T Lymphocytes and Eosinophils in Allergen-induced Late-phase Asthmatic Reactions in the Guinea Pig1 - 3

ANTHONY J. FREW, REDWAN MOQBEl, MAY AZZAWI, ADELE HARTNEll, JULIA BARKANS, PETER K. JEFFERY, A. BARRY KAY, RIK J. SCHEPER, JOHN VARLEY, MARTIN K. CHURCH, and STEPHEN T. HOLGATE

Introduction

The involvement of T lymphocytes in the pathogenesis of allergic tissue reactions is an area of considerable contemporary interest. It has long been known that T lymphocytes and their soluble mediators (lymphokines) control the production of IgE by B lymphocytes (1). More recently it has become clear that T cells can have direct proinflammatory roles and that T cell lymphokines are capable of attracting and activating nonspecific effector cells, such as monocytes, eosinophils, and neutrophils (2-4). We have recently provided evidence that CD4+ T lymphocytes infiltrate allergeninduced late-phase skin reactions in man, and have shown that there is a close relationship between the numbers of CD4 + T cells and activated eosinophils in such reactions (5). In the human airway, studies of allergen-induced T lymphocyte infiltration have been confined to indirect methods enumerating T cells in bronchoalveolar lavage (BAL) (6-8) or in peripheral blood (9). These studies were inevitably constrained in terms of the nature and number of samples that could be obtained. Nevertheless, they have provided evidence of a selective increase in CD4 + T lymphocytes in BAL accompanying late asthmatic reactions, with a complementary fall in the number of CD4 + T cells in the peripheral blood. Though these studies provide valuable insights into the cellular component of allergic airways inflammation, they are at best indirect measures of the processes occurring in the tissues. Several groups have developed animal models of allergen-induced early- and late-phase bronchoconstriction in which the cellular mechanisms of asthma can be studied. We have recently described a guinea pig model in which sensitization of the airways with inhaled ovalbumin leads to hyperresponsiveness (10) and

SUMMARY The kinetics and phenotype of T lymphocytes infiltrating the airways of guinea pigs undergoing late-phase asthmatic reactions (LAR) were studied with monoclonal antibodies, cytofluorimetry, and immunocytochemistry. Challenge of sensitized animals with aerosolized ovalbumin was followed by early (2 h) and late-phase (17 h) bronchoconstriction. The induction of hyp~r­ sensitivity, by aerosolized antigen, was associated with an increase in mucosal T cell numbers, which consisted almost entirely of CD8+ T cells. Following allergen challenge of fUlly sensitized animals, a biphasic rise in total T cell (CD3 +) numbers was observed in the bronchial mucosa, peaking at 17and 48 h. A similar pattern of T cell accumulation was observed in the bronchial adventitia but with an extra early peak at 2 h. In contrast to the T cell influx of the sensitization phase, the postchallenge infiltrate consisted largely of CD3 + ,CD8- cells. Eosinophil numbers were elevated in both submucosa and adventitia, with a single broad peak between 17and 48 h. T cell infiltration was compared with eosinophil accumulation: while correlations between T cell and eosinophil numbers varied over the 96 h of the experiment, strong associations were observed between CD8+ numbers and eoslnophils in the adventitia at 6 h (r 0.733, P < 0.01) and between CD3+ numbers and eoslnophlls in the submucosa at 72 h (r = 0.88, P < 0.001). No significant changes were detected in T cell or eosinophil numbers in the lung parenchyma. There was a postchallenge increase in eosinophils (but not T cells) in bronchoalveolar lavage (BAL). In contrast, analysis of blood leukocytes showed no changes in T cell total or subset numbers during the progression of the LAR. These results indicate that accumulation of T cells in the airways is a feature of the LAR in this model. These observations are consistent with human BALand skin studies of the allergen-induced late-phase reaction. These data also Illustrate that evaluation of distant compartments (blood and BAL) may not fully reflect rapidly evolving tissue infiltration in the bronchial mucosa.

=

AM REV RESPIR DIS 1990; 141:407-413

further challenge with aerosolized anti- ate, Levamisole, dimethylformamide, naphgen, under the protection of an Hl- thol AS-MX phosphate, DNase I, Fast Red receptor antagonist, results in early- and TR salt werepurchased from Sigma Chemical late-phase bronchoconstriction (11). The Co (Poole, Dorset, UK); RPMI-1640 tissue late phase of airways obstruction is ac- culture medium from Gibco (Paisley, Scotland); sodium pentobarbitone (60 mg/ml) companied by an influx of eosinophils into the airway wall and lumen, but the mechanisms responsible for their recruitment have not yet been elucidated (12). (Received in original form February 24, 1989 and In the present study, we have used this in revised form June 30, 1989) model to examine the kinetics of T lymFrom the Department of Allergy and Clinical phocyte and eosinophil accumulation Immunology and the Department of Lung Patholand migration in the bronchial mucosa ogy, National Heart and Lung Institute, Brompton following allergen challenge and have Hospital, London; the Pathological Institute, Free compared these responses with parallel University Hospital, Amsterdam, Netherland; and the Immunopharmacology Group, Southampton observations in blood and in BAL. General Hospital, Southampton, United Kingdom. I

Methods Chemicals Unless otherwise stated, all chemicals were purchased from BDH Ltd (Dagenham, Essex, UK). Bovine serum albumin (BSA)(grade VI), ovalbumin (grade III), pyrilamine male-

2 Supported by grants from the Wellcome Trust, the Asthma Research Council, and the Medical Research Council of Great Britain. 3 Address correspondence and reprint requests to A. B. Kay, Professor and Director, Department of Allergy and Clinical Immunology, National Heart and Lung Institute, Dovehouse Street, London, SW3 6LY, UK.

407

408

FREW, MOQBEL, AZZAWI, ET AL.

from May & Baker (Dagenham, Essex, UK); OCT cryoembedding medium from Raymond Lamb (London, England).

at Southampton University for I wk before sensitization.

Prepared Solutions

Measurements of airways mechanics, sensitization with aerosolized ovalbumin and allergen challenge were performed as previously described (11). Briefly, animals were placed in a clear perspex box and sensitized by exposure to aerosolized ovalbumin (1% wt/vol in 0.9% sodium chloride), administered via a Wright's jet nebulizer driven by compressed air at 8 L/min, for 3 min on 2 occasions separated by 7 days. Seven days after the second sensitizing exposure, animals were premedicated with pyrilamine maleate (10mg base/kg given by intraperitoneal injection, 30 min before challenge) and challenged with either aerosolized ovalbumin or isotonic saline diluent for 5 min. Airways resistance was measured in conscious animals by whole body plethysmography at intervals after challenge (2, 6, 17, 24, 48, 72, and 96 h). Animals were then killed by intraperitoneal injection of pentobarbitone (60 mg/kg) to obtain samples for immunopathology. Sixteen animals were studied at each time point (8 challenged with allergen and 8 with saline), except at 96 h, when 8 animals were studied (4 with allergen and 4 with saline). Three groups of unchallenged control animals were studied: immunologically naive animals (Pre I), animals sensitized once and killed after 7 days (Pre 2), and fully sensitized animals killed 7 days after the second sensitization exposure (Pre 3).

Sensitization and Challenge Procedure Tris-buffered saline (TBS): 0.05 M Tris-HCI buffered isotonic saline (pH 7.6), was prepared fresh daily from stock solutions. Lysis buffer. A total of 8.2 g ammonium chloride and 1.0 g potassium hydrogen carbonate was dissolved in I L of distilled water and adjusted to pH 7.3 Stains. Kimura stain was prepared fresh from stock solutions and filtered before use (13). Chromotrope 2R stain. An aqueous solution was prepared containing 0.50/0 phenol and 1% carbol chromotrope 2R. Alkaline phosphatase substrate. Naphthol AS-MX phosphate (2 mg) was dissolved in 200 ul dimethylformimide and then diluted in 9.8 ml 0.1 M tris-methylamine-HCl buffer pH 8.2. Ten microliters of I M levamisole was added to inhibit endogenous neutrophil alkaline phosphatase. Immediately before incubation of sections, 10 mg Fast Red TR salt was added to the substrate mixture, which was then filtered before use. Kaiser's mounting medium. Ten grams of gelatin and 0.25 g phenol was dissolved in 70 ml distilled water to which 60 ml glycerol was then added. The resultant mixture melts at 70° C, and sets at approximately 50° C.

Monoclonal antibodies to guinea pig T cell specificities. The murine monoclonal antibodies (mAb) CT7 and CT6 were prepared as ascites at the Free University Amsterdam. CT7 recognizes the guinea pig CD3 molecule present on all mature T cells whereas CT6 recognizes the guinea pig homologue of the CD8 molecule, present on the "Tcytotoxic/suppressor" cell subset (14). MSgpl2 (originally known as T4Bll) was the gift of Dr. D. Healey and Professor J. Turk (Royal College of Surgeons, London). MSgpl2 (T4B11) binds to cortical thymocytes and to a T cell subset that provides help for antibody production (15), but which probably does not identify the guinea pig CD4 molecule (which defines the classical "T-helper/inducer" subset), Optimum dilutions of mAb were determined in preliminary experiments using peripheral blood mononuclear cells and cryostat sections of mesenteric lymph nodes and lungs obtained from normal (i.e., immunologically naive) guinea pigs. Immunochemicals. Normal rabbit immunoglobulin (catalog # X903), rabbit antimouse immunoglobulin (catalog # Z256), FITC conjugated Ftab'), rabbit anti-mouse immunoglobulin (catalog # F313), and monoclonal APAAP reagent (catalog # D651) were purchased from Dakopatts (High Wycombe, England).

Guinea Pigs Specific pathogen-free (SPF) adult male outbred Dunkin Hartley guinea pigs (weight 500 to 550 g) were purchased from Porcellus (Sussex, UK) and maintained in clean conditions

Blood Leukocyte Preparation Blood was collected by cardiac puncture. A sample was taken into EDTA for differential counts using Kimura's stain. Mononuclear cells were separated by mixing heparinized blood with an equal volume of RPMI-1640, layering onto Ficoll-Paque (Pharmacia A.B., Uppsala, Sweden) and centrifuging at 1,200 g for 20 min. The cells remaining at the interface were washed twice in RPMI-1640 containing 20 ug/rnl DNase I and resuspended in RPMI-I640 containing 0.5% BSA at 2 x 106 cells/ml.

Bronchoalveolar Lavage Bronchoalveolar lavage (BAL) was performed by gently instilling 5 ml of RPMI-I640 into the trachea via a flexible plastic cannula (Butterfly-21, Abbott, Sligo, Ireland) and aspirating back into the syringe. This procedure was then repeated with a fresh 5-ml aliquot and the two aspirates were combined (return volume was consistently 7.5 to 8 ml). The BAL fluid was centrifuged at 400 g for 10 min and the supernatant was then removed and stored at - 80° C until assayed. The cell pellet was resuspended in lysis buffer to remove red cell contamination, washed twice in RPMI-I640 containing 20 ug/ml DNase I, and finally resuspended at 2 x 106 cells/ml in RPMI-I640 containing 0.5% BSA.

Lung Tissue The thoracic cavity was opened, and the lungs and heart removed en bloc. The trachea and the left lower lobe were dissected out, embedded in OCT medium, and snap-frozen ~I) - 80° C in isopentane, suspended in a baui of liquid nitrogen. Frozen sections of the lung hilum were cut (8 urn thick), air-dried, and then fixed in acetone for 5 min. Fixed slides were air-dried, wrapped in aluminum foil, and stored at - 30° C until stained.

Cytofluorimetric Analysis (FACS) Aliquots of cells (4 x 105 cells in 200 ul) were incubated with appropriately diluted mAbs for 15 min, washed twice in RPMI-1640, and then incubated for 15 min with FITC-conjugated F(ab12 rabbit anti-mouse immunoglobulin. After two further washes, the cells were fixed in 1% paraformaldehyde, and finally resuspended in phosphate-buffered saline containing 0.1% sodium azide. Fixed samples were stored in the dark at 4° C for 3 to 10 days before analysis; cells prepared in this way can be stored for up to 3 wk without significant loss of fluorescence. Samples were run on a FACS analyzer (Becton-Dickinson, Cowley, Oxford); an electronic gate was set on the lymphocyte peak and positively stained cells falling within the gated area were enumerated according to standard criteria. Background staining was assessed by omission of the primary antibody and averaged < I% for blood lymphocytes and < 2.5% for BAL lymphocytes.

Tissue Staining Methods Monoclonal antibody staining was detected by a modified alkaline phosphatase anti-alkaline phosphatase (APAAP) method, as previously described (5). System and specificity controls were included in each staining run, using guinea pig lymph node, and mouse IgGI and IgG2a myeloma proteins as a negative control. Tissue eosinophils were stained by Chromotrope 2R. Results

Airways Mechanics Airways resistance (Raw) was calculated as previously described (11) with modifications to comply with the method of Dubois and' coworkers (16). Vtg = i\Pbox(c) x (a - w) b x i\Pm(c) Raw

= b x .i\PA - d (em H

Raw

(1)

20·sec·L-l)

0)

V sGaw

c (ml)

1

x Vtg

(em H 2 0 · s t 1 (3)

i\Pa = i\Pm(e) x i\Pbox(0) , i\Pbox(c)

(4)

where V = flow at mouth (Lis), PA = alveolar pressure (ern H 20), Pm = mouth pressure (ern H 20), Pbox = box

409

T CELLS, EOSINOPHILS, AND THE GUINEA PIG LAR

pressure (em H 20), ~ = change in parameter, (0) = airways/shutter open, (c) = airways/shutter closed, a = box volume in ml (1,320ml), b = correction factor for differences between ambient and animal core temperature (1.045), c = dead space of apparatus in ml (4.5 ml), d = apparatus resistance (37 em H 2 0 ·L-l·s), and w = weight of animal in g. Before challenge and at each time point after challenge, airways resistance (Raw) was calculated as the mean of four measurements. Postchallenge Raw values were expressed as percentage changes from prechallenge values. Comparisons were made between ovalbumin and saline challenge responses and statistical significance assessed by Student's t test. (17).

Analysis: Immunohistology For each lung, at least two sections were stained with each antibody. Slides were counted blind using a Zeiss Universal II microscope at 200 x magnification and an eyepiece graticule. Positively stained cells were enumerated in three anatomical zones (figure 1). (l) Bronchialsubmucosa. Large airways with cartilage rings were studied. Positive cells were counted between the basement membrane and the muscle layer. Counts were divided by the external perimeter of the bronchial muscle layer and expressed as cells per unit muscle perimeter (l U being approximately 0.5 urn). (2) Bronchial adventitia. The same airways were assessed as for submucosa. Positive cells were enumerated in a cuff 50 urn deep, immediately adjacent to and outside the bronchial smooth muscle. Counts were again expressed as cells per unit muscle perimeter.

.. . .

Fig. 1. Anatomical areas studied in enumeration of cellular infiltrates. (1)Submucosa, between basement membrane and smooth muscle layer; (2) adventitia, cuff onegrid-square wide around smooth muscle layer; (3) parenchyma, random fields well away from bronchi. All counts performed at x200 magnification. Area of single grid square 2,025 J,tm 2 •

140

Fig. 2. Airways resistance following sensitization with ovalbumin, and after challengewith ovalbumin (0) or saline diluent (.). Pre 1,Pre 2, and Pre 3 values expressed as % of Pre 3 values; postchallenge values as % change from prechallenge resistance. Mean ± SEM of 8 animals for Pre 1, 2, and 3; 27 allergen-challenged and 27 salinechallenged animals for postchallenge points. (* = p < 0.05.)



W w

O+---±---+--+---=----+-.=F--+""::>O"=------l~

Co

o

LJ

+

rn

o

LJ

-5

-.-- ----r- --,--Pre1 Pre 2 Pre 3

I

I

I

I

i

6

17

24

48

72

Time (hours)

a correlation with eosinophils in the adventitiar = 0.733 (p