Teneurin-2 at the Synapse Construction Site is a Signpost for Cargo Unloading from Motor Proteins

In mature neurons, excitatory synapses are formed on the dendritic spine, whereas inhibitory synapses are formed on the dendritic shaft. Thus, it is primarily the accumulation of synaptic proteins that characterizes inhibitory synapses as distinct from non-synaptic regions. Protein accumulation is achieved by a combination of microtubule (MT)-based transport by kinesins and lateral diffusion across the plasma membrane; however, how and when proteins are released from kinesins remains unclear. Using primary cultured hippocampal neurons, we found that Teneurin-2 (TEN2) promotes synaptic protein accumulation by recruiting MTs via the representative MT plus end-tracking protein, EB1. MTs recruitment was enhanced when the extracellular domain of TEN2 successfully chose partners, and the lateral diffusion of TEN2 was inhibited. Conversely, if TEN2 partner choice is not achieved, MTs are not recruited, and thus synaptogenesis is not followed. Our study revealed that cargo release from kinesins through TEN2-MTs interactions supports the continuity from partner choice to synaptogenesis, which is a critical step in synaptic maturation.


29
Neurons connect to each other via synapses to generate dense neural circuits, and exchange 30 information via synaptic receptors to properly perform brain activities that underlie various 31 physiological functions. To this end, axons find their way to appropriate target regions (axon 32 guidance), select appropriate synaptic partners within those regions (synaptic specificity), and . Therefore, we hypothesized that TEN2 plays a more direct 158 causal role, although NLGN2 is closely associated with MAP2-rich synapses. In this study, we 159 focused on TEN2 to elucidate the mechanism by which MTs are recruited to inhibitory post-160 synapses.

162
Interaction with MTs via EB1 by two motifs in TEN2 163 The screening results suggest that TEN2 may bind to EB1. EB1 is localized at the MT plus end 164 at the endogenous expression level and is observed as the EB1 comet, but it detaches after the 165 polymerization of MTs is completed (Akhmanova & Steinmetz, 2015). Because the EB1 comet 166 can be seen transiently in living cells and its duration is approximately 10-20 seconds, the EB1 167 comet is not easy to detect in a fixed cell. However, when large amounts of EB1 are expressed, has an amino acid mutation in the EB1 binding motif, did not colocalize with EB1 ( Figures 2B   177 and 2D). Therefore, the intracellular domain of TEN2 is capable of interacting with MTs via EB1. 178 To confirm this interaction, we expressed TEN2N-L and TEN2TM in neurons and observed 179 their colocalization with endogenous EB1 in dendrites. As expected, colocalization of EB1 was To investigate whether the interaction between TEN2 and EB1 was related to the recruitment of 187 dynamic MTs, we first observed the arrangement of MTs. The fusion proteins TEN2N-L or 188 TEN2TM, with an HA tag added to the extracellular domain, were expressed in COS-7 cells.

189
Immunofluorescence staining of living cells with anti-HA tag antibodies confirmed that the EB1 190 binding motif was correctly located in the cytoplasm ( Figure 3B). Under these conditions, there 191 was no difference in MT arrangement between TEN2N-L and TEN2TM ( Figure 3C, lower panel).

192
TEN2 is endogenously bound to its binding partner at the synapse, immobilized, and restricted 193 from lateral diffusion. Therefore, to mimic the dynamics of endogenous TEN2 at synapses, we   These results suggest that TEN2 localizes and functions in the plasma membrane of dendrites.

236
On the other hand, presynaptic TEN2 has been reported to bind to latrophilin in the postsynaptic  However, it is impossible to detect whether TEN2 is located at the presynaptic or postsynaptic 243 membrane because of limitations in the resolution of the microscope. Therefore, we next 244 observed precise localization using one of super-resolution microscope (SRM), stochastic 245 optical reconstruction microscope (STORM). First, to determine whether TEN2 is present in the 246 presynapse or postsynapse at inhibitory synapses, we co-stained the cells with an antibody 247 against the intracellular domain of TEN2 (anti-ICD) and bassoon, a marker of presynapse, at results showed that the proximity between TEN2 and gephyrin was significantly greater than that 264 between normal IgG and gephyrin, which was used as the negative control ( Figures 4G and 4H).

265
This result supports our STORM data and suggests that the effect of signal misalignment 266 between channels owing to drift and other factors is minimal. Interestingly, the puncta of TEN2 267 and gephyrin were not always perfectly colocalization. Therefore, we measured the distance 268 between the centers of mass of the fluorescence intensities of each punctum and found that they  suggesting that postsynaptic TEN2 was involved in the maturation of inhibitory postsynapses.

292
Although we used gephyrin as a marker of inhibitory synapses throughout this study, we 293 examined how TEN2 knockdown affects the localization of GABAA receptors. There are 19 294 GABAA receptor subunits. Of these, the GABAA receptor localized at inhibitory synapses forms 295 a hetero-pentamer consisting of two α1-3 subunits, two β1-3 subunits, and one γ2 subunit, in 296 the order γ2-β-α-β-α, counterclockwise from the extracellular view. In TEN2 knockdown neurons,

297
The GABAA receptor subunits α1, α5, and γ2 were quantified. We found that only γ2 expression 298 was significantly reduced by TEN2 knockdown (Figures 5C and 5D). Since γ2 is present in all 299 synaptic GABAA receptors, a significant decrease in gephyrin would lead to a decrease in γ2.

300
However, α1 is a subunit present only in a specific subset of synaptic GABAA receptors, so TEN2 301 may not have a significant effect on synaptic maturation in this subset. α5 is a subunit of 302 extrasynaptic GABAA receptors and is not integrated into intrasynaptic GABAA receptors.

303
Therefore, it seems quite natural that TEN2, which functions via adhesion at inhibitory synapses,    Figure 6D).

319
To determine whether TEN2N-L interacts with neuronal MTs, we extracted membrane proteins 320 that did not interact with the cytoskeleton using saponins prior to methanol fixation. EGFP signals  We conclude that the interaction of TEN2 with MTs is critical for efficient transport of synaptic 348 components to develop inhibitory synapses. gephyrin were 85 nm apart. However, this is based on 2D observations; therefore, the actual 356 distance may be slightly longer, roughly estimated to be ~120 nm. Considering that the width of 357 postsynapses is generally 500 nm to 1 µm, we determined that they are present in the semi-358 periphery. This distance would be adequate because if it were located outside the postsynaptic 359 area, it would lose its connection to its presynaptic counterpart, resulting in a reduced MTs 360 trapping function. Indeed, dynein, a retrograde motor protein, has also been observed in this 361 region by EM, which supports our results (Fuhrmann et al., 2002).  At the same time, the origin of retrograde transport is also increased; therefore, our results can 407 be interpreted as an apparent absence of protein accumulation. Thus, one of our future tasks is 408 to elucidate whether adhesion molecules regulate the balance between anterograde and 409 retrograde transport.

410
Although the TEN2-MTs interaction has been confirmed to promote synaptogenesis, other 411 candidate molecules still need to be investigated. It has been suggested that NLGN2 also 412 localizes to the MAP2-rich synapses. We excluded NLGN2 because it is abundant in the center    These cells were kept at 37°C in a humidified atmosphere of 95% air and 5% CO2.   Using the following procedure, the molecules that do not interact with the cytoskeleton were 502 removed. First, a sufficient amount of saponin (Kanto Chemical) was dissolved in water to make 503 a saturated solution. Then, the saturated solution of saponin was added to the culture medium 504 to reach a final concentration of 0.1% and permeabilized in the incubator for 3 minutes. After 505 permeabilization, the cells were immediately immersed in methanol at -30°C and fixed on ice.

506
After that, additional permeabilization with Triton X-100 was performed. The rest of the procedure 507 was the same as for normal immunofluorescence staining.
(F) Time-lapse images of EB1 comet. Cells expressing TEN2N-L immobilized by antibodies show slower (blue arrows) or almost immobile (blue arrowheads) EB1 comets, compared to those (red arrows) in cells expressing TEN2TM without immobilization. Scale bar, 2 μm.

C A G G A T C C T A T C C T T A T G A C G T T C C A G A T T A C G C T G T T T A A C A A A A T A A C C T G C T G C 290
300 310 320 330 (B) Overview of the teneurin-2 full-length protein and antibody recognition sites. Teneurin is a type II transmembrane protein that is intracellular at its N-terminus and extracellular at its Cterminus. 3xHA, inserted just before the stop codon, is extracellular when translated.
(C) Typical genotyping results. A 300 bp band is seen in wild-type mice, while a +100 bp band is seen in knock-in mice; if both bands are seen, the mouse is heterozygous with only one allele being knock-in.
(D) Sequence confirmation by Sanger sequencing. Bands amplified by genotyping were purified and Sanger sequenced to confirm the knock-in sequence.
(E) No effect of HA knock-in on localization to inhibitory synapses. ICR-delivered wild-type neurons at DIV15 were co-stained with ICD and gephyrin antibodies, and HA knock-in neurons at DIV15 were co-stained with HA and gephyrin antibodies. mean ± SD were 0.30 ± 0.01 and 0.34 ± 0.02, respectively. Since there was no significant difference in the ratio of colocalization (p = 0.40), we concluded that HA knock-in had no effect on localization to inhibitory synapses.
(F) Images of immunofluorescence staining of HA tag and actin exposed on the cell membrane surface in the knock-in neuron. The red dashed box is magnified in (G).
(G) Confirmation that HA tag is exposed at the plasma membrane surface, suggesting that TEN2 functions at the plasma membrane surface.
(H) Images of immunofluorescence staining of HA tag and actin in the knock-in neuron. The red dashed box is magnified in (I).
(I) HA tag signals are also present in the dendritic shaft but are particularly strong near spinelike structures, suggesting that the molecule is more abundant at excitatory synapses. Arrows, representative colocalization.                 Table S6 The excitatory and inhibitory common synaptic cleft proteins with LxxPTPφ motif related to Figure 1