Viral vector-based improvement of optic nerve regeneration: characterization of individual axons' growth patterns and synaptogenesis in a visual target

Abstract

Lack of axon growth ability in the central nervous system poses a major barrier to achieving functional connectivity after injury. Thus, a non-transgenic regenerative approach to reinnervating targets has important implications in clinical and research settings. Previous studies using knockout (KO) mice have demonstrated long-distance axon regeneration. Using an optic nerve injury model, here we evaluate the efficacy of viral, RNA interference (RNAi) and pharmacological approaches that target the phosphatase and tensin homolog (PTEN) and signal transducer and activator of transcription-3 pathways to improve long-distance axon regeneration in wild-type mice. Our data show that adeno-associated virus (AAV) expressing short hairpin RNA (shRNA) against PTEN (shPTEN) enhances retinal ganglion cell axon regeneration after crush injury. However, compared with the previous data in PTEN KO mice, AAV-shRNA results in a lesser degree of regeneration, likely due to incomplete gene silencing inherent to RNAi. In comparison, an extensive enhancement in regeneration is seen when AAV-shPTEN is coupled to AAV encoding ciliary neurotrophic factor (CNTF) and to a cyclic adenosine monophosphate (cAMP) analog, allowing axons to travel long distances and reach their target. We apply whole-tissue imaging that facilitates three-dimensional visualization of single regenerating axons and document heterogeneous terminal patterns in the targets. This shows that some axonal populations generate extensive arbors and make synapses with the target neurons. Collectively, we show a combinatorial viral RNAi and pharmacological strategy that improves long-distance regeneration in wild-type animals and provide single fiber projection data that indicates a degree of preservation of target recognition.

Figure 1. AAV-shRNA mediated suppression of PTEN

(a) A map of AAV2-shPTEN4 construct. Control and shPTEN constructs contain EGFP reporter. (b) Immunocytochemistry showing expression of GFP, PTEN, and beta III tubulin (TUJ1) in cultured cortical neurons incubated with either AAV-anti-luciferase-EGFP (control) or AAV-RNA against PTEN (AAV-shPTEN4) for 7 days. These are representative micrographs from 2 separate experiments. (c) Retina sections following control AAV2-EGFP (EGFP) or AAV2-shPTEN4 injection show PTEN knockdown in ganglion cell layer. (d) Flat-mounted retina 2 weeks following intravitreal injection of AAV2-shPTEN4, stained with antibodies against GFP and TUJ1 shows >90% transduction efficiency in RGCs. (e) Western blot of protein lysates extracted from cultured cortical neurons showing PTEN level at 7 days after the addition of AAV-shPTEN4 or AAV-EGFP. (f) Quantification of PTEN level from the western blot. Values are represented as ratio of PTEN to βIII-tubulin (n=3 biological replicates). (g) Quantification of PTEN expression in RGCs in retinal sections, measured by ImageJ densitometry method. Intensity unit is an arbitary value. Error bars, SEM. Scale bars, 20 μm.

Figure 2. Assessment of RGC axon regeneration following AAV-shRNA mediated knockdown of PTEN

(a) Representative optic nerve sections of mice receiving AAV2-shPTEN4 or control AAV2–EGFP injection. Red asterisk, lesion site. (b) Quantification of regenerating axons in control and shPTEN treated animals (n=5/group) at different distances from the lesion site. The data are represented as mean number of axons per section. (c) Representative images of TUJ1 stained wholemount retinae for both groups. (d) Quantification of RGC density (n=5/group). Error bars, SEM; ****: p<0.0001; ***: p<0.001, *p<0.05 ANOVA with Bonferroni correction (b) and student’s unpaired t-test (d). Scale bars, 100 μm (a); 20 μm (b).

Figure 3. AAV-shPTEN, AAV-CNTF and cpt-cAMP further improves lengthy RGC axon regeneration

(a) Representative optic nerve sections of mice receiving various AAV treatments. Control, AAV2-EGFP; shPTEN4, AAV2-shPTEN4; CNTF, AAV-CNTF; CNTF/cAMP, AAV2-CNTF/cpt-cAMP; PCC, AAV2-shPTEN4/AAV2-CNTF/cpt-cAMP. Red asterisk, lesion site. (a′) Higher-magnification images of the boxed area in A. (b) Quantification of regenerating axons at different distances from the lesion site (n=6–8/group). *: p<0.05, Bonferroni test (PCC significantly different to AAV-CNTF and AAV-CNTF/cpt-cAMP groups). (c) Representative images of TUJ1 stained wholemount retinae from mice receiving various AAV treatments. (d) Quantification of RGC density (n=6/group). Error bars, SEM. Scale bars, 100 μm (a); 20 μm (d).

Figure 4. Tissue clearing and ultramicroscopic assessment of axonal trajectories in the optic chiasm of PCC animals

(a) Adult mouse brains with and without tissue clearing. The eyes and optic nerves are attached to the brain. (b) Top view of an intact mouse brain. Brain from an uninjured mouse shows the entire trajectory of CTB-labeled RGC axons: all visual targets are visible, including the suprachiasmatic nucleus, lateral geniculate nucleus, and superior colliculus, ipsi- and contralaterally to the injected eye. (c) A horizontal optical slice from an unsectioned brain of a PCC-treated animal shows CTB-labeled axons in the optic chiasm. (c′) Higher-magnification of the boxed area in c. (d) Reconstruction of traced axons through the chiasm. (e) Quantification of axonal trajectories into different regions in 5 different cases (i.e. ipsi- and contralateral optic tracts, contralateral optic nerve, and hypothalamic regions). Contra, contralateral; Hypothal. Hypothalamic region; Ipsi, ipsilateral; IOT, ipsilateral optic tract; LGN, lateral geniculate nucleus; ON, optic nerve; OT, optic tract; OX, optic chiasm; 3^rd V, 3^rd ventricle; SC, superior coliculus; SCN, suprachiasmatic nucleus. Scale bars, 100 μm.

Figure 5. Ultramicroscopic visualization of axonal projections in mouse visual target

(a) Optic chiasm, optic tract and SCN shown by CTB labeling in a cleared, uninjured brain. (b) Bottom view of a brain containing the SCN (i.e. dotted circles) imaged from a brain of an injured control mouse subjected to AAV-EGFP injection. None of the control animals had regenerating axons in the brain. (c) and (d) Bottom view of brains of two different PCC-treated animals showing regenerated axons in the chiasm and SCN. Yellow arrowhead indicates a regenerating axon extending into the SCN. Yellow arrow indicates distal end of the same axon showing continued growth beyond the SCN. White arrows in d indicate extensive arborization. (e–j), reconstructions of single axons showing various growth patterns. In (e), an axon bypasses the target. In (f) and (g), axons traverse the target. In (h), axon terminals are found within the SCN. In (i) and (j), axons enter the SCN and form complex arborization. ON, optic nerve; Uninj, uninjured; SCN, suprachiasmatic nucleus. Scale bars, 100 μm.

Figure 6. Regenerate axons following PCC treatment form synapses in the SCN

(a) Schematic coronal brain section showing the SCN marked in a dotted red rectangle. (b) CTB-labeled axons (red) in the SCN express presynaptic marker Vglut2 (green). (b′) High-magnifications of the boxed area in (b), with XZ and YZ projections at the yellow crosshairs. (c) Retinal cross section showing RGCs expressing WGA (green) six weeks following AAV-Cre injection in iZ/WAP mice. DAPI (blue) is used in these tissue sections for nuclear staining. (d) WGA⁺ cells in the SCN of an uninjured animal. WGA (green) and DAPI (blue). (e) WGA⁺ cells in the LGN (as marked by dotted line) of an uninjured animal. WGA (green) and DAPI (blue). (f) No WGA⁺ cells are detected in the injured AAV-EGFP control mice (n=5). (g) Low magnification coronal brain section showing the SCN of an injured PCC animal, marked in a yellow circles immunostained with antibodies against WGA. CTB, red; DAPI, blue. (h) High magnification of the dotted area in (g) showing CTB⁺ axons (red) wrapping or making close contact with WGA⁺ cells in the SCN. Scale bars, 50 μm.