Hubert group research

How muscle cell membranes resist to mechanical stress due to contraction and relaxation is a challenging question. A network of subsarcolemmal filamentous proteins interacts with the membrane bilayer through phospholipids and through transmembrane proteins that themselves associate to components of the extracellular matrix. Among this network, a key element is the protein dystrophin coded by the largest human gene DMD. Lack of dystrophin leads to the Duchenne Muscular Dystrophy (DMD) a severe genetic disease that affects 1/3500 male newborns. The less severe Becker Muscular Dystrophy (BMD) is caused by the expression in patients of truncated, shortened forms of dystrophin which retain a partial function of the wild type dystrophin. This is the pattern for the development of gene therapies which aim to restore the expression of short but efficient forms of dystrophin. However, very little is known about the structure of native as well as pathologic truncated dystrophins.

Our research project aims at better understanding the molecular basis for muscle membrane scaffolding in normal and DMD or BMD cells by in vitro and in silico structure to function approaches. We first showed that dystrophin interacts strongly with membrane phospholipids, a binding which is involved in the resistance to sarcolemmal rupture during muscle contraction (Sarkis et al., 2013).

Figure 1. The phospholipid monolayer DOPC:DOPS or PCPS is essentially elastic and when dystrophin (DYS R11-15) and actin F are added, a visco-elastic system is progressively obtained. The lipid monolayer is 10-fold more resistant to the shear stress when the complex lipid / R11-15 / F-actin is present compared to the lipid / R11-15 alone.


Recently, we combine 1) experimental data from studies of these truncated dystrophins in vitro by biochemical and biophysical tools such as Small angle X-ray scattering (SAXS) with 2) clinical data obtained by the clinicians and geneticists involved in these pathologies in France and with 3) with in silico homology modelling and molecular dynamics (Molza et al., 2014) to construct a clear understanding of the structure of normal and truncated dystrophin.

Figure 2. The interaction between dystrophin, nNOS and F-actin. Structural information for the nNOS PDZ domain and F-actin filaments are respectively provided by X-ray diffraction and cryo-EM experiments. PDB codes are 1qau for nNOS and 3mfp or 3g37 for F-actin. Dystrophin structural data were obtained in solution from Small-angle X-ray Scattering (SAXS) or in silico from homology modelling followed by molecular dynamics optimization.


We mainly showed that BMD variability of severity among the different mutations - deletions have a clear structural basis (Nicolas et al., 2015).

Figure 3. Structural consequences of the exon deletions obtained by in silico molecular homology modelling. Schematic representation of the exons and theircorresponding encoded proteins of interest, with their molecular weight. On the top of each drawing, the exons are noted by their numbering: a vertical bar indicates an ‘in-frame’ succession of repeats while the ‘>’ sign indicates an ‘out-of-frame’ succession of exons. The new bounded exons around a deletion are indicated in red. Repeats are coloured and numbered according to the alignment from Winder (1995). The green diamond represents the hinge 3. Molecular homology models are shown on the right of the schematic representation of proteins (cyan: R16, purple: R17, yellow: R18, grey: R19, red: R20, orange: R21). An inset shows that a tandem repeat is made of three helices (A–C) joined by loops and structured in a coiled-coil; the continuity of the filament is obtained by the common helix containing the C-helix of

the first repeat followed by the A′-helix of the following repeat. In the R16–21 model, each wild-type repeat has three α-helices gently wrapped in a coiled-coil. This filamentous structure is interrupted by the presence of hinge 3 in green. Models of the four truncated proteins are presented with the repeats coloured as in the wildtype R16–21. The truncated proteins all bear deletions starting at exon 45, which encodes the C-terminal half of repeat 17. The RΔ45–47 protein lacks part of repeats 17

and 18 and forms a structure clearly different from a true repeat at the site of the deletion. This incomplete repeat is known as a ‘fractional repeat’. RΔ45–48 is lacking part of repeat 17, all of repeat 18 and part of repeat 19 and a structure similar to a true repeat is formed at the site of the deletion, with three helices forming a new coiled-coil called ‘hybrid repeat’. RΔ45–49 is truncated from the middle of repeat 17 to the middle of repeat 19, and the two remaining parts of repeats 17 and 19 form two helices wrapped in a double coiled-coil; as the third helix of a conventional repeat is lacking in this mutant, the directionality of the molecule at the junction with hinge 3 is changed. This mutant therefore bears a ‘fractional repeat’ with a non-filamentous topology. RΔ45–51 lacks part of repeat 17 and all of repeats 18 and 19 and hinge 3; the remaining parts of repeat 17 and of repeat 20 make a ‘hybrid repeat’.


We reported for the molecular description of one of the most important complex of dystrophin with its partner nitrous oxide synthase (nNOS) and showed why certain BMD truncated dystrophins could not retain this complex, therefore explaining the increase of severity of the phenotype in these cases (Molza et al., 2015).

Figure 4. Final dystrophin R16-17: nNOS-PDZ complexes obtained after interactive flexible docking.


Our expertise in structural biology is appreciated by collaborators working on other aspects of dystrophin function (Rau et al., 2015).