Proteolyzed but not full-length galectin-9 in plasma predicts the prognosis of COVID-19

INTRODUCTION

Increased circulating galectin-9 (Gal-9) is associated with COVID-19 severity. Gal-9 structurally consists of two homologous carbohydrate-recognition domains (NCRD and CCRD) linked by a linker peptide highly susceptible to proteolysis. Plasma Gal-9 NCRD with an attached truncated linker peptide (N-cleaved-Gal9) is a severity marker for COVID-19, indicating higher AUC than full-length (FL)-Gal9. This study aims to determine whether N-cleaved-Gal9 in plasma on admission serve as a reliable predictor of the hospitalization period in COVID-19.

METHODS

We examined 44 COVID-19 patients admitted to Sendai City Hospital. FL-Gal-9 ELISA and Tr-Gal9 ELISA measured FL-Gal9 and both Gal-9 containing NCRD (FL-Gal9 and N-cleaved-Gal9), respectively. N-cleaved-Gal9 levels were calculated by subtracting FL-Gal9 levels from Tr-Gal9 levels. The time course of FL-Gal-9 and N-cleaved-Gal9 levels from the day of admission to after discharge were monitored.

RESULTS

FL-Gal-9 and N-cleaved-Gal9 levels on admission positively correlated with the hospitalization period (r = 0.3964 and 0.5676, respectively). N-cleaved-Gal9 levels on admission in the hospitalization ≥ 7 days group were higher than those in the hospitalization ≤ 6 days and the levels in both groups converged to the same extent at discharge and after. N-cleaved-Gal9 (AUC = 0.7900) but not FL-Gal9 on admission significantly discriminated both groups, whereas CRP and P/F ratio on admission indicated AUC values of 0.7727 and 0.8268, respectively. The sensitivity of N-cleaved-Gal9 was higher than that of CRP. N-cleaved-Gal9 levels positively correlated with CRP levels during hospitalization but not at discharge and after suggesting Gal9 proteolysis is upregulated along with abnormal CRP. N-cleaved-Gal9 levels negatively correlated with P/F ratio only on admission, suggesting that Gal-9 proteolysis is associated with respiratory failure.

CONCLUSION

Plasma N-cleaved-Gal9 on admission predicted the hospitalization period more accurately than CRP in COVID-19 and demonstrated its performance by simultaneously measuring the CRP and P/F ratio. These findings contribute to the management of prognosis in COVID-19 although larger, more diverse cohorts are needed to validate the findings.