Non Viral Gene Transfer Approaches for Lysosomal Storage Disorders

The lysosomal storage disorders (LSDs) are a group of almost 50 genetic diseases, characterized by mutations and loss of activity of lysosomal enzymes or, less frequently, non-lysosomal proteins that are involved in protein maturation or lysosomal biogenesis (Meikle et al, 2004). Most LSDs have an autosomal recessive inheritance, with some exceptions as Hunter syndrome (X-linked recessive), Danon disease (X-linked dominant) and Fabry disease (X-linked with a high proportion of heterozygous affected females). Storage of distinct undegraded or partially degraded material, usually the substrate of the defective enzyme, occurs in the lysosome. The substrate type is used to group the LSDs into general categories (table 1), including mucopolysaccharidoses (characterized by the storage of mucopolysaccharides, also called glycosaminoglycans), lipidoses (storage of lipids), glycogenoses (storage of glycogen) and oligosaccharidoses (storage of small sugar chains). Despite this categorization, many clinical similarities are observed between groups as well as within each group. Generally these diseases are multisystemic, and clinical features of many LSDs include organomegaly, central nervous system dysfunction and coarse hair and faces. Most LSDs are characterized by their progressive course with high morbidity and increased mortality, although there are significant variations between different diseases, and even among patients with the same disease (Walkley 2009). Lysosomal enzymes are ubiquitously distributed, but substrate storage is usually restricted to cells, tissues and organs with higher substrate turnover. Recently, it has been suggested that the primary gene defect and substrate storage are triggers of a complex cascade of events that lead to many of the disease manifestations (Bellettato & Scarpa, 2010). In this context, secondary substrate storage, perturbations of Calcium homeostasis and lipid trafficking would contribute to disease pathogenesis. Other manifestations, related to the lysosome’s role in vescicle trafficking, including antigen presentation, innate immunity, and signal transduction would cause inflammatory and auto-immune disturbances observed in the LSD (Parkinson-Lawrence et al., 2010). In addition, general mechanisms such as unfolded protein response, reticulum stress, oxidative stress and autophagy blockade would also play a role in the pathogenesis (Vitner et al., 2010). The incidence and prevalence of these diseases varies from different countries and regions. For example, the overall incidence of GM1 Gangliosidosis is considered to be 1:100,000-


Introduction
The lysosomal storage disorders (LSDs) are a group of almost 50 genetic diseases, characterized by mutations and loss of activity of lysosomal enzymes or, less frequently, non-lysosomal proteins that are involved in protein maturation or lysosomal biogenesis (Meikle et al, 2004).
Most LSDs have an autosomal recessive inheritance, with some exceptions as Hunter syndrome (X-linked recessive), Danon disease (X-linked dominant) and Fabry disease (X-linked with a high proportion of heterozygous affected females). Storage of distinct undegraded or partially degraded material, usually the substrate of the defective enzyme, occurs in the lysosome. The substrate type is used to group the LSDs into general categories (table 1), including mucopolysaccharidoses (characterized by the storage of mucopolysaccharides, also called glycosaminoglycans), lipidoses (storage of lipids), glycogenoses (storage of glycogen) and oligosaccharidoses (storage of small sugar chains). Despite this categorization, many clinical similarities are observed between groups as well as within each group. Generally these diseases are multisystemic, and clinical features of many LSDs include organomegaly, central nervous system dysfunction and coarse hair and faces. Most LSDs are characterized by their progressive course with high morbidity and increased mortality, although there are significant variations between different diseases, and even among patients with the same disease (Walkley 2009). Lysosomal enzymes are ubiquitously distributed, but substrate storage is usually restricted to cells, tissues and organs with higher substrate turnover. Recently, it has been suggested that the primary gene defect and substrate storage are triggers of a complex cascade of events that lead to many of the disease manifestations (Bellettato & Scarpa, 2010). In this context, secondary substrate storage, perturbations of Calcium homeostasis and lipid trafficking would contribute to disease pathogenesis. Other manifestations, related to the lysosome's role in vescicle trafficking, including antigen presentation, innate immunity, and signal transduction would cause inflammatory and auto-immune disturbances observed in the LSD (Parkinson-Lawrence et al., 2010). In addition, general mechanisms such as unfolded protein response, reticulum stress, oxidative stress and autophagy blockade would also play a role in the pathogenesis (Vitner et al., 2010). The incidence and prevalence of these diseases varies from different countries and regions. For example, the overall incidence of GM1 Gangliosidosis is considered to be 1:100,000www.intechopen.com 1:200,000, however in some countries as Malta (1:3,700) and the South of Brazil (1:13,317) it is considerably higher (Baiotto et al, 2011). Large population studies suggest that the overall incidence of the LSDs vary from 1:5,000-1:7,700 (Fuller et al, 2006). The treatment options available for the LSDs were restricted to support measures until a few decades ago. Nowadays, specific treatments are available for a certain number of LSDs, even though some of them are still in the experimental phase or have limited effects. Treatment options listed in table 1 consider those already approved or under clinical trial, including compassionate use. Support measures and palliative care are not considered treatment options in this context. The two most widely used treatment options are Hematopoietic Stem Cell Transplantation (HSCT) and Enzyme Replacement Therapy (ERT). HSCT has proven to be www.intechopen.com efficient to some of these diseases, especially if performed early enough to prevent irreversible lesions. Nevertheless, limitations such as the need of an early diagnosis, the difficulties to find a compatible donor in short time, and the high rates of morbidity and mortality associated with the procedure still limit this type of treatment (Malatack et al, 2003). Therefore, despite the many advances in this treatment over the last 30 years (Boelens et al 2010), its use has been deferred in favor of ERT whenever it is available. Enzyme replacement therapy is approved for a growing number of LSD, especially those without CNS involvement. It has proven to reduce some visceral symptoms as hepatosplenomegaly and improve respiratory function (Sifuentes et al, 2007), however difficult-to-reach organs such as the brain and the bones are still a major challenge.

Rationale for gene therapy in LSDs
The rationale for gene therapy and other enzyme-based approaches for treatment of LSDs was first introduced almost five decades ago by Christian de Duve, and can be summarized in the following sentence from his original work "In our pathogenic speculations and in our therapeutic attempts, it may be well to keep in mind that any substance which is taken up intracellularly in an endocytic process is likely to end up within lysosomes." (de Duve, 1964). His work and other pioneer studies showing cross-correction between fibroblasts from patients with Mucopolysaccharidosis type I and type II (deficient in alpha-Liduronidase and iduronate-sulphatase, respectively) established that the enzyme produced in one cell could be uptaken by a deficient cell, thus restoring its phenotype (Fratantoni et al, 1968). Later studies identified that this uptake was a saturable, receptor-mediated process, and the main actor of this process was the mannose-6-phosphate (M6P) receptor localized in the plasmatic membrane. The post-translational modification of addition of the M6P to the protein was discovered to be a signal not only to endocytosis but also for targeting nascent hydrolases to lysosomes (Fisher et al, 1980;Sly et al., 1981). These pivotal discoveries in the field of endocytosis and targeting of lysosomal enzymes provided the basis for treatments like HSCT and ERT. In the same way, LSDs may be considered good targets for gene therapy, despite their multisystemic involvement. The correction of a few cells could lead to the enzyme being secreted into the circulation and uptaken by the deficient cells, resulting in widespread correction of the biochemical defect (figure 1). Since long term gene expression is desirable, most clinical and preclinical trials used viral based vectors. Initial studies on fibroblasts showed promising data using retroviruses (Anson et al, 1992). However, when tests in animal models started, it became clear that some organs as the brain would not be easily corrected, as the enzyme could not cross the bloodbrain-barrier (Elinwood et al, 2004). This is a major hurdle as most LSD shows some degree of CNS involvement. Nevertheless, CNS targeted approaches could be envisaged to overcome this obstacle. For instance, Worgall et al. (2008) showed a slowing of progression of Neuronal Ceroid Lipofuscinosis 2 in ten children treated with serotype 2 adenoassociated virus expressing CLN2 cDNA. Another clinical trial, for Pompe disease, also used adeno-associated-based vector, but in this case serotype 1 (NCT00976352www.clinicaltrials.gov). Other two trials performed in the late 1990s used retroviral vectors for Gaucher (Dunbar et al., 1998) and Mucopolysaccharidosis type II (NCT00004454www.clinicaltrials.gov). Safety issues related to immune response of the adenoviral vectors (Wilson, 2009) and the possibility of insertional mutagenesis of the retroviral vectors (Hacein-Bey-Abina S et al., 2008) led researchers to develop a series of studies in parallel using non-viral approaches to treat lysosomal storage disorders.

Non-viral approaches
Non-viral vectors have important safety advantages over viral approaches, including their reduced pathogenicity and capacity for insertional mutagenesis, as well as their low cost and ease of production (Fraga et al., 2008). The application of non-viral vectors to humans has, however, been held back by the poor efficiency of their delivery to cells and the transient expression of the transgenes. As new strategies are being developed for the Fig. 1. Rationale for Gene Therapy in Lysosomal Storage Disorders: Cross-correction. The upper part of the figure shows two cells from a patient with LSD, with large lysosomes (green) due to accumulation of undegraded material. The gene transfer of a normal copy of the missing enzyme (red structure in the nucleus), allows the enzyme (red triangles) to be produced, and degrade the material accumulated in the lysosomes. Part of the enzyme is secreted from the recombinant cell and is captured by neighbouring cells via the mannose-6phosphate receptors (light blue structures located on the cell membrane), reaching the lysosomes and being able to correct their phenotype, in a process called cross-correction. Note that the cell on the right was never transduced, but is able to capture the enzyme from the circulation or from the neighbouring cell. application of non-viral vectors as nucleic acid delivery systems (Resina et al., 2009;Wasungu and Hoekstra, 2006), progress is being made in the application of this kind of therapy to LSDs.

Naked plasmid transfection
The direct injection of a plasmid containing the gene of interest and the regulatory mechanisms to ensure its expression is the simplest form of gene therapy. It has some advantages over the virus mediated gene transfer systems. First, DNA preparation is simple and can be performed at relatively low cost. In addition, the safety concerns are much lower and large amounts of DNA can be transferred. However, the major limitation of this method is that it requires a local administration and the level of transgene expression is relatively low and restricted to the injection site (Glover et al., 2005). Hydrodynamic injection (Liu et al, 1999) is an experimental method capable to achieve efficient gene transfer and high level of transgene expression by systemic administration. In this procedure, a large volume of saline containing plasmid DNA is injected in a short period of time. The large volume and high injection rate forces the DNA solution into the liver, probably due to the permeability of liver fenestrae. A small hepatotoxicity, probably due to the large volume of saline, is observed and resolves within a few days. Even though this procedure is widely used in mice by tail vein injection, its feasibility has been demonstrated in larger animal models using a balloon occlusion catheter-based method to mimic hydrodynamic injection (Brunetti-Pierri et al, 2007;Kamimura et al, 2009). This approach has been used in a proof-of-concept study to show the importance of coexpression of the formylglycine-generating enzyme for synthesis and secretion of functional Arylsulfatase A in a mouse model of Metachromatic Leukodystrophy (Takakusaki et al., 2005). This enzyme is a posttranslational modifying enzyme essential for activating multiple forms of sulfatases including Arylsulfatase A and therefore limits the amount of functional enzyme that can be secreted from transduced cells. It has also been used in Mucopolysaccharidosis (MPS) type I (Camassola et al, 2005) and type VII (Richard et al, 2009). In the MPS I animals, storage content was reduced and enzyme activity was elevated in the liver and spleen. For the MPS VII model, a beneficial effect on the pathology was also observed, as liver-directed gene transfer led to the correction of secondary enzymatic elevations and to the reduction of GAGs storage in peripheral tissues and brain, as well as to histological correction in many tissues.

Liposomes and nanotechnology
Liposomes are lipid particles that resemble the cell membrane. Liposome-based gene delivery was first reported by Felgner in 1987, and is still one of the major techniques for non viral gene delivery into cells (Niidome & Huang, 2002). Lipoplexes are formed by the interaction of anionic nucleic acids binding to the surface of cationic lipids, forming multilamellar lipid-nucleic acid complexes where the negatively charged nucleic acid remains trapped inside the lipid bilayer (Dass, 2004). Since its discovery, different lipid formulations have been tested and modified. For example, stealth liposomes sterically stabilized with methoxypoly(ethylene glycol)distearoylphosphatidylethanolamine conjugates (PEG-DSPE) have long circulation half-lives following intravenous injection (Moreira et al., 2001). In addition, linear polycations such as linear, branched and dendritic vectors based on poly(ethylenimine) (PEI), poly(L-lysine) (PLL) and a range of poly(ethylene glycol) (PEG) were also developed (Hunter, 2006). Although linear PEI shows greater in vivo efficiency because of a dynamic structure change of the complex under high salt concentrations as found in blood (Niidome & Huang, 2002), it also possesses greater toxicity (Morille et al., 2008). Cationic nanoemulsions have been more recently considered as potential systems for nucleic acid delivery. The interest in these systems is justified by the fact that they are biocompatible and able to form complexes with DNA protecting it from enzymatic degradation (Nam et al., 2009). Other practical advantages include ease of production and the potential for repeated administration (Al-Dorsari and Gao, 2009). We have investigated the influence of phospholipids on the properties of cationic nanoemulsions/pDNA complexes. Complexes containing the phospholipids DSPC (1,2-distearoyl-sn-glycero-3-phosphocholine) and DSPE (1,2-distearoyl-sn-glycero-3-phosphoethanolamine) were less toxic in comparison with the formulations obtained with lecithin, DOPC (1,2-dioleoyl-sn-glycero-3-phosphocholine) and DOPE (1,2-dioleoyl-sn-glycero-3-phosphoethanolamine). In addition, higher amounts of reporter DNA were detected for the formulation obtained with the DSPC phospholipid (Fraga et al., in press). These cationic macromolecules can readily condense DNA or RNA into stable nanostructures for use in gene delivery (Hunter, 2006) but also for other nanotechnology-based approaches useful for the treatment of LSDs (Muro, 2010). Nanomaterials, as a result of their small size and their large surface area offer great promise for neuro-therapeutics (Ragnail et al., 2011) and thus may be a valid option for a large number of LSDs that affect the CNS. Liposome-mediated gene transfer for LSD has been performed in vitro using patient's fibroblasts as target cells. Estruch et al. (2001) delivered therapeutic genes by integrinmediated uptake into fibroblasts from patients with Fucosidosis and Fabry disease. The vectors consisted of a complex of lipofectin and a peptide containing an integrin-targeting domain and a poly-lysine domain to which plasmid DNA was bound. Transfected cells produced the corresponding enzyme at levels which were 10-40% of the total activity in cultures of normal fibroblasts. Although 95-98% of this activity was secreted, it did not appear to affect the viability of the cells. Our group used Lipofectamine to transduce fibroblasts from GM1 Gangliosidosis patients with the beta-galactosidase gene. Treated cells showed 33 to 100-fold increases in enzyme activity compared to untreated fibroblasts. However, after seven days enzyme activity was back to uncorrected values (Balestrin et al., 2008). When Geneticin was added to the medium (figure 2), stable expression at therapeutic levels was observed (mean 300 nmoles/h/mg prot) for 30 days, although at values lower than the normal range (mean 1,300 nmoles/h/mg prot). In vivo, PEG-coated liposomes have been modified with monoclonal antibodies in order to reach the CNS. A liposome is coated with peptidomimetic monoclonal antibodies that undergo receptor-mediated transcytosis across the blood-brain barrier on the endogenous peptide receptor transporters (Pardridge, 2007). These Trojan horses (figure 3) may use the insulin or transferrin receptor, and since the MAb binding site is different from the binding site of the endogenous ligand, there is no interference of endogenous ligand transport (Skarlatos et al, 1995). This approach has been used to deliver a non-viral plasmid DNA to the brain across the blood-brain-barrier after intravenous administration of liposomes coated with monoclonal antibody to the mouse transferrin receptor in a mouse model of Mucopolysaccharidosis type VII (Zhang et al, 2008). The enzyme activity was increased greater than ten-fold in brain, liver, spleen, lung, and kidney, but not in heart. A similar strategy has been used by Osborn et al (2008) for Mucopolysaccharidosis type I, although they used a plasmid bearing a fusion gene consisting of Transferrin (Tf) and α-L-iduronidase. The fusion product consisted of an enzymatically active protein that was transported into the CNS by TfR-mediated endocytosis. Short-term treatment resulted in a decrease in GAGs in the cerebellum of Mucopolysaccharidosis type I mice.   3. Schematic view of a Trojan horse liposome. A stealth PEGylated liposome is complexed with monoclonal antibodies (MAb) that undergo receptor-mediated transcytosis across the blood-brain barrier.

Cell microencapsulation
Cell microencapsulation is an approach in which cells are trapped in a semipermeable membrane, allowing the exchange of metabolites and nutrients between them and the external environment. The membrane prevents the access of the immune system to the cells, without the need for continued immunosuppression of the host (Uludag et al., 2000). Furthermore, this technique allows the localized and controlled release, and long term duration of therapeutic products derived from the microencapsulated cells (Orive et al., 2002). Microencapsulation has become an important system for cellular preservation (Mayer et al., 2010) and a potential strategy for the controlled delivery of therapeutic products (Orive et al., 2003). Alginate has been the most important encapsulation polymer due to its abundance, easy gelling properties and biocompatibility. Agarose, chitosan, and hyaluronic acid are other polymers used for microencapsulation (Orive et al., 2003). Cell microencapsulation presents the potential to deliver the therapeutic product of interest directly to the Central Nervous System (CNS). This has been achieved by different groups for brain tumors (Kuijlen et al 2007) and neurodegenerative diseases (Spuch et al 2010). A phase I clinical trial was conducted in Huntington's patients without signs of toxicity (Bloch et al 2004). This approach delivers the gene of interest in the spinal fluid, similar to the intrathecal enzyme replacement therapy. Thus, cell encapsulation can be suitable for the treatment of LSD, once the deficient enzyme could be released for long term directly in the CNS . In order to obtain larger amounts of secreted enzyme, the encapsulated cells should be genetically modified to over-express the enzyme of interest. This enzyme would then be released to the extracapsular space (Bressel et al 2008) and uptaken by adjacent deficient cells ( figure 4). This strategy has been used experimentally for different LSD, especially the Mucopolysaccharidosis (MPS). Three in vitro studies were performed in LSDs other than the MPS, one in Fabry disease and the other two in Metachromatic Leukodystrophy (MLD). Naganawa et al (2002) co-cultured fibroblasts from patients with Fabry disease with microencapsulated recombinant Chinese Hamster Ovary cells (CHO) over-expressing alpha-galactosidase. The deficient cells were able to uptake the enzyme decreasing their levels of globotriaosylceramide storage. A similar approach was used by our group to test the ability of Baby Hamster Kidney (BHK) cells over-expressing ARSA to correct the deficiency of this enzyme in human skin fibroblasts from MLD patients. Fibroblasts co-cultured with the encapsulated cells for four weeks showed levels of enzyme activity higher than normal. Transmission electron microscopy showed evidence of normalization of the lysosomal ultra structure, suggesting that the secreted enzyme was able to degrade the substrate (Lagranha et al., 2008). Consiglio et al (2007) collected the conditioned media of C2C12 cells over-expressing ARSA encapsulated in polyether-sulfone polymer and used it to treat oligodendrocytes from MLD mice. The deficient cells internalized the enzyme and it was normally sorted to the lysosomal compartment, reaching 80% of physiological levels and restoring sulfatide metabolism. Both in vitro and in vivo studies have been performed in the MPS types I, II and VII. For MPS II two studies were performed. The first was a proof of principle in which Hunter primary fibroblasts were co-cultured with alginate microcapsules containing C2C12 cell clones overexpressing IDS. After 5 days of co-culture this strategy was able to increase IDS activity inside the deficient fibroblasts to levels similar to normal (Tomanin et al 2002). The second study was a pre-clinical experiment in which APA (alginate-poli-L-lysine-alginate) microcapsules containing 1.5 x 10 6 allogeneic C2C12 myoblasts over-expressing IDS were www.intechopen.com implanted in the peritoneum of the MPS II mouse model. An increase in IDS activity in plasma was observed, along with a reduction on urinary GAG between the fourth and the sixth week of treatment. After 8 weeks, a reduction of 30% in the amount of GAG accumulated in the liver and 38% in the kidney were shown (Friso et al 2005). To evaluate the usefulness of this technique to treat MPS VII, Ross et al (2000a) injected APA encapsulated non-autologous cells overexpressing Gusb in the peritoneum of MPS VII mice. The results showed the presence of Gusb in the plasma 24 hours after implantation, reaching 66% of physiological levels by 2 weeks post implantation. Activity of Gusb was also detected in liver and spleen for the duration of the 8-week experiment. Accumulation of GAG was significantly reduced in liver and spleen sections and urinary GAG content reached normal levels. In another study, enzyme released by APA encapsulated 2A50 fibroblasts implanted directly into the lateral ventricles of the brain of MPS VII mice was delivered throughout most of the CNS, reversing the histological pathology (Ross et al., 2000b). In vitro studies were performed by Nakama et al (2006) who encapsulated immortalized recombinant human amniotic epithelial cells with MPS VII human and mouse fibroblasts and high GUSB www.intechopen.com activity was detected in the medium. Addition of mannose-6-phosphate led to decreased enzyme activity, suggesting that enzyme uptake was mediated by mannose-6-phosphate receptor.
Our group has shown that the correction of MPS I fibroblasts by recombinant encapsulated BHK cells is also mediated by mannose-6-phosphate receptor. The effect of the ratio fibroblasts:encapsulated cells was also analysed (figure 5). The amount of enzyme uptaken by the fibroblasts is essentially the same under the different ratios (5:1; 1:1; 1:5) although the enzyme activity in the medium increases, as more enzyme is released. Our results also showed an increase in IDUA activity in MPS I fibroblasts after 15, 30 and 45 days of co-culture with the capsules. Cytological analysis showed a marked reduction in GAG storage within MPS I cells (Baldo et al., 2011). Ongoing experiments are under way in the MPS I mouse model. The capsules were implanted in the peritoneum (figure 6) and animals were sacrificed at 4 months later. Histological analysis showed a reduction on GAG storage although plasma and tissue enzyme activity levels were not increased. These results are quite different from those of Barsoum et al (2003) who implanted genetically modified Madin-Darby canine kidney cells (MDCK) over-expressing canine Idua in the brain parenchyma of one MPS I dog. Enzyme in plasma and cerebrospinal fluid was low but detectable for 21 days. Immunohistochemistry with anti-IDUA antibody showed the presence of the enzyme in various brain regions, however an extensive inflammatory reaction was noted, both at the sites of implantation and in the immediate vicinity. This may be the reason why histological correction of lysosomal inclusions has not been observed.

Transposon-based systems
For human gene therapy we can enumerate some important aspects of transposon systems (special emphasis in this chapter will be given to the sleeping beauty transposon) that make them appealing as a vector: (1) the integrated gene has stable expression, providing longlasting expression of a therapeutic gene, which, as already mentioned, is essential for lysosomal storage diseases; (2) the transposase directs the integration of single copies of a DNA sequence into chromatin and (3) the sys t e m i s b i n a r y ( t h e t r a n s p o s o n i s n o t autonomous or able to transpose on its own) (Hacket et al, 2005). The most studied transposon system, which has been used in pre-clinical studies for treatment of lysosomal disorders, is the Sleeping beauty (SB) system. Sleeping beauty transposon is a type of mobile element that belongs to the Tc1/Mariner class and that is able to transpose via movement of a DNA element in a simple cut-and-paste manner. For that, a precise piece of DNA is excised from one DNA molecule and moved to another site in the same or in a different DNA molecule (Plasterk, 1993). This reaction is catalyzed by the protein transposase, which can be supplied in trans by another plasmid for gene therapy purposes. The SB transposon system consists of two components: (i) a transposon, made up of the gene of interest flanked by inverted repeats (IRs), and (ii) a source of transposase ( figure  7). For Sleeping Beauty-mediated transposition, the transposase can recognize the ends of the IRs, excises the gene of interest from the delivered plasmid DNA, and then inserts it into another DNA site. Based on studies in about 2,000 integration events in either mouse or human genomes, transposons seem to integrate into random sites, including exons, introns and intergenic sequences Horie et al, 2003;Hacket et al, 2005). This is a potential problem, since it may lead to an event of insertional mutagenesis. A complete list of insertion positions of the SB tranposon can be found at the Mouse Transposon Insertion database at http://mouse.ccgb.umn.edu/transposon/ . The use of SB transposons as gene therapy approach in LSD is still recent. The first published work was conducted by Aronovich et al (2007) who studied the effects of intravenous hydrodynamic injection of the SB transposon into mice with Mucopolysaccharidosis types I or VII. Without immunomodulation, initial enzyme activities in plasma reached levels higher than 100-fold that of wild-type (WT). However, both GUSB (MPS VII) and IDUA (MPS I) levels fell to background within 4 weeks postinjection. A second group of animals was performed with immunomodulation only in MPS I mice. Plasma IDUA persisted for over 3 months at up to 100-fold WT activity in one-third of the mice, which was sufficient to reverse lysosomal pathology in the liver and, partially, in distant organs. Histological and immunohistochemical examination of liver sections i n I D U A t r a n s p o s o n -t r e a t e d W T m i c e r e v e aled inflammation 10 days post-injection consisting predominantly of mononuclear cells, which can be seen as a potential side-effect. A posterior study by the same group (Aronovich et al, 2009) was performed in another MPS I strain, which is immunodeficient (NOD/SCID mice). Using the same approach from the previous experiment, they were able to show a persistent elevation (100-fold normal) in the plasma IDUA levels for 18 weeks. Also, IDUA activity was present in all organs analyzed, including the brain. The SB transposon system proved efficacious in correcting several clinical manifestations of MPS I mice, including bone abnormalities, hepatomegaly, and accumulation of foamy macrophages in bone marrow and synovium. In 2008 the first human clinical trial using the SB transposon was approved in the USA for treatment of cancer (Williams, 2008), however although promising no clinical trials have been conducted in LSDs so far. Fig. 7. SB transposon system. This simplified version of the SB transposon system shows the cut-and-paste system used by the SB transposase to insert the DNA into the host genome. The gene of interest is flanked by two inverted repeats regions (IRs, arrows) which will be recognized by the transposase (usually given in trans in a second plasmid) and allow the transgene to be inserted into the host genome. This way only the transgene will have prolonged expression, as transposase gene expression is transient. The SB transposon is a non-viral method of gene delivery that allows integration of the transgene in the host cell.

Minicircle gene therapy
Sustained in vivo transgene expression from plasmids can be difficult to achieve due to gene silencing. The mechanism by which this process occurs was postulated to be due to the deposition of repressive heterochromatin on the noncoding bacterial backbone sequences required for plasmid bacterial preparation and propagation (Chen et al, 2008;Riu et al, 2005).
Based on those findings, a new technology has emerged, known as the minicircle (MC) gene therapy. This system uses a phiC31 integrase recombination event to remove the bacterial backbone elements of the plasmid resulting in a DNA circle (the MC), encoding the mammalian expression cassette of choice and a small attR footprint (Chen et al, 2003). This has proven to be resistant to gene silencing in vivo, is maintained as an extrachromosomal episome, and therefore represents an interesting platform for gene replacement strategies for lysosomal storage disorders (figure 8). This technology was recently used for treatment of MPS I mice in a proof-of-concept study (Osborn et al, 2011). In this study, the researchers performed a hydrodynamic injection of a minicircle plasmid containing the IDUA gene combined with immunomodulation, achieving stable expression of the transgene, increased IDUA tissue levels and reduction in GAG storage. As a recent technology, this is the only study performed on LSDs so far, but results are encouraging and should be tested on other diseases soon. Fig. 8. Production of a minicircle plasmid. This simplified version of the process shows the parental plasmid containing both the gene of interest and the bacterial components, including origin of replication (ORI), genes that confer resistance to antibiotic (Ab resist) and sites that allow attachment of the integrase (att B and P). After activation of the integrase, a cis-recombination event occurs, separating the gene of interest and its regulatory elements from the bacterial backbone, which then is degraded.

Conclusions
Still nowadays most lysosomal storage disorders do not have an effective treatment. Moreover, treatments currently available are not able to correct all the manifestations of these multisystemic diseases. Despite the small number of protocols (if compared to other areas, like oncology), gene therapy approaches have shown their potential to be helpful in many of these diseases. Not only proof of concept experiments have been performed, but clinically relevant results were obtained in some cases.
The limitations of non-viral gene transfer, i.e., transient expression of the transgene and low transfection efficiency, are being slowly overcome in the last decade using improved vector design and techniques, such as nanotechnology, transposons, and minicircle approaches (to name a few) as demonstrated throughout this chapter. Novel mechanisms to help the DNA to escape endosomal degradation and pass through the nuclear envelope are also under development but were not in the scope of this chapter. Nevertheless, these improvements help non-viral gene therapy to move towards clinical trials in LSDs, which are expected to happen in the years to come. Non-viral vectors are safer than viral particles, which make them an appealing alternative for treatment of lysosomal storage disorders and even other monogenic diseases. Yet, there is a long way to clinical application but the road is paved and the scientific community advances steadily.