Urinary Peptidomic Profiling in Post-Acute Sequelae of SARS-CoV-2 Infection: A Case-Control Study, 2025, Gülmez et al

[forestglip]

Now published, see post #4
--------

Urinary Peptidomic Profiling in Post-Acute Sequelae of SARS-CoV-2 Infection: A Case-Control Study

Dilara Guelmez, Justyna Siwy, Katharina Kurz, Ralph Wendt, Miroslaw Banasik, Bjorn Peters, Emmanuel Dudoignon, Francois Depret, Mercede Salgueira, Elena Nowacki, Amelie Kurnikowski, Sebastian Mussnig, Simon Krenn, Samuel Gonos, Judith Loffler-Ragg, Guenter Weiss, Harald Mischak, Manfred Hecking, Eva Schernhammer, Joachim Beige

Background
Post-acute sequelae of severe acute respiratory syndrome coronavirus 2-infection (PASC) is challenging to diagnose and treat, and its molecular pathophysiology remains unclear. Urinary peptidomics can provide valuable information on urine peptides that may enable improved and specified PASC diagnosis.

Methods
Using standardized capillary electrophoresis-MS, we examined the urinary peptidomes of 50 patients with PASC 10 months after COVID-19 and 50 controls including healthy individuals (n = 42) and patients with non-COVID-19-associated myalgic encephalomyelitis/chronic fatigue syndrome (n = 8). Based on peptide abundance differences between cases and controls, we developed a diagnostic model using a support vector machine.

Results
The abundance of 195 urine peptides among PASC patients significantly differed from that in controls, with a predominant abundance of collagen alpha chains. This molecular signature (PASC195), effectively distinguished PASC cases from controls in the training set [AUC of 0.949 (95% CI 0.900-0.998; p < 0.0001)] and independent validation set [AUC of 0.962 (95% CI 0.897-1.00); p < 0.0001)]. In silico assessment suggested exercise, GLP1-RA and MRA as potentially efficacious interventions.

Conclusions
We present a novel and non-invasive diagnostic model for PASC. Reflecting its molecular pathophysiology, PASC195 has the potential to advance diagnostics and inform therapeutic interventions.

Web | PDF | Preprint: MedRxiv | Open Access

 

[forestglip]

Competing Interest Statement

H.M. is the co-founder and co-owner of Mosaiques Diagnostics. J.S. is employed by Mosaiques Diagnostics GmbH.

Funding Statement

This project was supported by the Federal Ministry of Health (BMG) via grant number 2523FSB114; by the German Ministry for Education and Science (BMBF) via grant 01KU2309; by the Province of Tyrol via grant number GZ 75759; by Fisser Bergbahnen through a benefit gala donation (2022) under the project name: Projekt ME/CFS-Forschung; by the Province of Tyrol and the WE & ME Foundation (Scientific Commitment & Myalgic Encephalomyelitis Foundation) via grant number GZ 86686; by the Swedens innovation agency (Vinnova) via grant 2022-00542; by the National Centre for Research and Development (Narodowe Centrum Badan i Rozwoju) via grant number: PerMed/V/80/UriCov/2023; by the Austrian Science Fund (FWF) via Project number I 6464, Grant-DOI 10.55776/I6464; by the French National Research Agency-Agence Nationale de la Recherche (ANR)-under the grant ANR-22-PERM-0014; and in part by the Austrian Science Fund (FWF) via Project number I 6471, Grant-DOI 10.55776/I6471 under the frame of ERA PerMed. The funders were not involved in the study design, data collection, data analysis, interpretation of results, or manuscript preparation.

 

[ME/CFS Science Blog]

Of the 2,034 peptides detected in the urine peptidome analysis, 249 375 differed significantly in abundance (Figure 2, Panel A). Amino acid sequences could be 250 obtained for 243 peptides, which were further subjected to a take-one-out process. This 251 refinement resulted in a final pool of 195 peptides used to generate a classifier, PASC195, 252 consisting of 187 upregulated and eight downregulated peptides. Of these, 172 (88.2%) were 253 aligned to collagen alpha

Looks like they tested more than 2000 peptides in a cohort of only 30 cases and 30 controls (with ME/CFS patients in both cases and controls).

 

[SNT Gatchaman]

Published as —

Urinary Peptidomic Profiling In Post-Acute Sequelae of SARS-CoV-2 Infection: A Case-Control Study

Spoiler: Authors

Dilara Gülmez; Justyna Siwy; Katharina Kurz; Ralph Wendt; Miroslaw Banasik; Björn Peters; Emmanuel Dudoignon; Francois Depret; Mercedes Salgueira; Elena Nowacki; Amelie Kurnikowski; Sebastian Mussnig; Simon Krenn; Samuel Gonos; Judith Löffler-Ragg; Günter Weiss; Harald Mischak; Manfred Hecking; Eva Schernhammer; Joachim Beige

Post-acute sequelae of severe acute respiratory syndrome coronavirus 2-infection (PASC) is challenging to diagnose and treat, and its molecular pathophysiology remains unclear. Urinary peptidomics can provide valuable information on urine peptides that may enable improved and specified PASC diagnosis.

Using standardized capillary electrophoresis-MS, we examined the urinary peptidomes of 50 patients with PASC 10 months after COVID-19 and 50 controls, including healthy individuals (n = 42) and patients with non-COVID-19-associated myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) (n = 8). Based on peptide abundance differences between cases and controls, we developed a diagnostic model using a support vector machine.

The abundance of 195 urine peptides among PASC patients significantly differed from that in controls, with a predominant abundance of collagen alpha chains. This molecular signature (PASC195) effectively distinguished PASC cases from controls in the training set (AUC of 0.949 [95% CI 0.900–0.998; p < 0.0001]) and independent validation set (AUC of 0.962 [95% CI 0.897–1.00]; p < 0.0001]). In silico assessment suggested exercise, GLP-1RAs and mineralocorticoid receptor antagonists (MRAs) as potentially efficacious interventions.

We present a novel and non-invasive diagnostic model for PASC. Reflecting its molecular pathophysiology, PASC195 has the potential to advance diagnostics and inform therapeutic interventions.

STATEMENT OF SIGNIFICANCE OF THE STUDY
Despite the recent emergence of omics-derived candidates for post-acute sequelae of SARS-CoV-2 infection (PASC), the pending validation of proposed markers and lack of consensus result in the continuous reliance on symptom-based criteria, being subject to diagnostic uncertainties and potential recall bias.

Building upon prior findings of renal involvement in acute COVID-19 pathophysiology and PASC-associated alterations, we hypothesized that the use of urinary peptides for PASC-specific biomarker discovery, unlike conventional specimens that have been utilized thus far, may offer complementary information on putative disease mechanisms. In the present study, 195 significantly expressed peptides were used to form a classifier termed PASC195, which effectively discriminated PASC from non-PASC (p < 0.0001), including healthy individuals and non-COVID-19-associated myalgic encephalomyelitis/chronic fatigue syndrome, in both the derivation (n = 60) and an independent validation set (n = 40).

The peptidome profile associated with PASC was consistent with a shift in collagen turnover, with most PASC195 peptides derived from alpha chains. Ongoing inflammatory responses, hemostatic imbalances, and endothelial damage were indicated by cross-sectional variations in endogenous peptide excretion.

Web | DOI | PDF | PROTEOMICS | Open Access

 

[BrightCandle]

There are loads of these 100+ measure combined machine learned algorithm diagnosis tools coming out at the moment.195 measures in a study with 60 people and 8 ME/CFS controls, I have to question how likely it is this will just fail the moment more patients are tested. Maybe one day these highly dimensional tests will make it through and become accepted but I think everyone is holding out for something that is 1 or few measures.

 

[ScoutB]

I have to question how likely it is this will just fail the moment more patients are tested.

I have the same question. Sounds like they at least tested the classifier on a small validation set of 40 patients.


Also:

Case urine samples were collected at a median of 292 days after COVID-19 diagnosis.

Training data from the Innsbruck cohort, including 30 cases (27 PASC, 3 PASC-ME/CFS) and 30 controls (26 healthy controls, 4 non-COVID-19 ME/CFS) were used to define the PASC associated urinary peptides.

The LC patients had LC for <1 year on average. Interesting that they are trying to separate LC-ME/CFS from non-covid-ME/CFS. I'm curious where this thing would put someone who's had LC-ME/CFS for 4+ years.