Stellenbosch University: Proteomics analyses used to investigate long COVID
The SARS-CoV-2 virus called COVID-19 has paralysed the world since December 2019, and it was the scientific community that responded with treatment plans and accelerated vaccine development to treat and prevent this infection. Unfortunately, for some it became clear that SARS-CoV-2 infection could also have long-term effects. This lingering effect was first termed post-acute sequelae of COVID-19 (PASC) but is now commonly referred to as ‘long COVID’.
Prof Etheresia Pretorius (Stellenbosch University) was one of the first scientists to investigate this phenomenon, and she contacted the CAF Proteomics Unit to assist.
One feature of plasma obtained from long COVID patients is the formation of micro-clots. However, this feature is not unique to long COVID but is also observed in diabetes sufferers, although the clinical picture is vastly different. Up to this point, proteomic analyses have used either clarified plasma or fully denatured and solubilised plasma proteins and have been unable to distinguish between proteins associated in clots and soluble proteins. We derived a strategy using a two-step digestion to isolate the clots. The first, under nondenaturing conditions, fragmented the soluble proteins, which were removed after centrifugation, and the second used nondenaturing conditions to break up the protein aggregates prior to trypsin digestion. After this step, it became apparent why the diabetes and long COVID clinical profiles were so different. The clots from diabetes patients were cleared after the first digestion whereas long COVID patients had clots resistant to the first trypsinisation step.
The digested material was subjected to nanoflow liquid chromatography-mass spectrometry/mass spectrometry on a Thermo Scientific high-resolution mass spectrometer. The proteomes were separated using a 90-minute gradient on a charged surface hybrid (CSH) column (Figure 20 A) before high-resolution spectra were acquired for protein identification (Figure 20 B and C) and quantitation (Figure 20 B and C). This was done by matching the experimental spectra obtained (Figure 20 C) to theoretical spectra using a combination of software. The software used was Thermo Scientific Proteome Discoverer and Scaffold Q+ from Proteome Software. The theoretical sequences were obtained from a previously published human-COVID-19 database.
The study found that in addition to the expected proteins such as fibrinogen, many anticoagulation-associated proteins and inflammatory markers were also trapped. The change in protein expression can be visualised using a volcano plot, as presented in Figure 21. The fold change is depicted on the x-axis in Log2 scale and the confidence (p-value) on the y-axis in -Log10 scale. The increase in alpha-2-antiplasmin may explain why long COVID-associated clots are not broken down as this protein is one of the central components in the fibrinolytic pathway. In addition to this, the discovery of ‘trapped’ inflammatory markers may also explain why symptomatic individuals receive serological tests within normal parameters – the markers tested for are trapped in the clots. Furthermore, a comparison between acute COVID-19 patients and long COVID patients showed a change in protein expression as the condition progressed from COVID-19 to long COVID. There was a marked increase in fibrinogen chains, serum amyloid A1 and 4 as well as von Willebrand factor, pointing to increased platelet activation.