For the past two years, the entire world was caught unprepared by the COVID-19 pandemic. Up till now, the coronavirus has been raging in most parts of the globe. We have entered the season of influenza, which is complicated by worsening COVID-19 outbreaks. With the recent emergence of the more contagious mutant Omicron, many countries have become extremely vigilant toward this new wave of pandemic, and this is not a battle to be fought alone. Regardless of the prevalent variant, scientists have never stopped efforts of unfolding the mystery of this ever-evolving SARS-CoV-2 virus. To halt the COVID-19 crisis, accurate clinical diagnosis and classification of COVID-19 remains challenging due to multiple reasons, including timing, accessible equipment and supplies, well-trained medical personnel for sample collection, etc. Hence, it is vital to study the molecular phenotype of urine in patients with COVID-19.

On December 27, 2021, the Tiannan Guo research group, or Guomics, (https://guomics.com) in the School of Life Sciences at Westlake University and its collaborators published a paper entitled "Proteomic and metabolic profiling of urine uncovers immune responses in patients with COVID-19" in Cell Reports.

Link:

https://www.cell.com/cell-reports/fulltext/S2211-1247(21)01783-6

This study revealed that urine from patients with COVID-19 contained specific molecules that could sensitively reflect the pathological status in response to COVID-19. In this study, 20 protein markers were screened from the urine, and novel machine learning models were established, which successfully classified and predicted COVID-19 severity. The study also presented evidence supporting potential renal injury.

Urine can be obtained non-invasively without professional training (as opposed to serum, tissue, etc.) and can meet the requirements of daily real-time personal health monitoring. It is feasible and of significant clinical value to add urine as one of the routinely monitored health indicators in COVID-19 management, and in precision medicine at large.

In this study, 115 urine and serum samples from the COVID-19 patient group and healthy control group were systematically analyzed. The samples in each group were analyzed and compared using proteomics and metabolomics. The protein level per unit volume of the urine in the COVID-19 group was significantly higher than that in the healthy group, while for serum, it was approximately the same between these two groups. The results suggested that urine could precisely reflect pathological changes in diseases.

A total of 1494 serum proteins, 3854 urine proteins, 903 serum metabolites, and 1033 urine metabolites were quantified in this study. The protein molecular weight distribution in urine was consistent with that of the whole human proteome, suggesting that the urine sample did not miss any protein or lose any information.

      Profiling of serum and urine proteomics and metabolomics data

Can the urine protein reflect COVID-19 induced molecular changes? The machine learning results demonstrated that urine was as sensitive as serum in classifying COVID-19 by severity. A 20-urine-protein machine learning model was then established based on this finding. During the whole course of recovery in severe COVID-19 patients, the predicted value of the model decreased gradually with time. While in mild cases, the predicted value was relatively consistent without significant fluctuation. These results further confirmed that this 20-urine-protein machine learning model had great potential in classifying and predicting COVID-19 severity.

Classification of the patients with mild and severe COVID-19 at the proteomic level

The study then explored the correlation between the serum and the urine in patients with COVID-19. As the disease progressed (healthy, mild, severe), the relative abundance of 301 proteins showed an opposite expression pattern in the urine and the serum.

It was found that two important regulatory factors involved in renal tubular reabsorption, megalin (LRP2) and cubilin (CUBN), decreased in the urine of patients with COVID-19. The renal tubular reabsorption of patients with COVID-19 may be dysfunctional, resulting in different expression patterns between urine and blood. This phenomenon may also exist in other diseases, which awaits further investigation.

301 serum and urine proteins showed opposite expression patterns

Cytokine storm caused by uncontrolled innate inflammatory response is the main cause of mortality in patients with COVID-19. Therefore, this study also focused on the expression of cytokines in the serum and urine. The study quantified 124 and 197 cytokines in the serum and urine, respectively. CXCL14 presented in the urine had a significant correlation with lymphocyte counts and may be used to indicate the severity of COVID-19.

In addition, some proteins related to virus budding were specifically found in the urine proteome. These proteins showed a significant downward trend in the urine of patients with COVID-19 and were not detected in the serum. These results indicated that, with comparable MS methodology, the urine proteome exhibited higher information content than serum proteome.

The proteins related to virus budding quantified in urine showed significant difference between the healthy group and COVID-19 group

Several proteins were found to be frequently present in the obtained differential expression pathways. Among them, Cdc42, Rac1 / Rac2, and RhoA were most frequently detected. They were GTPase proteins, and dysregulation of these proteins may lead to glomerulosclerosis and kidney damage. The dynamic regulation of renal podocyte actin requires a large amount of ATP. The metabolomics data showed that the content of adenosine (a product of ATP metabolism) decreased significantly in the urine of patients with severe COVID-19, further confirming podocyte dyskinesia and potential renal injury.

Analysis of dysregulated proteins or pathways in serum and urine of patients with COVID-19

Like other viral infections, SARS-Cov-2 can trigger oxidative stress by breaking the balance between oxidative and antioxidant systems. From the metabolomic data of this study, it could be inferred that a variety of antioxidant factors, such as taurine, sub taurine and 1-methylnicotinamide (1-MNA), were significantly down-regulated in the serum of patients with COVID-19. At the protein level, the study also showed that many antioxidant enzymes such as SOD3 and GPx4 were significantly downregulated in the urine of the severe group. All findings showed that there might be ROS activation response in patients with COVID-19.

Dysregulated metabolite profiling in patients with COVID-19

The study had thus comprehensively interpreted the molecular and signaling pathways of abnormal changes in the urine and blood of patients with COVID-19 and proposed that the level of the molecules or pathways induced by COVID-19 in vivo was altered and regulated. Immune response triggered by inflammatory disorders, coagulation and cellular fibrosis ultimately damaged the kidney tissue. The clinical data also showed that, the indexes of various types of renal injury in severe patients, despite being within normal range, had changed significantly compared with the healthy control group. The above results suggested that SARS-Cov-2 might cause renal injury. The study suggested that we should pay close attention to clinical indicators of renal injury in patients with COVID-19, and follow-up on renal function after rehabilitation.

A model of renal injury induced by immune dysfunction and ROS activation in patients with severe COVID-19

Guomics focuses on high-throughput proteomics and clinical big data research. They apply pressure cycling technology to process trace amounts of clinical samples in high throughput, digitize their proteomes using DIA-MS technology, and develop machine learning algorithms to analyze proteomics big data. By exploring the quantitative patterns of protein expression and changes in various physiological and pathological states, Guomics strives to achieve its vision of a new era of AI-powered proteomics-drive precision medicine.