Fig. 4

α9-K1 axons regenerate to the brain stem. A Diagram showing the spinal cord segment from which the following sections (B–D) were taken—an animal from the α9-kindlin-1 group with a thoracic lesion and AAV injections into L4,5 DRGs. Created with BioRender.com. B Cross-section of the spinal cord (C3) with α9-V5 labelled axons. The diagram on the bottom left shows where the detail came from. The figure was Created with BioRender.com. Scale bar: 50 μm. White arrows point to axons at the boundary between the dorsal horn and the dorsal column. C Axons at level C3, labelled for V5 only. Most axons follow a path at the margin between the dorsal horn and the dorsal column. Many axons are seen in a layer between the dorsal horn and dorsal column white matter. In this figure there is some staining in the white matter, which is background staining on meninges and glial processes, but a few probable axons are seen in the dorsal horn white matter. D Axons rostral to the lesion were double stained for α9-V5 and kindlin-1-GFP, showing colocalisation of both markers. This suggests that both integrin and kindlin-1 are required for axon growth above the lesion. Scale bar: 100 μm. E, F Longitudinal sections of the medulla double stained for α9-V5 and kindlin-1-GFP with the cuneate nucleus (delineated by dotted line) at the top. The bundle of axons approaches the edge of the nucleus, and some grow towards the nucleus but not into it F compared to uninjured controls in E. There are no labelled axons in the nucleus, indicating that there are no uninjured sensory axons. Scale bar of longitudinal section: 100 μm. G, H Transverse sections of spinal cord showing uninjured control (G) and α9 integrin and kindlin-1 group (H) stained for α9-V5 with gracile and cuneate nuclei at the top. The transverse sections support the findings from the longitudinal sections in E, F. The staining around the meninges seen in some images is background, which appears differently in some sections. Scale bar of the transverse sections: 100 μm. The borders of the sensory nuclei are indicated by white dotted lines in E–H. I Frontal sections of rat medulla oblongata (created with BioRender.com modified from the Atlas of Paxinos and Watson (ISBN: 9780080475158); bregma: − 14.6 mm) showing the location of sensory nuclei. Gr—gracile nucleus, Cu—cuneatus nucleus. J Bar graphs show the number of axons after C4 cervical injury with C6, C7 DRG injections (left) and T10 thoracic injury with L4, L5 DRG injections (right). Approximately 900 sensory axons grew into the rostral spinal cord in the α9-kindlin-1 group, with their number decreasing slightly as they approached the hindbrain. The distance of the thoracic lesion from the medulla is 5 cm, where regenerated axons were observed. Bar graphs show data with mean ± SEM (n = 7–12 animals per group). K Schematic representation of the main results of this study. Only in the α9-kindlin-1 group was there substantial regeneration beyond the lesion into the rostral cord. In the kindlin-1 group, axons regenerated into the laminin-containing connective tissue at the core of the lesion. Created with BioRender.com. No difference was observed between cervical and thoracic injuries in terms of the pathways followed by axons under different AAV-expressing conditions