Loss of synaptotagmin IV results in a reduction in synaptic vesicles and a distortion of the Golgi structure in cultured hippocampal neurons

Abstract

Fusion of synaptic vesicles with the plasma membrane is mediated by the SNARE (soluble NSF attachment receptor) proteins and is regulated by synaptotagmin (syt). There are at least 17 syt isoforms that have the potential to act as modulators of membrane fusion events. Synaptotagmin IV (syt IV) is particularly interesting; it is an immediate early gene that is regulated by seizures and certain classes of drugs, and, in humans, syt IV maps to a region of chromosome 18 associated with schizophrenia and bipolar disease. Syt IV has recently been found to localize to dense core vesicles in hippocampal neurons, where it regulates neurotrophin release. Here we have examined the ultrastructure of cultured hippocampal neurons from wild-type and syt IV -/- mice using electron tomography. Perhaps surprisingly, we observed a potential synaptic vesicle transport defect in syt IV -/- neurons, with the accumulation of large numbers of small clear vesicles (putative axonal transport vesicles) near the trans-Golgi network. We also found an interaction between syt IV and KIF1A, a kinesin known to be involved in vesicle trafficking to the synapse. Finally, we found that syt IV -/- synapses exhibited reduced numbers of synaptic vesicles and a twofold reduction in the proportion of docked vesicles compared to wild-type. The proportion of docked vesicles in syt IV -/- boutons was further reduced, 5-fold, following depolarization.

Fig. 1

Syt IV −/− synapses have a decreased number of vesicles prior to stimulation and a decreased number of docked vesicles following stimulation, compared to WT synapses. All synapse images are from a 10 nm slice in the Z-plane of a tomographic reconstruction from 250 nm thick sections of cultured hippocampal neurons (WT or syt IV −/−). Scale bar in (i), (ii), and (iv) is 60 nm. Scale bar in (iii) is 70 nm. (a) (i) (L–R) WT hippocampal synapse (vesicle number is representative of the wild-type mean (504 vesicles)). Enlarged area to the left shows an example of a docked vesicle. Surface rendering of the full three-dimensional tomographic reconstruction generated using the IMOD software program, where blue indicates presynaptic membrane; yellow, postsynaptic membrane; red, mitochondria; and green, synaptic vesicles. (ii) (L–R) WT hippocampal synapse, stimulated with KCl for 90 s prior to processing for EM (vesicle number is representative of the WT mean (415 vesicles)). Enlarged area to the left shows an example of a docked vesicle. Surface rendering of the full three-dimensional tomographic reconstruction. (iii) (L–R) Syt IV −/− hippocampal synapse (vesicle number is representative of the knockout mean (177 vesicles)). Surface rendering of the full three-dimensional tomographic reconstruction. (iv) (L–R) Syt IV −/− hippocampal synapse, stimulated with KCl for 90 s prior to processing for EM (vesicle number is representative of the knockout mean (102 vesicles)). Surface rendering of the full three-dimensional tomographic reconstruction. (b) (L–R, top to bottom) Comparison of the mean number of vesicles per synaptic terminal, for WT versus syt IV −/−, under resting and stimulated (90 mM KCl, 90 s) conditions. WT mean=504 (resting) and 415 (stimulated) vesicles/terminal. Syt IV −/− mean=177 (resting) and 102 (stimulated) vesicles/terminal. Comparison of the mean number of docked vesicles per synaptic terminal, for WT versus syt IV −/−, under resting and stimulated conditions. WT mean=15 (resting) and 11 (stimulated) vesicles/terminal, syt IV −/− mean=8 (resting) and 2 (stimulated) vesicles/terminal. Comparison of the mean density of vesicles (vesicles/μm² of synaptic membrane). WT mean=402 (resting) and 359 (stimulated) vesicles/μm². Syt IV −/− mean=172 (resting) and 92 (stimulated) vesicles/μm². Comparison of the mean density of docked vesicles (docked vesicles/μm² of synaptic membrane). WT mean=13 (resting) and 9.6 (stimulated) vesicles/μm². Syt IV −/− mean=7.9 (resting) and 1.8 (stimulated) vesicles/μm². Error bars indicate SEM. (c) Comparison of the frequency distribution of the number of docked vesicles between WT (gray bars) and syt IV −/− (white bars) synapses under resting and stimulated conditions. (Number of samples imaged are as follows: WT unstimulated (20 boutons from 10 separate cultures), WT stimulated (17 boutons from eight separate cultures), syt IV −/− unstimulated (18 boutons from eight separate cultures), syt IV −/− stimulated (15 boutons from eight separate cultures; error bars indicate SEM; statistical significance was determined by a Student's t-test where * P<0.05, ** P<0.01, and *** P<0.001).

Fig. 2

Syt IV −/− mice show an enlarged Golgi apparatus. (a) Cultured hippocampal neurons (10 DIV) were fixed and labeled with an antibody directed against the Golgi marker GM130 (green) and the synaptic vesicle protein synaptophysin (red), then imaged using laser confocal microscopy (L: wild-type. R: Syt IV −/−). Bottom panel shows only the green channel highlighting the Golgi. Inset shows a single Golgi from WT or syt IV −/− neurons. (b) Average area occupied by WT and syt IV −/− Golgi, calculated from five 250 nm tomograms (top panel; wild-type mean=7.4×10⁷ nm³, syt IV −/− mean=10.4×10⁷ nm³), or from 22 confocal image reconstructions from each genotype (bottom panel; WT mean=81.56 μm³, syt IV −/− mean=116.75 μm³). Error bars indicate SEM. Statistical significance was determined by a Student's t-test where * P<0.05, ** P<0.01, and *** P<0.001. (c) Cultured hippocampal neurons from syt IV −/− and WT littermates immunostained for VGluT1 and VGAT to mark excitatory and inhibitory synaptic vesicles, respectively. (d) Quantitation of intensity of VGluT1 or VGAT signal at synapses in syt IV −/− and WT cultures. VGluT and VGAT channels were thresholded separately to include all recognizable puncta, and average intensity was determined using Metamorph software (n=9 images from three different cultures; error bars indicate SEM; statistical significance was determined by a Student's t-test where * P<0.05, ** P<0.01, and *** P<0.001; for VGAT P=.012, and for VGluT P=.026). (e) Cultured hippocampal neurons (10 DIV) were high-pressure frozen and processed for electron tomography. A slice through a tomographic reconstruction of a WT Golgi (top), and the subsequent model (bottom). Scale bar is 40 nm. (f) A slice through a tomographic reconstruction of a syt IV −/− Golgi (top) and subsequent model (bottom). Scale bars=35 nm.

Fig. 3

Syt IV −/− neurons have a large vesicle pool in close proximity to the Golgi. (a) Cultured hippocampal neurons (10 DIV) were high-pressure frozen and processed for electron tomography. Electron micrograph of a 250 nm slice through syt IV −/− neurons approximately 1.0 μm from the TGN showing a large vesicle pool. (b) A model of the tomographic reconstruction of the boxed region in panel A shows a mix of clear (blue) and clathrin-coated (red) vesicles. (c) Slice through a tomographic reconstruction of the boxed area in panel a. Arrows indicate clathrin-coated vesicles and arrowheads indicate clear vesicles. (d) 2×2 array of individual vesicles either with a characteristic clathrin triskilion (arrows, top row) or without (bottom row). Scale bars=35 nm.

Fig. 4

Syt IV coimmunoprecipitates with KIF1A from mouse brain. Whole mouse brains were homogenized, solubilized and incubated with either syt IV, syt I, or KIF1A primary antibodies overnight. Samples were incubated with protein A Sepharose beads for 1 h, pelleted and then analyzed by SDS-PAGE and western blot. Syt IV coimmunoprecipitates KIF1A. Similarly, KIF1A coimmunoprecipitates syt IV. Syt I coimmunoprecipitates both KIF1A and syt IV, but syt I and syt IV did not co-immunoprecipitate each other to a significant extent.