Stem Cell-Based Gene Therapy

Introduction

Many researchers and clinicians wonder if gene therapy remains a way to treat genetic or acquired life-threatening diseases. For the last few years, many experimental, preclinical, and clinical data have been published [1–3] showing that it is possible to transfer with relatively high efficiency new genetic information (transgene) in many cells or tissues including both hematopoietic progenitor cells and differentiated cells. Based on experimental works, addition of the normal gene to cells with deletions, mutations, or alterations of the corresponding endogenous one has been shown to reverse the phenotype and to restore (in some cases) the functional defect. In spite of very attractive preliminary results, however, suggesting the feasibility and safety of this process [4, 5], therapeutically efficient gene transfer and expression in targeted cells or tissues must be proven. In this review, we will focus primarily on the attempts to use gene transfer in hematopoietic stem cells as a model for more general genetic manipulations of stem cells. Hematopoietic stem cells are included in a subset of bone marrow, cord blood, or peripheral blood cells identified by the expression of the CD34 antigen on their membrane.

A Complex Network of Variables

Genetic manipulation of stem cells means the transgene has been maintained in the progeny and that gene expression has occurred in the terminal cells originating from the manipulated stem cells. Thus, stem cell-based gene therapy relies on the integration of the transgene in the stem cell genomic host, or episomal maintenance and replication of the expression vector. Numerous gene delivery systems called vectors have been tested (illustrated in Table 1 is a comparison of gene-delivery systems for hematopoietic cells), including lipid-based structures (liposomes) linked to plasmid vectors and physical treatment such as particle bombardment or shock waves. Only very low integration of the transgene can be obtained with these approaches, and in addition, extra chromosomal (episomal) replication in hematopoietic cells has not been successfully assessed [6]. Based on their natural properties to infect mammalian cells and for some of them, such as retroviruses, to be integrated in the genomic DNA-, adeno virus-, adeno-associated virus- (AAV), and retrovirus-based vectors have been developed. Adeno-associated virus-based vectors do not necessarily integrate the genome host, and mainly remain episomal, but transduction with AAV vectors done at high MOI (multiplicity of infection representing the ratio of viral particles per cell), appears to favor efficient integration of the transgene in parallel with short-term episomal maintenance [7, 8]. Despite the use of small transgene inserts (about 4 kilobases, presently considered the maximum size), and difficulties in preparing a high-titer supernatant devoid of contaminant adenovirus particles, it is anticipated that rapid evolution of these vectors for transduction of hematopoietic stem cell will occur. Therefore, murine retroviral vectors, the most widely used delivery system, remain the most powerful tool to drive genomic integration of the transgene [9]. Numerous in vitro and in vivo studies report retroviral infection and transduction of immature hematopoietic progenitors and the long-term maintenance of provirus in cells arising from transduced immature cells [2, 10–13]. When considering stem cell transduction, the challenge is complex since integration occurs only in cycling cells and most of the primitive stem cells are in the resting phase. Activation and proliferation of such cells can be obtained by addition of growth factors such as interleukin (IL)-3, IL-6, stem cell factor (SCF), granulocyte colony-stimulating factor (G-CSF), megakaryocyte growth and development factor (MGDF), flt3 ligand, and many others; or by cocultivating the target-cell population with the virus-producing cells. These strategies, however, also induce differentiation resulting in a decrease in the ability of cells to self-renew and to give rise to multipotential stem cells. HIV-1-derived and lentivirus-derived vectors able to transduce quiescent cells are foreseen as promising strategies to improve retroviral vector efficiencies in non-dividing cells [14].

Table 1

Clinical trials in the field of hematopoietic stem cell gene or T-lymphocyte therapy or gene marking: ex vivo manipulations. Adapted from [52]; used with permission.

Gene	Diseasea	Source of the cellsb	Referencesc
p47 Phox	CGD	BM	[26]#
ADA	SCID	T	[5]*
		BM	[47]*
		BM	[49]*
γ chain	X-SCID	PBSC	[51]#
RRE decoy	AIDS	BM	[46]#
Ribozyme	AIDS	BM	–
		BM	[44]#
NeoR	Cancer	BM	[48]#
	Breast cancer	BM/PBSC	[40]*
	AML, CML	BM	[32]*
nlsLacZ	Breast cancer	PBSC	[37]#
MDR	Ovarian cancer	PBSC	[50]*
GCB	Gaucher	BM	[27]#
			[45]#

^aCGD = chronic granulomatous disease; SCID = severe combined immune deficiency; X-SCID = chromosome X linked-SCID; AIDS = acquired immune deficiency syndrome; AML = acute myeloid leukemia; CML = chronic myeloid leukemia.

^bBM = bone marrow; PBSC = peripheral blood cell stem cells (CD34⁺); T = T lymphocytes.

^c(*) clinical trial report, (#) preclinical data.

Natural or artificial retroviral backbones are being developed that go beyond the classical Mo-MuLV-derived vectors and NIH-3T3-derived packaging cell lines presently in use. Transduction by viral vectors depends, in part, on the expression of receptors on the cell surface. Thus, improvements of retroviral vectors, in light of more efficient or cell-specific transduction, are based on chimeric viral backbones composed of retroviral elements from different origins, pseudotyped particles expressing non-retroviral env proteins (VSV-g), and/or proteins targeted to bind to cell-surface molecules such as growth-factor receptors [15–17]. Development of producing/packaging cell lines leading to the production of high virus titers either from murine or from human origin is also progressing [18].

Another approach may be to combine viral vectors exhibiting complementary features. For example, adenoviral vectors, although very efficient in infecting the majority of cells, do not fulfill the criteria for stable integration and low immunogenicity. They could, however, be used in combination with retroviral vectors when short-term expression of a gene whose activity could promote retroviral transfer is needed.

Stable integration is not the only limiting factor since long-term expression of the transgene also must be maintained at a level compatible with the synthesis and processing of the protein of interest. Use of new vector designs [19, 20], analysis, and modification of hypermethylation status of transgene or regulatory elements [21], construction of inducible or specific promoter-containing vectors [22], and combined expression of selection and therapeutic genes [23] are thought to improve both long-term expression and control of the expression of the transgene.

Most of the work has been done ex vivo using enriched or purified stem cells or progenitor cells. Gene transfer was accomplished by using high titer viral supernatants or by transducing the target cells by cocultivation with the vector producing cell line. Many other variables also are involved, such as the presence of fetal calf serum and the quality of the medium in which gene transfer and cell activation take place, the role of adhesion molecules such as fibronectin [24], and the growth factors selected for inducing cell cycling. In some cases, to avoid the possible deleterious effect of these factors on the multipotentiality of stem cells, attempts have been made to transduce the cells on a stromal layer able to support long-term hematopoiesis or, at least, survival of stem cells [25]. Cell cycling in long-term cultures of stem cells is induced by changing the medium.

Direct in vivo gene transduction would alleviate most of the problems encountered now. Many other problems, however, must be resolved before using this approach, the most important being the specific targeting of cells in vivo and the vector resistance to degradation by serum proteases [18].

Diseases and Therapeutic Genes

The first steps toward stem cell-based gene therapy for treatment of genetic disease is the characterization of the genetic activity directly involved in or related to hereditary disorders, able to revert a pathologic phenotype when transferred into cells with the genetic defect, or the ability to produce a therapeutic product. The transgene product should not interact with the normal biological processes of the target cells, including proliferation and differentiation, nor should it be immunogenic [26] if sequential administration of the vector and/or the modified cells is sought.

From a theoretical point of view, gene therapies could be developed for the treatment of a variety of hematological diseases caused by monogenic alterations [27]. Characterization of the adenosine deaminase (ADA) gene emphasized gene therapy treatment of ADA-linked severe combined immune deficiencies (SCID); the first clinical trial for genetic disease occurred in this field [4]. In addition to ADA deficiency, the lack of expression of γc IL-2 receptor chain gene in the X-linked SCID, the WASP protein in Wiskott-Aldrich syndrome, and the Zap 70 kinase in immunologic disorders are under characterization and could be restored by gene transfer. Other potential applications are chronic granulomatous disease secondary to a mutation of PHOX gene possibly corrected by expression of the missing enzyme in the myeloid progeny [28]. A complementary approach would be to make use of hematopoietic cells to provide a molecule missing or altered in some genetic diseases, such as lysosomal storage disease or Gaucher disease [29]. Fanconi's anemia due to a deletion of complementation group C gene (FAC) could possibly be corrected by the transfer of the gene to erythroid progenitors in affected patients [30]. Patients with β-thalassemia might be treated by transfer of the β-globulin gene if it is possible to control its expression by cotransferring the minimal locus control region [31].

Gene therapy might be of interest to treat patients with the acquired immune deficiency syndrome (AIDS), in spite of impressive progress obtained by combinations of antiviral drugs. Specifically, gene therapy might be useful in specific cases such as resistance to therapy or in utero or neonatal infection. Several steps of the human immunodeficiency virus (HIV) replication cycle can be inhibited by transfer of “anti-HIV” genetic activity into stem cells before they differentiate into macrophages, dendritic cells, or T-lymphocytes, cells susceptible to HIV infection. Genes of interest include those coding for antisenses, negative dominant-mutated viral proteins, protein fixation site decoys that compete with natural proteins encoded by the provirus, IFN-α and IFN-β, proteins required for the inhibition of viral replication, and ribozymes able to specifically cleave RNA viral genomes [32]. If efficient transduction of lymphoid and monocytic progenitor cells could be performed easily, gene therapy for the treatment of AIDS might become a possible alternative or an adjuvant therapy to chemotherapy.

In patients with leukemia or cancer, two approaches appear interesting. First, in the setting of autologous or allogeneic bone marrow or stem cell transplantation, gene marking can give information about the fate of the graft, the role of the graft in relapse of disease [33, 34], and, more importantly, will allow the accurate measurement of in vivo gene transfer and gene expression in the relevant cells. Second, transfer of therapeutic genes, e.g., the multidrug resistance gene (MDR) into normal progenitors or the transfer of anti-oncogenes into leukemic progenitors could represent effective adjuvant therapy [35]. One of the main concerns of using this approach is the possibility of transferring the MDR gene into cancer cells or transferring anti-leukemic genes into normal cells. These applications need to be validated by in vivo models before clinical trials.

Therapeutic Efficacy

The therapeutic efficiency of these approaches must be evaluated. Several problems, such as long-term expression of the transgene and transduction efficiency of long-term reconstituting cells, still remain to be resolved. For this purpose, in vitro experiments and gene-marking clinical trials using innocuous marker genes, such as NeoR, nlsLacZ, or those coding for alkaline phosphatase or humanized green fluorescent protein must be completed to better define the potential and the limits of this field [36–38].

In addition, all these approaches depend on ex vivo transfer of either the marker gene or the correct genetic version of the altered gene into hematopoietic stem cells, and the subsequent infusion of the transduced cells into patients to allow the “transmission” of the therapeutic information to all lineages, including lineages not directly affected by the disease. As previously emphasized, controls of the transgene expression in chosen tissues are important challenges for the evolution of the field.

Target Cells

To be useful, stem cells must be abundant and easily collected and purified in a clinically acceptable manner. The definition of “stem cells” is still a passionate and sometimes controversial subject among biologists. Using criteria of hematopoietic reconstitution observed both in animals and humans, in vitro clonogenic assays, and long-term bone marrow cultures, it has been established that the most important reconstituting cell populations express the CD34 surface antigen, as well as expression of Thy-1 (CD90) or lack of expression of CD38 combined with lack of expression of lineage-differentiated markers (Lin⁻). More recently, a population of cells with low expression of CD34 was observed to be able to reconstitute hematopoiesis in a murine model [39]. We learned, however, from our work with allogeneic stem cell transplantations and SCID-hu models that most of the progenitors of the hematopoietic lineage express CD34. These cells can be derived from fetal liver, cord blood, bone marrow, or peripheral blood. In the latter case, CD34⁺ can be increased in patients with cancer or in normal volunteers by mobilization with recombinant growth factors such as rG-CSF or recombinant granulocyte-macrophage colony-stimulating factor (rGM-CSF). Some other factors such as flt-3 ligand can also safely improve CD34 cell mobilization. Knowing this, the progenitor cell to be transduced could be chosen on its ability to produce the desired cell lineage. For example, genetic manipulation of dendritic cells could be approached by transducing CD34⁺ Lin⁻ CD45RA⁺ CD10⁺ bone marrow cells [40].

What about Preclinical and Clinical Results?

Animals models such as NOD-SCID or SCID-hu mouse and the transplantation of human stem cells into fetal sheep have clearly demonstrated that genetically modified stem cells participate in hematopoietic reconstitution [2, 12]. In humans, Brenner and colleagues [36] were the first to demonstrate that bone marrow cells infected with retovirus carrying the NeoR gene can be retrieved several months after transplantation. By using the same gene to tag CD34⁺ cell populations, other researchers arrived at the same conclusion [12, 34]. An important observation made by Nolta et al. [41] demonstrates that hematopoietic reconstitution is oligoclonal, thus supporting the hypothesis that efficient gene transfer in few progenitors can be therapeutically effective if ex vivo or in vivo selection is induced. An interesting possibility is that deficient cells transduced with the normal gene have selective advantages over nontransduced affected cells. This was the basis of the gene therapy of ADA deficiency and of AIDS. In the case of ADA, this objective was not completely achieved, as all the children originally treated by gene-transfer techniques still require systemic therapy with recombinant ADA.

From these clinical trials, it can be suggested that retroviral gene transfer is a safe process as long as replication competent retroviruses remain absent [42]. Insertional mutagenesis is still a theoretical drawback, but its frequency is probably very low and not relevant in patients with cancer. Several clinical trials have been proposed or have started in the fields of genetic diseases, cancer, and AIDS (Table 2), and more than 1,000 patients had received vectors or genetically manipulated cells.

Table 2

Comparison of delivery systems for stem cell gene therapy

	Retroviral (murine)	Retroviral (HIV-based)	AAV	Plasmid vector
Genomic integration	Yes	Yes	Yes (High MOI required)	Very low
Cytoplasmic location	No	No	Yes	Yes
Maximal insert size	8 Kb	8.5 Kb	4.5 Kb	?
Clinical experiencea	Yes	No	No	No
Target cell status	Cycling	Cycling and quiescent	Cycling and quiescent	Cycling and quiescent
Pre-stimulation step	Necessary	?	Not necessary	Not necessary
Safety concerns	Insertional mutagenesis	Insertional mutagenesis	Insertional mutagenesis	Insertional mutagenesis
	Presence of RCVb	Presence of RCVb	Presence of RCVb

^aIncludes clinical trial not based on CD34⁺ cell manipulation.

^bBased on preclinical data reports.

RCV = Replication competent virus

MOI = multiplicity of infection representing the ratio of viral particles per cell

Stem Cell Gene Therapy Concept

Stem cell gene therapy research has gone through a period of naive enthusiasm, and there is still much work to be done to discover safe and effective therapies for the treatment of cancer and leukemias. It required many decades of research to clearly demonstrate that allogeneic bone marrow transplantation in association with radiotherapy or chemotherapy could cure some forms of leukemias and tumors. In addition, stem cell gene therapy for inherited disorders will become a reality when concerns about its safety and its delivery problems are solved. This will require close association of basic, clinical, and applied research. Because of the difficulty to direct in vivo the transgene to the targeted tissues, gene therapy presently relies mostly on ex vivo approaches (Fig. 1). Direct in vivo gene transfer and expression in chosen target cells remain the ultimate goal, and progress remains to be made in vector design, control of gene expression, targeting of the cells or tissues to be corrected, and a combination of gene-delivery systems.

Figure 1

Stem cell harvest, transduction, infusion, and/or analysis: an overview of the principles in stem cell gene therapy.  HPP-CFC=High proliferative potential-colony forming cells.

Hematopoietic stem cell gene transfer is also a very attractive model that can be used as a basis for investigation of other stem cell-based gene therapies. For example, Prockop [43] recently emphasized the importance of mesenchymal stem cells present in the bone marrow stroma. In addition, it becomes evident that the specificity of growth factors and receptors is not restricted to unique tissue or cell lineage, and that pluripotent stem cells could be obtained from fetal source and induced to differentiate along different pathways.

The parallel increase in knowledge of non-hematopoietic stem cell identification and behavior and improvement of tools to transfer new genetic information or to modify gene expression in affected cells will certainly lead to new therapeutic approaches in inherited and acquired diseases.