The 14-3-3γ isoform binds to and regulates the localization of endoplasmic reticulum (ER) membrane protein TMCC3 for the reticular network of the ER

The reticular network of the endoplasmic reticulum (ER) is formed by connecting ER tubules through three-way junctions and undergoes constant remodeling through formation and loss of the three-way junctions. Transmembrane and coiled-coil domain family 3 (TMCC3), an ER membrane protein localizing at three-way junctions, has been shown to positively regulate formation of the reticular ER network. However, elements that negatively regulate TMCC3 localization have not been characterized. In this study, we report that 14-3-3γ, a phospho-serine/phospho-threonine-binding protein involved in various signal transduction pathways, is a negative regulator of TMCC3. We demonstrate that overexpression of 14-3-3γ reduced localization of TMCC3 to three-way junctions and decreased the number of three-way junctions. TMCC3 bound to 14-3-3γ through the N terminus and had deduced 14-3-3 binding motifs. Additionally, we determined that a TMCC3 mutant substituting alanine for serine to be phosphorylated in the binding motif reduced binding to 14-3-3γ. The TMCC3 mutant was more prone than wildtype TMCC3 to localize at three-way junctions in the cells overexpressing 14-3-3γ. Furthermore, the TMCC3 mutant rescued the ER sheet expansion caused by TMCC3 knockdown less than wild-type TMCC3. Taken together, these results indicate that 14-3-3γ binding negatively regulates localization of TMCC3 to the three-way junctions for the proper reticular ER network, implying that the negative regulation of TMCC3 by 14-3-3γ would underlie remodeling of the reticular network of the ER.


Abstract
The reticular network of the endoplasmic reticulum (ER) is formed by connecting ER tubules through three-way junctions, and undergoes constant remodeling through formation and loss of the three-way junctions. TMCC3, an ER membrane protein localizing at three-way junctions, has been shown to positively regulate formation of the reticular ER network. However, elements that negatively regulate TMCC3 localization have not been characterized. In this study, we report that 14-3-3γ, a phospho-serine/phospho-threonine-binding protein involved in various signal transduction pathways, is a negative regulator of TMCC3. We demonstrate that overexpression of 14-3-3γ reduced localization of TMCC3 to three-way junctions, and decreased the number of three-way junctions.
Additionally, we determined that a TMCC3 mutant substituting alanine for serine to be phosphorylated in the binding motif reduced binding to 14-3-3. The TMCC3 mutant was more prone than wild-type TMCC3 to localize at three-way junctions in the cells overexpressing 14-3-3γ. Furthermore, the TMCC3 mutant rescued the ER sheet expansion caused by TMCC3 knockdown less than wild-type TMCC3. Taken together, these results indicate that 14-3-3γ binding negatively regulates localization of TMCC3 to the three-way junctions for the proper reticular ER network, implying that the negative regulation of TMCC3 by 14-3-3γ would underlie remodeling of the reticular network of the ER.
While the earlier study has demonstrated that 14-3-3 proteins bound to TMCC3 (46), there is no study characterizing the interaction between 14-3-3 proteins and TMCC3 in the context of regulation of the ER morphology. In this study, we set out to examine the effect of 14-3-3 proteins on localization of TMCC3 to three-way junctions.
While human 14-3-3 family is composed of seven isoforms, the earlier study did not clarify whether TMCC3 bound to 14-3-3 proteins in isoform-dependent manner or not (46). On the other hand, BioGRID (56), the database for protein and genetic interactions, shows that 14-3-3γ and 14-3-3θ are identified as TMCC3-binding proteins. We, therefore, It is also noted that the PDI staining pattern in the peripheral ER seemed to be affected by overexpression of HA-14-3-3γ to some extent, suggesting that overexpression of 14-3-3γ affected the peripheral ER morphology. These results J o u r n a l P r e -p r o o f indicate that overexpression of 14-3-3γ reduced localization of TMCC3 to three-way junctions.
We next examined whether overexpression of 14-3-3γ affected the protein level of TMCC3.
HA-14-3-3γ or the control vector was transfected into U2OS cells. The total cell lystaes were subjected to immunoblotting with the anti-TMCC3 pAb, the anti-HA mAb, and the anti-GAPDH.
Collectively, these results suggest that

14-3-3 binding changes localization of TMCC3
without affecting the protein level.

Overexpression of 14-3-3γ decreases the number of three-way junctions.
Since TMCC3 is involved in formation of the reticular ER network (40), we reasoned that, if 14-3-3γ negatively regulated TMCC3 localization as shown in Figure 1B, overexpression of 14-3-3γ would affect the ER morphology. Indeed, we observed that overexpression of HA-14-3-3γ affected the peripheral ER morphology to some extent as mentioned above. To further assess this, HA-14-3-3γ or mCherry was transfected into Overall, these results indicate that overexpression of 14-3-3γ decreases the number of peripheral three-way junctions, and alters the peripheral ER morphology to some extent.

N-terminus.
We sought to determine which region of TMCC3 was responsible for binding to 14-3-3γ. Phosphorylated serine 15 is required for potent binding to 14-3-3γ.
We examined whether serine 15 was phosphorylated.
Collectively, these results indicate that phosphorylated serine 15 in the deduced 14-3-3 binding motif is required for potent binding to 14-3-3γ.
The TMCC3 mutant substituting alanine for serine 15 is prone to localize at three-way junctions against overexpression of 14-3-3γ.

HA
Collectively, these results indicate that 14-3-3γ negatively regulates localization of TMCC3 to the three-way junctions through binding to phosphoserine 15.

The negative regulation of TMCC3 by 14-3-3γ
is involved in formation of the reticular ER network.
We finally sought to assess the physiological importance of the negative regulation of TMCC3 by 14-3-3γ. To this end, we examined whether TMCC3-S15A could rescue the phenotype of Knockdown of endogenous TMCC3 was confirmed by immunoblotting ( Figure S5). The

Discussion
In this study, we demonstrate that 14-3-3γ binds to Overexpression of 14-3-3γ alters the peripheral ER morphology as shown in Figure 2 and Figure S2.  (TMCC3-S15D and TMCC3-S15E). However, TMCC3-S15D and TMCC3-S15E did not act as phosphomimetic mutants for 14-3-3 binding, because they did not enhance binding to 14-3-3γ and still localized at three-way junctions (data not shown). We do not think that these suggest inconsistency in our results, because the previous study of 14-3-3 binding sites states that glutamate and aspartate do not provide good phosphomimetic residues with respect to 14-3-3 binding to target proteins (66). Further studies will be required to address these concerns.
In summary, we demonstrate that 14-3-3γ binding negatively regulates localization of TMCC3 for the reticular ER network. for two days to allow endogenous TMCC3 to be depleted.

Quantification of the ER morphology
As for quantification of three-way junctions, Data availability: All data were contained within the article.

Supporting information
This article contains supporting information.       The siRNA targeting TMCC3 was transfected into U2OS cells to allow endogenous TMCC3 to be knocked down. HA-TMCC3, HA-TMCC3-S15A, or the control vector was then transfected to the TMCC3-knockdown cells. As a negative control, the control siRNA was transfected into U2OS cells,    (B) Localization of GFP-TMCC3-S15A at three-way junctions. The U2OS-GFP-TMCC3-S15A cells J o u r n a l P r e -p r o o f were immunostained with the anti-PDI mAb. GFP-TMCC3-S15A was detected by its fluorescence. Scale bar, 20 µm. The boxed areas are enlarged to highlight the peripheral ER and shown below each image.

Figure S6. Additional potential binding sites for 14-3-3 in TMCC3-D1
The sequence shown here is the N-terminal 210 aa of mouse TMCC3 (TMCC3-D1). Serine 166 and threonine 176, additional potential binding sites for 14-3-3 proteins which are predicted by 14-3-3 Pred with much lower scores than serine 15, are shown in green. Serine 15 that we characterized in this study is shown in red. The first coiled-coil domain is underlined. HA-TMCC3-D1 or HA-TMCC3-D1-S15A was transfected into HEK293 cells. The day after transfection, the culture medium was replaced with fresh one supplemented with 100 ng/mL nocodazole or DMSO as a negative control, and the cells were further cultured for 17 h. The cells were subjected to immunoprecipitation with the anti-HA mAb, followed by immunoblotting with the anti-phospho-14-3-3 binding motif pAb and the anti-HA mAb. The anti-phospho-14-3-3 binding motif pAb detected HA-TMCC3-D1 but did not recognize HA-TMCC3-D1-S15A, while the anti-HA mAb detected HA-TMCC3-D1 and HA-TMCC3-D1-S15A. Importantly, the immunoreactive band of the phosphorylated HA-TMCC3-D1 in the presence of DMSO was comparable to that in the presence of nocodazole, suggesting that mitosis will not affect the phosphorylation status of TMCC3.
J o u r n a l P r e -p r o o f