Dynamic subcompartmentalization of the mitochondrial inner membrane

Abstract

The inner membrane of mitochondria is organized in two morphologically distinct domains, the inner boundary membrane (IBM) and the cristae membrane (CM), which are connected by narrow, tubular cristae junctions. The protein composition of these domains, their dynamics, and their biogenesis and maintenance are poorly understood at the molecular level. We have used quantitative immunoelectron microscopy to determine the distribution of a collection of representative proteins in yeast mitochondria belonging to seven major processes: oxidative phosphorylation, protein translocation, metabolite exchange, mitochondrial morphology, protein translation, iron-sulfur biogenesis, and protein degradation. We show that proteins are distributed in an uneven, yet not exclusive, manner between IBM and CM. The individual distributions reflect the physiological functions of proteins. Moreover, proteins can redistribute between the domains upon changes of the physiological state of the cell. Impairing assembly of complex III affects the distribution of partially assembled subunits. We propose a model for the generation of this dynamic subcompartmentalization of the mitochondrial inner membrane.

Figure 1.

Localization of mitochondrial proteins by immuno-EM and in silico accumulation of gold particles onto an empiric model. S. cerevisiae wild-type (D273-10B) or matrix-targeted GFP-expressing cells were grown to early log phase in liquid complete media containing 2% lactate, chemically fixed, cryosectioned, and immunogold labeled. The location of gold particles found in mitochondria showing clearly resolvable CMs connected by cristae junctions to the IBM were plotted onto a single, empirically determined, drawn to scale model (see Materials and methods). (A) Model representing part of OM, IBM, CM, and matrix. The zones were defined as follows: OM/IBM, the center of a gold particle is ≤14 nm from the OM or IBM; CM, the center of a gold particle is ≤14 nm in distance from the CM and not in the OM/IBM zone; background, the center of a gold particle is neither in the OM/IBM nor the CM zone. Assignment to the matrix was done for gold particles that are in the background zone located on the matrix side of the inner membrane. (B and C) Exemplary alignment of the model with mitochondria that were immunogold labeled for Core1 or Tom20, respectively. Arrow points to gold particle plotted onto model shown. (D–F) Graphical representations of the distribution of Core1, Tom20, and mtGFP. Numbers at x and y axes represent distances in nanometers. Bars, 100 nm.

Figure 2.

Analysis and validation of subcompartmental localization of Tim23 and Cox2. (A) Distribution of Tim23 in wild-type cells under respiratory growth conditions was determined as described in Fig. 1 and Materials and methods. (B) Mitochondria from a GFP-Tim23–expressing strain were or were not treated with proteinase K and analyzed by Western blot analysis with indicated antibodies. (C) Distribution of GFP-Tim23 after immunogold labeling with antibodies raised against GFP. (D) Distribution of Cox2 in wild-type cells under respiratory growth conditions determined as described in Fig. 1. Numbers at x and y axes represent distances in nanometers. (E) Quantification of the distribution of Cox2, Tim23, and GFP-Tim23 in the OM/IBM and the CM zone. Signals were corrected for background and normalized per unit membrane length (see Materials and methods). For Cox2 and Tim23, the experiment was repeated three times using identical fixed cell material. Error bars represent the SD.

Figure 3.

Subcompartmental organization of mitochondrial proteins. Quantification of the distribution of mitochondrial proteins in wild-type cells under respiratory growth conditions was performed as described in Figs. 1 and 2 and Materials and methods. (A) Proteins involved in OXPHOS. (B) Proteins involved in the insertion of nuclear- and mitochondrial-encoded proteins.

Figure 4.

Subcompartmental localization of Mgm1 involved in mitochondrial fusion. Quantification of the distribution of mitochondrial proteins in wild-type cells under respiratory growth conditions was as described in Figs. 1 and 2 and Materials and methods. (A) Proteins involved in iron–sulfur biogenesis, Isd11 and Nfs1. (B) Proteins involved in protein degradation in mitochondria, prohibitin 1 (Phb1) and prohibitin 2 (Phb2). (C) Distribution of the mitochondrial fusion protein Mgm1. Inset, quantification of Mgm1 density in the OM/IBM and the CM zone. Numbers at x and y axes represent distances in nanometers.

Figure 5.

Dynamic redistribution of mitochondrial proteins involved in protein translocation and synthesis. Cells overexpressing mtGFP were or were not treated with puromycin for 30 min before fixation and immuno-EM. (A) Quantification of Tim23 distribution before or after 30 min of puromycin treatment. The experiment was repeated three times using identical fixed cell material. Error bars represent the SD. (B) Distribution of Mrpl36 before the addition of puromycin. (C) Distribution of Mrpl36 30 min after the addition of puromycin. (D) Location of Mrpl36 in the OM/IBM, CM, or matrix before and 30 min after the addition of puromycin. Numbers at x and y axes represent distances in nanometers.

Figure 6.

Subcompartmental distribution of assembly intermediates of complex III and IV. Wild-type and Δbcs1 cells were grown on complete media containing 2% galactose, chemically fixed, cryosectioned, and immunogold labeled for Cox2, Core1, and Rip. Quantification of the distribution of mitochondrial proteins was as described in Figs. 1 and 2 and Materials and methods. In the Δbcs1 strain, complex III is only partially assembled, and the supracomplexes III₂IV₂ are not formed.