Inflammation, Vol. 24, No. 4, 2000
MODULATION OF HYPEROXIA-INDUCED TNF-a EXPRESSION IN THE NEWBORN RAT LUNG BY THALIDOMIDE AND DEXAMETHASONE LEARIE LINDSAY,1 STEPHEN J. OLIVER,2 SHERRY L. FREEMAN,2 REGIS JOSIEN,2 ALFRED KRAUSS,1 and GILLA KAPLAN2 1Department of Neonatology New York Hospital-Cornell Medical Center New York, New York 10021 2Laboratory of Cellular Physiology and Immunology The Rockefeller University 1230 York Avenue New York, New York 10021
Abstract—The effect of high oxygen concentrations on lungs of neonatal rats was studied. In addition, some oxygen-exposed animals were treated with either dexamethasone or thalidomide. No gross histologic changes were noted in the lungs following exposure to 95% oxygen nor were there changes in the total number or the phenotypic distribution of BAL cells obtained from these lungs compared to lungs from air exposed (control) neonatal rats. The majority of the BAL cells were CD45+ leukocytes (macrophages). However, when BAL cells were exposed to LPS in vitro, TNF-a production was higher in cells from rats exposed to 95% oxygen compared to cells from rats exposed to ambient air. In addition, lung TNF-a and IL-6 mRNA levels were increased after exposure to 95% oxygen. In the lungs of animals treated with either dexamethasone or thalidomide, TNF-a mRNA levels were reduced, while only dexamethasone treatment also reduced IL-6 mRNA levels.
INTRODUCTION There are approximately 4 million births each year in the USA (1). About 1% of these infants are born prematurely and may develop respiratory distress syndrome (RDS) which often requires oxygen therapy. Of the infants who develop RDS and are treated with high concentrations of oxygen, approximately 20 to 30% go on to develop bronchopulmonary dysplasia (BPD), a form of chronic lung disease due to barotrauma, oxygen toxicity, and immaturity of the pulmonary structures (2). Exposure of the neonatal lung to high levels of oxygen results in pulmonary inflammatory injury very similar to the injury that occurs in 347 0360-3997/ 00/ 0800-0347$18.00/ 0 2000 Plenum Publishing Corporation
Lindsay et al.
348
the adult respiratory distress syndrome (ARDS) (3). At present, the mechanism underlying the lung damage in newborns due to high oxygen concentration as well as the effects of the duration of exposure are unknown. Lung disease caused by prolonged exposure to high concentrations of oxygen has been reported to be mediated by oxygen free-radicals generated by the macrophages of the lung during a respiratory burst (4). The oxygen metabolites can cause cell injury by lipid peroxidation of biological membranes through the formation of fatty acid radicals. These can further react with oxygen to form fatty acid peroxy radicals (5). In addition oxygen radicals may activate the nuclear transcription factor NF-k B (6, 7), which when translocated to the nucleus of the macrophage can activate the expression of genes involved in the inflammatory response including the genes for cytokines such as interleukin-1 beta (IL-1b), tumor necrosis factor-alpha (TNF-a) and IL-6 (7). In the present study, we examined the effects of hyperoxia in the newborn lung on TNF-a and IL-6 production by resident macrophage populations. Newborn rats were exposed to either ambient air or 95% oxygen. Bronchoalveolar lavage (BAL) was carried out and the cells obtained were counted, phenotyped, and evaluated for cytokine mRNA expression and lipopolysaccharide (LPS) induced cytokine production in vitro. In addition, some of the newborn rats were treated with either thalidomide or dexamethasone to determine whether these immunomodulatory drugs would alter the lung cytokine response to high oxygen concentration.
MATERIALS AND METHODS Animal Model. Sprague Dawley inbred pregnant rats (viral antibody-free) were obtained from Taconic Farms (German Town, New York). Spontaneously delivered newborns from the same litter each weighing 7.0 to 7.5 grams were used for each series of experiments. Within a few hours after birth, the newborn rats were divided into two groups and exposed to 95% oxygen or ambient air for either 24 h or 48 h. Newborn rats exposed to 95% oxygen were kept with their dams in plastic hoods (12 in. × 6 in. × 12 in.), equipped with continuous gas flow devices. Oxygen levels were monitored (Oxygen monitor 406, Instrument Lab, Lexington, Massachusetts). These hoods contained glove ports enabling the handling of dams and the newborn animals. Nursing dams were exchanged between control and oxygen exposed litters every 12 h in order to reduce oxygen intoxication of the dams. The newborn rats exposed to air were kept in similar plastic hoods supplied with ambient air. All nutritional needs of the animals were met and environmental conditions were stable. In some experiments newborn rats were treated with either dexamethasone (100 mg/ kg/ day) or thalidomide (100 mg/ kg/ day) at the start of the study. The protocol was approved by the Institutional Animal Care and Use Committee of Cornell University Medical College and the Laboratory Animal Research Center of The Rockefeller University. BAL Cells. Newborn rats were anesthetized with intraperitoneal injections of ketamine (80 mg/ kg) (Fort Dodge Animal Health, Fort Dodge, Iowa) and xylazine (7.5 mg/ kg) (Miles Laboratories, Shawnee, Kansas) using a 27 gauge needle. The area above the larynx was cleaned with
Modulation of Hypoxia
349
alcohol and the tissue covering the trachea was dissected away, care being taken to avoid damage to adjacent blood vessels. The trachea was then cannulated with size 24 gauge Angiocath (Becton Dickinson Vascular Access, Sandy, Utah). The lung was lavaged with 200 ml aliquots of sterile PBS (Gibco BRL, Grand Island, New York) and 1% lidocaine (Sigma Chemical Co., St. Louis, Missouri) to a total volume of 1000 ml. Due to the relatively low number of cell retrievable from BALS of individual neonatal rats, the BAL fluid from the total number of rats in each experimental group (air-exposed or 95% oxygen treated) were pooled and stored on ice (48 C). The cells obtained from the BAL were counted in a hemocytometer and stained with trypan blue to test for cell viability. Since the number of cells obtained was limited, the BAL cells obtained from each group of animals were used for either LPS stimulation in vitro, phagocytosis experiments or flow cytometric analysis. LPS Stimulation and TNF-a Assays. BAL cells were suspended at 1 × 106 cells/ ml in RPMI medium (Gibco BRL) supplemented with 10% heat inactivated fetal calf serum (R10) and 150 ml was added per well to 96 well plates. LPS at a concentration of 1 mg/ ml was then immediately added to each well. Plates were incubated at 378 C in 5% CO2 . The supernatants were collected after 1, 3 and 16 h. TNF-a levels were measured by enzyme-linked immunosorbent assays (ELISA) (Genzyme Corp., Boston, Massachusetts) according to the manufacturer’s specification. Phagocytosis of Latex Particles. 100 ml of BAL cell suspended at 1 × 107 / ml were placed onto sterile 13 mm glass coverslips and incubated at 378 C in 5% CO2 for 2 h. Non-adherent cells were removed by washing in prewarmed R10 and coverslips were placed in a 24-well culture plate. One ml of R10 media containing fluorescent latex beads [FluoSphere carboxylated-modified microspheres, 1.0 mm, yellow-green fluorescent (505/ 515 nm)] (Molecular Probes, Eugene, Oregon) at a concentration of 100 : 1 was added to each well and the plate was then incubated for 4 h at 378 C in 5% CO2 . The cells on coverslips were fixed with 10% formalin, washed in PBS, air-dried, and stained with Diff-Quik preparation. Cover slips were examined using fluorescent microscopy (Nikon, Garden City, New York). Immunofluorescent Staining of Cells for Flow Cytometry. FACS analyses were performed on BAL cells obtained from both air exposed and 95% oxygen exposed newborn rats. The phenotypic distribution of the cells was investigated by staining with monoclonal antibodies according to the manufacturer instructions (PharMingen Research Products, San Diego, California) and FACS analysis. Mouse anti-rat monoclonal antibodies used were: CD11b/ 11c, granulocyte, OX6* (RT1B(class II)], OX33*, (CD45 RA), OX34* (CD2), OX42* [CD11b(Mac-1)], OX62* (Integrin on DC + g d T cells), V65* (TcRg d), ED1* [CD68 like (Mac/ mono/ DC)], ED2* (Macroph), ED3* [Macroph. (Silaloadhesin)] and 3.2.3♦ [NKR-P1 (CD161)]. OX12* (Ig k chain), CD45* (OX1 + OX30), OX42* [CD11 b (Mac-1)], CD4* (W3–25), OX7* [Thy 1.1 (CD90)] (*ECACC—European Collection of Animal Cell Cultures,† provided by Dr. T. Hunig, Wurzburg, Germany, ♦Serotec, Oxford, English). Histology of the Lung. Newborn rats were euthanized by intraperitoneal injections of ketamine and xylazine. After instilling 200 ml of 10% formalin into the trachea using a 24 gauge Angiocath, enbloc dissections of both right and left lungs were performed. Sections of paraffin embedded tissue were stained with hematoxylin-eosin and examined by light microscopy. Cytokine mRNA Evaluation. Enbloc dissections of both lungs were performed as above, except without the addition of formalin, and samples were immediately snap frozen in alcohol and dry ice. Total cellular RNA was extracted from whole newborn rat lungs using a modified single-step procedure developed by Cinna/ TEL-Test, Inc (Friendswood, Texas). The RT-PCR reaction was carried out using specific primers for rat TNF-a and IL-6 mRNA (Clonetech, Palo Alto, California according to the protocol supplied with the Perkin Elmer Gene Amp RNA PCR kit, modified to incorporate 32 P dCTP during amplification). On completion of the amplification reaction, the cDNA was separated on a 1% agarose gel. The gel was then dried and exposed to Kodak film overnight and then densitometric analysis of the amplification bands was carried out in a phosphorimager (ImageQuant,
Lindsay et al.
350
Molecular Dynamics, Sunnyvale, Calfornia). The results obtained were normalized to the density of b-actin mRNA. Statistical Analysis. Due to the small sample sizes used in this study, formal statistical analysis was not done. The results are presented graphically with means and standard deviations.
RESULTS BAL Cell Numbers, Phenotype and Function. The average number of cells obtained from the BAL of the newborn rats did not differ significantly when the animals were exposed for 24 h to either ambient air or 95% oxygen (Table 1). Similar increases in the total number of cells were also seen in newborn rats exposed for 48 h to either ambient air or 95% oxygen. Flow cytometric analysis revealed that approximately 70% to 80% of the resident alveolar cells were CD45 positive (Figure 1 and Table 1). The percent of CD45+ cells in BAL as well as other cell phenotypes was not significantly affected by exposure of the rats to 95% oxygen compared to ambient air. About 1–2% of the BAL cells were CD4+ or CD8+ T cells. Similarly, about 1–2% of the BAL cells stained with a monoclonal antibody specific for polymorphonuclear leukocytes (PMN). Thus, exposure of the newborn rat lung to high concentrations of oxygen for up to 48 h modified neither the total number of leukocytes in the lungs nor their phenotypic distribution compared to ambient air exposure. To confirm that the 68–85% of BAL cells that stained for CD45 were indeed functionally active macrophages, the BAL cells from newborn rats exposed to ambient air for 24 h were introduced into tissue culture and incubated with fluo-
Fig. 1. Flowcytometric analysis of BAL cells obtained from newborn rat lungs exposed to ambient air or 95% oxygen for 24 or 48 h. The majority of the BAL cells were CD45+ (macrophages). CD4+, CD8+ and PMN were each less than 2% of total BAL cells (see Table 1).
Modulation of Hypoxia
351
Table 1. Phenotype of BAL Cells from the Newborn Rat Lung Leukocyte phenotype (% of total cells)b Treatment Air 24 hrs Air 48 hrs Oxygen 24 hrs Oxygen 48 hrs
Total Cell Count (cells/ animal × 105 )a
CD45+
CD4+
CD8+
PMN
± ± ± ±
84.8 68.6 78.4 78.1
1.1 0.7 1.6 1.5
NDc NDc 0.5 1.2
1.1 0.7 1.5 1.9
4.1 5.2 4.5 5.0
1.1 1.4 1.3 0.3
aTotal cell counts: results are means ± SD of 3–7 experiments with 5–7 animals per bCalculated from FACS analysis: results are from one representative experiment. cND: not done.
experiment.
rescent latex beads (FluoSphere), at a dilution of 100 : 1 for 1 h. The cells were then examined by normal light and fluorescent microscopy to confirm phagocytic ingestion of the spheres. In a separate experiment, phagocytic uptake of the fluorescent beads was evaluated using flow cytometry. The percent of cells that phagocytosed the particles was found to be 74.4 ± 4.1%, confirming that the majority of the BAL cells were indeed phagocytic macrophages. Effect of Hyperoxia on TNF-a Production by BAL Cells. We examined the effect of exposure of newborn rats to high concentrations of oxygen on LPS induced TNF-a production by BAL cells in vitro. BAL cells were obtained from newborn rats exposed to either ambient air or to 95% oxygen for 24 h or 48 h and introduced into culture. LPS stimulation of BAL cells resulted in increased TNF-a levels in the culture supernatants of all treatment groups. BAL cells from neonatal rats exposed to ambient air for 48 h released higher peak TNF-a levels compared to levels observed in supernatants of BAL cell from pups exposed for only 24 h (Figure 2). Following exposure to 95% oxygen, TNF-a production by LPS stimulated BAL cells was induced earlier and reached an even higher level. This response was also proportional to the duration of oxygen exposure (48 h > 24 h) (Figure 2). No detectable TNF-a was produced by unstimulated cells obtained from rats in all treatment groups. Effect of Hyperoxia on TNF-a mRNA Levels in the Lungs of Newborn Rats. We next obtained whole lungs from newborn rats exposed to either 95% oxygen or ambient air for 48 h. When the lungs were examined histologically, no gross differences were noted (not shown). We then assessed the induction of cytokine mRNA using RT-PCR for TNF-a and IL-6 mRNA. There was a significant increase in TNF-a mRNA in the whole lungs of the newborn rats exposed to 95% oxygen for 48 h compared to the level in lungs of newborn rats exposed to ambient air (Figure 3). The effects of the two immunomodulatory drugs, thalidomide or dexamethasone, on TNF-a and IL-6 mRNA production in the hyperoxic newborn
352
Lindsay et al.
Fig. 2. TNF-a levels induced in LPS stimulated adherent BAL cells (macrophages) from rats exposed to ambient air (triangles) or 95% oxygen (circles) for 24 h (open symbols) and 48 h (closed symbols). Results are means ± SD of 3 experiments containing 5–7 animals per experimental group.
rat lungs was next evaluated. When thalidomide or dexamethasone was administered to the newborn rats before and during exposure to 95% oxygen for 48 h, there was a decrease in TNF-a mRNA obtained from the lung in response to either drug (Figure 4). A 50% inhibition of TNF-a mRNA was observed in response to thalidomide treatment. Similarly, a 47% inhibition of TNF-a mRNA was seen in the dexamethasone treated rats, as compared to the oxygen exposed untreated rats. Dexamethasone treatment was found to also reduce IL-6 mRNA
Fig. 3. TNF-a mRNA levels in neonatal rat lungs exposed to ambient air or 95% oxygen for 48 h. Results are means of 7 animals/ per group ± SD.
Modulation of Hypoxia
353
Fig. 4. The effect of thalidomide and dexamethasone on TNF-a and IL-6 mRNA levels in the lungs of hyperoxic newborn rats exposed to 95% oxygen for 48 h. Results are means ± SD of 2 experiments with 2–3 animals per experiment.
levels in the whole lung (Figure 4). This effect on IL-6 mRNA was not seen with thalidomide. Thus thalidomide appeared to selectively inhibit TNF-a mRNA induction in the lungs of newborn rats exposed to 95% oxygen.
DISCUSSION One of the most serious problems accompanying premature birth is the lung damage which occurs following treatment of the neonate with high oxygen concentrations. To study the mechanisms underlying this toxicity we established a neonatal rat model of oxygen exposure. This model simulates conditions that are experienced by the human premature infant lung, where exposure to high concentrations of oxygen in conjunction with immaturity of the pulmonary structures can cause damage to the lung parenchyma (2). In our study, newborn rats were exposed to either ambient air or 95% oxygen for up to 48 h. We found that hyperoxia activated alveolar macrophages to express elevated levels of TNF-a mRNA. When ex vivo adherent BAL cells
354
Lindsay et al.
from these animals were further stimulated in vitro by LPS, there was a significantly elevated level of TNF-a production compared to BAL cells obtained from ambient air-exposed neonatal rats. This activation occurred in the absence of any gross histologic changes in the lung and without obvious recruitment of additional leukocytes to the lung. Thus it appears that high oxygen concentrations over a 48 h period primed resident alveolar macrophages to increase TNF-a synthesis. Endogenous TNF-a produced during hyperoxic exposure may have deleterious effects on the lungs for a number of reasons. TNF-a has been shown to be either cytostatic or cytotoxic to endothelial cells (8, 9). The cytokine causes distinct morphological changes, redistribution of actin filaments, and increased albumin permeability of endothelial monolayers (10, 11). TNF-a also enhances hyperoxia-induced endothelial cell injury (7). In addition TNF-a stimulates the production of reactive oxygen species by PMNs (12) and potentiates neutrophilmediated endothelial injury (13, 14). It has been shown that prolonged exposure of rats to hyperoxia for greater than 3 days results in recruitment of large numbers of PMNs into the alveolar spaces (15). PMNs migrating into the lungs may themselves become a source of TNF-a. TNF-a has autocrine effects and can stimulate PMNs to express IL-8, IL-6 and IL-1b. In addition TNF-a has been shown to induce the release of IL-8 from a variety of tissue cells. Finally, TNF-a can stimulate oxygen radical release from the PMNs. Oxygen radicals can then in turn cause damage to the structural components of the lung tissue (16). In the neonatal rat model described here, the effect of oxygen on the ability of lung macrophages to produce TNF-a preceded any recruitment of PMNs. Thus, PMN produced tissue damage may be secondary to TNF-a production by the resident cells of the lung, but once PMNs are recruited to the lung, they may amplify the damage. Dexamethasone is presently one of the mainstays of treatment in the management of chronic lung disease of premature infants (17). It is a powerful suppressor of the inflammatory response. However, its use is fraught with complications, such as hypertension, hyperglycemia and increased risk of infection. Thalidomide is an immune modulatory drug that has been shown both in vitro and in vivo to inhibit TNF-a production by LPS-stimulated macrophages (18–20). In our experimental model, treatment of the hyperoxic newborn rats with either dexamethasone or thalidomide caused an almost 50% reduction in TNF-a mRNA expression in the lungs. Dexamethasone also caused a reduction in IL-6 mRNA levels. Thus, thalidomide appeared to more selectively inhibit TNF-a production by resident alveolar cells in this model. The effects of early TNF-a inhibition by thalidomide on PMN migration and other downstream inflammatory processes that result in lung damage in the hypoxic neonatal rat model remain to be studied. Also, thalidomide’s use has not yet been evaluated in newborn human infants. However, its efficacy in inhibiting
Modulation of Hypoxia
355
TNF-a production by alveolar macrophages without depressing the production of other important cytokines such as IL-6, suggests a possible role in the prevention and treatment of lung disease associated with high oxygen therapy of premature infants. Acknowledgments—The authors thank Dr. Victoria H. Freedman for help with the manuscript, Marguerite Nulty for secretarial assistance and Judy Adams for her assistance with the figures. This study was supported in part by Celgene Corporation and by a U.S. Public Health Service grant AI 22626 to G.K.
REFERENCES 1. National Center for Health Statistic. Health United States 1991 and Prevention Profile Hyattsville, Maryland: Public Health Service, 1992. 2. NORTHWAY, W. H., JR. 1992. Bronchopulmonary dysplasia: twenty-five years later. Pediatrics, 89:969–973. 3. CONTRERAS, M., N. HARIHARAN, J. R. LEWANDOSKI, W. CIESIELSKI, R. KOSCIK, and J. J. ZIMMERMAN, 1996. Bronchoalveolar oxyradical inflammatory elements herald bronchopulmonary dysplasia. Crit. Care Med., 24:29–37. 4. MUSTAFA, M. G., and D. F. TIERNEY. 1978. Biochemical and metabolic changes in the lung with oxygen, ozone, and nitrogen dioxide toxicity. Am. Rev. Respir. Dis., 118:1061–90. 5. FANTONE, J. C., and P. A. WARD. 1982. Role of oxygen-derived free radicals and metabolites in leukocyte-dependent inflammatory reactions. Am. J. Pathol., 107:395–418. 6. TAKASUKA, N., K. MATSUURA, S. YAMAMOTO, and K. S. AKAGAWA. 1995. Suppression of TNF-a mRNA expression in LPS-primed macrophages occurs at the level of nuclear factork B activation, but not at the level of protein kinase C or CD14 expression. J. Immunol., 154:4803–4812. 7. SCHRECK, R., P. RIEBER, and P. A. BAEUERLE. 1997. Reactive oxygen intermediates as apparently widely used messengers in the activation of the NF-k B transcription factor and HIV-1. Embo. J., 10:2247–2258. 8. MARCHO, Z., J. E. WHITE, P. J. HIGGINS, and M. F. TSAN. 1992. Tumor necrosis factor enhances endothelial cell susceptibility to oxygen toxicity: role of glutathione. Am. J. Respir. Cell. Mol. Biol., 5:556–562. 9. SATO, N., T. GOTO, K. HARANAKA, N. SATOMI, H. NARIUCHI, Y. MANO-HIRANO, and Y. SAWASAKI. 1986. Actions of tumor necrosis factor on cultured vascular endothelial cells: morphological modulation, growth inhibition, and cytotoxicity. J. Natl. Cancer Inst., 76:1113–1121. 10. ROYALL, J. A., R. L. BERKOW, J. S. BECKMAN, M. K. CUNNINGHAM, S. MATALON, and B. A. FREEMAN. 1989. Tumor necrosis factor and interleukin 1 alpha increase vascular endothelial permeability. Am. J. Physiol., 257:L399–410. 11. STOLPEN, A. H., E. C. GUINAN, W. FIERS, and J. S. POBER. 1986. Recombinant tumor necrosis factor and immune interferon act singly and in combination to reorganize human vascular endothelial cell monolayers. Am. J. Pathol., 123:16–24. 12. SHALABY, M. R., B. B. AGGARWAL, E. RINDERKNECHT, L. P. SVEDERSKY, B. S. FINKLE, and M. A. PALLADINO, JR. 1985. Activation of human polymorphonuclear neutrophil functions by interferon-gamma and tumor necrosis factors. J. Immunol., 135:2069–2073. 13. TSAN, M. F., J. E. WHITE, P. B. MICHELSEN, and G. H. WONG. 1995. Pulmonary O2 toxicity: role of endogenous tumor necrosis factor. Exp. Lung Res., 21:589–597.
356
Lindsay et al.
14. GIBBS, L. S., L. LAI, and A. B. MALIK. 1990. Tumor necrosis factor enhances the neutrophildependent increase in endothelial permeability. J. Cell Physiol., 145:496–500. 15. HORINOUCHI, H., C. C. WANG, K. E. SHEPHERD, and R. JONES. 1996. TNF alpha gene and protein expression in alveolar macrophages in acute and chronic hyperoxia-induced lung injury. Am. J. Respir. Cell Mol. Biol., 14:548–555. 16. CARP, H., and A. JANOFF. In vitro suppression of serum elastase-inhibitory capacity by reactive oxygen species generated by phagocytosing polymorphonuclear leukocytes. J. Clin. Invest., 63:793–717. 17. ZIMMERMAN, J. J. 1995. Bronchoalveolar inflammatory pathophysiology of bronchopulmonary dysplasia. Clin. Perinatol., 22:429–456. 18. CORRAL, L. G., G. W. MULLER, A. L. MOREIRA, Y. CHEN, M. WU, D. STIRLING, and G. KAPLAN. 1996. Selection of novel analogs of thalidomide with enhanced tumor necrosis factor alpha inhibitory activity. Mol. Med., 2:506–515. 19. MOREIRA, A. L., E. P. SAMPAIO, A. ZMUIDZINAS, P. FRINDT, K. A. SMITH, and G. KAPLAN. 1993. Thalidomide exerts its inhibitory action on tumor necrosis factor alpha by enhancing mRNA degradation. J. Exp. Med., 177:1675–1680. 20. SAMPAIO, E. P., E. N. SARNO, R. GALILLY, Z. A. COHN, and G. KAPLAN. 1991. Thalidomide selectively inhibits tumor necrosis factor alpha production by stimulated human monocytes. J. Exp. Med., 173:699–703.