Human seasonal coronavirus neutralization and COVID-19 severity

Abstract

The virus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), responsible for the global coronavirus disease-2019 (COVID-19) pandemic, spread rapidly around the world causing high morbidity and mortality. However, there are four known, endemic seasonal coronaviruses in humans (HCoVs), and whether antibodies for these HCoVs play a role in severity of COVID-19 disease has generated a lot of interest. Of these seasonal viruses NL63 is of particular interest as it uses the same cell entry receptor as SARS-CoV-2. We use functional, neutralizing assays to investigate cross-reactive antibodies and their relationship with COVID-19 severity. We analyzed the neutralization of SARS-CoV-2, NL63, HKU1, and 229E in 38 COVID-19 patients and 62 healthcare workers, and a further 182 samples to specifically study the relationship between SARS-CoV-2 and NL63. We found that although HCoV neutralization was very common there was little evidence that these antibodies neutralized SARS-CoV-2. Despite no evidence in cross-neutralization, levels of NL63 neutralizing antibodies become elevated after exposure to SARS-CoV-2 through infection or following vaccination.

Keywords: SARS coronavirus; endemic infection; epidemiology; immune responses; neutralization; virus classification.

Figure 1

Neutralization IC₅₀ values for HCoVs and SARS‐CoV‐2 (A–C). Solid lines represent geometric means, dashed horizontal lines indicate the cutoff chosen to define detectable HCoV neutralization. Due to the different cell lines used for HKU1 (B) and 229E (C), data points were plotted on separate graphs. 98.6% of 282 plasma samples neutralized NL63 (A), 76.4% of 89 samples neutralized HKU1 (B), and 99% of 100 samples neutralized 229E (C). The SARS‐CoV‐2 data only includes samples from seropositive individuals. Panel (D) shows the percentage of samples with detectable HCoV neutralization, IC₅₀> 40. HCoVs, human coronaviruses; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 2

HCoV neutralization by demographic. Neutralizing IC₅₀ value for HCoVs against sex (A) and age (B). Black horizontal lines represent geometric means, colored lines in are simple linear regression lines. Sample sizes: NL63 n = 84, HKU1 n = 43, 229E n = 47. We found a significant difference in HKU1 neutralization between sex (p = 0.004) and age (p = 0.039). We did not see any statistical significance in HCoVs NL63 nor 229E between sexes. We observed a significant difference in 229E neutralization and age (p = 0.037). *p < 0.05, **p < 0.01 ***p < 0.001. HCoVs, human coronaviruses

Figure 3

Comparing neutralization between seropositive HCWs and seropositive patients for NL63 (A) (n = 84) (p = 0.150), HKU1 (B) (n = 43) (p = 0.162), 229E (C) (n = 47) (p = 0.155), and SARS‐CoV‐2 (D) (n = 101) (p < 0.0001). Horizontal black lines indicate geometric means. *p < 0.05, **p < 0.01 ***p < 0.001. HCWs, healthcare workers; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 4

Partial residual plots for the natural log of NL63 (n = 91) (A) and SARS‐CoV‐2 (n = 91) (B) pMN IC₅₀ values predicting COVID‐19 severity. These plots illustrate the effect of a variable on severity after accounting for other variables in the linear regression. This data suggests that NL63 neutralization is associated with COVID‐19 disease severity. COVID‐19, coronavirus disease‐2019; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 5

Comparing neutralization of SARS‐CoV‐2 seropositive and seronegative samples with the HCoVs (A–C). Our results revealed significant differences for NL63 (A) (n = 255) (p = 0.018) and SARS‐CoV‐2 (D) (n = 255) (p = <0.001). However, we observed no significance for HKU1 (B) (n = 75) (p = 0.143) and 229E (C) (n = 86) (p = 0.978). Horizontal black lines indicate geometric means. Samples were grouped as seropositive or seronegative regardless of being from patients or HCWs. Serostatus was based on IgG binding by Luminex assay or by SARS‐CoV‐2 pMN assay where a cut‐off reciprocal IC₅₀ value was derived using pre‐pandemic sera as described in the methods or in Castillo‐Olivares et al. A small number of samples were classed as seropositive by IgG binding assay despite low neutralization. *p < 0.05, **p < 0.01 ***p < 0.001. HCWs, healthcare workers; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 6

NL63 neutralization in HCWs before and approximately 1 month after first dose of SARS‐CoV‐2 vaccination (n = 21) (p = <0.001). *p < 0.05, **p < 0.01 ***p < 0.001. HCWs, healthcare workers; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 7

The relationship between SARS‐CoV‐2 neutralization and NL63 (n = 101) (A), HKU1 (n = 56) (B), and 229E (n = 60) (C) in SARS‐CoV‐2 seropositive samples with a linear line of best fit. Spearman's rank test reveals no significant correlation for each HCoV tested. HCoV, human coronaviruse; SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2

Figure 8

Correlation matrix of virus binding and neutralization. Larger darker circles indicate stronger correlations, as measured by Spearman's rank correlation coefficient. Blue circles indicate positive correlation and red circles indicate negative correlations. Sample sizes: SARS‐CoV‐2; n = 101. NL63; n = 101. HKU1; n = 56. 229E n = 60. SARS‐CoV‐2, severe acute respiratory syndrome coronavirus 2