A Bcl-xL-Drp1 complex regulates synaptic vesicle membrane dynamics during endocytosis - PubMed (original) (raw)

doi: 10.1038/ncb2791. Epub 2013 Jun 23.

Kambiz N Alavian, Emma Lazrove, Nabil Mehta, Adrienne Jones, Ping Zhang, Pawel Licznerski, Morven Graham, Takuma Uo, Junhua Guo, Christoph Rahner, Ronald S Duman, Richard S Morrison, Elizabeth A Jonas

Affiliations

A Bcl-xL-Drp1 complex regulates synaptic vesicle membrane dynamics during endocytosis

Hongmei Li et al. Nat Cell Biol. 2013 Jul.

Abstract

Following exocytosis, the rate of recovery of neurotransmitter release is determined by vesicle retrieval from the plasma membrane and by recruitment of vesicles from reserve pools within the synapse, which is dependent on mitochondrial ATP. The anti-apoptotic Bcl-2 family protein Bcl-xL also regulates neurotransmitter release and recovery in part by increasing ATP availability from mitochondria. We now find, that Bcl-xL directly regulates endocytic vesicle retrieval in hippocampal neurons through protein-protein interaction with components of the clathrin complex. Our evidence suggests that, during synaptic stimulation, Bcl-xL translocates to clathrin-coated pits in a calmodulin-dependent manner and forms a complex with the GTPase Drp1, Mff and clathrin. Depletion of Drp1 produces misformed endocytic vesicles. Mutagenesis studies suggest that formation of the Bcl-xL-Drp1 complex is necessary for the enhanced rate of vesicle endocytosis produced by Bcl-xL, thus providing a mechanism for presynaptic plasticity.

PubMed Disclaimer

Figures

Fig. 1

Fig. 1. Bcl-xL over-expression enhances the rate of release of styryl dyes in hippocampal neurons

a. Diagram of protocols for experiments (1-3) described in text. b. Examples of exponential fits to normalized fluorescence change of FM 5-95 puncta during stimulation with 90 mM KCl. Left panel: Bi-exponential fit to normalized fluorescence change from cells over-expressing Bcl-xL (ExpDec2 of Bcl-xL=bi-exponential decay of fluorescence). Center panel: Dotted lines model two different exponents of left panel (ExpDec of Bcl-xL=bi-exponential fit to fluorescence decay shown as one dotted line for each exponent). Right panel: Single exponential fit to the normalized fluorescence change from a GFP control neuron (ExpDec1 of control=single exponential decay of fluorescence). c. Time course of normalized fluorescence of 5 μM Fluo-4 measured in synaptic boutons during electrical stimulation (100 action potentials at 10 Hz; n=15 DsRed expressing cells, n=15 DsRed-Bcl-xL expressing cells). d. Peak fluorescence after loading with Exp 1 protocol (prior to unloading) of individual FM 5-95 puncta (n=10 GFP CTL puncta, 14 GFP- Bcl-xL puncta; **p < 0.0013; three independent cultures). In all experiments, Bcl-xL expressing and GFP-expressing controls are pairs from the same culture. e. Normalized fluorescence change in 90 mM KCl (n=10 GFP CTL puncta, 14 GFP- Bcl-xL puncta). f. Time constants for exponential fits to all experiments in e. g. Relative weights of exponential pools for experiments in e. h. Peak fluorescence of FM 5-95 puncta (protocol for Exp 2); n=14 GFP CTL puncta, 15 GFP-Bcl-xL puncta; three independent cultures; ***p < 0.0001). i. Normalized fluorescence change in 90 mM KCl (Exp 2 protocol; n=14 GFP CTL puncta, 15 GFP-Bcl-xL puncta). j. Time constants for exponential fits to experiments in i. k. Relative weights of exponential pools for experiments in i. l. Peak fluorescence of FM 5-95 puncta (Exp 3 protocol; n=11 GFP CTL puncta, 20 GFP-Bcl-xL puncta; three independent cultures). m. Normalized fluorescence change in 90 mM KCl (Exp 3 protocol; n=5 GFP CTL puncta, 10 GFP-Bcl-xL puncta). The initial 10 s of fluorescence decline is shown. 5 CTL puncta are represented to emphasize the bi-exponential effect found in a subgroup of CTL puncta. No punctal fluorescence changes from GFP-Bcl-xL expressing cells required bi-exponential fits. n. Time constants for exponential fits to experiments such as in m (n=9 GFP CTL puncta, n=10 GFP-Bcl-xL puncta). o. Relative weights of exponential pools for experiments in m. Statistics are represented as mean +/- S.E.M.

Fig. 2

Fig. 2. Endogenous Bcl-xL participates in normal vesicle pool dynamics

a. Diagram of protocol 1 for all experiments in Figure 2. b. Peak fluorescence of FM 5-95 puncta (n=13 scrambled shRNA, 10 Bcl-xL shRNA; *p<0.03; puncta from 5 different cells, three independent cultures). c. Normalized fluorescence in 90 mM KCl (n=9 scrambled shRNA, n=7 Bcl-xL shRNA; 5 different cells, three independent cultures. d. Time constants for exponential fits to experiments in c (n=12 scrambled shRNA, n=7 Bcl-xL shRNA ***p< 0.0001 comparing tau 2 for the two conditions). e. Relative weights of two exponential pools for experiments in c. f. Peak fluorescence of FM 5-95 in CTL and CaMi-treated cells (n=10 puncta each; **p<0.003; 5 different cells each from three independent cultures). Cells were exposed to 30 μM CaMi or vehicle for 30 min. prior to loading with FM 5-95. g. Normalized fluorescence in 90 mM KCl (n=10 CTL, n=8 puncta of CaMi-treated cells; 5 different cells for each condition from three independent cultures). h. Time constants for exponential fits to experiments in g (***p=0.0009). i. Relative weights of exponential pools for experiments in g. j. Peak fluorescence of FM 5-95 puncta in GFP-Bcl-xL expressing cells and GFP-Bcl-xL expressing cells treated with CaMi (n=8 each; 3 different coverslips in each condition). k. Normalized fluorescence in 90 mM KCl (n=7 GFP- Bcl-xL and 8 GFP- Bcl-xL + CaMi puncta; 3 different cells for each condition). l. Time constants for exponential fits to experiments in k (n=12 GFP- Bcl-xL and 7 GFP- Bcl-xL + CaMi puncta; **p=0.0069, comparing tau 1 for the two conditions). m. Relative weights of exponential pools for experiments in k. Statistics are represented as mean +/- S.E.M.

Fig 3

Fig 3. Bcl-xL increases the rate of mitochondrial ATP-resistant early endocytosis

a. Normalized fluorescence change of synaptopHluorin puncta before, during and after electrical stimulation (with 100 action potentials, 10 Hz; n=15 puncta from DsRed-Bcl-xL cells, n=15 mito-RFP CTL). Inset: time constants of change in fluorescence after stimulation (***p=0.0002; 3 independent cultures. Experiment repeated with 8 different cultures. b. Normalized fluorescence (n=15 DsRed control puncta before and after 5 μM bafilomycin). Non-normalized data: Suppl. Fig. 5a. c. Normalized fluorescence (n=15 DsRed Bcl-xL puncta before and after bafilomycin). d. Ratio of difference in peak fluorescence before and after bafilomycin to peak in bafilomycin of traces in b and c (***p< 0.0001). e. Normalized fluorescence (n=15 puncta from each DsRed-Bcl-xL shRNA or DsRed (CTL) cells). Inset: time course of change in fluorescence after stimulation (n=15 puncta each group; *p=0.0203; 3 independent cultures). f. Normalized fluorescence (n=15 puncta from DsRed controls before and after bafilomycin). g. Normalized fluorescence (n=15 puncta from DsRed Bcl-xL shRNA before and after bafilomycin). h. Ratio of difference in peak fluorescence before and after bafilomycin over peak in bafilomycin of traces in f and g (***p< 0.0001). i. Normalized fluorescence (n=15 puncta from DsRed control or DsRed Bcl-xL shRNA, or both DsRed Bcl-xL shRNA plus shRNA-resistant Bcl-xL; 3 independent coverslips in each group. j. Normalized fluorescence for vehicle or 5 min. after exposure to 10 μM ABT-737; N=6). Inset: time constants of fluorescence changes (*p<0.02). k. Phase image of single autaptic hippocampal neuron in culture DIV 16. Scale bar= 20 μm. l. Synaptic depression of GFP-Bcl-xL or control GFP (10 Hz stimulation, first and last 10 of 100 synaptic responses). m. Summary data of percentage change in average amplitude of last 10 compared to first 10 synaptic responses (n=8 control GFP and 9 GFP-Bcl-xL cells, two independent cultures; *p = 0.017). n. ATP levels (luciferin/luciferase) of lysed neurons expressing the indicated constructs; oligomycin=5 μg/ml for 5 min; n=12 wells each group,*** p<0.0001). o. Normalized fluorescence in 5 μg/ml oligomycin (stimulation 5 min. after oligomycin application n=10 puncta each from DsRed-Bcl-xL or mito-RFP (CTL) expressing cells). p. Time constants of change in fluorescence after stimulation for data in o (n=10 each for oligomycin; ***p<0.0001 comparing Bcl-xL to control; non-oligomycin puncta n=37 for each condition, ***p<0.0001 comparing Bcl-xL to control; oligomycin CTL data compared to non-oligo control data; ***p<0.0001; Bcl-xL oligomycin-treated compared to non-oligomycin ***p=0.0005). q. Normalized fluorescence before and after 5 μM bafilomycin in oligomycin-treated cells (n=10 puncta, three independent cover slips; mito-RFP controls). r. Normalized fluorescence before and after 5 μM bafilomycin (n=10 puncta, three independent cover slips, DsRed-Bcl-xL). s. Ratio of difference in peak fluorescence before and after bafilomycin over peak in bafilomycin of traces shown in q and r (***p< 0.0001). Statistics are represented as mean +/- S.E.M.

Fig. 4

Fig. 4. Calmodulin-dependent Bcl-xL translocation to synaptic vesicle membranes in stimulated neurons

a. Immunoblots for Bcl-xL of indicated sub-fractions of cell lysates of cultured unstimulated hippocampal neurons or stimulated with 90 mM KCl for 90 s with or without CaMi. GAPDH serves as protein control for cytosolic protein amount, COX IV as control for mitochondrial protein amount, synaptotagmin as control for synaptic vesicle membrane protein amount. b. Group data for protein densities for all experiments in a (n=3 blots; *p < 0.03, one-tailed unpaired student's t test of protein amount after stimulation compared to before stimulation; **p < 0.009, one-tailed unpaired student's t test of protein amount after stimulation compared to before stimulation). c. Immunoblots for Bcl-xL of indicated sub-fractions of cell lysates of cultured unstimulated or stimulated neurons. The right lane of each cell subfraction indicates level of Bcl-xL from cells transduced with DsRed calmodulin shRNA. GAPDH serves as protein control for cytosolic protein amount, COX IV as control for mitochondrial protein amount, synaptotagmin as control for synaptic vesicle membrane protein amount. d. Group data for protein densities for all experiments in c (n=3 blots; *p < 0.03, **p=0.0081 one-tailed unpaired student's t test of protein amount after stimulation compared to before stimulation). e. Normalized (to starting value) fluorescence change of synaptopHluorin puncta before, during and after electrical stimulation (with 100 action potentials, 10 Hz; n=10 puncta from DsRed expressing cells before and after 5 μM bafilomycin). f. Normalized fluorescence change (n=10 puncta from DsRed CaM shRNA expressing cells before and after bafilomycin. g. Ratio of difference in peak fluorescence change before and after bafilomycin to the peak fluorescence in bafilomycin of traces shown in e and f (**p=0.006). h. Normalized fluorescence change (n=10 puncta for calmodulin shRNA or scrambled shRNA; 3 cells each, one culture). Inset: bar graphs of time constants of fluorescence changes (*p=0.038). Statistics are represented as mean +/- S.E.M.

Fig. 5

Fig. 5. Drp1 is co-localized with clathrin and Mff on synaptic vesicles

a. Immuno-electron micrographs of Drp1 on clathrin-coated pits. Cultured hippocampal neurons were stimulated with 90 mM KCl for 90 s prior to fixation for electron microscopy (EM). b. Co-immunoprecipitation of whole rat brain lysate using indicated antibodies (IP refers to antibodies used for immunoprecipitation. Titles at right of blot refer to antibodies used for immunoblotting. Left lane represents cell lysate). c. Immunoblots of subcellular fractions (as indicated) of rat brain lysate using the indicated antibodies. Left lane shows cell lysate. d. Co-immunoprecipitation of purified synaptic vesicle fraction using the indicated antibodies (left lane represents cell lysate). e. Normalized fluorescence change before, during and after electrical stimulation (n=15 puncta from DsRed controls or DsRed Drp1 shRNA; 3 independent cultures each group. f. Normalized fluorescence change (n=15 puncta each DsRed control (red trace, same control data as for Bcl-xL shRNA in 3i), DsRed Drp1 shRNA expressing cells (black trace), or from cells expressing both DsRed Drp1shRNA plus Drp1 shRNA-resistant Drp1 overexpressing construct (blue trace). 3 independent cultures). g. Normalized fluorescence change (n=15 puncta from DsRed control expressing cells before and after bafilomycin). These control data are the same as the DsRed control data in Fig. 3f but are provided here for comparison to 5h. h. Normalized fluorescence change (n=15 puncta from DsRed Drp1 shRNA expressing cells before and after bafilomycin). i. Peak fluorescence change ratios of traces shown in g and h (***p< 0.0001). j. Co-immunoprecipitation of whole rat brain lysate using indicated antibodies (IP refers to antibodies used for immunoprecipitation; Titles at right of blot refers to antibodies used for immunoblotting. Left lane represents cell lysate). k. Normalized fluorescence change; n=12 puncta from scrambled shRNA expressing cells before and after 5 μM bafilomycin, 12 different coverslips, 2 independent cultures. l. Normalized fluorescence change; n=12 puncta from Mff shRNA expressing cells before and after bafilomycin, 13 different coverslips, 2 independent cultures. m. Peak fluorescence change ratios of traces shown in k and l (***p< 0.0001). Statistics are represented as mean +/- S.E.M.

Fig. 6

Fig. 6. Bcl-xL and Drp1 co-localize with synaptophysin on synaptic vesicles

a. Immuno-electron micrographs co-localizing clathrin and Bcl-xL. Cultured hippocampal neurons were stimulated with 90 mM KCl for 90 s just prior to fixation for EM. b. Co-immunoprecipitation using indicated antibodies (IP refers to antibody used for immunoprecipitation; antibodies used for immunoblotting are labeled at right). Left lane represents cell lysate. c. Average number of gold particles in electron micrographs of synapses. n=25 micrographs of non-stimulated cells, n=27 micrographs of stimulated cells. First two histograms represent clathrin-labeled particles, second two histograms represent Bcl-xL-labeled particles, third set of histograms represents co-localizing Bcl-xL- and clathrin-labeled particles on the same synaptic vesicle (p=0.0014). d. (Top) Immuno-electron micrographs showing co-localization of Drp1 and synaptophysin in unstimulated (left panel) and stimulated synapses (right panel). (Bottom) Immuno-electron micrographs showing co-localization of Bcl-xL and synaptophysin in unstimulated (left panel) and stimulated synapses (right panel). e. Average number of gold particles in electron micrographs of synapses treated with antibodies and stimulation as labeled (left to right: n=20, 17, 20, 17, 11, 11, 11, 11; **p=0.0081 (Drp1 nonstim. vs. Drp1 stim.). ***p<0.0001 (co-localization of Drp1 and synaptophysin non-stim. vs. stim.). *p=0.0356 (Bcl-xL non-stim. vs. Bcl-xL stim.). **p=0.0019 (co-localization of Bcl-xL and synaptophysin non-stim. vs. stim.). f. (Left top) Immuno-electron micrographs showing co-localization of Bcl-xL (10 nm gold balls) and synaptophysin (5 nm gold balls). (Left bottom) Immuno-electron micrographs showing co-localization of Drp1 (10 nm gold balls) and synaptophysin (5 nm gold balls). In the schematic diagrams to the right of, and superimposed on, the images, the presynaptic membrane is colored in light blue, synaptic vesicles are colored in red, gold balls indicating Bcl-xL or Drp1 antibody labeling are colored in dark blue, gold balls indicating synaptophysin antibody labeling are colored in yellow, green indicates mitochondria. Enlarged insets show synaptophysin, Drp1 and Bcl-xL immunolabeling (yellow arrows indicate synaptophysin antibody labeling, blue arrows indicate Bcl-xL (top two images) or Drp1 (bottom two images). The right-most panels picture mitochondria (top shows Bcl-xL immunolabeling, bottom shows Drp1 immunolabeling. Statistics are represented as mean +/- S.E.M.

Fig. 7

Fig. 7. Drp1 is required for formation of normal endocytotic vesicles

a. Electron micrographs of resting synapses of hippocampal neurons transduced with scrambled GFP-shRNA. Red arrows demonstrate representative control synaptic vesicles. Scale bar is shown in 7b. b. Electron micrographs of resting synapses of hippocampal neurons transduced with GFP-Drp1 shRNA. Black arrows indicate enlarged vesicles and (in lower right hand panel) a vesicle with an elongated neck. White arrows indicate clathrin-coated pits, which tend to be more frequently seen in micrographs of synapses of Drp1 shRNA expressing neurons. Scale bar = 500 nm. c. Histograms of all vesicle area measurements (n=6 micrographs of scrambled shRNA expressing neurons, 347 vesicles; n=8 micrographs of Drp1 shRNA expressing neurons, 667 vesicles. d. Electron micrographs of stimulated synapses of hippocampal neurons transduced with scrambled GFP shRNA (top image) or with GFP Drp1 shRNA (bottom image); blots were fixed just after stimulation with 90 mM KCl in the presence of horse radish peroxidase (HRP). Scale bar = 200 nm. Red arrows demonstrate HRP-labeled vesicles in controls, black arrows demonstrate HRP-labeled vesicles in Drp1 shRNA. White arrow shows an endocytotic pit pinching off the plasma membrane. e. Quantification of area (μm2) of all HRP-labeled vesicles (n=9 images, 98 total vesicles for control; n=11 images, 127 total vesicles for Drp1 shRNA, ***p<0.0001). f. Cryo-electron micrographs immuno-labeled with gold-tagged anti-VAMP antibodies after stimulation. Scale bar = 200 nm. g. Electron micrographs of synapses of hippocampal neurons transduced with scrambled GFP shRNA (top images) or with GFP Drp1 shRNA (bottom images). Left panels represent unstimulated synapses, right panels represent stimulated synapses. Scale bar = 500 nm. h-i. Histograms of all vesicle area measurements (μm2) from panels in f, g (n=8 micrographs of control scrambled shRNA unstimulated synapses, n=5 micrographs of control scrambled shRNA stimulated synapses, n=9 micrographs of Drp1 shRNA unstimulated synapses, n=12 micrographs of Drp1 shRNA stimulated synapses, n=19 micrographs of Drp1 shRNA synapses immunogold labeled with anti-VAMP antibodies). j. Quantification of number of vesicles in synapses such as shown in g (n=9 micrographs (synapses) each for unstimulated and stimulated scrambled shRNA, 9 micrographs for unstimulated Drp1 shRNA, 11 micrographs for stimulated Drp1 shRNA; *p = 0.0306, **p = 0.002, ***p<0.0001). k. Quantification of synapse area (μm2) of all synapses measured in j (*p=0.043; **p=0.007, ***p<0.0001). Statistics are represented as mean +/- S.E.M.

Figure 8

Figure 8. Mutations in BH2 domain of Bcl-xL disrupt physical and functional interaction with Drp1

a. CED-9 and Bcl-xL protein sequences. Green and yellow boxes specify BH1 and BH2 domains, respectively. Red boxes denote conserved mutated amino acids within each domain. b. Immunoprecipitation by an anti-FLAG antibody of the wild type and mutant (M1 and M2) FLAG-tagged Bcl-xL proteins. Samples were immunoblotted for Drp1 and Bcl-xL. Lysate studies are shown in the right panel. c. Normalized (to starting value) fluorescence change of synaptopHluorin puncta in hippocampal neurons expressing indicated constructs before, during and after electrical stimulation (with 100 action potentials, 10 Hz), n=10 puncta each from 3 cells each from one culture. d. Group data for all experiments shown in c (*p=0.0169; ***p=0.0004). e. Model showing order of events: 1) Neuronal stimulation leads to calcium influx. 2) Calcium binds to and activates Calmodulin (CaM), which is inhibited by calmidazolium (CaMi). 3) CaM causes the translocation of Bcl-xL and Drp1 to clathrin-coated pits where Bcl-xL binds to and activates Drp1. 4) Drp1 is in part responsible for the formation of normally curved endocytotic vesicles. Knock down or functional inhibition of Bcl-xL, Drp1 or CaM disrupts this process, slowing endocytosis.

References

    1. Adams JM, Cory S. The Bcl-2 apoptotic switch in cancer development and therapy. Oncogene. 2007;26:1324–1337. - PMC - PubMed
    1. Banasiak KJ, Xia Y, Haddad GG. Mechanisms underlying hypoxia-induced neuronal apoptosis. Prog Neurobiol. 2000;62:215–249. - PubMed
    1. Youle RJ, Strasser A. The BCL-2 protein family: opposing activities that mediate cell death. Nat Rev Mol Cell Biol. 2008;9:47–59. - PubMed
    1. Fannjiang Y, et al. BAK alters neuronal excitability and can switch from anti- to pro-death function during postnatal development. Developmental cell. 2003;4:575–585. - PubMed
    1. Kim H, et al. Hierarchical regulation of mitochondrion-dependent apoptosis by BCL-2 subfamilies. Nat Cell Biol. 2006;8:1348–1358. see comment. - PubMed

MeSH terms

Substances

Grants and funding

LinkOut - more resources