Oligomer-based functions of mitochondrial porin

Abstract

Porin, or the voltage-dependent anion channel (VDAC), is a primary β-barrel channel in the mitochondrial outer membrane. It transports small metabolites and ions through its β-barrel pore and plays key roles in apoptosis and inflammatory response. Here we report the cryo-electron microscopy structure of yeast porin (Por1) in its hexameric form at 3.2 Å resolution. This structure allows us to introduce various mutations at the protomer interfaces, uncovering three critical functions of Por1 assembly beyond transport. Por1 binds unassembled Tom22, a subunit of the mitochondrial protein import gate (the TOM complex), to facilitate protein import into the intermembrane space, maintains proper mitochondrial lipid composition in the outer membrane through lipid scramblase activity, and contributes to the retention and regulated loss of mitochondrial DNA, in cooperation with nucleases identified through screening enabled by the obtained Por1 mutant.

Fig. 1. Purification and structure of yeast Por1.

a Elution profile obtained by gel filtration of the purified Por1 in 0.02% GDN. b SDS-PAGE (left) and blue-native PAGE (right) gels of Por1 stained with Coomassie Brilliant Blue (CBB). c Cryo-EM density of the Por1 hexamer. d Cartoon backbone representation of the Por1 hexamer. e Isolated mitochondria with Por1 derivatives containing a pair of Cys residues (right) were subjected to crosslinking by M2M, and then solubilized and analyzed by SDS-PAGE and immunoblotting with anti-Por1 antibodies. Por1, Por1 monomer; Por1-Por1, crosslinked (Por1)₂ dimer; Por1-Por1-Por1, crosslinked (Por1)₃. Source data are provided as a Source Data file.

Fig. 2. Por1(KE) mutant is defective in Tom22 binding and impairs the MIApathway import.

a por1Δ cells expressing Por1 and mutant Por1 with the indicated mutations, or without expression (vector), were grown in SCD for at 30 °C. Cell extracts were prepared, and proteins were analyzed by SDS-PAGE and immunoblotting with the indicated antibodies. Pgk1, yeast phosphoglycerate kinase 1. b Serial dilutions of the por1Δ cells expressing Por1 and Por1 with the indicated mutations, or without expression (vector), were spotted on SCD and SCLac media and grown for 2 (SCD) or 3 (SCLac) days at 30 °C and 37 °C. c In this NADH transport assay, oxidation of externally added NADH was measured for the indicated reaction times for the indicated mitochondria (por1Δ cells expressing Por1 (WT) and mutant Por1 with the indicated mutations, or without expression (vector), which were isolated from cells cultured in YPD. Values are means ± SEMs (n = 3). d In vivo photocrosslinking of benzoylphenylalanine (BPA) at residue 118 of Tom22 in the indicated strains (por1Δ cells expressing WT Por1-HA and Por1(KE)-HA) was detected by anti-Tom22 antibodies (left) and anti-HA antibodies (right). 22, Tom22; 22-Por1HA, Tom22 crosslinked with Por1-HA. e Proteins in mitochondria isolated from the indicated strains were analyzed with the indicated antibodies. f, g The indicated mitochondria (WT, por1Δ, and por1Δ expressing Por1(KE)), isolated from cells grown in YPD, were incubated with the indicated radiolabeled precursors for the indicated times at 25 °C. Tim9 and Tim10 are IMS-localized MIA-pathway substrates, and Su9-DHFR and Hsp60 are matrix-localized TIM23-pathway substrates. Then the mitochondria were treated with 50 μg/mL proteinase K for 15 min on ice (for Su9-DHFR and Hsp60) or diluted 10-fold with ice-cold buffer containing 50 mM iodoacetamide and 50 μg/mL proteinase K for 15 min on ice (for Tim9 and Tim10), and PMSF was added to 1 mM. The proteins were analyzed by SDS-PAGE and radioimaging; quantification was plotted against time (bottom), where the maximum amount for WT mitochondria was set to 100%. Values are means ± SEMs (n = 3). C, 10% of the input. Source data are provided as a Source Data file.

Fig. 3. Por1 hexamer has a phospholipid scramblase activity.

a CGMD simulation of Por1(QE) and WT Por1 hexamers were conducted. Three independent simulations were run for 10 μs for Por1(QE) and 30 μs for WT Por1. The protein structures and their initial positions are those at the start of the simulation. Red dots indicate the positions of the phospholipid head groups (PO₄) that underwent scrambling, indicating that scrambling is more likely to occur at the B-C and B’-C’ interfaces. b Each atom of the residues frequently involved in scrambling at the B/C interface is represented by a sphere in both the side view (left) and top view (right), and the corresponding residues are listed (center). Hydrophilic or polar residues are highlighted in red. c Total phospholipids were extracted from the indicated cells grown in YPD containing [³²P]-Pi and analyzed by thin-layer chromatography and radioimaging (Left). CL, cardiolipin; PA, phosphatidic acid; PE, phosphatidylethanolamine; PS, phosphatidylserine: PI, phosphatidylinositol; PC, phosphatidylcholine. The relative amounts of each phospholipid are shown as a percentage of total phospholipids (Top Right) and compared to those in WT control cells (Bottom Right). Values are means ± SEMs (n = 3). ns, not significant. *p ≤ 0.046, p-values were obtained from the unpaired two-tailed t test. P-value for CL is 0.046. Source data are provided as a Source Data file.

Fig. 4. The Por1(Q194A) mutant causes mtDNA loss.

a Serial dilutions of the WT and por1Δ cells expressing Por1-HA and Por1(QA)-HA, or without expression (vector), were spotted on SD and SLac media and grown for 2 (SD) or 6 (SLac) days at 30 °C and 37 °C. b WT and por1Δ cells expressing Por1 and Por1(QA), or without expression (vector), together with the expression of Su9-RFP, were incubated with SYBR Green for mtDNA labeling in SCD media at 30 °C for 4 h and imaged by confocal microscopy. Scale bars: 5 μm (main images), 1 μm (inset). c–e Quantification of mtDNA in por1Δ cells expressing the indicated Por1 mutants from its own promoter via plasmid transformation. The amount of mtDNA was calculated based on the Cq value of the mtDNA-encoded COX2 (c, d) or 15S rRNA (e) gene, normalized to the nuclear-encoded ACT1 gene. The Y-axis represents mtDNA (Fold) normalized to the W303-1A WT cells. Values are means ± SEMs (n = 3). Source data are provided as a Source Data file.

Fig. 5. Identification of MDL proteins involved in the mtDNA loss by Por1(QA).

a–c Quantification of mtDNA in the double deletion mutants of POR1 and indicated genes without expression (a) or with expression of WT Por1 (b) and Por1(QA) (c) from its own promoter via plasmid transformation. The amount of mtDNA was calculated as in Fig. 4c. Values are means ± SEMs (n = 3). Source data are provided as a Source Data file.

Fig. 6. mtDNA is localized in mitochondria in the Por1 (QA) strains lacking the MDL proteins.

a, b Cells lacking Por1 (a) and those lacking Por1 and one of the MDL proteins (b) but expressing Por1(WT) (a, left) or Por1(QA) (a, right; b) from their own promoters via plasmid transformation, were incubated with SYBR Green for mtDNA labeling in SCD medium at 30 °C for 4 h. Z-max projection images are shown. The boxed regions are magnified. Experiments were performed independently three times. Scale bars: 5 μm (main images), 1 μm (inset). Fluorescence intensity (F.I. (arbitrary unit)) profile of mtDNA (green) and Su9-RFP (magenta) along the orange lines in the left panels are shown in the right panels. Orange arrowheads indicate mtDNA-positive and Su9-negative signals. c A proposed model for the mtDNA loss caused by Por1(QA). The roles of Por1(QA) and Yme2 in mtDNA escape from mitochondria are not clear. Source data are provided as a Source Data file.