Role of Pulmonary Inflammation in Altered Airway Responsiveness in Three Sheep Models of Acute Lung Injury1-3
JEAN F. SIMPSON, MARTHA J. BUTTERFIELD,4 PETER L. LEFFERTS, ERIC L. DYER, JAMES R. SNAPPER, and BARBARA MEYRICK
One of the currently more popular theories for the increased airway responsiveness to bronchial provocation associated with the human disease of asthma is that inflammatory cellssequestered within the lungs modulate airway responses to bronchial provocation (1-5). Although conflicting data exist (6, 7), increased airway responsiveness to bronchial challenge is seen in some normal subjects and asthmatics after upper respiratory tract infections (1) as well as in survivors of diffuse lung injury as is observed in the adult respiratory distress syndrome (ARDS) (8, 9). Experimental support for the role of pulmonary inflammation in altered airway responsivenesscomes from a variety of species including humans (1, 8, 9), dogs (3), rabbits (2), and, after endotoxemia, sheep (4, 5, 10). Infusions of Escherichia coli endotoxin, phorbol myristate acetate (PMA), or zymosan-activated plasma (ZAP) all cause pulmonary inflammation and qualitatively similar alterations in lung mechanics, pulmonary hemodynamics and lung fluid and solute exchange (4, 10-18). However, these same interventions have strikingly different effects on pulmonary responsiveness to aerosol histamine (figure 1).Airway responsiveness is strikingly increased after endotoxemia (4), whereas PMA causes only moderate increases in airway responsiveness to aerosol histamine (16). Repetitive bolus injections of ZAP, on the other hand, do not increase pulmonary responsiveness to aerosol histamine (18). The differences between the effects of endotoxin, PMA, and ZAP on pulmonary responsiveness to aerosol histamine could result from differences in the severity, location, or type of the inflammatory response induced by the three agents. We therefore undertook a series of experiments to examine the effects of en-
SUMMARY Pulmonary Inflammation may contribute to Increased airway responsiveness In ex· perlmental models of acute lung InjUry. Infusions of endotoxin, phorbol myrlstate acetata (PMA), or zymoSln-sctlvsted plasms (ZAP) all result In the sccumulatlon of polymorphonuclear leukocytes (PMNs) In the lung and alterations In lung mechanics. These threa Interventions have strikingly different effects on airway responses to aerosol histamine: ZAP does not Increase airway responslvsness, whereas endotoxin causes a greater Increase In airway responsiveness than does PMA. The present histologic study examines the question of whether the pattern and severity of PMN and mast cell accumulation In large· and medium-sized airways and lung periphery could contribute to the differences In airway responsiveness to histamine. Minimally Instrumented sheep WIre given either an Infusion of endotoxin (0.5 llg/lcg over 20 min), a bolus Injecllon of PMA (5 llgllcg), or repetl· tlve boluses of autologous ZAP (5 ml). Four and a half hours later, the animals WIre killed, and the left lung was removed snd fixed In the dlstanded state. Threa lavels of the left lung WIre examined by light microscopy: the large hllar bronchus, a medium-sized bronchus, snd parlpherallung. The number of PMNs snd mast cells In the airway WIll WIre expressed as cellalmm length of airway circumference and In the lung parlphery as cellsl100 alveolar profiles. Both endotoxin and PMA caused a significant 2· to 3-fold Increaee In number of PMNalmm of large airway circumference, the majority of PMNs being In the blood vessels of the lamina propria and submucoss; ZAP caused only minimal PMN accumulation In the blood vessels of the submucoSl. Similarly, endotoxin and PMA caused at least a 3-fold Increase In number of PMNsIn the submucosal blood vessels of mlddlesized bronchi and a 2· to 3-fold Increase In their numberl100 alveolar profiles. Mast cell numbers WIre unchanged by any of the Interventions. The £Ignlflcant Increases In PMNs after endotoxin, PMA, and ZAP grossly parallel the changes In airway responsiveness noted with these seme agents. On the other hand, absolute numbers of PMNs In large and medium airways and peripheral lung WIre not statistically different betWlen these Interventions. Such findings underscore the difficulty in establishing a cause·and-effect relationship betWlen Inflammation and airway responsiveness. AM REV RESPIR DIS 1991; 143:585-589
dotoxin, PMA, and ZAP on the severity and location of polymorphonuclear leukocyte (PMN) accumulation within the lungs. The location and number of mast cells werealso measured. The sheep were giventhe same doses of endotoxin, PMA, and ZAP previously used to quantify alterations in pulmonary responsiveness to aerosol histamine (4, 16, 18). Sheep were killed 4.5 h after receiving the intervention, and the lungs wereremoved for morphometric analyses. Wewere thus able to compare the pathologic changes caused by the three stimuli to the previously reported alterations in airway responsiveness to aerosol histamine. Methods Sheep Preparation Yearlingsheep of both sexesweighing between
30 ~d 40 kg were used for these experiments. To insure normal lung anatomy, instrumentation was limited to catheters placed through (Receivedin originalform January 5, 1990and in revised form September 4, 1990) 1 From the Center for Lung Research, Departments of Medicine and Pathology, Vanderbilt UniversitySchool of Medicine,Nashville,Tennessee. 2 Supported by Grant HL-27274 and SCOR in Pulmonary Vascular DiseasesGrant HL-19153 from the National Heart, Lung, and Blood Institute. The work was done during the tenure of an Established Investigatorship awarded to J. R. Snapper by the American Heart Association with funds contributed in part by the Middle Tennessee Chapter of the American Heart Association. 3 Correspondence and requests for reprints should be addressed to James R. Snapper, M.D., B-1308, Medical Center North, Vanderbilt University Hospital, Nashville, TN 37232. 4 Recipient of a grant from the American Lung Association of Tennessee.
SIMPSON, BUTTERFIELD, LEFFERTS, DYER, SNAPPER, AND MEYRICK
% CHANGE ED65Cdyn
LESS RESPONSIVE MoRERESI'ONSIVE -
Fig. 1. Effect of ZAP, PMA, and endotoxinon pulmonary responsiveness to aerosol histamine. The vertical axis is percentchange in ED.. Cdyn (the dose of histamine requiredto causea decreasein dynamiccomplianceof the lungsto 65% of controlvalue).Aveluegreaterthan 100indicates thatthe intervention hasmadethe sheepless responsive, whereas a value less than 100 indicatesthattheintervention hasmadethesheepmore responsive to aerosol histamine (seereferences 2,9,16).
a neck incision into the aorta via the carotid artery and into the superior vena cava via the external jugular vein. The sheep wereallowed 5 to 7 days to recover from the operation before experimentation.
ZAP Preparation Autologous ZAP was prepared on the day prior to experimentation and frozen at - 60° C. 100 ml of plasma were removed from heparinized blood (1,000 USP units of heparin sodium per 100ml of blood) and centrifuged for 15 min at 2,000 rpm. The plasma was incubated at 37° C for 45 min with 10 mg/ml of zymosan (Sigma Chemical Co., St. Louis, MO). The zymosan was removed after incubation by centrifugation (4,000 rpm for 15 min) followed by passing the plasma through a Millipore filter (0.45 IJ.m pore size) (Millipore Corp., Bedford, MA). On the day of experimentation, the ZAP was allowed to thaw and come to room temperature prior to experimentation. Autologous "control" plasma was prepared in the identical manner except zymosan was not added to the plasma during the 45-minincubation period at 37° C. Each sheep received either ZAP or control plasma prepared from its own blood. Repetitive rapid bolus injections of 5 ml of autologous ZAP or control plasma in the pulmonary artery were repeated every 30 min for a total of eight doses. This was the identical protocol to that used to define the effects of repetitive bolus injections of ZAP on airway responsiveness in chronically instrumented awake sheep (18). PMA Preparation Stock 4 p-phorbol 12-myristate 13-acetate (Sigma), lO mg, was dissolved in 1 ml of dimethyl sulfoxide (DMSO) (Sigma) to make a final concentration of lO IJ.g/1J.1. Then 30IJ.I aliquots ofthis solution were pipetted into
an inert storage tube and immediately frozen to - 60° C. Within 5 min of experimental use, the PMA was thawed, and 5 IJ.g/kg werediluted in 10 ml sterile 0.90/0 NaCI for injection. Control injectate for the PMA studies consisted of 2.5 ul/kg DMSO diluted into 10 ml of sterile 0.9070 NaCI. The PMA was given as a bolus into the jugular vein. This is the identical protocol to that used to define the effects of PMA on aerosol histamine responsiveness in chronically instrumented awake sheep (16).
Endotoxin Preparation E. coli endotoxin (0.5 ug/kg prepared by the Westphal method from E. coli 055:B5; Difco Laboratories, Detroit, MI) was dissolved in 30 ml of 0.9070 NaCI solution and infused intravenously over 20 min. In the control experiments, 30 ml of 0.9070 NaCI were infused over 20 min. This is the identical protocol to that used to define the effects of endotoxin on aerosol histamine responsivenessin chronically instrumented awake sheep (4). Protocols The sheep were killed 4.5 h after they began receiving endotoxin, PMA, or ZAP. The endotoxin infusion required 20 min, PMA was given as a single bolus, whereas the multiple bolus protocol was adhered to for ZAP. This meant that sheep in the ZAP group werekilled 4.5 h after the first ZAP bolus and 1 h after the last (eighth) ZAP bolus. This was the same time at which aerosol histamine dose-response curves had been repeated after the three stimuli (4, 16, 18).Sixsheep were killedafter repetitive bolus ZAP, and three sheep were killed after repetitive bolus injections of control plasma; six sheep werekilled after PMA, and four sheep were killed after DMSO in 0.9070 NaCI (pMA control); seven sheep were killed 4.5 h after endotoxin, and four sheep were killed 4.5 h after a 0.9070 NaCI infusion. Sheep were killed with an overdose of thiamylal sodium (Parke-Davis, Morris Plains, NJ), the sternum was split, and the heart and lungs were removed en bloc. The entire left lung was removed for pathologic analyses. Pathology Methods The left lung was fixed for 24 h in a bath containing 1.00/0 glutaraldehyde in lOO7o buffered formalin-saline with the mainstem bronchus perfused with the glutaraldehyde-formalin mixture at a constant pressure of 25 em H 2 0 . Tissue blocks for 4- and l-um sections were taken from adjacent sitesat three places along the axial airway, running from the hilus to the distal left lower lobe. One block was the first intrapulmonary bronchus at the hilus, the second block was taken from the peripheral lung at the base of the left lower lobe 0.5 mm from the pleural surface, and the third block was taken from the middle-sized bronchus halfway along the axial airway. Tissue blocks were processed for routine light microscopic examination and for examination of l-um sections; 4-lJ.m paraffin-embedded
sections werecut and stained with hematoxylin-eosin, and l-um Epon/Araldite sections were cut and stained with toluidine blue. Quantitative light microscopic techniques were used to identify the location and severity of PMN and mast cell infiltrates induced by the three interventions in at least two (4and l-um) sections from each of the three locations of the left lung. All slides were examined in a blinded fashion. The number of nucleated PMNs were counted in the hematoxylin-eosin stained 4-lJ.m sections, and the number of nucleated mast cells was counted in the l-um sections. Wealso noted the number of eosinophils in the 4-lJ.m sections. The number of PMN and mast cells and their location in the large- (left hilar bronchus) and middle-sized bronchi were counted. These counts included PMNs and mast cells within the airway lumen, airway epithelium, lamina propria (the layerof connectivetissue immediately beneath the epithelium and superficial to a dense layer of elastin), vesselsin the lamina propria, submucosa (the layer beneath the lamina propria), and submucosal vessels. The circumference of the luminal aspect of the airway was measured. Complete rings of airwayswereused when 4-lJ.m sections were examined, and the number of PMNs are presented as number per millimeter of airwaycircumferencefor these large- and middlesized bronchi. The length of airway in the l-um sections was usually between 2 and 3 mm, and thus mast cellnumbers are presented per millimeter length of airway. An approximation of the number ofPMNs and mast cells and their location in peripheral tissue was obtained by counting the number of nucleated PMNs and mast cells in lO (x40 objective lens) light microscopic fields and relating the type and number of nucleated PMNs and mast cells to the number of alveolar profiles in these same microscopic fields (11, 12, 19). As for the larger airways, inflammatory cells associated with the smallest conducting airways, the respiratory bronchioli, werecharacterized as to location (within the airway lumen, epithelium, or subepithelium) and expressed as number per millimeter length of respiratory bronchiolus.
Statistics The Mann-Whitney two-sample test was used to assess the effects of endotoxin, PMA, and ZAP on PMN and mast cell numbers. The null hypothesis was rejected for p < 0.05 (20). Because the data from the control animals of all three experimental groups (endotoxin, PMA, and ZAP) were not statistically different from each other when data from the largeand medium-sized bronchi and the peripherallung were examined, the data were pooled for statistical analysis.
Left Hi/ar Bronchus The circumference ofthe hilar bronchus ofthe left lung for the control sheep was
AIRWAY RESPONSIVENESS AND INFLAMMATION
I .. _ i
10J EN:?'!JOX IN
( cell s /mm c ite )
Fig. 2. Effects of endotoxin (n = 7). PMA(n - 6), and ZAP(n = 6) on polymorphonuclear cell (PMN) and mast cell numbersin the large hilar bronchi. Data from endotoxin, PMA, ZAp, and control animals are plotted. Because therewerenodifferences between control values, the control data from the differentexperiments are pooled together (n = 11). The vertical axesare PMN numbersnormalized to millimeter of airwaycircumference. The data for PMNs aregivenfor PMNswithinthe airwaylumen, the epithelium,within the lamina propria but outside of blood vessels, within the blood vesselsof the lamina propria,within the submucosabut outside of blood vessels, within the blood vesselsof the submucosa, and as total PMNs. Asterisks indicate valuesthat are significantlygreaterthancontrolvalues (p < 0.05).
0: _ PMN,
(cell ,/m m c;«l 0:
I .. j
MAST CELLS (cell s / m m l en q t h)
II : II ( '0
(cell,/~ I. ",, ' h)
Z AP N;6