Propionyl CoA—-y^-—Methylmaionyl CoA -> Succinyl Co A
Propiony l-Co A Methy I ma lony I-Co A
Carboxylase Mutase v
Figure 6 Enzymatic deficiencies in methylmalonic aciduria or propionic aciduria. TCA, tricarboxylic acid cycle.
etary therapy, they continue to have metabolic crises that can be life-threatening. Given the inadequacy of dietary therapy, a gene therapy approach is needed.
Both disorders are caused by enzyme deficiencies in the metabolism of 3-carbon species that are generated from the catabolism of amino acids and other metabolites (Fig. 6). Methylmalonic aciduria is caused by a deficiency of meth-ylmalonyl-CoA carboxylase activity that is a result of a defect either in the apoenzyme or in the active form of vitamin B12. Patients with the latter defect often respond well to treatment with large amounts of vitamin B12 and are therefore in less need of gene therapy.
The prominent target tissue for both disorders is presumed to be the liver. Hepatorenal transplantation has been successfully employed in a patient with a severe form of methylmalo-nic aciduria. Nonetheless, the major pathology associated with these organic acidurias is due to circulating toxic metabolites. The associated enzymes are normally expressed in many tissues including leukocytes, muscle, and fibroblasts. Therefore, these heterologous tissues should be explored in gene therapy preclinical studies. Unfortunately, animal models for these disorders do not exist at the present time.
Five percent of normal enzymatic activity in either propi-onic aciduria or methylmalonic aciduria is associated with a benign clinical course, indicating that this level of expression in a gene therapy should be sufficient to realize a large clinical benefit. This level of expression may have to be distributed over approximately 5% of the cells because overexpression of the relevant genes may not lead to a proportional increase in metabolic flux through the 3-carbon pathway. For example, in propionic aciduria, the propionyl-CoA carboxylase has 2
different subunits, a and p. In patients with a defect in the p subunit, overexpression would require gene transfer with both subunits. However, in patients with a defect in the a subunit, overexpression of the a subunit may be sufficient because the p subunit is produced in a 5-fold excess over that of the a subunit.
The common feature of lysosomal storage diseases is the inappropriate accumulation of normal cellular components within lysosomes. This storage of material is visible in cells by light microscopy as very large lysosomes that displace a large part of the cytoplasm. This class of disorders is caused by deficiency of specific lysosomal enzymes that are required for the degradation and recycling of glycoproteins and other cellular components. Without a specific degradative enzyme, the substrate for the reaction accumulates and cannot be removed from the lysosome. The clinical phenotype of each disease is dependent on the tissue type most affected by storage and by the accumulation rate. Physical findings that are suggestive of lysosomal storage include enlargement of the liver and spleen, anemia and thrombocytopenia due to replacement of normal bone marrow by stored material, destruction of bone, and for those enzymatic deficiencies that cause lysosomal storage in neurons, severe developmental regression, seizures, and other neurological symptoms. Not all of these problems are present in all lysosomal storage diseases; each different enzymatic deficiency presents with a specific phenotypic complex.
The challenges of gene therapy for lysosomal storage diseases are illustrated by the results of contemporary treatment with enzyme-replacement therapy or bone marrow transplantation. A major challenge to treating a lysosomal storage disease with gene therapy (in contrast to treatment of a liver enzymopathy) is the necessity of reversing lysosomal storage in multiple separate tissues. No currently available gene transfer technique is capable of delivering DNA to multiple target tissues efficiently. However, many lines of evidence demonstrate that lysosomal enzyme proteins can be produced in isolated tissues or even purified ex vivo and effectively delivered to most target tissues. For example, glucocerebrosidase, the enzyme deficient in Gaucher's disease can be produced in vitro using standard recombinant techniques, chemically modified to facilitate lysosomal targeting, and delivered to affected organs by simple intravenous infusion. This therapy if repeated periodically dramatically reduces liver and spleen size, corrects anemia and thrombocytopenia, and possibly prevents bone deterioration, all major debilitating features of Gaucher's disease. So, at least for gene therapy of Gaucher's disease, the enzyme would not need to be locally produced in all affected tissues. The enzyme could potentially be produced in a single target tissue, secreted into the circulation, and taken up by other diseased tissues. The major limitation of this approach is the difficulty of engineering a secreted form of the enzyme that would be efficiently taken up by other cells and incorporated into lysosomes.
Alternatively, the enzyme could be transferred from the site of production to diseased tissues via circulating blood cells. Seminal experiments demonstrated that functional lyso-somal enzymes may be transferred directly from a normal cell to an enzyme-deficient cell in tissue culture. Bone marrow transplantation in the treatment of lysosomal storage diseases exploits this phenomenon. Replacement of enzyme-deficient host bone marrow with enzyme-sufficient donor bone marrow yields a population of circulating blood cells of the reticuloen-dothelial lineage that infiltrates tissues and transfers lysosomal enzyme to the native cells. Bone marrow transplantation has been employed successfully in Gaucher's disease and in select other storage diseases that do not exhibit brain involvement. Apparently either insufficient numbers of corrected cells penetrate the central nervous system or insufficient enzyme is transferred to neurons to successfully ameliorate the neurological phenotype of many lysosomal storage disorders. Presumably, difficulties with correcting enzyme deficiency in the brain will also be a major obstacle to successful gene therapy.
Gene therapy for lysosomal storage diseases has to date focused on gene transfer into bone marrow stem cells for the purpose of supplying enzyme via circulating reticuloendothe-lial cells (6). Enzymatic correction of Gaucher bone marrow cells in culture has been accomplished with recombinant ret-roviral vectors. Similar experiments using other lysosomal enzymes in both cultured bone marrow and fibroblasts have been successful. Persistent production of enzyme in circulating blood cells has been demonstrated in rodents. Phenotypic improvement following retroviral-mediated gene transfer into bone marrow has been shown in gusmpslgusmps mice, a p-glucuronidase-deficient mouse model of human mucopoly-saccharidosis type VII. As expected, enzymatic correction of bone marrow resulted in amelioration of the somatic symptoms but did not arrest progressive neurological deterioration in this model. However, lysosomal storage in the brain did decrease in mice that had received intracerebral p-glucuroni-dase-expressing fibroblast implants. Clinical trials of ret-roviral-mediated bone marrow stem cell-directed gene therapy are underway in humans with Gaucher's disease and in patients with Hunter's syndrome (mucopolysaccharidosis type II) who have little central nervous system involvement.
This X-linked syndrome is caused by a deficiency in hypoxan-thine phosphoribosyl transferase (HPRT), an enzyme required for salvaging purines (Fig. 7). It is characterized clinically by increased blood and urine uric acid, mental retardation, choreoathetoid movements, and, most extraordinarily, self-mutilation. It is not understood how a deficiency in HPRT leads to these remarkable neurological sequelae. A genetic mouse model completely lacking HPRT activity does not exhibit any neurological dysfunction except when stressed with amphetamine administration or inhibition of adenine phos-phoribosyl transferase (APRT) with 9-ethyladenine. The cho-reoathetoid movement disorder, however, is postulated to be due to dysfunction within the basal ganglion secondary to disturbed dopamine metabolism.
Although the hyperuric acidemia and its sequela can be controlled with allopurinol, the absence of treatment for the neurological symptoms has prompted the search for gene therapy approaches. Historically, Lesch-Nyhan syndrome has played an important role in the development of gene therapy. One of the first demonstrations of the ability of retroviral vectors to correct a genetic mutation was done using the human HPRT gene. The first animal experiment in which a foreign gene was expressed in the brain was done by intracere-brally transplanting fibroblasts genetically modified to express the human HPRT (7). Although HPRT is expressed in all cells, its high levels in the basal ganglia suggest that this area of the brain should be targeted for gene transfer. Prevention of the mental retardation may require more global expression within the brain.
The amount of normal gene expression required to effect relief can be extrapolated from clinical experience. Although it was previously believed that the severity of the syndrome was not correlated with residual enzymatic activity, it is now realized that its severity does correlate with the amount of HPRT activity in whole cells. Patients with 1.6% to 8% of normal activity had choreoathetosis but not mental retardation or self-mutilation.
In summary, this syndrome is an example of a genetic disorder in which therapy is lacking even when so much is known about its genetic and molecular basis. The development of effective gene transfer methods into the brain may not only provide a therapy but will be quite revealing about its patho-genesis.
Gene therapy has the potential to significantly improve the clinical status of patients with familial hypercholesterolemia (FH), which is caused by a defect in the LDL receptor (LDLR) (Fig. 8). Deficiency in this receptor leads to reduced clearance of LDL by the liver and higher blood levels of LDL. In addition, affected individuals synthesize more cholesterol because the inhibitory effect of LDL on cholesterol synthesis is lost. This inhibition results from decreased HMG CoA reductase activity, the rate-limiting step in cholesterol synthesis.
Heterozygotes with LDLR deficiency occur at a frequency of 1:500 (as common as insulin-dependent diabetes mellitus), making it one of the most common genetic disorders in the United States, Europe, and Japan. Such patients have a 2-fold elevation in plasma cholesterol levels (300-600 mg/dL) and may develop coronary artery disease by the fourth decade of life. Three to 6% of survivors of myocardial infarctions are heterozygotes for FH.
Homozygotes with LDLR deficiency occur much more infrequently (1 in a million), but have much higher cholesterol levels (600-1000 mg/dL) and invariably die from coronary artery disease in their 20s. The severity of the sequelae is attenuated in the homozygotes by a few percent residual LDLR activity. Deaths were much less frequent in those homozygotes who had at least 10% of normal LDLR activity. This indicates that clinical benefit could be achieved by a gene
therapy in which only a small percentage of LDLR activity is restored. Furthermore, the severity of this disorder increases the benefit-to-risk ratio of clinical trials and thereby facilitates them. A gene therapy protocol can be first tested in the homozygotes (aided by Orphan Drug Status) and then extended to the more common heterozygotes.
Liver transplantation in children has proven that correction of the LDLR defect in the liver can normalize cholesterol levels. For this reason, gene therapy techniques for FH have been directed at the hepatocyte. Based on preclinical studies in mouse and rabbit LDLR-deficient models, ex vivo gene therapy in 5 homozygous FH patients using retrovirus-mediated LDL receptor gene transfer was performed. This technically challenging protocol yielded a highly variable metabolic response with some improvement in only 1 of the patients (8). This study indicates that important modifications must be made to the ex vivo gene transfer method before gene therapy can be used as a general therapeutic procedure for such patients (9).
Given the borderline results of the human clinical trial, efforts were initiated with adenoviral vectors carrying the LDLR gene. In vivo adenovirus-mediated transfer of the LDLR was shown to be highly effective in reversing the hy-percholesterolemia in LDLR knockout mice and WHHL rabbits (10). The important limitation of adenoviral-mediated gene transfer remains the transient expression in vivo after infection of somatic cells with recombinant adenovirus. Nonetheless, these studies demonstrate the proof of principle for the gene therapy of FH by the transfer of the normal LDLR gene and highlight the inadequacies of current gene transfer methods.
Current therapy for hypercholesterolemia (not limited to homozygotic FH) includes the use of HMG-CoA reductase inhibitors, which work by secondarily inducing expression of the LDLR, thereby lowering plasma LDL levels. These agents not only lower serum cholesterol, but also lower all-cause mortality by at least 30% in men and women who have coronary disease and total cholesterol levels of 215 to 300 mg/ dL. However, 2% of patients suffer liver toxicity and 0.2% develop muscle disease requiring cessation of drug administration. These drugs have to be taken once or twice every day for extended periods of time, and compliance is often difficult. A gene therapeutic agent that is administered less than every month (even by intravenous injection) would offer substantial benefit to the patient.
At high efficiencies of liver gene transfer, LDLR gene transfer into the liver could be used to prevent coronary artery disease in the general population. Taking into account all types of hypercholesterolemias, the third National Health and Nutrition Examination Survey (NHANES III) concluded that lipid-lowering therapy was required for 29% of Americans over 20 years of age (11). The Cholesterol and Recurrent Events study showed that patients with coronary artery disease but having ''normal'' LDL cholesterol levels benefited from treatment with a single statin therapy (12). The positive correlation between LDL levels and coronary artery disease is a continuum. In addition, overexpression of the normal LDL receptor in the liver of transgenic mice (4 to 5 times that of the endogenous receptor) prevented diet-induced hypercholesterolemia, suggesting that unregulated overexpression of the LDLR by liver gene therapy would be therapeutic in humans with hypercho-lesterolemia of various causes (13).
Many individuals develop coronary artery disease from other causes not amenable to statin therapy but that are potentially treatable by gene therapy. Liver gene therapy using the apoB mRNA editing enzyme (Apobec 1) or the VLDL receptor genes could modify LDL cholesterol levels. Other lipopro-tein factors besides LDL cholesterol levels influence the onset of coronary artery disease and are amenable to modulation by liver gene transfer. Additional expression of apoA-I in the liver by foreign gene transfer could raise high-density lipopro-tein levels and prevent atherosclerosis, as has been demonstrated in mouse and rabbit models. Hypertension, a predisposing factor for coronary artery disease, could be treated by delivering the kalikrein gene to the liver.
Was this article helpful?