Impact of tobacco smoke on interleukin-16 protein in human airways, lymphoid tissue and T lymphocytes

Abstract

CD4⁺ and CD8⁺ lymphocytes are mobilized in severe chronic obstructive pulmonary disease (COPD) and the CD8⁺ cytokine interleukin (IL)-16 is believed to be important in regulating the recruitment and activity of CD4⁺ lymphocytes. In the current study, we examined whether tobacco smoke exerts an impact not only on IL-16 in the lower airways but also in CD4⁺ or CD8⁺ lymphocytes or in lymphoid tissue. The concentration of IL-16 protein was measured by enzyme-linked immunosorbent assay (ELISA) in concentrated bronchoalveolar lavage fluid (BALF) collected from 33 smokers with chronic bronchitis (CB), eight asymptomatic smokers (AS) and seven healthy never-smokers (NS). The concentrations of IL-16 and soluble IL-2 receptor alpha (sIL-2Rα) protein were also measured in conditioned medium from human blood CD4⁺ and CD8⁺ lymphocytes stimulated with tobacco smoke extract (TSE) in vitro. IL-16 mRNA was assessed in vitro as well, using reverse transcription–polymerase chain reaction (RT-PCR). Finally, the intracellular immunoreactivity for IL-16 protein (IL-16IR) was assessed in six matched pairs of palatine tonsils from smokers and non-smokers. BALF IL-16 was higher in CB and AS than in NS. TSE substantially increased the concentration of IL-16 but not sIL-2Rα in conditioned medium from CD4⁺ and CD8⁺ lymphocytes. There was no corresponding effect on IL-16 mRNA. IL-16IR in tonsils was lower in smokers than in non-smokers. The current findings demonstrate that tobacco smoke exerts a wide impact on the CD8⁺ cytokine IL-16, in the airway lumen, in blood CD4⁺ and CD8⁺ lymphocytes and in lymphoid tissue. The effect on IL-16 release may be selective for preformed IL-16 in CD4⁺ lymphocytes. New clinical studies are required to evaluate whether tobacco smoke mobilizes T lymphocytes via IL-16 in the lower airways and whether this mechanism can be targeted in COPD.

INTRODUCTION

Recent clinical studies show that patients with severe chronic obstructive pulmonary disease (COPD) display an increase not only in CD8⁺ but also in CD4⁺ lymphocytes in the airway wall [1]. Furthermore, the ratio between CD4⁺ and CD8⁺ lymphocytes is increased in the blood of smokers with COPD and emphysema [2]. However, the underlying cellular mechanisms behind this mobilization of CD4⁺ lymphocytes are not known. What is known is that interleukin (IL)-16 is a very potent chemoattractant for CD4⁺ lymphocytes that can be released by CD8⁺ lymphocytes [3–5]. Interestingly, IL-16 may also mediate non-specific airway hyperresponsiveness, at least in a mouse model of airway allergy [6]. IL-16 may enhance the proliferative response of co-stimulated blood CD4⁺ lymphocytes [7,8]. Thus, IL-16 constitutes a plausible candidate cytokine for mobilizing CD4⁺ lymphocytes in severe COPD.

Earlier clinical studies on smokers with COPD compared to non-smokers have shown a negative correlation between the number of T lymphocytes and lung function, an observation compatible with T lymphocytes playing a pathogenetic role in the development of COPD [9,10]. In addition to this, we recently published clinical evidence that the local concentration of soluble IL-16 protein is increased in the lower airways of tobacco smokers [8]. We demonstrated that this local IL-16 correlates negatively with the percentage of circulating CD4⁺ lymphocytes in blood, but it remains uncertain whether this is due to CD4⁺ lymphocytes being recruited from the blood into the airways [8]. In view of these findings, we hypothesized that tobacco smoke exerts a wide impact on IL-16 by acting directly on its cellular source and that this can alter the traffic of CD4⁺ lymphocytes within and outside the lower airways.

In the present study, we examined whether tobacco smoke exerts an impact not only in lower airways but also in CD4⁺ or CD8⁺ lymphocytes or in lymphoid tissue.

METHODS

IL-16 protein in lower human airways in vivo

Protocol

The clinical protocol of these examinations has been described in detail elsewhere [11]. The study was approved by the Ethics Committee at Göteborg University (Diary no. 246–94). All subjects gave their written consent after receiving both written and oral information.

Study population

Three different groups of subjects were studied: 33 smokers with chronic bronchitis (CB), eight asymptomatic smokers (AS) and seven healthy never-smokers (NS). Some data from this patient material, on end-points not included in the present publication, have been published elsewhere [11]. The clinical characteristics of the subjects included are presented in Table 1.

Table 1.

Characteristics of 7 never-smokers, 8 asymptomatic smokers and 33 smokers with chronic bronchitis (CB). All values are presented as median (range)

	Never smokers	Asymptomatic smokers	Smokers with CB
n	7	8	33
Age (years)	51 (37–64)	45 (34–60)	50 (37–68)
Male/female	4/3	3/5	6/27
FEV1 % pred	105 (91–120)	108 (95–129)	83 (61–111)
Pack-years	0	25 (11–62)	34 (12–88)
Current smoking cigarettes/day	0	18 (10-30)	20 (10–40)
Duration of CB (years)	0	0	10 (4–30)
No. of exacerbations in 2 years	0	0	7 (4–20)
Recovery (%)	68 (36–82)	71 (64–75)	61 (34–82)

For inclusion, all smokers (CB and AS) had to be current smokers with a tobacco exposure corresponding to at least 5 pack-years. The CB subjects had symptoms of chronic bronchitis as defined by the American Thoracic Society [12]. Furthermore, CB subjects also had a history of two or more acute exacerbations during the past 12 months, as defined by Boman et al. [13] and as recorded in Table 1. Co-existing chronic airway obstruction defined as forced expiratory volume during 1 s (FEV₁) <80% of predicted was allowed. The subjects in NS and AS all displayed a normal ventilatory lung function, as defined by FEV₁ (>80% of predicted). The exclusion criteria have been described in detail elsewhere [11]. Two patients were excluded because their bronchoalveolar lavage fluid (BALF) samples contained >1% squamous cells (indicating a more proximal origin in the airways). All examinations were performed during a clinically stable phase of the disease (i.e. CRP <5 mg/l and no clinical symptoms).

Lung function

FEV₁ (% of predicted) was measured with a spirometer (Model Alpha, Vitalograph Ltd, Buckingham, UK), in accordance with the directions of the ERS [14].

BAL samples

Fibreoptic bronchoscopy was conducted as described previously [11]. Harvests of BALF were obtained by instillation and aspiration of totally 160 ml sterile, pyrogen-free phosphate-buffered saline (PBS: pH 7·3, temperature 37°C). The first aliquot (20 ml) was discarded, to minimize cell contamination from the upper airways. The remaining volume (80 + 60 ml) of PBS was instilled, aspirated immediately and pooled in a siliconized glass container on ice. Recovery was recorded (Table 1), and the fluid was subsequently filtered through a nylon web (100 µm pore size) for retention of mucus. The BALF was then centrifuged (250 g, 10 min, +4°C). Finally, the BALF supernatant was separated and stored in frozen aliquots (−80°C).

IL-16 protein in BALF

For all subgroups, BALF was thawed and centrifuged (400 g, 10 min, + 4°C) to remove cell debris. Each BALF sample was concentrated ( × 10, Centricon Centrifugal Filter Device, 10 kDa cut-off; Millipore, Bedford, MA, USA) using centrifugation (3500 g, 60 min, + 4°C) of 1800 µl. Finally, the concentration of soluble IL-16 protein was measured using a commercial enzyme-linked immunosorbent assay (ELISA) kit (BioSource, International Inc., Camarillo, CO, USA). In all concentrated BALF samples, the concentration of IL-16 exceeded the lowest concentration of the standard curve of the ELISA kit (i.e. > 23·4 pg/ml). The presented data on BALF IL-16 are compensated for the concentration procedure.

Release of IL-16 protein and sIL-2Rα protein in blood CD4⁺ and CD8⁺ lymphocytes in vitro

Tobacco smoke extract

The water soluble components of tobacco smoke were extracted as described previously [15]. Briefly, the tobacco smoke extract (TSE) was generated by leading the mainstream smoke from one commercial available cigarette (tar 10 mg, nicotine 0·8 mg and CO 10 mg/cigarette) (Marlboro, Philip Morris, Neuchâtel, Switzerland) be drawn through 15 ml of cell culture medium (RPMI-1640, Sigma-Aldrich, St Louis, MO, USA) by vacuum at room temperature. Each cigarette was smoked during 4 min. The solution was then passed through a sterile filter and this solution was referred to as TSE 1 : 1. The endotoxin concentration in TSE was not detectable (detection limit: 10 pg/ml) using a commercial endotoxin-specific test (Seikagu Co, Tokyo, Japan).

Enrichment of blood CD4⁺ and CD8⁺ lymphocytes and adherent mononuclear cells

Fresh human buffy coat was obtained from healthy blood donors at the Department of Clinical Chemistry and Transfusion Medicine, Sahlgrenska University Hospital, Göteborg, Sweden. The blood sample was diluted in PBS 1 : 4 and the peripheral blood mononuclear cells (PBMC) were collected by density centrifugation over a Ficoll gradient (Pharmacia Biotech, Uppsala, Sweden). CD4⁺ and CD8⁺ lymphocytes were isolated separately using a magnetic separation technique using a commercially available assay (Miltenyi Biotech, Bergish Gladbach, Germany). In short, all cells except CD4⁺ or CD8⁺ lymphocytes were labelled magnetically using a cocktail of antibodies for negative isolation of CD4⁺ and CD8⁺ lymphocytes. Cytospins were made and slides were air-dried and stored frozen until immunocytochemistry was performed. To obtain a single cell layer of adherent mononuclear cells, PBMC were seeded (1 × 10⁶ cells in 200 µl of RPMI-1640) in a 96-well cell culture plate (BD Bioscience, Bedford, MA, USA).

The monocytes in the PBMC sample were allowed to adhere to the dishes in a humified incubator for 2 h (37°C, 7% CO₂). The wells were then washed three times with RPMI-1640 and non-adherent cells were removed. Approximately 80% of the adherent mononuclear cells (AMC) were monocytes (i.e. true AMC, as revealed by May–Grünewald–Giemsa staining).

Cell culture

The separated CD4⁺ and CD8⁺ lymphocytes were resuspended in RPMI-1640 medium supplemented with fetal calf serum (FCS: 10%), l-glutamin (4 m m), sodium pyruvate (100 µg/ml) and penicillin/streptomycin (100 units/100 µg/ml) (all from Sigma-Aldrich) referred to as complete medium. CD4⁺ or CD8⁺ lymphocytes were seeded (0·5 × 10⁶ cells/well) in 96-well tissue plates (BD Bioscience), with or without AMC. The cell cultures were stimulated either with TSE (1 : 300 in complete medium) or with calcium ionophore (CI: 1 µm, batch A 23187) and phorbol-12–myristate-13–acetate (PMA: 2 ng/ml) (positive control) (Sigma-Aldrich) or with complete medium alone (vehicle, negative control). The chosen concentration of TSE caused reproducible IL-16 responses with intact cell viability (see Results). The cells were incubated for 20 h in a humified incubator (7% CO₂, 37C°).

After incubation in vitro, the conditioned medium was collected and frozen immediately (−80°C).

IL-16 protein and sIL-2Rα protein ELISA

IL-16 protein was assessed using the same commercial ELISA kit as for BALF (see above). Soluble IL-2Rα protein was utilized as a marker of general activity in T lymphocytes in vitro [16]. It was analysed using a commercial ELISA kit (R&D Systems Europe, Abingdon, UK). The values below the lowest concentration of the standard curve for IL-16 (23·4 pg/ml) were referred to as the mean value of 0 and 23·4 (i.e. 11·7 pg/ml). The values below the lowest concentration of the standard curve for sIL-2Rα (i.e. 39 pg/ml) were referred to as the mean value of 0 and 39 (i.e. 19·5 pg/ml).

Purity and viability of isolated cells prior to experiments

The purity of CD4⁺ and CD8⁺ fractions were analysed by immunohistochemistry staining. Cytospins of CD4⁺ and CD8⁺ lymphocytes were fixed in acetone (50% followed by 100%) and subsequently dried in air. Endogenous peroxidase was blocked in a solution of Tris-buffered saline (TBS) at 37°C containing sodium azide (0·0064%) glucose (0·18%), saponin (0·1%) and glucoseoxidase (0·1%). Non-specific staining was blocked with 10% rabbit serum (Dako A/S, Glostrup, Denmark), followed by incubation with mouse antihuman CD4⁺ antibody (clone MT310, Dako), mouse antihuman CD8⁺ antibody (clone C8/144B, DakO) or mouse IgG1 antibody (BD Biosciences). A horseradish peroxidase (HRP) conjugated rabbit antimouse IgG1 antibody (Zymed Laboratories Inc., San Fransisco, CA, USA) was used as secondary antibody. Bound antibodies were visualized by 3,3-diaminobenzidine (DAB) substrate chromogen system (Dako) and were counterstained with Mayer's haematoxylin (Sigma-Aldrich).

The median (range) purity for CD4⁺ lymphocytes was 94% (94–95%) and for CD8⁺ lymphocytes 87% (85–90%) (n = 3). The viability of CD4⁺ and CD8⁺ lymphocytes and AMC was assessed using exclusion of trypan blue dye. Before stimulation in vitro, the median (range) viability was: CD4⁺ 98% (97–100%), CD8⁺ 99% (99–100%) and AMC 94% (93–94%) (n = 3 for CD4⁺ and CD8⁺; n = 2 for AMC).

Expression of IL-16 mRNA in CD4⁺ lymphocytes

CD4⁺ lymphocytes (1 × 10⁶ cells/well) from buffy coat were cultured as described above and harvested after 5, 10 and 20 h. RNeasy Protect minikit was used according to the manufacturer's protocol, with the addition of DNase I treatment of storage cells and purification of total RNA (Qiagen GmbH, Hilden, Germany). The quantity and purity of the eluted total RNA was measured using a spectrophotometer (SpectraMaxPlus®Molecular Devices Corporation, Sunnyvale, CA, USA) (absorbance 260/280) and was then frozen (−80°C).

The reverse transcription–polymerase chain reaction (RT-PCR) ELISA technique was employed [17] to quantify relative changes in the IL-16 mRNA gene transcript. A one-step RT-PCR was performed with a Gene Amp PCR system 2400 (Perkin Elmer, Wellesley, MA, USA) for amplification. Each RT-PCR of 50 µl was conducted using 10 ng of total cellular RNA and 30 pmol of each primer, 10 U RNase inhibitor (recombinant RNasin, Promega Corporation, Madison, WI, USA), 200 µm PCR digoxigenin (DIG) labelling mix (20 µm dATP, dGTP, dCTP each plus 19 µm dTTP plus 1 µm DIG-dUTP), 5 m m DDT, 10 µl RT-PCR reaction buffer (1·5 m m Mg⁺), 1 µl Titan enzyme mix (AMV reverse transcriptase +Taq DNA polymerase + Pwo DNA polymerase) or 0·5 µl expand high fidelity PCR system (Taq DNA polymerase + Pwo DNA polymerase) for DNA controls (all reagents from Roche, Basel, Switzerland). Reverse transcription was performed (30 min) at 50°C. The annealing temperature for both IL-16 and the housekeeping gene hypoxantine guanine phosphoribol transferase (HPRT) was 55°C. IL-16 and HPRT were amplified (25 cycles) and the DIG-labelled PCR product was stored frozen (−20°C) before use in the detection step. Separate RT-PCRs were performed in duplicate. Gene sequences were accessed from the NCBI database and the accession number used was for human (h) HPRT V00530 and for hIL-16 U82972. Scandinavian Gene Synthesis AB (Köping, Sweden) provided the oligonucleotides.

The gene primer sequences used for RT-PCR and ELISA detection were as follows.

hHPRT

Sense [5′] 5′CGT CGT GAT TAG TGA TGA TGA AC3′; antisense [3′] 5′GCA AAG TCT GCA TTG TTT TGC CA3′; internal probe 5′GAG GCC ATC ACA TTG TAG CCC TCT GTG-3′.

hIL-16

Sense [5′] 5′-TAG AAT CTA CAG CAG AGG CCA-3′; antisense [5′] 5′-TTT GTT CTG AGG CTG CTC CTT-3′; internal probe 5′-GAG AGG CTT GTC TCC GTG TAG GGA G-3′.

The DIG-labelled PCR products were denaturated and hybridized with the biotinylated internal probe designed specifically to hybridize with each gene PCR product, and immobilized on streptavidin-coated microtitre plates (PCR ELISA DIG-detection kit: 3 h, 42°C; Roche). After washing, the bound PCR products were incubated (30 min, 37°C) with an anti-DIG HPRT antibody followed by reaction with the substrate 2,2′-azino-di(3-ethyl benzthiazoline sulphonate) (ABTS). The absorbance (405/492 nm) was measured after 20 min in an ELISA plate reader (Labsystems Multiscan Multisoft, Vanda, Finland). The expression of transcripts for IL-16 mRNA was normalized to the expression of HPRT transcripts and shown as the ratio between the two. In parallel with tested samples, control samples for PCR, probe specificity, DNA contamination, hybridization and sample dilution were run.

IL-16 protein, CD4⁺ and CD8⁺ lymphocytes in palatine tonsils

Human palatine tonsils were obtained during anaesthesia from patients undergoing routine tonsillectomy because of hypertrophy or chronic tonsillitis at the Department of Othorhinolaryngology, Malmö University Hospital. Tonsils from patients with ongoing smoking (at least 3 pack-years) were compared with tonsils from matched, non-smoking subjects. Tonsils were matched for age. The clinical characteristics of the included patients are presented in Table 2. All patients were examined before the tonsillectomy to exclude clinical signs of infection.

Table 2.

Clinical characteristics of six non-smokers and six smokers donating palatine tonsils; tobacco exposure is presented as median (range)

	Non-smokers	Smokers
Numbers	6	6
Age (years)	30 (17–49)	28 (21–55)
Male/female	3/3	3/3
Pack-years	0 (0–4)*	10 (3–15)

^*One of the non-smokers stopped smoking >9 years ago.

After surgery, tonsil specimens were kept cold in Histocon (HistoLab, Göteborg, Sweden) overnight, cut (10 × 10mm), embedded in OTC™ compound (Sakura, Zoeterwoude, the Netherlands) and snap-frozen in liquid nitrogen. Cryostat sections (5 µm) were fixed in acetone followed by inhibition of endogenous peroxidase (see cytospin preparations above).

Staining for immunoreactivity (IR) of IL-16, IL-16 plus CD4+ or IL-16 plus CD8⁺ was conducted in part using an automated immunostainer (TechMate™ Horizon, LJL Biosystems Inc., CA, USA). To block non-specific staining, the tissue slides were first incubated with rabbit serum (10%, 20 min, Dako) followed by single staining of IL-16 or double staining of either IL-16 plus CD4⁺ or IL-16 plus CD8⁺ using monoclonal antibodies (MoAb) during 1 h. The following MoAbs were used: mouse antihuman IL-16 (clone 14·1 for all: 10 µg/ml, Research Diagnostics Inc., Flanders, NJ, USA or 5 µg/ml, BD Bioscience), mouse antihuman CD4⁺ MoAb and mouse antihuman CD8⁺ MoAb (see cytospin preparations above). As controls, species- and isotype-matched immunoglobulins were used (for IL-16, mouse IgG2a; for CD4⁺ and CD8⁺, mouse IgG1, all from BD Bioscience).

The sections were then incubated (45 min) with a peroxidase conjugated rabbit antimouse IgG2a antibody and an alkaline phosphatase conjugated rabbit antimouse IgG1 antibody (1%, Zymed). To permeabilize cell membranes, saponin was included in both Dako ChemMate™ antibody diluent and to buffer kit (Dako) solutions, until the enzyme conjugated antibodies were added.

For visualization of IR, the peroxidase was developed (20 min) by amino ethyl carbazole (AEC) substrate chromogen system (Dako), followed by washing in distilled H₂O. The alkaline phosphatase was developed (30 min) manually in fast blue (40 ml of 100 m m Tris-HCl buffer at pH 8·5) containing Naphtol-AS-MX-phosphate dissolved in N,N-dimethylformamide, fast blue BB salt (8 mg, Sigma-Aldrich) and levamisole solution (20 µl, Vector Laboratories, Inc., Burlingame, CA, USA).

The matched pairs of tonsils were sectioned and stained simultaneously. Two matched pairs of tonsils were read for IL-16IR after incubation with a lower concentration of IL-16 antibody (5 µg/ml), but this did not alter the visual signal, thus confirming that an IL-16 antibody concentration of 5 µg/ml is sufficient. The stained samples were assessed in a blinded manner using light microscopy (Axioplan 2, Carl Zeiss, Jena, Germany) either with a magnification of 20× (phase contrast lens) for the quantitative analysis or a magnification of 20× and 100× for the qualitative analysis.

First, a quantitative IR analysis of the tissue slide samples was performed: IL-16 expression was analysed in samples stained for IL-16 alone, whereas CD4⁺ and CD8⁺ expression was analysed in samples co-stained for IL-16 plus CD4⁺ and IL-16 plus CD8⁺, respectively. The epithelium was identified as a starting point. Beneath the epithelium layer, 10 subsequent areas were analysed. This procedure was repeated twice in a fixed distance from the first area. Thus, in total, up to 30 areas were analysed in each sample. Areas with more than one-third non-lymphoid tissue were excluded. Secondly, a qualitative IR analysis of co-expression of IL-16 plus CD4⁺ and IL-16 plus CD8⁺ was performed on samples co-stained as above. If one or more double-stained lymphocytes was in the slide, i.e. IL-16 plus CD4⁺ or IL-16 plus CD8⁺, the result was counted positive, otherwise negative. An image analysis system was used for reading the signal (KS 400, Kontron Elektronik, Eching, Germany).

Analysis of data

Non-parametric descriptive statistics are presented as median values (with range text and tables) unless stated otherwise. Non-parametric analytical statistics (Wilcoxon's signed rank test and Mann–Whitney) were applied as indicated using the StatView 4·01 (Abacus Concepys, Berkeley, CA, USA) software.

RESULTS

IL-16 protein in lower human airways in vivo

The specific characteristics of all included subjects are presented in Table 1. The BALF IL-16 protein concentration for tobacco smokers (CB plus AS) was significantly higher than in NS (Fig. 1). There was no statistically significant difference between CB and AS in this respect (Fig. 1).

Fig. 1.

Concentration of IL-16 protein in bronchoalveolar lavage fluid from never-smokers (NS), from asymptomatic smokers (AS) and from smokers with chronic bronchitis (CB). For CB alone, filled triangles represent patients with an FEV₁ <80% predicted and open squares represent FEV₁ >80% predicted. Data are shown as individual plus median values (bold lines). *P < 0·05 (Kruskal–Wallis test followed by Mann–Whitney U-test).

Release of IL-16 protein and sIL-2Rα protein in blood CD4⁺ and CD8⁺ lymphocytes in vitro

TSE increased the concentration of IL-16 protein in conditioned medium from CD4⁺(Fig. 2a) and from CD8⁺ lymphocytes (Fig. 2b) cultured in vitro. In neither case was this release potentiated by co-culture with AMC.

Fig. 2.

Concentration of IL-16 protein in conditioned medium from either isolated human CD4⁺ lymphocytes (Fig. 2a) or human CD8⁺ lymphocytes (Fig. 2b) cultured in vitro, with TSE (tobacco smoke extract) or without TSE (negative control) stimulation or CI plus PMA (positive control). Filled symbols represent data from co-cultures with adherent mononuclear cells (AMC). Data are shown as individual values plus median values (bold lines) *P < 0·05(Wilcoxon's signed-rank test).

TSE did not increase the concentration of sIL-2Rα protein in conditioned medium from the CD4⁺ or CD8⁺ lymphocytes in vitro. All values for the negative controls and TSE were below the lowest concentration of the standard curve (i.e. 39 pg/ml), while all the positive controls were detectable. Co-culture with AMC did not alter this pattern.

By the end of the 20-h experiment, the viability of CD4⁺ was 100% (no variability) for vehicle (negative control), for TSE and for PMA + IP-stimulation (positive control) (n = 3). Correspondingly, the viability of CD8⁺ lymphocytes was 100% (no variability), 100% (99–100%) and 98% (96–100%) for vehicle, for TSE and for PMA + IP-stimulation (n = 3). The viability of AMC was 100% (no variability), 100% (no variability) and 86% (76–99%) for vehicle, for TSE and for PMA + IP-stimulation (n = 3).

IL-16 mRNA in blood CD4⁺ lymphocytes

The expression of mRNA in CD4⁺ lymphocytes for IL-16 tended to increase over time at three different time points (5, 10 and 20 h), both with and without stimulation by TSE. TSE did thus not increase IL-16 mRNA markedly at any time-point (Fig. 3).

Fig. 3.

Expression of mRNA for IL-16 normalized to the housekeeping gene (HPRT) in isolated CD4⁺ lymphocytes cultured in vitro, with tobacco smoke extract (TSE, solid line) or without (dashed line). Data represents three different time-points (5, 10 and 20 h).

IL-16IR in palatine tonsils

The specific characteristics of the patients donating tonsils are shown in Table 2. Tonsils from active smokers displayed less intracellular IL-16IR per area than did tonsils from non-smokers (Fig. 4). In contrast, there was no corresponding, pronounced difference in CD4⁺ IR or CD8⁺ IR (data not shown). Interestingly, double staining revealed that CD4⁺ as well as CD8⁺ lymphocytes displayed IL-16IR and this was evident in tonsils from smokers as well as from non-smokers (Fig. 5). Three slides of six from smokers displayed IR for IL-16 plus CD4⁺ compared with five of six from non-smokers. In smokers, four of six slides displayed IR for IL-16 plus CD8⁺ compared with two of six slides in non-smokers.

Fig. 4.

Immunoreactivity for intracellular IL-16 (IL-16IR) in matched human palatine tonsils from non-smokers and from smokers (n = 6 in each group). The data is expressed as IL-16IR in percentage of total tissue area (% area). Data are shown as individual values plus median values (bold lines). *P < 0·05 (Wilcoxon's signed-rank test).

Fig. 5.

Immunohistochemical double-staining from a never-smoking subject showing immunoreactivity (IR) for IL-16 in a CD4⁺ lymphocyte (Fig. 5a) and IL-16 in a CD8⁺ lymphocyte (Fig. 5b). Further, immunohistochemical double staining from a tobacco-smoking subject showing IR for IL-16 in a CD4⁺ lymphocyte (Fig. 5c) and IL-16 in a CD8⁺ lymphocyte (Fig. 5d). The IL-16IR is shown in brown and the CD4⁺ IR as well as the CD8⁺ IR is shown in blue. Arrows indicate the double-staining cells. The magnification factor was 100×.

DISCUSSION

The current study confirms that smokers, with (CB) or without (AS) chronic bronchitis, have an elevated concentration of free, soluble IL-16 protein in BALF, thus forwarding additional clinical evidence that tobacco smoke, in the lower airways, exerts an impact on a cytokine controlling the mobilization of CD4⁺ and, most probably, also reflecting activity in CD8⁺ lymphocytes [8]. Even though both sexes were represented in the current study, we cannot exclude a gender-specific difference, because women dominated the CB group. Furthermore, in line with our previous study in this area [8], it was not possible to link the increase in IL-16 concentration directly to airway symptoms; there was no pronounced difference in IL-16 for patients with chronic bronchitis (CB) and asymptomatic smokers (AS). To some extent these observations contrast with the experimental data, suggesting that IL-16 contributes to non-specific airway hyperreactivity in allergic mice [6], even though this end-point was not assessed, and therefore such an impact cannot therefore be ruled out based upon the current study.

Interestingly, there is clinical evidence that IL-16 may exert an inhibitory effect on CD4⁺ lymphocytes, in addition to being a chemoattractant for these key regulatory immune cells. Thus, we demonstrated recently a negative correlation between IL-16 and sIL-2Rα in BAL fluid in a study on lung allograft recipients, in which patients with acute allograft rejection also displayed a lack of increase in airway IL-16 when compared with matched control patients [18]. Others have demonstrated that IL-16 exerts an inhibitory effect on the release of interferon (IFN)-γ, IL-1β and tumour necrosis factor (TNF)-α in a murine in vivo model of rheumatoid synovitis [19]. Thus, IL-16 may act both as a pro- and an anti-inflammatory mediator; it should therefore be pointed out that the currently demonstrated increase in local IL-16 need not be detrimental only.

The current study also shows that the water-soluble extract from tobacco smoke (TSE) does increase the release of free, soluble IL-16 protein, but not soluble IL-2Rα protein, in isolated human blood CD4⁺ and CD8⁺ lymphocytes in vitro; forwarding a plausible and seemingly selective cellular mechanism that may explain the impact tobacco smoke exerts on local IL-16 in the lower airways of tobacco smokers. The IL-16 response tended to be somewhat higher in CD4⁺ than in CD8⁺ lymphocytes in vitro, but this difference was not statistically significant and there was a clear response in both cell types, pointing out the CD4⁺ lymphocyte itself as a potential source for IL-16. This novel mechanism in blood T lymphocytes clearly deserves to be evaluated further in T lymphocytes from the lower airways in future studies.

Unexpectedly, the current study indicates that the presence of co-stimulatory ‘macrophage-like’ cells (AMC) do not potentiate the TSE-induced release of IL-16 protein in CD4⁺ or CD8⁺ lymphocytes. It is noteworthy this was observed even though these macrophage-like cells displayed very high viability. In line with this failure of macrophage-like cells to co-stimulate T lymphocytes, a previous study did show that tobacco smoke extract inhibits the release of TNF-α in murine alveolar macrophages, without any substantial decrease in viability [20]. It can thus be speculated that exposure to tobacco smoke selectively inhibits macrophage secretion of cytokines normally facilitating CD4⁺ or CD8⁺ lymphocyte activation without causing massive cell death, even though it is uncertain whether the method used in our study (i.e. trypan blue exclusion) discriminates early signs of apoptosis. It can also be added that the lipid soluble components of tobacco smoke were not included in our current in vitro experiments, and the putative effects of these therefore remain unknown.

We also evaluated whether de novo synthesis of IL-16 protein is affected by the water-soluble components of tobacco smoke in the current study. However, the referred water-soluble components did not alter the level of IL-16 mRNA markedly in CD4⁺ lymphocytes at either of the three time-points investigated. Thus, because the water-soluble components of tobacco smoke do not alter the inherent IL-16 production, these components probably cause mainly the release of preformed IL-16 protein. This type of mechanism would be in line with a previous study, where secretion of bioactive IL-16 protein occurred in peripheral blood CD4⁺ lymphocytes without a corresponding increase in IL-16 mRNA, after stimulation with anti-CD3 [21].

When we analysed the immunoreactivity for intracellular IL-16 protein in palatine tonsils from smokers and non-smokers, we found that tonsils from tobacco smokers display less immunoreactivity for IL-16 than do tonsils from non-smokers. This suggests that tobacco smoke does exert an impact on a type of lymphoid tissue that is not involved directly in the drainage of the lower airways [22]. We believe that this decrease in IL-16 immunoreactivity among smokers can be explained in at least two ways: either there is a subset of T lymphocytes becoming depleted of IL-16 or the subset expressing IL-16 migrates away from lymphoid tissue, the latter possibly migrating to the target organ (i.e. the lower airways).

Finally, our current study provides evidence that certain CD8⁺ lymphocytes, just as CD4⁺ lymphocytes, do express intracellular IL-16 protein in lymphoid tissue, in smokers as well as in non-smokers. This observation thus expands on the previous idea that CD4⁺ lymphocytes would constitute a major source of IL-16 protein in human palatine tonsils [23].

In conclusion, this study on human material suggests that tobacco smoke exerts a wide impact on the expression of IL-16, a cytokine believed to be important in regulating the recruitment and activity of CD4⁺ lymphocytes. Our study forwards the possibility that a local increase in IL-16 protein in the lower airways is due to a direct stimulatory effect by the water-soluble components of tobacco smoke on CD4⁺ and CD8⁺ lymphocytes and that the lymphocytes in lymphoid tissue are also affected. New and interventional studies on T lymphocytes from lower airways are motivated to evaluate whether targeting IL-16 is beneficial for limiting local inflammation caused by tobacco smoke.