Taihe Lu

Interleukin-7 receptor mutants initiate early T cell precursor leukemia in murine thymocyte progenitors with multipotent potential

Two interleukin-7 receptor mutants identified in human early T cell precursor leukemia are sufficient to induce disease in mice when expressed in primitive, Arf-null thymocytes.

Self-Selection by Genetically Modified Committed Lymphocyte Precursors Reverses the Phenotype of JAK3-Deficient Mice without Myeloablation

Janus kinase 3 (JAK3) is an essential component of cytokine receptor signal transduction pathways required for normal lymphocyte development and function. JAK3 deficiency in both mice and humans results in severe combined immunodeficiency (SCID) and increased susceptibility to opportunistic infections. We have previously shown that JAK3 gene transfer into irradiated recipients could restore immune function. However, since this toxic conditioning would be undesirable for infants in a clinical application, we have tested whether immune function could be restored in nonmyeloablated JAK3-deficient (-/-) mice. Murine JAK3 retroviral vectors were transduced into hematopoietic stem cells from the livers of newborn JAK3(-/-) mice. These cells were then injected intraperitoneally into nonirradiated JAK3(-/-) neonates. Transduced cells were detectable in these mice at time points 4 to 6 months after injection and resulted in significant correction of T and B lymphocyte numbers and circulating immunoglobulin (Ig) levels. After immune challenge with a dose of influenza A virus that was lethal to nonmanipulated JAK3(-/-) mice, mice injected with transduced cells showed development of circulating virus-specific IgG and enhanced survival. This work shows that the large selective advantage for JAK3-corrected lymphoid cells may be sufficient to overcome the need for myeloablative conditioning in JAK3 gene therapy protocols.

Functional interactions between Lmo2, the Arf tumor suppressor, and Notch1 in murine T-cell malignancies

LMO2 is a target of chromosomal translocations in T-cell tumors and was activated by retroviral vector insertions in T-cell tumors from X-SCID patients in gene therapy trials. To better understand the cooperating genetic events in LMO2-associated T-cell acute lymphoblastic leukemia (T-ALL), we investigated the roles of Arf tumor suppressor loss and Notch activation in murine models of transplantation. Lmo2 overexpression enhanced the expansion of primitive DN2 thymocytes, eventually facilitating the stochastic induction of clonal CD4(+)/CD8(+) malignancies. Inactivation of the Arf tumor suppressor further increased the self-renewal capacity of the primitive, preleukemic thymocyte pool and accelerated the development of aggressive, Lmo2-induced T-cell lympholeukemias. Notch mutations were frequently detected in these Lmo2-induced tumors. The Arf promoter was not directly engaged by Lmo2 or mutant Notch, and use of a mouse model in which activation of a mutant Notch allele depends on previous engagement of the Arf promoter revealed that Notch activation could occur as a subsequent event in T-cell tumorigenesis. Therefore, Lmo2 cooperates with Arf loss to enhance self-renewal in primitive thymocytes. Notch mutation and Arf inactivation appear to independently cooperate in no requisite order with Lmo2 overexpression in inducing T-ALL, and all 3 events remained insufficient to guarantee immediate tumor development.

Hemgn is a direct transcriptional target of HOXB4 and induces expansion of murine myeloid progenitor cells

HOXB4, a member of the Homeobox transcription factor family, promotes expansion of hematopoietic stem cells and hematopoietic progenitor cells in vivo and ex vivo when overexpressed. However, the molecular mechanisms underlying this effect are not well understood. To identify direct target genes of HOXB4 in primary murine hematopoietic progenitor cells, we induced HOXB4 function in lineage-negative murine bone marrow cells, using a tamoxifen-inducible HOXB4-ER(T2) fusion protein. Using expression microarrays, 77 probe sets were identified with differentially changed expression in early response to HOXB4 induction. Among them, we show that Hemogen (Hemgn), encoding a hematopoietic-specific nuclear protein of unknown function, is a direct transcriptional target of HOXB4. We show that HOXB4 binds to the promoter region of Hemgn both ex vivo and in vivo. When we overexpressed Hemgn in bone marrow cells, we observed that Hemgn promoted cellular expansion in liquid cultures and increased self-renewal of myeloid colony-forming units in culture, partially recapitulating the effect of HOXB4 overexpression. Furthermore, down-regulation of Hemgn using an shRNA strategy proved that Hemgn contributes to HOXB4-mediated expansion in our myeloid progenitor assays. Our results identify a functionally relevant, direct transcriptional target of HOXB4 and identify other target genes that may also participate in the HOXB4 genetic network.

Reduction in Hematopoietic Stem Cell Numbers with in Vivo Drug Selection can be Partially Abrogated by HOXB4 Gene Expression

In vivo selection of hematopoietic stem cells (HSCs) offers an approach to enrichment of genetically modified blood cells in the context of gene therapy for blood disorders. We have previously demonstrated efficient HSC selection in mice using retroviral vectors expressing dihydrofolate reductase (DHFR) or methylguanine methyltransferase (MGMT) drug resistance genes. In this study, we examined whether drug selection was followed by subsequent HSC regeneration and, if not, whether regeneration could be augmented by enforced expression of HOXB4, which has previously been shown to enhance HSC regeneration after transplant. Using a murine competitive repopulation model, we found that selection using either the DHFR or the MGMT system was accompanied by a significant overall reduction in repopulating activity in secondary transplant assays, although hematopoiesis remained normal after recovery. Inclusion of a HOXB4 expression cassette in the DHFR vector resulted in a partial restoration of HSC numbers following selection and was associated with an increase in HSC selection efficiency. These results illustrate that while drug resistance vectors can protect transduced HSC from cytotoxic drugs, the self-renewal capacity of transduced HSCs is limited following in vivo selection. Strategies that increase self-renewal capacity could increase the efficiency and safety of in vivo selection.

Mouse Transplant Models for Evaluating the Oncogenic Risk of a Self-Inactivating XSCID Lentiviral Vector

Hematopoietic stem cell gene therapy requires the use of integrating retroviral vectors in order to stably transmit a therapeutic gene to mature blood cells. Human clinical trials have shown that some vector integration events lead to disrupted regulation of proto-oncogenes resulting in disordered hematopoiesis including T-cell leukemia. Newer vectors have been designed to decrease the incidence of these adverse events but require appropriate pre-clinical assays to demonstrate safety. We have used two distinct mouse serial transplant assays to evaluate the safety of a self-inactivating lentiviral vector intended for use in X-linked severe combined immunodeficiency (XSCID) gene therapy trials. These experiments entailed 28 months of total follow-up and included 386 mice. There were no cases in which the XSCID lentiviral vector clearly caused hematopoietic malignancies, although a single case of B cell malignancy was observed that contained the lentiviral vector as a likely passenger event. In contrast, a SFFV-DsRed γ-retroviral vector resulted in clonal transformation events in multiple secondary recipients. Non-specific pathology not related to vector insertions was noted including T cell leukemias arising from irradiated recipient cells. Overall, this comprehensive study of mouse transplant safety assays demonstrate the relative safety of the XSCID lentiviral vector but also highlight the limitations of these assays.

Evaluating the Safety of Retroviral Vectors Based on Insertional Oncogene Activation and Blocked Differentiation in Cultured Thymocytes.

Insertional oncogenesis due to retroviral (RV) vector integration has caused recurrent leukemia in multiple gene therapy trials, predominantly due to vector integration effects at the LMO2 locus. While currently available preclinical safety models have been used for evaluating vector safety, none have predicted or reproduced the recurrent LMO2 integrations seen in previous X-linked severe combined immunodeficiency (X-SCID) and Wiskott–Aldrich clinical gene therapy trials. We now describe a new assay for assessing vector safety that recapitulates naturally occurring insertions into Lmo2 and other T-cell proto-oncogenes leading to a preleukemic developmental arrest in primary murine thymocytes cultured in vitro. This assay was used to compare the relative oncogenic potential of a variety of gamma-RV and lentiviral vectors and to assess the risk conferred by various transcriptional elements contained in these genomes. Gamma-RV vectors that contained full viral long-terminal repeats were most prone to causing double negative 2 (DN2) arrest and led to repeated cases of Lmo2 pathway activation, while lentiviral vectors containing these same elements were significantly less prone to activate proto-oncogenes or cause DN2 arrest. This work provides a new preclinical assay that is especially relevant for assessing safety in SCID disorders and provides a new tool for designing safer RV vectors.

571. Safe Harbor Targeting of IL2RG Expression Cassettes for Gene Therapy of X-Linked Severe Combined Immunodeficiency (SCID-X1)

SCID-X1 is caused by mutations in the IL2RG gene and results in severe defects in T, B, and NK-cell mediated immunity. Transduction of hematopoietic stem and progenitor cells with IL2RG-expressing gamma-retroviral vectors can be clinically effective but has also caused leukemia in several trials due to vector integrations that activate T-cell proto-oncogenes. One approach to eliminate these genotoxic integrations is to use genome editing techniques to insert an IL2RG transgene into a genomic safe harbor loci. The choice of a genomic safe harbor rather than the endogenous IL2RG locus offers several advantages including correction of a wide variety of mutations with a single therapeutic cassette, potentially higher editing efficiency in selected safe harbor sites, and the ability to adapt the specific nuclease reagents to other diseases and therapeutic templates. We are testing two different IL2RG-expression cassettes for targeting and expression into the AAVS1 locus. First, we generated a recombination template that contains a codon-optimized IL2RG cDNA driven by the short elongation factor alpha (EF1a) promoter based on clinical data with this cassette being transmitted via a lentiviral vector and showing immune correction in human SCID-X1 subjects in an ongoing trial at the NIH Clinical Center. We first tested whether a sufficient level of IL2RG expression can be achieved when a single copy of the EF1α-IL2RG-cDNA cassette was inserted into the AAVS1 locus in human ED7R T-cells. This transgene cassette was flanked by 500bp homologous regions from the AAVS1 locus and transfected into ED7R cells, along with a pair of TALEN endonucleases that cleave within the AAVS1 locus. Three weeks after transfection, a distinct 7% subpopulation of cells expressed IL2RG on the cell surface, which was subsequently sorted by flow cytometry. We analyzed 10 single cell subclones from this population and confirmed that homologous recombination and single allele targeting had occurred in each of these clones, demonstrating efficient homologous recombination at the AAVS1 site. Flow analysis showed that the EF1α-IL2RG-cDNA cassette was expressed at approximately the same level as that seen in cells transduced with a single copy of the clinical EF1α-IL2RG-cDNA lentiviral vector, verifying that the AAVS1 is permissive for therapeutic expression levels at the single copy level. One potential disadvantage of the safe harbor approach is the potential loss of precise regulation of gene expression resulting from the use of such synthetic cDNA constructs. For this reason, we are also currently targeting the full 5.3 kb genomic IL2RG locus and endogenous promoter to the AAVS1 safe harbor to compare their IL2RG expression level with that obtained from the cDNA construct. We are also comparing editing efficiency at AAVS1 using several TALEN pairs as well as 4 Staphylococcus Aureus Cas9 guide RNAs, which are deliverable by “all-in-one” single stranded AAV6 vectors. To date, the best guide has led to 37 % allele editing in 293T cells when transfected as plasmid. We are currently producing AAV vectors that contain this guide RNA and will test this AAV vector, along with IDLV vectors that will transfer either the cDNA or genomic IL2RG templates, for editing efficiency and IL2RG gene expression in ED7R and human CD34+ cells.