Altered mitochondrial parameters in bcl-x–deficient neurons. (A) Immunofluorescence microscopy of control (bcl-x^+/flox;NEX^+/cre) and cKO (bcl-x^flox/flox;NEX^+/cre) cortical cultures (DIV6) stained with anti–cytochrome c (green; 7H8.2C12 [1:80; BD] and goat anti–mouse Alexa Fluor 488 [Invitrogen]) to detect mitochondria and costained for Cre recombinase (red; anti-Cre [1:2,000; EMD] and Cy3 goat anti–rabbit [1:1,000; Jackson ImmunoResearch Laboratories, Inc.). Images were captured with a real-time camera (Diagnostic Instruments, Inc.) and a microscope (Eclipse E800; Nikon). Bars, 4 µm. (B) Summary of mean fluorescence intensities ± SEM from two-photon microscopy images of live cortical cultures recorded simultaneously in three channels to assess ΔΨ_m (100 nM TMRM), ROS accumulation (2 µM CM-DCF), and NAD(P)H (intrinsic fluorescence) in one mitochondria-enriched ROI per cell for multiple cells (control, n = 50; cKO, n = 56) from multiple cultures (control, n = 9; cKO, n = 11), with each culture prepared from a different embryo. a.u., arbitrary unit. Student’s t test was used; *, P < 10⁻⁶; **, P < 10⁻¹⁰; ***, P < 10⁻¹². (C) Examples of two-photon microscopy images marking example ROI. Fluorescence intensity is scaled with pseudocolors (filled arrows). N, nucleus. Bars, 10 µm. (D) A diagram of three major determinants of mitochondrial membrane potential (dashed boxes). Electron flow from NADH to O₂ (red arrow) via the ETC complexes (I–IV), proton (H⁺) paths across the membrane (blue arrows), ATP/ADP + Pi exchange via ANT and phosphate carrier (PC; gray arrow), and inner mitochondrial membrane (IM) and outer mitochondrial membrane (OM). OSCP, oligomycin sensitivity–conferring protein.

Submitochondrial localization of endogenous Bcl-x_L. (A) Immunoblots of total cell lysates (clarified), cytosolic fractions, and heavy membranes prepared from dissected cortexes pooled from three 3-d-old mice per sample. Blots were probed with antibodies to Bcl-x_L (1:1,000; provided by L. Boise), cytochrome c (Cyt c; 7H8.2C12 [1:1,000]), cytochrome oxidase subunit IV (COX IV; 1:1,000; Invitrogen), SMAC (1:1,000; Invitrogen), voltage-dependent anion channel (VDAC; 1:1,000; EMD), actin (1:1,000; MP Biomedicals), and UCP-2 (6525 [1:100; Santa Cruz Biotechnology, Inc.]). A representative of three independent experiments is shown. (B) Summary of immunogold EM staining for Bcl-x_L and ATP synthase β proteins in control and cKO mouse brain representing three independent experiments. (C) EM of microdissected CA1 hippocampus (where CA1 synapses onto CA3) from mouse brains stained with gold-labeled Bcl-x_L antibody detects Bcl-x_L on mitochondrial inner membranes/cristae (black arrows), outer membrane polar clusters (arrowheads), and adjacent membranes (line arrows). Bars, 0.1 µm. (D) Immunogold staining of bcl-x cKO CA1 mouse brain prepared in parallel as in C. Bar, 0.1 µm. (E) Immunoblot analysis of Percoll-purified nonsynaptic adult rat brain mitochondria. Mitochondria were incubated with the indicated proteases (for 30 min) with or without 0.01% digitonin to permeabilize the outer membrane and were blotted for Bcl-x_L (1:1,000; Abcam), Tom20 (11415 [1:2,000; Santa Cruz Biotechnology, Inc.]), and ATP synthase β subunit (A21351 [1:1,000; Invitrogen]).

Bcl-x_L association with inner mitochondrial membrane components. (A) Summary of yeast two-hybrid interactions between BCL-2 family proteins and the ATP synthase β subunit, including wild-type (WT) and Bcl-x_L mutants mt1 (F131V/D133A), mt7 (V135A/N136I/W137L), and mt8 (G138E/R139L/I140N). (B) Yeast (BY4741) transformed with the indicated plasmids were heat ramp treated and presented as colony counts from four determinations in two independent experiments. Data are presented as the mean ± SEM. Student’s t test was used; *, P = 0.012 compared with control; **, P = 0.001. (C) Immunoblots for Bcl-x_L (provided by L. Boise) or β subunit in total rat liver mitochondria (T), inner membrane vesicles (F1), and highly purified ATP synthasomes (F2 and F3) of increasing purity (Ko et al., 2003). The replicate blot in the middle was probed with Bcl-x_L antibody preincubated with recombinant Bcl-x_L (rBcl-x_L) protein purified from Escherichia coli (lanes F1–F3 only). (D) Coomassie blue–stained preparative SDS gel of final purification step for Bcl-x_L–binding partners from WEHI 7.1 cells. (E) Immunoblots for Bcl-x_L and β subunit in size column chromatography fractions (F) and total lysates (T). Eluted size markers are indicated.

Bcl-x_L–deficient mitochondria have fluctuating membrane potentials. (A) Continuous recordings (3.5-s intervals) of TMRM intensities per ROI of individual cortical neurons for at least 250 s (DIV4–5; Fig. 1). Traces are representative of multiple neurons in three independent experiments. (B) Fluorescence intensity (arbitrary units [a.u.]) of TMRM stain per pixel was determined for the marked mitochondria-rich region of a single neuron in bcl-x–deficient (cKO) and control (Cont) cultures shown in Fig. 1 C. (C) SDs were calculated for mean TMRM fluorescence intensities per pixel for 200 individual mitochondria derived from 20 different cells per genotype monitored every 2.5 s for at least 90 s. Each symbol represents one mitochondrion. An F-test for variance comparing control and bcl-x–deficient neuronal mitochondria was performed; P < 0.0001. (D) Continuous recordings of intracellular calcium levels at 4-s intervals in DIV3–4 cortical neurons. Initial intracellular calcium levels from four independent experiments are graphed. *, P = 0.039. (E) Mitochondrial membrane potential fluctuation increases with ABT-737. Fluorescence intensities were measured in small puncta (estimated to be one mitochondrion) near the soma in cultured rat hippocampal neurons (DIV14–16) stained with 5 nM TMRE. Relative fluorescence intensities (collected at 1/s) for the same puncta/mitochondria treated with 0.1% DMSO before and after addition of 1 µM ABT-737 (in 0.1% DMSO) for 10 min are shown. (F) SDs of TMRE intensity measurements as in E; data are for 30 measurements for each of 12 puncta in six cells in two independent experiments and are similar to three additional experiments with protocol variations. Paired t test was used; *, P = 0.02. (G) SD of TMRE as in F, except 4 d after transfection with shRNA vector with scrambled (n = 10) or bcl-x–specific shRNAs (n = 17). Fluorescent images were taken every 3 s for 8 min. Paired t test was used; **, P = 0.00027. (F and G) Data are presented as the mean ± SD.

Bcl-x_L stabilizes the mitochondrial membrane potential to conserve energy. (A) A simplified model of ion flux across the inner mitochondrial membrane. Ions may enter or leave the matrix through leakage channels at point a (e.g., the F₁F_O ATP synthase through which protons enter the mitochondrial matrix), and overall stability and membrane potential are maintained by active pumps at point b (e.g., the ETC). The double arrow represents the source of fluctuation in potential. (B) Numerical simulations of the additional ion flux that occurs as a result of fluctuations in membrane potential using the 1-µm membrane vesicle model in A. An external perturbation with a fixed amplitude between 0 and 10 pA (5-ms duration) was allowed to occur randomly with a mean interval of 1 s, and the total ion flux was integrated over 20-s periods. The graph plots the increase in total integrated flux that results from fluctuations of increasing amplitude. (C) Numerical simulations of changes in mitochondrial inner membrane potential produced by stochastic opening of a nonselective cation channel. No openings occur when probability of channel opening (Po) equals 0, and the membrane is maintained at a steady level (−180 mV; dashed lines). Indicated opening rates of the channel produce fluctuations in membrane potential accompanied by net hyperpolarization (more negative potentials). (D) Mean ± SD of membrane potentials for traces in C.

Leaky energetics in Bcl-x_L–deficient mitochondria. (A and B) Total ATP levels corrected for total protein in paired DIV3–6 cortical cultures after addition of 5 µg/ml oligomycin (control [n = 8] and cKO [n = 7] at each of six time points in three independent experiments) or 20 µg/ml aurovertin B (control [n = 5] and cKO [n = 4] at each of six time points from two independent experiments). Vertical dashed lines mark the time frame evaluated for TMRM fluorescence in C and D. Data are presented as the mean ± SEM. (C and D) Time course of TMRM fluorescence intensities (mean ± SEM) after addition of 5 µg/ml oligomycin. Analyses of 15–35 neurons from three to five fields per genotype per time point are presented, and results are representative of five similar independent experiments. Parallel experiments were performed with 20 µg/ml F₁ inhibitor aurovertin B (n = 13–35 cells from two to three fields at each time point; total of 136 control and 193 cKO cells) from two independent experiments. (E) Two-photon microscopy images of cortical neurons (DIV4) stained with 100 nM TMRM to assess ΔΨ_m as described for Fig. 1 C after addition of 6 nM oligomycin (Oligo.). A representative of five independent experiments is shown. Bar, 10 µm. (F and G) Relative NAD(P)H levels were determined in the same cellular subregions analyzed in C and D, as described for but not included in Fig. 1 B. Horizontal dashed lines mark starting NAD(P)H levels. Data are presented as relative ratios for direct comparisons. Mean ± SEM is represented, and values are the same as in C and D.

Nonapoptotic function of Bcl-x_L. (A) Two-photon microscopy of bcl-x–deficient cultured neurons was performed as described for Fig. 1 B, except using a flow chamber. The measurements shown begin with the flow of oligomycin-free medium after 40 min of 5 µg/ml oligomycin (n = 4–11 per time point per genotype). Data are presented as the mean ± SD. (B) A model depicting an increase in membrane leak in the absence of Bcl-x_L. M, matrix. (C) Heat ramp cell death assay of yeast strains with mutant ATP2 or FIS1/whi2-1 (Cheng et al., 2008; Teng et al., 2011). Representative images of yeast growth are shown; arrows indicate fivefold dilutions. Data are presented as mean ± SD for four independent strains per plasmid, each tested in duplicate in each of two independent experiments and plotted as the ratio of colony numbers for Bcl-x_L/empty vector (FIS1/whi2, n = 10; ATP2, n = 16). Student’s t test was used; *, P = 5.7 × 10⁻¹⁰.