A. Types of IEM
One common type of IEM is caused by deficiency of an enzyme that catalyzes the conversion of one chemical to another (Fig. 2a). Deficiency of a specific enzyme can cause disease through 3 separate mechanisms: (1) excessive accumulation of substrate to toxic levels, (2) deficiency of an essential product, or (3) metabolism of the substrate through alternative biochemical pathways leading to toxic secondary metabolites. Examples of such IEM include phenylketonuria and meth-ylmalonic aciduria.
IEM can also be caused by deficiency of protein that is involved in the transport of metabolite (Fig. 2b). Examples include the cystine transporter in cystinosis and the LDL (low-density lipoprotein) receptor in familial hypercholesterolemia.
Other genes relevant to IEM are required for the proper formation of organelles (Fig. 2c). Neonatal adrenoleuko-dystrophy and Zellweger syndrome are caused by defects in genes that are required for the proper formation of peroxi-somes.
The pathogenesis of IEM can be explained by several models (Fig. 3). One major category includes IEM in which organ dysfunction occurs by a circulating toxic metabolite (Fig. 3a). Another major category is organ dysfunction resulting from a cell autonomous process (Fig. 3b). Although these concepts are useful in formulating gene therapy approaches, it should be appreciated that they are only models and that our understanding of the pathogenesis for many IEM is incomplete. In fact, gene therapy trials may provide decisive information concerning the mechanism by which the metabolic defect leads to the diseased state.
In this class of disorder, a metabolite accumulates in one tissue as result of an enzymatic deficiency (Fig. 3a). This leads to increased metabolite levels in the blood and toxicity in other tissues. The prototype for this type of disorder is phenylketon-uria in which deficiency of hepatic phenylalanine hydroxylase leads to increased blood levels of phenylalanine and toxic effects to the developing brain. Familial hypercholesterolemia is another IEM that fits this model. Deficiency of the LDL receptor in the liver leads to increased levels of LDL and subsequent damage to the coronary arteries.
A corollary of this model is intraorgan toxicity from a metabolite that accumulates in an extracellular space within the affected tissue. It is particularly applicable to the central nervous system (CNS). Some IEM associated with neurological dysfunction may be caused by a toxic metabolite that accumulates within the brain and circulates in the cerebral spinal fluid (CSF). Gene therapy could then be predicated on providing gene expression in any cell within the brain as long as the expressed enzyme could lower levels of the toxic metabolite in the CNS. The cerebral spinal fluid could provide the conduit for such exchange.
A metabolic defect in one tissue could also harm another tissue by decreasing the circulating level of a metabolite. For example, a defect in gluconeogenesis that occurs in the liver and muscle (e.g., glycogen storage disorder) can cause hypo-glycemia and damage to the brain.
In other IEM, the metabolic defect only leads to toxicity to the cell that has the metabolic deficiency (Fig. 3b). Cellular toxicity results from either increased or decreased levels of a metabolite within the affected cell. For these disorders, gene therapy would be effective only if the normal gene is targeted to the dysfunctional cell.
A variety of parameters of expression are important determinants of the ability of gene therapy to treat specific disorders. Some generalizations can be made concerning the expression requirements for IEM (Table 1). Most IEM are recessive conditions, and addition of a single gene copy is sufficient to correct the disease phenotype. In effect, gene addition converts the patient to a biochemical state analogous to that of a carrier. For those patients with single-point mutations, targeted gene correction using gene conversion or homologous recombination is a possible therapy, but gene correction is not necessary if a functional gene can be added. The obvious therapeutic gene to be added in IEM is the human gene that is defective in the disease state, but it is conceivable that a therapeutic effect could be achieved using another gene. For example, a gene from another species could metabolize a toxic metabolite by a different mechanism.
F. Requirements for Expression Persistence
For most IEM, gene expression does not have to be regulated and can be constant. Most genes involved in IEM are considered ''housekeeping'' genes. In contrast, in diabetes mellitus, insulin expression has to be regulated in response to blood glucose levels.
Given that IEM are chronic conditions, persistent expression is needed. It would be best if gene correction and therefore a ''cure'' could be done with 1 or few administrations. If expression cannot be persistent after 1 gene dose, then repetitive administrations are required. Repetitive administrations can be problematic for some vectors such as adenoviral vectors that induce neutralizing antibodies. Loss of expression from vectors can be a result of removal of the foreign DNA, promoter suppression, or rejection of the foreign gene product.
Immune effects can arise even if the gene product is intracellular because all parts of proteins are presented to the immune system via the MHC I complex. The important issue is whether, in the disease state, the patient expresses any residual native protein and is immunologically tolerant to the normal gene product. One measure of this is whether tissues from the patient exhibit cross-reactive material (CRM), protein that cross-reacts with antibodies against the native protein. This is best determined by performing immunoblot (Western blot) analysis. Even if protein is not present, native protein could have been produced but be unstable. Expression of the foreign gene in such a patient may not induce an immune effect because the protein is not recognized as foreign. Further experi-
Table 1 Generalizations Concerning Expression Requirements for Gene Therapy of IEM
Method of modification Therapeutic gene
Gene addition is sufficient Normal gene that is defective in patient Persistent Not needed >5% of normal levels Liver, CNS, blood cells, muscle, heterologous expression possible in some disorders ence is necessary to determine whether the immune system will prevent stable expression of the normal gene in patients with IEM.
The level of expression is a critical determinant for the success of a gene therapy. For most IEM, foreign gene expression only has to be greater than 5% of normal levels in order to attenuate the majority of the diseased state. This is based on clinical experience in which the percent of residual enzyme activity is correlated with the phenotype. In many IEM, people with more than 5% of normal enzymatic activity are free of symptoms. If enzymatic activity is between 1% and 5%, their clinical course is less severe than patients with 0% of enzymatic activity.
Although the total enzymatic activity is one measure, the percent of cells expressing the foreign gene may also be important. Overexpression in a few cells may not lead to a therapeutic effect if the expressed enzyme alone cannot completely produce the metabolic conversion. The protein deficient in the patient may be part of an enzymatic complex so that overexpression of 1 component would not necessarily lead to higher activity of the complete complex. Similarly, other enzymatic steps, cofactors, or transport of metabolites may limit the ability of the cell to perform the required metabolic conversion at a rate higher than the normal level. If so, the therapeutic gene has to be expressed in more than 5% of the target cells.
Details of the pathogenesis for the IEM need to be understood, and the target tissue has to be tailored for each disorder. For IEM that fit the ''circulating toxic metabolite'' model, the therapeutic gene does not necessarily have to be targeted to the tissue that normally expresses the affected gene (Fig. 4). Although correction of the deficient enzymatic activity in the affected organ would be most straightforward, expression within a heterologous tissue (different from that which normally expresses the enzyme) could clear the circulating toxic metabolite and attenuate the disease state. For this approach to be effective, the enzyme must be functional within the het-erologous tissue. Restrictions on enzymatic function can include requirements for protein subunits, cofactors, substrate, and clearance of product. Given the ability for several gene transfer systems (e.g., plasmid DNA, adenoviral vectors, and AAV vectors) to express foreign genes stably in muscle, it will be a useful tissue for many heterologous gene therapy approaches. Blood cells derived from genetically modified stem cells are another candidate tissue for heterologous gene expression if the problems associated with stable foreign expression are solved.
For IEM that fit the ''cell autonomous'' model, expression within the affected cell is generally required. The exception is for the lysosomal storage disorders in which the enzyme can be transferred from one cell to another.
For IEM that affect the brain, ''global'' gene expression throughout the brain may be required. Alternatively, specific neurological symptoms could be treated by targeting specific regions of the brain. For example, in Lesch-Nyhan syndrome, choreathetoid movements could be treated by targeting the basal ganglia.
I. Mitochondrial Disorders of Oxidative Phosphorylation
Several IEM are caused by defective oxidative phosphoryla-tion within the respiratory chain complex of mitochondria. A unique feature of mitochondria is that 13 of the more than 80 respiratory chain subunits are encoded within the mitochon-drial genome. The mitochondria also contain 22 transfer RNA (tRNA) and 2 ribosomal RNA that enable protein synthesis within the mitochondria. The remaining 70 or so respiratory chain subunits are encoded within the nuclear genome. These proteins are produced within the cytoplasm and contain an amino terminus that targets their entry into the mitochondria by interacting with a number of chaperone and transport proteins.
For disorders caused by mutations in the nuclear encoded respiratory chain genes, the gene therapy approaches described above are germane. However, disorders caused by mutations in the mitochondrial genome offer additional challenges for gene therapy. One approach would be to express the deficient subunit within the nucleus, regardless of its native mitochondrial origin. The subunit could be modified to contain an amino leader sequence to enable entry into the mitochondria.
The other approach of genetically modifying the mitochondrial genome is at an early conceptual stage. Toward this end, a peptide mitochondria-targeting sequence has been covalently attached to oligonucleotide to enable mitochondrial entry. The oligonucleotide could correct a point mutation by some type of gene conversion or recombination process. Point mutations occur in mitochondrial disorders such as Leber hereditary optic atrophy (LHON), myoclonic epilepsy and ragged-red fiber disease (MERRF), and mitochondrial encephalomyopa-thy, lactic acidosis, and strokelike episodes (MELAS). An alternative treatment approach would be the addition of functional tRNA genes to patients with mitochondrial disorders such as MERRF or MELAS that are caused by tRNA mutations. The treatment of mitochondrial DNA deletion diseases would require the delivery of larger DNA sequences (>5 kb), which would be more challenging. Deletions occur in disorders such as Kearns-Sayre syndrome. Another option would be to deliver normal mitochondria en toto.
Different mitochondria can proliferate in a tissue at different rates. This may explain why inborn and somatic (acquired) mitochondrial defects often present in later life. Any genetic modification of mitochondria must enable the corrected mitochondria to have a proliferation advantage over the abnormal mitochondria in order to achieve a permanent cure. A final challenge for mitochondrial disorders is that they often involve the nervous system, which is less accessible than other organs to therapeutic endeavors.
Gene therapy for IEM will have a significant impact on newborn screening programs and vice versa. Screening for IEM at birth enables gene therapy to be initiated prior to the onset of symptoms and any irreversible tissue damage and thereby increases the value of the gene therapy. Irreversible brain damage occurs in many IEM when a neonatal metabolic crisis is not prevented. For example, the extent of perinatal hyperam-monemia in a urea cycle defect, ornithine transcarbomylase deficiency (OTC), has been directly correlated with intelligence in later life.
One criterion for the initiation of newborn screening for a particular disorder is whether an effective treatment exists. The development of effective gene therapy for a disorder could satisfy this criterion. Another criterion is the availability of a reliable, inexpensive laboratory method for disease detection.
Currently, most states in the United States and many other nations are screening for phenylketonuria and galactosemia. Screening for maple syrup urine disease or homocysteinemia is less common. Tandem mass spectroscopy procedures are being developed for analyzing blood spots in amino acids and organic acids conjugated to carnitine (acylcarnitines) in order to detect many of the disorders in amino acid and fat metabolism and organic acidurias. Such comprehensive newborn screening programs developed in conjunction with new gene therapies will have a major impact on the morbidity and mortality of IEM.
Many IEM can be reliably diagnosed in the prenatal period. As intrauterine gene therapy approaches are developed, IEM will be good candidates for such approaches. One potential advantage of prenatal approaches may be a decreased chance of an immune recognition of the therapeutic gene product.
Was this article helpful?