Mst3b mediates the response of retinal ganglion cells (RGCs) to growth factors. (a–g) Representative photos of RGCs transfected with plasmids expressing the shRNA constructs and/or Mst3b variants shown on the left. Cells were exposed to media alone or to a combination of oncomodulin, mannose, and forskolin (Ocm/m/f) as indicated. Cells were fixed 3 days later and immunostained for EGFP, a marker for transfected cells, and either βIII-tubulin, a marker for RGCs (antibody TuJ1), or the myc-his-tag present on recombinant human Mst3b. (h) Quantitation of axon outgrowth. RGCs transfected with control shRNA extended axons when exposed to Ocm/m/f (a,b,h), whereas cells expressing mst3b shRNA did not (c,h). Outgrowth was restored when RGCs expressing mst3b shRNA were co-transfected with a plasmid expressing his-tagged human Mst3b (d,h). Expression of human Mst3b was verified by immunostaining for the his tag (e). Expression of kinase-dead (k/d) Mst3b blocked the effects of growth factors (f,h), whereas expression of constitutively active (c/a) Mst3b resulted in outgrowth even with growth factors absent (g,h). Scale bar for a–g: 40 μm. Each experiment included 4 blinded, independent observations (10–100 cells per well), and each experiment was repeated 3 times. *, **, ***increase relative to untreated controls significant at P < 0.05, P < 0.01, or P < 0.001, respectively. ^†††decrease relative to control cells treated with growth factors significant at P < 0.001. Error bars represent s.e.m.

Mst3b is essential for optic nerve regeneration. RGCs were infected with AAV2 expressing EGFP plus either control or mst3b shRNA. Four weeks later, the optic nerve was crushed and the lens was injured to induce inflammation and transform RGCs into an active growth state. Regeneration was quantified 2 weeks later. (a) Longitudinal sections through the optic nerve stained for GAP-43 (red, a marker for regenerating axons) and EGFP (green, a marker for transfected neurons). Asterisk denotes the site of nerve injury; box shows the region magnified in b. Scale bar: 100 μm. (b) Area distal to the injury site from a case expressing mst3b shRNA. Merged image shows that GAP-43-positive axons that arise from uninfected RGCs (red), but not double-labeled axons (yellow) that arise from transfected neurons, extend distal to the injury site. Arrows indicate regenerating axons from uninfected RGCs. Scale bar: 100 μm. (c) Mean number of axons that regenerate ≥ 500 μm beyond the injury site. The set of bars on the left shows axons that arise from transfected RGCs. The number of regenerating axons was normalized by the percentage of transfected cells to account for differences in infection efficiency. The set of bars on the right shows axons of noninfected cells, normalized by the percentage of noninfected cells. Results are based on 7 animals per group, 4-6 sections per animal, and 0-24 axons per section. ^†††decrease relative to cases transfected with control shRNA significant at P < 0.001. Error bars represent s.e.m.

Mst3b knockdown in DRG neurons. Sections through the rat DRG four weeks after injecting AAV2 expressing EGFP and either control or mst3b shRNA. (a) Infection of DRG neurons is demonstrated by co-immunostaining (arrows) with antibodies to the neuronal marker, βIII tubulin (TuJ1, red), and to virally expressed EGFP (green). Overall transfection rates were 21% for the control shRNA group (N = 10) and 35% for the mst3b shRNA group (N = 13). (b) AAV2 expressing EGFP and mst3b shRNA infects small IB4-positive (red) non-peptidergic neurons (arrows) as well as small IB-4-negative neurons (arrowheads). (c) AAV2 also infects large neurons (arrow) that stain positively with antibodies to the neurofilament protein NF200 (red). Scale bar: 40 μm. (d) Knockdown of Mst3b in DRG neurons. Ganglia were immunostained with antibodies to Mst3b (red) and EGFP (green) 4 weeks after infecting DRGs with AAV2 expressing EGFP and either control (top) or mst3b shRNA (bottom). Mst3b can be visualized in cells expressing control shRNA (arrows) (top), but not in cells expressing mst3b shRNA (arrowheads) (bottom). Scale bar: 40 μm.

Mst3b activation and regulation of p42/44 MAPK phosphorylation in vivo. (a) Mst3b was immunoprecipitated from lysates of DRG neurons 3 days after peripheral nerve crush or sham surgery. Kinase assays were performed using HF1 as a pseudo-substrate and γ-[³²P]-ATP, with or without 6-thioguanine present. Reaction mixes were separated by SDS-PAGE and kinase activity was visualized by autoradiography (top). Protein loading was visualized by Coomassie Blue staining (bottom). Position of molecular weight markers are shown on the right. (b) Quantitation of Mst3b kinase activity based on densitometry of autoradiograms using Image J (NIH) averaged from 3 separate experiments. (c,d) Cross-sections through the DRG of animals injected 4 weeks earlier with AAV2 expressing EGFP and either control shRNA (top) or mst3b shRNA (bottom). Arrows point to infected, EGFP-positive neurons. (c) Double-immunostaining with antibodies to phospho-p42/44 MAPK (red) and EGFP (green) reveal a selective reduction of phospho-MAPK in DRG neurons expressing the mst3b shRNA. (d) Double-immunostaining with antibodies to p42/p44 MAPK (red) and EGFP (green) reveal that Mst3b knock-down does not alter overall p42/44 MAPK levels. (e) Quantitation of changes in the overall expression and phosphorylation of p42/p44 MAPK based on 4-6 sections from 3 animals per condition. Scale bar: 40 μm. **Increase relative to controls significant at P < 0.01; ^{†, ††}Decrease relative to induced state significant at P < 0.05 or P < 0.01, respectively. Error bars represent s.e.m.

Infection of RGCs with AAV2 expressing shRNAs in vivo. (a) Animals received intraocular injections of AAV2 expressing EGFP and either a control shRNA or mst3b shRNA 4 weeks prior to dissection. Cross-sections through the retina show double-labeled cells (arrows) expressing TuJ1 (to identify RGCs: red) and EGFP (a marker for transfected cells: green). Transfection efficiencies averaged 58% for the virus expressing control shRNA and 68% for the virus expressing mst3b shRNA. (b) Suppression of Mst3b expression. Cross-sections through the retinas of animals injected 4 weeks earlier with AAV2 expressing EGFP and either control shRNA (top) or mst3b shRNA (bottom). Double-immunostaining with antibodies to Mst3b (red) and EGFP (green) shows co-expression of the two markers in cells expressing control shRNA (arrows, top), but a loss of Mst3b in RGCs expressing mst3b shRNA (arrowheads, bottom). Scale bar: 20 μm. (c) Quantitation of Mst3b expression. EGFP-positive cells expressing either control or mst3b shRNA were analyzed for intensity of Mst3b immunostaining in cross-sections through all cases using Image J. (d) Lack of macrophage infection: Four weeks after intravitreal injections of AAV2 expressing mst3b shRNA and EGFP, the lens was injured to induce an inflammatory response. Double immunostaining shows ED-1-positive macrophages to be uninfected, as evidenced by an absence of EGFP. Scale bar: 40 μm. Error bars represent s.e.m.

Mst3b mediates the response of DRG neurons to NGF. (a–f) Representative photos of DRG neurons transfected with plasmids expressing shRNA constructs and/or variant forms of Mst3b as shown on the left. Cells were exposed to NGF (50 ng/ml) or media alone. Three days later, cells were fixed and immunostained to detect EGFP, a marker for transfected cells, and βIII-tubulin to identify neurons (antibody TuJ1). (g) Quantitation of axon growth. (a,b,g) Cells expressing the control shRNA extend neurites when treated with NGF (b,g). (c,g) Cells expressing mst3b shRNA (arrowheads) fail to respond to NGF; an nontransfected cell in the same field shows outgrowth (arrow). (d,g) NGF-induced outgrowth is restored when cells expressing mst3b shRNA (green) are co-transfected with a plasmid expressing human Mst3b. (e,g) Cells expressing kinase-dead (k/d) Mst3b (arrowheads) fail to extend neurites when treated with NGF; a non-transfected cell (arrows) in the same field shows outgrowth. (f,g) expression of constitutively active (c/a) Mst3b (arrowhead) is sufficient to induce outgrowth in the absence of NGF; arrow points to an untransfected cell in the same field that is not growing. Scale bar in a–f: 100 μm. Each experiment included 4 blinded, independent observations of about 10 cells per well, and each experiment was performed three times. **,***increase relative to baseline (control shRNA-transfected cells without NGF) significant at P < 0.01 or P < 0.001, respectively. ^{†, †††} decrease relative to control shRNA-transfected cells treated with NGF significant at P<0.05 or P<0.001, respectively. Error bars represent s.e.m.

Mst3b knockdown attenuates peripheral nerve regeneration. (a–d) The radial nerve was crushed 4 weeks after infecting cervical DRG neurons with AAV2 expressing EGFP and either mst3b- or control shRNA. Axons extending beyond the injury site (arrowheads) were visualized 3 days later in sections stained with antibodies to detect GAP-43 (red) and EGFP (green). The box in b is magnified in panel e. Only profiles that are labeled with both markers are considered as regenerating axons that arise from infected neurons (arrows). Scale bar for a–d: 200 μm. (e) High-power images show regenerating axons that arise from neurons expressing control shRNA: many axons are positive for both GAP-43 (red) and EGFP (green) (arrows). Scale bar: 100μm. (f) Quantitation of GAP-43⁺/EGFP⁺ axons. The number of regenerating axons was normalized by the percent of infected cells to account for differences in infection efficiency. Note the marked attenuation of axon growth in cells expressing mst3b shRNA. (g) Quantitation of GAP-43⁺/EGFP^- axons from non-infected neurons shows similar levels of regeneration irrespective of the virus infecting neighboring neurons. Data show mean number of axons at the indicated distances from the injury site (8-9 animals per group, 4-6 sections per animal, 0-10 axons per section). **^,***Differences between number of axons arising from cells expressing control vs. mst3b shRNA are significant at P < 0.01, P < 0.001, respectively. Error bars represent s.e.m.