Identification of a novel MIPEP splice variant with altered substrate-binding properties

Abstract

Mitochondrial intermediate peptidase (MIPEP) is a mitochondrial signal peptidase that removes N-terminal amino acids from mitochondrial matrix proteins. We have identified a novel Mipep splice variant that lacks exons 15 and 16, which we termed "ΔMIPEP". We characterized the molecular features of ΔMIPEP by investigating its expression level in numerous mouse tissues and by performing a computer simulation that allows the prediction of protein structures and substrate-binding properties. ΔMipep mRNA was detected in all mouse tissues examined but at much lower levels than full-length Mipep. Structure prediction and docking simulation of full-length MIPEP and ΔMIPEP with substrates of MIPEP, such as malate dehydrogenase 2 (MDH2) and cytochrome c oxidase subunit 4, showed that entry of these substrates into ΔMIPEP with a low binding energy was greatly restricted. To determine levels of MIPEP substrates in the presence or absence of full-length MIPEP or ΔMIPEP, we created Mipep and ΔMipep overexpression 3T3-L1 cells and Mipep knockout (KO) cells. Western blotting showed that in Mipep KO cells Mipep overexpression slightly decreased the molecular weight of MDH2 and Sirtuin 3, another MIPEP substrate, whereas ΔMipep overexpression did not. These results indicate that ΔMIPEP fails to recognize MIPEP substrate proteins. Together, our findings indicate that ΔMIPEP is a novel splice variant that can contribute to mitochondrial signal peptidase-mediated regulation of mitochondrial protein homeostasis.

Keywords: AlphaFold2; Mitochondrial intermediate peptidase; Processing; Splice variant; ΔMIPEP.

Fig. 1

ΔMipep is a splice variant of Mipep. (A) Sequences from exon 14 to 17 and the exon structures of mouse Mipep and ΔMipep. The shaded area represents exons 15 and 16. (B) The domain structure of MIPEP and ΔMIPEP. Amino acids 493, 497, and 522 are zinc-binding sites. (C) The positions of primers for semiquantitative RT-PCR analysis of Mipep and ΔMipep cDNAs. (D) Agarose gel electrophoresis images of semiquantitative RT-PCR for full-length Mipep and ΔMipep in mouse tissues. Full-length Mipep plasmid (pMXs-AMNN-Mipep-Puro) and ΔMipep plasmid (pMXs-AMNN-ΔMipep-puro) were used as positive controls of full-length Mipep and ΔMipep, respectively. (E) The positions of primers and the probe for quantitative RT-PCR analysis of Mipep and ΔMipep cDNAs. (F) Copy numbers of full-length Mipep and ΔMipep mRNAs per total RNA μg in mouse tissues. Values are shown as the mean ± standard deviation (n = 3 or 4).

Fig. 2

Predicted structures and structural characteristics of MIPEP and ΔMIPEP proteins. (A) Structure prediction of MIPEP (magenta) and ΔMIPEP (blue) proteins using AlphaFold2. The pocket entrance size is shown for each structure. The upper image shows the merged structure of MIPEP and ΔMIPEP. (B) The cartoon representation of MIPEP is shown with specific exons highlighted. Residues 479–515 (exon 14) and 582–622 (exon 17) are colored forest green, while residues 516–581 (exons 15–16) are colored orange to emphasize the spliced region of interest. (C) The cartoon representation of ΔMIPEP (711 residues with deletion of residues 516–581, corresponding to exons 15–16) is shown. Residues 479–556, encompassing exons 14 and 17, are highlighted in forest green to emphasize the continuous region formed after exon 15–16 deletion. (D) The root-mean-square deviation (RMSD) of Cα atoms was calculated during 10 ns molecular dynamics simulations. The RMSD profile of wild-type MIPEP (blue) and ΔMIPEP (orange) is shown as a function of simulation time. The overall stability of the trajectories was assessed based on the convergence of RMSD values. (E) The root-mean-square fluctuation (RMSF) of Cα atoms was computed from the same 10 ns simulations. Per-residue fluctuations are plotted for MIPEP (blue) and ΔMIPEP (red). Peaks indicate flexible regions, whereas valleys correspond to structurally stable domains.

Fig. 3

Docking simulation between MIPEP or ΔMIPEP and substrates. (A, F) N-terminal residues of malate dehydrogenase 2 (MDH2) (A) and cytochrome c oxidase subunit 4 (COX4) (F). The cleavage sites of MDH2 and COX4 by MIPEP are shown. The yellow area indicates the 15 amino acid residues used for docking simulation. (B, D, G, I) Docking models of MDH2 (B–E) and COX4 (G–J) 15-aa fragments (yellow or orange) with AlphaFold2 models of MIPEP (magenta) and ΔMIPEP (blue), prepared using the HPEPDOCK server. The left images show space filling models, while the right images show ribbon models. The three zinc-binding sites of MIPEP are colored white. (C, E, H, J) The top 100 conformations with the highest scores of each docking simulation, respectively (B, D, G, I).

Fig. 4

Changes in molecular sizes of MIPEP substrates in Mipep knockout cells overexpressing full-length Mipep or ΔMipep. (A) The lack of MIPEP was confirmed in Mipep knockout (KO) cells. (B) Full-length MIPEP and ΔMIPEP protein levels in mock, full-length Mipep- and ΔMipep-overexpressing 3T3-L1 and Mipep KO cells. ∗ indicates a non-specific band. (C) Sirtuin 3 (SIRT3), malate dehydrogenase 2 (MDH2) and cytochrome c oxidase 4 (COX4) protein levels in mock, full-length Mipep- and ΔMipep-overexpressing cells. ∗ indicates the higher molecular weight bands, while ∗∗ indicates the lower molecular weight bands. LaminB1 was used as a loading control.