Fig 2: ICV-ERT with rhß-gal for 8 weeks normalizes ß-gal activity in systemic tissues, which coincides with a partial reduction in urinary A2G2' substrate. A and B, ß-gal activity detected in liver tissue (A) or bone marrow (B) from WT vehicle (n = 4–7), KO vehicle (n = 6–8), and KO rhß-gal–treated mice (n = 6), with results expressed as individual values error bars represent the standard deviation (S.D.). ****, p < 0.0001. Also indicated above each graph is a representative Western blot of ß-gal protein levels in pooled liver homogenates (A) or bone marrow lysates (B) prepared from WT or GLB1 KO mice treated with vehicle or rhß-gal. For comparison, 4 ng of purified rhß-gal was also included on gels as an indicator of the precursor, nonlysosomal enzyme. Also included was 5 µg of cell lysate prepared from WT human fibroblasts (WT fibro) as an indicator of mature ß-gal successfully delivered to lysosomes. Enolase was used as loading control. C–F, correlation between ß-gal activity levels in brain (C) and liver (D) with A2G2' substrate levels detected in brain (E) and urine (F) from individual mice from the 8-week PoC study. Note that the individual ß-gal activity and A2G2' levels in C–F are shown for this small subset of mice from the 8-week PoC study, the reason being that we only managed to successfully collect urine samples from these mice 24 h after the final ICV dose of rhß-gal. Mouse numbers are internal mouse ID numbers assigned by the GM1 mouse study subgroup.

Fig 3: ICV administered rhß-gal exhibits broad bilateral bio-distribution throughout the brain, which coincides with clearance of multiple substrates in GM1 gangliosidosis mice. A, detection of rhß-gal in WT mouse brain tissue (n = 3 mice) 3 h after a single, unilateral injection (100-µg dose) into the lateral ventricle utilizing MS. Colored bars represent detection of enzyme activity in each macro-dissected area. Mouse left and right brain hemispheres were each dissected into the following regions: OFB, CBX, HS, CB, and MB (containing pons and medulla) and analyzed individually. Each pair of bars represents signal in the left (L) or right (R) sagittal hemispheres. B, summary of the 2-week and 8-week short-term proof-of-concept (2 wk or 8 wk PoC) ICV-ERT studies evaluated in the GM1 gangliosidosis mouse model. The 2-week PoC study commenced at 8 weeks of age, with mice receiving four ICV doses (100 µg/dose) of rhß-gal or vehicle (aCSF) over 2 weeks. The 8-week PoC study commenced at 12 weeks of age, with mice receiving weekly ICV dosing (100 µg/dose) with rhß-gal or vehicle for 8 weeks. Mice were taken down 24 h after the final ICV dose of enzyme. C, representative ß-gal Western blotting of ß-gal protein levels in pooled brain homogenates prepared from the left brain hemisphere of WT or GLB1 KO mice treated with vehicle or rhß-gal, as summarized in B. For comparison, 4 ng of purified rhß-gal was also included on gels as an indicator of the precursor, nonlysosomal enzyme. Also included was 5 µg of cell lysate prepared from WT human fibroblasts (WT fibro) as an indicator of mature ß-gal successfully delivered to lysosomes. D, quantification of mature lysosomal ß-gal protein levels in brain homogenates from Western blots of individual samples, standardized to an enolase loading control of Western blots described in C. Quantified results are expressed as individual values, error bars represent the standard deviation (S.D.). 2 wk PoC, WT vehicle, n = 5; KO vehicle, n = 6; KO-rhß-gal, n = 7. 8 wk PoC, WT vehicle, n = 4; KO vehicle n = 7; KO rhß-gal, n = 9). E, corresponding ß-gal activity levels detected in brain homogenates prepared from mice, with results expressed as individual values with the S.D. indicated. 2 wk PoC, WT vehicle, n = 5; KO vehicle, n = 9; KO rhß-gal, n = 7. 8 wk PoC, WT vehicle, n = 4; KO vehicle, n = 6; KO rhß-gal, n = 9. F–I, GM1 and GA1 ganglioside levels (F and G) and A2G2' glycan levels (H and I) detected in brain homogenates from the 2-week PoC (WT vehicle, n = 5; KO vehicle, n = 9; KO-rhß-gal, n = 7) and 8-week PoC mice (WT vehicle, n = 4; KO vehicle, n = 7; KO rhß-gal, n = 9). ****, p < 0.0001.

Fig 4: ICV-ERT with rhß-gal augments lysosomal ß-gal activity in patient fibroblasts and hippocampal neurons of the GM1 gangliosidosis mouse brain. A, in situ detection at ×40 magnification of ß-gal activity observed in normal fibroblasts at pH 4.5 and 7 after 18 h using X-gal substrate. X-gal staining was also performed in control GM1 gangliosidosis untreated fibroblasts (GM05653) or following 3 h of cellular uptake with 100 nm rhß-gal. Alternatively, after cellular uptake, cells were washed and chased in growth medium without enzyme for a further 48 h. B, ß-gal Western blotting of cells described in A. Enolase was used as a loading control. C, in situ detection, at ×20 magnification, of ß-gal activity at acidic pH using X-gal substrate in cryo-frozen sagittal brain sections from a 16-week-old control WT mice, a control GLB1 KO mouse, or GLB1 KO mice following a single ICV dose (100 µg) of rhß-gal. rhß-gal ICV-ERT–treated mice were taken down 3, 24, or 48 h after ICV administration. Blue signal indicates areas of ß-gal activity. Black arrows indicate development of signal in the CA3 region of the hippocampus (neuronal) and in vascular structures. D, Western blotting of brain homogenates (left and right hemispheres) prepared from individual WT-vehicle (n = 4), KO-vehicle (n = 7), or KO-rhß-gal (n = 7) mice after receiving four ICV doses of rhß-gal (100 µg/dose) over 2 weeks, with mice being taken down 24 h after the last dose. Mouse numbers are internal mouse ID numbers assigned by the GM1 mouse study subgroup. For comparison, 2.5, 5, or 7.5 ng of rhß-gal was also included on gels as an indicator of the precursor, nonlysosomal enzyme. 2, 4, or 8 µg of fibroblast lysate prepared from normal human fibroblasts was used as an indicator of mature, lysosomal ß-gal. E, quantification of mature ß-gal protein levels shown in D, standardized to the enolase loading control and expressed as individual values, error bars represent the standard deviation (S.D.). F, quantification of Grp78 protein levels shown in D, standardized to the enolase loading control and expressed as individual values with the S.D. indicated.

Fig 5: ICV administered rhß-gal reverses secondary neuropathology in GM1 gangliosidosis mice. A, representative LAMP2 IHC images (20× magnification) of sagittal hippocampus sections prepared from the left brain hemisphere from mice dosed weekly for 8 weeks (dosing regimen indicated in Fig. 3B). Black arrows, presence (KO vehicle) or absence (WT vehicle and KO rhß-gal-treated) of LAMP2 signal in the CA1–CA3 and hippocampal neurons. B, quantitative analysis of the percentage of area of LAMP2-positive signal measured over the cortex, hippocampus, thalamus, brain stem and cerebellum. n = 4–6 mice for each group. C, representative ß-gal, LAMP2, and enolase Western blots of pooled brain homogenates prepared from the left hemispheres from WT or GLB1 KO mice treated with eight weekly doses of vehicle or rhß-gal: WT vehicle, n = 4; KO vehicle, n = 6; KO rhß-gal, n = 9. D, quantification of LAMP-2 protein levels from Western blots of individual samples, standardized to an enolase loading control, with individual values indicated error bars represent the standard deviation (S.D.). WT vehicle, n = 4; KO vehicle, n = 6; KO rhß-gal, n = 9. E–H, representative GFAP IF (E) or IBA1 IF (G) images of cortical mouse brain tissue. Corresponding percentage area analyses for cortical GFAP (F) or IBA1 (H) are indicated. All quantitative data are represented as individual values error bars represent the standard error of the mean (SEM). GFAP IF, WT vehicle, n = 4; KO vehicle, n = 5; KO rhß-gal, n = 5. IBA1 IF, WT vehicle, n = 3; KO vehicle, n = 3, KO rhß-gal, n = 5. ****, p < 0.0001. **, p < 0.01; *, p < 0.05.