This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rabinovici, R. Articles by Neville, L.F. Search for Related Content PubMed PubMed Citation Articles by Rabinovici, R. Articles by Neville, L.F.

(Circulation Research. 1996;78:329-336.)
© 1996 American Heart Association, Inc.

Articles

Locally Produced Tumor Necrosis Factor- Mediates Interleukin-2Induced Lung Injury

R. Rabinovici, G. Feuerstein, F. Abdullah, M. Whiteford, P. Borboroglu, E. Sheikh, D.-R. Phillip, P. Ovadia, L. Bobroski, O. Bagasra, L.F. Neville

From the Departments of Surgery (R.R., F.A., M.W., P.B., E.S., D.-R.P., P.O., L.F.N.) and Medicine (L.B., O.B.), Jefferson Medical College, Philadelphia, Pa, and SmithKline Beecham (G.F.), King of Prussia, Pa.

Correspondence to Reuven Rabinovici, MD, Department of Surgery, Jefferson Medical College, 1025 Walnut St, Philadelphia, PA 19107-5083.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Abstract Interleukin (IL)-2–induced microvascular lung injury is an experimental paradigm commonly used to investigate the pathogenesis of the adult respiratory distress syndrome. Since tumor necrosis factor- (TNF-) is known to induce such an injury in vivo and since TNF- is involved in other models of lung injury, we postulated that it might also mediate pulmonary toxicity after IL-2 administration. The present study tested this hypothesis by evaluating the effect of TNF- inhibition on IL-2–induced lung injury in the rat. Recombinant human IL-2 (10⁶ U IV per rat, n=6) elevated lung water, myeloperoxidase activity, and protein accumulation in bronchoalveolar lavage fluid and induced tissue hypoxia. Also, IL-2 enhanced lung tissue TNF- mRNA and peptide (1543±496 pg/g lung wet weight) localized to alveolar macrophages by in situ hybridization. In marked contrast, IL-2 failed to affect serum TNF-, which remained at undetectable levels. Pretreatment with anti–TNF- monoclonal antibody (25 mg/kg IV, n=7) or the TNF- synthesis inhibitor rolipram (200 µg/kg IV, n=7) attenuated lung injury and reverted tissue hypoxia. Furthermore, TNF- inhibition prevented the upregulation of lung tissue IL-1ß, IL-6, cytokine-induced neutrophil chemoattractant, and E-selectin (ELAM-1) but not intercellular adhesion molecule-1 mRNAs in response to IL-2. These data imply that locally produced TNF- mediates IL-2–induced lung inflammation and tissue injury and point to the potential utilization of TNF- inhibitors in treating the pulmonary toxicity of IL-2 immunotherapy.

Key Words: cytokines • adult respiratory distress syndrome • immunotherapy

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Recombinant human IL-2 is currently under investigation as a new treatment modality for patients with advanced metastatic cancer.^{1 2 3} The use of IL-2 immunotherapy in these patients is severely hampered by the development of microvascular lung injury, which parallels ARDS.^{1 3} This clinical observation led to the use of IL-2 infusion as an experimental animal model aimed to elucidate the pathogenic mechanisms of ARDS.

The pathophysiology of IL-2–induced clinical and experimental ARDS is still unclear. Nevertheless, the reduced toxicity of IL-2 in immunocompromised mice,⁴ together with the lack of any demonstrable direct effects of IL-2 on endothelial cells,⁵ indicates that the toxicity of IL-2 is mediated via activation of effector systems. Indeed, reports suggest that IL-2 induces lung edema through lymphocyte activation,^{6 7} production of inflammatory mediators such as IL-1,⁸ thromboxane B₂,⁹ and PAF,¹⁰ and the complement system.^{11 12}

Another putative mechanism of IL-2–induced microvascular lung injury could be mediated through the production of TNF-.^{13 14} Several observations support this hypothesis. First, TNF- has been shown to be produced by human peripheral blood mononuclear cells in response to IL-2 stimulation in vitro.¹⁵ Second, peripheral blood lymphocytes obtained from patients undergoing IL-2 therapy produced significantly larger amounts of TNF- compared with lymphocytes obtained before treatment.¹⁶ Third, TNF- has been shown to directly increase pulmonary endothelial permeability in vivo¹⁷ and to interfere with the alignment of endothelial monolayers in vitro.¹⁸ Fourth, TNF- inhibition exerted protective effects in other models of microvascular lung injury.¹⁹ Fifth, an increasing body of evidence indicates that TNF- may promote tissue injury by activating neutrophils to produce oxygen radicals,²⁰ to express adhesion molecules,²¹ to promote neutrophil adherence,²² and to stimulate adhesion molecule production on endothelial cells (for review, see Reference 23).

Therefore, the present study was designed to test the hypothesis that TNF- mediates lung injury produced by IL-2. To that end, the effect of rolipram, a TNF- synthesis inhibitor in vitro²⁴ and in vivo,²⁵ or specific anti–TNF- mAb on the development of IL-2–induced lung injury was evaluated in the rat.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Drugs
IL-2
Recombinant human IL-2, kindly provided by Hoffman LaRoche Inc, was reconstituted just before use with 1 mL of sterile 0.9% NaCl per 10⁶ U of IL-2.

Anti–TNF- mAb
Hamster anti-murine TNF- mAb (Genzyme) was used in these studies. For the 25 mg/kg dose, a stock solution of 4.35 mg/mL of hamster anti-murine TNF- mAb (Genzyme) in PBS supplemented with 1% BSA was used.

Hamster IgG
Hamster IgG was purchased from Rockland Inc and diluted with 1% BSA to give a concentration similar to the anti–TNF- mAb concentration. This IgG was used in control studies to match the IgG dose in the anti–TNF- mAb group.

Rolipram
Rolipram (SmithKline Beecham) was dissolved in a solution containing 26% dimethyl sulfoxide and 74% sterile distilled water to give a stock concentration of 1.5 mg/mL.

Plasmid cDNAs
Rat TNF-, IL-1ß, IL-6, and ELAM-1 and mouse KC and ß-actin cDNA clones were generously provided by Dr P. Young (SmithKline Beecham, King of Prussia, Pa) and Dr T. Collins (Brigham and Women's Hospital, Boston, Mass).

Animal Preparation
Male Sprague-Dawley rats (278 to 330 g, Taconic Farms, Germantown, NY) were used. After anesthesia with pentobarbital (30 mg/kg IP), catheters (PE-50) were introduced into the femoral vein and artery for drug infusion and blood sampling, respectively. The rats were put in a stereotaxic frame and immobilized. A 1-cm-long left paramedian abdominal incision was made, and the rectus abdominis muscle was exposed. A miniaturized thin-film oxygen sensor was placed in the muscle tissue and fixed by the arm of the stereotaxic apparatus.

Experimental and Control Groups
All experimental and control groups consisted of 5 to 8 animals. After a 30-minute recovery period, a blood sample (0.1 mL, exchanged with an equivolume of 0.9% NaCl) for TNF- assay was collected, and basal muscle oxygen tension was recorded. Anti–TNF- mAb (2.5 or 25 mg/kg IV) or nonspecific hamster IgG (at equivalent doses) was administered. Fifteen minutes later, IL-2 (10⁶ U per rat) or vehicle was infused intravenously over 1 hour. Blood samples for TNF- assay were taken at hourly intervals from the initiation of IL-2 infusion to the termination of the experiment 5 hours later. Thereafter, lungs were harvested and used to determine MPO activity as well as lung wet and dry weights. Additional control groups were established by repeating the above protocol but with the administration of hamster IgG or vehicle, followed by IL-2 or vehicle. Also, to further confirm the role of TNF- in our model, the TNF- synthesis inhibitor rolipram (100 or 200 µg/kg IV) or vehicle was administered, followed by IL-2 (10⁶ U per rat) infusion over 1 hour, and the previous protocols were repeated. A similar protocol including all control groups was repeated solely for BAL to determine protein concentration in BAL fluid. Control values of lung weights, MPO activity, and protein concentration in BAL fluid were determined using lungs harvested from sham rats.

Preliminary data derived from the above protocols indicated that TNF- inhibition attenuated IL-2–induced lung injury in the absence of elevated serum TNF-, which remained at basal levels (<25 pg/mL). These data raised the possibility that locally produced TNF- within the lung may be implicated in the pulmonary inflammation and tissue injury induced by IL-2. To test this postulate, IL-2 or vehicle (n=4) was administered as described above, and lungs were harvested at 4 hours and processed for TNF-, IL-1ß, IL-6, ELAM-1, and ICAM-1 mRNA as well as for TNF- peptide levels. Also, since lung injury in the IL-2 paradigm is neutrophil-mediated⁹ and since previous studies have shown the importance of chemokines in the development of lung injury,^{26 27} lungs were assayed for CINC mRNA levels using a KC cDNA probe. KC is the mouse homologue of the human chemokine gro²⁸ and exhibits 92% sequence homology to CINC.²⁹ To further investigate the role of TNF- in IL-2–induced lung injury, the latter protocol was repeated with anti–TNF- mAb (25 mg/kg IV) or rolipram (200 µg/kg IV) given 15 minutes before IL-2 infusion.

Finally, a time course of TNF- mRNA was established in a separate group of rats given IL-2 at 10⁶ U per animal. In these animals, lungs were harvested at 0.5, 1, 2, and 4 hours after IL-2 infusion (n=2) and assayed for TNF- mRNA signal. Also, lungs harvested at 4 hours were assayed for cellular localization of TNF- using in situ hybridization.

Assays
Tissue Oxygen Tension
Miniaturized thin-film oxygen sensor (Otto Sensors Co) was used as previously described.³⁰

Lung Injury
Previously described methods were used to perform BAL^{10 11} and to assess lung edema^{10 11} and MPO activity.^{10 11}

Serum TNF- ELISA
Serum levels of TNF- were measured using a "sandwich" ELISA as described previously.^{10 11}

Lung TNF- ELISA
Lung aliquots (0.3 g) were homogenized in 2 mL ice-cold suspension buffer (0.1 mol/L NaCl, 0.01 mol/L Tris-HCl [pH 7.6], 0.001 mol/L EDTA [pH 8], 2 µg/mL aprotinin, and 100 µg/mL phenylmethylsulfonyl fluoride) using a tissue tearer (Biospec Products). The resultant homogenate was centrifuged at 10 000g at 4°C, and the clear supernatant was transferred to a sterile Eppendorf tube. Aliquots (100 µL) were added to Maxi-Sorp Immulon plates (Nunc Inc) precoated with anti–TNF- mAb. TNF- levels were thereafter measured in a manner identical to that described for serum TNF-.^{10 11} To account for loss of lung tissue TNF- during preparation, a "correction-curve" assay was performed. In brief, 0.3 g aliquots of naive lungs were "spiked" with recombinant mouse TNF- at 35, 140, 560, 840, and 1120 pg/mL. Lung homogenates were prepared and centrifuged, and the supernatants were taken for TNF- ELISA. The respective values for absorbance at the above TNF- doses were 51%, 57.8%, 62.5%, and 63.6% of the readings for the same TNF- concentrations in the standard curve. These data allowed for the use of a correction factor to more accurately calculate tissue TNF- levels.

Northern Blots
Lungs harvested from IL-2 or 0.9% NaCl (vehicle)–challenged rats were cut into 0.3 g aliquots and taken for RNA isolation as described previously.³¹ Thirty micrograms of capillary-blotted total RNA was hybridized to random-prime prepared rat [³²P]TNF-, IL-1ß, IL-6, ICAM-1, and ELAM-1 or mouse KC and ß-actin cDNA probes overnight and washed at high stringency.

In Situ Hybridization
Lungs harvested from IL-2–challenged or vehicle-challenged rats (n=3) were washed with PBS and fixed in 1% paraformaldehyde overnight. Paraffin-embedded sections (3 µm) were cut, deparaffinized by heating at 80°C, and then dipped sequentially in xylene, 100% ethanol, 70% ethanol, 50% ethanol, and diethyl pyrocarbonate (DEPC)–treated water, and endogenous peroxidase was inactivated by incubating slides in 0.3% H₂O₂ in PBS overnight. The sections were treated with proteinase K (20 µg/mL) at room temperature for 10 minutes, heated at 95°C for 2 minutes, and hybridized to a random prime–generated biotin-tagged rat TNF- cDNA probe (Bioprime labeling kit, GIBCO BRL). Sections were washed in 2x SSC and incubated with 33 µg/mL streptavidin peroxidase in PBS (Sigma Chemical Co), followed by detection with 0.3% H₂O₂–containing 3-amino-9-ethylcarbazole dissolved in 50 mmol/L acetate buffer. To confirm the cellular source of lung TNF- mRNA–positive cells, the same cryosections of lung tissue were immunostained before the in situ hybridization using macrophage/granulocyte-specific anti-CD11b/c mAb conjugated with FITC (Pharmingen) and with anti–von Willebrand factor (factor VIII, Sigma), a microvascular endothelial cell–specific polyclonal antibody. von Willebrand Ag/Ab complexes were detected after the in situ hybridization using FITC-tagged polyclonal goat anti-rabbit secondary antibodies (Sigma). Nonspecific hybridization was evaluated by using an unrelated probe, HIV-I gag (SK19) conjugated with biotin.

Data Analysis
Data in text and figures are mean±SEM for the indicated number of animals. One-way ANOVA followed by the Student-Newman-Keuls test was used for statistical analysis. A value of P<.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Effect of IL-2 on Lung Water
Lung water content in sham rats was 401±45 mg. IL-2 infusion elevated lung water content by 40±7% (P<.05, Fig 1A). Pretreatment with anti–TNF- mAb (Fig 1A) at 25 mg/kg (but not vehicle or IgG) and rolipram (but not vehicle) attenuated this response by 40±6% (P<.05, Fig 1A) and 52±4% (P<.05; data not shown), respectively. The lower doses of anti–TNF- and rolipram did not affect the IL-2–induced pulmonary edema (data not shown). Also, when injected alone, anti–TNF- mAb (Fig 1A), IgG (Fig 1A), and rolipram (data not shown) did not alter basal lung water content.

View larger version (16K): [in this window] [in a new window]   Figure 1. Effect of anti–TNF- mAb on IL-2–induced pulmonary edema (A), lung MPO activity (B), or BAL fluid protein concentration (C). Values are percent change from the anti–TNF- mAb vehicle+IL-2 vehicle group. *P<.05 vs all groups except for the IgG+IL-2 group; **P<.05 vs all groups except for anti–TNF- mAb vehicle+IL-2 group.

Effect of IL-2 on Lung MPO Activity
Lung MPO activity in sham rats was 9.5±1.2 U/mg wet lung weight. Administration of IL-2 increased pulmonary MPO activity by 61±5% (P<.05, Fig 1B). This response was eliminated after pretreatment with anti–TNF- mAb (but not vehicle) at 25 mg/kg (12±6%, P<.05, Fig 1B) or rolipram (but not vehicle) at 200 µg/kg (15±5%, P<.05; data not shown). The lower doses of anti–TNF- mAb or rolipram had no effect on MPO response to IL-2 (data not shown). The injection of anti–TNF- mAb (Fig 1B), IgG (Fig 1B), and rolipram (data not shown) alone did not affect basal lung MPO activity.

Effect of IL-2 on BAL Fluid Protein
BAL fluid protein concentration was 322±29% higher in IL-2–treated rats compared with a concentration of 0.25±0.52 mg/mL in the negative control group (anti–TNF- mAb vehicle plus IL-2 vehicle) (Fig 1C). Anti–TNF- mAb (25 mg/kg) or rolipram (200 µg/kg) attenuated this response by 66±8% (P<.05, Fig 1C) and 70±9 (P<.05; data not shown), respectively. The lower doses of anti–TNF- mAb and rolipram had no effect on BAL fluid protein response to IL-2 (data not shown). No significant changes in BAL fluid protein concentration were noted after the infusion of IL-2 vehicle (Fig 1C), anti–TNF- mAb (Fig 1C), IgG alone (Fig 1C), or rolipram (data not shown) alone.

Effect of IL-2 on Tissue PO₂
IL-2 infusion reduced basal tissue PO₂ (93±10 mm Hg) to 59±10 mm Hg (Fig 2). Pretreatment with anti–TNF- mAb at 25 mg/kg attenuated the IL-2–induced hypoxia (PO₂, 82±8 mm Hg; Fig 2). A similar beneficial effect was observed after pretreatment with rolipram at 200 µg/kg (PO₂, 84±8 mm Hg; data not shown), whereas anti–TNF- mAb at 2.5 mg/kg or rolipram at 100 µg/kg did not exert any protective effect (data not shown). Tissue PO₂ remained unchanged in all other control groups.

View larger version (16K): [in this window] [in a new window]   Figure 2. Effect of anti–TNF- mAb on skeletal muscle oxygen tension response to IL-2. Values are amount of change from the anti–TNF- mAb vehicle+IL-2 vehicle group. *P<.05 vs all groups except for IgG+IL-2 group; **P<.05 vs all groups except for anti–TNF- mAb vehicle+IL-2 group.

Effect of IL-2 on Serum TNF-
The basal serum TNF- level was below the sensitivity of the ELISA (25 pg/mL). Administration of IL-2 failed to elevate serum TNF- beyond the minimal detection level of the ELISA (data not shown). Also, no elevation of serum TNF- was detected in any of the other experimental or control groups (data not shown).

Effect of IL-2 on Lung Tissue TNF- mRNA and Peptide
Lung TNF- mRNA (Fig 3A) and peptide (Table) from IL-2–treated rats were significantly elevated compared with IL-2 vehicle–treated control values. In the latter group of animals, lung tissue TNF- levels were below the minimal detection level of the ELISA assay (<25 pg/mL, Table). Pretreatment with rolipram (Fig 3B) or anti–TNF- mAb (Fig 4) attenuated the elevation of lung TNF- mRNA and peptide in response to IL-2.

View larger version (29K): [in this window] [in a new window]   Figure 3. IL-2–induced lung TNF- mRNA production: effect of rolipram and time course. A, Total RNA was isolated from lungs of separate rats challenged with either 106U IL-2 (lanes 1 to 4) or IL-2 vehicle (lanes 5 to 8) and hybridized to a [32P]TNF- cDNA probe as described in "Materials and Methods." Note the pronounced increase in TNF- mRNA signal from IL-2–treated rats (lanes 1 to 4) as opposed to control rats (lanes 5 to 8). A positive control of 5 µg total RNA extracted from lipopolysaccharide (1 µg/mL)–stimulated AMJ2C11 cells, a mouse alveolar macrophage cell line,32 was used in lane 9. B, Pretreatment with rolipram at 200 µg/kg (n=4, lanes 5 to 8) markedly suppressed IL-2–induced (lanes 1 and 3) lung TNF- mRNA production. Rolipram alone (lanes 2 and 4) did not upregulate lung TNF- mRNA. C, Time course study of lung TNF- mRNA production shows a rapid appearance with signal observed 1 hour after IL-2 infusion.

View this table: [in this window] [in a new window]   Table 1. Lung TNF- Response to IL-2: Effect of Pretreatment With Anti–TNF- mAb or Rolipram

View larger version (35K): [in this window] [in a new window]   Figure 4. Effect of anti–TNF- mAb on IL-2–induced lung TNF-, IL-1ß, IL-6, CINC, ELAM-1, and ICAM-1 mRNA. Total IL-2–induced lung RNA from rats pretreated with anti–TNF- mAb (25 mg/kg, n=3) or nonspecific IgG (25 mg/kg, lane 4) was hybridized to rat TNF-, IL-1ß, IL-6, ELAM-1, and ICAM-1 and mouse KC [32P]cDNA probes and reprobed with a ß-actin. Note the marked induction, after IL-2 administration (lane 4), of all mediators whose expression was suppressed by pretreatment with anti–TNF- mAb (lanes 1 to 3), except for ICAM-1.

Lung TNF- mRNA was upregulated as early as 60 minutes after IL-2 infusion (Fig 3C). This precedes lung injury, which does not appear until 4 hours after the administration of IL-2.¹¹

Effect of IL-2 on Lung IL-1ß, IL-6, KC, ELAM-1, and ICAM-1 mRNAs
IL-2, but not vehicle, produced a robust induction of lung IL-1ß, IL-6, and CINC (Fig 5) and ELAM-1 and ICAM-1 (Fig 6) mRNAs. Pretreatment with rolipram (Figs 5 and 6) or anti–TNF- mAb (Fig 4) totally prevented IL-2–enhanced upregulation of all inflammatory mediators, with the exception of ICAM-1.

View larger version (20K): [in this window] [in a new window]   Figure 5. Effect of rolipram (Rol) on IL-2–induced upregulation of lung IL-1ß, IL-6, and CINC mRNA. Total RNA was isolated from lungs of rats challenged with Rol vehicle+IL-2 (n=2), Rol+IL-2 vehicle (n=2), or Rol+IL-2 (n=4) hybridized to rat [32P]IL-1ß, IL-6, or mouse KC cDNA probes and reprobed with ß-actin. Note the marked induction, from only IL-2–challenged rats, of IL-1ß, IL-6, and CINC mRNAs, which were suppressed by Rol pretreatment.

View larger version (30K): [in this window] [in a new window]   Figure 6. Effect of rolipram on IL-2–induced upregulation of lung ELAM-1 and ICAM-1 mRNA. Total RNA was isolated from lungs of rats challenged with rolipram vehicle+IL-2 (n=2), rolipram+IL-2 vehicle (n=2), or rolipram+IL-2 (lanes 1 to 4, n=4) hybridized to rat [32P]ELAM-1 and ICAM-1 cDNA probes and reprobed with ß-actin. Note the marked induction of ELAM-1 and ICAM-1 mRNAs from IL-2 (but not vehicle)–challenged rats. Pretreatment with rolipram totally suppressed ELAM-1 but not ICAM-1 upregulation in response to IL-2.

Cellular Localization of IL-2–Induced Lung TNF-
IL-2 (Fig 7a), but not vehicle (Fig 7b), induced pulmonary TNF- mRNA production in alveolar macrophages as detected by in situ hybridization of paraffin-embedded tissue. These macrophages (Fig 7d) also fluoresced with anti-CD11b/c antibodies (Fig 7c), confirming cell type. It should be noted that although anti-CD11b/c antibodies can also bind to granulocytes, morphologically, the cells exhibiting dual positivity in these experiments were macrophages/monocytes. In contrast to the macrophages, microvascular endothelial cells identified by their positivity toward anti–von Willebrand factor antibodies (Fig 7e) did not exhibit a TNF- signal (Fig 7f). No nonspecific hybridization was observed when biotinylated HIV-1 gag (SK19) probe was used (data not shown).

View larger version (0K): [in this window] [in a new window]   Figure 7. Cellular localization of pulmonary TNF- mRNA by in situ hybridization. a and b, Paraffin sections of lungs harvested from IL-2 (a) or vehicle-treated (b) rats. Numerous signal-positive macrophages were observed only in animals infused with IL-2. A representative signal-positive macrophage from IL-2–stimulated tissue (a, arrow) and a representative signal-negative macrophage from IL-2 vehicle–treated rats (b, arrow) can be seen. Original magnification x40. c and d, Cryosections of lungs treated with IL-2, labeled with FITC-conjugated anti-CD11b/c antibodies (c), and taken for in situ hybridization using biotinylated TNF- cDNA probe (d). Macrophages staining for anti-CD11b/c also exhibited a positive TNF- signal, confirming cell type. A representative macrophage is seen in panels c and d. Original magnification x100. e and f, Cryosections from IL-2–challenged rats dually labeled with a polyclonal anti–von Willebrand factor antibody, followed by FITC detection (e) and a biotinylated TNF- cDNA probe (f). Microvascular endothelium that stained positively for von Willebrand factor (e, arrows) showed no detectable signal for TNF- (f, arrows). Original magnification x40. Results described in this figure were uniformly present in all lungs tested.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The data in the present study confirm previous observations from our laboratory^{10 11} and others^{9 17} that the systemic administration of IL-2 causes lung injury as early as 4 hours after infusion. Pulmonary injury is characterized by increased permeability (evidenced by edema and elevated BAL fluid protein concentration), leukosequestration (evidenced by enhanced lung MPO activity), and peripheral tissue hypoxia.

One key finding of the present study is that TNF- inhibition by rolipram, a repressor of TNF- mRNA synthesis,^{24 25} or anti–TNF- mAb attenuated lung injury produced by the systemic administration of IL-2. Since IL-2 infusion enhanced lung tissue TNF- mRNA and peptide levels in the absence of detectable systemic TNF-, it is highly plausible that locally produced TNF- mediates the pulmonary toxicity of IL-2. Indeed, in situ hybridization localized TNF- mRNA production to alveolar macrophages, suggesting that these cells might be a primary source of TNF- in IL-2–challenged rats. This possibility is strongly supported by the wide documentation that cultured alveolar macrophages produce large amounts of TNF- in response to a variety of stimuli including IL-2.^{33 34 35}

The elevation of lung tissue but not serum TNF- sharpens the debate as to the significance of circulating cytokine levels in shock states. It is generally accepted that elevated serum cytokine levels are indicative of their roles as inflammatory mediators in stress situations such as sepsis and hemorrhage. Therefore, a variety of anti-cytokine therapies have been developed to combat these clinical syndromes. The failure of cytokine inhibition to improve the outcome of human sepsis,³⁶ however, cast doubt on the relevance of serum cytokines in disease. In contrast to circulating cytokines, the contribution of tissue-associated cytokines in inflammatory conditions has been clearly overlooked. This is surprising, since circulating cytokines represent "the tip of the iceberg," which reflects a delicate balance between cytokine production and clearance, levels of endogenous inhibitors, and receptor occupancy at target cells.³⁷ Therefore, the failure to detect serum cytokines may not negate their involvement in an inflammatory condition. Indeed, the protective effect of TNF- inhibition on IL-2–induced lung injury was associated with elevated lung but not serum TNF-. In that respect, it should be noted that failure to measure TNF- in the serum does not imply its absence but may rather reflect the limitation of the assay used. Nevertheless, basal serum TNF- levels in animals^{10 11} and humans³⁸ administered with IL-2 were reported to be below the sensitivity of the various bioassays and ELISAs commonly used.

An important finding presented herein is that systemically infused IL-2 elevated TNF- production in lung tissue as early as 1 hour after the termination of IL-2 infusion. The upregulation of pulmonary TNF- mRNA before the presentation of lung injury at 4 hours¹¹ clearly suggests a causative link between IL-2–induced TNF- production and lung injury. The exact mechanism by which IL-2 induces lung TNF- production is still obscure. Nevertheless, the presence of TNF-–positive signals in alveolar macrophages, in agreement with previous studies,³⁹ along with the inhibition of IL-2–induced upregulation of TNF- mRNA by anti–TNF- mAb or rolipram would indicate that IL-2 acts directly on macrophages to trigger TNF- mRNA transcription.³⁵

To our best knowledge, no previous studies with IL-2 have evaluated pulmonary levels of TNF-. Nevertheless, elevated levels of this cytokine were detected in mesenteric lymph nodes, but not in the serum, of mice that were administered IL-2.⁴⁰ The observation that IL-2 did not increase serum TNF- levels is in accord with numerous previous studies in various species, including humans.³⁸ Interestingly, other investigators reported elevated serum TNF- levels in humans undergoing IL-2 immunotherapy.¹⁶ Minimal endotoxin contamination and the altered immunological reactivity of patients with advanced cancer could account for these discrepancies.

The implication of TNF- in the pathogenesis of IL-2–induced pulmonary toxicity is in agreement with previous studies in which IL-2 has been shown to stimulate TNF- production in a variety of in vitro systems. For example, culture supernatants of IL-2–activated human monocytes,¹⁵ neutrophils,⁴¹ lymphocytes,⁴² and alveolar macrophages^{33 34 42} displayed elevated TNF- concentrations. Also, lymphocyte-activated killer cells stimulated with IL-2 were reported to produce TNF-.⁴³ Furthermore, in vivo studies have shown that passive immunization against TNF-^{44 45} or treatment with soluble TNF- receptor⁴⁶ can attenuate IL-2–induced permeability defects in the rat.

Another key finding in the present study is that the pulmonary inflammatory response to IL-2 involves multiple cytokines (eg, IL-1ß and IL-6) apart from TNF-, adhesion molecules (ELAM-1 and ICAM-1), and the chemokine CINC. It seems that TNF- is central in the production of all these mediators except for ICAM-1, since rolipram or anti–TNF- mAb prevented their upregulation in response to IL-2. The failure of TNF- inhibition to modify the IL-2–induced expression of ICAM-1 would indicate that IL-2 induces this adhesion molecule directly or via generation of yet an unknown mediator(s). In support of the former hypothesis, fluorescence photobleaching studies have indicated a direct physical interaction between the IL-2 receptor and ICAM-1 on a T-lymphocytic cell line.⁴⁷ The differential effect of TNF- inhibition on the expression of ICAM-1 and ELAM-1 is in agreement with recent in vitro studies in which phosphodiesterase inhibitors attenuated ELAM-1, but not ICAM-1, upregulation from TNF-–stimulated human umbilical vein endothelial cells.^{48 49}

The inhibitory effect of anti–TNF- mAb on lung TNF- mRNA production in response to IL-2 is further evidence supporting the purported existence of a positive autocrine-feedback mechanism by which TNF- promotes its own production.⁵⁰ This finding is in agreement with previous data from our laboratory in which anti–TNF- mAb reduced serum TNF- in a rat model of ARDS induced by a combination of otherwise noninjurious doses of lipopolysaccharide and PAF.⁵¹

It should be noted that anti–TNF- mAb did not provide complete protection against IL-2–induced lung injury. The residual injury might have been mediated by other proinflammatory agents produced in response to IL-2 stimulation. For example, PAF, a very potent phospholipid mediator of inflammation,⁵² has been reported to modulate the development of IL-2–induced microvascular injury in the rat.¹⁰ Also, complement factors have been implicated in the pathogenesis of IL-2 pulmonary toxicity in laboratory animals¹¹ and humans.¹² Additionally, IL-2 might facilitate the adhesion of natural killer cells to vascular endothelium, which will promote extravasation of plasma and tissue edema.⁵³

In conclusion, the present study indicates that lung-derived TNF- is central in the pathogenesis of IL-2–induced pulmonary inflammatory response and injury. Therefore, organ-specific therapy, ie, local administration (intratracheal) of TNF- inhibitors may represent a novel and feasible approach to prevent the pulmonary toxicity of IL-2.

	Selected Abbreviations and Acronyms

 

ARDS = adult respiratory distress syndrome BAL = bronchoalveolar lavage CINC = cytokine-induced neutrophil chemoattractant ELAM-1 = endothelial-leukocyte adhesion molecule-1 ELISA = enzyme-linked immunosorbent assay FITC = fluorescein isothiocyanate ICAM-1 = intercellular adhesion molecule-1 IL = interleukin KC = murine homologue of human chemokine gro mAb = monoclonal antibody MPO = myeloperoxidase PAF = platelet-activating factor TNF- = tumor necrosis factor-

	Acknowledgments

 
The authors would like to thank Albert Kovatich (Department of Immunopathology, Jefferson Medical College) for fruitful discussions regarding the in situ hybridization data. This work was supported in part by an American Heart Association grant.

Received June 20, 1995; accepted November 1, 1995.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 
1. Rosenberg SA, Lotze MT, Muul LM, Leitman S, Chang AE, Ettinghausen JE, Motory YL, Skibber JM, Shiloni E, Vetto TJ, Scipp CA, Simpson C, Richer CM. Observations on the systemic administration of autologous lymphokine-activated killer cells and recombinant interleukin-2 to patients with metastatic cancer. N Engl J Med. 1985;313:1485-1492. [Abstract]

2. Rosenberg SA, Lotze MT, Mule JJ. New approaches to the immunotherapy of cancer using interleukin-2. Ann Intern Med. 1988;108:853-858.

3. Rosenberg SA, Yang JC, Topalian SL, Schwartzentrubber DJ, Weber JS, Parkinson DR, Seipp CA, Einhorn JH, White DE. Treatment of 283 consecutive patients with metastatic melanoma or renal cell cancer using high dose IL-2. JAMA. 1994;271:907-913. [Abstract/Free Full Text]

4. Rosenstein M, Ettinghausen SE, Rosenberg SA. Extravasation of intravascular fluid mediated by the systemic administration of recombinant interleukin 2. J Immunol. 1986;137:1735-1742. [Abstract]

5. Cotran RS, Pober JS, Gimbrone MA Jr, Springer TA, Wiebke EA, Gaspari AA, Rosenberg SA, Lotze MT. Endothelial activation during interleukin 2 immunotherapy: a possible mechanism for the vascular leak syndrome. J Immunol. 1987;139:1883-1888.

6. Anderson TD, Hayes TJ, Gately MK, Bontempo JM, Stern LL, Truitt GA. Toxicity of human recombinant interleukin-2 in the mouse is mediated by activated lymphocytes: separation of efficacy and toxicity by selective lymphocyte subset depletion. Lab Invest. 1988;59:598-612. [Medline] [Order article via Infotrieve]

7. Damle NK, Doyle LV, Bender JR, Bradley EC. Interleukin-2-activated human lymphocytes exhibit enhanced adhesion to normal vascular endothelium cells and cause their lysis. J Immunol. 1987;138:1779-1785. [Abstract]

8. Numeroff RP, Aronson FR, Mier JW. IL-2 stimulates the production of IL-1 alpha and IL-1 beta by human peripheral blood mononuclear cells. J Immunol. 1988;141:4250-4257. [Abstract]

9. Welbourn R, Goldman G, Kobitz L, Valeri CR, Shepro D, Hechtman HB. Involvement of thromboxane and neutrophils in multiple-system organ edema with interleukin-2. Ann Surg. 1990;212:728-733. [Medline] [Order article via Infotrieve]

10. Rabinovici R, Sofronski MD, Renz JF, Hillegas LM, Esser KM, Vernick J, Feuerstein G. Platelet activating factor mediates interleukin-2 induced lung injury in the rat. J Clin Invest. 1992;89:1669-1673.

11. Rabinovici R, Sofronski MD, Borboroglu P, Spirig AM, Hillegas LM, Levine J, Vernick J, Scesney SM, Feuerstein N, Feuerstein G. Interleukin-2–induced lung injury: the role of complement. Circ Res. 1994;74:329-335. [Abstract/Free Full Text]

12. Thijs LG, Hack CE, Strack van Schijndel RJ, Nuijens JH, Wolbink GJ, Eerenberg-Belmer AJ, Van der Vall H, Wagstaff J. Activation of the complement system during immunotherapy with recombinant IL-2. J Immunol. 1990;144:2419-2424. [Abstract]

13. Stephens KE, Ishizaka A, Larrick JW, Raffin TA. Tumor necrosis factor causes increased pulmonary permeability and edema: comparison to septic acute lung injury. Am Rev Respir Dis. 1988;137:1364-1370. [Medline] [Order article via Infotrieve]

14. Hocking DC, Phillips PG, Ferro TJ, Johnson A. Mechanisms of pulmonary edema induced by tumor necrosis factor-. Circ Res. 1990;67:68-77. [Abstract/Free Full Text]

15. Owen-Schaub LB, de Mars M, Murphy EC Jr, Grimm EA. IL-2 dose regulates TNF alpha mRNA transcription and protein secretion in human peripheral blood lymphocytes. Cell Immunol. 1991;132:193-200. [Medline] [Order article via Infotrieve]

16. Heslop HE, Gottlieb DJ, Bianchi ACM, Meager A, Prentice HG, Mehta AB, Hoffbrand AV, Brenner MK. In-vivo induction of gamma interferon and tumor necrosis factor by interleukin-2 infusion following intensive chemotherapy or autologous marrow transplantation. Blood. 1989;74:1374-1380. [Abstract/Free Full Text]

17. Klausner JM, Morel N, Paterson IS, Kobitz L, Valeri CR, Eberlein TJ, Shepro D, Hechtman HB. The rapid induction by interleukin-2 of pulmonary microvascular permeability. Ann Surg. 1989;209:119-128. [Medline] [Order article via Infotrieve]

18. Stolpen A, Guinan E, Fiers W, Pober J. Recombinant tumor necrosis factor and immune interferon act singly and in combination to recognize human vascular endothelial cell monolayers. Am J Pathol. 1986;123:16-24. [Abstract]

19. Rabinovici R, Bugelski PJ, Esser KM, Hillegass LM, Vernick J, Feuerstein GZ. ARDS-like lung injury produced by endotoxin in platelet activating factor primed rats. J Appl Physiol. 1993;74:1791-1802. [Abstract/Free Full Text]

20. Shalaby MR, Aggarwal BB, Rinderknecht E, Svedersky LP, Finkle BS, Palladino MR. Activation of human polymorphonuclear neutrophil functions by interferon gamma and TNF. J Immunol. 1985;135:2069-2073. [Abstract]

21. Jutila MA, Rott L, Berg EL, Butcher EC. Function and regulation of the neutrophil MEL-14 antigen in-vivo: comparison with LFA-1 and Mac-1. J Immunol. 1989;143:3318-3324. [Abstract]

22. Gamble JR, Harlan JM, Klebanoff SJ, Lopez AF, Vadas MA. Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc Natl Acad Sci U S A.. 1985;82:8667-8671. [Abstract/Free Full Text]

23. Pober JS, Cotran RS. Cytokines and endothelial cell biology. Physiol Rev. 1990;70:427-451. [Free Full Text]

24. Semmler J, Wachtel H, Endres S. The specific type IV phosphodiesterase inhibitor rolipram suppresses tumor necrosis factor- production by human mononuclear cells. Int J Immunopharmacol. 1993;15:409-413. [Medline] [Order article via Infotrieve]

25. Turner CR, Esser KM, Wheeldon EB. Therapeutic intervention in a rat model of ARDS, IV: phosphodiesterase IV inhibition. Circ Shock. 1993;39:237-245.[Medline] [Order article via Infotrieve]

26. Mulligan MS, Jones ML, Bolanowski MA, Baganoff MP, Deppeler CL, Meuers DM, Ryan US, Ward P. Inhibition of lung inflammatory reactions in rats by an anti-human IL-8 antibody. J Immunol. 1993;150:5585-5595. [Abstract]

27. Ulich TR, Howard SC, Remick DG, Wittwer A, Yi ES, Yin S, Guo K, Welply JK, Williams JH. Intra-tracheal administration of endotoxin and cytokines, VI: antibodies to CINC inhibit acute inflammation. Am J Physiol. 1995;268:L245-L250. [Abstract/Free Full Text]

28. Zagorski J, DeLorco JE. Rat CINC (cytokine-induced neutrophil chemoattractant) is the homologue of the human GRO proteins but is encoded by a single gene. Biochem Biophys Res Commun. 1993;190:104-110. [Medline] [Order article via Infotrieve]

29. Huang S, Paulauskis JD, Godleski JJ, Kobzik L. Rat CINC/KC cDNA and mRNA expression in lung macrophages and fibroblasts. Biochem Biophys Res Commun. 1992;184:922-929. [Medline] [Order article via Infotrieve]

30. Rabinovici R, Rudolph AS, Vernick J, Feuerstein GZ. A new salutary resuscitative fluid: liposome encapsulated hemoglobin/hypertonic saline solution. J Trauma. 1993;35:121-127. [Medline] [Order article via Infotrieve]

31. Chomczynski P, Sacchi N. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156-159. [Medline] [Order article via Infotrieve]

32. Palleroni AV, Varesio L, Wright RB, Brunda MJ. Tumoricidal alveolar macrophage and tumor infiltrating macrophage cell lines. Int J Cancer. 1991;49:296-302. [Medline] [Order article via Infotrieve]

33. Renz H, Gong J-H, Schmidt A, Nain M, Gemsa D. Release of tumor necrosis factor- from macrophages: enhancement and suppression are dose-dependently regulated by prostaglandin E₂ and cyclic nucleotides. J Immunol. 1988;141:2388-2394. [Abstract]

34. Hancock WW, Muller WA, Cotran RS. Interleukin-2 receptors are expressed by alveolar macrophages during pulmonary sarcoidosis and are inducible by lymphokine treatment of normal human lung macrophages, blood monocytes and monocyte cell lines. J Immunol. 1987;138:185-192. [Abstract]

35. Espinoza-Delgado I, Bosco MC, Musso T, Gusella GL, Longo DL, Varesio L. Interleukin-2 and human monocyte activation. J Leukoc Biol. 1995;57:13-21. [Abstract]

36. Lamy M, Thijs LG, eds. Mediators of Sepsis, Volume 16: Update in Intensive Care and Emergency Medicine. New York, NY: Springer-Verlag; 1992.

37. Cavaillon JM, Munoz C, Fitting C, Misset B, Carlet J. Circulating cytokines: the tip of the iceberg? Circ Shock. 1992;38:145-152. [Medline] [Order article via Infotrieve]

38. Gemlo BT, Palladino MA Jr, Jaffe HS, Espevik TP, Rayner AA. Circulating cytokines in patients with metastatic cancer treated with recombinant IL-2 and LAK cells. Cancer Res. 1988;48:5864-5867. [Abstract/Free Full Text]

39. Dubinnett SM, Huang M, Lichtenstein A, McBride WH, Wang J, Markovitz G, Kelley D, Grody WW, Mintz LE, Dhanani S. Tumor necrosis factor-alpha plays a central role in interleukin-2-induced pulmonary vascular leak and lymphocyte accumulation. Cell Immunol. 1993;157:170-180.

40. Butler LD, Mohler KM, Layman NK, Cain RL, Ridel PE, Puckett LD, Bendele AM. IL-2-induced systemic toxicity: induction of mediators and immunopharmacologic intervention. Immunopharmacol Immunotoxicol. 1989;11:445-487. [Medline] [Order article via Infotrieve]

41. Wei S, Blanchard DK, Liu JH, Djeu JY. Activation of tumor necrosis factor production from human neutrophils by IL-2 via IL-2 receptor beta. J Immunol. 1993;150:1979-1987. [Abstract]

42. Strieter RM, Remick DG, Lynch JP, Spengler RN, Kunkel SL. Interleukin-2-induced TNF gene expression in human alveolar macrophages and blood monocytes. Am Rev Respir Dis. 1989;139:335-342. [Medline] [Order article via Infotrieve]

43. Aramburu J, Balboa MA, Rodriguez A, Melero I, Alonso M, Alonso JL, Lopez-Botet M. Stimulation of IL-2-activated natural killer cells through the Kp-43 surface antigen up-regulates TNF production involving the LFA-1 integrin. J Immunol. 1993;151:3420-3429. [Abstract]

44. Edwards MJ, Abney DL, Heniford T, Miller FN. Passive immunization against TNF inhibits IL-2-induced microvascular alternations and reduces toxicity. Surgery. 1992;112:480-486. [Medline] [Order article via Infotrieve]

45. Fraker DL, Langstein HN, Norton JA. Passive immunization against tumor necrosis factor partially abrogates interleukin-2 toxicity. J Exp Med. 1989;170:1015-1020. [Abstract/Free Full Text]

46. Quinn TD, Miller FN, Wilson MA, Garrison RN, Anderson JA, Lenz LG, Edwards MJ. Interleukin-2-induced lymphocyte infiltration of multiple organs is differentially suppressed by soluble tumor necrosis factor receptor. J Surg Res. 1994;56:117-122. [Medline] [Order article via Infotrieve]

47. Edidin M, Aszalos A, Damjanovich S, Waldmann TA. Lateral diffusion measurements give evidence for association of the Tac peptide of the IL-2 receptor with the T27 peptide in the plasma membrane of HUT-102-B2 T cells. J Immunol. 1988;141:1206-1210. [Abstract]

48. Pober JS, Slowik MR, DeLuca LG, Ritchie AJ. Elevated cAMP inhibits endothelial cell synthesis and expression of TNF-induced ELAM-1 and VCAM-1 but not ICAM-1. J Immunol. 1993;150;5114-5123.

49. Ghersa P, Van Huijsduijnen RH, Whealan J, Cambet Y, Pescini R, DeLamarter JF. Inhibition of E-selectin gene transcription through a cAMP-dependent protein kinase pathway. J Biol Chem. 1994;269:29129-29137. [Abstract/Free Full Text]

50. Sherry B, Cerami A. Cachectin/tumor necrosis factor exerts endocrine, paracrine and autocrine control of inflammatory responses. J Cell Biol. 1988;107:1269-1277. [Free Full Text]

51. Rabinovici R, Bugelski PJ, Esser KM, Hillegass LM, Griswold DE, Vernick J, Feuerstein GZ. Tumor necrosis factor- mediates endotoxin-induced lung injury in platelet activating factor-primed rats. J Pharmacol Exp Ther. 1993;267:1550-1557. [Abstract/Free Full Text]

52. Braquet P, Touqui L, Shen T, Vargaftig BB. Perspectives in platelet activating factor. Pharmacol Rev. 1987;39:97-145. [Medline] [Order article via Infotrieve]

53. Aronson FR, Libby P, Brandon EP, Janicka MW, Mier JW. IL-2 rapidly induces natural killer cell adhesion to human endothelial cells: a potential mechanism for endothelial injury. J Immunol. 1988;141:158-163.[Abstract]

This article has been cited by other articles:

				R. H. Sohn, C. B. Deming, D. C. Johns, H. C. Champion, C. Bian, K. Gardner, and J. J. Rade Regulation of endothelial thrombomodulin expression by inflammatory cytokines is mediated by activation of nuclear factor-kappa B Blood, May 15, 2005; 105(10): 3910 - 3917. [Abstract] [Full Text] [PDF]

				K. Odoms, T. P. Shanley, and H. R. Wong Short-term modulation of interleukin-1{beta} signaling by hyperoxia: uncoupling of I{kappa}B kinase activation and NF-{kappa}B-dependent gene expression Am J Physiol Lung Cell Mol Physiol, March 1, 2004; 286(3): L554 - L562. [Abstract] [Full Text] [PDF]

				T. KOSHIKA, A. ISHIZAKA, I. NAGATOMI, Y. SUDO, N. HASEGAWA, and T. GOTO Pretreatment with FK506 Improves Survival Rate and Gas Exchange in Canine Model of Acute Lung Injury Am. J. Respir. Crit. Care Med., January 1, 2001; 163(1): 79 - 84. [Abstract] [Full Text]

				R. A. Tio, J. Nieken, E. G.E. de Vries, C. Pfeiffer, M. J.L. de Jongste, E. Pieper, H. Moshage, P. C. Limburg, and N. H. Mulder Negative inotropic effects of recombinant interleukin 2 in patients without left ventricular dysfunction Eur J Heart Fail, June 1, 2000; 2(2): 167 - 173. [Abstract] [Full Text] [PDF]

				D. HÄFNER and P.-G. GERMANN Additive Effects of Phosphodiesterase-4 Inhibition on Effects of rSP-C Surfactant Am. J. Respir. Crit. Care Med., May 1, 2000; 161(5): 1495 - 1500. [Abstract] [Full Text]

				R. Rabinovici, K. Kabir, M. Chen, Y. Su, D. Zhang, X. Luo, and J.-H. Yang ADAR1 Is Involved in the Development of Microvascular Lung Injury Circ. Res., May 25, 2001; 88(10): 1066 - 1071. [Abstract] [Full Text] [PDF]

This Article Abstract Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rabinovici, R. Articles by Neville, L.F. Search for Related Content PubMed PubMed Citation Articles by Rabinovici, R. Articles by Neville, L.F.