Search

Enzyme replacement has been under consideration as a therapeutic strategy for patients with Gaucher disease for more than two decades. Previous studies indicated that single injections of purified glucocerebrosidase reduced the amount of storage material in the liver. It was important to determine whether administration of exogenous enzyme on a regular basis would be of clinical benefit. We report here that weekly i.v. infusions of a macrophage-targeted preparation of human placental glucocerebrosidase in a child with type 1 Gaucher disease increased hemoglobin from 6.9 +/- 0.8 g/dl (+/- 1 SD) to 10.2 +/- 0.4 g/dl (+/- 1 SD) over a 20-week period. The platelet count also increased from a pretreatment value of 30,000 +/- 7000/mm3 (+/- 1 SD) to 54,000 +/- 11,000/mm3 (+/- 1 SD). Phagocytic activity in the spleen decreased during the period of enzyme administration, and there was radiographic evidence of skeletal improvement. These observations document objective clinical responses to enzyme supplementation in a patient with a sphingolipid storage disorder.

Krabbe's disease is caused by a deficiency of galactocerebrosidase in organs and tissues. Determinations of galactocerebrosidase activity had required the use of galactocerebroside labeled with radiocarbon or radiohydrogen. These materials are expensive and their use is restricted to laboratories with radioactive counting facilities. An analogue of galactocerebroside, 2-hexadecanoylamino-4-nitrophenyl-beta-D-galactopyranoside, was synthesized. The hydrolysis of this analogue by extracts of tissues and cells from patients with Krabbe's disease is greatly reduced from normal levels. Cultured skin fibroblasts preparations derived from heterozygous carriers of Krabbe's disease have an intermediate level of hydrolytic activity. Thus, the analogue is a reliable chromogenic reagent for the diagnosis of patients with Krabbe's disease and for the detection of heterozygous carriers of the Krabbe trait.

Enzyme replacement therapy for the alleviation of Gaucher's disease has been impeded because of the difficulty in preparing large amounts of glucocerebrosidase, the enzyme that is deficient in patients with this disorder. A large-scale procedure for the purification of human placental glucocerebrosidase has been developed. The method uses cholate extraction, ammonium sulfate fractionation, acid precipitation, butanol extraction, and hydrophobic chromatography; the final enzyme preparation has a specific activity of more than 10(6) units/mg of protein with an overall recovery of 30%. In addition, the contamination of enzyme preparations, intended for human infusion, prepared by isolation procedures involving concanavalin A columns has been studied and is reported here.

Multiple molecular forms of beta-glucocerebrosidase that permit discrimination between neurologic and non-neurologic phenotypes of Gaucher disease have been identified radioimmunologically in fibroblasts and human brain tissue. In normal human fibroblasts these forms have been shown by NaDodSO4/polyacrylamide gel electrophoresis to have apparent Mr of 63,000 (form A1), 61,000 (form A2), and 56,000 (form B). The Mr 63,000 form may be a precursor of the Mr 56,000 form. Non-neurologic Gaucher disease (type 1) fibroblasts and normal brain tissue are characteristic in that they contain only one major immunoreactive protein, the Mr 56,000 form. In contrast, fibroblast extracts and brain tissue from neurologic Gaucher disease phenotypes contain only the higher molecular weight forms A1 and A2. These data and the low residual activity of the enzyme in all the variants of Gaucher disease suggest that the mutations of beta-glucocerebrosidase are allelic and involve the active site.

The clearance of native human placental glucocerebrosidase by rat liver shows the presence of two distinct enzyme forms with different recognition characteristics. The clearance and uptake of native enzyme by liver cells was compared to that of glucocerebrosidase sequentially treated with neuraminidase, beta-galactosidase and beta-N-acetylglucosaminidase. The initial rate of clearance of infused enzyme was increased greater than 10-fold for the asialo-, agalacto- and ahexoenzymes over that of native glucocerebrosidase. Incorporation of asialo enzyme was increased in hepatocytes over that of native enzyme, while the distribution of agalacto- and ahexoenzyme preparations was increased in non-parenchymal cells. This observation is consistent with the identification of a galactose receptor on hepatocytes and N-acetylglucosamine/mannose receptors on Kupffer cells. These data and inhibition studies by specified monosaccharide-terminal glycoprotein derivatives demonstrate the importance of these sugars in the uptake of this lysosomal enzyme by receptor-mediated endocytosis. Modification of the enzyme to expose certain monosaccharide moieties results in increased delivery to specific cell types. Therefore, naturally occurring receptors can be utilized for targeting glucocerebrosidase to the non-parenchymal cell in liver.

The N-asparagine-linked oligosaccharide chains of homogeneous human placental beta-glucocerebrosidase were released by hydrazinolysis, and their structures were analyzed. The sequence of sugars, linkage, and anomeric configuration of the glycosidic bonds were determined. Approximately 20% of the released sugar chain was of the typical high mannose type. The balance was of complex type with biantennary and triantennary structures. As a major component, the following trisialylated oligosaccharide was detected: (formula; see text).

At this point in time, enzyme replacement therapy for Gaucher disease appears both biochemically effective (patients 1-3, and 10) as well as clinically promising (patients 4 and 5), although these salutary responses have not been obtained consistently (patients 6-9). The discrepancy may be due to the method employed for the isolation of the enzyme. One current opinion is that glucocerebrosidase prepared by conventional fractionation and ion exchange chromatographic procedures may have been more effective in vivo than enzyme prepared by butanol extraction and hydrophobic column chromatography, which we developed because of the difficulty in obtaining large quantities of glucocerebrosidase by the conventional method. It should be remembered that the enzyme prepared by the latter procedure was fully active on glucocerebroside in liver biopsy specimens in vitro. Furthermore, glucocerebrosidase prepared by the second method exerted a positive effect in 2 young patients, who received relatively large amounts of enzyme, and in an adult who was treated with corticosteroid prior to infusion of the enzyme. At this moment, a possible explanation is that a necessary associated factor, perhaps a lipid, may have been removed by the large-scale isolation procedure. We are attempting to improve the biochemical and clinical responses to enzyme infusion by investigating the effects of lysosomal modifying agents and by enhancing the uptake and localization of the infused enzyme in lysosomes of cells that contain the stored glucocerebroside.

The present studies were undertaken to evaluate the possibility that hexosaminidase A, the enzyme deficient in Tay--Sachs disease, could be effectively delivered to brain. Previous studies from our laboratory have shown that hypertonic mannitol can be used to osmotically produce reversible disruption of the blood--brain barrier in animals (rat and dog) and man without significant neurotoxicity and that such barrier modification significantly increases the delivery of cytoreductive chemotherapy agents to selected areas of brain. By using the rat model of blood--brain barrier modification and radiolabeled enzyme, increased hexosaminidase A delivery to brain has been demonstrated in more than 85 animals. The time of injection of hexosaminidase A after blood--brain barrier disruption is critical for maximum delivery. Rapid (over 30 sec) intra-arterial administration of hexosaminidase A immediately after blood--brain barrier disruption resulted in a marked increase in enzyme delivery to the brain when compared with controls without prior barrier disruption. When the enzyme was administered 15-20 min after barrier disruption, approximately 50% less hexosaminidase A was delivered; when given 60-120 min after barrier modification, the amount delivered was the same as in control animals. This critical time course is very different than that seen in trials of low molecular weight chemotherapeutic agents (methotrexate and adriamycin). These preliminary studies suggest that hexosaminidase A can be delivered to the brain by blood--brain barrier modification and may be indicative of the potential for enzyme replacement in patients who hae Tay--Sachs disease.