PI 3‐kinase delta enhances axonal PIP 3 to support axon regeneration in the adult CNS

Abstract Peripheral nervous system (PNS) neurons support axon regeneration into adulthood, whereas central nervous system (CNS) neurons lose regenerative ability after development. To better understand this decline whilst aiming to improve regeneration, we focused on phosphoinositide 3‐kinase (PI3K) and its product phosphatidylinositol (3,4,5)‐trisphosphate (PIP 3). We demonstrate that adult PNS neurons utilise two catalytic subunits of PI3K for axon regeneration: p110α and p110δ. However, in the CNS, axonal PIP 3 decreases with development at the time when axon transport declines and regenerative competence is lost. Overexpressing p110α in CNS neurons had no effect; however, expression of p110δ restored axonal PIP 3 and increased regenerative axon transport. p110δ expression enhanced CNS regeneration in both rat and human neurons and in transgenic mice, functioning in the same way as the hyperactivating H1047R mutation of p110α. Furthermore, viral delivery of p110δ promoted robust regeneration after optic nerve injury. These findings establish a deficit of axonal PIP 3 as a key reason for intrinsic regeneration failure and demonstrate that native p110δ facilitates axon regeneration by functioning in a hyperactive fashion.

. Gene expression profile of p110 isoforms in the nervous system from previously published RNAseq datasets.
Panels A to C show data from Tedeschi et al (2016), Panels D and E from Koseki et al (2017), and Panels F to M from Brain-RNAseq databases (http://www.brainrnaseq.org/), Zhang et al (2014) and (2016).

A
Normalised mean expression values of p110 genes in adult mouse DRG neurons isolated after sciatic nerve lesion compared to a sham control. B Normalised mean expression values of p110 genes in cultured mouse DRG neurons 6-, 12, 24and 36 h post-plating (representing the shift from arborising to elongating axon growth). C Normalised mean expression values of p110 genes in cultured mouse DRG neurons from at embryonic days 12.5 and 17.5. D Relative abundance of p110 mRNA levels in cortical neurons cultured from E18 rat embryos at increasing periods of time in vitro. E Relative abundance of p110 mRNA levels in cortical neurons cultured from E18 rat embryos at increasing periods of time in vitro, also showing expression levels other regeneration-associated genes. F-M Relative abundance (FPKM) of p110 genes in various mouse and human brain cell types (astrocytes, neurons oligodendrocyte precursor cells, newly formed oligodendrocytes, myelinating oligodendrocytes, microglia/macrophages and endothelial cells). Each replicate consists of pooled cortices from 3 to 12 mice. For human samples, n = 6-12 for each cell type. Data are shown as the mean AE SEM. See Zhang et al (2014) and (2016)  A Time-lapse images (single confocal section imaged by spinning disc microscopy) of a DRG growth cone cultured from adult AKT-PH-GFP mice. Arrows point to hotspots/regions of increased fluorescence indicative of AKT-PH recruitment. See also associated Movie EV1. B Time-lapse images (single confocal section imaged by spinning disc microscopy) of a DRG growth cone cultured from adult AKT-PH-GFP mice, stained with cell-mask orange (magenta colour) to detect the membrane. Arrows indicate a dynamic region of AKT-PH-GFP recruitment that does not label for membrane aggregation. See also associated Movie EV2. C Adult DRG growth cone cultured from AKT-PH-GFP mice, fixed for PIP 3 immobilisation (see methods section) and labelled with an antibody to PIP 3 (magenta). Arrows and dotted circles indicate colocalisation. D Non-neuronal cell from a dissociated DRG culture from AKT-PH-GFP mice, fixed for PIP 3 immobilisation (see methods section), and labelled with an antibody to PIP 3 (magenta). Arrows indicate colocalisation. Inset highlights colocalisation at a large region of AKT-PH-GFP recruitment. E N1E neuroblastoma cells stimulated with insulin and labelled for PIP 3 in the presence (lower panels) or absence (upper panels) of the pan-PI3K inhibitor GDC-0941.
Cells were co-labelled for F-actin to show cell density (right four panels). Left four panels are heatmap to show fluorescence intensity. Middle four panels are grey scale images of PIP 3 fluorescence . significance as measured by two-tailed Student's t-test. C Representative images showing expression of AAV2.GFP and AAV2.p110d in eye cup and retinal sections 2 weeks after AAV injection. Retinae from eyes injected with AAV2.p110d (right-hand panels) were labelled for TUJ1 (green) and p110d (red). Blue colour is DAPI to indicate nuclei. Lower panel inset images show individual colours at the same scale as the full image. D Quantification of the p110d immunofluorescence in RGCs transduced with AAV2.p110d. n = 41 for AAV2.GFP, 31 for AAV2.p110d. Data are shown as the mean AE SEM.
P-values indicate statistical significance as measured by two-tailed Student's t-test.