Abstract
Abstract. Erythropoietin (Epo) is a glycoprotein hormone produced by genetic engineering. Many pathologic conditions could benefit from its administration, such as chronic renal failure or hemoglobinopathies. Epo secretion from genetically modified tissued could be proposed to patients only if the protocol is low cost and low risk. For that purpose, retroviral vectors and adeno-associated vectors expressing the Epo cDNA were developed. Gene transfer was performed into skeletal muscles. To avoid polycythemia, a tetracycline-regulated system was used to control the levels of protein secretion in vivo. β-thalassemias are among diseases that could benefit from an Epo gene transfer. β-thalassemias are attributable to deficient synthesis of β-globin and accumulation of unpaired α-chains. Stimulation of fetal globin synthesis is one strategy to correct the globin chain imbalance. There is evidence that Epo could play this role. In a mouse model of β-thalassemia, an adeno-associated vector expressing the Epo cDNA was injected intramuscularly. Epo was secreted continuously during at least 1 yr. Erythropoiesis was improved in those mice by increasing the synthesis of fetal hemoglobin.
Erythropoietin (Epo) is a glycoprotein hormone secreted by the kidney and, in some situations, by the liver. Its main role is to induce terminal differentiation of erythroid precursors into red blood cells (1,2). Increased secretion of Epo in kidney or liver tumors induces polycythemia, whereas decreased secretion of Epo in chronic renal failure induces severe anemia. Recombinant human Epo (rHuEpo) was the first hematopoietic growth factor produced by genetic engineering. rHuEpo is administered to patients with chronic renal failure. It induces erythropoietic activity, with increases in reticulocyte cell counts and hemoglobin concentrations. The extension of rHuEpo treatment to other pathologic conditions, such as anemia associated with AIDS, cancer, inflammatory origins, self-transfusions, or hemoglobinopathies, is limited by costs. The future use of Epo for such patients requires either a reduction of production costs or the development of alternative methods of administration. Epo secretion from genetically modified tissues could play a role in this context. We think that a gene therapy protocol should be proposed to patients only if gene transfer is performed as a low-cost, low-risk, simple protocol, allowing long-term secretion of the protein at controlled levels.
Gene Transfer Approaches for Epo Delivery
Studies conducted in our laboratory demonstrated that Epo secretion from primary fibroblasts embedded into neo-organs or from primary myoblasts genetically modified with retroviral vectors induced long-term, stable polycythemia in normal mice (3,4). Adenoviral vectors containing Epo cDNA were injected into mice (5,6) and primates (7). Very high levels of Epo secretion were measured, but expression was transient and was eventually followed by prolonged secretion at very low levels when the protein was autologous (8). The most interesting results were obtained with vectors derived from adeno-associated virus (AAV). Long-term secretion at high levels was measured after a single intramuscular injection of a recombinant AAV coding for Epo (9,10). Stable, long-lasting Epo secretion was also achieved in mice after naked DNA injection and muscle electroporation (11).
Regulation of Epo Secretion
With all of these gene transfer approaches, Epo secretion levels were high enough to be considered for the treatment of anemias (including human hemoglobinopathies), which require very high doses of Epo. Nevertheless, human applications require in vivo control of Epo secretion, to ensure biologic efficacy and avoid toxic effects. Four major systems are currently under development, i.e., those regulated by the antibiotic tetracycline, the insect steroid ecdysone or its analogues, the antiprogestin RU486, and chemical dimerizers (such as the immunosuppressant rapamycin and its analogues) (12).
In our work, we used the tetracycline system described by Bujard and co-workers (13,14). This system relies on a chimeric transactivator expressed under the control of a constitutive promoter. The transactivator specifically recognizes a tetracycline-inducible promoter, which controls the expression of the Epo cDNA. We introduced this system into retroviral vectors and performed ex vivo gene transfer into immortalized (15) or primary (16) myoblasts. For mice transplanted with primary transduced cells, iterative on/off switching of Epo secretion, depending on the administration of a tetracycline derivative (doxycycline) in the drinking water of the animals, was observed for 5 mo. Long-lasting regulation of Epo secretion was also observed for mice implanted with immunoprotective capsules containing allogeneic transduced fibroblasts (17). Importantly, there was no apparent immune response to the chimeric transactivator in these gene transfer approaches.
The system was also introduced into a single AAV vector (18) or into two separate AAV vectors (19). Long-term regulation of hematocrit levels and of Epo concentrations was observed in mice in both cases. Muscle electroporation of two plasmids expressing the two components of the tetracycline system allowed regulation of Epo secretion in vivo for at least 3 mo (11). Another system of regulation, using rapamycin, was introduced into two AAV vectors and allowed iterative regulation of Epo secretion for at least 6 mo in mice and 3 mo in rhesus monkeys (20).
Ideally, the delivery of adequate amounts of Epo for replacement therapy in Epo-dependent anemias would reproduce the physiologic regulation. This gene therapy approach is not currently available, but regulation by an exogenous inducer, such as tetracycline, is presumably acceptable for hemoglobinopathies, at least in an initial phase of human applications. In contrast, a gene transfer protocol can be considered for applications in other Epo-responsive anemias only if Epo secretion is regulated by its physiologic stimulus, namely hypoxia.
Epo Delivery for β-Thalassemia
Thalassemias are the most common monogenic diseases. They are very prevalent in the Mediterranean area and in southeast Asian countries, where there are presumably more than 1 million severely affected individuals. Because of rapid population increases and decreases in childhood mortality rates in these areas, thalassemias are likely to present a severe world health problem in the next few years.
β-Thalassemias are attributable to deficient synthesis of β-globin and the accumulation of unpaired α-chains. Severe forms of the disease are characterized by a complete absence of β-globin synthesis. Genotypes responsible for less severe globin chain imbalances result in usually milder but very heterogeneous phenotypes (thalassemia intermedia). Clinically, the homozygous βo form is responsible for Cooley's disease, which is associated with microcytic anemia, hypochromia, hemolysis, iron overload, and an enlarged spleen.
Unmatched α-chains form insoluble complexes that precipitate in erythroblasts, releasing iron (21). Binding of denatured α-globin chains to membranes and redistribution of cellular iron are responsible for alterations in membrane lipids and proteins through oxidative mechanisms, causing the loss of membrane function and contributing to premature cell destruction (22,23). Bone marrow hemolysis and decreased survival of adult erythrocytes in the peripheral blood induce anemia, which stimulates ineffective erythropoiesis. Potential therapeutic approaches include correction of β-globin chain synthesis by bone marrow transplantation or gene therapy and reduction of the levels of unpaired α-globin chains by stimulation of fetal globin chain synthesis.
β-Globin gene transfer in hematopoietic progenitors was performed with retroviral vectors in mice as soon as the first experiments in gene therapy were performed (24,25,26,27). However, after > 12 yr, it still appears extremely difficult to obtain suitable expression of β-globin in genetically modified erythroid cells (28,29,30), although the strategies represent very promising approaches.
Reactivation of fetal hemoglobin (HbF) synthesis would be a reasonable approach to improving globin chain balance. However, trials conducted with 5-azacytidine, hydroxyurea, butyrate compounds, and combination therapies demonstrated few clinical effects, although HbF synthesis was slightly induced (31,32,33,34).
Experiments in anemic and nonanemic baboons suggested that HbF production was markedly increased after the administration of high doses of rHuEpo (35,36). Therefore, trials with rHuEpo were conducted in some patients with β-thalassemia intermedia (37,38,39) or with β-thalassemia major (40,41). Those trials led to the following conclusions: (1) high doses of rHuEpo did not significantly increase HbF levels in patients, perhaps with the exception of combination therapies with hydroxyurea (42); (2) clinical benefits were nevertheless observed for some patients, although positive responses to rHuEpo could not be predicted and the mechanisms responsible for improved erythropoiesis remain unclear; and (3) potential drawbacks of the treatment are the costs, which will not allow it to be used worldwide, and the risk of worsening bone marrow expansion and bone disease.
Epo Gene Transfer in a Mouse Model of β-Thalassemia
Homozygous β-thalassemic mice exhibit microcytic hypochromic anemia, highly dysmorphic red blood cells, extensive intramedullary and intrasplenic destruction of erythroid progenitors, and erythroid hyperplasia (43). Similar abnormalities in membrane functions have been documented in murine and human β-thalassemias (23). Although β-minor globin is expressed at significant levels in adult mice, expression is much lower than in fetuses (44,45). Similarly to primate HbF, the synthesis of β-minor was increased in response to in vitro culture with erythroid progenitors (46), under stress erythropoiesis conditions (47), after hydroxyurea treatment (48), and transiently in mice injected with high doses of rHuEpo (49).
The effectiveness of Epo delivery from genetically modified hematopoietic stem cells (50) or encapsulated myoblasts (51) was examined in a mouse model of β-thalassemia. A transient correction of anemia, with improved red blood cell phenotype, was observed in both cases.
In our studies (52), we demonstrated that a recombinant AAV vector containing murine Epo cDNA, under the control of a constitutive cytomegalovirus promoter, could induce robust Epo secretion from engineered skeletal muscles of β-thalassemic mice. After 1 yr of follow-up monitoring, 12 treated mice demonstrated dramatic stable improvement of erythropoiesis. Correction of anemia was associated with improved red blood cell morphologic features, decreased amounts of α-globin chains bound to erythrocyte membranes, and increased β-minor chain synthesis. More effective erythropoiesis probably accounted for a reduction in erythroid cell proliferation, as indicated by decreased proportions of circulating reticulocytes and by the reduction of 59Fe incorporation into erythroid tissues. We determined that Epo concentrations adequate for maintaining normocythemia and reducing erythroid hyperplasia were in the range of 250 to 350 mU/ml. Higher concentrations resulted in the expansion of phenotypically improved red blood cells, with subsequent polycythemia. Therefore, the study highlighted the necessity for control of Epo secretion in engineered cells.
The significant increase in β-minor chain synthesis and the subsequent reduction in the levels of unpaired α-globin chains incorporated into membranes probably played important roles in improving the red blood cell phenotype. These effects were likely the result of the action of Epo at the level of primitive erythroid progenitors. The hypothesis is that, in β-thalassemic mice with strong Epo stimulation, the preferential mobilization of erythroid blast-forming units programmed for β-minor synthesis may account for the observed emergence of effective erythropoiesis, which slows erythroid cell proliferation, whereas the expansion of late erythroid progenitors, which would aggravate dyserythropoiesis, remains absent or limited.
- © 2000 American Society of Nephrology
References
