Activation of the Carboxypeptidase U (CPU, TAFIa, CPB2) System in Patients with SARS-CoV-2 Infection Could Contribute to COVID-19 Hypofibrinolytic State and Disease Severity Prognosis

Abstract

Coronavirus disease 2019 (COVID-19) is a viral lower respiratory tract infection caused by the highly transmissible and pathogenic SARS-CoV-2 (severe acute respiratory-syndrome coronavirus-2). Besides respiratory failure, systemic thromboembolic complications are frequent in COVID-19 patients and suggested to be the result of a dysregulation of the hemostatic balance. Although several markers of coagulation and fibrinolysis have been studied extensively, little is known about the effect of SARS-CoV-2 infection on the potent antifibrinolytic enzyme carboxypeptidase U (CPU). Blood was collected longitudinally from 56 hospitalized COVID-19 patients and 32 healthy controls. Procarboxypeptidase U (proCPU) levels and total active and inactivated CPU (CPU+CPUi) antigen levels were measured. At study inclusion (shortly after hospital admission), proCPU levels were significantly lower and CPU+CPUi antigen levels significantly higher in COVID-19 patients compared to controls. Both proCPU and CPU+CPUi antigen levels showed a subsequent progressive increase in these patients. Hereafter, proCPU levels decreased and patients were, at discharge, comparable to the controls. CPU+CPUi antigen levels at discharge were still higher compared to controls. Baseline CPU+CPUi antigen levels (shortly after hospital admission) correlated with disease severity and the duration of hospitalization. In conclusion, CPU generation with concomitant proCPU consumption during early SARS-CoV-2 infection will (at least partly) contribute to the hypofibrinolytic state observed in COVID-19 patients, thus enlarging their risk for thrombosis. Moreover, given the association between CPU+CPUi antigen levels and both disease severity and duration of hospitalization, this parameter may be a potential biomarker with prognostic value in SARS-CoV-2 infection.

Keywords: COVID-19; carboxypeptidase B2; carboxypeptidase U; coronavirus; thrombin-activatable fibrinolysis inhibitor.

Figure 1

Plasma procarboxypeptidase U (proCPU) levels (A) and total active and inactivated carboxypeptidase U (CPU+CPUi) antigen levels (B) in COVID-19 exposed (N = 14) and COVID-19 non-exposed (N = 18) controls. COVID-19 exposed controls are individuals that previously tested positive for SARS-CoV-2 (PCR-confirmed SARS-CoV-2 infection (at least 2 months before inclusion) or a positive serological test result), while COVID-19 non-exposed controls are individuals without any evidence of SARS-CoV-2 exposure. Data are presented as mean ± SD. Mann–Whitney U test; ns = not significant.

Figure 2

(A,B) Time course of plasma procarboxypeptidase U (proCPU) levels (A) and total active and inactivated carboxypeptidase U (CPU+CPUi) antigen levels (B) in hospitalized COVID-19 patients (N = 12–56) at inclusion (ranging from 1–5 days after hospital admission), 1 week after inclusion (ranging from 5–8 days after inclusion), 2 weeks after inclusion (ranging from 12–15 days after inclusion), and at discharge (ranging from 17–61 days after inclusion), and in clinically healthy controls at inclusion and 28 days later (N = 32). Data are presented as mean ± SD. Mann–Whitney U test (unpaired data); ## p < 0.01; and ### p < 0.001. Wilcoxon Matched-Pairs Signed Rank test (paired data); * p < 0.05; ** p < 0.01. The horizontal dotted line represents the mean proCPU level or CPU+CPUi antigen level of the healthy controls with the corresponding confidence interval (2*SD; grey area). (C,D) Time course of plasma proCPU levels (C) and CPU+CPUi antigen levels (D) at inclusion (ranging from 1–5 days after hospital admission), 1 week after inclusion (ranging from 5–8 days after inclusion), 2 weeks after inclusion (ranging from 12–15 days after inclusion), and at discharge (ranging from 17 to 61 days after inclusion) in COVID-19 patients with critical disease (right; N = 12) versus non-critical disease (left; N = 44). Data are presented as mean ± SD. The horizontal dotted line represents the mean proCPU level or CPU+CPUi antigen level of the healthy controls with the corresponding confidence interval (2*SD; grey area). Mann–Whitney U test (unpaired data); # p < 0.05; ## p < 0.01; and ### p < 0.001. Wilcoxon Matched-Pairs Signed Rank test (paired data); * p < 0.05; ** p < 0.01; *** p < 0.001. (E,F) Individual proCPU (E) and CPU+CPUi antigen level (F) profiles of six critically ill patients (samples at ≥4 time points available). The horizontal dotted line represents the mean proCPU level or CPU+CPUi antigen level of the healthy controls with the corresponding confidence interval (2*SD; grey area).

Figure 3

Relationship between the highest recorded CRP levels and corresponding proCPU levels (A) or CPU+CPUi antigen levels (B). For the majority of the patients, the highest CRP value was measured within one week after enrolment in the study. Spearman correlation coefficient r was determined for both correlations. In case of a significant correlation (p > 0.05), linear regression analysis was performed, and the best-fit line (solid line) with 95% confidence bands was plotted (dashed lines).

Figure 4

Correlation between the duration of hospitalization and both baseline (inclusion time point shortly after hospital admission) procarboxypeptidase U (proCPU) (A) and total active and inactivated carboxypeptidase U (CPU+CPUi) antigen levels (B) in hospitalized COVID-19 patients (N = 56). Spearman correlation coefficient r was determined. For statistically significant correlations (p < 0.05), linear regression analysis was performed and the best-fit line (solid line) with 95% confidence bands was plotted (dashed lines). Baseline proCPU levels (C) and baseline CPU+CPUi antigen levels (D) of hospitalized COVID-19 patients grouped by disease severity (WHO COVID-19 disease severity categorization): moderate (N = 39), severe (N = 5), and critical (N = 12). Mann–Whitney U test (unpaired data); * p < 0.05; ** p < 0.01.