SCID-X1 Gene Therapy Research Articles

Transduction of Human Primitive Repopulating Hematopoietic Cells With Lentiviral Vectors Pseudotyped With Various Envelope Proteins

Lentiviral vectors are useful for transducing primitive hematopoietic cells. We examined four envelope proteins for their ability to mediate lentiviral transduction of mobilized human CD34(+) peripheral blood cells. Lentiviral particles encoding green fluorescent protein (GFP) were pseudotyped with the vesicular stomatitis virus envelope glycoprotein (VSV-G), the amphotropic (AMPHO) murine leukemia virus envelope protein, the endogenous feline leukemia viral envelope protein or the feline leukemia virus type C envelope protein. Because the relative amount of genome RNA per ml was similar for each pseudotype, we transduced CD34(+) cells with a fixed volume of each vector preparation. Following an overnight transduction, CD34(+) cells were transplanted into immunodeficient mice which were sacrificed 12 weeks later. The average percentages of engrafted human CD45(+) cells in total bone marrow were comparable to that of the control, mock-transduced group (37-45%). Lenti-particles pseudotyped with the VSV-G envelope protein transduced engrafting cells two- to tenfold better than particles pseudotyped with any of the gamma-retroviral envelope proteins. There was no correlation between receptor mRNA levels for the gamma-retroviral vectors and transduction efficiency of primitive hematopoietic cells. These results support the use of the VSV-G envelope protein for the development of lentiviral producer cell lines for manufacture of clinical-grade vector.

Dynamics of gene-modified progenitor cells analyzed by tracking retroviral integration sites in a human SCID-X1 gene therapy trial

AbstractX-linked severe-combined immunodeficiency (SCID-X1) has been treated by therapeutic gene transfer using gammaretroviral vectors, but insertional activation of proto-oncogenes contributed to leukemia in some patients. Here we report a longitudinal study of gene-corrected progenitor cell populations from 8 patients using 454 pyrosequencing to map vector integration sites, and extensive resampling to allow quantification of clonal abundance. The number of transduced cells infused into patients initially predicted the subsequent diversity of circulating cells. A capture-recapture analysis was used to estimate the size of the gene-corrected cell pool, revealing that less than 1/100th of the infused cells had long-term repopulating activity. Integration sites were clustered even at early time points, often near genes involved in growth control, and several patients harbored expanded cell clones with vectors integrated near the cancer-implicated genes CCND2 and HMGA2, but remain healthy. Integration site tracking also documented that chemotherapy for adverse events resulted in successful control. The longitudinal analysis emphasizes that key features of transduced cell populations—including diversity, integration site clustering, and expansion of some clones—were established early after transplantation. The approaches to sequencing and bioinformatics analysis reported here should be widely useful in assessing the outcome of gene therapy trials.

Gene Correction in Human Embryonic and Induced Pluripotent Stem Cells: Promises and Challenges Ahead

The sequencing of the human genome has greatly facilitated our ability to identify the gene candidates that are critical for developmental regulation and disease progression. However, investigators seeking to use such data to gain a better understanding of the developmental biology and therapeutic potential of pluripotent stem cell types may be frustrated by the limitations of current genetic engineering techniques. Targeted genomic manipulation of mammalian cells is inefficient, and unintended effects of such manipulation on the cells’ developmental and proliferative potential remain uninvestigated. To this end, the recent demonstration of targeted gene correction in human induced pluripotent stem cells (hiPSCs) before and after reprogramming, as reported in this issue of Molecular Therapy, warrants special mention.1

Gene Correction by Homologous Recombination With Zinc Finger Nucleases in Primary Cells From a Mouse Model of a Generic Recessive Genetic Disease

Zinc Finger nucleases (ZFNs) have been used to create precise genome modifications at frequencies that might be therapeutically useful in gene therapy. We created a mouse model of a generic recessive genetic disease to establish a preclinical system to develop the use of ZFN-mediated gene correction for gene therapy. We knocked a mutated GFP gene into the ROSA26 locus in murine embryonic stem (ES) cells and used these cells to create a transgenic mouse. We used ZFNs to determine the frequency of gene correction by gene targeting in different primary cells from this model. We achieved targeting frequencies from 0.17 to 6% in different cell types, including primary fibroblasts and astrocytes. We demonstrate that ex vivo gene-corrected fibroblasts can be transplanted back into a mouse where they retained the corrected phenotype. In addition, we achieved targeting frequencies of over 1% in ES cells, and the targeted ES cells retained the ability to differentiate into cell types from all three germline lineages. In summary, potentially therapeutically relevant frequencies of ZFN-mediated gene targeting can be achieved in a variety of primary cells and these cells can then be transplanted back into a recipient.

Safe and Efficient Expansion of Human HSC with Sendai Virus Vector Expressing HoxB4 in Fetal Sheep.

Abstract 693 Background:Homeobox B4 (HoxB4) has been shown to be a potent stem cell self-renewal gene, especially in hematopoietic stem cells (HSC). Accumulating evidence from murine studies indicates that the overexpression of HoxB4 enhances in vivo and ex vivo expansion of HSC. Although no leukemia has been observed after transplantation of HoxB4-transduced cells in murine models, the study using large animals such as dogs and non-human primates with retroviral vectors expressing HoxB4 showed the frequent development of leukemia. Regarding retroviral vectors expressing HoxB4, there is another concern, that is, insertional leukemogenesis, which has been elucidated in the hematopoietic stem cell gene therapy for X-SCID. To avoid the insertional mutagenesis, other vectors may be considered, including Epstein-Barr nuclear antigen (EBNA)-1 based episomal vectors or the transposon; however, problems are left, i.e. low transduction efficiency with EBNA vectors and unclear safety with transposon vectors. To avoid both the persistent HoxB4 expression and insertional mutagenesis leading to leukemogenesis, we have developed a new type of Sendai virus (SeV) vector; it lacks the polymerase gene, namely P-defective SeV (SeV/δP) vector. SeV is an enveloped virus with a non-segmented, negative-stranded RNA genome. SeV-based vectors are non-integrating, cytoplasmic vectors. They replicate exclusively in the cytoplasm of transduced cells, and do not go through a DNA phase; therefore, there is no concern about the unwanted integration of foreign sequences into chromosomal DNA of the host. We have previously shown that the transduction efficiency of human CD34+ cells with the SeV vector was very high; around 70% (100 multiplicity of infections). On the other hand, SeV/δP vectors are incapable of self-replication, thus enabling transient gene expression without spoiling their ability to efficiently transduce CD34+ cells. Here, using the SeV/δP vector expressing HoxB4 (SeV/δP/HoxB4 vector), we examined the effectiveness and safety of human HSC expansion after in utero transplantation to fetal sheep. Methods:After enrichment of CD34+ cells from cryopreserved human umbilical cord blood, these cells were repeatedly exposed to SeV/δP/HoxB4 vector every 24 h for 4 days. The transduced cells (3.2–11.7 × 105) were transplanted into the abdominal cavity of fetal sheep at 45–50 gestational days (full term, 147 days) that have premature immune system (HoxB4 group, n = 4; control group, n = 4). The engraftment of hematopoietic cells derived from human HSC in the lambs after birth was quantitatively evaluated by colony PCR of the bone marrow. The development of leukemia was assessed by regular sampling of peripheral blood and bone marrow. Results:The human–sheep chimeric ratio in the bone marrow of HoxB4 group was calculated 4.8-times higher than that of control group after birth, as assessed by colony PCR. The SeV genome was no longer detectable in the bone marrow and peripheral blood of lambs as assessed by RNA-PCR, confirming the SeV vectors were cleared. No leukemia developed in any of the sheep in either group at present (at 12 months after transplantation). Conclusion:The SeV/δP vector would be suitable for transient expression of HoxB4 in human CD34+ cells, enabling 4.8-times expansion of human HSC as assessed by their repopulating ability in sheep. The expansion of human HSC with the SeV vector was comparable to that with HoxB4-retroviral vectors. In addition, the SeV/δP vector is free of concern about transgene-related and insertional leukemogenesis and should be safer than retroviral vectors. Disclosures:No relevant conflicts of interest to declare.

Mechanism of Reduction in Titers From Lentivirus Vectors Carrying Large Inserts in the 3′LTR

Self-inactivating (SIN) lentiviruses flanked by the 1.2-kb chicken hypersensitive site-4 (cHS4) insulator element provide consistent, improved expression of transgenes, but have significantly lower titers. The mechanism by which this occurs is unknown. Lengthening the lentiviral (LV) vector transgene cassette by an additional 1.2 kb by an internal cassette caused no further reduction in titers. However, when cHS4 sequences or inert DNA spacers of increasing size were placed in the 3'-long terminal repeat (LTR), infectious titers decreased proportional to the length of the insert. The stage of vector life cycle affected by vectors carrying the large cHS4 3'LTR insert was compared to a control vector: there was no increase in read-through transcription with insertion of the 1.2-kb cHS4 in the 3'LTR. Equal amount of full-length viral mRNA was produced in packaging cells and viral assembly/packaging was unaffected, resulting in comparable amounts of intact vector particles produced by either vectors. However, LV vectors carrying cHS4 in the 3'LTR were inefficiently processed following target-cell entry, with reduced reverse transcription and integration efficiency, and hence lower transduction titers. Therefore, vectors with large insertions in the 3'LTR are transcribed and packaged efficiently, but the LTR insert hinders viral-RNA (vRNA) processing and transduction of target cells. These studies have important implications in design of integrating vectors.

Targeting LMO2 with a Peptide Aptamer Establishes a Necessary Function in Overt T-Cell Neoplasia

LMO2 is a transcription regulator involved in human T-cell leukemia, including some occurring in X-SCID gene therapy trials, and in B-cell lymphomas and prostate cancer. LMO2 functions in transcription complexes via protein-protein interactions involving two LIM domains and causes a preleukemic T-cell development blockade followed by clonal tumors. Therefore, LMO2 is necessary but not sufficient for overt neoplasias, which must undergo additional mutations before frank malignancy. An open question is the importance of LMO2 in tumor development as opposed to sustaining cancer. We have addressed this using a peptide aptamer that binds to the second LIM domain of the LMO2 protein and disrupts its function. This specificity is mediated by a conserved Cys-Cys motif, which is similar to the zinc-binding LIM domains. The peptide inhibits Lmo2 function in a mouse T-cell tumor transplantation assay by preventing Lmo2-dependent T-cell neoplasia. Lmo2 is, therefore, required for sustained T-cell tumor growth, in addition to its preleukemic effect. Interference with LMO2 complexes is a strategy for controlling LMO2-mediated cancers, and the finger structure of LMO2 is an explicit focus for drug development.

A Novel Model of SCID-X1 Reconstitution Reveals Predisposition to Retrovirus-induced Lymphoma but No Evidence of γC Gene Oncogenicity

The emergence of leukemia following gene transfer to restore common cytokine receptor gamma chain (gammaC) function in X-linked severe combined immunodeficiency (SCID-X1) has raised important questions with respect to gene therapy safety. To explore the risk factors involved, we tested the oncogenic potential of human gammaC in new strains of transgenic mice expressing the gene under the control of the CD2 promoter and locus control region (LCR). These mice demonstrated mildly perturbed T-cell development, with an increased proportion of thymic CD8 cells, but showed no predisposition to tumor development even on highly tumor prone backgrounds or after gamma-retrovirus infection. The human CD2-gammaC transgene rescued T and B-cell development in gammaC(-/-) mice but with an age-related delay, mimicking postnatal reconstitution in SCID-X1 gene therapy subjects. However, we noted that gammaC(-/-) mice are acutely susceptible to murine leukemia virus (MLV) leukemogenesis, and that this trait was not corrected by the gammaC transgene. We conclude that the SCID-X1 phenotype can be corrected safely by stable ectopic expression of gammaC and that the transgene is not significantly oncogenic when expressed in this context. However, an underlying predisposition conferred by the SCID-X1 background appears to collaborate with insertional mutagenesis to increase the risk of tumor development.

Efficient construction of producer cell lines for a SIN lentiviral vector for SCID-X1 gene therapy by concatemeric array transfection

AbstractRetroviral vectors containing internal promoters, chromatin insulators, and self-inactivating (SIN) long terminal repeats (LTRs) may have significantly reduced genotoxicity relative to the conventional retroviral vectors used in recent, otherwise successful clinical trials. Large-scale production of such vectors is problematic, however, as the introduction of SIN vectors into packaging cells cannot be accomplished with the traditional method of viral transduction. We have derived a set of packaging cell lines for HIV-based lentiviral vectors and developed a novel concatemeric array transfection technique for the introduction of SIN vector genomes devoid of enhancer and promoter sequences in the LTR. We used this method to derive a producer cell clone for a SIN lentiviral vector expressing green fluorescent protein, which when grown in a bioreactor generated more than 20 L of supernatant with titers above 107 transducing units (TU) per milliliter. Further refinement of our technique enabled the rapid generation of whole populations of stably transformed cells that produced similar titers. Finally, we describe the construction of an insulated, SIN lentiviral vector encoding the human interleukin 2 receptor common γ chain (IL2RG) gene and the efficient derivation of cloned producer cells that generate supernatants with titers greater than 5 × 107 TU/mL and that are suitable for use in a clinical trial for X-linked severe combined immunodeficiency (SCID-X1).

Correction of Laminin-5 Deficiency in Human Epidermal Stem Cells by Transcriptionally Targeted Lentiviral Vectors

Deficiency of the basement membrane component laminin-5 (LAM5) causes junctional epidermolysis bullosa (JEB), a severe and often fatal skin adhesion defect. Autologous transplantation of epidermal stem cells genetically corrected with a Moloney leukemia virus (MLV)-derived retroviral vector reconstitutes LAM5 synthesis, and corrects the adhesion defect in JEB patients. However, MLV-derived vectors have genotoxic characteristics, and are unable to reproduce the physiological, basal layer-restricted expression of LAM5 chains. We have developed an alternative gene transfer strategy based on self-inactivating (SIN) or long terminal repeat (LTR)-modified lentiviral vectors, in which transgene expression is under the control of different combinations of promoter-enhancer elements derived from the keratin-14 (K14) gene. Analysis in human keratinocyte cultures and in fully differentiated skin regenerated onto immunodeficient mice showed that gene expression directed by K14 enhancers is tissue-specific and restricted to the basal layer of the epidermis. Transcriptionally targeted lentiviral vectors efficiently transduced clonogenic stem/progenitor cells derived from a skin biopsy of a JEB patient, restored normal synthesis of LAM5 in cultured keratinocytes, and reconstituted normal adhesion properties in human skin equivalents transplanted onto immunodeficient mice. These vectors are therefore an effective, and potentially more safe, alternative to MLV-based retroviral vectors in gene therapy of JEB.Molecular Therapy (2008) 16 12, 1977-1985 doi:10.1038/mt.2008.204.

Partially corrected X-linked severe combined immunodeficiency: long-term problems and treatment options

Rapid progress has been made from the identification of the molecular defects causing X-linked severe combined immune deficiency (X-SCID) to the development of cutting-edge therapeutic approaches such as hematopoietic stem cell transplant and gene therapy for XSCID. Successful treatment of XSCID has created a new population of patients, many of whom are now adolescents and young adults and are facing a variety of chronic problems secondary to partial correction of their underlying disease. This review focuses on the clinical challenges facing these patients (and their caregivers) and provides an overview of some of the treatment options available, including gene therapy.

Reduced Genotoxicity of Avian Sarcoma Leukosis Virus Vectors in Rhesus Long-term Repopulating Cells Compared to Standard Murine Retrovirus Vectors

Insertional mutagenesis continues to be a major concern in hematopoietic stem-cell gene therapy. Nonconventional gene transfer vectors with more favorable integration features in comparison with conventional retrovirus and lentivirus vectors are being developed and optimized. In this study, we report for the first time a systematic analysis of 198 avian sarcoma leukosis virus (ASLV) insertion sites identified in rhesus long-term repopulating cells, and a comparison of ASLV insertion patterns to Moloney murine leukemia virus (MLV) (n = 396) and simian immunodeficiency virus (SIV) (n = 289) using the newly released rhesus genome databank. Despite a weak preference toward gene-coding regions, ASLV integration is nonclustered, does not favor gene-rich regions, transcription start sites, or CpG islands. There was no propensity for ASLV insertions within or near proto-oncogenes, and most importantly, no insertions close to or within the Mds1-Evi1 locus, which is in contrast to the significant over-representation of this insertion site for MLV vectors in the same transplantation model. Furthermore, ASLV long terminal repeats (LTRs) do not have detectable promoter and enhancer activity in a quantitative luciferase assay to measure neighboring gene activation. The combination of these features is unique for ASLV and suggests that optimized vectors based on this virus could be useful and safe for gene transfer to hematopoietic stem cells and progenitor cells.

Stable and Functional Lymphoid Reconstitution in Artemis-deficient Mice Following Lentiviral Artemis Gene Transfer Into Hematopoietic Stem Cells

Patients with mutations in the Artemis gene display a complete absence of T- and B lymphocytes, together with increased cellular radiosensitivity; this leads to a radiosensitive severe combined immunodeficiency (RS-SCID). Allogenic hematopoietic stem-cell (HSC) transplantation is only partially successful in the absence of an human leukocyte antigen-genoidentical donor, and this has prompted a search for alternative therapeutic approaches such as gene therapy. In this study, a self-inactivated lentiviral vector expressing Artemis was used to complement the Artemis knockout mouse (Art(-/-)). Transplantation of Artemis-transduced HSCs into irradiated Art(-/-) mice restored a stable (over a 15-month period of follow-up) and functional T- and cell repertoire that was comparable to that of control mice. The success of secondary transplantations demonstrated that the HSCs had been transduced. One of thirteen mice developed a thymoma 6 months after gene therapy. Although thymic cells were seen to be carrying two lentiviral integration sites, there was no evidence of lentivirus-driven oncogene activation. The Art(-/-) mice were found to be prone to develop T-cell lymphomas, either spontaneously or after irradiation. These data indicate that the observed lymphoproliferation was probably the consequence of the chromosomal instability associated with the Artemis-deficient background. As a whole, our work provides a basis for supporting the gene therapy approach in Artemis-deficient SCID.

Development of an in vitro cell culture assay system for measuring the activation of a neighbouring gene by the retroviral vector

The use of retroviral vectors has shown an actual clinical benefit in a few inherited diseases. However, the occurrence of cases of leukemia after the X-SCID gene therapy trial raised concerns about the safety of insertional mutagenesis inherent to the biology of the retrovirus. Although the retrovirus has long been known to integrate into the host chromosome, and thus have the potential to activate the nearby gene, there has been no convenient method of studying or assaying such a cis-activation phenomenon. In the present study, we report an in vitro assay system in which the effect of retroviral integration on the expression of the neighbouring gene can be studied. In this system, a retroviral vector and the neighbouring reporter gene were constructed in a single plasmid as if it had integrated into the chromosome. Using this assay, we found that the full-length long terminal repeat (LTR) could indeed activate the neighbouring gene expression from a distance and the magnitude of its activation was highly increased when this LTR was placed in the vicinity of the transcription start site of the gene, whereas the truncated LTR exerted little influence. This assay system might provide a useful tool for selecting the appropriate vector structure, as well as for studying the molecular mechanism underlying the cis-activation by the viral LTR.

297. Molecular Analysis of a Severe Adverse Event in the UK SCID-X1 Gene Therapy Clinical Trial

297. Molecular Analysis of a Severe Adverse Event in the UK SCID-X1 Gene Therapy Clinical Trial

An antibody inhibitor of the LMO2-protein complex blocks its normal and tumorigenic functions

The LIM-domain protein LMO2 is a T-cell oncogenic protein first recognized by gene activation through chromosomal translocations, but it is also responsible for leukaemias arising as secondary, adverse effects in an X-SCID gene therapy trial. There are no specific reagents currently available to analyse the LMO2 multiprotein complex or to combat LMO2-dependent leukaemias. Accordingly, we have isolated an anti-LMO2 single chain Fv antibody fragment to determine if intracellular interference with LMO2-protein complexes can avert LMO2-dependent functions in normal and cancer settings. The anti-LMO2 single chain Fv, obtained using Intracellular Antibody Capture (IAC) technology, is specific for LMO2 among the LIM-only protein family and binds LMO2 through the third and fourth LIM fingers. Using vector-mediated expression of anti-LMO2 scFv, we show inhibition of Lmo2-dependent erythropoiesis but not endothelial development. We also demonstrate inhibition of Lmo2-dependent leukaemia in a mouse T-cell tumourigenesis transplantation assay with retroviral-mediated expression of anti-LMO2 scFv. Our studies establish that interference with the LMO2 multiprotein complex inhibits both normal and tumourigenic roles. The antibody fragment is a tool for dissecting LMO2 function in haematopoiesis and leukaemia and is a lead for development of therapeutics against LMO2-dependent T-ALL.

SMaRT Technology Enables Gene Expression Repair in Skin Gene Therapy

In this issue, Wally et al. (2008) report successful gene expression repair by spliceosome-mediated RNA trans-splicing (SMaRT), a novel achievement in molecular medicine. In their model, SMaRT was able to replace a mutation of the plectin gene in epidermolysis bullosa simplex with muscular dystrophy. This approach is particularly attractive for skin gene therapy of dominant-negative mutations present in a number of blistering genodermatoses.

Improved Human β-globin Expression from Self-inactivating Lentiviral Vectors Carrying the Chicken Hypersensitive Site-4 (cHS4) Insulator Element

Effective gene therapy for beta-thalassemia major (beta-TM) requires consistent, high expression of human beta-globin (hbeta-globin) in red blood cells (RBCs). Several groups have now shown that lentiviral (LV) vectors stably transmit the hbeta/hgamma-globin genes and large elements of the locus control region, resulting in correction of the murine thalassemia intermedia (TI) phenotype and survival of mice with the TM phenotype. However, current LVs show variable hbeta/hgamma-globin expression and require a high number of vector copies/cell for a therapeutic effect. To address this, we designed LVs flanked by the chicken hypersensitive site-4 (cHS4) chromatin insulator element and compared them with their "un-insulated" counterparts. We observed a consistent twofold-higher hbeta expression from insulated vectors in single-copy mouse erythroleukemia cell clones, an increase that resulted from reduced position effect variegation (PEV) and increased probability of expression from individual integrants. This effect was confirmed in vivo: an approximately twofold increase in hbeta expression was seen in the RBC progeny of murine hematopoietic stem cells, with significantly higher numbers of hbeta-expressing cells in individual secondary spleen colony-forming units. In summary, cHS4-insulated hbeta-globin LVs showed distinct chromatin barrier activity, resulting in higher, consistent hbeta expression. These studies have important implications for vector design for clinical trials for gene therapy for hemoglobinopathies.

Vector integration is nonrandom and clustered and influences the fate of lymphopoiesis in SCID-X1 gene therapy

Recent reports have challenged the notion that retroviruses and retroviral vectors integrate randomly into the host genome. These reports pointed to a strong bias toward integration in and near gene coding regions and, for gammaretroviral vectors, around transcription start sites. Here, we report the results obtained from a large-scale mapping of 572 retroviral integration sites (RISs) isolated from cells of 9 patients with X-linked SCID (SCID-X1) treated with a retrovirus-based gene therapy protocol. Our data showed that two-thirds of insertions occurred in or very near to genes, of which more than half were highly expressed in CD34(+) progenitor cells. Strikingly, one-fourth of all integrations were clustered as common integration sites (CISs). The highly significant incidence of CISs in circulating T cells and the nature of their locations indicate that insertion in many gene loci has an influence on cell engraftment, survival, and proliferation. Beyond the observed cases of insertional mutagenesis in 3 patients, these data help to elucidate the relationship between vector insertion and long-term in vivo selection of transduced cells in human patients with SCID-X1.

Anti-CD25 Targeted Killing of Bicistronically Transduced Cells: A Novel Safety Mechanism Against Retroviral Genotoxicity

Gene therapy for Fabry disease, a deficiency in alpha-galactosidase A (alpha-gal A) activity, has the potential to provide a cure for the disorder with a single treatment. Despite modifications to existing vectors, concerns have arisen regarding the risk of genotoxicity associated with the use of retroviruses. To address safety concerns, we propose that expression of a cell surface protein, human CD25 (huCD25) in a bicistronic format, with any therapeutic gene such as alpha-gal A can provide a target that can be used to kill transduced cells selectively should transformative events occur. We show that an anti-CD25 antibody and immunotoxin can specifically target and eliminate transduced leukemia cells expressing CD25. In a murine leukemia model, antibody treatment reduced tumor burden 32-fold and increased survival compared with untreated mice. Furthermore, after a bone marrow transplant of therapeutically transduced cells into Fabry mice, antibody treatment reduced the number of retrovirally transduced huCD25-expressing cells in the peripheral blood. A systemic loss of transduced cells with functional consequences was also evident in the liver and spleen. This proof-of-principle study demonstrates that a targeted antibody can reduce tumor burden and selectively clear bicistronically transduced hematopoietic cells that express a target antigen, thus acting as a built-in safety mechanism.