Activity of the SNARE Protein SNAP29 at the Endoplasmic Reticulum and Golgi Apparatus

Abstract

Snap29 is a conserved regulator of membrane fusion essential to complete autophagy and to support other cellular processes, including cell division. In humans, inactivating SNAP29 mutations causes CEDNIK syndrome, a rare multi-systemic disorder characterized by congenital neuro-cutaneous alterations. The fibroblasts of CEDNIK patients show alterations of the Golgi apparatus (GA). However, whether and how Snap29 acts at the GA is unclear. Here we investigate SNAP29 function at the GA and endoplasmic reticulum (ER). As part of the elongated structures in proximity to these membrane compartments, a pool of SNAP29 forms a complex with Syntaxin18, or with Syntaxin5, which we find is required to engage SEC22B-loaded vesicles. Consistent with this, in HeLa cells, in neuroepithelial stem cells, and in vivo, decreased SNAP29 activity alters GA architecture and reduces ER to GA trafficking. Our data reveal a new regulatory function of Snap29 in promoting secretory trafficking.

Keywords: Golgi apparatus; SEC22B; SNAP29 gene; SNARE protein; Syntaxin 5; endoplasmic reticulum; vesicle fusion.

FIGURE 1

SNAP29 is required to support correct Golgi apparatus (GA) architecture. (A) Immunoblotting of total proteins from mock and SNAP29-depleted (KD) HeLa cell protein with the indicated antibodies. The asterisk indicates the presence of a non-specific band recognized by anti-SNAP29. Snap29 is efficiently depleted upon KD. (B) Max Projections of mock and SNAP29 KD HeLa cells, stained as indicated. (B’) Measurement of the Golgin97-positive area of cells as in panel (B). The area of each measured GA is shown on the x-axis, while the ratio of the length of the major axis over the minor axis of the GA is shown on the y-axis. Depleted cells display a larger and rounder GA. (C–F) Max Projections of mock (C,E) and SNAP29 KD HeLa cells (D,F) or, in addition, over-expressing the indicated transgene (E,F), stained as indicated. (G) Quantification of the number of Giantin-positive objects. The mean with standard error of the mean is shown, and the p-value is obtained by one-way ANOVA with Tukey’s multiple-comparisons analysis. GA alterations upon SNAP29 depletion are rescued expression of GFP–SNAP29, which per se does not alter GA architecture. (H–J) Electron microscopy sections of mock (H) and SNAP29-depleted HeLa cells (I,J). 3D tomographic reconstruction of encompassing sections is shown below. In addition to the indicated pseudo-coloring, COPI-coated vesicles are in white, and COPII-coated buds and tubules are in light brown. Clathrin-dependent vesicles are in light blue. SNAP29 depletion leads to GA vesiculation, endoplasmic reticulum (ER)–Golgi intermediate compartment tabulation, and ER enlargement.

FIGURE 2

GFP–SNAP29 partially colocalizes with endoplasmic reticulum (ER) and Golgi apparatus (GA) markers. (A–I) Single sections of HeLa cells over-expressing GFP–SNAP29 for 6 h stained as indicated and acquired by stimulated emission depletion microscopy. The dashed and the continuous lines delimit the nucleus and the plasma membrane, respectively. The boxed GA area is magnified in the insets. Yellow arrows point to an example of co-localization between GFP–SNAP29 and ER or GA markers. (J,K) Cryo-immuno-EM sections of HeLa cells stably transfected to express GFP–SNAP29, stained, and revealed as indicated. Some GFP–SNAP29 localize to the ERGIC area and colocalizes with a COPI marker.

FIGURE 3

SNAP29 contributes to ManII–SBP–GFP trafficking to the Golgi apparatus (GA). (A–E) Single confocal sections of HeLa cells stably expressing ManII–SBP–GFP, treated and stained as indicated. The EGFP pattern has been imaged before addition of biotin (no biotin) and 20 min after addition of biotin (20 min biotin). The insets show close-ups of the GA and surrounding areas. (F) Quantification of the ratio of the Giantin-positive EGFP signal over total, relative to the experiment in panel (A–E). SNAP29, as well as the endoplasmic reticulum and GA SNAREs STX18, STX5, and SEC22B, appears to support ManII–SBP–GFP trafficking to the GA. The median with interquartile range is shown, and the p-value is obtained by Dunn’s multiple-comparisons test.

FIGURE 4

SNAP29 interacts with the endoplasmic reticulum (ER) and the Golgi apparatus (GA) SNAREs. (A) Immunoblotting of proteins immunoprecipitated from HeLa protein extracts with the indicated antibodies and related inputs. Endogenous SNAP29 interacts with the ER and GA SNAREs STX18, STX5, and SEC22B. (B) Immunoblotting of proteins immunoprecipitated using GFP Trap from protein extracts of HeLa cells expressing GFP–SNAP29 or GFP as a control and related inputs and supernatants. GFP–SNAP29 interacts with the ER and GA SNAREs. (C) Immunoblotting of protein extracts from Drosophila S2 cells over-expressing the indicated transgenes immunoprecipitated with the indicated antibody and related controls. Endogenous Drosophila Snap29 interacts with HA-Syx18 and HA-Sec22. (D) Drosophila S2 cells over-expressing the indicated transgenes and stained as indicated. Endogenous Drosophila Snap29 colocalizes with HA-Syx18 and HA-Sec22.

FIGURE 5

SNAP29 interacts primarily with STX18 and is required to stabilize interactions with SEC22B. (A) HeLa cells expressing GFP–SNAP29^Q1Q2. Cells stained with anti-Golgin97 to mark the Golgi apparatus (GA) show that GFP–SNAP29^Q1Q2 forms enlarged bodies at the cell periphery and that the GA is fragmented. (A’) Quantification of Golgin97-positive GA object upon expression of the indicated transgenes reveals that GFP–SNAP29^Q1Q2 induces GA fragmentation, thereby acting as a dominant negative SNAP29 form. The mean with standard error of the mean is shown, and the p-value is obtained by one-way ANOVA with Tukey’s multiple-comparisons analysis. (B) Representative images of a CLEM analysis of a HeLa cell expressing GFP–SNAP29^Q1Q2. Single sections of HeLa cells expressing GFP–SNAP29^Q1Q2 collected at phase contrast (bright field) and by confocal microscopy (GFP) to visualize the cell morphology and GFP–positive bodies. (B’) Electron microscopy image of the cell indicated by the arrow in panel (B). The GFP–SNAP29^Q1Q2 bodies are composed of vesicular material and fragmented GA cisternae as highlighted in a close-up of the cytoplasmic portion boxed in panel (B’). (C,D) Single sections of HeLa cells over-expressing the indicated SNAP29 forms stained as indicated and acquired by stimulated emission depletion microscopy. The dashed and the continuous lines delimit the nucleus and the plasma membrane, respectively. High magnifications of the boxed areas are shown in the insets. The GFP–SNAP29^Q1Q2 bodies are highly decorated with N-ethylmaleimide-sensitive fusion. (E) Immunoblotting (IB) with the indicated antibodies of proteins immunoprecipitated using GFP Trap from protein extracts of HeLa cells expressing the indicated transgenes and related control. Interactions with Qb, Qc, and R-SNAREs, but not with Qa-SNARE STX18, are weakened by the expression of GFP–SNAP29^Q1Q2. (F,G) IB with the indicated antibodies of protein extracts from HeLa cells over-expressing the indicated transgenes (F) or treated as indicated (G) and immunoprecipitated with anti-SEC22B and related controls. The asterisk indicates a non-specific band recognized by anti-SNAP29. GFP–SNAP29^Q1Q2 is not included in SEC22B immunoprecipitations, and SNAP29 depletion impairs the interaction of SEC22B with STX18.

FIGURE 6

SNAP29 depletion in neuroepithelial stem (NES) cells causes Golgi apparatus (GA) and spindle alteration and formation of micronuclei. (A) Maximal confocal projections of NES cells treated and stained as indicated. Depleted NES cells display GA fragmentation. (B) Quantification of the number of Giantin-positive objects. The median with interquartile range is shown, and the p-value is obtained by Mann–Whitney test. (C) Maximal confocal projections of NES cells treated and stained to detect α-tubulin and p-Histone3. The depleted NES cells in metaphase show an altered mitotic spindle. The arrows indicate spindle poles. (D) Maximal confocal projections of NES cells treated and stained as indicated. The depleted NES cells possess several micronuclei (arrows). (E) Quantification of the percentage of cells with at least one micronucleus. The median with interquartile range is shown, and the p-value is obtained by Mann–Whitney test.

FIGURE 7

A model of SNAP29 activity at the endoplasmic reticulum (ER) and Golgi apparatus (GA). SNAP29 forms elongated structures that could assist the tethering of vesicles and/or that could regulate STX5/18 fusion competence. Some of SNAP29 are re-localized from the ER–GA area to form the outer kinetochore of mitotic chromosomes in prophase.