Cell-free DNA maps COVID-19 tissue injury and risk of death and can cause tissue injury
Temesgen E. Andargie, Naoko Tsuji, Fayaz Seifuddin, Moon Kyoo Jang, Peter S.T. Yuen, Hyesik Kong, Ilker Tunc, Komudi Singh, Ananth Charya, Kenneth Wilkins, Steven Nathan, Andrea Cox, Mehdi Pirooznia, Robert A. Star, Sean Agbor-Enoh
Temesgen E. Andargie, Naoko Tsuji, Fayaz Seifuddin, Moon Kyoo Jang, Peter S.T. Yuen, Hyesik Kong, Ilker Tunc, Komudi Singh, Ananth Charya, Kenneth Wilkins, Steven Nathan, Andrea Cox, Mehdi Pirooznia, Robert A. Star, Sean Agbor-Enoh
View: Text | PDF
Clinical Research and Public Health COVID-19 Inflammation

Cell-free DNA maps COVID-19 tissue injury and risk of death and can cause tissue injury

Abstract

INTRODUCTION The clinical course of coronavirus 2019 (COVID-19) is heterogeneous, ranging from mild to severe multiorgan failure and death. In this study, we analyzed cell-free DNA (cfDNA) as a biomarker of injury to define the sources of tissue injury that contribute to such different trajectories.METHODS We conducted a multicenter prospective cohort study to enroll patients with COVID-19 and collect plasma samples. Plasma cfDNA was subject to bisulfite sequencing. A library of tissue-specific DNA methylation signatures was used to analyze sequence reads to quantitate cfDNA from different tissue types. We then determined the correlation of tissue-specific cfDNA measures to COVID-19 outcomes. Similar analyses were performed for healthy controls and a comparator group of patients with respiratory syncytial virus and influenza.RESULTS We found markedly elevated levels and divergent tissue sources of cfDNA in COVID-19 patients compared with patients who had influenza and/or respiratory syncytial virus and with healthy controls. The major sources of cfDNA in COVID-19 were hematopoietic cells, vascular endothelium, hepatocytes, adipocytes, kidney, heart, and lung. cfDNA levels positively correlated with COVID-19 disease severity, C-reactive protein, and D-dimer. cfDNA profile at admission identified patients who subsequently required intensive care or died during hospitalization. Furthermore, the increased cfDNA in COVID-19 patients generated excessive mitochondrial ROS (mtROS) in renal tubular cells in a concentration-dependent manner. This mtROS production was inhibited by a TLR9-specific antagonist.CONCLUSION cfDNA maps tissue injury that predicts COVID-19 outcomes and may mechanistically propagate COVID-19–induced tissue injury.FUNDING Intramural Targeted Anti–COVID-19 grant, NIH.

Authors

Temesgen E. Andargie, Naoko Tsuji, Fayaz Seifuddin, Moon Kyoo Jang, Peter S.T. Yuen, Hyesik Kong, Ilker Tunc, Komudi Singh, Ananth Charya, Kenneth Wilkins, Steven Nathan, Andrea Cox, Mehdi Pirooznia, Robert A. Star, Sean Agbor-Enoh

×

Figure 11

Purified plasma cfDNA stimulation of mtROS production and the effect of TLR9 inhibition.

Options: View larger image (or click on image) Download as PowerPoint
Purified plasma cfDNA stimulation of mtROS production and the effect of ...
mPPTC culture was treated with plasma of COVID-19 patient B or purified cfDNA from the same plasma samples. (A) mtcfDNA and ncfDNA dynamics in plasma from COVID-19 patient B. (B) mPPTCs (n = 15–44 cells from 3–5 fields/group) treated with 3% healthy plasma or serial 3% COVID-19 patient B plasma and (C) with control DNA buffer and equivalent amount of cfDNA purified from serial COVID-19 patient B plasma; MitoSOX Red intensity measured after 6 hours incubation. Data summarized as in Figure 9. Statistical significance determined using Dunnett’s multiple-comparison test after Kruskal-Wallis test. *P < 0.05 versus control, #P < 0.05 versus day 7. (D) Comparison of MitoSOX Red intensity with serial COVID-19 patient B plasma (red) and cfDNA purified from COVID-19 patient B plasma (purple) after 6 hours incubation. Each point represents total MitoSOX Red background-corrected intensity for 1 tubule cell. Line graph expressed as mean ± SEM. Data analyzed using Mann-Whitney test after Kruskal-Wallis test. (E) MitoSOX Red intensity in mPPTCs (n = 26–47 cells from 3–6 fields/ group) treated with 3% healthy plasma; 3% COVID-19 patient plasma with 10 μM control oligodeoxynucleotide (ODN); or 0, 0.1, 1, or 10 μM TLR9 antagonist (ODN2088) after 6 hours incubation. Statistical significance determined using Dunnett’s multiple-comparison test after Kruskal-Wallis test. *P < 0.05 versus 3% healthy plasma, #P < 0.05 versus no ODN, †P < 0.05 versus 10 μM control ODN. (F) Comparison of MitoSOX Red intensity in mPPTCs (n = 17–63 cells from 3–6 fields/group) treated with serial COVID-19 plasma or COVID-19 plasma with ODN2088, or (G) serial cfDNA purified from the COVID-19 patient B plasma with and without ODN2088 after 6 hours incubation. Statistical significance determined using a Mann-Whitney test after Kruskal-Wallis test. *P < 0.05. mtROS production of patient B shown in Supplemental Figure 4.