Determination of the Concentration of IgG against the Spike Receptor-Binding Domain That Predicts the Viral Neutralizing Activity of Convalescent Plasma and Serum against SARS-CoV-2

Round 1

Reviewer 1 Report

Line 306 the limit of detection was established at 1 ng/ml. However what is the background of negative human sera run on the test? This is important to be able to identify positive and negative samples and what is the sensitivity and specificity of the ELISA?

Line 315 what was the cutoff used to determine the 92% positive samples. It looks that there a more positive samples around the cutoff.

Can the VNT assay be more sensitive if it is done as a plaque reduction assay?

Author Response

We would like to thank the reviewer for the suggestions that have significantly improved our manuscript.

1- We agree with reviewer that the limit of detection of IgG is not useful unless this limit is calculated using the human negative samples. We how now calculated the limit of detecion of our ELISA employing the  values obtained in the samples from the human real negative donors (LOD= mean+ 3*sd of the values in negative samples collected before 2018) and used this limit (0,08 mg/L= 80 ng/ml) as the cutoff to discriminate between positive and negative test.  In addition we have calculated the sensitivity (71,3%) and specificity (100%) of the ELISA using this cutoff and the 2x2 table method. Al these results have now been included in the MS.

2- As indicated by reviewer viral plaque assay is an alternative to TICD50 to analyse  viral reduction during VNT assays. VNT and TCID50 present advantages and disadvantages . We have not compared boht protocols but in previous studies it has been found a higher sensitivity for TCID50 (J Virol Methods. 2013 Nov;193(2):565-71). In order to increase the sensitivity of our method we performed the TCID50 assay using log2 dilutions of plasma samples so we believe that the sensitivity would not be increased by using viral plaque assay albeit albeit we have not formally tested it. Albeit a direct comparison (as previously performed for other viruses) would be interesting, we fell that performing viral plaque assays will not change the main conclusions of our study correlating the amount of RBD IgG with the VNT activity of convalescent plasma and serum.

Reviewer 2 Report

In this study Santiago et al. present the results of the validation of an in-house quantitative RBD IgG ELISA test and show that the quantity of IgG antibodies specific to the Spike RBD of SARS-CoV-2 vary significantly between convalescent donors, reaffirming the results of prior studies. The authors also show that the concentration of anti-RBD IgG in plasma is higher in convalescent patients who had required hospitalization, suggesting this as a criterion to optimize donor selection. Finally,  anti-RBD IgG plasma concentration were shown to be significantly associated with the plasma neutralizing activity against SARS-CoV-2 in vitro. Overall, even though the findings of this study are not novel, the manuscript is very well written and the results are clearly presented. Minor comments/edits follow:

Comment #1: Lines 88-93: Based on the most recent evidence, EUA has been granted only for high-titer convalescent plasma among hospitalized patients with COVID-19 who are early in the disease course or have impaired humoral immunity. The available evidence does not support a clear role of convalescent plasma in patients with severe disease. Please rephrase this sentence.

Comment#2 Lines 343-348: Previous studies have suggested a correction between the RBD IgG in plasma and duration of symptoms as well as gender (for example PMID 32555388, 32917729). Was this noted in your study? Maybe a multivariate analysis would be helpful here to see if the above are independent predictors of the levels of RBD IgG. 

Comment #3 Could you comment on how the SARS-CoV-2 variants that have emerged (eg the SARS-CoV-2 VOC 202012/01) could affect the performance of your in-house ELISA but also the neutralizing activity of the anti-RBD IgG antibodies?

Author Response

We thank the reviewer for the constructive criticism and all suggestions that have considerably improved our study.

Al changes are highlighted in red in the revised version of our MS.

1- We are sorry for this mistake that has now been modified according to reviewer suggestion.

2- We have analysed other factors like hospitalisation, fever, symptoms and oxygenation but unfortunately we have not recorded the duration of symptoms. Regarding gender, we have now analysed it and confirmed that male patients present significantly higher RBD IgG than female patients. Thank you for this excellent suggestion that has confirmed previous studies. This analysis has been included in figure 2D.

3- This is a very good question, especially since the emergence of new variants in the last months containing mutations in the RBD of Spike that might affect the performance of ELISA tests based on the Spike RBD antigen, as ours. It is very difficult to predict how the efficacy of our current ELISA test to detect IgG in plasma from patients infected with these strains will be affected. Specially, since the available experimental evidences have not provided clear results yet. The B1.1.7 UK original strain contains, at least, the mutation N501Y in RBD although new mutations have been detected in recent studies. The South African B1.351 and Brazilian P1 strains contain two additional mutations in RBD, K417 and E484K. All these mutations might affect the binding of IgG directed against non mutated RBD although as far as we know it has not been formally tested the performance of ELISA test based on wild type RBD to detect IgG generated against mutated RBDs. We can offer some speculations based on the viral neutralizing activity against strains containing RBD mutations of antibodies present in convalescent plasma from COVID19 patients infected with “original” strains or in individuals vaccinated with wild type Spike. We could predict that our ELISA test might still detect B1.1.7 RBD specific IgG as it has been shown by different studies that the presence of single N501Y mutation has low impact of the neutralizing activity of antibodies generated against the original strains including convalescent plasma or plasma from mRNA vaccinated humans. Regarding the presence of E484K mutation present in the B1.351 and P1 strains, some studies have found that its presence reduces the efficacy of antibodies generated by mRNA vaccines although contradictory findings have been reported. In addition, a recent report has shown a significant reduction in the viral neutralizing activity of sera from convalescent COVID19 patients infected with the original strain against the E484K mutation. Thus, our ELISA might not properly detect IgG in plasma from patients infected with B1.351 or P1 strains but still be effective in those from patients infected with B1.1.7 strain. However, to confirm these speculations we should compare the results obtained using plasma from convalescent patients infected with the different strains and a ELISA test based on wild type and mutated RBD antigens. We have now indicated this limitation in the discussion.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.