The Vsr-like protein FASTKD4 regulates the stability and polyadenylation of the MT-ND3 mRNA

Abstract

Expression of the compact mitochondrial genome is regulated by nuclear encoded, mitochondrially localized RNA-binding proteins (RBPs). RBPs regulate the lifecycles of mitochondrial RNAs from transcription to degradation by mediating RNA processing, maturation, stability and translation. The Fas-activated serine/threonine kinase (FASTK) family of RBPs has been shown to regulate and fine-tune discrete aspects of mitochondrial gene expression. Although the roles of specific targets of FASTK proteins have been elucidated, the molecular mechanisms of FASTK proteins in mitochondrial RNA metabolism remain unclear. Therefore, we resolved the structure of FASTKD4 at atomic level that includes the RAP domain and the two FAST motifs, creating a positively charged cavity resembling that of the very short patch repair endonuclease. Our biochemical studies show that FASTKD4 binds the canonical poly(A) tail of MT-ND3 enabling its maturation and translation. The in vitro role of FASTKD4 is consistent with its loss in cells that results in decreased MT-ND3 polyadenylation, which destabilizes this messenger RNA in mitochondria.

Graphical Abstract

Figure 1.

Atomic structure of FASTKD4 and the impact of mutations on its function. (A) Schematic structure of FASTKD4 containing an N-terminal mitochondrial targeting sequence and the C-terminal FAST_1, FAST_2 and RAP domains. (B) Atomic structure and electrostatic potential of the surface of FASTKD4. (C) Structural alignment of FASTKD4 (shown in grey) with the Vsr endonuclease (red). (D) The potential catalytic pocket of FASTKD4 in comparison with Vsr. (E) Positions of flanking charged residues that contribute to the positively charged cavity. (F) Positions of the ‘Lock’ and ‘Key’ helices with key residues indicated. (G) Helical wheel projections for the ‘Lock’ and ‘Key’ helices (generated by pepwheel). (H) Detection of protein and RNA levels in FASTKD4 (D4^−/−) mutants targeting either the positively charged surface (K337D and K453W), the central lock helix (I457C, T460C and L4663W) or C-term key helix (Q612D, 616YLKtoALD and 620KMRKtoDMDD). Six micrograms of RNA or 20 μg of protein were separated by size via agarose formaldehyde gel electrophoresis or SDS-PAGE, respectively, transferred to a respective membrane and probed for FASTKD4, β-actin, MT-ND3 and MT-ND5.

Figure 2.

FASTKD4 is required for non-canonical RNA processing and MT-ND3 mRNA stability. (A) Detection of RNA levels through northern blotting after chloramphenicol treatment and in the presence (WT) and absence of FASTKD4 (D4^−/−). Northern blotting was performed using 10 μg of RNA and probed for MT-ND3, MT-ND5 and MT-CYB mRNAs. (B) Detection of MT-ND3 mRNA length in WT and D4^−/− cells after incubation with ssDNA (50 μM, 5 μM or 0.5 μM) and RNase H. Northern blotting was performed using 10 μg RNA and probed for MT-ND3, MT-ND5 and MT-CYB mRNAs. (C) HEK293T mitochondrial lysates were incubated with 4 μg of the following in vitro transcribed MT-ND3 RNAs: 51 nt poly(A) tail, 25 nt poly(A) tail, without poly(A) tail, tRNA-R or mirror MT-ND3 RNA. Protein levels were detected by SDS-PAGE using GRSF1, FASTKD4 and PNPT1 antibodies.

Figure 3.

Inhibition of translation protects MT-ND3 mRNA from degradation in the absence of FASTKD4. Detection of RNA levels via northern blotting in untreated or treated WT, FASTKD4 knockout (D4^−/−), ANGEL2 single or double knockout (A) or Nocturin single or double knockout (NOCT^−/−) cells (B). Cells were either treated with 20 μg/ml chloramphenicol or untreated and 10 μg isolated RNA was used for northern blotting and probed for MT-ND3, MT-ATP8/6, MT-CO3, MT-CYB, MT-ND6 mRNAs and 18S rRNA. Representative blots are shown of three independently run biological experiments.

Figure 4.

FASTKD4 protects the polyadenylated end of MT-ND3 required for the stability of this mRNA. (A) Levels of alternate and canonical MT-ND3 mRNA ends or (B) MT-ND4L/4 and MT-ATP8/6 mRNA ends in HAP1 cells or in CAL51 cells (C) were measured by poly(A) capture and next generation sequencing. (D) Comparison of polyadenylation of MT-ND3 RNA ends. WT or FASTKD4^−/− cells were either untreated or treated with 20 μg/mL chloramphenicol prior to poly(A) capture and Sanger sequencing. Three biologically independent samples were used in each experiment in Figure 4, the results are mean ± SD (Student's t test, two-tailed unpaired t-test) and P values are shown. (E) Model of FASTKD4 function, where FASTKD4 protects the length of the poly(A) tail from the MT-ND3 mRNA, and in its absence there is reduced polyadenylation of this mRNA, compromising its stability.