Glycosylation-independent lysosomal targeting of acid α-glucosidase enhances muscle glycogen clearance in pompe mice

Abstract

We have used a peptide-based targeting system to improve lysosomal delivery of acid α-glucosidase (GAA), the enzyme deficient in patients with Pompe disease. Human GAA was fused to the glycosylation-independent lysosomal targeting (GILT) tag, which contains a portion of insulin-like growth factor II, to create an active, chimeric enzyme with high affinity for the cation-independent mannose 6-phosphate receptor. GILT-tagged GAA was taken up by L6 myoblasts about 25-fold more efficiently than was recombinant human GAA (rhGAA). Once delivered to the lysosome, the mature form of GILT-tagged GAA was indistinguishable from rhGAA and persisted with a half-life indistinguishable from rhGAA. GILT-tagged GAA was significantly more effective than rhGAA in clearing glycogen from numerous skeletal muscle tissues in the Pompe mouse model. The GILT-tagged GAA enzyme may provide an improved enzyme replacement therapy for Pompe disease patients.

FIGURE 1.

SDS-PAGE and Western blots. 100 ng of untagged rhGAA and HEK293-produced BMN 701 used in studies in this report (lanes 1 and 2, respectively) were separated by reducing SDS-PAGE, and gels were silver-stained (A) or transferred and probed on Western blots with polyclonal anti-GAA PAb2443 (B) or anti-IGF-II (C). Sizes of molecular mass markers (lane M) are labeled in kDa.

FIGURE 2.

Uptake of recombinant GAA in L6 rat myoblasts. HEK293-produced BMN 701 and rhGAA were added at concentrations of 97 and 251 nm, respectively, to wells containing L6 cells. Inhibitors (5 mm M6P, 2.4 mm IGF-II, or both M6P and IGF-II) were also included in some wells. Uptake of the recombinant proteins is expressed as units of GAA activity/mg of cellular lysate. Each data point represents the average of duplicate samples. Error bars represent standard deviation.

FIGURE 3.

Uptake inhibition of HEK293-produced BMN 701 by the IGF-II binding domain of the CI-MPR. Uptake assays with 10 nm loadings of rhGAA (A) or HEK293-produced BMN 701 (B) were performed in the presence of increasing amounts (0–250 nm) of soluble receptor fragments 1288 (circles, wild-type IGF-II binding domain) or 1355 (squares, mutated IGF-II binding domain). Enzyme uptake is expressed as a percentage of the uptake in wells lacking receptor fragment inhibitors. Each data point represents the average of duplicate samples. Error bars represent standard deviation.

FIGURE 4.

Determination of the half-life of rhGAA and HEK293-produced BMN 701 within L6 cells. Multiple wells with L6 cells were loaded with rhGAA (circles) at 235 nm or BMN 701 (squares) at 90 nm. Following an 18-h uptake incubation, all cells were washed four times with growth medium. Day 0 samples were then lysed and frozen, and the remaining samples were cultured for up to 14 days with triplicate wells harvested and frozen daily (Days 1–14). GAA units present in cell lysates are expressed as a percentage of activity of the Day 0 samples. Error bars display standard deviation. Data were fit to a first-order exponential decay curve using KaleidaGraph software to determine the enzymatic half-life. This experiment was repeated three times, and representative results are shown.

FIGURE 5.

Intracellular processing of rhGAA and HEK293-produced BMN 701. Cell lysates from half-life uptake assays (Fig. 4) on rhGAA (upper panel) and BMN 701 (lower panel) were probed on Western blots with monoclonal antibody 3A6-1F2, which recognizes an epitope within amino acid residues 204–482 of human GAA. Sizes of molecular mass markers are labeled in kDa on the left of the panels. Lane 1, L6 lysate from cells grown in the absence of GAA protein; lane 2, rhGAA (upper panel) and FS701 (lower panel) uptake load material; lanes 3–17, L6 lysate harvested on days 0–14, respectively. This experiment was repeated three times, and representative results are shown.

FIGURE 6.

Intracellular localization of CHO-produced BMN 701. C2C12 cells grown in the presence (upper images) or absence (lower images) of BMN 701 were examined using immunofluorescence microscopy. Whole cells were imaged with differential interference contrast (DIC) filters. Nuclear DNA was imaged with a DAPI filter. CHO-produced BMN 701 was imaged with anti-human GAA primary antibody 3A6-1F2 (α-BMN 701 panels), anti-mouse IgG AF594 secondary antibody, and a Texas Red filter. Lysosomal LAMP1 was imaged with FITC-conjugated anti-LAMP1 primary antibody (α-LAMP1 panels) and a FITC filter. Fluorescent images were overlaid (merged panels).

FIGURE 7.

Determination of K_uptake for CHO-produced BMN 701, HEK293-produced rhGAA, and alglucosidase alfa. Uptake assays with increasing amounts of GAA protein BMN 701 (circles), rhGAA (squares), or alglucosidase alfa (CHO-produced rhGAA manufactured by Genzyme) (triangles) were performed. Each point is the average of duplicate samples. Data points were fit to the Michaelis-Menten equation to determine K_uptake values of 5.4, 141, and 147 nm for BMN 701, rhGAA, and alglucosidase alfa, respectively.

FIGURE 8.

Determination of affinity of BMN 701 for the IGF-II binding domain of the CI-MPR by surface plasmon resonance. Similar amounts of the biotinylated CI-MPR constructs 1355 and 1288 were immobilized on the surface of a SA sensor chip (280 and 300 RU, respectively). IGF-II and BMN 701 were injected in a volume of 120 μl over the 1355 and 1288 coupled flow cells at a rate of 40 μl/min. After 2 min, the solutions containing the analyte were replaced with buffer, and the complexes were allowed to dissociate for 3 min. An average of the R_eq values was determined for each analyte concentration using BIAevaluation version 4.0.1 software. Shown are equilibrium plots for IGF-II (0.5, 1, 2, 4, 8, 12, 20, and 40 nm) (A) and BMN 701 (5, 10, 25, 50, 100, 200, and 400 nm) (B). Equilibrium constants were calculated using nonlinear regression (SigmaPlot version 10.0) and are summarized in Table 1.

FIGURE 9.

Glycogen clearance in Pompe mice. Glycogen levels were measured in various muscle tissues following four injections with CHO-produced BMN 701 or rhGAA. See “Experimental Procedures” for details.