Effect of administration of a recombinant adenovirus expressing the genes for IFN-gamma and interleukin-12 on acute murine toxoplasmosis

The effect of recombinant murine interferon-gamma (rMuIFN-gamma) produced from an adenovirus construct on Toxoplasma gondii in tissue culture and on the outcome of a T. gondii infection in mice was determined. Supernatants from AdCMVMuIFN-gamma-infected mouse lung epithelial (MuLE) cells were evaluated for the ability to produce biologically active IFN-gamma by measuring the capacity of the supernatants to activate peritoneal macrophages for killing of T. gondii. The bioactivity of IFN-gamma in supernatants increased with increasing multiplicity of infection (moi). Replication was inhibited 43%, 67%, and 70% by supernatants from MuLE cells infected with AdCMVMuIFN-gamma moi 5, 10, and 50, respectively, (p < 0.01 compared with controls). Bioactivity of IFN-gamma also increased as the length of time after infection increased. T. gondii replication was inhibited 28% and 36%, respectively, by AdCMVMuIFN-gamma-infected MuLE cell supernatants recovered at 24 and 48 h (p < 0.01 compared with control). In vivo administration of AdCMVMuIFN-gamma exhibited 33% mortality by day 9 in mice acutely infected with T. gondii compared with 100% mortality in control mice (p = 0.045). Administration of AdCMVIL-12 reduced mortality to 40% compared with control mice. However, this reduction was not significant (p = 0.08). Overall survival was extended 2 days with AdCMVMuINF-gamma administration and 5 days with AdCMVIL-12. AdCMVMuIFN-gamma in vitro inhibits T. gondii, and in vivo AdCMVMuIFN-gamma and AdCMVIL-12 lead to increased survival in mice.

Pulmonary venous drainage into the left atrial appendage facilitates transplantation of the left lung with difficult exposure

Heterotopic implantation of the pulmonary venous confluence into the left atrial appendage during left lung transplantation is a reasonable alternative technique to reestablish venous drainage when exposure of the native left pulmonary veno-atrial connection may be problematic. We used this approach in a 39-year-old woman with chronic bronchiectasis who underwent bilateral sequential lung transplantation through a clam-shell approach. Dense hilar scarring and a small left atrial size made exposure of the native left pulmonary veno-atrial connection difficult.

Neutrophil inhibitory factor abrogates neutrophil adhesion by blockade of CD11a and CD11b beta(2) integrins

We studied the basis of inhibition of polymorphonuclear leukocyte (PMN) adhesion induced by neutrophil inhibitory factor (NIF), a 41-kDa CD11/CD18 beta(2) integrin-binding protein isolated from the canine hookworm (Ancylostoma caninum). NIF blocked PMN adhesion in a concentration-dependent manner with complete blockade occurring at approximately 10 nM NIF. Because CD11a and CD11b beta(2) integrins are functionally active on stimulated PMNs, and yet NIF is postulated to inhibit only CD11b integrin by binding to its I domain, we evaluated the contributions of CD11a and CD11b beta(2) integrins in the mechanism of inhibition of PMN adhesion to endothelial cells. We observed an additive inhibitory effect (>90% inhibition) of PMN adhesion to endothelial cells when NIF was used in combination with anti-CD11b monoclonal antibodies, which alone at saturating concentrations reduced PMN adhesion by only 50%. NIF also prevented aggregation of phorbol ester-stimulated JY lymphoblastoid cells that expressed only the functionally active CD11a, suggesting that NIF also can inhibit CD11a-dependent response. We transduced the NIF cDNA into human dermal microvessel endothelial cells in which NIF synthesis and release prevented PMN adhesion to the transduced human dermal microvessel endothelial cells. These data indicated that the potent antiadhesive effect of NIF may be the result of inhibition of CD11a and CD11b beta(2) integrins on PMNs. Moreover, the strategy of NIF release from transduced endothelial cells suggests the feasibility of blocking the CD11a- and CD11b beta(2) integrin-dependent PMN adhesion and PMN migration responses specifically at sites of endothelial cell activation.

Expression of PKA inhibitor (PKI) gene abolishes cAMP-mediated protection to endothelial barrier dysfunction

We investigated the hypothesis that cAMP-dependent protein kinase (PKA) protects against endothelial barrier dysfunction in response to proinflammatory mediators. An E1-, E3-, replication-deficient adenovirus (Ad) vector was constructed containing the complete sequence of PKA inhibitor (PKI) gene (AdPKI). Infection of human microvascular endothelial cells (HMEC) with AdPKI resulted in overexpression of PKI. Treatment with 0.5 microM thrombin increased transendothelial albumin clearance rate (0.012 +/- 0.003 and 0.035 +/- 0.005 microl/min for control and thrombin, respectively); the increase was prevented with forskolin + 3-isobutyl-1-methylxanthine (F + I) treatment. Overexpression of PKI resulted in abrogation of the F + I-induced inhibition of the permeability increase. However, with HMEC infected with ultraviolet-inactivated AdPKI, the F + I-induced inhibition was present. Also, F + I treatment of HMEC transfected with reporter plasmid containing the cAMP response element-directed transcription of the luciferase gene resulted in an almost threefold increase in luciferase activity. Overexpression of PKI inhibited this induction of luciferase activity. The results show that Ad-mediated overexpression of PKI in endothelial cells abrogated the cAMP-mediated protection against increased endothelial permeability, providing direct evidence that cAMP-dependent protein kinase promotes endothelial barrier function.

Selective inhibition of collagen gene expression in fibroblasts by an interferon-gamma transgene

Because interferon-gamma (IFN-gamma) selectively inhibits collagen gene expression, we hypothesized that expression of IFN-gamma cDNA in fibroblasts might be a useful strategy to inhibit the development of fibrosis. A replication-deficient E1-, E3- adenovirus vector encoding murine IFN-gamma (AdCM VmIFN gamma) was constructed. Infection of murine fibroblasts with AdCM VmIFN gamma in vitro was well tolerated. The results showed that IN F-gamma mRNA was expressed in infected cells, and as much as 17.7 ng of mIFN-gamma/10(6) cells was secreted into culture supernatants. Steady-state levels of alpha 1 (I) procollagen mRNA were decreased by 90% in infected cells compared to uninfected cells. The inhibition of collagen mRNA expression was partially abrogated with a neutralizing anti-mIFN-gamma antibody. Secretion of total collagen by AdCM VmIFN gamma-infected fibroblasts was decreased by 60% compared to uninfected cells. Induction cells. Induction of cytokine responsive gene-2 expression in AdCM VmIFN gamma-infected cells demonstrated that suppression of collagen production was a selective response. The results suggest a novel strategy of cytokine gene transfer and expression for the treatment of fibrotic lung diseases.

Augmentation of lung liquid clearance via adenovirus-mediated transfer of a Na,K-ATPase beta1 subunit gene

Previous studies have suggested that alveolar Na,K-ATPases play an important role in active Na+ transport and lung edema clearance. We reasoned that overexpression of Na,K-ATPase subunit genes could increase Na,K-ATPase function in lung epithelial cells and edema clearance in rat lungs. To test this hypothesis we produced replication deficient human type 5 adenoviruses containing cDNAs for the rat alpha1 and beta1 Na,K-ATPase subunits (adMRCMValpha1 and adMRCMVbeta1, respectively). As compared to controls, adMRCMVbeta1 increased beta1 subunit expression and Na,K-ATPase function by 2. 5-fold in alveolar type 2 epithelial cells and rat airway epithelial cell monolayers. No change in Na,K-ATPase function was noted after infection with adMRCMValpha1. Rat lungs infected with adMRCMVbeta1, but not adMRCMValpha1, had increased beta1 protein levels and lung liquid clearance 7 d after tracheal instillation. Alveolar epithelial permeability to Na+ and mannitol was mildly increased in animals infected with adMRCMVbeta1 and a similar Escherichia coli lacZ-expressing virus. Our data shows, for the first time, that transfer of the beta1 Na,K-ATPase subunit gene augments Na,K-ATPase function in epithelial cells and liquid clearance in rat lungs. Conceivably, overexpression of Na,K-ATPases could be used as a strategy to augment lung liquid clearance in patients with pulmonary edema.

Overexpression of the Na+,K+-ATPase alpha1 subunit increases Na+,K+-ATPase function in A549 cells

We hypothesized that viral mediated transfer of Na+,K+-ATPase subunit genes to alveolar epithelial cells to overexpress Na+, K+-ATPase could increase Na+,K+-ATPase function. We produced replication-deficient human type 5 adenoviruses that contained cytomegalovirus (CMV)-driven cDNAs for the rat alpha1 and beta1 subunits of Na+,K+-ATPase (AdMRCMValpha1 and AdMRCMVbeta1, respectively). These viruses were used to transduce human adenocarcinoma cells (A549) in culture. Na+,K+-ATPase function was increased by 2.5-fold in the AdMRCMValpha1-infected cells. Sham and AdMRCMVbeta1-infected cells, and cells infected by a CMV-driven beta-galactosidase-expressing adenovirus, had no increases in Na+, K+-ATPase activity. A549 cells infected with multiplicities of infection of 10-200 of AdMRCMValpha1 demonstrated expression of a rat alpha1 mRNA and increased alpha1 protein; no change in beta1 message or protein was noted. Ouabain sensitivity was measured in A549 cells following infection with AdMRCMValpha1. In contrast to controls, AdMRCMValpha1-infected cells demonstrated two IC50s. The first was similar to the IC50s of the controls; the second IC50 was 2 logs greater than the first, consistent with the presence of both the rat and human alpha1 isozymes. These results demonstrate for the first time that adenoviruses can be used to augment Na+,K+-ATPase function.

Aqueous smokeless tobacco extract impairs endothelium-dependent vasodilation in the oral mucosa

The purpose of this study was to determine whether an aqueous extract of smokeless tobacco (moist snuff) modulates vasomotor tone in the oral mucosa in situ and, if so, to determine the mechanisms that mediated these responses. Using intravital microscopy, we found that the extract had no significant effects on diameter of resistance (second-order) arterioles [44 +/- 5 (SD) microns] in the hamster cheek pouch. However, it significantly attenuated vasodilation elicited by two endothelium-dependent agonists, acetylcholine and bradykinin (P < 0.05). These effects were specific because smokeless tobacco extract had no significant effects on vasodilation elicited by nitroglycerin, an endothelium-independent agonist. Indomethacin, a cyclooxygenase inhibitor, and SQ-29548, a thromboxane A2/prostaglandin H2-receptor antagonist, abrogated the attenuating effects of smokeless tobacco extract on acetylcholine- and bradykinin-induced vasodilation. These data indicate that an aqueous extract of smokeless tobacco impairs endothelium-dependent vasodilation in the oral mucosa in situ in a specific fashion and that these effects are mediated, in part, by cyclooxygenase products of arachidonic acid metabolism that stimulate thromboxane A2/prostaglandin H2 receptors.

Stable VIP analogue Ro-24-9981 potentiates substance P-induced plasma exudation in hamster cheek pouch

The purpose of this study was to determine whether vasoactive intestinal peptide (VIP; 300 nM) and a stable cyclic analogue of VIP, Ro-24-9981 (226 nM), modulated neurogenic plasma exudation in the oral cavity in situ and, if so, to determine the mechanisms that mediated these responses. With the use of intravital microscopy, we found that suffusion of substance P induced a significant concentration-dependent formation of fluorescein-isothiocyanate-dextran (mol wt 70 kDa) leaky sites in the hamster cheek pouch (P < 0.05). These effects were significantly and stereospecifically attenuated by NG-nitro-L-arginine methyl ester, an inhibitor of NO synthase, and restored by L-arginine, the substrate for NO synthase (P < 0.05). Topical application of human VIP and Ro-24-9981 had no significant effects of leaky site formation. In addition, human VIP had no significant effects on substance P-induced responses. By contrast, Ro-24-9981 significantly potentiated substance P- and capsaicin-induced leaky site formation (P < 0.05). The effects of Ro-24-9981 on substance P-induced responses were significantly attenuated by NG-nitro-L-arginine methyl ester and restored by L-arginine (P < 0.05). Indomethacin had no significant effects on Ro-24-9981-induced responses. Ro-24-9981 had no significant effects on adenosine- and calcium ionophore A-23187-induced leaky site formation. Collectively, these data suggest that VIP plays no significant role in modulating neurogenic plasma exudation in the oral mucosa. By contrast, Ro-24-9981 amplified this response in a specific receptor-mediated fashion.

Gene therapy for the respiratory manifestations of cystic fibrosis

Cystic fibrosis (CF) is caused by mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. The major manifestations are on the airway epithelial surface, with purulent mucus, recurrent infections, chronic inflammation, and loss of lung function. Consequent to mutations in both parental genes, airway epithelial cells have insufficient CFTR function. Because this can be corrected in vitro by transfer of the normal CFTR gene into airway epithelial cells, it is reasonable to hypothesize that the respiratory manifestations of CF could be prevented by transfer of the normal human CFTR cDNA to the airway epithelium in vivo. Over the past 6 years, our laboratory has developed a strategy to accomplish this goal using a replication deficient E1-E3- recombinant adenovirus (Ad) serotype 5 vector containing the normal human CFTR cDNA (AdCFTR). Studies with experimental animals demonstrate that with administration of such a vector to the airways, the human CFTR cDNA could be transferred to the airway epithelium, with expression of the human CFTR cDNA for at least 6 weeks. Extensive preclinical studies in vitro and in vivo demonstrated that the risks to humans were sufficiently low to initiate a Phase I trial using the AdCFTR vector to treat the respiratory manifestations of CF in humans. Following approval by the National Heart, Lung, and Blood Institute Institutional Review Board, the National Institutes of Health Biosafety Committee, the National Institutes of Health Recombinant DNA Advisory Committee, and the Food and Drug Administration, we initiated the first human trial of gene therapy for CF on April 17, 1993. The clinical study is still ongoing, with safety and efficacy data being evaluated, but there is clear evidence that it is feasible to transfer and express the normal CFTR cDNA to the airway epithelium in vivo in individuals with CF.

Administration of an adenovirus containing the human CFTR cDNA to the respiratory tract of individuals with cystic fibrosis

We have administered a recombinant adenovirus vector (AdCFTR) containing the normal human CFTR cDNA to the nasal and bronchial epithelium of four individuals with cystic fibrosis (CF). We show that this vector can express the CFTR cDNA in the CF respiratory epithelium in vivo. With doses up to 2 x 10(9) pfu, there was no recombination/complementation or shedding of the vector or rise of neutralizing antibody titres. At 2 x 10(9) pfu, a transient systemic and pulmonary syndrome was observed, possibly mediated by interleukin-6. Follow-up at 6-12 months demonstrated no long term adverse effects. Thus, it is feasible to use an adenovirus vector to transfer and express the CFTR cDNA in the respiratory epithelium of individuals with CF. Correction of the CF phenotype of the airway epithelium might be achieved with this strategy.

Intraperitoneal in vivo gene therapy to deliver alpha 1-antitrypsin to the systemic circulation

The utility of replication-deficient recombinant adenovirus vector-mediated transfer and expression of the alpha 1-antitrypsin (alpha 1AT) cDNA to peritoneal mesothelial tissues was evaluated as a means of delivering alpha 1AT to the systemic circulation. Preliminary studies with Ad.RSV beta gal, an adenovirus vector expressing the Escherichia coli lacZ gene (beta-galactosidase), showed that intraperitoneal injection of 10(9) plaque-forming units (pfu) to cotton rats resulted in beta-galactosidase activity in mesothelial cells lining the peritoneal cavity. After intraperitoneal administration of 10(9) pfu of Ad alpha 1AT (an adenovirus vector containing the human alpha 1AT cDNA), human alpha 1AT was detectable in serum for up to 24 days, with a maximal level of 3.4 micrograms/ml at 4 days. Expression of the exogenous gene was localized to the peritoneal mesothelium as PCR analyses detected no evidence of expression of the exogenous gene in any other tissues evaluated. Anti-adenovirus vector antibodies were detectable in serum after intraperitoneal administration of the recombinant vectors, including antibodies with neutralizing activity. Repeat administrations of adenovirus vectors to the peritoneal cavity at 1 wk and 1 mo after the initial dose failed to show gene expression, but repeat administration 3 mo after demonstrated measurable gene transfer and expression. Together these observations suggest replication-deficient adenovirus-mediated gene transfer to the peritoneal mesothelium offers a promising means to transfer alpha 1AT to the systemic circulation, although immunity induced against the adenovirus may limit frequent repetitive dosing.

Direct in vivo gene transfer and expression in malignant cells using adenovirus vectors

To evaluate the ability of replication-deficient, recombinant adenovirus vectors to transfer genes to human tumor cells in vivo, adenovirus vectors containing the Escherichia coli lacZ (Ad.RSV beta gal) gene (coding for beta-galactosidase; used as a cell marker for gene transfer) or the human alpha 1-antitrypsin (Ad-alpha 1AT) cDNA (used as an example of a secreted protein) were administered intraperitoneally to nude mice with human malignant mesothelioma cell (H-MESO-1) malignant ascites. Preliminary in vitro studies showed that both vectors effectively transferred genes to H-MESO-1 cells. Tumor cells recovered from ascites of animals intraperitoneally administered a control adenovirus revealed no evidence of beta-galactosidase (beta-gal) activity 3 or 14 days later. In contrast, beta-gal activity was detected at the same time points in tumor cells from animals receiving intraperitoneal Ad.RSV beta gal. Flow cytometric quantification of beta-gal activity in recovered cells showed < 3% beta-gal-positive cells in animals administered control virus, but in animals administered intraperitoneal Ad.RSV beta gal there was a mean of 71 +/- 18% positive cells at 3 days and 56 +/- 27% at 14 days. Human alpha 1AT was not detected by enzyme-linked immunosorbent assay (ELISA) in ascites of animals receiving a control virus; however, in ascites of animals administered Ad-alpha 1AT, 21,000 +/- 3,800 ng/ml of human alpha 1AT was detected at 3 days and 4,900 +/- 1,700 ng/ml at 14 days. These data demonstrate that replication-deficient recombinant adenovirus vectors can be used to transfer genes to malignant cells in vivo and suggest a new strategy for genetic modification for antitumor therapy.

Ex vivo and in vivo gene transfer to the skin using replication-deficient recombinant adenovirus vectors

The skin has the potential for a variety of gene therapy applications. In addition to local delivery, it is the largest organ of the body, and highly vascular, and thus is an ideal site for systemic delivery of gene products. To evaluate the potential for adenovirus-mediated skin gene transfer, the replication-deficient recombinant adenovirus vectors Ad.RSV beta gal (coding for Escherichia coli beta-galactosidase) and Ad alpha 1AT (coding for human alpha 1-antitrypsin) were used in both ex vivo and in vivo approaches. Following in vitro infection with Ad.RSV beta gal, murine keratinocytes expressed beta-galactosidase. Parallel in vitro studies with Ad alpha 1AT documented de novo synthesis and secretion of human alpha 1AT as shown by [35S]methionine labeling and immunoprecipitation. Quantification of human alpha 1AT in the culture supernatants demonstrated 0.1-0.3 microgram human alpha 1AT secreted/ml-24 h. Evaluation of the serum of mice receiving transplants (10(5) cells/mouse) of Ad alpha 1AT-infected syngeneic keratinocytes demonstrated human alpha 1AT for at least 14 d with maximum levels of 41 ng/ml. To demonstrate the feasibility of direct adenovirus-mediated in vivo transfer of genes to the skin, Ad.RSV beta gal or Ad alpha 1AT were administered subcutaneously to mice. Histologic evaluation after 4 d demonstrated expression of beta-galactosidase in various types of skin cells. Quantification of human alpha 1AT in serum of animals infected subcutaneously with Ad alpha 1AT showed levels of 53 ng/ml at day 4, with human alpha 1AT detectable for at least 14 d. These observations support the feasibility of ex vivo and in vivo gene transfer to the skin mediated by replication-deficient adenovirus vectors.

Activation of alveolar macrophages in asymptomatic HIV-infected individuals

After the initial infection with HIV, there is evidence of immune dysfunction despite an apparent normal clinical state. In the context that the lung is a major site affected by opportunistic infection during the progression of this immune dysfunction, and that some components of the immune system are activated during early HIV infection, we hypothesized that there may be activation of alveolar macrophages (AM), a key component of the pulmonary host defense system, during the asymptomatic phase of HIV infection. Compared to normals, in HIV-infected individuals the class II MHC molecules DR, DQ, and DP were all expressed more frequently and in greater cell surface density on AM (p &lt; 0.03, all comparisons), and there was increased spontaneous release of superoxide anion (O2-.) by AM (p &lt; 0.002). To gain insight into whether the activation of the AM was an inherent property of the cells or dependent on the in vivo milieu, AM were evaluated after 24 h in culture for O2-. release. In contrast to the findings in fresh AM, after 24 h in culture, O2-. release by HIV AM was not different from normals (p &gt; 0.7), suggesting that these AM had been activated in vivo. To assess whether IFN-gamma could be mediating these effects, mRNA levels of the IP-10 gene (a gene specifically induced by increased concentrations of IFN-gamma) were quantified in AM. Strikingly, the IP-10 gene was expressed only in AM of HIV-seropositive individuals, suggesting the AM had been exposed to IFN-gamma in vivo. Overall, these observations are consistent with the concept that the HIV-seropositive state is associated with activation of AM, in part due to local exposure to IFN-gamma.

Activation of alveolar macrophages in asymptomatic HIV-infected individuals

After the initial infection with HIV, there is evidence of immune dysfunction despite an apparent normal clinical state. In the context that the lung is a major site affected by opportunistic infection during the progression of this immune dysfunction, and that some components of the immune system are activated during early HIV infection, we hypothesized that there may be activation of alveolar macrophages (AM), a key component of the pulmonary host defense system, during the asymptomatic phase of HIV infection. Compared to normals, in HIV-infected individuals the class II MHC molecules DR, DQ, and DP were all expressed more frequently and in greater cell surface density on AM (p < 0.03, all comparisons), and there was increased spontaneous release of superoxide anion (O2-.) by AM (p < 0.002). To gain insight into whether the activation of the AM was an inherent property of the cells or dependent on the in vivo milieu, AM were evaluated after 24 h in culture for O2-. release. In contrast to the findings in fresh AM, after 24 h in culture, O2-. release by HIV AM was not different from normals (p > 0.7), suggesting that these AM had been activated in vivo. To assess whether IFN-gamma could be mediating these effects, mRNA levels of the IP-10 gene (a gene specifically induced by increased concentrations of IFN-gamma) were quantified in AM. Strikingly, the IP-10 gene was expressed only in AM of HIV-seropositive individuals, suggesting the AM had been exposed to IFN-gamma in vivo. Overall, these observations are consistent with the concept that the HIV-seropositive state is associated with activation of AM, in part due to local exposure to IFN-gamma.

Adenovirus-mediated in vivo gene transfer and expression in normal rat liver

Replication deficient, recombinant adenovirus (Ad) vectors do not require target cell replication for transfer and expression of exogenous genes and thus may be useful for in vivo gene therapy in hepatocytes. In vitro, primary cultures of rat hepatocytes infected with a recombinant Ad containing a human alpha 1-antitrypsin cDNA (Ad-alpha 1AT) synthesized and secreted human alpha 1AT for 4 weeks. In rats, in vivo intraportal administration of a recombinant Ad containing the E. coli lacZ gene, was followed by expression of beta-galactosidase in hepatocytes 3 days after infection. Intraportal infusion of Ad-alpha 1AT produced detectable serum levels of human alpha 1AT for 4 weeks. Thus, targeted gene expression has been achieved in the liver, albeit at low levels, suggesting that adenovirus vectors may be a useful means for in vivo gene therapy in liver disorders.

Adenovirus-mediated transfer of a recombinant human alpha 1-antitrypsin cDNA to human endothelial cells

To evaluate the feasibility of using a replication-deficient recombinant adenovirus to transfer human genes to the human endothelium, human umbilical vein endothelial cells were infected in vitro with adenovirus vectors containing the lacZ gene or a human alpha 1-antitrypsin (alpha 1AT) cDNA. After in vitro infection with the lacZ adenovirus vector, cultured endothelial cells expressed beta-galactosidase. In parallel studies with the alpha 1AT adenovirus vector, infected cells expressed human alpha 1AT transcripts, as evidenced by in situ hybridization and Northern analysis, and de novo synthesized and secreted glycosylated, functional alpha 1AT within 6 hr of infection, as shown by [35S]methionine labeling and immunoprecipitation. Quantification of the culture supernatants demonstrated 0.3-0.6 micrograms of human alpha 1AT secreted per 10(6) cells in 24 hr, for at least 14 days after adenovirus vector infection. To demonstrate the feasibility of direct transfer of genes into endothelial cells in human blood vessels, lacZ or alpha 1AT adenovirus vectors were placed in the lumen of intact human umbilical veins ex vivo. Histologic evaluation of the veins after 24 hr demonstrated transfer and expression of the lacZ gene specifically to the endothelium. alpha 1AT adenovirus infection resulted both in expression of alpha 1AT transcripts in the endothelium and in de novo synthesis and secretion of alpha 1AT. Quantification of alpha 1AT in the vein perfusates showed average levels of 13 micrograms/ml after 24 hr. These observations strongly support the feasibility of in vivo human gene transfer to the endothelium mediated by replication-deficient adenovirus vectors.

Interleukin-8 gene expression in human bronchial epithelial cells

The capacity of cells of the human bronchial epithelium to express the gene for interleukin-8 (IL-8) was evaluated in bronchial epithelium derived cell lines, HS-24 and BET-1A, using tumor necrosis factor-alpha (TNF) as a model inflammatory stimulus. As in other epithelium, TNF markedly increased the level of the 1.8-kilobase IL-8 mRNA transcripts in both bronchial epithelial cell lines. In HS-24 cells, nuclear run-on analyses showed the IL-8 gene transcription rate was dramatically increased, more than 30-fold, after TNF stimulation. The half-life of IL-8 mRNA transcripts in these cells was approximately 40 min and did not change after TNF stimulation, suggesting that TNF up-regulated IL-8 gene expression mainly at the transcriptional level. DNase I hypersensitivity site mapping of chromatin DNA in resting HS-24 cells demonstrated two hypersensitivity sites within 400 base pairs (bp) 5' to exon I and one site within exon I. However, after TNF stimulation, the exon I hypersensitivity site disappeared and a new site approximately 120 bp 5' to exon I emerged. Consistent with these observations, transfection studies with HS-24 cells using fusion genes composed of the 5'-flanking sequences of the IL-8 gene and a luciferase reporter gene demonstrated potent promoter activity in a 174-bp segment (-130 to +44 relative to the transcription start site), which also exhibited a response to TNF, while a segment from -112 to +44 showed very low promoter activity and no response to TNF. Thus, human bronchial epithelial cells can express the IL-8 gene, with expression in response to the inflammatory mediator TNF regulated mainly at the transcriptional level, and with elements within the 5'-flanking region of the gene that are directly or indirectly modulated by the TNF signal.

Organ specific cytokine therapy. Local activation of mononuclear phagocytes by delivery of an aerosol of recombinant interferon-gamma to the human lung

In the context of the central role of the alveolar macrophage in host defense of the respiratory epithelial surface, and the ability of IFN-gamma to activate mononuclear phagocytes, we have evaluated strategies to use rIFN-gamma to activate human alveolar macrophages in vivo. To accomplish this, rIFN-gamma was administered to nonsmoking normals, the amounts of IFN-gamma quantified in serum and respiratory epithelial lining fluid (ELF) and the status of IFN-gamma related activation of blood monocytes and alveolar macrophages was evaluated by quantifying the expression of mRNA transcripts of IP-10, a gene induced specifically by IFN-gamma. Systemic administration (subcutaneous) of maximally tolerated amounts of rIFN-gamma (250 micrograms) was followed by detectable levels of IFN-gamma in serum but not ELF, the expression of IP-10 transcripts in blood monocytes but not alveolar macrophages, and multiple systemic adverse effects. To circumvent the inability of systemic administration to reach respiratory ELF and activate alveolar macrophages, rIFN-gamma (250-1,000 micrograms) was inhaled as an aerosol once daily for 3 d. Strikingly, while IFN-gamma was not detected in serum it was detectable in respiratory ELF in a dose-dependent fashion. Further, alveolar macrophages, but not blood monocytes, expressed IP-10 mRNA transcripts and, importantly, inhalation of aerosolized rIFN-gamma was not associated with local or systemic adverse effects. Thus, it is feasible to use rIFN-gamma to activate alveolar macrophages by targeting the cytokine directly to the lung. These data suggest a potential strategy for targeted cytokine therapy, without systemic side effects, to augment respiratory tract defenses in individuals at risk for or with lung infection.

Systemic glutathione deficiency in symptom-free HIV-seropositive individuals

To find out whether systemic glutathione deficiency is associated with human immunodeficiency virus (HIV) infection, thus contributing to the immunodeficiency state, glutathione concentrations in venous plasma and lung epithelial lining fluid (ELF) of symptom-free HIV-seropositive and normal individuals were measured. Total and reduced glutathione concentrations in the plasma of the HIV-infected subjects were about 30% of those in the normal individuals. Concentrations of these substances in the ELF of HIV-infected subjects were about 60% of those in the controls. There was no correlation between ELF and plasma concentrations of total or reduced glutathione. Since glutathione enhances immune function, glutathione deficiency may contribute to the progressive immune dysfunction of HIV infection.

Relationships between pulmonary inflammation, plasma transudation, and oxygen metabolite secretion by alveolar macrophages

We have previously shown that alveolar macrophages from normal rabbit lungs do not synthesize reactive oxygen intermediates unless first conditioned by culture in vitro in the presence of serum for 24 to 48 hr. This conditioning process is mediated by a serum constituent that partitions on gel exclusion columns with an apparent m.w. of 30,000 to 50,000 daltons. Alveolar macrophage conditioning in vitro requires protein synthesis, is associated with the generation of membrane NADPH oxidase activity, and is reversible. We have predicted therefore that during the course of pulmonary inflammation, as observed 3 wk after i.v. injection of M. butyricum in oil, alveolar macrophages might similarly become conditioned in vivo through exposure to plasma protein transudates reaching the alveolus. In support of this hypothesis we show that after experimental production of granulomatous pulmonary inflammation in rabbits, alveolar macrophages showed an augmented capacity to secrete superoxide anion when stimulated with phorbol ester, and this enhancement increases exponentially with increased plasma transudation. This augmented enhancement was reversible, and decreased after culture in vitro in the absence of serum. Mature alveolar macrophages were responsible for this enhanced superoxide anion production rather than freshly emigrated monocytes. Moreover, superoxide anion production in this model of pulmonary inflammation appears to be an "all-or-none" phenomenon, with superoxide anion production associated with a subpopulation of optimally conditioned alveolar macrophages, whereas the remaining unconditioned alveolar macrophages produce little or none. We feel that these two classes of alveolar macrophages may be derived from inflamed and noninflamed regions of the lung, respectively, thereby reflecting the discontinuous nature of the inflammatory lesions themselves. Thus we propose that measurements of reactive oxygen intermediate production by lavaged alveolar macrophages may provide a semi-quantitative measure of chronic pulmonary inflammation.

Relationships between pulmonary inflammation, plasma transudation, and oxygen metabolite secretion by alveolar macrophages

We have previously shown that alveolar macrophages from normal rabbit lungs do not synthesize reactive oxygen intermediates unless first conditioned by culture in vitro in the presence of serum for 24 to 48 hr. This conditioning process is mediated by a serum constituent that partitions on gel exclusion columns with an apparent m.w. of 30,000 to 50,000 daltons. Alveolar macrophage conditioning in vitro requires protein synthesis, is associated with the generation of membrane NADPH oxidase activity, and is reversible. We have predicted therefore that during the course of pulmonary inflammation, as observed 3 wk after i.v. injection of M. butyricum in oil, alveolar macrophages might similarly become conditioned in vivo through exposure to plasma protein transudates reaching the alveolus. In support of this hypothesis we show that after experimental production of granulomatous pulmonary inflammation in rabbits, alveolar macrophages showed an augmented capacity to secrete superoxide anion when stimulated with phorbol ester, and this enhancement increases exponentially with increased plasma transudation. This augmented enhancement was reversible, and decreased after culture in vitro in the absence of serum. Mature alveolar macrophages were responsible for this enhanced superoxide anion production rather than freshly emigrated monocytes. Moreover, superoxide anion production in this model of pulmonary inflammation appears to be an "all-or-none" phenomenon, with superoxide anion production associated with a subpopulation of optimally conditioned alveolar macrophages, whereas the remaining unconditioned alveolar macrophages produce little or none. We feel that these two classes of alveolar macrophages may be derived from inflamed and noninflamed regions of the lung, respectively, thereby reflecting the discontinuous nature of the inflammatory lesions themselves. Thus we propose that measurements of reactive oxygen intermediate production by lavaged alveolar macrophages may provide a semi-quantitative measure of chronic pulmonary inflammation.

Studies on the transfer of protective immunity with lymphoid cells from mice immune to malaria sporozoites

In an effort to understand the mechanisms involved in the protective immunity to malarial sporozoites, an A/J mouse/Plasmodium berghei model was studied. Protective immunity could consistently be adoptively transferred only by using sublethal irradiation of recipients (500 R); a spleen equivalent (100 X 10(6))of donor cells from immune syngeneic mice; and a small booster immunization (1 X 10(4)) of recipients with irradiation-attenuated sporozoites. Recipient animals treated in this manner were protected from lethal challenge with 1 X 10(4) nonattenuated sporozoites. Immune and nonimmune serum and spleen cells from nonimmune animals did not protect recipient mice. Fewer immune spleen cells (50 X 10(6)) protected some recipients. In vitro treatment of immune spleen cells with anti-theta sera and complement abolished their ability to transfer protection. This preliminary study suggests that protective sporozoite immunity can be transferred with cells, and that it is T cell dependent.