Terminals of the major thalamic input to visual cortex are devoid of synapsin proteins

Abstract

Synapsins are nerve-terminal proteins that are linked to synaptic transmission and key factors in several forms of synaptic plasticity. While synapsins are generally assumed to be ubiquitous in synaptic terminals, whether they are excluded from certain types of terminals is of interest. In the visual pathway, synapsins are lacking in photoreceptor and bipolar cell terminals as well as in retinogeniculate synapses. These are the terminals of the first three feedforward synapses in the visual pathway, implying that lack of synapsins may be a common property of terminals that provide the primary driver activity onto their postsynaptic neurons. To further investigate this idea, we studied the fourth driver synapse, thalamocortical synapses in visual cortex, using glutamatergic terminal antibody markers anti-VGluT1 and VGluT2, anti-Synapsin I and II, and confocal microscopy to analyze co-localization of these proteins in terminals. We also used pre-embedding immunocytochemical labeling followed by electron microscopy to investigate morphological similarities or differences between terminals containing synapsins or VGluT2. In visual cortex, synapsin coincided extensively with non-TC-neuron marker, VGluT1, while thalamocortical terminal marker VGluT2 and synapsin overlap was sparse. Morphologically, synapsin-stained terminals were smaller than non-stained, while VGluT2-positive thalamocortical terminals constituted the largest terminals in cortex. The size discrepancy between synapsin- and VGluT2-positive terminals, together with the complementary staining patterns, indicates that thalamocortical synapses are devoid of synapsins, and support the hypothesis that afferent sensory information is consistently transmitted without the involvement of synapsins. Furthermore, VGluT2 and synapsins were colocalized in other brain structures, suggesting that lack of synapsins is not a property of VGluT2-containing terminals, but a property of primary driver terminals in the visual system.

Fig. 1

SynI and SynII staining patterns in visual cortex are interdigitated with that of VGluT2. A, VGluT2 staining using a Cy5 coupled secondary antibody. A strong band of staining appears in Layer 4. B, Image of SynI staining using a Cy2 coupled secondary antibody. A band of weaker staining corresponds to Layer 4. C, Merged images with both SynI and VGluT2 labeling. D,E, VGluT2 staining as in (A), but staining with VGluT2 (Cy2 secondary) and SynII (Cy3 secondary) on a slice fixed with GA-containing fixative. Similar to SynI, staining intensity is the weakest in Layer 4. F, Merged images with both SynII, and VGluT2 labeling. Scale bar = 100μm (applies to all panels).

Fig. 2

In primary visual cortex (V1) Layers 4 and 6, VGluT1 (red puncta) labeled fibers also contain SynI and SynII (green puncta). Overlapping pixels appear yellow. A, VGluT1-SynI colocalization in Layer 4 (L4), area V1. Putative Layer 4 (L4) / Layer 3 (L3) border indicated by white arrowhead. B, VGluT1- SynII colocalization in Layer 4. C, VGluT1 and SynI colocalization in Layer 6. Arrowhead marks putative Layer 5 (L5) and Layer 6 (L6) border. D, VGluT1 and SynII coloalization in L6. Scale bar =10 μm (applies to all panels). E, Pearson correlation coefficient (PC) analysis between VGluT1 and synapsin containing pixels in subsequent scans with each fluorophore (avg ± SD). PC values between -1 and 1 indicate the strength of colocalization; an index value of 1 can only be attained between two identical scans. A value of 0 indicates two randomly distributed sets of pixels, that is, chance level of pixel overlap between two scans. A PC value of -1 indicates that the pixels obtained in two scans are systematically non-overlapping. Values between -0.5 and 0.5 occur as an outcome of strong background and subthreshold fluorescence, thus inconclusive for colocalization.

Fig. 3

In primary visual cortex (V1) Layer 4, VGluT2 (red puncta) labeled fibers do not contain SynI or SynII (green puncta). A, Lack of colocalization between VGluT2 and SynI positive puncta in Layer 4. B, Lack of colocalization between VGluT2 and SynII positive puncta in Layer 4. C, Lack of colocalization between VGluT2 and SynI positive puncta in Layer 6. Scale bar = 10 μm (applies to all panels). D, In Layer 6, some VGluT2 positive puncta also contained SynII (yellow puncta). E, Cross-correlation function analysis between VGluT2 and synapsin containing pixels in subsequent scans with each fluorophore, averaged from 18 sections (±SD).

Fig. 4

Control experiments. A,B, Transgenic mouse lacking SynI or SynII proteins are used as negative controls. Genetic deletion of SynI or SynII proteins eliminated SynI (A) and SynII (B) labeling in cortical fibers, while VGluT2 stain was unaffected. C,D, Colocalization of synapsin and VGluT2 occurs in other brain areas. C, In hippocampal subiculum, VGluT2 and SynI were colocalized in a narrow band close to the medial border (Hc sub). PC analysis yielded high positive correlation within a band that was marked within two yellow stippled lines. Scale bar =10 μm (applies to all panels). D, VGluT2 and SynII were colocalized in a region apical to the pyramidal cells of hippocampal lucidum (Hc luc).

Fig. 5

The size of synapsin labeled terminals do not match the size of thalamocortical terminals. A, Electron micrograph from primary visual cortex Layer 4. DAB-labeled, SynI positive terminals appear grey, apposed to a dendrite (d), mitochondria (m), or spine (s). Synaptic specializations indicated by black arrowheads. B, Similar section treated with SynII antibodies. An unstained, larger terminal (t-) forms an asymmetric synapse onto a spine. C, A large, VGluT2 positive terminal in primary visual cortex, forms synapses onto both a dendrite (d) and a spine (s). The terminal contains an unlabeled inclusion (*), a characteristic of TC terminals. Such inclusions are not encountered in synapsin-positive terminals. Scale bar = 250 nm. D, Terminal cross-section area distributions of terminals labeled for SynI, SynII or VGluT2. SynI and SynII terminals are similar in size, whereas VGluT2 positive terminals are significantly larger. While non-parametric comparisons were used for analysis, descriptive statistics for each population are indicated in the inset (mean ±SE). E, Histogram comparing cross-section areas of SynII positive terminals and unlabeled terminals that are found in the same regions as SynII terminals. Terminals that were labeled for SynII were significantly smaller than neighboring synapsin-unlabeled terminals.