CO and NO bind to Fe(II) DiGeorge critical region 8 heme but do not restore primary microRNA processing activity.

Abstract

The RNA-binding heme protein DiGeorge critical region 8 (DGCR8) and its ribonuclease partner Drosha cleave primary transcripts of microRNA (pri-miRNA) as part of the canonical microRNA (miRNA) processing pathway. Previous studies show that bis-cysteine thiolate-coordinated Fe(III) DGCR8 supports pri-miRNA processing activity, while Fe(II) DGCR8 does not. In this study, we further characterized Fe(II) DGCR8 and tested whether CO or NO might bind and restore pri-miRNA processing activity to the reduced protein. Fe(II) DGCR8 RNA-binding heme domain (Rhed) undergoes a pH-dependent transition from 6-coordinate to 5-coordinate, due to protonation and loss of a lysine ligand; the ligand bound throughout the pH change is a histidine. Fe(II) Rhed binds CO and NO from 6- and 5-coordinate states, forming common CO and NO adducts at all pHs. Fe(II)-CO Rhed is 6-coordinate, low-spin, and pH insensitive with the histidine ligand retained, suggesting that the protonatable lysine ligand has been replaced by CO. Fe(II)-NO Rhed is 5-coordinate and pH insensitive. Fe(II)-NO also forms slowly upon reaction of Fe(III) Rhed with excess NO via a stepwise process. Heme reduction by NO is rate-limiting, and the rate would be negligible at physiological NO concentrations. Importantly, in vitro pri-miRNA processing assays show that both CO- and NO-bound DGCR8 species are inactive. Fe(II), Fe(II)-CO, and Fe(II)-NO Rhed do not bear either of the cysteine ligands found in the Fe(III) state. These data support a model in which the bis-cysteine thiolate ligand environment of Fe(III) DGCR8 is necessary for establishing proper pri-miRNA binding and enabling processing activity.