Designing Protein Therapies: Lessons Learned from Cerezyme® for Gaucher Disease

Abstract

The development of carbohydrate-modified human glucocerebrosidase (Ceredase® and Cerezyme® [Genzyme Corporation, Cambridge, Massachusetts]) for the treatment of Gaucher disease has transformed the lives of patients and led to the development of enzyme replacement therapies (ERTs) for many other lysosomal storage disorders (LSDs). Gaucher disease and the use of Cerezyme as an ERT serve as a useful model for the design of protein and other therapies for the treatment of LSDs. Like other LSDs, Gaucher disease is a progressive and lethal metabolic disorder that results in irreversible endorgan damage when untreated. Early intervention is key to preventing irreversible complications. Several hundred different mutations have been identified that give rise to Gaucher disease,1 but there is still no reliable correlation between genotype and phenotype. Response to therapy is organ specific; that is, liver and spleen demonstrate a rapid response at a relatively low dose2 but a higher dose, earlier intervention, and long-term therapy are necessary to demonstrate benefit for the skeletal disease. Gaucher disease differs from other LSDs in the low percentage of patients that develop antibodies3; this reflects the fact that many patients produce residual, though inactive or partially active, protein. Cerezyme has an excellent safety and efficacy record, demonstrated over 15 years of clinical use.4 Nevertheless, we have evaluated alternative production methods in an attempt to improve mannose receptor targeting and enzyme delivery to target tissues. To evaluate the effect of increased amounts of terminal mannose residues on the uptake and biodistribution of human glucocerebrosidase, we produced human glucocerebrosidase with mannose chain lengths of 3 to 9 residues.5 Because Cerezyme differs from the human placental enzyme used in the production of Ceredase by 1 amino acid, we also evaluated the effect of that amino acid difference on uptake and biodistribution (even though no clinical difference has been observed between Cerezyme and Ceredase).6 Although terminal mannose is essential for efficient uptake into target cells, more mannose does not alter the uptake or biodistribution of human glucocerebrosidase. More mannose did, however, increase binding to the mannose-binding lectin in vitro, which could have negative consequences on clearance and uptake in vivo. The single amino acid difference between Cerezyme and placental glucocerebrosidase had no effect on half-life, uptake, or tissue targeting.5 In addition to engineering the carbohydrate residues on glucocerebrosidase, it is also possible to engineer the protein backbone to improve biological properties. Although the crystal structure of glucocerebrosidase has been determined,7 it is of little use for predicting how to improve the protein. It does not explain why most mutations cause disease or how the protein is activated by saposin C or phospholipids and is inconsistent with biochemical data. This most likely reflects the fact that the crystal structure is just a static image of a dynamic molecule. The use of competitive inhibitors as chemical chaperones has recently been proposed as a therapeutic strategy.8 This is based on the hypothesis that mutations which cause LSDs result in misfolded proteins, which are degraded by the proteasome, an effect that is likely to be mutation specific. We have studied the effect of chemical chaperones on the uptake and stabilization of Cerezyme as well as recombinantly produced N370S mutant glucocerebrosidase.9 These studies demonstrated that the increased enzyme levels observed are due to lysosomal stabilization. Chaperones, which are competitive inhibitors, remain bound in the lysosome and inhibit enzyme activity, as also demonstrated by Steet et al.10 Because these competitive inhibitors tend to have the same pH optima as the natural substrate—that is, they bind better at low pH—there is likely to be a narrow (if any) therapeutic window. As our results demonstrate, the effects of chaperones on protein levels reported in the literature can be accounted for by reduced lysosomal degradation. Gaucher disease is the first and most effectively treated of the LSDs, and Cerezyme is the gold standard against which to measure future LSD therapies. Alternative manufacturing platforms are unlikely to offer any improvement over Cerezyme. Small molecule chaperones are at best likely to be limited to a few of the several hundred disease-causing mutations. More promising approaches for the future include the development of a secondgeneration substrate inhibition therapy with improved specificity and reduced adverse effects, as well as gene therapy for LSDs with central nervous system involvement. Clinical Therapeutics/Volume 29, Supplement C, 2007