Synaptic vesicle endocytosis is detected by light microscopy of pHluorin fused with synaptic vesicle protein and by electron microscopy of vesicle uptake.
During endocytosis, fused synaptic vesicles are retrieved at nerve terminals, allowing for vesicle recycling and thus the maintenance of synaptic transmission during repetitive nerve firing. Impaired endocytosis in pathological conditions leads to decreases in synaptic strength and brain functions. Here, we describe methods used to measure synaptic vesicle endocytosis at the mammalian hippocampal synapse in neuronal culture. We monitored synaptic vesicle protein endocytosis by fusing a synaptic vesicular membrane protein, including synaptophysin and VAMP2/synaptobrevin, at the vesicular lumenal side, with pHluorin, a pH-sensitive green fluorescent protein that increases its fluorescence intensity as the pH increases. During exocytosis, vesicular lumen pH increases, whereas during endocytosis vesicular lumen pH is re-acidified. Thus, an increase of pHluorin fluorescence intensity indicates fusion, whereas a decrease indicates endocytosis of the labelled synaptic vesicle protein. In addition to using the pHluorin imaging method to record endocytosis, we monitored vesicular membrane endocytosis by electron microscopy (EM) measurements of Horseradish peroxidase (HRP) uptake by vesicles. Finally, we monitored the formation of nerve terminal membrane pits at various times after high potassium-induced depolarization. The time course of HRP uptake and membrane pit formation indicates the time course of endocytosis.
Neurotransmitters are stored in synaptic vesicles and released by exocytosis. The synaptic vesicle membrane and protein are then internalized by endocytosis, and reused in the next round of exocytosis. Endocytosis of synaptic vesicles is important for maintaining synaptic vesicle pools and removes protruding vesicles from the plasma membrane. The pH-sensitive green fluorescent protein pHluorin, which is quenched in acidic circumstances and dequenched in neutral pH, has been used to measure endocytosis time courses in live cells1,2,3. The pHluorin protein is typically attached to the lumenal side of synaptic vesicle proteins, such as synaptophysin or VAMP2/synaptobrevin. At rest, pHluorin is quenched in the 5.5 pH lumen of synaptic vesicles. Vesicle fusion to the plasma membrane exposes the vesicular lumen to the extracellular solution where the pH is ~7.3, resulting in an increase in pHluorin fluorescence. After exocytosis, the increased fluorescence decays, due to endocytosis of synaptic vesicle proteins followed by vesicle re-acidification within those recovered vesicles. Although the decay reflects both endocytosis and vesicular re-acidification, it mostly reflects endocytosis, because re-acidification is faster than endocytosis in most conditions1,4. The time constant of re-acidification is 3-4 s or less5,6, which is generally faster than the 10 s or more required for vesicle endocytosis4,5. If experiments are needed to distinguish endocytosis from re-acidification, acid quenching experiments using the 4-Morpholineethanesulfonic acid (MES) solution (25 mM) with a pH of 5.5 can be used to determine whether synaptic vesicle proteins are retrieved from the plasma membrane via endocytosis1,3,4. Thus, the pHluorin fluorescence intensity increase reflects a balance of exo- and endocytosis, and the decrease after nerve stimulation specifically reflects endocytosis.
pHluorin imaging may be used not only to measure the time course of endocytosis, but also the size of synaptic vesicle pools7,8, and the probability of evoked release and spontaneous release9. Many factors and proteins involved in regulating endocytosis, such as calcium, soluble NSF-attachment protein receptor (SNARE) proteins, brain-derived neurotrophic factor(BDNF), and calcineurin have been identified using pHluorin imaging1,2,10,11,12,13,14,15,16. Moreover, release of neurotransmitter could be detected in not only primary neurons but in neuroblastoma cells with TIRFM17. Recently, pHluorin variants, dsRed, mOrange and pHTomato were developed for monitoring simultaneous recordings of multiple factors in a single synapse18,19. For example, pHTomato has been fused with synaptophysin and used with a genetically encoded calcium indicator (GCaMP5K) to monitor presynaptic vesicle fusion and Ca2+ influx in the postsynaptic compartment20. Therefore, pHluorin attached to synaptic proteins provides a useful method to analyze the relationship between endocytosis and exocytosis.
EM is another method commonly used to study endocytosis, due to the high spatial resolution that shows ultrastructural changes during endocytosis. Two general areas are the ability to visualize pathological changes within neuronal cells21 and track vesicle proteins22. In particular, the observation of synaptic vesicle uptake, membrane curvature coated by clathrin in the periactive zone, and endosomal structures are possible with EM3,23,24,25,26,27,28. While EM involves potential artifacts, such as fixative-induced malformations, that may affect endocytosis, and data analysis is labor intensive, the resolution provides an attractive opportunity to visualize cellular structure. Potential fixative problems and the limitation in EM temporal resolution can be overcome by high pressure freezing, providing a fast and non-chemical method of stabilizing the delicate structures present during endocytosis27.
NOTE: The following protocol describes the pHluorin imaging methods and EM methods used in cultured hippocampal neurons. pHluorin monitors synaptic vesicle protein uptake in living cells and EM detects uptake of synaptic vesicle and ultrastructural changes.
Animal care and procedure followed NIH guidelines and were approved by the NIH Animal Care and Use Committee.
1. pHluorin Imaging
2. Electron Microscopy
Using the lipid carrier method, SpH was expressed in hippocampal neurons, allowing for the identification of boutons (Figure 1a). Electrical stimulation of the cells induced exocytosis, and a corresponding increase in fluorescence intensity. The increase in fluorescence (ΔF) was stopped by ending the stimulus (Figure 1b). The increased fluorescence was followed by a slow decrease, due to endocytosis. In the case of VAMP2-pHluorin, VAMP2 diffuses along the axon from a bouton after stimulus4. The raw data used arbitrary units, and was normalized to the baseline to obtain the rate of decay and τ (Figure 2a–c). Since the measurements were made from a normalized trace, the rate of decay reflects the initial decay of fluorescence in the percentage of the peak ΔF per second (ΔF/s). In our experimental conditions, endocytosis (usually longer than 10 s) was much slower than reacidification (~3-4 s)4,5. Thus, the fluorescence decay of pHluorin primarily reflects endocytosis4,5. In presynaptic boutons, ΔF is larger than the axon except bouton and the beginning time-of-increase was matched with initiation of electrical stimulus. In axon except bouton, ΔF was lower in comparison to regions with boutons and beginning time-of-increase was delayed compared to the initiation of electrical stimulus (Figure 2b). ΔF induced in bouton and axon except bouton was 82.9 ± 9.0% (n = 5 experiments) and 23.2 ± 4.0% (n = 5 experiments), respectively (Figure 2b). ΔF of SpH decayed mono-exponentially with a τ of 17.7 ± 0.3 (n = 5 experiments) and 20.7 ± 0.2 (n = 4 experiments) in 50 AP and 200 AP, respectively (Figure 1b). In 200 AP, τ was not significantly different before or after normalizing with the baseline. In the case of VAMP2-pHluorin transfected cells, the expression level of VAMP2-pHluorin was higher than SpH (Figure 3a–c). Traces of VAMP2-pHluorin were also obtained by normalizing to the baseline (Figure 3b).
EM was performed in cultured neurons and the clathrin coated vesicles were examined in comparison to control samples, which were incubated with 5 mg/mL of HRP for 90 s in absence of the stimulus (Figure 4a–b). The clathrin coated pits were observed at a probability of 0.05 per bouton in the absence of stimulus. Stimulation with high K+ induced bulk endosome and synaptic vesicles uptake, which were identified by HRP labeling (Figure 5a). Synaptic vesicles were defined as vesicles with a diameter less than 50 nm and bulk endosomes were defined as having a diameter over 80 nm or with a cross sectional area of more than an 80 nm vesicle. Induced bulk endosome uptake was decreased by recovery with normal saline (Figure 5b).
Figure 1: Images of SpH transfected neurons and traces of SpH transfected cell responses to electrical stimulation. (a) A sample image of cultured hippocampal neurons transfected with SpH by liposome delivery. SpH was highly expressed in boutons, presenting circle or oval patterns, with relatively lower expression along the axon. White boxes indicate ROI containing boutons (1.5 x 1.5 µm). Scale bar: 20 µm. (b) Responses of SpH transfected cells to an electrical stimulus of 50 AP at 20 Hz (n = 5 experiments) or 200 AP at 20 Hz (n = 4 experiments). Each experiment was based on 20-60 boutons. Please click here to view a larger version of this figure.
Figure 2: Rate of decay and tau value for endocytosis. (a) Responses to 50 AP at 20 Hz were converted from arbitrary units to normalized traces (n = 5). Fluorescence change was normalized to the baseline. This figure has been modified from Wu et al., 20163. (b) Normalized responses to 200 AP at 20 Hz in boutons (black, n = 4 experiments, left panel) and axon except bouton (red, n = 4 experiments, left panel). Fluorescence change was normalized to the baseline. Dashed box is magnified in the right panel. (c) Using normalized baseline, the rate of decay or τ was calculated in 50 AP (n = 5 experiments) and 200 AP (n = 4 experiments) at 20 Hz. Before calculation, ΔF was normalized to 100%. τ of endocytosis was obtained from normalized trace between the time of maximum fluorescence and the end of experiment. Rate of decay (mean + s.e.m., upper panel) and τ (mean + s.e.m., lower panel) induced by 50 AP and 200 AP. Please click here to view a larger version of this figure.
Figure 3: Images of VAMP2-pHluorin transfected neurons and examples of traces. (a) A sample image of cultured hippocampal neurons transfected with VAMP2-pHluorin by liposome delivery. White boxes indicate the ROI containing boutons (1.5 x 1.5 µm). Scale bar: 20 µm. (b) Traces for VAMP2-pHluorin responses to 200 AP at 20 Hz (n = 4 experiments). Fluorescence change was normalized to the baseline. (c) Fbase (a.u.) (mean + s.e.m.) in VAMP2-pHluorin (n = 4 experiments) and SpH (n = 7 experiments) transfected boutons. **, p <0.01. Please click here to view a larger version of this figure.
Figure 4: Images of EM in presynaptic terminal. (a) Synaptic vesicles were shown in cultured hippocampal neurons using EM. This figure has been modified from Wu et al., 20163. Scale bar = 200 nm. (b) Magnified image showing clathrin coated pits. Scale bar: 50 nm. Please click here to view a larger version of this figure.
Figure 5: Images of presynaptic terminal in stimulated neurons with high potassium. (a) EM images of neurons. The cells were fixed without stimulus (R), immediately after stimulation with 90 mM KCl for 90 s containing soluble HRP (K+), or after stimulation and then incubated for 10 min in normal saline. This figure has been modified from Wu et al., 20163. Scale bar = 200 nm. (b) Compared to control, HRP (-) synaptic vesicles were decreased and HRP (+) vesicles were increased by KCl. Bulk endosome uptake was induced by KCl and gradually decreased by recovery (mean + s.e.m., each group was from 40-100 synaptic profiles). *p <0.1; **p <0.01, significant to resting; ##p <0.01, significant to K+. Please click here to view a larger version of this figure.
Here we demonstrate two methods for monitoring synaptic vesicle endocytosis. In the first method, we monitored pHluorin fused with a synaptic vesicle protein in transfected neurons and subsequently electrically stimulated. Secondly, we used EM imaging of HRP uptake as induced by KCl. We used different stimuli for two reasons. First, high potassium application induces depolarization of all neurons in the culture. This facilitates EM examination, given that our EM methodology could not distinguish between non-stimulated and stimulated neurons. Second, ultra-structural morphological changes were more reliably observed after intense stimulation, such as the high potassium stimulation, whereas pHluorin fluorescence changes can be observed after a brief train of action potentials or even a single action potential.
Light microscopy imaging using pHluorin shows synaptic vesicle protein fusion and uptake with high temporal resolution in living cells. It can be used to study mechanisms that regulate the endocytosis time course and the size of synaptic vesicle pools. pHluorin imaging has not only been used for monitoring synaptic protein endocytosis, but also for monitoring evoked or spontaneous single vesicle release, short-term synaptic plasticity, and for the measurement of the release probability35. Using pHluorin has shown that the synaptic vesicular protein pools being released are not the same as those being retrieved36,37. Recently, pHluorin fused with synaptic proteins was used to detect the fate of newly fused vesicles38, and bulk endocytosis39. Therefore, pHluorin fused with a synaptic protein is a valuable tool for studying exocytosis and endocytosis in living synapses.
The limitation of this technique is that the decay of the fluorescence intensity reflects not only endosomes, but also vesicular re-acidification. These two components could potentially be separated by treatment with bafilomycin, an inhibitor of V-ATPase10,40. Moreover, the fluorescence increase during electrical stimulation reflects the net outcome of exocytosis and endocytosis. Although pHluorin imaging could not detect membrane invagination, this limitation can be overcome by other techniques, particularly EM.
In EM, strongly stimulated cells showed increased HRP-positive synaptic vesicles and bulk endosomes. These bulk endosomes gradually disappeared by then regenerating into synaptic vesicles24. This technique shows synaptic vesicle uptake by revealing the ultrastructure in the presynaptic terminal in both control and stimulated cells. A drawback to this technique is assessing synaptic vesicle uptake in fixed cells rather than living cells. Additionally, the possibility of a fixative induced artifact cannot be excluded. Known artifacts affecting the membrane include blebbing, membrane erosion, and protein changes41. These differences from the naturally occurring state of a cell are an inherent concern to all methods of visualizing cells, and particular attention was paid to the EM data presented to assess the possibility of such artifacts41,42,43. While this technique alone cannot show functional endocytosis of synaptic proteins, combining it with light microscopy, which can provide live images of the endocytosis process, creates a more complete view of this cellular mechanism.
The authors have nothing to disclose.
We thank Dr. Yong-Ling Zhu for providing synaptophysin-pHluorin2x construct, and Dr. James E. Rothman for providing VAMP2-phluorin. We thank Dr. Susan Cheng and Virginia Crocker of NINDS Electron Microscopy Facility for their technical support and help. This work was supported by the National Institute of Neurological Disorders and Stroke Intramural Research Program in USA and a grant from the KRIBB Research Initiative Program (Korean Biomedical Scientist Fellowship Program), Korea Research Institute of Bioscience and Biotechnology, Republic of Korea.
Lipofectamine LTX with Plus | Thermo Fisher | 15338-100 | Transfection of plasmid DNA including synaptophysin or VAMP2-pHluorin |
neurobasal medium | Thermo Fisher | 21103-049 | Growth medium for neuron, Warm up to 37°C before use |
B27 | Thermo Fisher | 17504-044 | Gradient for neuronal differentiation |
Glutamax | Thermo Fisher | 35050-061 | Gradient for neuronal culture |
Poly-D-Lysine coated coverslip | Neuvitro | GG-25-pdl | Substrate for neuronal growth and imaging of pHluorin |
Trypsin XI from bovine pancrease | Sigma | T1005 | Neuronal culture-digest hippocampal tissues |
Deoxyribonuclease I from bovine pancreas | Sigma | D5025 | Neuronal culture-inhibits viscous cell suspension |
pulse stimulator | A-M systems | model 2100 | Apply electrical stimulation |
Slotted bath with field stimulation | Warner Instruments | RC-21BRFS | Apply electrical stimulation |
stimulus isolation unit | Warner Instruments | SIU102 | Apply electrical stimulation |
lubricant | Dow corning | 111 | pHluorin imaging-seal with coverslip and imaging chamber, avoid leak from chamber |
AP5 | Tocris | 3693 | Gradient for normal saline, selective NMDA receptor antagonist, inhibit postsynaptic activity which have potential for recurrent activity |
CNQX | Tocris | 190 | Gradient for normal saline, competitive AMPA/kainate receptor antagonist, inhibit postsynaptic activity which have potential for recurrent activity |
Illuminator | Nikon | C-HGFI | Metal halide light source for pHluorin |
EMCCD camera | Andor | iXon3 | pHluorin imaging, detect pHluorin fluorescence intensity |
Inverted microscopy | Nikon | Ti-E | Imaging for synaptophysin or VAMP2 pHluorin transfected cells |
NIS-Elements AR | Nikon | NIS-Elements Advanced Research | Software for imaging acquisition and analysis |
Igor Pro | WaveMetrics | Igor pro | Software for imaging analysis and data presentation |
imaging chamber | Warner Instruments | RC21B | pHluorin imaging, apply field stimulation on living cells |
poly-l-lysine | Sigma | P4832 | Electron microscopy, substrate for neuronal growth, apply on multiwell plate for 1 h at room temperature then wash with sterilized water 3 times |
Horseradish peroxidase(HRP) | Sigma | P6782 | Electron microscopy, labeling of endocytosed synaptic vesicles by catalyzing DAB in presence hydrogen peroxide, final concentration is 5 mg/mL in normal saline, make fresh before use |
Na cacodylate | Electron Microscopy Sciences | 12300 | Electron microscopy, buffer for fixatives and washing, final concentration is 0.1 N |
3,3′-Diaminobenzidine(DAB) | Sigma | D8001 | Electron microscopy, labeling of endocytosed synaptic vesicles, substrate for HRP, final concentration is 0.5 mg/mL in DDW and filtered, make fresh before use |
Hydrogen peroxide solution | Sigma | H1009 | Electron microscopy, labeling of endocytosed synaptic vesicles by inducing HRP-DAB reaction, final concentration is 0.3% in DDW, make fresh before use |
glutaraldehyde | Electron Microscopy Sciences | 16365 | Electron microscopy, fixatives, final concentration is 4% in Na-cacodylate buffer, make fresh before use, shake well before to use |
TEM | JEOL | 200CX | Electron microscopy, imaging of endocytosed vesicles and ultrastructural changes |
CCD digital camera | AMT | XR-100 | Electron microscopy, capturing images |
Lead citrate | Leica microsystems | 16707235 | Electron microscopy, grid staining |