Clinical and Experimental Pharmacology and Physioiogy (1992) 19, 509-5 15

PLATELET-IQCTIVATINGFACTOR IN BRONCHOALVEOLAR LAV,AGE FLUID OF PATIENTS WITH ADULT RESPIRATORY DISTRESS SYNDROME Kohei Matsumoto, Fumio Taki, Yasuhiro Kondoh, *Hiroyuki Taniguchi and Kenzo Takagi Second Department of Internal Medicine, Nagoya University School of Medicine, Tsuruma-cho, Showa-ku, Nagoya and *Department of Respiratory Medicine, TosleiGeneral Hospital, Nishioiwake-cho, Seto, Aichi, Japan (Received 12 November 1991; revision received 12 February 1992)

SUMMARY 1. To clarify the role of platelet-activating factor (PAF) in the development of adult respiratory distress syndrome (ARDS), we performed bronchoalveolar lavage (BAL) in 19 patients with ARDS and examined cell populations, albumin concentrations and PAF levels. PAF levels were measured by a newly developed radioimmunoassay. 2. In the BAL fluid of ARDS patients, neutrophil percentages and albumin concentrations markedly increased compared with control subjects. 3. PAF was detected in 14 of 19 patients with ARDS, whereas it did not exist in the control subjects. 4. Furthermore, we investigated the priming effect of recombinant human tumour necrosis factor-a (TNF,), which is known to be one of the most important mediators in the development of ARDS, on PAF production induced by the calcium ionophore in neutrophils. 5. Pre-incubation with TNF, dose-dependently increased both extracellular and intracellular PAF production in neutrophils. 6. These results suggest that P A F might play an important role in the development of ARDS.

Key words: adult respiratory distress syndrome, bronchoalveolar lavage, neutrophils, platelet activating factor.

INTRODUCTION Adult respiratory distress syndrome (ARDS), which is characterized by severe systemic hypoxemia and pulmonary oedema, results in high mortality and morbidity. Increased permeability as a result of diffuse damage to pulmonary microvasculature is generally considered to be involved in the pathogenesis of ARDS. There is increasing evidence for a role of cellular elements, such as neutrophils and platelets, and humoral mediators, such as complement, prosta-

glandins, leucotrienes, thromboxanes, interleukin-I and tumour necrosis factor (TNF) (Spragg & Smith 1991). Platelet activating factor (PAF) is a potent acetylated alkyl phosphoglyceride autacoid released by a variety of activated cells, including platelets, neutrophils, basophils, alveolar macrophages, renal mesangial cells and cultured endothelial cells (Benveniste & Vargaftig 1983). PAF has an ability to enhance neutro-

Correspondence: Kohei Matsumoto, Second Department of Internal Medicine, Nagoya University School of Medicine, Tsuruma-cho, Showa-ku, Nagoya 466, Japan.

5 10

K. Matsumoto et al.

phi1 functions such as chemotaxis, aggregation, degranulation, adherence and the production of leucotriene C4, D4 and superoxide anion (McManus & Deavers 1989). Exogenous PAF causes vasoconstriction (Hamasaki et al. 1984) and bronchoconstriction (Vargaftig et al. 1980). Pathological studies have documented pulmonary infiltration of neutrophils and platelets after PAF injection in rabbits (Lewis ez al. 1983). Moreover the infusion of PAF into rabbits induced neutrophil accumulation in bronchoalveolar ravage (BAL) fluid and increased microvascular permeability (Worthen et al. 1983). PAF can also initiate platelet aggregation and the production of thromboxane A2, and it increases the procoagulant activity, and induces microembolism in pulmonary vasculature, which is considered to be one of the major aetiologic factors in the development of lung injury (McManus & Deavers 1989). Despite data establishing that PAF can induce these pathophysiologic changes associated with acute lung injury, it is uncertain at present whether PAF actually serves as a mediator of lung injury in the clinical condition of ARDS, because no accurate method of assay has been available for the quantitation of PAF. Recently a novel, facile and sensitive radioimmunoassay has been developed. The present study measured PAF extracted from BAL fluid of patients with ARDS using the Du Pont NEN PAF radioimmunoassay kit. Furthermore, the priming effect of recombinant

TNFu, which is known as a major factor in the pathogenesis of ARDS (Marks et al. 1990), on PAF production in neutrophils was also investigated.

METHODS Recombinant human TNF, was supplied by Asahi Chemical Industry Co. Ltd (Tokyo, Japan); "PAF [1*5I] radioimmunoassay kit was purchased from E. J. Du Pont de Nemours & Co. Inc. (Boston, MA, USA); Bond-Elut cl8 columns were from Analytichem International (Harbour City, CAYUSA). Other chemicals were obtained from Sigma Chemical Co. (St Louis, MO, USA).

Clinical study Patient selection Nineteen patients with ARDS identified in the intensive care units of the Nagoya University Hospital and the Tosei General Hospital were studied. Diagnostic criteria for ARDS included the following: (i) lung injury (LI) score greater than 2.5, as described by Murray et al. (1988); (ii) excluding exacerbations of chronic disease or left ventricular failure. The characteristics of ARDS subjects are shown in Table 1. Normal volunteers (n = 8, mean age 3 3 f 6 , four males and four females) free of lung disease and taking no medication were studied as the control POUP.

Table 1. Clinical characteristics in 19 patients with ARDS

Patient

Age (vears'l

Sex

Trigger

56 73

M M M F

Sepsis Sepsis Aspiration Aspiration Bacterial pneumonia Fat embolism Fat embolism Fat embolism Radiation Radiation Leukaemia Radiation Radiation Bacterial pneumonia Unknown Unknown Unknown Unknown Unknown

~~

~

1

2 3

74

4 5

58 69

6 7 8 9 10

19 16 20

M

14

78 63 66 66 13

M F M M M F M F M

15 16 17 18

20 62 70 12

F M M M

19

59

M

11 12 13

LI Outcome score

58

3.5 3.8 3.5 2.8 3.0

Expired Expired Expired Survived Survived

4.0 2.8 2.8 3.5 4.0 3.8 3.0 4.0 4.0

Survived Survived Survived Expired Expired Expired Survived Survived Survived

2.5 3.0 4.0 4.0 3.5

Survived Survived Expired Expired ExDired

Bronchoalveolar lavage After formal consent was obtained, BAL was performed with a flexible fibreoptic bronchoscope in patients with ARDS and in the control subjects. All patients with ARDS underwent BAL within 24 h of satisfying the criteria for ARDS. The distal end of the bronchoscope was wedged in a single segment or subsegment of the right middle lobe. Three 50 mL aliquots of sterile 0.9% NaCl were instilled through the suction port and returned by gentle aspiration into the injecting syringe. The lavage procedure in the d in a similar fashion. uid were cultured for bacteria The BAL fluid was passed through sterile surgical gauze, and immediately centrifuged at 500 g for 10 min at 4OC to separate the cellular and non-cellular components.

Cell counts and protein analysis in BAL fluid The number of cells was determined by washing the recovered cellular pellet with Hank's Balanced Salt

ARDS and PAF

Solution, and by counting with a haemacytometer. Absolute numbers of cells were expressed as cells X lO4/ mL unconcentrated BAL fluid. The cell analysis was performed by direct counting using Kovhcs' staining fluid (Kovlcs et al. 1961). By this method, basophils, eosinophils, neutrophils and mononuclear leucocytes, including lymphocytes and monocytes, can be recognized easily and identified simultaneously. Albumin concentrations in the supernatants of BAL fluid were determined according to the method of Mancini et al. (1965). The supernatants of BAL fluid were mixed with equal volumes of 20% acetic acid to inactivate acetylhydrolase activity, and stored at 4" C.

Measurement of PAP in BALf2uid The Du Pont NEN PAF [l25I] radioimmunoassay kit was used to measure PAF extracted from the samples (Smal el al. 1990). This PAF radioimmunoassay kit consists of PAF primary antibody, secondary antibody (Smal et al. 1989), [*251]-PAFtracer concentrate, PAF standard concentrate, 4 X assay buffer and [3H]PAF recovery tracer. Acetic acid-treated samples (2 mL) were applied to the pre-treated Cle columns. The columns were washed with two 1 mL aliquots of 10% acetic acid. Three aliquots (2 mL each) of ethyl acetate were applied to the column. PAF was eluted from the column by applying 6 mL of methanol. The elute was collected in a siliconized glass tube. The DEAE cellulose (100 mg) was preweighed into the methanol collection tube. Further chloroform and water were added to separate aqueous and chloroform phase (final ratio of chloroform/ methanol/ water is 1 : 1 :0.9). After centrifugation (100 g for 10 min) the chloroform layer was collected. The solution was evaporated to dryness under a gentle stream of nitrogen and reconstituted with 1 mL of working assay buffer (0.1% sodium azide and 0.05% Tween 20 in 50 mmol/L sodium citrate buffer, pH 6.3). A PAF standard concentrate (6 pg/mL in 1 : 1 ethanol/ water) was diluted in working assay buffer to give a standard solution within the range of 30-3000 pg/O. 1 mL. ['*5I]-PAF tracer concentrate was diluted 1:40 (v/v) with secondary antibody. One hundred microlitres of each sample or diluted standard was added to appropriate tubes. PAF primary antibody (100 pL) was added to all tubes, except total counts and blanks, and incubated at room temperature for 15 min. Secondary antibody/ tracer solution (100 pL) was added to all tubes and incubated at room temperature for 24 h. After the incubation, 2 mL of working assay buffer was added to all tubes except total counts. The tubes except total counts were centrifuged at 2000 g for 30 min. The supernatants of tubes

511

were decanted and allowed to drain on absorbent paper for I min. Tubes were counted in a gamma counter. After counting had been completed, the concentration of PAF in the samples was determined from a standard curve (Fig. 1). The normalized per cent bound (%B/Bo) for each standard and sample was as follows:

%B'Bo =

Net ct/min of standard or sample x 100 Net ct/min of '0' standard

Using semilogarithmic graph paper, %B/ Bo for each standard vs the corresponding concentration of PAF added was plotted. The concentration obtained was corrected for dilution, concentration and recovery to determine the original concentration in the sample. The detection limit of this assay was 30 pg/O. 1 mL.

Experimental study Human neutrophil suspensions were prepared from healthy adult male volunteers by dextran sedimentation and the Percoll method with slight modifications. In brief, after dextran sedimentation, neutrophils were suspended in calcium-free HEPES-buffered saline containing 17 mmol/L HEPES, 120 mmol/L NaC1, 5 mmol/L KCI, 5 mrnol/L glucose and 0.3% bovine serum albumin, and they were layered onto a discontinuous Percoll density gradient (densities of 1.10,

100-

-s? 8 0 v

z

A

5! z

60-

I

*

4020 U

20

50 1 0 0 2 0 0

500

2000

PAF (pgl0.l rnL) Fig. 1. A typical standard curve obtained for the PAF

radioimmunoassay.

K. Matsumoto et al.

512

1.08 and 1.04). Cells at the interface between 1.10 and 1.08 were collected and washed twice. Cell analysis was performed by using the same method described above. The population of cell types was more than 98% neutrophils, and the trypan blue exclusion test indicated that more than 99% of the cells were viable. The effect of recombinant TNFa on the extracellular and intracellular PAF production was assessed. Neutrophils (1 X 107) in 2 mL MEM/20 mmol/L HEPES were incubated with 100 pL of TNF, (final concentration 100 U/mL, 1000 U/mL) or vehicle for 15 min. At the end of the incubation period, neutrophils were stimulated with 2.5 pmol/L Ca ionophore A 23187 for 15 min. The reaction was stopped by rapid cooling. After centrifugation, cell pellets and supernatants were separated. The supernatants were mixed with equal volumes of 20% acetic acid and stored at 4OC. The pellets mixed with 2 mL of methanol were placed at room temperature for 24 h and stored at 4OC. PAF in the supernatant was measured by using the same method described above. Cell extraction can be concentrated and applied to thin layer chromatography (TLC) with chloroform/ methanol/ water (65 :35 :6) as a developing solvent. The silica area containing PAF was scraped off, extracted with chloroform/ methanol/ water (1 :2:0.8 v/v) and then phased by adding the appropriate volume of chloroform and water to a final ratio of 1:1:0.9 v/v. The chloroform layer was evaporated to dryness and reconstituted with assay buffer.

Statistical analysis In the clinical study, statistical analyses of neutrophil percentages and albumin concentrations between the ARDS group and the control group were performed using the Mann-Whitney U test. Correlations between variables were achieved using the Pearson correlation. In the experimental study, statistical analyses were performed using the analysis of variance. Statistically, P values of 0.05 or less were considered to be significant. All data in this paper are expressed as mean f s.e.m.

RESULTS The mean volume of recovered BAL fluid of the ARDS patients was 94.7f6.6 mL and that of the control subjects was 118.4f 6.5 mL. The mean total cell number in the ARDS patients was 82.7 f 12.8X l O 4 / mL, and that in the control subjects was 11.4 f 3.0 X lO4/mL. The mean neutrophil percentage in BAL fluid of the ARDS patients (48.0f6.496) was significantly different from that of the control subjects (0.6f0.2%; P