© Birkhäuser Verlag, Basel, 2004 Inflamm. res. 53 (2004) 004– 12 1023-3830/04/010004-09 DOI 10.1007/s00011-003-1215-3

Inflammation Research

Original Research Papers Upregulation of bronchioloalveolar carcinoma-derived C-X-C chemokines by tumor infiltrating inflammatory cells M. Wislez 1,2, C. Philippe 3, M. Antoine 4, N. Rabbe 2, J. Moreau 5, A. Bellocq 3, C. Mayaud 1, 2, B. Milleron 1, 2, P. Soler 5, J. Cadranel 1,2 1

2 3 4 5

Service de Pneumologie et de Réanimation Respiratoire, AP-HP, Hôpital Tenon, 4 rue de la chine, 75020, Paris, France, Fax: ++33 1 56 01 69 68, e-mail: [email protected] Laboratoire de Biologie Cellulaire et d’Immunopathologie Pulmonaire, UPRES EA 3493, UFR Saint-Antoine, Université Paris VI, Paris, France Service d’Explorations Fonctionnelles Multidisciplinaires, AP-HP, Hôpital Tenon, Paris, France Service d’Anatomie Pathologique, AP-HP, Hôpital Tenon, Paris, France Unité Inserm 408, Hôpital Bichat, Paris, France

Received 20 November 2003; returned for revision 28 May 2003; accepted by G. Letts 19 August 2003

Abstract. Objective and design: The presence of increased numbers of tumor-infiltrating neutrophils is associated with poorer outcome in patients with adenocarcinoma of the bronchioloalveolar (BAC) subtype. We evaluated the role of inflammatory environment on C-X-C chemokine tumor production. Materials: Bronchoalveolar lavage from 31 consecutive patients with adenocarcinoma of the BAC subtype as well as tumor and normal pulmonary tissue samples. A549 BAC cell line. Peripheral blood mononuclear cells (PBMC), polymorphonuclear neutrophils (PMN) and alveolar macrophages (AM). Methods: Elisa measurements and immunohistochemical studies of ENA-78, IL-8, IL-1b and TNF-a. RNA isolation, reverse transcription, and PCR amplification of ENA-78 and IL-8. Results: C-X-C peptides were expressed by tumor cells of all the tumor specimens tested. ENA-78 and IL-8 were also expressed by AM. To better understand the regulation of the C-X-C production, BAC cell line was cultured alone or with inflammatory cells. PBMC upregulated both tumor ENA-78 and IL-8 mRNA expression and protein release whereas AM only upregulated ENA-78 mRNA expression and protein release; PMN had no effect. Anti-human IL-1b antibodies (ab) inhibited the A549 ENA-78 and IL-8 production stimulated by PBMC-CM. Anti-human TNF-a ab inhibited A549 ENA-78 production stimulated by AM-CM. IL-1b and TNFa were expressed in vivo by inflammatory cells, although TNF-a was also expressed by tumor cells. Conclusions: This work emphasizes the role of the host inflammatory response in promoting tumor growth in vivo. Correspondence to: J. Cadranel

Key words: Adenocarcinoma – chemokines – lung – macrophages – peripheral blood mononuclear cells

Introduction C-X-C chemokines, especially interleukin-8 (IL-8) and epithelial cell derived neutrophil-activating protein-78 (ENA-78), seem to have a pivotal role on tumor progression and survival of patients with non small cell lung carcinoma (NSCLC) [1–4]. They mediate their actions via specific cell surface receptors (CXC-CR-1 and -2) expressed on endothelial cells and tumor cells but also on neutrophils. C-X-C chemokines are among the most important angiogenic factors in NSCLC. NSCLC-derived IL-8 and ENA78 stimulated the chemotaxis of endothelial cells and promoted angiogenesis in corneal pocket assay [5, 6]. They also enhanced the growth and angiogenesis of two NSCLC cell lines in murine tumor models [6, 7]. C-X-C should also act on tumor progression through an autocrine pathway. However, addition of neutralizing anti-IL8 antibody to the culture medium of several NSCLC cell lines did not reduce their proliferation [7]. Another possible mechanism for the association of C-X-C chemokines with tumor progression is through the chemoattraction of inflammatory cells. In this context, we have previously shown that tumor cells drive local neutrophil recruitment via C-X-C chemokines release and that increased numbers of tumor-infiltrating neutrophils are linked to poorer outcome in patients with adenocarcinoma (ADC) of the bronchioloalveolar carcinoma subtype (BAC) [2]. Inflammatory cells could act as promoting forces in NSCLC tumor development by several ways [2, 8–11]. They could provide a first defense line against tumor cells releasing

Vol. 53, 2004

Tumor derived C-X-C chemokines and inflammatory cells

5

soluble chemotactic factors that guide the recruitment of both nonspecific and specific effector cells [12, 13]. Conversely, they could promote tumor progression producing several potent angiogenic cytokines, growth factors, and proteases [13, 14] and could also contribute to the burden of genetic abnormalities associated with tumor progression inducing DNA damage in epithelial cells through their generation of reactive oxygen and nitrogen species [15]. Finally, it has been shown that they could favorize a chronic bypass of p53 regulatory functions creating an environment with a deficient reponse to DNA damage and accumulation of potential oncogenic mutation [16]. In the present study, we evaluated the potential role of inflammatory cells on the C-X-C chemokine production by ADC of the BAC subtype. We first analysed the role of peripheral blood mononuclear cells (PBMC) and alveolar macrophages (AM) on IL-8 and ENA-78 tumor cell release in vitro and then looked for the implication of IL-1b and TNF-a as regulatory mediators. We also evaluated the expression of these cytokines in vivo in BAC tissue.

donors. Platelets were separated by isolation of platelet-rich plasma. Cells were purified by Ficoll (Eurobio, Les Ulis, France) density gradient centrifugation. The PBMC layer thus obtained was washed three times in sterile saline solution. PBMC included 65 ± 10% lymphocytes and 35 ± 13% monocytes, as evaluated by flow cytometry. PMN were separated from erythrocytes by hypotonic exposure then washed three times in sterile saline solution. This method yielded > 97% pure PMN, as assessed by May Grünwald Giemsa (MGG) staining. ‘Fresh’ alveolar macrophages (AM) were recovered from 10 subjects who underwent fiberoptic bronchoscopy and BAL during during diagnosis procedures. They consisted in 6 men and 4 women; mean age: 48 ± 15 years; 7 smokers and 3 non smokers. Cells were separated from BAL fluid by centrifugation and were then resuspended as described below. Alveolar macrophages, as assessed by MGG staining, accounted for more than 95% of the recovered cell population. Peripheral blood mononuclear cells, PMN, and AM were resuspended at densities of 25 ¥ 103/ml to 5 ¥ 105/ml in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 2% fetal bovine serum (FBS), 5 mM HEPES, 2 mM L-glutamine, 105 units/L penicillin and 100 mg/L streptomycin, referred to below as complete medium (Gibco™, Invitrogen, Cergy-Pontoise, France). They were plated in 24-well cell culture dishes (500 ml/well) and cultured at 37°C in humidified 5 % CO2/95% air.

Materials and methods

BAC cell line culture

Clinical samples Patients with bronchioloalveolar carcinoma (BAC) Between 01/1992 and 01/1999, BAC was diagnosed in a total of 31 patients, who were treated and followed-up in our chest department. The diagnosis was based on published criteria i.e. a peripheral pleural-based tumor manifesting a well-differentiated adenocarcinomatous infiltration pattern along alveolar septa, with an intact interstitial framework of the lung and no evidence of primary ADC at an extrapulmonary [17]. The degree of intraalveolar proliferation, including papillary proliferation manifested by carcinoma cells and the presence of solid areas of moderately to poorly differentiated ADC did not exclude the tumor from the BAC category if areas of classic BAC were also present [17]. In 26 patients, bronchoalveolar lavage (BAL) was performed to confirm the diagnosis [18]. Briefly, 200 ml of sterile saline solution (four 50-ml aliquots) was infused into the radiologically abnormal segment or lobe. The fluid was recovered by gentle suction, pooled and filtered through sterile gauze, and used for total and differential cell counts, tumor cell examination, and microbiological analysis. After centrifugation, BAL supernatants were stored at –80°C for chemokine assays. The patients were 18 men and 8 women aged 59 ± 11 years (mean ± SD; range 34 to 81 years). There were 14 active smokers (43 ± 4.1 pack-years; range, 15 to 120 pack-years) and 12 nonsmokers. No additional samples or tissues were collected specifically for this study. Data were analyzed anonymously.

Control group We used data on 5 healthy volunteers whose served in a previous study as a control group for cellularity and chemokine measurements [2]. The control subjects were 2 men and 3 women aged 50 ± 4 years. Two were smokers (34 and 42 pack-years, respectively). None had a history of pulmonary or neoplastic disease. Bronchoalveolar lavage was carried out in the middle lobe, and the supernatants were stored at –80°C after centrifugation.

Inflammatory cell recovery, preparation, and generation of conditioned media Peripheral blood mononuclear cells (PBMC) and polymorphonuclear neutrophils (PMN) were isolated from peripheral blood of healthy

The A549 cell line (American Type Culture Collection, Rockville, MD) was originally established in culture from a patient with BAC carcinoma [19]. A549 cells were resuspended at a density of 105/ml in complete medium and cultured in 24-well cell culture plates (500 ml/well) for 72 h at 37°C. Medium was then harvested and confluent monolayers of A549 cells were exposed to complete medium or to PMN, PBMC or AM-CM. In some experiments PMN, PBMC and AM were directly placed on the A549 monolayer. Cell-free supernatants were collected after 30 min to 24 h of incubation and stored at –80°C. In some experiments total A549 cell RNA was isolated for reverse transcription and PCR amplification of IL-8 and ENA-78 mRNA, as described below. In neutralization studies, CM was preincubated for 45 min with (a) a neutralizing goat polyclonal antihuman TNF-a antibody (ab) (1:250 to 1:100 dilution) (Genzyme Diagnostics, Cergy Saint-Christophe, France), (b) a neutralizing goat polyclonal antihuman IL-1b ab (1:250 to 1:100 dilution) (R&D systems, Abingdon, UK), (c) both anti-TNF-a and anti-IL-1b ab, or (d) control goat polyclonal ab (R&D systems). Antibody-treated CM was placed on A549 monolayers for 24 h at 37 °C. Cell-free supernatants were then collected and stored at –80°C.

ELISA measurements ENA-78, IL-8, IL-1b and TNF-a concentrations were measured in BAL fluid and cell culture supernatants by using commercial ELISA kits (R&D systems) as indicated by the manufacturer; the kits consistently detected ENA-78, IL-8, IL-1b and TNF-a at concentrations > 15, 10, 1, 4 pg/ml, respectively, in a linear fashion.

RNA isolation, reverse transcription, and PCR amplification Total RNA from 5 ¥ 106 cells was extracted using the TRIZOL® Reagent (Gibco™) according to the manufacturer’s recommendations. Complementary DNAs (cDNAs) were prepared from 1 mg of total RNA using Oligo (dT)12–18 primers and Moloney murine leukemia virus reverse transcriptase (MMLV-RT) (200 units/ml) (Gibco™). Then, cDNA samples were used for specific PCR amplification with ENA-78 and IL8 primers designed from the following cDNA sequences (Gibco™): ENA78-F, 5¢-GCCCGTGTCCCCGGTCCTTCGAG-3¢; ENA-78-R, 5¢-CTG GATCAAGACAAATTTCCTTC-3¢; IL8-F, 5¢-CTGCGCCAACACA

6

M. Wislez et al.

Inflamm. res.

GAAATTA-3¢; IL8-R, 5¢-ATTGCATCTGGCAACCCTAC-3¢. Co-amplification of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was performed in each experiment with GAPDH primers (Gibco™) to compare the amounts of RNA in samples (GAPDH-F, 5¢CCACCCATGGCAAATTCCATGGCA-3¢; GAPDH-R, 5¢-TCTGACG GCAGGTCAGGTCCACC-3¢). For ENA-78, the reaction conditions were 1X (94°C, 5 min), 28X (94°C, 1 min 15 s: 55°C, 2 min: 72°C, 2 min), and 1X (72°C, 10 min). For IL-8, the reaction conditions were 1 X (94°C, 5 min), 30X (94°C, 1 min 15 s: 60°C, 2 min: 72°C, 2 min), and 1 X (72°C, 10 min). PCR products were analyzed by electrophoresis in 1.5% agarose gel containing 0.2 mg/ml ethidium bromide (Sigma, Saint Quentin Fallavier, France). In all PCR amplifications, we used the reverse transcription reaction mixture without MMLV-RT and the PCR mixture without a cDNA template (blank) as controls (data not shown). The bands were scanned densitometrically on the gels, using an Imager scanner (Appligene, Illkirch, France) and NIH image densitometry software (version 1.44). The sizes of the PCR products for ENA, IL8, and GAPDH were 255, 238, and 598 bases, respectively.

intensity of immunostaining was graded from – (absent) to +++ (strongly positive); complete agreement was obtained between two independent observers. Immunoreactivity with anti-CD14 and anti-CD3 identified cells (i.e. monocytes/macrophages and T lymphocytes, respectively) that expressed the cytokines.

Immunohistochemical studies of IL-8, ENA-78, IL-1b, TNF-a expression in BAC and normal pulmonary tissue samples

BAL cell results

We tested tumor tissues and normal tissue distant from the tumor from 5 patients with BAC. Tissue fragments were immediately frozen in liquid nitrogen and stored at –80°C. Sections (4 mm thick) were fixed in acetone and reacted with appropriate dilutions of antibodies. Monoclonal antihuman TNF-a (Serotec, Oxford, UK), IL-1b (Genzyme, Cambridge, MA, USA), CD3 (Leu-4, Becton Dickinson, San Jose, CA, USA), and CD14 (Pharmingen, BD Biosciences, Le Pont de Claix, France) abs were used. Isotype-matched abs were used as controls (MOPC 21, IgG1; UCP10, IgG2a, Sigma and IgM, BD Biosciences). Positive cells were revealed using the Vectastain ABC-alkaline phosphatase kit system (Vector, Burlingame, CA, USA) and fast red substrate. For IL-8 and ENA-78 immunostaining, paraffin sections from routinely processed samples were used. They were first pretreated by microwave heating for antigen retrieval. They were then reacted with appropriate dilutions of polyclonal rabbit anti-human IL-8 (Peprotech, Rocky Hill, NJ, USA) or goat anti-human ENA-78 ab (R&D systems). Rabbit and goat polyclonal immunoglobulin G were used as controls (R&D systems). Sections were then reacted with a biotinylated secondary antibody (DAKO, Trappes, France) and peroxidase-conjugated streptavidin (DAKO). Positive cells were revealed by reaction with the substrate chromogen (diaminobenzidine and hydrogen peroxide) (DAKO). To test the specificity of immunostaining, abs were replaced by control ab; no positive cells were identified in these conditions. The

Fig. 1. Distribution of IL-8, ENA-78 in BAL supernatants from controls (n = 6) (open bars) and BAC (n = 26) (closed bars) assayed by ELISA.

Statistical analysis Spearman’s r coefficient was used for correlation studies between quantitative variables. Comparisons of quantitative variables between cases and controls were made using the Mann-Whitney non parametric test. Comparisons between results of experiments with the cell line were made using the Wilcoxon non parametric test. Data are presented as means ± SEM with p values. P values below 0.05 were considered significant.

Results

The total BAL cell count was increased in patients with BAC as compared with controls (944 ± 242 vs 242 ± 101 cells/ml; p = 0.038). This resulted mainly from a marked increase in total (240 ± 62 vs 4 ± 2 cells/ml; p = 0.007) and differential (26 ± 6% vs 1 ± 0.2%; p = 0.004) neutrophil counts. Macrophage and lymphocyte total counts did not differ between patients and controls. ENA-78 and IL-8 production in ADC airspace We determined the concentrations of the two C-X-C chemokines, ENA-78 and IL-8, in vivo in lung airspace from patients with BAC and controls using specific ELISA. As shown in Fig. 1, all the two chemokines were detected in BAL fluid from the patients, while only ENA-78 was detected in the controls. IL-8 and ENA-78 BAL fluid concentrations were significantly higher in BAC patients (285 ± 88 pg/ml vs 0 pg/ml; p = 0.003 and 478 ± 129 pg/ml vs 17 ± 3 pg/ml; p = 0.0004, respectively). Mean BAL IL-8 (286 ± 16 vs 267 ± 153 pg/ml in smokers and non smokers, respectively) and ENA-78 (346 ± 144 vs 647 ± 231 pg/ml in smokers and non smokers, respectively) levels did not differ between smokers and nonsmokers in BAC patients. The cellular sources of IL-8 and ENA-78 were identified by immunohistochemical techniques in lung tissue sections from 5 patients with BAC (Table 1, Fig. 2). The results for the five different specimens were similar. Interestingly, IL-8 and ENA-78 staining was observed in tumor cells in each specimen, as well as in AM associated within the tumor. The staining was homogeneous and cytoplasmic. IL-8 expression was sometimes observed in macrophages present in the alveolar lumen from the adjacent normal lung, and in a minority of neutrophils and plasma cells. Normal alveolar epithelial cells, endothelial cells and fibroblasts did not exhibit IL-8 or ENA-78 immunostaining. As the C-X-C chemokines are chemoattractant for neutrophils, we tested IL-8 and ENA-78 BAL fluid concentrations for correlations with alveolar neutrophil count. A strong positive correlation was observed between IL-8 and ENA-78 BAL fluid concentrations and the neutrophil percentage in BAC patients (R2: 0.782; p < 0.0001 and R2: 0.662; p =

Vol. 53, 2004

Tumor derived C-X-C chemokines and inflammatory cells

7

Table 1. IL-8, ENA-78, TNF-a and IL-1b expression in BAC and normal lung.

Tumor area – tumor cells – monocytes/macrophages – lymphocytes – polymorphonuclear cells – endothelial cells – smooth muscle cells – fibroblasts Normal lung – alveolar, bronchiolar and bronchial cells – monocytes/macrophages – lymphocytes – endothelial cells – smooth muscle cells – fibroblasts

IL-8

ENA-78

TNF-a

IL-1b

+++ ++ – +* – – –

+++ + – – – – –

+++ –/+ – – ++ – ++

– + – – + – –

– + – – – –

– – – – – –

++ –/+ – ++ – ++

– + – + – –

IL-8, ENA-78, TNF-a and IL-1b immunoreactivity was graded from – (absent) to +++ (strongly positive). *a minority of polymorphonuclear cells expressed IL-8.

Fig. 2. Immunohistochemical evaluation of cytokine expression in BAC tissues. A: H&S-stained parrafin section showing typical aspect of BAC tumor. B and C: immunostaining using polyclonal antibodies and the peroxydase technique on paraffin sections. Although the use of parrafin embedded material is known to give usually less intense positive signals, tumor cells were found to unequivocally express IL-8 (B) and ENA-78 (C) (compare with D), whereas intertitial cells were negative. D: negative control using non-immune rabbit IgG. E– H: immunostaining using monoclonal antibodies and the alkaline phosphatase technique on frozen sections. E: immunodetection of TNF-a. Tumor cells are strongly positive, whereas interstitial cells and alveolar macrophages express less intensely this cytokine. F: immunostaining with anti-CD3 antibody showing the infiltration of alveolar walls by numerous T lymphocytes. G: immunostaining with anti-CD14 antibody. Numerous strongly positive macrophages are present within tumoral alveolar lumen. H: immunodetection of IL-1b showing that macrophages present either in normal bronchiolar lumen or in tumoral alveolar lumen (inset) express this cytokine. (Original magnification: A: ¥ 200 ; B–G, ¥ 125 ; H: ¥ 300).

8

0.0007, respectively). IL-8 and ENA-78 concentrations did not correlate with AM counts. Altogether, these data suggested that C-X-C chemokines produced by both tumor cells and tumor-infiltrating macrophages were biologically active, but that alveolar macrophages were not the predominant source of production in the tumor microenvironment. Effect of inflammatory cells on tumor-derived ENA-78 and IL-8 production To assess in vitro the potential role of inflammatory cells in the regulation of tumor IL-8 and ENA-78 production in vivo, we evaluated the effect of PBMC, PMN and AM on the constitutive release of these chemokines by A549 cells.

M. Wislez et al.

Inflamm. res.

culture supernatant. Conversely, exposure A549 cells to PBMC or AM during 24 h allowed ENA-78 protein increase that was time- and dose-dependent, and ENA-78 protein accumulation was significantly increased above control values at 4 h and at all time points thereafter, reaching a plateau at 24 h (Fig. 3A). However, ENA-78 protein increase was higher in coculture supernatant of A549 with PBMC than with AM (581 ± 85 ng/ml vs 348 ± 100 ng/ml, p = 0.03). The production was significantly increased from a PBMC- or an AM- to tumor cell ratio as low as 1:20 (Fig. 3B). The protein accumulation was related to A549 cells and not to PBMC or AM, as the accumulation was similar when tumor cells were cultured directly with PBMC or AM and with their 24 hrespective CM (Fig. 3C). This result was also confirmed by the A549 ENA-78 mRNA accumulation along the 24 h when A549 cells were exposed to PBMC or AM 24 h-CM (Fig. 3D).

Effect on tumor-derived ENA-78 production ENA-78 was constitutively produced in large amounts by A549 cells (22 ± 1 ng/ml). Inflammatory cells, i.e. AM and PBMC, showed lower release (14 ± 4 ng/ml and 2 ± 0.6 ng/ml, respectively) and PMN showed no release. Exposure A549 cells to PMN during 24 h did not allow ENA-78 protein increase as measured by ELISA in the cell

Effect on tumor-derived IL-8 production By contrast to ENA-78, unstimulated A549 tumor cells had moderate constitutive IL-8 production, releasing 0.5 ± 0.1 ng/ ml in 24 h. Unstimulated PBMC and AM constitutively released large amounts of IL-8 (34 ± 6 ng/ml and 182 ±

Fig. 3. A: Time course of ENA-78 accumulation in the medium of A549 cells cultured without (open circles) or with PBMC (closed circles), or AM (crosshatched circles). Results are from 3 experiments for PBMC and expressed as means ± SEM. *means p < 0.05. Results are from 2 experiments for AM. B: Dose dependent ENA-78 extra-cellular protein accumulation by A549 cells with PBMC (closed bars) or AM (crosshatched bars). Results are from 3 experiments and are expressed as means ± SEM. * means p < 0.05. C: ENA-78 was measured in the supernatant of A549 cells, PBMC or AM alone or in the supernatant of A549 cells exposed to PBMC or AM directly or to their respective conditioned media. At 24 h, supernatants were collected and assayed by ELISA as described in Methods. Results are from 6 experiments and are expressed as means ± SEM. *means p < 0.05 with cells alone. D: Time course of ENA-78 mRNA expression by A549 cells cultured with or without PBMC or AM conditioned media. Total RNA was isolated after various periods of time and reverse transcribed. Resulting complementary DNA was subjected to reverse transcriptase polymerase reaction (RT-PCR) to evaluate steady state levels of mRNA for ENA-78 and GAPDH. AM: alveolar macrophage. PBMC: peripheral blood mononuclear cell. AM-CM: alveolar macrophage conditioned media; PBMC-CM: peripheral blood mononuclear cell conditioned media.

Vol. 53, 2004

Tumor derived C-X-C chemokines and inflammatory cells

41 ng/ml in 24 h, respectively), in contrast to PMN (1.5 ± 0.05 ng/ml). Exposure A549 cells to PMN during 24 h did not allow IL-8 protein increase as measured by ELISA in coculture supernatant. Exposure A549 cells to PBMC or AM during 24 h allowed IL-8 protein increase that was time- and dosedependent, and IL-8 protein accumulation was significantly increased above control values at 4 h and at all time points thereafter, reaching a plateau at 24 h (248 ± 44 ng/ml, p = 0.04 and 182 ± 41 ng/ml) (Fig. 4A). Production was significantly increased from a PBMC or AM to tumor cell ratio as low as 1:20 (Fig. 4B). However, protein accumulation in coculture supernatant was related to A549 cells and not to PBMC in A549-PBMC coculture, and to AM and not A549 cells in A549-AM coculture (Fig. 4C). A549 IL-8 mRNA increased along the 24 h when A549 cells were exposed to PBMC 24 h CM while moderately increased and only at 4 h when A549 cells were cultured with AM (Fig. 4D). Effect of inflammatory cells derived-cytokines on ENA-78 and IL-8 production by tumor cells The fact that IL-8 and ENA-78 mRNA and protein accumulation was similar when tumor cells were cultured with inflam-

9

matory cells directly or with their 24-h CM also suggested that the regulation of A549-derived IL-8 or ENA-78 depended mainly on the presence of soluble factors. As the proinflammatory cytokines IL-1b and TNF-a have previously been shown to enhance IL-8 and ENA-78 production by epithelial cells, we performed neutralization studies on IL-1b and TNF-a to investigate the role of these two cytokines in the regulation of A549-derived IL-8 and ENA-78. Anti-IL-1b ab, alone and in combination with antiTNF-a ab, inhibited 39% and 70% of PBMC-CM stimulated ENA-78 release, respectively. In contrast, PBMC-CM stimulated ENA-78 release was not inhibited by anti-TNF-a ab alone (Fig. 5A) or by the control goat ab (data not shown). We then performed the same experiment with AM-CM. AntiTNFa ab, alone and in combination with anti-IL-1b ab, inhibited 53% and 75% of AM-CM stimulated ENA-78 release, respectively, while anti-IL-1b ab alone (Fig. 5A) and the control goat ab had no effect (data not shown). Neutralization studies on A549-derived IL-8 yielded a similar inhibition pattern. Indeed, anti-IL-1b ab, alone and in combination with anti-TNF-a ab, inhibited 47% and 75% of PBMC-CM stimulated tumor cell IL-8 release, respectively (Fig. 5B). PBMC-CM stimulated tumor IL-8 release was not inhibited by anti-TNF-a ab alone (Fig. 5B) or by the control goat ab (data not shown). The fact that A549 cells did not

Fig. 4. A: Time course of IL-8 accumulation in the medium of A549 cells cultured without (open circles) or with PBMC (closed circles), or AM (crosshatched circles). Results are from 3 experiments for PBMC and expressed as means ± SEM. *means p < 0.05. Results are from 2 experiments for AM. B: Dose dependent IL-8 extra-cellular protein accumulation by A549 cells with PBMC (closed bars) or AM (crosshatched bars). Results are from 3 experiments and are expressed as means ± SEM.*means p < 0.05. C: IL-8 was measured in the supernatant of A549 cells, PBMC or AM alone or in the supernatant of A549 cells exposed to PBMC or AM directly or to their respective conditioned media. At 24 h, supernatants were collected and assayed by ELISA as described in Methods. Results are from 6 experiments and are expressed as means ± SEM. * means p < 0.05 with cells alone. D: Time course of IL-8 mRNA expression by A549 cells cultured with or without PBMC or AM conditioned media. Total RNA was isolated after various periods of time and reverse transcribed. Resulting complementary DNA was subjected to reverse transcriptase polymerase reaction (RT-PCR) to evaluate steady state levels of mRNA for IL-8 and GAPDH. AM: alveolar macrophage. PBMC: peripheral blood mononuclear cell. AM-CM: alveolar macrophage conditioned media; PBMC-CM: peripheral blood mononuclear cell conditioned media.

10

M. Wislez et al.

Inflamm. res.

histochemical study of both tumor tissue and adjacent normal lung tissue (Table 1 and Fig. 2). The results obtained with three different specimens were similar. The TNF-a-producing cells were mainly tumor cells and AM. The staining was strong, homogeneous and cytoplasmic. It should be noted, however, that interstitial mononuclear cells did not express this cytokine, contrary to normal epithelial cells, fibroblasts and endothelial cells. IL-1b expression was restricted to AM in tumor area as well as in the distant normal tissue, but AM reactivity was inconsistent and faint. IL1-b expression was not found in interstitial mononuclear cells (mainly T lymphocytes). Taken together, these data and those obtained in vitro suggest that TNF-a and IL-1b could participate in the upregulation of tumor-derived C-X-C chemokines. However, it should be stressed that the cellular source of these proinflammatory cytokines in vivo was not only monocyte/ macrophages but also resident cells and tumor cells themselves. Discussion

Fig. 5. Inhibition of ENA-78 (A) and IL-8 (B) protein release from A549 cells, stimulated with PBMC (closed bars) or AM (crosshatched bars) conditioned media, by antihuman TNF-a or IL-1b antibodies (ab). At 24 h, supernatants were collected and IL-8 and ENA-78 were measured by ELISA as described in Methods. Results are from 3 experiments and are expressed as means ± SEM. *means p < 0.05.

Table 2. TNF-a and IL1-b concentrations in cell-free supernatants.

A549 PBMC A549 + PBMC-CM AM A549 + AM-CM

TNF-a (pg/ml)

IL-1b (pg/ml)

nd 318 ± 117 98 ± 33 661 ± 384 227 ± 122

nd 102 ± 56 80 ± 44 1±1 0.6 ± 0.6

Peripheral blood mononuclear cells (PBMC) and alveolar macrophages (AM) were cultured in 24-well plates for 24 h. PBMC and AM conditioned media (CM) were then collected and stored at –80°C. A549 cells were cultured in 24-well plates with complete medium, PBMC-CM or AM-CM. At 24 h, cell-free supernatants were collected and assayed by ELISA as described in Methods. Results are from 3 experiments and are expressed as means ± SEM. nd, not detectable.

release detectable amounts of IL-1b or TNF-a either constitutively or after exposure to mononuclear cell CM (Table 2) eliminated the possibility of autocrine production of IL-1b and TNF-a by A549 cells. Furthermore, IL-1b and TNF-a proteins appeared to be degraded or internalized by A549 cells (Table 2). To determine the cellular origin of IL-1b and TNF-a in the tumor environment in vivo, we performed an immuno-

We observed that C-X-C peptides (detected by immunohistochemical studies) were present in tumor cells of all the specimens tested, confirming previous reports by us and others [2, 5, 6]. ENA-78 and IL-8 were also expressed by AM, suggesting the participation of this cell type in total chemokine production. As the C-X-C chemokines are chemoattractant for neutrophils, the fact IL-8 and ENA-78 BAL fluid concentrations strongly correlated with alveolar neutrophil count suggested that C-X-C chemokines produced by both tumor cells and tumor-infiltrating macrophages were biologically active. These findings were not specific of the BAC condition. Indeed, an increased production of C-X-C chemokines by normal epithelial cells or alveolar macrophages had also been described in several other chronic inflammatory lung diseases, such as chronic bronchitis or idiopathic pulmonary fibrosis [20, 21]. Inflammatory cells, i.e. neutrophils, lymphocytes and monocytes/macrophages, might also indirectly participate to the total chemokine production in vivo by stimulating tumor cell production. Tumor cells have close contacts with a large number of inflammatory cells such as neutrophils and macrophages (apical pole) and mononuclear cells (baso-lateral pole) [2, 22] which are known to produce a large number of mediators in vivo that upregulate the release of C-X-C chemokines in several cellular systems in vitro [23–25]. To test this hypothesis in vitro, we evaluated the effect of PMN, PBMC and AM on the spontaneous release of ENA-78 and IL-8 by a tumor cell line derived from a patient with BAC (A549-cell line). We found that ENA-78 and IL-8 production was mostly attributable to tumor cells after exposure to inflammatory cells. Interestingly, the inflammatory cells stimulated different sets of tumor-derived C-X-C chemokines to different extents. PMN failed to enhance C-X-C chemokine production whatever the PMN to A549 cell ratio. Although PBMC were able to stimulate ENA-78 and IL-8 production to a similar extent, AM induced only half the increase in ENA-78 protein induced by PBMC, and were unable to induce mesurable IL-8 mRNA expression or protein increase by A549 cells. These differences between ENA-

Vol. 53, 2004

Tumor derived C-X-C chemokines and inflammatory cells

78 and IL-8 mRNA expression and protein accumulation are consistent with previous studies showing substantial differences in the intracellular signaling pathways of ENA-78 and IL-8 regulation [23, 26]. The absence of clear induction of C-X-C chemokines production by PBMC and AM exposed to A549 cells in vitro was probably related to the lack of constitutive IL-1b and TNFa production by this cell line (Table 2). However, AM can probably be stimulated in vivo by tumor cells to express C-X-C chemokines (and especially IL-8) as we showed that tumor cells expressed TNFa (Fig. 2) and GM-CSF [27]. This hypothesis was also supported by the results of immunohistochemical studies performed on tumor and adjacent normal lung from patients showing an IL-8 staining in AM within the tumor area while faint and inconstantly in AM present in the adjacent normal lung. To better understand how AM and PBMC stimulate C-XC chemokine production by the A549 cell line, we performed neutralizing studies with anti-IL-1b and anti-TNF-a antibodies. Interestingly, AM and PBMC exhibited different patterns of activity, as they operated via generation of TNF-a and IL-1b, respectively. However, IL-1b and TNF-a had synergistic activities, as shown by experiments with both cytokine-specific antibodies. Surprisingly, TNF-a and IL-1b concentrations in the AM- and PBMC-CM used to stimulate A549-derived IL-8 and ENA-78 release were 50-fold and 200-fold lower than those used in previous studies of the effect of recombinant cytokines on epithelial cells in vitro [23, 24]. This suggests that the PBMC and AM-CM contained other factors that might directly regulate A549 cell chemokine production or enhance the expression of IL-1b and TNF-a receptors on A549 cells. To confirm in vivo the potential role of tumor-infiltrating host cells in the amplification of tumor ENA-78 and IL-8 production, we finally evaluated TNF-a and IL1-b protein expression in lung tissue from patients with BAC. Using immunohistochemical studies, we detected TNF-a and IL-1b both in tumor areas and normal distant tissue of all the specimens tested. Expression of both TNF-a and IL-1b by AM suggested that these cells could participate in the upregulation of tumor-derived C-X-C chemokines. Furthermore, the predominant role of macrophage-derived TNF-a in the C-X-C chemokine upregulation observed in vitro was also probable in vivo. Indeed, TNF-a expression by AM was strong and homogeneous, whereas that of IL-1b was faint and inconsistent. The fact that AM are relatively poor IL-1 producers compared with blood monocytes has previously been reported [28, 29]. Interestingly, epithelial tumor cells also contributed to local TNF-a production, as shown by immunohistochemical studies and in keeping with previous reports [30]. Furthermore, the fact that TNF-a was detected in both tumor areas and normal distant tissue while ENA-78 and IL-8 were detected in tumor but not normal epithelial cells prompted the question on a different sentivity of these cells in response to TNF-a. In this setting, Shimomoto et al. have previously reported the specific expression of both the 55 kDa and 75 kDa TNF-a receptors in lung tumor but not normal alveolar epithelial cells [31]. In conclusion, this work emphasizes the role of the host inflammatory cells response promoting tumor growth in vivo. Inflammatory cells release cytokines or other products

11

acting directly on tumor cell growth or metastasis as well as indirectly stimulating tumor cell cytokine synthesis and release. This is in keeping with several reports providing evidence that constitutive and increased TNF-a inducible NF-kB activation in tumor cells of the squamous cell carcinoma type allow proinflammatory cytokines expression then promoting tumor growth and metastasis in vivo whereas not in vitro [32, 33]. All these data are also consistent with the fact that host response could promote tumor progression and had to be a new targets for therapy [34]. Acknowlegments. This study was supported by grants from ‘Leg Poix’, ‘Fondation pour la recherche médicale’ and the ‘Pole d’Etude du Microenvironnement Tumoral’, ‘Ligue Nationale Contre le Cancer’. M. Wislez is a doctoral fellow of the ‘Le Ministère de l’Education Nationale, de la Recherche et de la Technologie’.

References [1] Freemantle SJ, Dmitrovsky E. Clinical link between p53 and angiogenesis in lung cancer. J Clin Oncol 2002; 20: 883–4. [2] Bellocq A, Antoine M, Flahault A, Philippe C, Crestani B, Bernaudin JF et al. Neutrophil alveolitis in bronchioloalveolar carcinoma: induction by tumor-derived interleukin-8 and relation to clinical outcome. Am J Pathol 1998; 152: 83–92. [3] Yuan A, Yang PC, Yu CJ, Chen WJ, Lin FY, Kuo SH et al. Interleukin-8 messenger ribonucleic acid expression correlates with tumor progression, tumor angiogenesis, patient survival, and timing of relapse in non-small-cell lung cancer. Am J Respir Crit Care Med 2000; 162: 1957–63. [4] Yuan A, Yu CJ, Luh KT, Kuo SH, Lee YC, Yang PC. Aberrant p53 expression correlates with expression of vascular endothelial growth factor mRNA and interleukin-8 mRNA and neoangiogenesis in non-small-cell lung cancer. J Clin Oncol 2002; 20: 900–10. [5] Smith DR, Polverini PJ, Kunkel SL, Orringer MB, Whyte RI, Burdick MD et al. Inhibition of interleukin 8 attenuates angiogenesis in bronchogenic carcinoma. J Exp Med 1994; 179: 1409–15. [6] Arenberg DA, Keane MP, DiGiovine B, Kunkel SL, Morris SB, Xue YY et al. Epithelial-neutrophil activating peptide (ENA-78) is an important angiogenic factor in non-small cell lung cancer. J Clin Invest 1998; 102: 465–72. [7] Arenberg DA, Kunkel SL, Polverini PJ, Glass M, Burdick MD, Strieter RM. Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice. J Clin Invest 1996; 97: 2792–802. [8] Kamimura A, Kamachi M, Nishihira J, Ogura S, Isobe H, DosakaAkita H et al. Intracellular distribution of macrophage migration inhibitory factor predicts the prognosis of patients with adenocarcinoma of the lung. Cancer 2000; 89: 334–41. [9] Kasuga I, Makino S, Kiyokawa H, Katoh H, Ebihara Y, Ohyashiki K. Tumor-related leukocytosis is linked with poor prognosis in patients with lung carcinoma. Cancer 2001; 92: 2399–405. [10] Nakahashi H, Yasumoto K, Nagashima A, Yaita H, Takeo S, Motohiro A et al. Antitumor activity of macrophages in lung cancer patients with special reference to location of macrophages. Cancer Res 1984; 44(12 Pt 1): 5906–9. [11] Ylisirnio S, Hoyhtya M, Makitaro R, Paaakko P, Risteli J, Kinnula VL et al. Elevated serum levels of type I collagen degradation marker ICTP and tissue inhibitor of metalloproteinase (TIMP) 1 are associated with poor prognosis in lung cancer. Clin Cancer Res 2001; 7: 1633–7. [12] di Carlo E, Iezzi M, Pannellini T, Zaccardi F, Modesti A, Forni G et al. Neutrophils in anti-cancer immunological strategies: old players in new games. J Hematother Stem Cell Res 2001; 10: 739– 48. [13] Di Carlo E, Forni G, Lollini P, Colombo MP, Modesti A, Musiani P. The intriguing role of polymorphonuclear neutrophils in antituor reactions. Blood 2001; 97: 339–45.

12 [14] Wahl LM, Kleinman HK. Tumor-associated macrophages as targets for cancer therapy. J Natl Cancer Inst 1998; 90: 1583–4. [15] Sandhu JK, Privora HF, Wenckebach G, Birnboim HC. Neutrophils, nitric oxide synthase, and mutations in the mutatect murine tumor model. Am J Pathol 2000; 156(2): 509–18. [16] Hudson JD, Shoaibi MA, Maestro R, Carnero A, Hannon GJ, Beach DH. A proinflammatory cytokine inhibits p53 tumor suppressor activity. J Exp Med 1999; 190: 1375–82. [17] Barsky SH, Cameron R, Osann KE, Tomita D, Holmes EC. Rising incidence of bronchioloalveolar lung carcinoma and its unique clinicopathologic features. Cancer 1994; 73: 1163–70. [18] Springmeyer SC, Hackman R, Carlson JJ, McClellan JE. Bronchiolo-alveolar cell carcinoma diagnosed by bronchoalveolar lavage. Chest 1983; 83: 278–9. [19] Lieber M, Smith B, Szakal A, Nelson-Rees W, Todaro G. A continuous tumor-cell line from a human lung carcinoma with properties of type II alveolar epithelial cells. Int J Cancer 1976; 17: 62– 70. [20] Keane MP, Belperio JA, Burdick MD, Lynch JP, Fishbein MC, Strieter RM. ENA-78 is an important angiogenic factor in idiopathic pulmonary fibrosis. Am J Respir Crit Care Med 2001; 164: 2239–42. [21] Corvol H, Fitting C, Chadelat K, Jacquot J, Tabary O, Boule M et al. Distinct cytokine production by lung and blood neutrophils from children with cystic fibrosis. Am J Physiol Lung Cell Mol Physiol 2003; 284: L997–1003. [22] Tosi P, Sforza V, Santopietro R, Lio R, Gotti G, Paladini P et al. Bronchiolo-alveolar carcinoma: an analysis of survival predictors. Eur J Cancer 1992; 28A: 1365–70. [23] Standiford TJ, Kunkel SL, Basha MA, Chensue SW, Lynch JPd, Toews GB et al. Interleukin-8 gene expression by a pulmonary epithelial cell line. A model for cytokine networks in the lung. J Clin Invest 1990; 86: 1945–53. [24] Schmouder RL, Streiter RM, Walz A, Kunkel SL. Epithelialderived neutrophil-activating factor-78 production in human renal tubule epithelial cells and in renal allograft rejection. Transplantation 1995; 59: 118–24. [25] Gutman M, Singh RK, Xie K, Bucana CD, Fidler IJ. Regulation of interleukin-8 expression in human melanoma cells by the organ environment. Cancer Res 1995; 55: 2470–5.

M. Wislez et al.

Inflamm. res.

[26] Chang MS, McNinch J, Basu R, Simonet S. Cloning and characterization of the human neutrophil-activating peptide (ENA-78) gene. J Biol Chem 1994; 269: 25277–82. [27] Wislez M, Fleury-Feith J, Rabbe N, Moreau J, Cesari D, Milleron B et al. Tumor-derived granulocyte-macrophage colony-stimulating factor and granulocyte colony-stimulating factor prolong the survival of neutrophils infiltrating bronchoalveolar subtype pulmonary adenocarcinoma. Am J Pathol 2001; 159: 1423–33. [28] Bernaudin JF, Yamauchi K, Wewers MD, Tocci MJ, Ferrans VJ, Crystal RG. Demonstration by in situ hybridization of dissimilar IL-1 beta gene expression in human alveolar macrophages and blood monocytes in response to lipopolysaccharide. J Immunol 1988; 140: 3822–9. [29] Wewers MD, Rennard SI, Hance AJ, Bitterman PB, Crystal RG. Normal human alveolar macrophages obtained by bronchoalveolar lavage have a limited capacity to release interleukin-1. J Clin Invest 1984; 74: 2208–18. [30] Tran TA, Kallakury BV, Ambros RA, Ross JS. Prognostic significance of tumor necrosis factors and their receptors in nonsmall cell lung carcinoma. Cancer 1998; 83: 276–82. [31] Shimomoto H, Hasegawa Y, Nozaki Y, Takagi N, Shibagaki T, Nakao A et al. Expression of tumor necrosis factor receptors in human lung cancer cells and normal lung tissues. Am J Respir Cell Mol Biol 1995; 13: 271–8. [32] Dong G, Chen Z, Kato T, Van Waes C. The host environment promotes the constitutive activation of nuclear factor-kappaB and proinflammatory cytokine expression during metastatic tumor progression of murine squamous cell carcinoma. Cancer Res 1999; 59: 3495–504. [33] Duffey DC, Chen Z, Dong G, Ondrey FG, Wolf JS, Brown K et al. Expression of a dominant-negative mutant inhibitor-kappaBalpha of nuclear factor-kappaB in human head and neck squamous cell carcinoma inhibits survival, proinflammatory cytokine expression, and tumor growth in vivo. Cancer Res 1999; 59: 3468–74. [34] Loukinova E, Dong G, Enamorado-Ayalya I, Thomas GR, Chen Z, Schreiber H et al. Growth regulated oncogene-alpha expression by murine squamous cell carcinoma promotes tumor growth, metastasis, leukocyte infiltration and angiogenesis by a host CXC receptor-2 dependent mechanism. Oncogene 2000; 19: 3477–86.