Comprehensive molecular characterization of [LC]: Immunoglobulin suppression and persistent SARS-CoV-2 antigens…, 2025, Seco-Gonzalez+

[SNT Gatchaman]

Comprehensive molecular characterization of post-COVID condition: Immunoglobulin suppression and persistent SARS-CoV-2 antigens as key pathophysiological drivers
Seco-Gonzalez; Antelo-Riveiro; Bravo; Domínguez-Santalla; Rodríguez-Ruiz; Piñeiro; Garcia-Fandino

BACKGROUND
Post-COVID condition (PCC), or long COVID, affects a significant proportion of individuals following SARS-CoV-2 infection, yet its molecular framework remains poorly understood. This study aimed to define the molecular profile of PCC by integrating broad proteomic analysis using Sequential Window Acquisition of All Theoretical Mass Spectra (SWATH-MS) with targeted antigen quantification through targeted mass spectrometry (MRM/SRM).

METHODS
Plasma and pellet fractions from 65 PCC patients, classified as symptomatic or asymptomatic, were analyzed using SWATH-MS with updated SARS-CoV-2 protein libraries (v2022 and v2024), enabling a comprehensive profiling of immune-and viral-related proteins. The presence of viral antigens, specifically spike and nucleocapsid proteins, was quantified using MRM/SRM. A protein-concentration-based severity metric using clustering analysis and dimensionality reduction methods was proposed to assess correlations between proteomic alterations and clinical symptoms.

RESULTS
A key finding in PCC patients, particularly in symptomatic cases, was a pronounced downregulation of immunoglobulins, including kappa and lambda light chains. SWATH-MS analysis identified six proteins (corresponding to UniProt entries Q8N5F4, LV147, KV311, KVD20, A0A5C2G1U0, and KV315) that strongly correlated with disease severity (R² > 0.9), highlighting their potential as biomarkers. In pellet samples, the protein G1SG72 (ABCE1) emerged as a marker associated with severity, suggesting possible alterations in antiviral responses. The severity metric proposed showed a strong correlation with clinical symptoms, providing a quantifiable measure of PCC severity and enabling effective patient stratification. Additionally, MRM/SRM analysis detected the persistent presence of SARS-CoV-2 antigens, specifically the Spike and Nucleocapsid proteins, in symptomatic PCC patients.

CONCLUSIONS
This study defines a molecular profile of PCC, marked by immunoglobulin downregulation and the persistence of SARS-CoV-2 antigens, which may contribute to ongoing immune alterations in PCC. The severity metric derived from proteomic profiling provides a tool for categorizing PCC patients based on symptom severity and could support future studies aimed at targeted interventions.

Link | PDF | Journal of Infection and Public Health [Open Access]

 

[Utsikt]

METHODS
Plasma and pellet fractions from 65 PCC patients, classified as symptomatic or asymptomatic

How can you have asymptomatic PCC patients?

 

[Eleanor]

Blood samples from 65 PCC patients were separated into plasma (supernatant) and pellet fractions for proteomic analysis. Proteomic profiling using SWATH-MS was conducted with updated SWATH-MS SARS-CoV-2 protein libraries (SARS-CoV-2 proteomic library v2022 and SARS-CoV-2 proteomic library v2024) to identify molecular alterations in both fractions. The SWATH-MS data underwent unsupervised clustering, categorizing patients into subgroups based on their molecular profiles, which were subsequently aligned with clinical symptomatology. Statistical analyses identified significantly dysregulated proteins, culminating in the development of a proteomics-based severity scale to differentiate symptomatic and asymptomatic patients and uncover molecular patterns linked to PCC severity.

The optimal number of clusters was determined by maximizing the average silhouette score, ensuring consistent clustering methodology across sample types and libraries. Although various clustering algorithms were considered, k-means was selected due to its superior performance across all datasets. Clusters were primarily differentiated along PC1, providing distinct separation between patient groups. Cluster assignments were then validated by comparing them to previous patient classifications and clinical symptomatology (Table S2).

This empirical evaluation revealed that plasma samples analyzed with the v2024 library exhibited the strongest alignment with symptom-based classifications, as PC1 explained 68.7 % of variance, with a high Silhouette score (0.522) and a strong Cramér’s V value (0.654). In contrast, plasma samples analyzed with v2022, pellet samples analyzed with v2024, and pellet samples analyzed with v2022 did not show strong associations with clinical severity, as PC1 explained only 44.9 %, 23.0 %, and 22.7 % of variance, respectively, and these datasets exhibited weaker clustering performance (Silhouette score ≤0.122, Cramér’s V ≤0.232). Given these results, these datasets were not prioritized for the main analysis and are instead included in the Supporting Information (Figures S2-S3, S6-S7, Tables S8-S9).

To a lay person, this all sounds a bit circular?