EMD ΔLEM does not traffic in U2OS cells, while the TMD alone does.

A) U2OS cells expressing mouse EMD-GFP and ΔLEM-GFP were induced with 1 ug/mL doxycycline for 24 hours, treated with 100nM lysosome blocking agent Bafilomycin A1 (BafA1) overnight, and fixed and stained for lysosome marker Lamp1. Scale bar, 10 um.

B) Summary of EMD domains and truncation mutants.

C) EMD enriches at the INM by binding to the nuclear lamina, or exits the ER into the secretory pathway. Before reaching the lysosome, EMD transiently accesses the cell surface where lumenal GFP is exposed to anti-GFP antibody.

D) Example FACS plot of mouse WT vs ΔLEM surface labeling; fluorescent anti-IgG secondary antibody included as background control. Antibody signal is divided by total GFP+ signal to yield histograms on the right.

E) Quantification of the antibody:GFP ratio from 4 independent experiments, with error bars representing SD. * indicates adjusted P-value = 0.0451 by one-way ANOVA with Šídák’s multiple comparisons test.

F) U2OS cells expressing mouse TMD-GFP (lumenal tag) or GFP-TMD (cytosolic tag) were treated overnight with BafA1, then fixed and stained with an anti-GFP antibody.

G) Example surface labeling FACS histogram of U2OS cells expressing WT, LEM-TMD-GFP, or TMD-GFP.

Emerin’s trafficking depends on its TMD, not on the LEM domain.

A) Surface anti-GFP antibody:GFP histogram of human WT EMD, ΔLEM, and EMD with mouse cytochrome B5 TMD (EMDCyb5 TMD) chimera.

B) U2OS cells expressing mouse EMD-GFP and EMDCyb5 TMD-GFP were induced with 1 ug/mL doxycycline for 24 hours, treated with 100nM lysosome blocking drug bafilomycin A1 (BafA1) overnight, and fixed and stained for lysosome marker Lamp1. Scale bars, 10 um.

C) TMD mutation strategy. The TM alpha helix of human EMD was divided into N-terminal and C-terminal halves, and the aromatic residues in each half were mutated to alanine to generate mutants with similar ΔG-insertion values. For the full TMD mutant, all aromatic residues were mutated to alanine, yielding a predicted ΔG-insertion of -2.299.

D) Surface antibody:GFP FACS histogram and E) quantification of the mutants diagramed in (C). N = 4 independent experiments

F) Surface antibody:GFP FACS histogram and G) quantification of mouse WT EMD, RA, ΔLEM, and ΔLEMΔQRRR truncation surface expression. N = 4 independent experiments.

H) Summary of emerin constructs that do or do not traffic.

I) Western blot analysis of emerin constructs induced with 2 ug/mL doxycycline for 48 hours, then washed and incubated for an 18 hour chase.

J) Quantification of Western blot band intensity from (I). GFP antibody signal after washout was divided by the respective unwashed condition to yield the fraction GFP remaining after doxycycline washout across 3 independent replicates.

For all panels: *** indicates adjusted P-value <0.0005, * indicates P = 0.0104. All P-values were obtained by one-way ANOVA with Šídák’s or Tukey’s multiple comparisons tests.

A) Lengths and predicted ΔG-insertion values of ER protein cytochrome B5, PM protein syntaxin 3, and emerin INM proteins emerin and Lap2β.

B) Single pass human transmembrane proteins from Membranome database plotted by free energy of transfer into a lipid bilayer. Lines represent median; **** indicates P < 0.0001 by one-way Anova with Tukey’s multiple comparisons test.

C) Transmembrane domain length (by amino acid count) of the proteins in (B). ** indicates P = 0.0028; **** indicates P < 0.0001.

D) Protein sequence alignment of emerin’s LEM domain across species. Conserved LEM domain amino acid class indicated in blue. QRRR motif is conserved in mammals and is exposed in the ΔLEM construct.

RXR motif limits TMD-dependent trafficking of emerin and LAP2β without influencing COPI binding.

A) Surface anti-GFP:GFP histogram and B) quantification of indicated RA + TMD mutant combinations. ** indicates P = 0.0021. N = 3 independent experiments.

C) Diagram of LAP2β domain structure and position of RXR motifs. APEX2 fusion contains no RXR motifs.

D) Antibody:GFP histogram and E) quantification of the highest 25% GFP-expressing cells diagramed in (C). RXRmut : LAP2β RXR1 mutated to AAA. * indicates P = 0.0173; ** indicates P = 0.0072

F) Spectral counts of COPI proteins immunoprecipitated by WT FLAG-EMD normalized to negative control IP. Spectral counts of the mouse EMD bait and common contaminant filamin A (FLNA) plotted for comparison. N = 2 independent experiments.

G) Spectral counts of proteins immunoprecipitated by FLAG-WT and FLAG-RA mouse EMD normalized to each respective bait. Dotted line represents equal co-immunoprecipitation with the two constructs. COPI proteins and endogenous human EMD highlighted in yellow and green, respectively.

All P-values were obtained by one-way ANOVA with Šídák’s or Tukey’s multiple comparisons tests.

A) GFP versus surface anti-GFP FACS plots of negative control anti-IgG, WT LAP2β / RXR1mut, and WT EMD / LAP2β APEX2-TMD.

B) Median anti-GFP antibody:GFP ratio of the highest 25% GFP-expressing cells in (A).

C) Western blot analysis of anti-FLAG immunoprecipitations. U2OS cells expressing no FLAG construct (neg), FLAG-WT mouse EMD, and FLAG-RA mutant were lysed and incubated with anti-FLAG magnetic beads. Equal volume equivalents were separated by SDS-PAGE and blotted with an anti-EMD antibody to visualize tagged and untagged EMD.

D) U2OS cells expressing FLAG-EMD were incubated with lysosome blocker bafilomycin A1 (BafA1) overnight, then fixed and permeabilized with 0.1% Triton-X-100, 0.02% SDS, 10 mg/ml BSA in PBS. Cells were stained with anti-FLAG antibody to visualize relocalization during BafA1 treatment. Scale bars, 20 um.

Safe harbor expression reveals that C-terminal GFP destabilizes emerin.

A) Diagram of emerin integration into the AAVS1 locus. Landing pad BFP is exchanged for mCherry-P2A-emerin +/-GFP via Bxb1 and PhiC31 integrases. Integrases irreversibly recombine landing pad attP and donor attB sites into attR and attL sites.

B) Strategy to compare overexpressed emerin to knockout (KO) rescue.

C) Western blot analysis WT and EMD KO DICE hiPSCs expressing N- and C-terminally GFP-tagged emerin. Untagged endogenous EMD and GFP-tagged EMD are detected by the same anti-EMD antibody.

D) Anti-GFP surface labeling histogram and E) quantification of C-terminally tagged emerin integrants from (B-C). Statistical significance was determined using one-way ANOVA with Šídák’s multiple comparisons test. N = 3 independent experiments.

F) FACS plot and G) quantification of steady state GFP abundance. Data were normalized to the GFP-EMD signal in the EMD WT background. N = 3 independent experiments. ** indicates P = 0.0079 by two-way ANOVA with Tukey’s multiple comparisons test.H) DICE landing pad WT, EMD KO, and KO re-integrated with untagged constructs were lifted, fixed, and stained with anti-EMD antibody. Fluorescence was quantified by flow cytometry and normalized to the WT DICE integrant. MFI, median fluorescence intensity of antibody signal. N = 5 replicates over 4 independent experiments. **** indicates P <0.0001 by mixed effects analysis.

A) EMD KO DICE hiPSCs were integrated with GFP-tagged EMD constructs. Cells were treated with vehicle or lysosome blocker bafilomycin A1 (BafA1) overnight, then fixed and stained with anti-LAMP1 antibody.

B) EMD KO DICE hiPSCs were integrated with untagged EMD WT, TMD mutant, or RA mutant. Cells were treated overnight with BafA1, then fixed and stained with anti-EMD antibody. Scale bars, 20 um.

C) We mined a dataset of matched RNAseq and quantitative protein abundances across 29 human tissues to evaluate EMD transcript and protein abundances. EMD protein and transcript abundance are largely uncorrelated (R2 ∼ 0), while, for comparison, the enzyme spleen tyrosine kinase (SYK) has well-correlated protein and RNA levels (R2 = 0.89)