Use of retroviral vectors for gene transfer and expression

Posttranscriptional regulation of hepatitis B virus replication by the precore protein

Hepadnaviruses encode two core-related open reading frames. One directs the synthesis of the p21 core protein, which subsequently becomes a structural component of the viral nucleocapsid. The other produces a p25 precore protein that is targeted by a signal peptide to a cell secretory pathway where N-terminal processing will create a p22 species. This molecule will be further modified at the C-terminal region to generate p17, and the truncated protein is secreted from the cell as hepatitis B e antigen (HBeAg). The function of the precore gene in the biology of hepadnaviruses is unknown. We found that ablation of the precore gene resulted in the generation of a hepatitis B virus (HBV) species with a high-replication-level phenotype. More important, expression in trans of physiologic levels of p25 restored viral replication to wild-type levels. Moreover, transient or stable overexpression of the precore gene resulted in striking inhibition of HBV replication. The molecular species responsible for this viral inhibitory effect was identified as the p22 nonsecreted HBeAg precursor protein. By sucrose gradient sedimentation analysis, we determined that expression of p22 leads to the formation of nucleocapsids similar to those made with wild-type p21 core protein. Immunoprecipitation experiments revealed that the p21 and p22 physically interact and form hybrid nucleocapsid structures devoid of pregenomic viral RNA. These experiments suggest that expression of the precore gene may be important in the regulation of HBV replication and describe a possible molecular mechanism(s) for this effect.

Ribonucleotide reductase R2 component is a novel malignancy determinant that cooperates with activated oncogenes to determine transformation and malignant potential

Ribonucleotide reductase is a highly regulated cell cycle-controlled activity that is essential for DNA synthesis and repair. A retroviral vector for the R2 component of mammalian ribonucleotide reductase, the rate-limiting protein for enzyme activity and DNA synthesis in proliferating cells, was constructed and introduced into mammalian cells. Expression of Myc epitope-tagged R2 protein in benign BALB/c 3T3 and NIH 3T3 cells leads to a greatly increased frequency of focus formation in cooperation with H-ras transformation. Four lines of H-ras-transformed mouse 10T1/2 fibroblasts showed increased growth efficiency in soft agar after infection with the recombinant R2 expression virus vector. Furthermore, cells with altered R2 expression also exhibited significantly reduced subcutaneous tumor latency and increased tumor growth rates in syngeneic mice, and showed markedly elevated metastatic potential in lung metastasis assays. The results indicate that altered R2 gene expression cooperates with ras in mechanisms of malignant progression. A major Ras pathway involves the Raf-1 protein, which is recruited to the plasma membrane for activation. We show that recombinant R2 expression leads to significant increases in membrane-associated Raf-1 protein and mitogenactivating protein kinase-2 activity suggesting a mechanism for the observed Ras/R2 synergism. In support of this finding, we observed that activated Rac-1, which operates parallel to Raf-1 and cooperates with Raf-1 in Ras activated pathways, also cooperates with R2 in cellular transformation. These studies demonstrate that the R2 protein can participate in other critical cellular functions in addition to ribonucleotide reduction, and that deregulated R2 is a novel tumor progressor determinant that cooperates in oncogene-mediated mechanisms, which control malignant potential.

Complete restoration of glucocerebrosidase deficiency in Gaucher fibroblasts using a bicistronic MDR retrovirus and a new selection strategy

Retrovirus-mediated gene transfer is currently the most common method for the application of genetic therapy to cancer and many inherited and acquired disorders. Here we report the generation of an amphotropic producer cell line (CA2) that synthesizes viral particles carrying a bicistronic cassette in which the selectable MDR1 cDNA encoding P-glycoprotein (P-gp) a multidrug efflux pump, and the human glucocerebrosidase (GC) gene are transcriptionally fused. Transduction of human Gaucher fibroblasts with this recombinant virus allowed coordinate expression of P-gp and GC. Treatment of the transduced fibroblasts with various cytotoxic substrates of P-gp selected for cells with increased expression of GC, which paralleled the stringency of drug selection. Thus, selection of the genetically modified Gaucher fibroblasts in 1 microgram/ml colchicine raised their GC activity levels from nearly undetectable to those present in WI-38 normal human fibroblasts, correcting the enzyme deficiency present in Gaucher cells. Moreover, by simultaneously inhibiting the P-gp pump, it was possible to use much lower concentrations of colchicine to select for high-level expression of MDR1 and GC. Thus, selection with colchicine at 5 ng/ml in combination with the P-gp inhibitors verapamil or PSC 833 produced a complete correction of the GC deficiency in the CA2-transduced fibroblasts. These combination regimens, already in clinical use for the treatment of multidrug-resistant malignancies, may prove useful in gene therapy trials when utilized for high level selection of a nonselectable gene such as glucocerebrosidase when transcriptionally fused to the MDR1 gene.

Construction of retroviral vectors to induce strong hepatoma cell-specific expression of cytokine genes

Continuing advances in molecular biology have provided tools for a promising approach to the treatment of cancer. Among the various strategies of gene therapy for cancer, many are aimed at killing tumour cells indirectly by the induction or reinforcement of a host immune response by gene transduction of various cytokines, major histocompatibility complex or immune accessory molecules. In the present study, we selected the tumour necrosis factor-alpha, interleukin-2 and interleukin-3 genes as potential cytokine genes to induce antitumour effects. We constructed retroviral vectors carrying these cytokine genes under the control of the murine albumin enhancer and promoter and retrovirally transduced these genes into hepatoma and non-hepatoma cell lines. Strong expression of the cytokine genes was induced in transduced hepatoma cells, while no evident expression was detected in transduced non-hepatoma cells. These results demonstrate the hepatoma-specific expression of cytokine genes and imply the feasibility of in vivo gene transfer into hepatomas without affecting any other tissues. Furthermore, these cytokine genes were expressed much more intensively when they were derived from the albumin enhancer and promoter than when derived from the simian virus 40 early region promoter. These results indicate that transcriptional regulatory sequences specific for the target tissues could be preferable to viral promoters for the gene therapy of cancer.

Endothelial cell death induced by tumor necrosis factor-alpha is inhibited by the Bcl-2 family member, A1

Endothelial cells play a central role in the inflammatory process. Tumor necrosis factor-alpha (TNF) is a multifunctional cytokine which elicits many of the inflammatory responses of endothelial cells. While TNF directly causes apoptosis of tumor cells and virally infected cells, normal cells are generally resistant. However, most resistant cells, including human endothelial cells, can be rendered susceptible to TNF by inhibiting RNA or protein synthesis. This finding suggests that TNF provides a cell survival signal in addition to a death signal. We have previously cloned a human Bcl-2 homologue, A1, and shown that it is specifically induced by proinflammatory cytokines but not by endothelial growth factors. In this study, we show that retroviral-mediated transfer of the A1 cDNA to a human microvascular endothelial cell line provides protection against cell death initiated by TNF in the presence of actinomycin D. The induction of A1 by TNF in this system is mediated via a protein kinase C pathway. Since TNF signaling has also been shown to proceed via ceramides, we tested whether exogenous ceramides could induce A1. Our findings indicate that ceramides do not induce A1 but do up-regulate c-jun and induce endothelial death. Ceramide-activated endothelial death is also inhibited by A1, suggesting that TNF may initiate divergent survival and death pathways via separate lipid second messengers.

Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector

We describe the construction of a safe, replication-defective and efficient lentiviral vector suitable for in vivo gene delivery. The reverse transcription of the vector was found to be a rate-limiting step; therefore, promoting the reaction inside the vector particles before delivery significantly enhanced the efficiency of gene transfer. After injection into the brain of adult rats, sustained long-term expression of the transgene was obtained in the absence of detectable pathology. A high proportion of the neurons in the areas surrounding the injection sites of the vector expressed the transduced beta-galactosidase gene. This pattern was invariant in animals sacrificed several months after a single administration of the vector. Transduction occurs by integration of the vector genome, as it was abolished by a single amino acid substitution in the catalytic site of the integrase protein incorporated in the vector. Development of clinically acceptable derivatives of the lentiviral vector may thus enable the sustained delivery of significant amounts of a therapeutic gene product in a wide variety of somatic tissues.

Cell-surface receptors for retroviruses and implications for gene transfer

Retroviruses can utilize a variety of cell-surface proteins for binding and entry into cells, and the cloning of several of these viral receptors has allowed refinement of models to explain retrovirus tropism. A single receptor appears to be necessary and sufficient for entry of many retroviruses, but exceptions to this simple model are accumulating. For example, HIV requires two proteins for cell entry, neither of which alone is sufficient; 10A1 murine leukemia virus can enter cells by using either of two distinct receptors; two retroviruses can use different receptors in some cells but use the same receptor for entry into other cells; and posttranslational protein modifications and secreted factors can dramatically influence virus entry. These findings greatly complicate the rules governing retrovirus tropism. The mechanism underlying retrovirus evolution to use many receptors for cell entry is not clear, although some evidence supports a mutational model for the evolution of new receptor specificities. Further study of factors that govern retrovirus entry into cells are important for achieving high-efficiency gene transduction to specific cells and for the design of retroviral vectors to target additional receptors for cell entry.

Endogenous murine leukemia virus DNA sequences in murine cell lines: implications for gene therapy safety testing by PCR

Safety testing for replication-competent retrovirus (RCR) is an important requirement in gene transfer clinical trials using retroviral vectors. A sensitive polymerase chain reaction (PCR) method is one approach to RCR detection. Only in the presence of RCR will the pol-env encoding sequences, necessary for viral replication and packaging, be amplified from proviral DNA in infected indicator cells. To avoid false-positive results in this assay it is crucial that indicator cell lines are free of endogenous retroviral sequences that could also be amplified with pol-env PCR primers. We screened candidate murine indicator cell lines and determined that while Mus dunni is free of detectable pol-env sequences, endogenous retroviral sequences do indeed exist in several cell lines and lead to false-positive results in the PCR assay for RCR. Furthermore, these endogenous retroviral sequences are expressed as RNA transcripts in NIH 3T3 and SC-1 cell lines, as determined by PCR amplification of cDNA but, nevertheless, do not give rise to replication-competent particles. We recognize the potential for murine cell lines to undergo spontaneous rearrangements of endogenous viral sequences in culture and give rise to recombinants containing newly acquired contiguous pol-env sequences. Indicator cell lines should thus be carefully selected and monitored on an ongoing basis when used in safety testing using PCR approaches for the detection of RCR.

Recombinant adeno-associated virus-mediated high-efficiency, transient expression of the murine cationic amino acid transporter (ecotropic retroviral receptor) permits stable transduction of human HeLa cells by ecotropic retroviral vectors

Adeno-associated virus has a broad host range, is nonpathogenic, and integrates into a preferred location on chromosome 19, features that have fostered development of recombinant adeno-associated viruses (rAAV) as gene transfer vectors for therapeutic applications. We have used an rAAV to transfer and express the murine cationic amino acid transporter which functions as the ecotropic retroviral receptor, thereby rendering human cells conditionally susceptible to infection by an ecotropic retroviral vector. The proportion of human HeLa cells expressing the receptor at 60 h varied as a function of the multiplicity of infection (MOI) with the rAAV. Cells expressing the ecotropic receptor were efficiently transduced with an ecotropic retroviral vector encoding a nucleus-localized form of beta-galactosidase. Cells coexpressing the ecotropic receptor and nucleus-localized beta-galactosidase were isolated by fluorescence-activated cell sorting, and cell lines were recovered by cloning at limiting dilution. After growth in culture, all clones contained the retroviral vector genome, but fewer than 10% (3 of 47) contained the rAAV genome and continued to express the ecotropic receptor. The ecotropic receptor coding sequences in the rAAV genome were under the control of a tetracycline-modulated promoter. In the presence of tetracycline, receptor expression was low and the proportion of cells transduced by the ecotropic retroviral vector was decreased. Modulation of receptor expression was achieved with both an episomal and an integrated form of the rAAV genome. These data establish that functional gene expression from an rAAV genome can occur transiently without genome integration.

Clonal analysis of stably transduced human epidermal stem cells in culture

We have transduced normal human keratinocytes with retroviral constructs expressing a bacterial beta-galactosidase (beta-gal) gene or a human interleukin-6 (hIL-6) cDNA under control of a long terminal repeat. Efficiency of gene transfer averaged approximately 50% and 95% of clonogenic keratinocytes for beta-gal and hIL-6, respectively. Both genes were stably integrated and expressed for more than 150 generations. Clonal analysis showed that both holoclones and their transient amplifying progeny expressed the transgene permanently. Southern blot analysis on isolated clones showed that many keratinocyte stem cells integrated multiple proviral copies in their genome and that the synthesis of the exogenous gene product in vitro was proportional to the number of proviral integrations. When cohesive epidermal sheets prepared from stem cells transduced with hIL-6 were grafted on athymic animals, the serum levels of hIL-6 were strictly proportional to the rate of secretion in vitro and therefore to the number of proviral integrations. The possibility of specifying the level of transgene expression and its permanence in a homogeneous clone of stem cell origin opens new perspectives in the long-term treatment of genetic disorders.

Chemoprotection of normal tissues by transfer of drug resistance genes

The effectiveness of many types of antitumour agent is limited by (i) acute dose limiting cytotoxicity, principally myelosuppression but also lung, liver and gastrointestinal tract toxicity, (ii) the risk of therapy related secondary malignancy and (iii) the inherent or acquired drug-resistance of tumour cells. As the management of the acute toxic effects improve, the more insidious effects, and particularly haematological malignancies, are anticipated to increase. Furthermore, attempts to overcome tumour cell resistance to treatment can lead to increased collateral damage in normal tissues. One approach to circumventing both the acute toxic and chronic carcinogenic effects of chemotherapy would be to use gene therapy to achieve high levels of expression of drug resistance proteins in otherwise drug-sensitive tissues. To date the products of the multi-drug resistance (MDR-1) and the human O6-alkylguanine-DNA-alkyltransferase (ATase) gene have been used in preclinical experiments to demonstrate proof of principle, and the former of these is now being tested in a clinical situation. Here we discuss the potential of drug-resistance gene therapy to provide chemoprotection to normal tissues and examine the prospects for a dual approach which combines this with pharmacological sensitisation of tumours to chemotherapeutic agents.

Retrovirus packaging cells based on 10A1 murine leukemia virus for production of vectors that use multiple receptors for cell entry

10A1 murine leukemia virus can enter cells by using either of two different cell surface phosphate transport proteins, the gibbon ape leukemia virus receptor Glvr-1 (Pit-1) or the amphotropic retrovirus receptor Ram-1 (Pit-2). Glvr-1 and Ram-1 are widely expressed in different tissues, but the relative amounts of each are highly variable. We have developed retrovirus packaging cell lines based on 10A1 virus to take advantage of this dual receptor utilization to improve gene transfer rates in somatic cells of animals and humans, in which the relative levels of the two receptors are not always known. Optimization of the Env expression vector allowed the generation of packaging lines that produce helper-free vector titers up to 10(7)/ml. By interference analysis, we found that a 10A1 pseudotype retroviral vector can utilize Ram-1 for efficient entry into mouse, rat, and human cells and can utilize Glvr-1 for entry into mouse and human cells but not for entry into rat cells. The 10A1 pseudotype vector efficiently enters mouse cells by using Glvr-1, while entry into human cells is much less efficient. Thus, the 10A1 pseudotype packaging cells may be advantageous compared with the standard amphotropic packaging cells because vectors produced by the cells can use an additional receptor for cell entry. These packaging cells will also be useful to further explore the complicated pattern of receptor usage conferred by the 10A1 viral surface protein.

Cardiovascular gene therapy: current concepts

Gene therapy techniques are under development for many areas of medicine, including cardiovascular disease. Identifying appropriate gene targets will require more detailed knowledge of the molecular pathophysiology of these disorders, and choosing appropriate vectors and delivery systems will contribute significantly to the challenge of developing this approach for clinical use. The concepts of toxicology and therapeutic drug monitoring will need to be broadened to account for the unique chemical, biological, and genetic characteristics of gene therapeutic agents. This review will provide an overview of strategy development, currently available vectors, and examples of their application to cardiovascular gene transfer. Considerations of the potential toxicities associated with particular vectors and delivery systems, as well as the types of genetic modifications possible, will provide some guidelines regarding appropriate monitoring of their clinical application.

Improved retroviral vectors for hematopoietic stem cell protection and in vivo selection

Therapeutic gene transfer into hematopoietic cells is critically dependent on the evolution of methods that allow ex vivo expansion, high-frequency transduction, and selection of gene-modified long-term repopulating cells. Progress in this area needs elaboration of defined culture and transduction conditions for long-term repopulating cells and improvement of gene transfer systems. We have optimized retroviral vector constructions based on murine leukemia viruses (MuLV) to overcome the transcriptional repression encountered with the use of conventional Moloney MuLV (MoMuLV) vectors in early hematopoietic progenitor cells (HPC). Novel retroviral vectors, termed FMEV (for Friend-MCF/MESV hybrid vectors), were cloned that mediate greatly improved gene expression in the myeloerythroid compartment. Transfer of the selectable marker multidrug resistance 1 (mdr1), FMEV, in contrast to conventional MoMuLV-related vectors currently in use for clinical protocols, mediated background-free selectability of transduced human HPC in the presence of myeloablative doses of the cytostatic agent paclitaxel in vitro. Furthermore, FMEV also greatly improved chemo-protection of hematopoietic progenitor cells in a murine model system in vivo. Finally, when a second gene was transferred along with mdr1 in an FMEV-backbone, close to 100% coexpression was observed in multidrug-resistant colonies. These observations have significant consequences for a number of ongoing and planned gene therapy trials, for example, stem cell protection to reduce the myelotoxic side effects of anticancer chemotherapy, correction of inherited disorders involving hematopoietic cells, and antagonism of HIV infection.

The pCL vector system: rapid production of helper-free, high-titer, recombinant retroviruses

We describe the construction and characterization of retroviral vectors and packaging plasmids that produce helper-free retrovirus with titers of 1 X 10(6) to 5 X 10(6) within 48 h. These vectors contain the immediate early region of the human cytomegalovirus enhancer-promoter fused to the Moloney murine leukemia virus long terminal repeat at the TATA box in the 5' U3 region, yielding the pCL promoter. By selecting vectors designed to express genes from one of four promoters (dihydrofolate reductase, Rous sarcoma virus, long terminal repeat, or cytomegalovirus), the pCL system permits the investigator to control the level of gene expression in target cells over a 100-fold range, while maintaining uniformly high titers of virus from transiently transfected producer cells. The pCL packaging plasmids lack a packaging signal (delta-psi) and include an added safety modification that renders them self-inactivating through the deletion of the 3' U3 enhancer. Ecotropic, amphotropic (4070A), and amphotropic-mink cell focus-forming hybrid (10A1) envelope constructions have been prepared and tested, permitting flexible selection of vector pseudotype in accordance with experimental needs. Vector supernatants are free of helper virus and are of sufficiently high titer within 2 days of transient transfection in 293 cells to permit infection of more than 50% of randomly cycling target cells in culture. We demonstrated the efficacy of these vectors by using them to transfer three potent cell cycle control genes (the p16(INK4A), p53, and Rb1 genes) into human glioblastoma cells.

Prolonged expression of therapeutic levels of human granulocyte colony-stimulating factor in rats following gene transfer to skeletal muscle

Gene transfer to skeletal muscle was examined as a means of gene therapy for neutropenia. A recombinant retrovirus containing a human granulocyte colony-stimulating factor (G-CSF) gene was introduced into primary human or rat myoblasts, which were then shown to produce biologically active G-CSF. Transplantation of G-CSF-producing rat myoblasts into the muscle of syngeneic rats resulted in a 15-fold increase in absolute neutrophil counts. This increase correlated with detection of circulating human G-CSF protein throughout the 6-month duration of the experiment. These results clearly demonstrate long-term production of therapeutically relevant amounts of a human protein by normal cells in vivo.

Regulation of Btk by Src family tyrosine kinases

Loss of function of Bruton's tyrosine kinase (Btk) results in X-linked immunodeficiencies characterized by a broad spectrum of signaling defects, including those dependent on Src family kinase-linked cell surface receptors. A gain-of-function mutant, Btk*, induces the growth of fibroblasts in soft agar and relieves the interleukin-5 dependence of a pre-B-cell line. To genetically define Btk signaling pathways, we used a strategy to either activate or inactivate Src family kinases in fibroblasts that express Btk*. The transformation potential of Btk* was dramatically increased by coexpression with a partly activated c-Src mutant (E-378 --> G). This synergy was further potentiated by deletion of the Btk Src homology 3 domain. Downregulation of Src family kinases by the C-terminal Src kinase (Csk) suppressed Btk* activation and biological potency. In contrast, kinase-inactive Csk (K-222 --> R), which functioned as a dominant negative molecule, synergized with Btk* in biological transformation. Activation of Btk* correlated with increased phosphotyrosine on transphosphorylation and autophosphorylation sites. These findings suggest that the Src and Btk kinase families form specific signaling units in tissues in which both are expressed.

Matrix metalloproteinase 2 (gelatinase A) regulates glomerular mesangial cell proliferation and differentiation

A biologic role for the 72-kDa gelatinase A (matrix metalloproteinase 2; MMP-2), beyond simple extracellular matrix turnover, was evaluated in glomerular mesangial cells. To determine the significance of MMP-2 secretion for the acquisition of the inflammatory phenotype, we reduced the constitutive secretion of MMP-2 by cultured mesangial cells with antisense RNA expressed by an episomally replicating vector or with specific anti-MMP-2 ribozymes expressed by a retroviral transducing vector. The phenotype of the transfected, or retrovirally infected, cells was profoundly altered from the activated state and closely approximated that of quiescent cells in vivo. The prominent differences included a change in the synthesis and organization of the extracellular matrix, loss of activation markers, and a virtually total exit from the cell cycle. Reconstitution with exogenous active, but not latent MMP-2, induced a rapid return to the inflammatory phenotype in vitro. This effect was specific to MMP-2, because the closely related MMP-9 did not reproduce these changes. Furthermore, this pro-inflammatory effect of MMP-2 is dependent upon the active form of the enzyme, which can be produced by an autocatalytic activation process on the mesangial cell plasma membrane. It is concluded that MMP-2 acts directly upon mesangial cells to permit the development of an inflammatory phenotype. Specific inhibition of MMP-2 activity in vivo may represent an alternate means of ameliorating complex inflammatory processes by affecting the phenotype of the synthesizing cells, per se.

In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector

A retroviral vector system based on the human immunodeficiency virus (HIV) was developed that, in contrast to a murine leukemia virus-based counterpart, transduced heterologous sequences into HeLa cells and rat fibroblasts blocked in the cell cycle, as well as into human primary macrophages. Additionally, the HIV vector could mediate stable in vivo gene transfer into terminally differentiated neurons. The ability of HIV-based viral vectors to deliver genes in vivo into nondividing cells could increase the applicability of retroviral vectors in human gene therapy.

Retrovirus-mediated gene transfer into salivary glands in vivo

In the present report, we show prolonged expression of beta-galactosidase (beta-Gal) in the acinar cells of the submandibular and sublingual glands of rats following retrograde ductal injection of the retroviral vector BAG. To facilitate integration of viral DNA, cell division in the gland was induced by the administration of the beta-adrenergic agonist isoproterenol prior to the delivery of the vector. The frequency of cells stained for beta-Gal was higher if the virus was injected 4-20 hr after the two injections of isoproterenol given 24 hr apart than after the injection of only one dose of the drug. Without stimulation of cell division, no integration of the viral DNA was observed. Expression of the marker enzyme was observed up to 43 days, the limit of the observation period. The data indicate that salivary glands are potential targets of retrovirus-mediated gene transfer for somatic gene therapy.

A link between ferritin gene expression and ribonucleotide reductase R2 protein, as demonstrated by retroviral vector mediated stable expression of R2 cDNA

We have constructed a retroviral expression vector for the mammalian ribonucleotide reductase R2 component. Stable infectants, which express a myc epitope tagged R2 protein from the vector cDNA were obtained and described for the first time. Cells containing the recombinant protein exhibited increased ribonucleotide reductase activity, and were resistant to the antitumour agent hydroxyurea, which targets the R2 component of ribonucleotide reductase. Furthermore, a direct link between ferritin gene expression and R2 protein was observed, since cells containing vector expressed recombinant R2 protein exhibited increased H-chain and L-chain ferritin gene expression.

Loss of expression of a retrovirus-transduced gene in human keratinocytes

Retroviral-mediated transfer of new genetic information into keratinocytes is a key step in epidermal gene therapy. An obstacle to the use of retroviruses for gene therapy is that although high levels of expression of the transduced gene can be maintained in tissue culture, expression is often lost when the cells are transplanted to an animal host. To examine some of the factors involved in this instability of expression, we transduced keratinocytes with a retrovirus encoding the gene for human factor IX and monitored secretion of the transduced gene. We observed continued secretion of factor IX through five passages in culture. When, however, sheets of these cells were grafted to athymic mice, factor IX expression was reduced or lost within 6 wk. We show that the reduction of factor IX expression in grafted keratinocytes did not result from a loss of grafted cells, nor was there a block to systemic delivery of a secreted endogenous product.

Sindbis virus DNA-based expression vectors: utility for in vitro and in vivo gene transfer

Several DNA-based Sindbis virus vectors were constructed to investigate the feasibility and potential applications for initiating the virus life cycle in cells transfected directly with plasmid DNA. These vectors, when transfected into mammalian cells, have been used to produce virus, to express heterologous genes, and to produce infectious vector particles. This approach involved the conversion of a self-replicating vector RNA (replicon) into a layered DNA-based expression system. The first layer includes a eukaryotic RNA polymerase II expression cassette that initiates nuclear transcription of an RNA which corresponds to the Sindbis virus vector replicon. Following transport of this RNA from the nucleus to the cytoplasm, the second layer, autocatalytic amplification of the vector, proceeds according to the Sindbis virus replication cycle and results in expression of the heterologous gene. The Sindbis virus DNA vectors expressed reporter genes in transfected cells at levels that were comparable to those of in vitro-transcribed RNA replicons and were approximately 10-fold higher than the levels produced by conventional RNA polymerase II-dependent plasmids in which the promoter and reporter gene were linked directly. Reporter gene expression was also observed in rodent muscle following injection with Sindbis virus DNA vectors. In a second application, packaged vector particles were produced in cells cotransfected with complementing replicon and defective helper DNAs. The Sindbis virus-derived DNA vectors described here increase the utility of alphavirus-based vector systems in general and also provide a vector with broad potential applications for genetic immunization.

Induction of apoptosis by c-Fos protein

The role of c-Fos in apoptosis was examined in two Syrian hamster embryo cell lines (sup+I and sup-II) and a human colorectal carcinoma cell line (RKO), using the chimeric Fos-estrogen receptor fusion protein c-FosER. As previously reported, contrasting responses were observed when these two cell lines were placed under growth factor deprivation conditions; sup+I cells were highly susceptible to apoptosis, whereas sup-II cells were resistant. In this report, we show that the activated c-FosER protein induces apoptosis in sup-II preneoplastic cells in serum-free medium, indicating that c-Fos protein can induce apoptotic cell death in these cells. c-Fos-induced apoptosis was not blocked by the protein synthesis inhibitor cycloheximide, suggesting that the c-Fos transcriptional activation activity is not involved. This conclusion was further supported by the observation that overexpression of v-Fos, which is highly proficient in transcriptional activation but deficient in the transcriptional repression activity associated with c-Fos, did not induce apoptosis. Constitutively expressed Bcl-2 delayed the onset of low-serum-induced apoptosis in sup+I cells and enhanced survival in sup-II cells. Further, coexpression of Bcl-2 and c-FosER in sup+I or sup-II cells protected the cells from c-FosER-induced apoptosis. The possibility that c-FosER-induced apoptosis requires a p53 function was examined. Colorectal carcinoma RKOp53+/+ cells, which do not normally undergo apoptosis in serum-free medium, showed apoptotic DNA fragmentation upon expression and activation of c-FosER. Further, when the wild-type p53 protein was diminished in the RKO cells by infection with the papillomavirus E6 gene, subsequent c-FosER-induced apoptosis was blocked. The data suggest that c-Fos protein plays a causal role in the activation of apoptosis in a p53-dependent manner. This activity does not require new protein synthesis and is blocked by overexpression of Bcl-2 protein.

Novel retroviral vectors for efficient expression of the multidrug resistance (mdr-1) gene in early hematopoietic cells

We present data that retroviral gene expression in early hematopoietic cells is subjected to transcriptional controls similar to those previously described for embryonic stem cells. Transient transfection experiments revealed that both the viral enhancer region in the U3 region of the long terminal repeat as well as a repressor element coincident with the primer binding site of Moloney leukemia viruses are limiting for expression in hematopoietic cells in a differentiation-dependent manner. Within the group of Moloney leukemia virus-related viruses, only the myeloproliferative sarcoma virus showed high enhancer activity in myeloid (including erythroid) cells. In contrast, enhancer regions related to the Friend mink cell focus-forming viruses mediate much higher gene expression levels in both multipotent and lineage-committed myeloid cells. In addition, transcriptional repression related to sequences in the primer binding site of Moloney leukemia virus-derived vectors is also found in early hematopoietic cells and can be overcome by using the corresponding sequences of the murine embryonic stem cell virus. On the basis of these results, two types of novel retroviral hybrid vectors were developed; they combine the U3 regions of either the Friend mink cell focus-forming virus family or the myeloproliferative sarcoma virus with the primer binding site of the murine embryonic stem cell virus. When used to express the human multiple drug resistance gene, these vectors substantially improve protection to cytostatic drugs in transduced hematopoietic cell lines FDC-Pmix, TF-1, and K-562 in comparison with Moloney leukemia virus-derived vectors presently used for the stem cell protection approach in somatic gene therapy.

Bone marrow and clinical gene therapy

Remarkable progress has been made in the last 5 years in the use of gene therapy for the treatment of inherited diseases and acquired disorders. This article reviews these applications with particular emphasis on the use of genetically modified hematopoietic cells.

Expression and signaling specificity of the IFNAR chain of the type I interferon receptor complex

The IFNAR chain of the type I interferon (IFN) receptor (IFNIR) undergoes rapid ligand-dependent tyrosine phosphorylation and acts as a species-specific transducer for type I IFN action. Using the vaccinia/T7 expression system to amplify IFNAR expression, we found that human HeLa-S3 cells transiently express high levels of cell surface IFNAR chains (approximately 250,000 chains per cell). Metabolic labeling and immunoblot analysis of transfected HeLa cells show that the IFNAR chain is initially detected as 65-kDa and 98-kDa precursors, and then as the 130-kDa mature protein. Due to variation in N-glycosylation, the apparent molecular mass of the mature IFNAR chain varies from 105 to 135 kDa in different cells. IFNIR structure was characterized in various human cell lines by analyzing 125I-labeled IFN cross-linked complexes recognized by various antibodies against IFNIR subunits and JAK protein-tyrosine kinases. Precipitation of cross-linked material from Daudi cells with anti-IFNAR antibodies showed that IFNAR was present in a 240-kDa complex. Precipitation of cross-linked material from U937 cells with anti-TYK2 sera revealed a 240-kDa complex, which apparently did not contain IFNAR and was not present in IFN-resistant HEC1B cells. The tyrosine phosphorylation and down-regulation of the IFNAR chain were induced by type I IFN in several human cell lines of diverse origins but not in HEC1B cells. However, of type I IFNs, IFN-beta uniquely induced the tyrosine phosphorylation of a 105-kDa protein associated with the IFNAR chain in two lymphoblastoid cell lines (Daudi and U266), demonstrating the specificity of transmembrane signaling for IFN-beta and IFN-alpha through the IFNAR chain.

The membrane-bound and soluble forms of HLA-G bind identical sets of endogenous peptides but differ with respect to TAP association

The class Ib antigen HLA-G is expressed as a membrane-bound protein like classical class Ia molecules (M.HLA-G) but, unlike typical class I, is also expressed as a soluble protein (S.HLA-G) with a unique C terminus. Our results show that, similar to classical class I proteins, the membrane-bound form of HLA-G associated with TAP, as evidenced by the ability to immunoprecipitate HLA-G class I heavy chain with TAP antisera. In contrast, the soluble G protein did not appear to associate with TAP in the same manner, since similar immunoprecipitation experiments failed to detect soluble G complex. A detailed analysis of peptides bound to the soluble and membrane HLA-G proteins expressed in the B lymphoblastoid cell line 721.221 showed that, like class Ia complexes, both HLA-G proteins consist of heavy and light chains complexed with nonameric peptides in a 1:1:1 ratio. The two proteins bind essentially the same set of peptides, which are derived from a variety of intracellular proteins and define a peptide motif for HLA-G. The peptides contain Leu at the C terminus and Pro or small hydrophobic amino acids in position 3 followed by Pro or Gly in position 4. The complexity of the bound peptides is lower than that found for some class Ia complexes, but is more similar to class Ia than to the limited repertoire of some murine class Ib molecules.

Generation of targeted retroviral vectors by using single-chain variable fragment: an approach to in vivo gene delivery

We report the generation of a retroviral vector that infects human cells specifically through recognition of the low density lipoprotein receptor. The rationale for this targeted infection is to add onto the ecotropic envelope protein of Moloney murine leukemia virus, normally trophic for murine cells, a single-chain variable fragment derived from a monoclonal antibody recognizing the human low density lipoprotein receptor. This chimeric envelope protein was used to construct a packaging cell line producing a retroviral vector capable of high-efficiency transfer of the Escherichia coli beta-galactosidase gene to human cells expressing low density lipoprotein receptor. This approach offers a generalized plan to generate cell and tissue-specific retroviral vectors, an essential step toward in vivo gene therapy strategies.

PML suppresses oncogenic transformation of NIH/3T3 cells by activated neu

The chromosomal translocation t(15;17)(q22;q12) is a consistent feature of acute promyelocytic leukemia (APL) that results in the disruption of genes for the zinc finger transcription factor PML and the retinoic acid receptor alpha (RAR alpha). We have previously shown that PML is a growth suppressor and is able to suppress transformation of NIH/3T3 by activated neu oncogene. In the study presented here, the full-length PML cDNA was transfected into B104-1-1 cells (NIH/3T3 cells transformed by the activated neu oncogene) by retrovirally mediated gene transfer. We found that expression of PML could reverse phenotypes of B104-1-1 including morphology, contact-limiting properties, and growth rate in both transient-expression and stable transfectants. We also demonstrated that PML is able to suppress clonogenicity of B104-1-1 in soft agar assay and tumorigenicity in nude mice. These results strongly support our previous finding that PML is a transformation or growth suppressor. Our results further demonstrate that expression of PML in B104-1-1 cells has little effect on cell cycle distribution. Western blot analysis demonstrated that suppression of neu expression in B104-1-1 by PML was insignificant in the transient transfection experiment but significant in the PML stable transfectants. This study suggests that PML may suppress neu expression and block signaling events associated with activated neu. This study supports our hypothesis that disruption of the normal function of PML, a growth or transformation suppressor, is a critical event in APL leukomogenesis.

Tumor necrosis factor alpha-induced apoptosis in human neuronal cells: protection by the antioxidant N-acetylcysteine and the genes bcl-2 and crmA

Tumor necrosis factor alpha (TNF-alpha) is a candidate human immunodeficiency virus type 1-induced neurotoxin that contributes to the pathogenesis of AIDS dementia complex. We report here on the effects of exogenous TNF-alpha on SK-N-MC human neuroblastoma cells differentiated to a neuronal phenotype with retinoic acid, TNF-alpha caused a dose-dependent loss of viability and a corresponding increase in apoptosis in differentiated SK-N-MC cells but not in undifferentiated cultures. Importantly, intracellular signalling via TNF receptors, as measured by activation of the transcription factor NF-kappa B, was unaltered by retinoic acid treatment. Finally, overexpression of bcl-2 or crmA conferred resistance to apoptosis mediated by TNF-alpha, as did the addition of the antioxidant N-acetylcysteine. These results suggest that TNF-alpha induces apoptosis in neuronal cells by a pathway that involves formation of reactive oxygen intermediates and which can be blocked by specific genetic interventions.

A family of retroviruses that utilize related phosphate transporters for cell entry

The amphotropic murine retrovirus receptor Ram-1 shows significant sequence similarity to the gibbon ape leukemia virus (GALV) receptor Glvr-1, and both of these cell surface virus receptors normally function as sodium-dependent phosphate symporters. However, Ram-1 from humans or rats does not serve as a receptor for GALV, and Glvr-1 from humans does not serve as a receptor for amphotropic virus. Here we show that the murine retrovirus 10A1 can enter cells by using either Glvr-1 or Ram-1. Furthermore, we have constructed Ram-1/Glvr-1 hybrid receptors that allow entry of both GALV and amphotropic virus. While GALV and amphotropic virus are in separate interference groups when assayed on human cells, they do interfere with each other in cells expressing the hybrid receptor. These results indicate a close functional relationship between retroviruses that utilize members of this newly defined receptor family and provide a molecular explanation for nonreciprocal and cell type-specific interference observed for some retrovirus classes.

Gene therapy of rat 9L gliosarcoma tumors by transduction with selectable genes does not require drug selection

9L rat glioma cells have been used as a model for brain tumor therapies. It has been reported that in vivo infection of 9L cells with a replication-defective retrovirus expressing the herpes simplex thymidine kinase gene resulted in decreased tumor formation following treatment with the antiviral drug ganciclovir. In the study reported here, rats were injected either intracerebrally or subcutaneously with 9L glioma cells expressing a chimeric hygromycin phosphotransferase-thymidine kinase fusion protein or with unmodified 9L cells. Tumor formation was decreased in the rats receiving modified cells, even in the absence of treatment with ganciclovir. Suppression of tumor growth was also observed with cells modified to express the intracellular selectable marker neomycin phosphotransferase. These results indicate that an intracellular selectable marker, in the absence of pharmacologic selection, can inhibit tumor growth of 9L cells. The demonstration that intracellular marker genes can negatively influence the survival of transplanted cells has important implications for in vivo studies that use genetically modified cells.

Csk suppression of Src involves movement of Csk to sites of Src activity

Csk phosphorylates Src family members at a key regulatory tyrosine in the C-terminal tail and suppresses their activities. It is not known whether Csk activity is regulated. To examine the features of Csk required for Src suppression, we expressed Csk mutants in a cell line with a disrupted csk gene. Expression of wild-type Csk suppressed Src, but Csk with mutations in the SH2, SH3, and catalytic domains did not suppress Src. An SH3 deletion mutant of Csk was fully active against in vitro substrates, but two SH2 domain mutants were essentially inactive. Whereas Src repressed by Csk was predominantly perinuclear, the activated Src in cells lacking Csk was localized to structures resembling podosomes. Activated mutant Src was also in podosomes, even in the presence of Csk. When Src was not active, Csk was diffusely located in the cytosol, but when Src was active, Csk colocalized with activated Src to podosomes. Csk also localizes to podosomes of cells transformed by an activated Src that lacks the major tyrosine autophosphorylation site, suggesting that the relocalization of Csk is not a consequence of the binding of the Csk SH2 domain to phosphorylated Src. A catalytically inactive Csk mutant also localized with Src to podosomes, but SH3 and SH2 domain mutants did not, suggesting that the SH3 and SH2 domains are both necessary to target Csk to places where Src is active. The failure of the catalytically active SH3 mutant of Csk to regulate Src may be due to its inability to colocalize with active Src.

Csk suppression of Src involves movement of Csk to sites of Src activity

Csk phosphorylates Src family members at a key regulatory tyrosine in the C-terminal tail and suppresses their activities. It is not known whether Csk activity is regulated. To examine the features of Csk required for Src suppression, we expressed Csk mutants in a cell line with a disrupted csk gene. Expression of wild-type Csk suppressed Src, but Csk with mutations in the SH2, SH3, and catalytic domains did not suppress Src. An SH3 deletion mutant of Csk was fully active against in vitro substrates, but two SH2 domain mutants were essentially inactive. Whereas Src repressed by Csk was predominantly perinuclear, the activated Src in cells lacking Csk was localized to structures resembling podosomes. Activated mutant Src was also in podosomes, even in the presence of Csk. When Src was not active, Csk was diffusely located in the cytosol, but when Src was active, Csk colocalized with activated Src to podosomes. Csk also localizes to podosomes of cells transformed by an activated Src that lacks the major tyrosine autophosphorylation site, suggesting that the relocalization of Csk is not a consequence of the binding of the Csk SH2 domain to phosphorylated Src. A catalytically inactive Csk mutant also localized with Src to podosomes, but SH3 and SH2 domain mutants did not, suggesting that the SH3 and SH2 domains are both necessary to target Csk to places where Src is active. The failure of the catalytically active SH3 mutant of Csk to regulate Src may be due to its inability to colocalize with active Src.

Astrogliosis in culture: III. Effect of recombinant retrovirus expressing antisense glial fibrillary acidic protein RNA

Injury to the central nervous system (CNS) either from trauma or due to demyelinating/degenerating diseases results in a typical response of astrocytes, termed astrogliosis. This reaction is characterized by astrocyte proliferation, extensive hypertrophy of nuclei, cell body, and cytoplasmic processes and an increase in immunodetectable glial fibrillary acidic protein (GFAP). GFAP accumulation may cause a physical barrier preventing the reestablishment of a functional environment. Our studies have aimed at modulating astrogliosis by inhibiting or delaying GFAP synthesis in damaged and reactive astrocytes. The present study investigates the use of a recombinant retrovirus expressing antisense GFAP RNA in controlling the response of mechanically injured astrocytes. A 650 bp fragment from the coding region of mouse GFAP cDNA was cloned in the antisense orientation under the control of long terminal repeat (LTR) promoter of Moloney murine leukemia virus. Increase in GFAP as detected by immunocytochemical staining in injured astrocytes was inhibited by treatment with retrovirus expressing antisense GFAP RNA. Also, astrocytes at the site of injury in these scratched cultures did not show cell body hypertrophy compared to control cultures. These observations demonstrate that the increase in GFAP at the site of injury can be inhibited using retroviral treatment and indicate the potential of retrovirus-mediated gene transfer in modulating scar formation in the CNS in vivo. These studies also shed light on the role of GFAP in maintaining the morphology of astrocytes.

Retroviral delivery of DNA into the livers of transgenic mice bearing premalignant and malignant hepatocellular carcinomas

To develop gene therapy for hepatocellular carcinoma (HCC), we infused mice through the portal vein with retrovirus carrying the Escherichia coli beta-galactosidase reporter gene under the transcriptional control of the viral long terminal repeat (LTR) and the promoter from the mouse multidrug resistance gene mdr1b. Two transgenic mouse HCC models were used, one bearing the human hepatitis B viral envelope protein and the other SV40 T antigen. These animals develop HCC with predictable pathological manifestations. The viral transduction efficiency appeared to depend upon the stage of the disease in the animals. The most efficient transduction occurred when the livers had developed microscopic nodular hyperplasia; in some cases as many as 0.01-0.1 copies/cell were transduced. The transduction efficiency was lower in the late stage of the disease when livers had a heavy tumor burden and in the early stage when no lesion was evident. Low viral transduction efficacy was also seen in nontransgenic animals but was significantly increased by partial hepatectomy. The expression of the reporter gene in these animals was very low, as determined by histological staining. These results suggest that hepatocarcinogenesis can enhance retroviral delivery of foreign genes into the liver. Further development by increasing the viral transducing efficiency and the level of expression of transduced gene is required.

The cytoplasmic domain of CD4 is sufficient for its down-regulation from the cell surface by human immunodeficiency virus type 1 Nef

Human immunodeficiency virus type 1 Nef down-regulates surface expression of murine and human CD4 but not human CD8. We recently reported that the cytoplasmic domain of CD4 is required for its down-regulation by Nef. Using a chimeric molecule composed of the extracellular and transmembrane domains of human CD8 fused to the cytoplasmic domain of human CD4, we show here that the cytoplasmic domain of CD4 is sufficient for down-regulation by Nef. Since the cytoplasmic domain of CD4 is also the site of its association with p56lck, we used a series of CD4 mutants to determine whether the regions of the cytoplasmic domain of CD4 required for down-regulation by Nef are the same as those required for p56lck binding. Our results indicate that the portion of the cytoplasmic domain required for the down-regulation of CD4 by Nef overlaps with the binding site of p56lck, but the cysteine residues which are essential for the association of CD4 with p56lck are not required. This observation raised the possibility that Nef competes with p56lck for binding to CD4. However, under conditions which are considerably milder than those permissive for coimmunoprecipitation of CD4 and p56lck, we found no evidence for an association between Nef and CD4. While a decrease in total CD4 was observed in lysates of cells expressing Nef, the levels of p56lck were not significantly affected. Pulse-chase experiments further revealed a decrease in the half-life of CD4 in Nef-expressing cells. These results show that the decrease in surface CD4 expression induced by Nef is mediated at least in part by a decrease in the half-life of CD4 protein. These results also indicate that a large portion of p56lck is free of CD4 in T cells expressing Nef, which could have a significant effect on T-cell function.

Inhibition of myogenic differentiation in myoblasts expressing a truncated type II TGF-beta receptor

Transforming growth factor-beta (TGF-beta) is thought to play a role in mesenchymal cell development and, specifically, in muscle differentiation, yet its precise role in the latter process remains unclear. TGF-beta has been shown to both inhibit and induce myoblast maturation in vitro, depending on the culture conditions. Whether the type I or type II TGF-beta receptor mediates the various TGF-beta effects on myogenesis is not known. In the present study, C2C12 myoblasts were transfected with an expression vector for a truncated type II TGF-beta receptor, which has been shown to act as a dominant negative inhibitor of type II receptor signaling. In contrast to the parental cells, the transfected clones did not efficiently form myotubes or induce expression of MyoD, myogenin and several other differentiation markers following incubation in low serum media. However, some muscle differentiation markers continued to be expressed in the transfected cells suggesting that at least two pathways are involved in muscle cell differentiation. These cells could still growth arrest in low serum media, showing that decreased proliferation can be dissociated from differentiation. Unlike several oncogenes known to block myogenic differentiation, expression of the truncated TGF-beta receptor did not result in myoblast transformation. Injection of the parental or the transfected C2C12 cells into the limb muscle of nude mice revealed quantitative and qualitative differences in their behavior, and suggested that myoblasts expressing the truncated TGF-beta receptor cannot fuse in vivo. Finally, retrovirus-mediated expression of MyoD in the transfected cells restored their ability to form myotubes in vitro, indicating that inhibition of myoblast differentiation by the truncated TGF-beta receptor may depend on decreased MyoD expression. We propose that TGF-beta signaling through the type II receptor is required for several distinct aspects of myogenic differentiation and that TGF-beta acts as a competence factor in this multistep process.