Fig. 8

DUX4 interacts with C1qBP in a normal size myofiber with a cluster of nuclei and in a proximal close cell. C1QBP and DUX4 interaction was determined by PLA as described in Fig. 7 (healthy control sections and negative controls are presented in Fig. S14), using a mouse C1QBP antiserum and the rabbit anti-DUX4 E5-5 MAb. A (upper panel) 3D reconstruction of a muscle section area. Box 1 surrounds clear PLA signals with large dots in clusters and the circles indicate nonspecific signals (as determined in Fig. 7). A (bottom panels) Different focal depths (44 images with steps of 0.09 μm in the Z axis) showing PLA signal (arrowheads) in two close cells surrounded by laminin-α2 staining. The upper cell is small and could correspond to an activated satellite cell fusing with the below myofiber (arrow points to a lack of laminin-α2 only at a specific depth: images Z34/44). PLA signals are detected inside the nuclei (images Z10 and Z20) but also in the thin cytoplasm and in cluster at one nucleus side. Moreover, PLA signals are also observed next to the membrane and at the periphery of a very close nucleus residing inside the adjacent normal-size fiber (images Z10 and Z20), and next to the ‘fusion’ area (arrow). The myofiber present a very large cluster of nuclei, several are aligned and very close at the plasma membrane. B Cluster of PLA signals detected at a myofiber tip (triangular unusual shape) close to a myofiber having a central nucleus (#). The box is magnified for a better laminin-α2 detection at the fiber tip. A fuzzy laminin staining is also observed within the ‘triangular’ tip (arrow), overlapping with the PLA signals. PLA was performed with the specific anti-DUX4 and-C1qBP primary antisera pairs on muscle sections from 5 patients with FSHD (also see Fig. S13E-F) (in parallel to muscle sections from 3 healthy controls, see Fig. S14A)