A rapid detection method for SARS-CoV-2 spike protein is essential for control of COVID19. We investigated various combinations of engineered nanobodies in a sandwich ELISA to detect the Spike protein of SARS-CoV-2. We have identified an optimal combination of nanobodies. These were selectively functionalised to further improve antigen capture. This dataset contains data from ELISA experiments described in the manuscript.                                                                                                                                          

Plate coating of nanobodies for ELISA by passive adsorption vs biotinylation was compared. A series of nanobody pairings (two cluster 2 ACE2-binding epitope and two cluster 1 CR3022 epitope) were screened for optimum sensitivity. The optimal pair were then tested against a series of SARS-COV-2 antigens: recombinant spike 1 protein; recombinant receceptor binding domain (RBD); pseudotyped HIV-1 and heat-empigen inactivated SARS-CoV-2 virus. X-ray irradiated SARS-CoV-2 was also tested. Sensitivity to these antigens was compared with nanobodies biotinylated a) site-selectively and b) in a non-specific stochastic manner. Batch-to-batch viral variation and effects of inactivating agents were investigated. Limit of detection was compared against delta and beta viral mutants. Combining optimal nanobody pairing and site-selective biotinylation, we observed a limit of detection of 147 pg/mL for Spike protein; 33 pg/mL for RBD; 16 TCID50/mL of pseudovirus and 15 ffu/mL of heat-Empigen inactivated SARS-CoV-2. The pairing also showed sensitivity towards delta variant. We have demonstrated the use and sensitivity of nanobodies in ELISA by detection of recombinant and viral SARS-CoV-2 antigens.

The data supplied is raw absorbance data for ELISA. The data was collected on a Spectramax M3 microplate reader (Molecular Devices).
The data was processed using Prism software. Variable slopes were modelled using Prism's sigmoidal 4-parameter logistic curve. To determine slope gradient, values in linear range were modelled using simple linear regression. Standard deviation was calculated using the built-in prism analysis.

All ELISA data is supplied in one Prism file, each dataset is given in the order displayed in the main manuscript, followed by ELISA data from supporting information.
A word document describing which figure relates to which prism dataset is also supplied for clarity.

Some anomalous replicates were omitted for the final analyses; these are included in the supplied datasets and highlighted in blue, and are: Figure 3, for Biotinx-C5-Fc/C1-Fc-HRP: all third replicates. Figure 4b (same data also in SOI Figure 2) - one replicate value is removed.