Fig 2: Synthesis of rhGAAs containing modified terminal sialic acids.A, a schematic of chemical modification of terminal sialic acids on rhGAA complex N-glycan structures. rhGAA is oxidized with sodium metaperiodate in the first step of synthesis. Oxidized rhGAA is then reacted with AOAA to obtain rhGAA–AOAA. Glycan annotations are represented by symbols according to Symbol Nomenclature for Glycans (SNFG) conventions. B, chemical structures of modified sialic acid. Step 1 of the synthesis showing periodate oxidation of sialic acid at C7 to form an aldehyde. Oxidized sialic acid is further modified in the second step with AOAA to form a conjugate that contains an oxime bond. AOAA, aminooxyacetic acid; GAA, acid alpha-glucosidase; rhGAA, recombinant human GAA.

Fig 3: Durable gene-corrected cell therapy for Pompe disease requires persistent proliferative cells.a Plots showing the degree of phenotypically normal cardiac tissue when either 10 billion non-proliferative single- or double-corrected cells are dosed or 0.1 billion proliferative single- or double-corrected cells are dosed. For these plots, the GAA production for double-corrected cells is 150% excess that of single corrected cells for four different half-lives of dosed cells. For long-lasting correction, proliferative progenitor cells need to have a half-life exceeding 100 days. b Heatmap showing the proliferative cell dose (assuming double-corrected cells with 75% excess GAA production compared to single-corrected cells) against the half-life of dosed cells. If the dosed cells have a half-life similar to endogenous hepatocytes, a 75 million cell dose is sufficient for matching ERT. c Heatmap showing the proliferative cell dose (assuming that the dosed cells have 250-day half-life) against the GAA production rate of dosed cells, demonstrating that dose has a higher effect on therapeutic efficacy than GAA production gained from double correction. d Heatmap showing that the GAA production against the half-life of the dosed cells (assuming that 250 million cells were dosed), indicating that the half-life of the dosed cells has a higher effect on therapeutic efficacy than the GAA production gained from double correction. e Sensitivity analysis for model parameters relevant to cell therapy. In addition to growth and differentiation rates of the transplanted cells, outcomes are highly sensitive to the proliferative cell half-life, cell dose and excess GAA produced by double gene correction. These later parameters, relevant to the design of therapy with autologous gene-corrected cells, are bolded.

Fig 4: In vivo somatic cell gene correction strategies involve tradeoffs between efficiency, precision, progenitor affinity, and editor stability.a In silico Gene Therapy Efficacy Model (GETEM) for Pompe disease correction in a developing infant. Schematic showing gene correction for two diseased alleles in a liver indicating correction of alleles, a1 and a2, by genome editors 1 and 2 to form gene-corrected cells capable of secreting GAA to enzymatically correct other unedited cells. Secreted GAA is also absorbed by striated muscle tissue (both heart and skeletal muscle). b Percentage of normal cardiac tissue within a developing heart of a Pompe diseased infant after the administration of six doses of genome editors at 23.9 mg/kg. c Cell numbers indicating growth of diseased, normal, and precisely-edited cells in the gene-edited liver depot after the administration of 6 doses of genome editors at 23.9 mg/kg. d Distribution of genotypes in the gene-edited liver depot after the administration of 6 doses of genome editors at 23.9 mg/kg. e Sensitivity analysis of the model indicating the absolute values of parameter sensitivity of tissue morphogenesis factors, genome editor factors, and cell/tissue biology intrinsic factors. f Tradeoff between genome editor efficiency and genome editor stability, focusing on the percentage of enzymatically cross-corrected heart tissue. Heatmap indicates that lower efficiency editors could be efficacious if the extracellular editor stability increases. g Tradeoff between genome editor efficiency and precision, focusing on the percentage of enzymatically cross-corrected heart tissue. Heatmap indicates that lower efficiency editors can be efficacious if higher precision editors are used. h Tradeoff between increasing genome editor dose and progenitor affinity, focusing on the percentage of enzymatically cross-corrected heart tissue. i Using GETEM, heatmap indicating tradeoff in heart muscle correction in the developing infant between the degradation rates of GAA in the serum and cellular GAA, indicating that stabilization of GAA in the serum can improve clinical outcome (gray arrowhead indicates pre-stabilization, black arrowhead indicates post-stabilization). Simulation results for liver and skeletal muscle are shown in Supplementary Figs. 11 and 12.

Fig 5: Histological, ultrastructural and transcriptome analyses show a similarly affected skeletal muscle in Gaa KODBA and Gaa KOB6;129 mice.