Tissue-specific pathway activities: A retrospective analysis in COVID-19 patients

Abstract

The ACE2 receptors essential for SARS-CoV-2 infections are expressed not only in the lung but also in many other tissues in the human body. To better understand the disease mechanisms and progression, it is essential to understand how the virus affects and alters molecular pathways in the different affected tissues. In this study, we mapped the proteomics data obtained from Nie X. et al. (2021) to the pathway models of the COVID-19 Disease Map project and WikiPathways. The differences in pathway activities between COVID-19 and non-COVID-19 patients were calculated using the Wilcoxon test. As a result, 46% (5,235) of the detected proteins were found to be present in at least one pathway. Only a few pathways were altered in multiple tissues. As an example, the Kinin-Kallikrein pathway, an important inflammation regulatory pathway, was found to be less active in the lung, spleen, testis, and thyroid. We can confirm previously reported changes in COVID-19 patients such as the change in cholesterol, linolenic acid, and arachidonic acid metabolism, complement, and coagulation pathways in most tissues. Of all the tissues, we found the thyroid to be the organ with the most changed pathways. In this tissue, lipid pathways, energy pathways, and many COVID-19 specific pathways such as RAS and bradykinin pathways, thrombosis, and anticoagulation have altered activities in COVID-19 patients. Concluding, our results highlight the systemic nature of COVID-19 and the effect on other tissues besides the lung.

Keywords: COVID-19; SARS-CoV-2; WikiPathways; pathway activity; tissue-specific proteomics data.

Figure 1

Protein presence in pathway models. (A) 46% of the measured proteins in the dataset were found in at least one pathway in the combined pathway collection. (B) The distribution of the number of pathways in which a protein participates shows that most proteins are only present in very few pathways.

Figure 2

Changes in pathway activities in COVID-19 patients. (A) The box plot shows the number of less and more active pathways in each tissue. (B–H) For each tissue, the effect size of all significant pathways from the Wilcoxon test are shown. The cutoff of >0.3 for effect size is indicated with a dashed line. Only the 69 pathways, which show significant changes in protein abundance in at least one tissue are included on the x-axis.

Figure 3

Protein-pathway network of 69 changed pathways in COVID-19 patients. Pathways are shown as rectangles and the proteins present in a pathway are connected to the pathway node and shaped as circles. The pathway nodes visualize the heatmap of pathway activity change in COVID-19 patients from thyroid, lung, heart, liver, spleen, kidney and testis in that order. Blue indicates a lower pathway activity in COVID-19 patients (protein abundance distribution shifted to the left) and red indicates an increased activity (protein abundance distribution shifted to the right). Most of the pathways are highly connected with many proteins in common.

Figure 4

Heatmap of pathway activities per tissue (p-value < 0.05, Wilcoxon test). Pathway activities are highly tissue specific and only few pathways show significant changes in multiple tissue. All significant changed pathway are shown and the color gradient indicates the effect size. Blue: the pathway is less active in COVID-19 patients, red: the pathway is more active in COVID-19 patients. Grey: unmeasured or non-significant pathways (p-value >= 0.05).

Figure 5

COVID-19 protein expression visualized on the Kinin Kallikrein pathway (WP5089). The pathway has activities changed in most tissues in COVID-19 patients. (A) Protein abundance is visualized on protein nodes in the pathway. Top row is data from COVID-19 patients and bottom row is data from non-COVID-19 patients. The higher the abundance the darker the value. The data nodes are split in seven columns for the seven different tissues. (B) The pathway figure shows the differential expression of the proteins (log2 fold change) between COVID-19 and non-COVID-19 patients. All proteins are downregulated or not changed in all tissues. The data nodes are split in seven columns for the seven different tissues. Left to right: thyroid, lung, heart, liver, spleen, kidney, testis.

Figure 6

COVID-19 protein expression visualized in the cholesterol biosynthesis pathway (WP197) in the testis. All proteins in the pathway are less abundant in COVID-19 patients.