Synergistic growth factors enhance rat liver proliferation and enable retroviral gene transfer via a peripheral vein.

BACKGROUND & AIMS: Genetic diseases reflecting abnormal hepatocyte function are potentially curable through gene therapy. Retroviral vectors offer the potential for permanent correction of such conditions. These vectors generally require cell division to occur to allow provirus entry into the nucleus, initiated in many experimental protocols by partial hepatectomy. We have explored methods to improve the efficiency of retroviral gene transfer that avoid the need for liver damage. METHODS: Triiodothyronine (T3) and keratinocyte growth factor (KGF) were used to induce hepatic proliferation in rats. The effects of intraportal and peripheral administration of a modified retrovirus that encoded the Lac Z gene during growth factor-induced liver hyperplasia were analyzed. RESULTS: T3 initiated hepatocyte proliferation midzonally; after KGF, proliferation was more diffuse. Optimal concentrations of T3 and KGF acted synergistically to induce proliferation in 61% of hepatocytes in the intact liver. This enabled in vivo hepatocyte transduction, leading to gene expression by up to 7.3% of hepatocytes after intraportal retroviral vector administration and 7. 1% after peripheral venous administration. CONCLUSIONS: T3 and KGF act synergistically to induce hepatocyte proliferation in undamaged liver. The liver can be simply transduced with integrating vectors via the peripheral venous system during a wave of growth factor-induced proliferation.

Long-term restoration of striatal L-aromatic amino acid decarboxylase activity using recombinant adeno-associated viral vector gene transfer in a rodent model of Parkinson's disease.

As a potential treatment for Parkinson's disease, viral vector-mediated over-expression of striatal L-aromatic amino acid decarboxylase was tested in an attempt to facilitate the production of therapeutic levels of dopamine after peripheral L-dihydroxyphenylalanine administration. The results of microdialysis and enzyme activity assays indicate that striatal decarboxylation of peripherally administered L-dihydroxyphenylalanine was enhanced by recombinant adeno-associated virus-mediated gene transfer of L-aromatic amino acid decarboxylase in unilateral 6-hydroxydopamine-lesioned rats. This gene transfer-induced increase in striatal decarboxylase activity was shown to remain undiminished over a six-month period and transgene expression was demonstrated to persist for at least one year. Unlike previous approaches involving delivery of either tyrosine hydroxylase, or tyrosine hydroxylase and L-aromatic amino acid decarboxylase transgenes together to accomplish unregulated dopamine delivery, the current study proposes a pro-drug strategy (peripheral L-dihydroxyphenylalanine administration after L-aromatic amino acid decarboxylase transduction). This strategy for dosage control could potentially allow lowered L-dihydroxyphenylalanine doses and potentially obviate complicated transcriptional regulation paradigms. These data suggest that the use of the non-pathogenic adeno-associated virus to transfer the L-aromatic amino acid decarboxylase gene into the striatum of Parkinson's disease patients may be an attractive gene therapy strategy.

Retrovirus-mediated HO gene transfer into endothelial cells protects against oxidant-induced injury.

Heme oxygenase (HO)-1 is a stress protein that has been implicated in defense mechanisms against agents that may induce oxidative injury, such as endotoxins, heme, and cytokines. Overexpression of HO-1 in cells might, therefore, protect against oxidative stress produced by certain agents, specifically heme, by catalyzing its degradation to bilirubin, which by itself has antioxidant properties. We report for the first time the successful transduction of human HO-1 gene into rat lung microvessel endothelium using replication-defective retroviral vector. Cells transduced with human HO-1 gene exhibited a 2.1-fold increase in HO-1 protein level, which was associated with a 2.3-fold elevation in enzyme activity compared with that in nontransduced cells. The cGMP content in transduced endothelial cells was increased by 2.9-fold relative to that in nontransduced cells. Moreover, human HO-1 gene-transduced endothelial cells acquired substantial resistance to toxicity produced by exposure to heme and H(2)O(2) compared with that in nontransduced cells. The protective effect of enhancement of HO-1 activity against heme and H(2)O(2) was reversed by pretreatment with stannic mesoporphyrin, a competitive inhibitor of HO. These data demonstrate that the induction of HO-1 in response to injurious stimuli represents an important mechanism for moderating the severity of cell damage. Regulation of HO activity in this manner may have clinical applications.

Phospholamban-to-SERCA2 ratio controls the force-frequency relationship.

The force-frequency relationship (FFR) describes the frequency-dependent potentiation of cardiac contractility. The interaction of the sarcoplasmic reticulum Ca2+-adenosinetriphosphatase (SERCA2) with its inhibitory protein phospholamban (PLB) might be involved in the control of the FFR. The FFR was analyzed in two systems in which the PLB-to-SERCA2 ratio was modulated. Adult rabbit cardiac myocytes were transduced with adenovirus encoding for SERCA2, PLB, and beta-galactosidase (control). After 3 days, the relative PLB/SERCA2 values were significantly different between groups (SERCA2, 0.5; control, 1.0; PLB, 4.5). SERCA2 overexpression shortened relaxation by 23% relative to control, whereas PLB prolonged relaxation by 39% and reduced contractility by 47% (0.1 Hz). When the stimulation frequency was increased to 1.5 Hz, myocyte contractility was increased by 30% in control myocytes. PLB-overexpressing myocytes showed an augmented positive FFR (+78%), whereas SERCA2-transduced myocytes displayed a negative FFR (-15%). A more negative FFR was also found in papillary muscles from SERCA2 transgenic mice. These findings demonstrate that the ratio of phospholamban to SERCA2 is an important component in the control of the FFR.

The role of TGFbeta1 in initiating hepatic stellate cell activation in vivo.

BACKGROUND/AIMS: The activation of hepatic stellate cells is a key initiating event in hepatic fibrogenesis. Although TGFbeta1 is a potent inducer of collagen alpha1(I) expression in vitro and elevated levels of TGFbeta1 are found in patients and experimental animals with hepatic fibrosis and cirrhosis, the role of increased TGFbeta1 in the initiation of hepatic stellate cell activation in vivo is unknown. We used two experimental approaches to study this relationship: 1) Induction of an acute liver injury with carbon tetrachloride (CCl4) in normal and TGFbeta1-knockout (ko) mice, and 2) overexpression of TGFbeta1 in the liver of wild-type mice using a recombinant replication-deficient adenovirus encoding human TGFbeta1 (Ad-TGFbeta1). METHODS: TGFbeta1-ko mice (n=6) and normal mice (n=6) were injected once intraperitoneally (i.p.) with CCl4 (1 microl/g BW) or mineral oil. Wild-type mice (n=3) were injected intravenously with Ad-TGFbeta1 (10(10) pfu) or a control virus expressing beta-galactosidase (Ad-LacZ, 10(10) pfu). Animals were sacrificed after 3 days and total liver RNA was prepared. The expression of collagen alpha1(I) mRNA normalized to GAPDH mRNA was measured by RNase protection assay, asmooth muscle actin (alpha-sma) protein expression was analyzed by Western blotting. The expression of TGFbeta1, TGFbeta2, and TGFbeta3 mRNAs were determined semi-quantitatively with RT-PCR. RESULTS: The collagen alpha1(I) mRNA was increased 10-fold in CCl4-treated wild-type mice compared to the controls. This increase was reduced about 80% in the TGFbeta1-ko mice. The TGFbeta1 mRNA levels in the wild-type mice were proportional to the collagen alpha1(I) mRNA levels. a-sma, a marker of hepatic stellate cell activation, was expressed earlier and at a higher level in wild-type mice than TGFbeta-ko mice after CCl4 treatment. The Ad-TGFbeta1 infected mice had 14-fold higher hepatic TGFbeta protein levels and 15-fold higher collagen alpha1(I) mRNA levels than the Ad-LacZ-infected control mice. Collagen alpha1(I) mRNA levels were proportional to the transgenic TGFbeta1 mRNA levels, while the endogenous TGFbeta1 was only slightly higher than in the controls. TGFbeta2 and TGFbeta3 mRNA levels were elevated in CCl4-treated wild-type and TGFbeta1-ko mice and in Ad-TGFbeta1-infected mice compared to the controls. CONCLUSIONS: Absence of TGFbeta1 inhibits hepatic collagen alpha1(I) mRNA and alpha-sma protein expression by the toxic stimulus CCl4, and targeted TGFbeta1 overexpression increases collagen alpha1(I) mRNA and alpha-sma protein levels in the liver in vivo. Other TGFbeta family members do not compensate for the TGFbeta1 deficiency. This indicates that TGFbeta1 accelerates, but is not absolutely required, for the activation of hepatic stellate cells.

Nonthrombogenic, adhesive cellular lining for left ventricular assist devices.

BACKGROUND: The textured, blood-contacting surfaces of the Thermocardiosystems HeartMate left ventricular assist device (LVAD) promote the passivation of the biomaterial caused by the accumulation of an integral coagulum. Commonly, acute, postimplantation thrombocytopenia causes significant bleeding, requiring surgery or blood transfusions. Chronic complications include thromboembolic microevents that can affect central nervous system function. Pumps, explanted during donor organ transplantation, are often found to have an extensive cellular panus associated with the blood-contacting surfaces of the device. This natural cellular lining suggests a possible strategy for improving the blood biocompatibility of the HeartMate. Therefore, seeding of LVADs with cells genetically engineered to enhance their antithrombotic properties before implantation was investigated as a means to improve biocompatibility for long-term use. METHODS AND RESULTS: Bovine vascular smooth muscle cells genetically engineered to produce nitric oxide were seeded on LVAD biomaterials and exposed to elevated shear stresses to determine cell-adhesive capabilities. Comparative studies were performed with vascular endothelial cells isolated from the same vessel. To assess the thrombogenic potential of the genetically engineered smooth muscle cells, monolayers were exposed to whole blood in parallel plate flow chambers and were platelet-adhesion quantified. This procedure used scanning electron microscopy and computer image-capture software. Endothelial cell monolayers and mock-transduced smooth muscle cells were assayed in a comparative manner. LVADs were seeded with genetically engineered smooth muscle cells and maintained under cell culture conditions for 96 hours. Thereafter, seeded LVADs were incorporated into in vitro flow loops. Cell retention within the pump was determined by sampling the effluent culture medium downstream of the pump and cell counting in a Coulter counter. After 18 hours of in vitro flow, a seeded pump was implanted into the abdominal cavity of a calf and anastomosed to the apex of the heart and to the descending aorta. More genetically engineered smooth muscle cells were retained on the surface of LVAD biomaterials when they were subjected to shear stresses up to 75 dyne/cm than endothelial cells assayed in the identical manner. Adherence of platelets to the surface of smooth muscle cells was significantly reduced after their transduction with nitric oxide synthase with GTP cyclohydrolase genes. Platelet deposition on the genetically modified myocyte layers was similar to that associated with endothelial cell layers. Cell loss from cell-seeded LVADs incorporated into in vitro flow loops remained < 5% of the total cell number seeded regardless of the duration of flow. CONCLUSIONS: LVADs seeded with smooth muscle cells, transduced with the genes to optimize nitric oxide production, adhered well to the pump surface under in vitro and in vivo flow conditions.

Fas-mediated apoptosis is involved in the elimination of gene-transduced hepatocytes with E1/E3-deleted adenoviral vectors.

Gene-transduced hepatocytes with E1/E3-deleted adenoviral vectors are eliminated immediately and the expression of transduced genes disappears rapidly following the vector administration. In this report, we analysed the involvement of apoptotic cell death in the elimination of hepatocytes infected with adenoviral vectors. An E1/E3-deleted adenoviral vector expressing Escherichia coli beta-galactosidase (LacZ) was injected via the portal vein into congenitally Fas-deficient mice (lpr), Fas ligand-deficient mice (gld) and their control mice, MRL and C3H. 5-Bromo-4-chloro-3-indolyl-beta-D-galactoside (X-gal) staining of the liver specimens showed that 80-100% of hepatocytes were LacZ positive at 7 days after virus administration, suggesting that most of the hepatocytes received the injected adenoviral vectors. In normal mice, the number of LacZ-positive cells decreased dramatically at 14 and 21 days after transduction and few positive cells were observed at day 28. Beta-galactosidase activity, quantified by the O-nitrophenyl-beta-D-galactopyranoside assay, gave comparable results to X-gal staining. At days 14 or 21, many apoptotic hepatocytes and apoptotic infiltrating cells were detected with the terminal deoxyribonucleotidyl transferase-mediated dUTP-digoxigenin nick end-labelling (TUNEL) in situ apoptosis detection method. This observation suggested that the apoptotic process was associated with the elimination of adenovirus-infected hepatocytes. To test the involvement of the Fas-Fas ligand interaction in this apoptotic process, the period of transgene expression was measured in lpr and gld mice, which had received the same amount of AxCALacZ. X-Gal histochemical analysis detected many LacZ-positive cells in lpr or gld mice liver even at 21 or 28 days after AxCALacZ injection. There were significant differences in the reduction rates of beta-galactosidase activity of liver homogenates between lpr and MRL, or gld and C3H mice. Based on these observations, we conclude that the Fas-mediated apoptotic process is involved in the elimination of hepatocytes infected with E1/E3-deleted adenoviral vectors.

Long-term expression of the gene encoding green fluorescent protein in murine hematopoietic cells using retroviral gene transfer.

BACKGROUND: A major goal in retroviral-based gene therapy is to establish methods that allow for selection and tracking of transduced cell populations. Green fluorescent protein (GFP) may be useful for gene therapy applications because it is a naturally fluorescent protein that can be detected using conventional flow cytometers facilitating rapid analysis and purification of transduced cell populations. However, it is unknown whether GFP can be stably expressed in vivo, particularly in multiple bone marrow-derived cell lineages. METHODS: A murine retrovirus carrying the gene encoding GFP was used to infect murine bone marrow cells (BMCs). These studies were conducted to (1) directly determine whether GFP could be used as a marker of BMC transduction, (2) determine whether GFP is capable of being expressed in multiple bone marrow-derived hematopoietic cell lineages, and (3) determine whether GFP could be used to follow the fate of transduced cells in vivo. RESULTS: Infection of BMCs with retroviruses carrying the gene encoding GFP resulted in a fluorescent signal in viable transduced cells that was detectable by flow cytometry. Expression of GFP was detected in multiple bone marrow-derived cell lineages after transduction, including stem cell antigen-positive (Sca-1+), lineage marker-negative (Lin-) cells. Using GFP as a selectable marker, we were able to enrich for transduced cells by cell sorting. Mice reconstituted with enriched populations of GFP+ cells showed a significant increase in the percentage of cells expressing GFP in the periphery when compared with mice reconstituted with unenriched transduced bone marrow. CONCLUSIONS: These data indicate that GFP can be used to select for transduced BMCs in vitro, expressed in multiple bone marrow-derived cell lineages, used to select transduced cells, and follow the fate of transduced cells long-term in vivo.

A role for p21ras in the angiotensin II AT2 receptor transduction pathway.

We have previously shown that the activation of the AT2 receptor of Ang II induced neurite outgrowth in NG108-15 cells. We also found that stimulation of NG108-15 cells with Ang II induced a rapid decrease in GTP-bound p21ras. In order to investigate the possible role of p21ras in Ang II-induced neuronal differentiation, we have established NG108-15 sublines which inducibly express a dominant inhibitory form of p21ras (p21N17Ras). We observed that IPTG-induced expression of p21N17Ras in these NG108-15 sublines induced the same morphological changes as does Ang II in control untransfected cells. Immunofluorescence labeling of beta-tubulin showed that expression of p21N17Ras induced neurite outgrowth and elongation. These observations were supported by Western blot analysis of the level of polymerized tubulin. These results strongly support the hypothesis that AT2 receptor-induced neuronal differentiation in NG108-15 cells is mediated by the inhibition of p21ras.

Adenovirus-mediated beta-galactosidase gene delivery to the liver leads to protein deposition in kidney glomeruli.

The many cell types of the kidney, precisely arranged, allow this organ to perform its complex physiologic functions. However, this architectural complexity makes gene transfer into the kidney difficult. One approach to delivering a therapeutic protein to the kidney is to transfer a gene to a non-renal tissue. Release of the protein into the circulation might then result in deposition in the kidney, if the protein has the appropriate molecular properties. In this study, we found that parenterally administered replication deficient adenovirus carrying the beta-galactosidase gene resulted in intense beta-galactosidase gene expression in hepatocytes. As a result of immune attack on transduced hepatocytes, beta-galactosidase protein from these cells is released into the circulation, transported, and deposited almost exclusively in kidney glomeruli. Intense beta-galactosidase activity was noted in both kidneys with a peak at two weeks following viral administration, concurrent with loss of beta-galactosidase positive hepatocytes. Consistent with our hypothesis of protein transfer, no beta-galactosidase mRNA was detected in glomeruli. Moreover, systemically administered protein generated similar glomerular beta-galactosidase activity. Finally, co-administration of murine CTLA4 Ig, an immunomodulator of T cell activation, with the adenovirus protected infected hepatocytes and markedly diminished glomerular beta-galactosidase activity. Collectively, these findings suggest that a therapeutic protein can be "targeted" to the renal glomerulus, utilizing the liver as a gene transfer organ.

In situ retroviral-mediated gene transfer for the treatment of brain tumors in rats.

Gene transfer with vectors derived from murine retroviruses is restricted to cells which are proliferating and synthesizing DNA at the time of infection. This suggests that retroviral-mediated gene transfer might permit targeting of gene integration into malignant cells in organs composed mainly of quiescent nonproliferating cells, such as in the brain. Accordingly, selective introduction of genes encoding for susceptibility to otherwise nontoxic drugs ("suicide" genes) into proliferating brain tumors may be used to treat this cancer. We investigated the efficacy and dynamics of in vivo transduction of growing brain tumors with the herpes simplex-thymidine kinase gene followed by administration of the antiviral drug ganciclovir. Ganciclovir is phosphorylated by thymidine kinase to toxic triphosphates that interfere with DNA synthesis, resulting in the preferential death of the transduced tumor cells. Rats inoculated with 4 x 10(4) 9L gliosarcoma cells into the frontal lobe were treated 7 days later with an intratumoral stereotaxic injection of murine fibroblasts (NIH 3T3 cells) that were producing a retroviral vector containing the herpes simplex-thymidine kinase gene. Controls received vector producer and nonproducer NIH 3T3 cell lines containing the Escherichia coli lacZ (beta-galactosidase) gene as well as nonproducer NIH 3T3 cells containing the thymidine kinase gene. The animals were rested for 7 days to allow time for in situ transduction of the proliferating tumor cells with the herpes-thymidine kinase retroviral vector. The animals were then treated with ganciclovir, 15 mg/kg i.p. twice a day for 14 days. Gliomas receiving an injection of 3-5 x 10(6) thymidine kinase producer cells regressed completely in 23 of 30 rats given ganciclovir therapy, while 25 of 26 control rats developed large tumors. Intratumoral injection of a lower concentration of thymidine kinase vector producer cells (1.8 x 10(6)) resulted in a lower frequency of tumor regression (5 of 13 rats). To estimate the efficiency of in vivo gene transfer, 9L brain tumors were given injections of 5 x 10(6) beta-galactosidase vector producer cells. 5-Bromo-4-chloro-3-indolyl-beta-D-galactopyranaside staining revealed maximal staining of beta-galactosidase within the tumor 7-14 days after injection of the vector producer cells. In vivo transduction rates in harvested tumors ranged from 10 to 70%. There was no evidence of transduction of the surrounding normal neural tissue. Occasional blood vessel endothelial cells within or adjacent to the tumor were observed to be 5-bromo-4- chloro-3-indolyl-beta-D-galactopyranaside positive.(ABSTRACT TRUNCATED AT 400 WORDS)

Lack of coupling of muscarinic receptors to phosphoinositide metabolism and adenylyl cyclase in human lymphocytes and polymorphonuclear leukocytes: studies in healthy subjects and allergic asthmatic patients.

A change in muscarinic receptor-induced phosphoinositide (PI) metabolism or inhibition of adenylyl cyclase activity could play a role in cholinergic hyperresponsiveness in patients with asthma. In the present study, we considered the possibility of using peripheral blood lymphocytes or polymorphonuclear leukocytes (PMN) as easily accessible models to investigate this hypothesis. Previous studies have indicated the presence of muscarinic binding sites on these cells; however, their transduction mechanisms are still unclear. We found that neither lymphocytes nor PMN from healthy donors, nor lymphocytes from allergic asthmatic patients accumulated inositol phosphates after stimulation with methacholine, whereas positive responses were found for phytohemagglutinin (PHA) in the lymphocytes and for formyl-methionyl-leucyl-phenyl-alanine in the PMN. Normal basal and PHA-stimulated inositol phosphates levels were found in the lymphocytes of the patients. Both levels were significantly enhanced after allergen challenge, indicating that activation of PI metabolism could play a role in the immune response of these cells. Since methacholine was also unable to inhibit adenylyl cyclase activity in both lymphocytes and PMN, we reevaluated the presence of muscarinic binding sites on these cells. Radioligand binding studies with the hydrophilic ligand [3H]-N-methyl-scopolamine surprisingly showed only a very small population of high affinity binding sites on lymphocytes (Bmax 57-127 sites/cell; Kd 45-133 pM), while no high affinity sites could be detected on PMN. The results indicate that leukocytes are not a suitable model to investigate the muscarinic receptor function in relation to bronchial asthma.

Mechanism of transduction by retroviruses.

Retroviruses can capture cellular sequences and express them as oncogenes. Capture has been proposed to be a consequence of the inefficiency of polyadenylation of the viral genome that allows the packaging of cellular sequences flanking the integrated provirus in virions; after transfer into virions, these sequences could be incorporated into the viral genome by illegitimate recombination during reverse transcription. As a test for this hypothesis, a tissue culture system was developed that mimics the transduction process and allows the analysis and quantitation of capture events in a single step. In this model, transduction of sequences adjacent to a provirus depends on the formation of readthrough transcripts and their transmission in virions and leads to various recombinant structures whose formation is independent of sequence similarity at the crossover site. Thus, all events in the transduction process can be attributed to the action of reverse transcriptase on readthrough transcripts without involving deletions of cellular DNA.

Gene therapy model for stromal precursor cells of hematopoietic microenvironment.

Marker bacterial Neor gene was transduced by retroviral gene transfer into stromal precursor cells making up the hematopoietic microenvironment in murine long-term bone marrow cultures (LTBMC). Cultures were infected six times during the first 3 weeks of cultivation. At 4 weeks, the adherent cell layers (ACLs) were implanted under the renal capsule of syngeneic unirradiated and irradiated mice. Cells from newly formed ectopic foci were explanted into secondary LTBMC. ACLs containing the marker gene were detected by polymerase chain reaction. About 74% of stromal cells in ACLs contained Neor gene. The possibility of stable gene transduction into stromal precursor cells competent to transfer the hematopoietic microenvironment was established.

Expression of a human complementary DNA for the multidrug resistance gene in murine hematopoietic precursor cells with the use of retroviral gene transfer.

In multidrug resistance, cells become simultaneously resistant to anthracyclines, vinca alkaloids, epipodophyllotoxins, and certain other natural product cytotoxic drugs. Resistance results from synthesis of a multidrug transporter (P-glycoprotein) encoded by the MDR1 gene (also known as the PGY1 gene). In the present study, a retrovirus vector containing a complementary DNA for the human multidrug resistance gene HaMDR1/A was used to transfer the multidrug resistance phenotype to bone marrow cells of the DBA/2J mouse. A high proportion of transduced bone marrow cells showed resistance to both colchicine and vinblastine, as determined by in vitro colony formation of hematopoietic precursor cells. In addition, brief culturing of the cells in a cytotoxic drug following exposure to the retrovirus vector could be used to increase the proportion of bone marrow cell colonies that were resistant. These results may serve as a model for the generation and selection of bone marrow cells resistant to the toxic effects of chemotherapeutic agents in vivo.