FIGURE SUMMARY
Title

FKRP-dependent glycosylation of fibronectin regulates muscle pathology in muscular dystrophy

Authors
Wood, A.J., Lin, C.H., Li, M., Nishtala, K., Alaei, S., Rossello, F., Sonntag, C., Hersey, L., Miles, L.B., Krisp, C., Dudczig, S., Fulcher, A.J., Gibertini, S., Conroy, P.J., Siegel, A., Mora, M., Jusuf, P., Packer, N.H., Currie, P.D.
Source
Full text @ Nat. Commun.

<italic>fkrp</italic><sup><italic>−/−</italic></sup> fish possess a basement membrane failure phenotype.

ad Muscle fibre-crossing and detachment phenotypes at the muscle basement membrane of fkrp−/ mutant larvae. ab” Z-projected images from the entire mediolateral extent of the myotome centred on the region of the anal pore. Phalloidin conjugated to TRITC (red) stains for f-actin as a marker of skeletal muscle and paxillin (green) visualises vertical myosepta muscle basement membrane integrity. White arrowheads: muscle fibres crossing at defective muscle basement membranes, asterisks: muscle fibre detachment. 3 dpf, scale bar = 100 μm. 5 dpf, scale bar = 75 μm, white boxes in a’, b’ denote the high-magnification views presented in a” and b”. c, d Quantitation of fibre detachment (c) and basement membrane crossing events (d) per myosepta in wild-type (n = 9) or fkrp mutant larvae (n = 9) over five independent experiments, analysed using two-way ANOVA multiple comparisons assuming non-parametric data, the significance of ***(P = < 0.0001), error bars represent SEM. e Physiological analysis of muscle function. The maximum active force (mN) generated over a specific time interval (ms) was measured in individual genotypes of 6 dpf larvae represented in single-twitch recordings of fkrp+/+ sibling (sib) black line (n = 8), fkrp−/ green line (n = 6) and dag1/− purple line (n = 5).

The fibronectin–collagen axis is disrupted at the muscle basement membrane in <italic>fkrp</italic> mutants.

ab” Fibronectin and collagen immunochemistry. Whole-mount staining and confocal imaging of zebrafish myotomes centred on myotome 10 (1 dpf) and myotome 12 (5 dpf). Images are z projections of the mediolateral extent of the myotome stained for f-actin (red) to mark muscle fibres and DAPI (blue) for nuclei. a Fibronectin (Fn) staining (green) at the vertical myosepta at 1 dpf, images representative of n = 9, three larvae from three separate clutches on different weeks, most severe and weakest phenotypes excluded (scale bar = 30 μm). a’dag1/−, a”fkrp/−. b Collagen-1 (Col1a, green) staining of the muscle basement membrane in wild-type sibling fish. b’dag1−/− and b”fkrp−/−, white arrow: absent collagen, scale bar = 40 μm. ab” Lateral views anterior to the left. c Transmission electron micrographs (TEM) of longitudinal sections of 7 dpf zebrafish myotome centred on the muscle basement membrane (red arrowheads). Red arrows: absent collagen fibrils, red star: fibre detachment from vertical myosepta, scale bar = 0.5 μm. Wild-type siblings, c’, dag1−/c”, fkrp−/−, micrographs representative of (n = 9), three larvae from three separate clutches on different weeks, most severe and weakest phenotypes excluded. d Schematic of fibre-crossing and detachment model at the Myotendinous Junction (MTJ) and disease progression, d’ canonical DGC axis in dag1/− and, d” combined detachment crossing in fkrp/−. e, f Measurement of max intensity from fibronectin (e) and collagen (f) staining analysed by two-way ANOVA, ***(P = <0.0001), plotted points outside 95% confidence interval, box represents 5–95%, median centre line, whiskers = SEM, Tukey’s multiple comparison analysis, three independent experiments. g Passive force measurement at 6 dpf, calculated by plotting the maximum active force at given loads, as passive tension (mN) against external stretch (Lo%), with representative linear regression analysis of the plotted points of fkrp+/+ sibling (sib); black line (n = 8), fkrp−/, red line (n = 6) and dag1−/ blue line (n = 5) was used to test the significance of the differences (***P = <0.0001), error bars = SEM.

FKRP-dependent N-linked sialyation of fibronectin is required for collagen binding and muscle basement membrane localisation in vivo.

aN-glycan analysis of muscle fibronectin (Fn) isolated from control and LGMD2I and CMD patient cells. Blue box highlights the area of change at m/z 965.9 and m/z1111.4. b Quantification n = 3 averaged from three experimental replicates, each repeated with three technical replicates (*P < 0.05). c Co-immunoprecipitation of collagen released from purified fibronectin, normalised to healthy human controls (***P < 0.0001); three technical repeats repeated three times, box represents 5–95%, median centre line, whiskers = SEM. d Collagen and fibronectin binding quantitated via Biacore, before and after treatment with neuraminidase, an enzyme that specifically removes terminal sialic acid residues. e Schematic of the fibronectin rescue experiment. f Injection of human fibronectin into the single-cell stage of 1-dpf embryos. Fibronectin tagged with Alexa488 localises correctly to the myosepta in wild-type larvae, white arrows. Scale bar = 100 μm. Only fully sialylated fibronectin can rescue collagen deficits evident in fkrp/ larvae. fkrp+/+ and fkrp−/− larvae uninjected or injected, with either desialylated fibronectin or fully sialylated fibronectin, visualised for collagen localisation via z projections of the entire mediolateral extent of the myotome centred on the anal pore. Note also that fkrp+/+ control fish, injected with sialylated fibronectin, exhibited a lower collagen deposition at the myoseptal boundaries. This lower collagen deposition is attributed to a previously described fibronectin–collagen feedback loop, scale bar = 100 μm. g Quantitation of the relative max intensity of collagen deposition at muscle basement membranes, 5 dpf, (***P < 0.0001), plotted points outside 95% confidence interval analysed by two-way ANOVA, Dunn’s multiple comparisons, from five repeats, box represents 5–95%, median centre line, whiskers = SEM. h Physiological rescue after fibronectin injection. Force transducer measurement of passive tension over stretch load % (n = 5 for each experimental condition) (***P < 0.0001), error bars = SEM.

The relative localisation of fibronectin and myosin10 is altered in the Golgi of FKRP-deficient cells.

Fibronectin and myosin10 protein localisation within the trans-Golgi, as determined by STED microscopy and Gaussian distribution line profile analyses, images representative of a minimum of three repeats separated by a minimum of a week and three technical repeats. a STED images of myosin10: blue, fibronectin: red and Golgi 58 k protein: green. Merged image marked for Golgi with dashed lines and white solid lines marking the location of the line profiles quantitated in b. b Individual line profiles, normalised to maximum and using Gaussian distribution curve in Fiji image analysis software, fibronectin red, myosin10 blue, means marked with dotted lines. c Co-localisation of myosin10 and fibronectin from the Golgi identified by a 58 K antibody stain, Pearson correlation analysis (*P < 0.05), (***P < 0.0001) (n = 9), box-and-whisker plot, middle line = mean, box = 95% confidence interval, error bars = SEM, one-way ANOVA analysis. d Quantitation of the difference between individual line profile means, analysed from Gaussian fit curves in b (n = 9), (*P < 0.05), (***P < 0.0001). Box-and-whisker plot, middle line = mean, box = 95% confidence interval, error bars = SEM. eh Analysis of Airyscan data in Imaris image analysis software. e Images from the Airyscan rendered in 3D, myosin10 in yellow, fibronectin in blue and Golgi: Golgi Reassembly Stacking Protein 2 (GORASP2) in green, scale bar = 2 µm. f Pearson’s correlation of fibronectin and myosin10 using Golgi marker: GORASP2 was used as a mask for analysis. Box-and-whisker plots, error bars represent a 95% confidence interval and middle box line represents mean (***P < 0.0001). One-way ANOVA analysis (n = 9 for each sample), one-way ANOVA analysis. g Sphericity of fibronectin within Golgi, Imaris generated surfaces and software analysis. Box-and-whisker plots, error bars represent a 95% confidence interval and middle box line represents mean (***P < 0.0001). One-way ANOVA analysis (n = 9 for each sample). h Airyscan data with cells rendered in 3D and statistically coded for sphericity in a heat map (with a value of one (red) indicating a perfect sphere) reveal the altered Golgi structure in patient cells, scale bar 2 µm. i Schematic of FKRP action. The schematic illustrates that in healthy control cells, FKRP is required to correctly localise myosin10 within the terminal Golgi, a process important for sialylation of fibronectin, that in turn regulates its binding to collagen. In FKRP deficiency, myosin10 is no longer anchored correctly and fibronectin is consequently not sialylated correctly. This lack of sialylation results in a failure of collagen–fibronectin binding at the MBM, which ultimately leads to a loss of MBM stability and an inability of individual muscle fibres to resist passive force.

ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.

ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.

ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.

Acknowledgments
This image is the copyrighted work of the attributed author or publisher, and ZFIN has permission only to display this image to its users. Additional permissions should be obtained from the applicable author or publisher of the image. Full text @ Nat. Commun.