5) followed by a mean viral count for Clade G was 13.8(SD+/-9.2). Conversely in all the other cities besides Beijing, the viral counts averaged 45.8 for Clade D and 161 for Clade G. The variation in viral counts between cities suggests the strong possibility of variation in the availability of sampling between cities.

The newer variant, Clade G demonstrated viral counts initially appearing in mid-February in Beijing to later displace Clade D as the dominant viral Clade. The appearance of Clade G coincided with the decreasing gradient of PM2.5 levels. A number of significant correlations were obtained between PM2.5 levels and the viral count in all the cities reviewed.

Conclusion

COVID-19 viral counts appear to increase concomitant with increasing PM2.5 levels. Viral counts of both Clades correlated differentially with PM2.5 levels in all the cities assessed. The significantly highly elevated PM2.5 levels in Beijing resulted in correlating mainly with Clade D, however Clade G began to appear with decreasing PM2.5 levels, suggesting the beginnings for the initial SARS-CoV-2 Clade evolution. Clade G, the newer variant was able to flourish at lower levels of PM2.5 than Clade D. Clade G may possibly have utilized other sources of particulate matter as a viral vector, such as that derived from tobacco smoking, whereby 66% of Chinese males are smokers and 70% of the Chinese non-smoking population are exposed to 2 ^nd hand smoking.