Fig 2: Genetic compensation in reep5 CRISPR knockout zebrafish mutant.a Sequencing analysis generated by Sanger Synthego shows reep5 gRNA targeted and edited homozygous CRISPR mutant sequences in the region around the guide sequence. The horizontal black underlined region represents the guide sequence and the horizontal red underline indicates the protospacer adjacent motif (PAM) site with the vertical black line representing the cut site. b, c CRISPR/Cas9-mediated reep5 homozygous mutant compared to control zebrafish embryos, n = 3 independent biological crosses. Scale, 500 μm. d, e Optical imaging analysis of control and reep5 CRISPR mutant embryos shows normal cardiac chamber orientation and development in the reep5 CRISPR mutant hearts. Scale, 10 μm. f Heart rate analysis of control and reep5 homozygous mutant embryos upon 600 μM verapamil treatment for 30 min; a total of 46 control, 24 reep5 CRISPR mutants, 32 control + verapamil, and 36 reep5 CRISPR mutants + verapamil embryos were analyzed, n = 3 independent biological crosses; data are presented as mean ± SEM. g, h reep5 CRISPR mutants injected with REEP5 MO (1 ng) appears phenotypically normal compared to overt cardiac abnormalities in the REEP5 morphant embryos. Images shown are representative of 287 REEP5 MO and 232 reep5 CRISPR mutants injected with REEP5 MO embryos, n = 3 independent biological crosses. Scale, 500 μm. i, j Optical imaging analysis of REEP5 MO and reep5 CRISPR mutants injected with REEP5 MO embryo hearts shows normal cardiac chamber orientation and development in the reep5 CRISPR mutant hearts. Scale, 10 μm. k Bar graph showing phenotypic penetrance represented as percentage from 287 REEP5 MO and 232 reep5 CRISPR mutants injected with REEP5 MO embryos, n = 3 independent biological crosses; data are presented as mean ± SEM. l Immunoblot analysis of cardiac-enriched ER structure proteins (RTN4, ATL3, and CKAP4) in reep5 CRISPR homozygous mutants shows upregulation of RTN4 in reep5 CRISPR mutant hearts. Source data containing original uncropped immunoblots are provided as a Source Data file. Asterisks indicate a statistically significant p value in a Tukey’s multiple comparison analysis where **p < 0.01 and ***p < 0.001.

Fig 3: Rtn4 and Climp63 jointly define the R2 tubule width. (A) A simple model of the R2 tubule based on an elliptical cross-section, in which the luminal height h is varied as the membrane curvature at the two edges is held at 1/re = 0.08 nm−1. Color represents the local curvature. (B) The expected R2 tubule width w as a function of h according to this simple model (line). Colored squares mark the experimental w for cells expressing the different Climp63 mutants in C, calculated as the STORM-determined Rtn4–Rtn4 separations minus 5-nm lipid-bilayer thickness. (C) Distribution of center-to-center separations between opposing Rtn4 lines in R2 tubules, for COS-7 cells expressing mCherry-tagged mouse Climp63 mutants of different intraluminal sizes. Whiskers and boxes show 10, 25, 50, 75, and 90% quantiles. For each data point, n = 200 local distances were pooled from five cells. (C, Inset) Schematics of the full-length (FL) and truncated Climp63 versions examined. CC1 to CC3, intraluminal coiled-coil domains; Cyto, cytosolic domain; FP, mCherry fluorescent protein; TM, transmembrane domain. (D–I) Representative STORM images of the immunolabeled endogenous Rtn4 in COS-7 cells expressing mCherry-tagged full-length (1–575) (D), 2xlumen (E), 1–437 (F), 1–336 (G), 1–301 (H), and 1–192 (I) mutants of mouse Climp63. (D–I, Insets) Epifluorescence images of the mCherry channel. Filled and hollow arrowheads point to examples of R2 and R1 tubules, respectively. In H, the arrow points to an untransfected cell in the same view, which facilitates a direct comparison. (Scale bars, 1 µm.)

Fig 4: Mass spectrometry analysis identifies REEP5 interactions with known cardiac SR shaping proteins.a Immunoblot analysis of nickel-His purification of CnVA-REEP5 from transfected HEK293 cells shows stable expression of CnVA-REEP5 and successful immunoprecipitation of CnVA-REEP5 monomer (~34 kDa) and dimer (~68 kDa). b Identification of known SR/ER-shaping proteins as REEP5 interacting proteins by mass spectrometry analysis. Average precursor MS1 peak areas (peptide m/z signal) as defined by iBAQ (intensity based absolute quantification) are shown, n = 3 independent mass spectrometry runs. To calculate fold change, average null values (n.d.—not detected) were inputted with a value of 10. RTN reticulon, ATL atlastin, CKAP4 cytoskeleton-associated protein 4, ACTC1 alpha cardiac muscle actin 1, MYL6 myosin light chain 6, GAPDH glyceraldehyde-3-phosphate dehydrogenase. Asterisks indicate a statistically significant p value in a Tukey’s multiple comparison analysis where *p < 0.05. c RNASeq analysis of the RTN, ATL families of proteins, and CKAP4 in human fetal heart and adult heart tissue data using data from Human Protein Atlas. d Immunoblot analysis of nickel-His REEP5 immunoprecipitation lysates for RTN4/Nogo-A/B, ATL3, and CKAP4, n = 3. e HEK293 cells were transfected with myc-tagged Nogo-A and Nogo-B plasmids and detected with myc and α-tubulin antibodies. f Co-immunoprecipitation assay of cotransfected HEK293 cells with anti-REEP5 antibody (left panel) and anti-RTN4/Nogo-A/B antibody (right panel) demonstrated interaction between REEP5 and RTN4, n = 3. Eluates were collected on ice and loaded directly for blotting, or samples were boiled prior to blotting. g–j Co-immunoprecipitation and reverse order co-immunoprecipitation assays with g anti-REEP5, h anti-RTN4/Nogo-A/B, i anti-ATL3, and j anti-CKAP4 antibodies in adult mouse cardiac microsomes (input), followed by immunoblots analysis, n = 5. Source data containing original uncropped immunoblots are provided as a Source Data file.

Fig 5: In vivo AAV9-mediated REEP5 depletion in mice results in increased cardiac fibrosis, activated cardiac ER stress, cardiac dysfunction, and death.a Immunoblot analysis of AAV9-mediated REEP5 depletion in mice 4 weeks post viral infection. b Histological analyses of AAV9 Scram shRNA and REEP5 shRNA injected mouse hearts at 4 weeks post viral infection. Top panels: H&E, bottom panels: Masson’s trichrome stain. Scale, 20 μm. All images shown are representative of approximately 20 total images captured per condition, n = 3 independent biological replicates. c Transmission electron microscopy analysis of AAV9 REEP5 shRNA-injected myocardium reveals degeneration of muscle fibers and disrupted SR membranes and organization compared to scram controls. M mitochondria, SR sarcoplasmic reticulum, V vacuoles. Scale, 0.5 nm. All images shown are representative of approximately 20 total images captured per condition, n = 3 independent biological replicates. d In vivo echocardiographic heart function assessment of AAV9 scram and AAV9 REEP5 shRNA-injected mice at 4 weeks post viral infection, n = 5. e Echocardiographic M-mode and B-mode measurements showed significantly compromised cardiac function with reduced cardiac ejection fraction and cardiac output, n = 5 independent biological replicates. Asterisks indicate a statistically significant p value in a Tukey’s multiple comparison analysis where ***p < 0.001; data are presented as mean ± SEM. f Immunoblotting analysis of cardiac ER stress markers (GRp78, GRp94, XBP1, ATF4, and CHOP) and ER-dependent apoptosis (caspase 12) upon AAV9-induced REEP5 depletion in vivo in mice, n = 3; p values are shown. g Immunoblots of RTN4, ATL3, and CKAP4 upon AAV9-induced REEP5 depletion in vivo in mice, n = 3 independent biological replicates. Source data containing original uncropped immunoblots are provided as a Source Data file.