Quantifying neutralising antibody responses against SARS-CoV-2 in dried blood spots (DBS) and paired sera

Abstract

The ongoing SARS-CoV-2 pandemic was initially managed by non-pharmaceutical interventions such as diagnostic testing, isolation of positive cases, physical distancing and lockdowns. The advent of vaccines has provided crucial protection against SARS-CoV-2. Neutralising antibody (nAb) responses are a key correlate of protection, and therefore measuring nAb responses is essential for monitoring vaccine efficacy. Fingerstick dried blood spots (DBS) are ideal for use in large-scale sero-surveillance because they are inexpensive, offer the option of self-collection and can be transported and stored at ambient temperatures. Such advantages also make DBS appealing to use in resource-limited settings and in potential future pandemics. In this study, nAb responses in sera, venous blood and fingerstick blood stored on filter paper were measured. Samples were collected from SARS-CoV-2 acutely infected individuals, SARS-CoV-2 convalescent individuals and SARS-CoV-2 vaccinated individuals. Good agreement was observed between the nAb responses measured in eluted DBS and paired sera. Stability of nAb responses was also observed in sera stored on filter paper at room temperature for 28 days. Overall, this study provides support for the use of filter paper as a viable sample collection method to study nAb responses.

Figure 1

Serum stored on filter paper retains neutralising capacity against SARS-CoV-2 spike. Pseudo-virus particles (PVP) expressing SARS-CoV-2 spike were used to measure the neutralisation capacity of paired human sera stored in direct aliquots and eluted from dried serum spots (DSS) stored on filter paper. Neutralisation activity was defined as the serum dilution that reduced PVP infectivity by 50%, 70% or 90% (IC50, IC70 or IC90, respectively). (a–c) Slope charts from left to right display on the y-axis PVP neutralisation as (a) IC50, (b) IC70 and (c) IC90. The x-axes show the sample type with sera represented by orange circles and DSS eluates represented by blue circles. IC50, IC70 and IC90 values for paired sample types are connected via grey lines. The dotted lines across the charts represent the lower limits of detection, with the limit for direct sera in orange and DSS in blue (d–f) Scatter plots from left to right show direct sera IC50, IC70 and IC90 plotted against DSS eluates IC50, IC70 and IC90 values. Simple linear regression analysis were performed and found significant positive relationships between sera aliquots and DSS eluates IC50 values [n = 43, R² = 0.7432, p < 0.0001], IC70 values [n = 47, R² = 0.8381, p < 0.0001] and IC90 values [n = 42, R² = 0.8286, p < 0.0001]. The dotted lines represent the upper and lower 95% confidence intervals (CI) of the line of best fit. (g–i) Bland–Altman plots display on the x-axes the average IC50 values (n = 43), IC70 values (n = 47) and IC90 values (n = 42) for sera aliquots and DSS eluates and the difference between the two values on the y-axes. Red lines represent bias, and the dotted lines represent the upper and lower 95% CI.

Figure 2

Neutralising capacity of serum against SARS-CoV-2 spike is not significantly affected by storage on filter paper at room temperature for up to 28 days. Single-round infectious pseudo-virus particles (PVP) expressing SARS-CoV-2 spike were used to measure the neutralisation capacity of 7 human sera samples (squares-CCP-UK01, downwards traiangles-CCP-UK10a, upwards triangles CCP-UK10b, circles-CCP-UK05, diamonds-CCP-UK16, hexagons-CCP-UK17 and stars-CCP-UK31). Sera were stored on filter paper kept at room temperature (RT) as dried serum spots (DSS) for 0–28 days before elution and compared to sera stored as direct aliquots at − 80 °C (serum). Due to availability of sera only 4 out of 7 had DSS left at RT for 28 days. Bar charts a-f show neutralisation activity on the y-axes defined as the serum dilution that reduced PVP infectivity by 50% (a and d), 70% (b and e) and by 90% (c and f) (IC50, IC70 or IC90, respectively). Error bars represent standard deviation from the mean. Wilcoxon t-tests were run, and no significant differences (ns) were found between the mean IC values for serum controls and the mean IC values for DSS stored at RT from 2 to 28 days (p > 0.05).

Figure 3

Venous whole blood stored on filter paper shows comparable serum neutralisation of SARS-CoV-2 spike to paired sera aliquots. Pseudo-virus particles (PVP) expressing SARS-CoV-2 spike were used to measure the neutralisation capacity of paired human sera with venous blood eluates that were stored on filter paper as dried blood spots (VDBS). Neutralisation activity was defined as the serum dilution that reduced PVP infectivity by 70% (IC70). As whole blood contains approximately 55% serum this was accounted for when calculating IC70 values for VDBS eluates. (a) Violin plot displays on the y-axis PVP neutralisation as IC70 values. The x-axis shows paired sample types (n = 10) with sera represented by orange circles and 3 pipetted volumes of VDBS (15 μL, 25 μL & 50 μL) represented by pink circles. The dotted lines across the plot represent the lower limits of detection, with the limit for sera in orange and VDBS eluates in pink. A repeated measures one-way ANOVA test was run to compare the mean IC70 value for sera against the mean IC70 values for VDBS eluates. A significant difference was found (n = 10, p < 0.0001) and a Holm-Šidák’s multiple comparisons test identified it to be between the mean sera and mean 15 μL VDBS eluates IC70 values (p < 0.0001). No significant differences (ns) were found between the mean IC70 values for sera and the 25 μL and 50 μL VDBS eluates (p > 0.05) (b-d) Scatter plots show the IC70 values for sera plotted against the IC70 values for VDBS eluates (b) 15 μL VDBS (c) 25 μL and (d) 50 μL VDBS. Simple linear regression analysis found significant relationships between IC70 values for sera and 15 μL VDBS eluates [n = 10, R² = 0.6780, p = 0.0034], 25 μL VDBS eluates [n = 12, R² = 0.7429, p = 0.0003] and 50 μL VDBS eluates [n = 12, R² = 0.8655, p < 0.0001]. The dotted lines represent the upper and lower 95% confidence intervals (CI) of the line of best fit. (3e–3g.) Bland–Altman plots display on the x-axes the average IC70 values for sera and VDBS eluates (e) 15 μL (f) 25 μL and (g) 50 μL and the difference between the IC70 values on the y-axes. The red lines represent bias, and the dotted lines represent the upper and lower 95% CI.

Figure 4

Dried blood spots obtained via fingerstick show comparable serum neutralisation of SARS-CoV-2 spike to paired serum aliquots. Pseudo-virus particles (PVP) expressing SARS-CoV-2 spike were used to measure the neutralisation capacity of human sera and eluted dried blood spots obtained via fingerstick (FDBS) for 11 participant samples. Neutralisation activity was defined as the serum dilution that reduced PVP infectivity by 50%, 70% or 90% (IC50, IC70 or IC90, respectively). As whole blood contains approximately 55% serum this was accounted for when calculating IC values for FDBS eluate. For each FDBS the exact volumes of blood blotted from the participants’ fingers were not measured therefore two volumes were estimated for FDBS volume. The dotted lines across the graph represent the lower limits of detection, with the limit for sera in orange, DSS eluate in blue,  ≈ 15 μL FDBS eluate in yellow and ≈ 25 μL FDBS eluate in brown.