Fig 1: Biochemical and Cell Biological Characterization of TMEM109 as hSnd3. (A) Venn diagram summarizing the numbers and overlaps of candidate proteins found after siRNA-mediated hSnd2 depletion (yellow circle), siRNA-mediated hSnd2 plus Wrb double depletion (purple circle), and proteins that were previously published as hSnd2 interaction partners using co-immunoprecipitation (turquoise circle) [28]. Protein names are listed in Table S16. The single candidate is TMEM109. Proteins from the LFQ MS analysis that showed a log2 fold value <−1 and a p-value < 0.01 for both siRNAs were considered as candidates. (B) Co-immunoprecipitation using the supernatant of CHAPS-solubilized HeLa lysates after centrifugation (Input). α-hSnd2 (left panels) or α-TMEM109 (right panels) antibodies were used for immunoprecipitation in combination with protein A/G sepharose. A sample of the supernatant after immunoprecipitation (IP) is shown as well as a sample of the first washing step (Wash) and the eluate after the acidic pH shift. Proteins of interest indicated on the right were detected by Western blotting. The original blots with molecular mass markers are shown in Figure S11. (C) Aplhafold structural models of TMEM109 and hSnd2. The N- and C-termini are indicated. The first 45 residues (including the 33-amino-acid-long signal peptide) were deleted from the N-terminus of the TMEM109 model (N*). Polar and charged residues of hSnd2 reminiscent of a hydrophilic vestibule are highlighted as stick models in red. The orange circles highlight the oligopeptides within the coiled-coil domain of TMEM109 (224EELRWRQRR232) and in the C-terminal domain of hSnd2 (156EHNEKRQRR164), respectively, that show sequence similarity. (D) Peptide spots that cover the primary structure of hSnd2 (top array) and TMEM109 (bottom array). The binding of solubilized TMEM109 or hSnd2 proteins from the RM lysate to particular spots was evaluated by antibody detection. Light orange or blue rectangles represent parts of the primary structure that encode a TMH or SP, respectively. Spots that are likely accessible in the native protein and repeatedly show a strong interaction (n = 3) are highlighted by dashed rectangles in red. (E) Bimolecular luminescence complementation was used to detect protein–protein interactions between TMEM109 and various partners based on relative luminescence units (RLU). TMEM109 was tagged at its C-terminus with the LargeBit component of the NanoLuc luciferase (TMEM109-CL). Putative interaction partners of interest were N- or C-terminally fused with the SmallBiT component of the NanoLuc luciferase (protein-NS or protein-CS). The heterologous Halo-CS fusion protein served as a non-interacting negative control for normalization and was set to 1 [71]. (F) The interaction between TMEM109-CL and hSnd2-NS (blue bars) as well as TMEM109-CL and TMEM109-CS (green bars) was challenged by complementation with an untagged interaction partner protein. In comparison to the empty vector (EV) control, plasmids that encode for an untagged variant of one of the interaction partners reduced the RLU. (G) Same as in (E), but Sec61α-NL served as a reporter component. (H) In vitro transport of 35S-Met-labeled TMEM109. The TMEM109 precursor (pTMEM109) was synthesized in the absence (No ER) or presence of semi-permeabilized cells pretreated with control siRNA (siCtrl) or HSND2 siRNA (siSND2) for 96 h before semi-permeabilization. The orange color highlights the oligopeptide 224EELRWRQRR232 within the coiled-coil domain of TMEM109 that shows sequence similarity to an oligopeptide in the C-terminal domain of hSnd2 (156EHNEKRQRR164). Statistical comparison of multiple conditions was performed using one-way ANOVA followed by Dunnett’s multiple comparison post-test (E,G). Statistical comparison of the competition experiment (F) was based on a Student’s t-test comparing the EV treatment versus the corresponding hSnd2 or TMEM109 competition. p-values are indicated by asterisks with p < 0.05 (*) < 0.01 (**) < 0.001 (***).
Fig 2: Functional Characterization of TMEM109 as hSnd3. (A,B) Western blot analysis of the indicated proteins upon treatment of HeLa cells with siRNAs targeting HSND2 (SND2 UTR) or TMEM109 (T109 UTR). An untargeted control siRNA was used for comparison. Cells were harvested 96 h (A) or 72 h (B) post-silencing. The original blots with molecular mass markers are shown in Figures S13 and S14. (C) Quantification of the relative protein abundance based on Western blot analysis upon silencing TMEM109 for 72 h. (D) In vitro transport of the 35S-Met-labeled model substrates preprolactin (ppl), invariant chain (invc), Cytochrome b5 (Cytb5), or Sec61β. Precursors are indicated by a ‘p’, and the imported, glycosylated proteins by a ‘g’. Upon transport and cleavage of the signal peptide, the indicated prolactin form is visible (pl). Precursors were synthesized in the absence (No ER) or presence of semi-permeabilized cells pretreated with control siRNA (siCtrl) or TMEM109 siRNA (siT109) for 72 h before semi-permeabilization. (E) Quantification of the relative transport efficiency for multiple repeats of the substrates tested in (D). For each repeat, the substrate transport in control siRNA-treated ER fractions was set to 100%. (F) In vitro transport after 72 h of different siRNA treatments targeting SRα, hSnd2, TMEM109, or combinations thereof. Following in vitro translation of the OPG2-tagged model substrates Cytb5, glycophorin C (GypC), tumor necrosis factor receptor superfamily member 17 (BCMA, containing an artificially inserted N-glycosylation site), and synaptotagmin 1 (Syt1), the reactions were solubilized using Triton X−100. Total radiolabeled products (i.e., membrane-associated and non-targeted nascent chains) were recovered by immunoprecipitation via the OPG2-tag and analyzed by SDS-PAGE and phosphorimaging. The N-glycosylation of lumenal domains, previously confirmed by treatment with endoglycosidase H [42,43], indicates successful membrane translocation/insertion (1 g or 2 g) in contrast to non-inserted precursor proteins (0 g). (G) Relative membrane insertion efficiencies were calculated using the ratio of N-glycosylated protein to non-glycosylated protein, relative to the control siRNA-treated cells (set to 100% insertion efficiency). Quantifications are given as means ± SEM for independent insertion experiments from separate siRNA treatments performed in triplicate (n = 3). (H) Carbonate extraction of the 35S-Met labeled model substrate Reep5. The total translation reaction (T) was centrifuged through a sucrose cushion, and the pellet was resuspended in alkaline carbonate solution before a second centrifugation step through an alkaline sucrose cushion occurred. Samples of supernatants after the first (S1) and second (S2) centrifugation as well as the final pellet (p) that includes the membrane-integrated Reep5 are shown. Statistical comparison of multiple conditions was performed using a one-way ANOVA followed by Dunnett’s multiple comparison post-test (G). Statistical comparison of TMEM109 silencing (C,E) was based on a Student’s t-test comparing against the control siRNA treatment that was normalized to 100% for each tested protein (C) or substrate (E). p-values are indicated by asterisks with p < 0.05 (*) < 0.01 (**) < 0.001 (***).
Supplier Page from MilliporeSigma for Anti-TMEM109 antibody produced in rabbit