Mario C. Rattazzi

Excretion-Reuptake Route of β-Hexosaminidase in Normal and I-Cell Disease Cultured Fibroblasts

It has been proposed that in cultured fibroblasts the final packaging of enzymes in lysosomes requires excretion followed by pinocytosis by neighboring cells via a carbohydrate-specific receptor mechanism. It has also been proposed that the abnormally high activity of lysosomal enzymes in the medium of cultured fibroblasts from patients with I-cell disease (mucolipidosis II) results from an altered carbohydrate recognition residue on the enzymes which prevents reuptake into the cells. With beta-hexosaminidase as a marker, and competitive inhibition of uptake by 2 mM mannose-6-phosphate, we have determined that only 12% of the total (intra- and extracellular) beta-hexosaminidase in normal fibroblasts is channeled through the excretion-reuptake route. After 9 d of exposure to mannose-6-phosphate, normal fibroblast cultures accumulated in the medium only a fraction of the enzyme excreted by I-cell disease fibroblasts in the same period. Furthermore, this minimal loss of enzyme to the medium did not result in a decrease of intracellular enzyme activity. Finally, if the defect in I-cell disease were only because of an impairment of a reuptake mechanism that involves only 12% of the total enzyme, then 88% of the newly synthesized enzyme should be retained by I-cell fibroblasts, resulting in intracellular activity three to nine times higher than that which is observed. These data are consistent with our previous proposal that excessive lysosomal enzyme activity in the medium of I-cell disease fibroblasts results from preferential exocytosis.

Genetic Polymorphisms of Human Mitochondrial Glutamic Oxaloacetic Transaminase

In a survey of 860 unselected human placental extracts, three variants of mitochondrial glutamic oxaloacetic transaminase were found, all of which were common enough to be considered polymorphisms. Family studies showed that this enzyme is under the control of nuclear rather than mitochondrial DNA.

Prenatal diagnosis of metachromatic leukodystrophy by electrophoretic and immunologic techniques.

Summary: Electrophoretic examination of extracts of cultured amniotic fluid cells from a pregnancy at risk for metachromatic leukodystrophy (MLD) showed absence of arylsulfatase A (AS-A) activity. Immunodiffusion with anti-human AS-A immune serum failed to show enzymatically active arcs of immune precipitate. Electrophoretic studies and quantitative assay of extracts of organs from the aborted fetus confirmed the diagnosis of MLD.Electrophoresis of amniotic fluid from this and one additional fetus with MLD showed an arylsulfatase pattern qualitatively and quantitatively indistinguishable from normal. In both normal and MLD fluids, the AS-A band was replaced by a band with lower anodal mobility. Only the anodal band of normal amniotic fluid, however, reacted with the anti-AS-A immune serum in immunoelectrophoresis. Assay of amniotic fluid with p-nitrocatechol sulfate (PNCS) as a substrate showed marked deficiency of “AS-A” activity in the fluids from the two MLD pregnancies.An optimal procedure for prenatal detection of MLD should include electrophoresis of extracts of cultured amniotic fluid cells with visual demonstration of absence of AS-A activity. Immunologic techniques applied to cell-free amniotic fluid may be of help in the rapid identification of the fetal genotypes.Speculation: Qualitative methods such as electrophoresis and immunodiffusion allow reliable prenatal diagnosis of metachromatic leukodustrophy using cultured amniotic fluid cells. The peculiaritites of arylsulfatase in amniotic fluid render this material less suitable for prenatal diagnosis, but may provide an insight in the structural and functional defect of arylsulfatase A in this disease.

Toward enzyme therapy in Gm2 gangliosidosis: beta-hexosaminidase infusion in normal cats.

Summary: As a first step toward the use of cats with Gm2 gangliosidosis as models for enzyme replacement therapy, we studied plasma clearance, organ disposition, and subcellular localization of human β-hexosaminidase in normal cats. Plasma half-life of placental β-hexosaminidase at low doses was 3–4 min; both Hex A and Hex B were cleared at approximately the same rate. The half-life of human plasma β-Miexosaminidase in contrast, was >60 min. Clearance curves at higher doses approached zero order kinetics, suggesting the existence of a saturable clearance mechanism. Injection of periodate-treated placental β-Miexosaminidase resulted in a plasma half-life of ∼50 min, strongly suggesting that rapid clearance of both Hex A and Hex B was mediated by carbohydrate-specific mechanisms. Circulatory bypass of liver resulted in plasma half-life of the enzyme of ∼60 min, indicating that the liver was the main clearing organ. As both main feline β-Miexosaminidase isozymes did not crossreact with antihuman β-Miexosaminidase immune sera, independent evidence of preferential hepatic uptake was obtained by immunofixation clectrophoresis; immunoelectrophoresis, and immunotitration. The human enzyme detected in liver accounted for ∼80% of the injected dose; small amounts of exogenous enzyme were detected in spleen and kidney. Subcellular fractionation of liver showed that human Hex A and Hex B had entered the lysosomal-vacuolar apparatus of hepatic cells.Speculation: Therapeutic applications of lysosomal enzyme replacement in patients with storage diseases are fraught with difficulties. Normal cats infused with human β-Miexosaminidase can be used to develop enzyme replacement methodologies to be tested in cats with genetic Gm2 gangliosidosis. This unique animal model makes it possible to explore in vivo rational approaches to therapeutic intervention in human patients.

Immunoaffinity Chromatography of Human β-Hexosaminidase A

A highly specific method for the purification of human beta-hexosaminidase A employing immunoaffinity chromatography is described. Using an antiserum against the unique antigenic determinant, alpha, of beta-hexosaminidase A, and elution with 8.0M urea, a 283-or 417-fold purification of the enzyme was obtained in a single step from crude human placental homogenate.

Delivery of Compacted DNA Across the Blood-Brain Barrier |[diams]| 732

Delivery of DNA to the central nervous system via the cerebral circulation for a possible treatment of disorders affecting the entire brain is prevented by the blood-brain barrier (BBB). Although BBB disruption (BBBD) can be obtained by intracarotid infusion of hyperosmolar solutions, the size of the resulting interendothelial gaps does not allow extravasation of cDNA transgenes. We have assessed brain delivery of DNA complexed with the DNA-compacting, cationic proteins poly-L-lysine (PLL) and salmon sperm protamine (SSP) after osmotic BBBD in 26 rats, including 7 controls. The 5kb plasmid, pGL encoding A. victoria green fluorescent protein (GFP) was reacted with PLL or SSP under carefully controlled conditions to obtain DNA-protein complexes 10-20 nm in diameter, as assessed by electron microscopy. The DNA suspensions were infused into the common carotid artery of rats after unilateral BBBD. This was obtained by ligation of the ipsilateral occipital and pterigopalatine arteries, temporary clamping of the external and common carotid arteries distal and proximal to an indwelling catheter, and infusion of 1.7 molal arabinose (0.1 ml in 10 s). The presence of plasmid DNA in homogenates of brain hemispheres and cerebellum from 16 animals sacrificed and saline-perfused at 2 hours was assessed by the polymerase chain reaction(PCR). PCR product was detected in 15 of these 16 animals, usually with an intensity corresponding to the degree of BBB permeabilization, as judged by Evans blue-albumin (EBA) extravasation. Separation of DNA from endothelial cells and plasma membranes by a modified capillary depletion method supported the contention that most of the plasmid DNA was present in the brain parenchyma. In fact, little or no PCR product was detected in the brain of 4 animals infused with compacted DNA in the absence of BBBD, indicating minimal binding or internalization of DNA-protein complexes by vascular endothelial cells. Predictably, no PCR product was detectable after BBBD and infusion of uncomplexed DNA. Neural cell transduction by PLL-cDNA resulting in GFP expression was assessed by fluorescence- and confocal microscopy in 2 rats sacrificed 5 days after treatment. Paravascular and parenchymal cells exhibiting green cytoplasmic fluorescence were visible in diverse brain regions, usually coincident with red-fluorescing, extravasated EBA. These initial data indicate that transduction of neural cells can be obtained by delivery of compacted DNA-protein complexes to the brain parenchyma across a reversibly permeabilized BBB.

752 ENZYME REPLACEMENT IN FELINE GM 2 GANGLIOSIDOSIS: CATABOLIC EFFECTS OF HUMAN β-HEXOSAMINIDASE A (HEX A)

To assess the feasibility of enzyme therapy in human Gm2 gangliosidosis, hampered by hepatic uptake of, and blood-brain barrier (BBB) impermeability to exogenous Hex A, kittens with Gm2 gangliosidosis (a model for human Sandhoff disease) were injected IV or intracarotid (IC) with ∼5 mg purified human placental Hex A. Hepatic uptake was reduced with IV mannan; reversible BBB permeability compatible with survival without gross neurologic sequelae was induced by 1 ml oxygen IC. At 12 to 72 hrs, residual exogenous enzyme activity in liver, spleen, kidney and brain cortex was 100-50%, 14-5%, 22-5% and 50-10%, respectively, of normal endogenous activity. TLC quantitation showed time-and dose-dependent reduction of liver GL4 globoside and Gm2 ganglioside to 20% of affected controls, and increase of Gm3 ganglioside. A reduction of GL4 globoside to 65% of controls was observed in spleen and kidney. These results demonstrate for the first time a catabolic effect of Hex A in vivo at organ level, even at the relatively low extrahepatic levels obtained by hepatic uptake depression. Although oxygen-induced BBB permeability allowed delivery of comparable enzyme activity to the CNS, no effects on Gm2- or Gm3 ganglioside were evident in brain. Higher enzyme doses, longer exposure, and clarification of neuronal uptake specificity are needed to assess catabolic effects in CNS, a prerequisite for therapeutic attempts in humans.

556 ENZYME THERAPY IN GM2 GANGLIOSIDOSIS: CARBOHYDRATE-SPECIFIC HUMAN β-HEXOSAMINIDASE UPTAKE BY FELINE LIVER

In our studies on cats with genetic Gm2 gangliosidosis as models for enzyme therapy, we characterized a mechanism responsible for rapid plasma clearance (t½ ∼ 3 min) of human β-hexosaminidase (β-hex)injected into normal cats. Preferential hepatic uptake was shown by recovery studies; circulatory bypass of liver markedly impaired enzyme clearance (t3/2 ∼ 60 min). Carbohydrate-specific uptake, suggested by slow clearance of periodatetreated β-hex (t½ ∼ 60 min) was confirmed by impaired clearance of normal β-hex obtained by injection of terminal N-Acetyl glucosamine- and mannose-rich glycoproteins. Ovomucoid, ovalbumin and ribonuclease-B (final plasma concentration, fpc 0.1, 0.4 and 0.3 mM, resp.) markedly inhibited clearance (t½ ∼ 60 min). Terminal galactose-rich,desialylated orosomucoid and fetuin had no significant effect. Marked inhibition of β-hex clearance (t½ ∼ 60 min) was also obtained by injection of mannose, N-Acetyl glucosamine and L-fucose (fpc 0.15, 0.7, and 0.7M, resp.), but not of glucose or galactose (fpc 0.7M). Thus the feline liver receptor involved in β-hex clearance recognizes terminal mannosyl-and N-Acetyl glucosaminyl-, but not galactosyl residues on glycoproteins. A hepatic receptor with similar specificity, also clearing exogenous lysosomal enzymes, has recently been described in the rat. This suggests that the same mechanism may be present in other mammalian species including man, and may be of critical importance in lysosomal enzyme replacement in humans.

THE CAT AS A MODEL FOB ENZYME THERAPY IN GM 2 GANGLIOSIDOSIS

A genetic neurodegenerative disorder with lysosomal storage of Gm2 ganglioside, its asialo derivative, and globoside, and β-hexosaminidase deficiency has been described in domestic cats (Cork et al, Science, 1977, in press). To strengthen the apparent analogy with human Gm2 gangliosidosis type II, feline β-hexosaminidase has been studied. Two major forms of this lysosomal enzyme, Hex A and Hex B, are present in organs and body fluids from normal animals, with biochemical and kinetic properties remarkably similar to those of the corresponding human enzymes. These data, and the deficiency of both components in affected cats, support the analogy between feline and human diseases. Thus it is possible to utilize this animal model for enzyme therapy experiments. Structural differences between human and feline β-hexosaminidase, however, may result in different cellular uptake, endocellular activity and stability, limiting the relevance of replacement experiments with feline enzyme. Thus, plasma clearance and organ disposition of human placental Hex A and Hex B has been studied in normal cats, utilizing the apparent absence of interspecific immunologic cross-reactivity. The results so far indicate a rapid, preferential hepatic uptake, possibly via a saturable receptor mechanism, with recovery of the exogenous enzymes in a lysosomal fraction. Further studies using human enzymes should result in enzyme replacement strategies, to be tested in diseased cats, relevant to therapeutic attempts in human Gm2 gangliosidosis.