Validation of a MALDI-TOF MS method for SARS-CoV-2 detection on the Bruker Biotyper and nasopharyngeal swabs. A Brazil - UK collaborative study. (preprint)

We had developed a MALDI-TOF mass spectrometry method for detection of SARS-CoV-2 virus in saliva-gargle samples using Shimadzu MALDI-TOF mass spectrometers in the UK. This was validated in the USA to CLIA-LDT standards for asymptomatic infection detection remotely via sharing protocols, shipping key reagents, video conference and data exchange. In Brazil, more so than in the UK and USA, there is a need to develop non-PCR dependent rapid affordable SARS-CoV-2 infection screening tests, which also identify variant SARS-CoV-2 and other virus infections. Travel restrictions necessitated remote collaboration with validation on the available Clinical MALDI-TOF - the Bruker Biotyper (microflex LT/SH) - and on nasopharyngeal swab samples, as salivary gargle samples were not available. The Bruker Biotyper was shown to be almost log10^3 more sensitive at detection of high molecular weight spike proteins. A protocol for saline swab soaks out was developed and duplicate swab samples collected in Brazil were analysed by MALDI-TOF MS. The swab collected sample spectra varied from that of gargle-saliva in three additional mass peaks in the mass region expected for IgG heavy chains and human serum albumin. A subset of clinical samples with additional high mass, probably Spike-related proteins, were also found. Spectral data comparisons and analysis, subjected to machine learning algorithms in order to resolve RT-qPCR positive from RT-qPCR negative swab samples, showed a 78% agreement with RT-qPCR scoring for SARS-CoV-2 infection.

Degree of immunoglobulin kappa light chain glycosylation of anti-spike SARS CoV-2 antibodies correlates with COVID-19 severity. (preprint)

Glycosylation of antibodies and the effects this has on inflammatory responses has concentrated predominately on the study of glycosylation moieties found in the Fc region of heavy chains. Light chain glycosylation and their ratios are relatively understudied. Nevertheless, variable glycosylation and ratio of {kappa} and {lambda} light chains have been associated with worse prognosis in myeloma and in tissue deposition amyloidosis. The {kappa} {lambda} light chains, of antibodies binding to SARS-CoV2 nucleocapsid and spike protein were analysed, using MALDI-ToF MS, in respect to their intensity, ratios, glycosylation patterns and any pattern changes correlating with COVID-19 severity. The molecular masses and signal intensity of {kappa} and {lambda} glycosylated and non-glycosylated light chains were measured for immunoglobulins isolated from plasma of seropositive and seronegative health care workers (HCW), and convalescent patients who had suffered from acute respiratory distress syndrome (ARDS). Overall, there was no significant changes in {kappa} to {lambda} ratio of total IgG (via protein G capture) antibodies between the groups. A non-statistically significant trend towards {lambda} light chains was found in antibodies against SARS CoV-2 Nucleocapsid and Spike proteins. However, detailed analysis of the molecular forms found a significant increase and bias towards un-glycosylated light chains and in particular un-glycosylated {kappa} light chains, in antibodies against SAR-CoV-2 spike protein, from convalescent COVID-ARDS patients. Here we have demonstrated a bias towards un-glycosylated {kappa} chains in anti-spike antibodies in those who suffered from ARDS as a result of SARS-CoV2 infection 3 months after recovery. How this relates to the immunopathology of COVID-19 requires further study.