New Paradigm of Gene Therapy: Skeletal-Muscle-Targeting Gene Therapy for Kidney Disease

The annual number of patients who were introduced to chronic dialysis has been increasing in Japan year by year and was over 30,000 in 1998 [1]. The total number of chronic hemodialysis patients exceeded 180,000 at the end of 1998 in Japan [2]. This phenomenon is not limited to Japan but occurs in the rest of the world. This fact suggests that we have no effective therapy against the progression of renal diseases. In order to halt the explosion of patient numbers with end-stage renal failure, we clearly need an innovative therapy for progressive renal diseases. In this view, we review an emerging gene therapy targeting skeletal muscle and discuss, in particular, the potential application to progressive renal diseases.By the end of 1998, 3,089 patients had been treated with gene therapy according to 367 protocols. However, 99% of the challenges seem to fail in treating disease. The first human gene therapy for adenosine deaminase deficiency was not concluded to cure the disease by itself, although the first case, a child, was cured and safely returned to school [3]. Some protocols against malignancies may halt the progression of certain types of cancer, but it cannot be concluded that gene therapy can treat all cancers. It is natural that cancer, a life-threatening disease, should be managed by multimodality therapy since failure to induce remission culminates in the patient’s death. Only one approach performed by Baumgartner et al. [4], however, clearly succeeded in human gene therapy for critical limb ischemia caused by Burger’s disease and arteriosclerosis obliterans. Gene therapy using simple naked plasmid DNA and skeletal muscle can make a breakthrough by widening the application to non-life-threatening diseases.Why has human gene therapy failed so far? It should be pointed out that the application of human gene therapy has principally been limited to hereditary diseases and incurable diseases, e.g. life-threatening malignancies and HIV infection. The main reasons, however, are technical problems: imperfection of vectors, difficulties of gene delivery and shutdown of the expression of the gene successfully transferred. First, vectors available to date have definite faults including low expression of transgene and evocation of cytotoxicity by the inevitably and needlessly expressing viral proteins. In adenoviral vector, the cytotoxicity and immunogenicity of viral vectors and its products are obstacles as they cause tissue injury. Secondly, targeting a specific organ and ideally particular cells is still an unresolved major issue in the application of gene therapy. No selective gene transfer technique to particular cells in the kidney has been developed so far. In the kidney, more than 20 kinds of cells constitute the glomerulus, tubule and vasculature, and form the nephron as an isolated functional unit. In case of the approach from the renal artery, most of the cells are sequestered from the blood stream except for glomerular cells. Proximal tubules and Henle’s thin loop cells are located away from the pelvis when retrogradely approached from the ureter. Thus, the accessibility to the particular kidney cells seems not to be so easy anatomically. Finally, inactivation of transgene occurs in a few months after gene transfer. Instability of the transgene in vivo can be explained in part by shutdown of the gene expression of the vectors, which are not integrated into the genome. However, integrated retrovirus vector into the genome also terminates gene expression in a few months after transfection in vivo. It is suggested that an alteration of the chromatin structure may be the cause of the shutdown of transcriptional activity of the transgene. One possible explanation is that the methylation of cytosine residues in the DNA sequence may cause silencing of the transcription of the transgene. Recently, it is reported that methylation-dependent transcriptional silencing can be overcome by adding trichostatin A, which specifically inhibits histone deacetylase [5]. Histone deacetylation is guided to specific chromatin domains by underlying genomic methylation patterns, and methylation-dependent transcriptional repression, in part, relies on histone acetylation.Delivery of the heterologous genes into skeletal muscle has been challenged since 1990. Wolff et al. [6]firstly demonstrated that plasmids encoding reporter genes for chloramphenicol acetyltransferase, luciferase and β-galactosidase were separately injected into mouse skeletal muscle to estimate the efficiency of gene expression. The expression of the luciferase gene, which is driven by Raus sarcoma virus promoter, persisted for at least 2 months. Thereafter, various similar observations have been published using expression vectors encoding secreted proteins, e.g. erythropoietin [7], vascular epithelial growth factor (VEGF) [8]and endostatin [9]. Generally, quite an amount of naked plasmid is required for gene expression in skeletal muscle and its expression increases in a dose-dependent manner according to the administered plasmid. Regenerating muscle increases uptake of the plasmid DNA and improves the efficiency of transfection. Tokui et al. [10]reported that gene expression was increased with pretreatment of muscle with bupivacaine.The strength of transgene expression depends on the promoter activity in muscle cells. The activity of various promoters has been tested in skeletal muscle. Among them several strong and muscle-specific promoters have been reported. Niwa et al. [11]reported that CAG promoter, containing cytomegalovirus (CMV) enhancer and chicken β-actin promoter, provides strong and ubiquitous transgene expression in most of the cells including skeletal muscle. Actually, transgenic mice carrying CAG promoter driving green fluorescence protein showed prominent expression of green fluorescence protein in muscle [12]. Anwer et al. [13]compared CMV enhancer/promoter and chicken skeletal muscle actin promoter regarding human growth hormone expression in the mouse. Interestingly, chicken skeletal muscle actin promoter with 3′-UTR from human growth hormone was strongly expressed in skeletal muscle but not lung tissue. In contrast, the CMV enhancer/promoter was strongly expressed in both muscle and lung. Blezinger et al. [9]reported that the combination of CMV enhancer/promoter and human growth hormone 3′-UTR provides strong expression of endostatin in skeletal muscle. Li et al. [14]developed a synthetic muscle-specific promoter composed of several myogenic regulatory elements and reported that the artificial promoter yielded a 3- to 8-fold increase in transcription efficiency.The hemagglutinating virus of Japan (HVJ) liposome method has been used for skeletal-muscle-targeting gene transfer. The HVJ-liposome-mediated gene transfer into skeletal muscle improves to a 10 times higher gene expression than the simple naked plasmid injection. Recently, Saeki et al. [15]have developed an artificial viral envelope (AVE)-type HVJ liposome, in which lipid composition is changed to mimic the viral envelope. The AVE-type HVJ liposome provided a 6 times higher gene transfection efficiency compared with the prototype HVJ liposome. Moreover, Kaneda and colleagues challenged the AVE-type HVJ-liposome-mediated gene transfer to muscle of primates. They injected AVE-type HVJ liposome containing human hepatic growth factor (HGF) plasmids to monkey skeletal muscle and then identified human HGF in the serum [pers. commun.]. Serum transaminases, LDH and creatinine were not changed during 4 weeks of observation. Only CPK was initially increased but normalized in a week. Histological examination of skeletal muscle, liver, pancreas, intestine, kidney, bone marrow, testis/ovary and lymph nodes revealed neither obvious inflammation nor tissue injury. These results imply that the application of HVJ liposome to humans could be harmless.Electroporation has been widely used to introduce DNA to various types of cells in vitro and to skin, liver and melanoma in vivo. Aihara and Miyazaki [16]injected the plasmid encoding interleukin 5 into tibialis anterior muscle and electric pulses were subsequently delivered through a pair of needles. An increase in interleukin 5 concentration in the serum was observed over 20 ng/ml (2 ng/ml in normal serum). Mir et al. [17]reported similar results using low-field strength, long-duration, square-wave electric pulses. They obtained effective gene transfer into skeletal muscle with low variability among individuals and long-term foreign gene expression for at least 9 months. Electroporation allows a much higher gene transfer efficiency compared with simple naked DNA injection although the procedures are rather invasive.Adeno-associated virus (AAV) has emerged as an attractive alternative. AAV is a single-strand DNA virus which has the ability of site-specific integration on chromosome 19, if the Rep proteins are supplied in trans [18]. However, a recent study demonstrated that AAV is not always integrated into chromosome 19 [19]. Skeletal muscle is a good target for long-term gene expression of AAV. Kessler et al. [20]demonstrated that the sustained expression of erythropoietin gene lasted for 32 weeks after single intramuscular injection of the AAV vector. During observation, the serum level of erythropoietin reached up to 700 mU/ml and the hematrocrit was sustained over 80%. Gene therapy with AAV vectors targeting skeletal muscle may be feasible for constant supply of individual protein through circulation.In case of AAV, some device for the regulatory expression of the transgene may be required since the transfected gene may overproduce the coding product. Ye et al. [21]created the rapamycin-inducible system in the context of gene transfer to skeletal muscle with AAV as a vector. A character of rapamycin which binds to two proteins, FKBP12 and FRAP, is applied to the system. Transcriptional activation is achieved in vivo through rapamycin-induced reconstitution of a transcription factor complex that is formed by coupling of two independently expressed artificial proteins: ZFHD1-3xFKBP which is formed by a unique DNA-binding domain of ZFHD1 genetically fused with FKBP, and FRB-p65 which consists of the activation domain of p65 of NFkB fused with the rapamycin-binding domain of FRAP, termed FRB. One AAV vector containing CMV promoters drives FRB-p65 and ZFHD1-3xFKBP. The other AAV vector contains erythropoietin gene driven by a promoter recognized by the ZFHD1 DNA binding domain. Administration of rapamycin resulted in a 200-fold induction of plasma erythropoietin.A successful human gene therapy for critical limb ischemia has been presented by Baumgartner et al. [4]. They injected naked plasmid DNA encoding the 165-amino-acid isoform of human VEGF directly into the skeletal muscle of an ischemic limb. In consequence, intramuscular injection of the plasmid achieved constitutive overexpression of VEGF sufficient for inducing therapeutic angiogenesis in the limbs. In 7 out of 10 limbs, newly visible collateral blood vessels were demonstrated by angiogenesis and the ankle-brachial index improved significantly. VEGF levels were apparently elevated in the systemic circulation as evidenced by transient peaks of VEGF in the serum and by edema of the ischemic as well as the opposite limb. Interestingly, the increased collateral vessels were localized to the ischemic limb and did not develop in other parts of the body. As pointed out by the authors, this may reflect the upregulation of VEGF receptor in an ischemic limb and the short half-life of VEGF in the circulation.Skeletal-muscle-targeted gene therapy may create a new paradigm of human gene therapy. Gene therapy has several potential advantages over recombinant protein therapy. Gene therapy can produce sustained delivery of therapeutic protein to the lesion compared with the intermittent administration of drugs or recombinant proteins. Of note, protein produced by the host may be more natural because it contains important posttranslational modifications, which may lack in a recombinant protein. Furthermore, the fact that Isner’s group [4]prepared the plasmid for human gene therapy at the university medical center laboratory impressed with the simple and rapid clinical application of the idea born in the laboratory. From the financial aspect, producing cDNA gene is less expensive and overcomes the need to highly purify a recombinant protein. In contrast to therapeutic proteins on drugs, they require expensive manufacturing facilities and many years of scaled-up effort.Transforming Growth Factor β. As discussed above, the gene delivery to skeletal muscle is a promising strategy for the systemic secretion of therapeutic protein. Application of the strategy to experimental glomerulonephritis has been challenged. Isaka et al. [22]proved that the intervention of overexpressed transforming growth factor β (TGF-β) ameliorates the extracellular matrix expansion of anti-Thy1 glomerulonephritis. They used a proteoglycan, decorin [22], and a soluble receptor for TGF-β which is composed of the extracellular domain of TGF-β type II receptor and IgG-Fc (TGF-RII/Fc) [23]. Decorin and the soluble receptor TGF-RII/Fc bind to active TGF-β, resulting in its inactivation. A plasmid encoding decorin cDNA was transfected to the skeletal muscle and the synthesized decorin was accumulated in glomeruli. The same behavior was also observed by the transfection of TGF-RII/Fc cDNA. Interestingly, the increment of extracellular matrix was reduced in comparison with the diminution of TGF-β mRNA, suggesting an interruption of the autocrine production of TGF-β. These results suggest that manipulation of overexpressed TGF-β may ameliorate the progression of fibrotic changes of chronic progressive glomerulonephritis.Hepatocyte Growth Factor. HGF, a pleiotropic growth factor mainly produced by mesenchymal cells, is mitogenic for a broad range of epithelial and endothelial cells. HGF is also known as a morphogen [24]. When MDCK tubular epithelial cells in monolayer are exposed to HGF, they form branching tubules. HGF induces mesenchyme-to-epithelium conversion [25]. In IMCD cells, HGF reduces apoptosis of tubular cells exposed to cisplatin [26]. HGF ameliorates acute renal failure caused by ischemic injury and toxic reagents such as cisplatin and mercury chloride HgCl2 [27]. Mizuno et al. [28]established a nephritic mouse strain, which develops chronic renal failure and dies by the age of 17 weeks. The apparent renal fibrosis and glomerulosclerosis were observed at 14 weeks. Mizuno et al. [28]challenged HGF therapy in congenitally nephrotic mice. They injected recombinant HGF into the intraperitoneal space for a period of 4 weeks. HGF inhibited the expression of TGF-β, smooth muscle α-actin and type I collagen, and significantly suppressed glomerular lesions in hereditarily nephrotic mice. The interstitial fibrosis and tubular atrophy of the mouse was also inhibited by the administration of HGF. HGF is also effective in other fibrotic diseases including liver cirrhosis. Ueki et al. [29]challenged gene therapy using HGF gene against experimental liver cirrhosis induced by dimethylnitrosamine administration. They transduced HGF cDNA in skeletal muscle 3 weeks after the administration of dimethylnitrosamine. The rats transfected with HGF gene significantly reduce fibrotic lesions in the liver. Intriguingly, HGF suppressed TGF-β expression, which plays an essential role in the progression of liver cirrhosis. These observations strongly suggest that HGF is an antifibrotic factor in progressive fibrosis including progressive renal diseases and a promising factor for gene therapy.Systemic delivery of the therapeutic protein produced by the transferred gene is an alternative method of gene therapy. Skeletal-muscle-targeting gene therapy is easy to perform and highly efficient at the expression of the transgene. The emerging image of the secreting protein, for example, an inhibitor, an activator or a soluble receptor, can be realized by the basic gene technology in the laboratory and can test the action by skeletal-muscle-targeting gene transfer in experimental animals. The clinical application of the gene transfer has come true in vascular diseases, and the kidney may not be an exception because vascular diseases involve the renal arteries. We believe skeletal-muscle-targeting gene therapy will become a common therapeutic strategy in the near future.We thank Dr. Yasufumi Kaneda for helpful discussion to create the paper.