Fig 1: gp78 mediates Cx32 degradation. a Overexpression of a nonfunctional gp78 mutant (C341/378S) significantly decelerates Cx32L90H degradation (mean ± SEM, N ≥ 3, *P value < 0.05, two-tailed Student’s t tests). b Mutant gp78 (C341/378S) overexpression inhibits polyubiquitination of Cx32. FLAG-tagged Cx32L90H was expressed either alone or in the presence of V5-tagged wild type gp78 or gp78C341/378S as indicated. Samples were immunoprecipitated against FLAG (Cx32L90H) and blotted against Cx32 or ubiquitin (Ub). Where shown, MG132 was added to inhibit proteasomal degradation of Ub-modified Cx32L90H. As can be seen from the blots and quantifications in c, MG132 leads to increased ubiquitination of Cx32L90H in the absence or presence of gp78wt overexpression, while gp78C341/378S overexpression inhibits Cx32L90H polyubiquitination. c Quantifications of the data shown in b (mean ± SEM, N = 4, ns: nonsignificant, **P value < 0.01, two-tailed Student’s t tests)
Fig 2: A model for the biogenesis and quality control of multipass TM proteins. A red asterisk indicates a mutation that leads to membrane misintegration for Cx32. Chaperones and quality control factors identified in this study to act on Cx32 are shown
Fig 3: Cx32L90H shows defects in gap-junction formation and rapid degradation. a COS-7 cells were transfected with the indicated constructs and immunostained for FLAG-tagged Cx32 (magenta), PDI (yellow) as an ER marker, or GM130 (yellow) as a Golgi marker. Nuclei were stained with DAPI (blue). Anti-FLAG immunofluorescence data are depicted as maximum intensity projections from deconvoluted z-stacks, while PDI, GM130, and nuclei are shown as a central cell plane from the same, nondeconvoluted images. Gap-junction plates lining cell–cell boundaries are indicated with white arrowheads. Pictures are representative of cells from at least three different biological replicates. Scale bars correspond to 20 µm. b HEK293T cells transfected with the indicated constructs were incubated with either cycloheximide (CHX) alone, or additionally with the proteasome inhibitor MG132 where indicated. Arrowheads indicate monomer (M) and dimer (D) bands quantified to determine Cx32 turnover. Quantifications are shown below the immunoblots (mean ± SEM, N = 3). c Half-lifes without MG132 for Cx32wt and Cx32L90H as derived from b are shown
Fig 4: Single-point mutations lead to failures in membrane integration for Cx32. a Schematic of Cx32, showing its predicted topology. b Predicted free energies for helix insertion for apolar-to-polar missense mutations investigated in this study. For ?G < 0 (green) an energetically favorable membrane integration is predicted, while ?G > 0 (red) indicates an unfavorable insertion reaction. Transmembrane helices and ?G values were predicted according to Hessa et al.17. c Side and top view of the modeled hexameric Cx32 connexon. Disease-causing mutants investigated in this study are shown in a CPK representation on a single monomer. Transmembrane helices are shown in black. Individual Cx32 monomers are numbered from 1 to 6. d Cx32wt and e Cx32L90H with individually introduced glycosylation sites in the indicated regions were transfected into HEK293T cells, lysates treated with or without EndoH as indicated and analyzed by immunoblotting. The schematics below each indicate the location of the individually assessed glycosylation sites (N) and possible topologies deduced from the observed glycosylation. An orange arrow in the blots and an orange hexagon in the schematic indicate sites that became glycosylated. f HEK293T cells transfected with the indicated Cx32 TM2 mutant constructs were analyzed as in d, e. Constructs all carried a glycosylation reporter site in loop 2 (L2:NVT). Predicted free energies of membrane insertion for TM2 of the respective proteins as well as a quantification of relative loop 2 reporter site glycosylation (mean ± SEM, N = 3) are shown below the immunoblots
Fig 5: Connexin mutants are recognized by the ER quality control system. a Mass spectrometry volcano plot for FLAG-tagged Cx32wt, immunoprecipitated in 1% digitonin from transfected HEK293T cells. Enriched proteins are denoted with their Uniprot gene name. Cx32 is shown in orange and ER chaperones investigated further in this study are highlighted in blue (EMC10) and green (Cnx). Either a rabbit monoclonal anti-FLAG antibody or a rabbit IgG isotype control was used. b Representative blots from immunoprecipitation experiments from HEK293T cells transfected with the indicated Cx32 constructs. Interaction of Cx32 with endogenous Cnx and EMC subunits was detected and increased for Cx32L90H with both EMC4 and EMC10 (mean ± SEM, N = 3, ns: nonsignificant, *P value < 0.05, two-tailed Student’s t tests). Quantifications were performed as described in the Methods section. c Transient knockdown of EMC5/10 by siRNA (average knockdown (KD) efficiencies are shown below the blots) increases glycosylation of a reporter site in loop 2 for Cx32L90H. Monomeric species ±glycosylation are shown on the blot, indicative of the topologies depicted on the right. Changes in glycosylation, quantified as described in the Methods section, are shown on the right (mean ± SEM, N = 5, **P value < 0.01, two-tailed Student’s t tests). d Same as in b for co-transfected hamster BiP (mean ± SEM, N = 3, **P value < 0.01, two-tailed Student’s t tests). e Schematic of a reporter construct to assess BiP-TM region binding. The nonglycosylated immunoglobulin ? light chain CL domain with its own ER import sequence is followed by a flexible linker connected to the TM sequence of interest shown below the schematic. TM segment 1 was inverted to allow for a type I topology. A C-terminal glycosylation site (NVT, marked in red) allows to assess membrane integration (no glycosylation) versus ER import (possible glycosylation). Membrane integration/ER import was assessed for Cx32 TM segment 1, 2, and 2 carrying the L90H mutation by transfection of the constructs into HEK293T cells and EndoH deglycosylation where indicated. f CL-TM constructs were co-transfected with hamster BiP into HEK293T cells and their interaction was analyzed by co-immunoprecipitation experiments coupled to immunoblots
Supplier Page from MilliporeSigma for Anti-Connexin-32 (106-124) antibody produced in rabbit