Tobacco Documents Online

Airway Permeability to Horseradish Peroxidase in Guinea Pigs: the Repair Phase after Injury by Cigarette Smoke1-3 WILLIAM Co HULBERT, DAVID C. WALKER, ANNE JACKSON, and JAMES C. HOGG Introduction Several years ago, Simani and associ- ates (1) suggested that the increase they observed in the permeability of the re- spiratory epithelium after exposure to tobacco smoke was due to damaged ep- ithelial tight junctions. More recently, Boucher and co-workers (2) demon- strated that this altered permeability was associated with structural damage to the sealing strands of the tight junc- tions. Their study also showed that to- bacco smoke in relatively small doses caused this damage to the epithelium and that increasing the dose increased the severity of the injury. These changes in mucosal permeability could be of considerable importance because they cause the underlying structures in the airway wall to be more exposed to substances present in the lumen. Al- though the permeability is increased, for example, the irritant receptors, which are located just beneath the tight junctions, are more exposed. These structures are thought to initiate rapid shallow breathing and bronchocon- striction (3), and they may be more easily stimulated when the mucosa is disrupted. In addition, the mucosal damage allows antigens to have greater access to the large number of mast cells present in the submucosa (4). The al- tered state of permeability associated with the mucosal injury must eventual- ly return to normal, and the time for repair will determine the period of ab- normal mucosal function. The present study was therefore designed to ex- amine the changes in epithelial permea- bility as the epithelium returned toward normal in the 24-h period after exposure to tobacco smoke. Methods Thirty Carom Hartlcy guinea pigs of either sex (average weight, 500 g) were used in this 320 SUMMARY Airway pemleshlllty was examined In the 24.h period immediately liter Injury by cigarette smoke In 30 guinea p~gs studied in groups ol five at 30 rain and 1, 6, 12, and 24 h after smoke exposure, end In 1 control group. The animals were anesthetized) trscheostomlzod, a carotid csnnula Inserted, and purified horseradish pemxidase was instilled on the airway sudaoe via the tracheostomy tube. Blood samples (0.8 ml) were drawn and replaced with haparlnlzed saline before and st 10, 15, 20, 30, and 40 rain after horseradish peroxldsse instillation. The animals ware then kilted, end samples ot trachea and lung tissue taken for weUdry wt determina- tions and for light and electron microscopic examination. The HRP concentrations In the blood were determined using an Ells= plate essay. We Iound the acute exposuro to 100 pulls cigarette smoke resulted In a Irenslsnt Increase In airway aipthellal permeability to HRP with e maximum at 30 rain and a return to control values by 12 h altar insult, Those chonges In mucosal pewneeblll- ty occurred In relation to s well-dellnad Inflammotory reaction where Increased permeability o¢- cuaed during the exudatlve phsso, which was monitored by measuring airway wet wUdry wl ratios and the InfiRration of polymorphonuclssr cells. The return to the control value of permeability was associated with the repair phase of the inflammatory reaction, which was measured by monitoring basal cell mitoses. AM REV RESPIR DIS 1981; 123:320-326 study. Twenty-five guinea pigs were exposed to 100 puffs cigarette smoke in the awake, restrained state using methods that have been previously described (2), and 5 control animals were exposed to 100 puffs air using the same exposure system. The smoking machine delivered 15 m[ of whole smoke every 20 s into a perforated manifold in which the head of the guinea pig was se- cured. The 25 experimental animals were studied in groups of 5, commencing at intervals of 30 rain and 1, 6, 12, and 24 h after exposure. The test cigarettes were sup- plied by the Tobacco Manufacturers Coun- cil of Canada and were composed of high quality flue-cured tobacco. At each of the predetermined time points after challenge, the animals were intraperi- toneally anesthetized with Nembutal® (25 mg/kg) and a tracheostomy tube (PE-10) and carotid cannula (PE-200) were inserted. Purified horseradish peroxidase (HRP) (1 mg in 0.2 ml tyrodes) was introduced through the tracheostomy tube over a 4-rain period. Blood samples (0.8 ml) were taken before and at 10, 15, 20, 30, and 40 min af- ter placing the HRP on the tracheal surface with the blood volume being replaced with heparinized saline. Because of the 40-min blood sampling time, the morphologic anal- ysis was conducted at 70 rain, 1 h 40 rain, 6 h 40 rain, 12 h 40 rain, and 24 h 40 mins after exposure. The animals were then killed with a carotid injection of saturated KCI. and the trachea and lung tissue rapidly re- moved for microscopic examination and wet wt/dry wt determination. The HRP used in this study was Sigm; Type 11 (P8250), (Sigma Chemical Co., St. Louis, MO), which was purified accordin to the method of Moroz and associates (5) Assay .for HRP in serum. An Elisa assa was used to measure HRP in serum. Tt- antiserum was obtained from rabbits in munized with 100-pg aliquots of HRP in 0 ml phosphate buffered saline (PBS) in equal volume of complete Freunds ad) vant and boosted at 6-wk intervals. antibody concentration in various samples was tested by Ouehterlony analy ~th 1 mg/ml of the purified enzyme. The with titers of i/8 or greater were pooi (Received in origina/ form Jul.)" 29, 1980 one. revised form November 24, 1980) ~From ti~e University of British Colur." Pulmonary Research Laboratory, St. Hospital, Vancouver, B.C., Canada. = Supported by the MRC of Canada and Tobacco Manufacturers Council of Canada = Requests for r¢prints should be addresse Dr. J.C. Hogg. St. Paul's Hospital, 1081 Burr Street, Vancouver. B.C.. Canada. V6Z IY6 T!04231001

AIRWAY MOCOSAI. PERME~BILIT~ ~,~1 and the serum fraetionated by precipitation with 40% saturated ammonium sulphate. This antiserum precipitated 2 lines of en- zyme-active malerial in both the purified and crude preparations on immunoelectro- phoretic assay. Standards were prepared by diluting pu- rified HRP to l to 320 ng/ml in PBS, or, in some instances, in one-fifth normal guinea pig serum in PBS. These were prepared fresh daily from I mg/ml stock solutions in H 20. The stock solutions were stored with- out preservatives and were replaced at month- ly intervals. The antiserum fraction was diluted from 1/500 to 1/30~ in 0.05 Mcax- bonate buffer (pH, 9.6), and 100 ~1 aliquots were added to the wells of microtiter plates (Elisa-Cooke Engineering 1-223-229), the plates were incubated overnight at 2 to 5° C, the contents were removed, and the plates were washed 3 times with PBS. Quadruplicate I00-~I aIiquots of the stan- dards were incubated on the antibody coated plates for 3 to 4 h, and removed by washing. Enzyme activity was measured by the method of Saunders (6), using 0.25 M 2'-azino-di-(zet hylbenzthiazloine sulfonic acid) or a dye and 0.1% HzOz as substrate in 0,05 M citrate buffer (pH, 4.0). When nec- essary, the reaction was stopped by the ad- dition of HF:0.1 M in EDTA 0.001 M, NaOH 0.01 M. The plates were read at 2 to 5 rain intervals in a spectrophotometer (Multiskan Flow Laboratories), and the average optical density for each concentra- tion plotted on semilog graph paper. Rate averages were:also plotted to assess the linearity of the color development. Guinea pig serum samples were diluted 1/5 to 1/10 in PBS, and 100-/~l aliquots were incubated in quadruplicate in 96 well plates, as described previously. A 6-poim standard curve was included on each, as well as randomly located individual stan- dards and PBS as a blank. Color develop- ment was assessed at 3 or more intervals after substrate and dye addition and un- known optical densities read at each time from the appropriate standard curve. Cor- rection for nonspecific serum effects was carried out. Tissue preparation. Trachea from the control animals and from the l-h, 6-h, 12-h, and 24-h experimental animals was fixed by immersion in ice-cold 2% glutaral- dehyde in 0. i M Na-cacodylate buffer (pH, 7.4). However. adequate fixation of the tra- chea for the 30-rain animals required retro- grade perfasion of the abdominal aorta caudal to the diaphragm. After primary fix- ation, the tissues were washed overnight in the same buffer and processed according to Karnovsky (7) to demonstrate the presence of HRP. The tissues were then postfixed in 2% O~O,, stained on block with saturated uranyl acetate in distilled water, dehydrated in a graded ethanol series, and embedded in Spurr. Thin sections were cut using glass knives on a Reichert ultramicrotome, mounted on naked 200-mesh copper grids, gained with uranyl acetate (saturated in 70% methanol) and lead citrate (8), and viewed with a Phillips 400EM. l~valuation of epithelial penetration. Penetration of the tight junctions was de- termined by examining for the presence of an electron-dense reaction product indicat- ing HRP in the intercellular space. At least 8 grid spaces (between 4 and 10 cells per grid space) from 2 separate grids were ex- amined for each block and at least 2 blocks were examined for each animal. The results were presented as a ratio of the number of junctions penetrated/total counted. This analysis was conducted by ! investigator (WCH) who was unaware of the experimen- tal condition. Cell counts. Cell counts were conducted by 2 investigators (WCH, DCW) who were unaware of the experimental treatment of the animals. For each animal, at least 3 tis- sue blocks with 2 slides (i-~ sections stained with Toluidine Blue O) per block were ex- amined yielding an epithelial length in ex- cess of 1 cm for all experimental groups. Goblet cell~, e~slnophils, plasma cells, polymorphonuclear leukocytes, and mast cells were counted, and the totals were ex- pressed as cells/mm epithelium. Mitotic fig- ures were taken as distinct anaphase, meta- phase, and telophase stages and are ex- pressed as the number of mitoses per/ram of epithelium. Assessment of basement raembrane. The basement membrane area was defined as that region bounded by the elastic lamina and basal cells. Area determinations were made using planimetry on light micro- graphs and the measurements are presented as basement membrane area mml/mm epi- thelium. Statistical analysis. Dunnett's modifica- tion of the t test (9), which allows multiple comparisons to a single control, was used to ascertain significance between the control and the experimental groups. This test was used to compare the rate of appearance of HRP in the blood, the change in wet wt/dry Wl ratio, changes in various cell types, the mitotic index of the basal cell layer, and the thickening of the basement membrane. Results The acute exposure of guinea pigs to 100 puffs whole tobacco smoke pro- duced electron microscopic evidence of increased airway epithelial permeabili- ty, which was maximal in the 30-rain and l-h groups of animals (and hence between 70 and 100 rain), and returned to control values by 12 h after exposure (figure 1). The mean junctional pene- tration ratio - SE (tight junctions pen- etrated/total tight junctions counted) for the 30-rain, 1- and 6-h postsmoking groups was found to be 0.9_+ 0.03, 0.7 Hours Post Chollengewl|h 100 Pulls Whole C=garene Smoke Fig. 1. This histogram plots the penetration ratio (number of tight junctions penetrated/total num- ber counted) for the 24-h examination period after exposure to 100 puffs whole cigarette smoke. Standard error bars are indicated, as well as the number of junctions penetrated over the total counted. ± 0.15, and 0.2 __- 0.17, respectively. In contrast, no evidence of junctional penetration was obtained by electron microscopic examination of the control 12- and 24-h groups. In figure 2A and B, the electron micrographs are of a 30-rain postsmoking animal. These 2 micrographs demonstrate the presence of the electron-dense tracer HRP in the intercellular space (arrows), and fur- ther show an epithelial nerve in longi- tudinal section (figure 2A) and in cross-section (figure 2B). Note the close positioning of these nerves to the tight junction and airway surface. The insets of Figure 2A and B are high magnification enlargements of the epi- thelial nerves and readily demonstrate the electron-dense tracer HRP in the intercellular space, as well as longitudi- nal and cross-sectional views of nerve fibers. An electron micrograph from a control animal (figure 3) shows the presence of HRP on the airway sur- face, and the inset clearly shows that there is no HRP in the intercellular space. The appearance of HRP in the blood during the 40-rain period after it was placed on the trachea is presented in figure 4. This shows that in animals not exposed to tobacco smoke the rate of appearance was slow and the value re- vealed at 40 rain was similar (0.1 ± 0.06% injected dose) to that reported by Boucher and co-workers (2) who used a radioimmunoassay technique. In contrast, in the animals examined 30 min after smoke exposure, the rate of appearance was faster and the 40-min blood concentration increased seven- fold to 0.73 ± 0.3~0 injected dose. The mean rate of appearance ± SE of HRP in the serum during the 40 min after Ti04231002

";22 HULBERT, WALKER, JACKSON. AHD HOGG F~g. 2 A. Eleclron micrograph Irom an ammal examined 30 m~n after smoke exposure. Note espemally an ep~thehal nerve in longitudinal secbon and its close proximity to the tight junclion and atrway surface. and the HRP m the intercellular space (arrows). The inset shows a high magnification enlargement ot the nerve in tong~tudmal section; bar = 0.55 p. B. Note the cross-sectional views o! the nerve lihers and thmr close proximity to the mrway surface. The inset shows a h~gh magnification enlargement of the nerves in cross-section: bar = 0.55 ~. (C = ciha; G = golgi complex; HRP = horseradish peroxidase; M --- mitochondna; MV .= microvilli; NF = neural filaments; N = nerve: NU = nucleus; RER = rough en- ooplasm~c ret=culum; TJ = t~ght junct=on.) placement of HRP on the tracheal sur- face is shown for each of the groups in figure 5. This shows that HRP accum- ulated more quickly in the blood of animals examined at 30 rain, 1, and 6 h after smoke exposure than it did in the control group or in those animals ex- amined 12 and 24 h al'ter smoke expo- sure. The data concerning the exudation of fluid (tracheal wet wt/dry wt ratio), recruitment of polymorphonuclear ce.lls, the onset of the repair phase (mi- totic index), and basement membrane measurements are presented in figure 6. As shown in figure 6A, the tracheal wet wt/dry wt ratio increased from control values of 2.4 ± 0.16 to 3.2 ± 0.18 at 30-rain postsmoking (p < 0.05) and re- turned to control values 12 h after the insult. By 24 h the ratio decreased to 1.9 ± 0.06 (p < 0.05) below that of the control group. As shown in figure 6B, the number of polymorphonuclear cells increased fivefold from control values of 2.8 ± 0.7 to 14.6 ± 2.9 cells/mm epithelium 6 h after injury, and then decreased slightly to 9.5 ± 1.6 cells/mm epithelium by 24 h postexpo- sure. Initiation of the repair phase (fig- ure 6C) was marked by a 12-fold in- crease in basal cell mitoses (from 0.5 _.+ 0.2 to 6.2 ± 0.4/ram epithelium) at 12 h postchallenge, with further increases to 15 mitoses/ram epithelium by 24 h postchallenge. In contrast to the tra- cheal wet wt/dry wt ratios, the thick- ness of the basement membrane (figure 6D) initially decreased from 8.9 -+ 0.4 to 6.0 ± 0.3 mm~/mm epithelium, and then gradually increased, reestablish- ing control values by 12 h postsmoking and then significantly (p < 0.05) ex- ceeded control values by 24 h. The data concerning the numbers of goblet cells, mast cells, eosinophils, and plasma cells are shown in figure 7. This figure shows that goblet cells (fig- ure 7A) decreased from 26.8 + 10.0 to 7.3 _+ 2.1 cells/ram epithelium during the 24-h examination period with values of 5.1 + 1.8and2.5 _+ 1.g recorded at 30 rain and 1 h postsmoking. Epithelial mast ceils (figure 7B) generally increased in number to a maximum of 6.7 + 0.8 cells/ram of epithelium by 6 h postexpo- sure. Eosinophils (figure 7C) decreased from 19.4 +_. 3.9 to 0.7+ 0.3 at 24 h postchallenge. Plasma cells (figure 7D) showed an immediate threefold de- crease from 16.7 ± 2.3 to approxi- mately 5 cells/ram epithelium, an amount that was maintained through. out the remaining examination period. The observations concerning celt counts made by 2 independent obser- vers (WCH, DCW) unaware of the e×. perimental treatment are compared i~ table 1. The table shows that the corre lations are all greater than 0.986 an. that the slopes of the regression are a" close to 1. Discussion Previous studies (1,2) have shown th, tobacco smoke injures the respirato; epithelium by damaging the struetur. components of the tight junctio: which in turn increases mucosal pernf. ability. Our work extended these obse: vations and showed lhat this incre-a ; TIO4231003

• IRWAY MUCO~I,,L PERMF.ABILri'Y 323 Fig. 3. Electron micrograph Irom a control animal demonstrating the presence ot HRP on the airway sur- face and no intercellular penetration (arrow~). The inset is a high magnification enlargement of the boxed area and clearly shows no penetratton of HRP Into the intercellular space, (C = cilia; GC = goblet cell; HRP = horseradish peroxidase; M = mitochondria; MV = microvilli: NU = nucleus;TJ = tight junction; bar = 2,0 ~.) ":" in mucosal permeability is transient in nature, with maximal permeability oc- curring between 30 rain a~d 6 h after exposure and a return to control values by 12 h after injury. In contrast, Marin and associates (10) fourtd that perme- ability returned to normal as early as 6 h after mechanical injury of the trache- al mucosa. They showed that while only point contacts were present between epithelial cells 6 h after injury, the col- Iodial intercellular tracer lanthanum hydroxide was excluded from the inter- cellular space (11). Total reformation of the epithelial tight junctions was found to occur by 12 h after mechani- cal injury. The 6-h difference in the time required to restore the barrier after permeability between mechanical and smoking injury is of interest. It may be that the initial injury after smoke exposure in our study was much greater than the mechanical injury studied by Gordon and Lane (11). It could also be that the initial injuries were the same, but complete restora- tion of the junction was required to prevent penetration of the HRP in our study. A careful investigation using the freeze cleave technique will be required to settle this point. Increased alveolar permeability in symptomless chronic smokers was re- cently demonstrated by Jones and col- leagues (12). They measured the rate of transfer of [99mTc]diethylenetraimine penta-acetate from the lung into the blood and found a significantly higher transfer rate in the symptomless cigar- ette smokers when compared with non- smokers. That chronic smoking in- creases airway permeability is of interest with respect to our findings that corre- late increased airway permeability with inflammation. Thus, the increase in permeability that Jones and colleagues (12) observed may be related to airway inflammation occurring before the onset of the classic symptoms of bronchitis. It is of interest that mueosal perme- ability was increased at the same points in time (30 min, 1, and 6 h), that the tracheal wet wt/dry wt ratios were ele- vated. This suggested that mucosal per- 6 t'o ~o ;o 20 ~'o Hmules Fig. 4. This figure shows the accumulation of HRP (as % inlected dose) in the blood during the 40-rain sampling time and demonstrates the marked increase in HRP accumulation in the animals ex. amined 30 rain after smoke exposure. Time 0 is the time at which the HRP was placed on the tracheal surface. Standard error bars are indlcalad, meability is increased during the exuda- tive phase of the it~flammatory response induced by ~he tobacco smoke and re- turns to normal as the exudation sub- sides. A change in mucosal permeabil- ity associated with acute inflammation could have important implications with respect to airway function. A number of studies have shown that a wide vari- ety of insults, including viral infections (13), NO~ (14), and ozone (15), cause increased bronchial hyperactivity, and Empey and colleagues (16) have sug- ~. ooo= In0 Pulls Whole C,garelle Smoke, Fig. 5. This histogram shows Ihe mean slope or rate of HRP penetration (as % injected dose) at each of the time periods studied alter exposure to 100 puffs whole cigarette smoke. Each bar ~epre- serifs the data lrom 5 animals. Standard error bars are indicated, and experimental values signifi- cantly different at the 95% confidence level are aslerisked. T!04.231004

:~24 - HULBERT, WALKER. JACKSON. &ND HOGG ~ E Hours Posl Cholfe~ge wdh 100 Pulls Whole Clgorelle ~moke Fig. 6 A. ~ulnea pi9 trachea wet wt/dry wt ratio for 1he 24-~ examination period after exposure to 100 puffs whole cigarette smoke, Experimental values significantly different (p < 0.05) from 1hose ol the con- trol group a~e asterisked, B. The number of PMNImm of tracheal epithelium Ihroughoul the 24-h I~on perio~ alter exposure to 1~ Duffs whole mgarette smoke. Standard error bars are indicate~, and perzmental values slgmbcantly different (p < 0.05) from those of the control group are asterisked. G. The mitoU¢ m~ex or number of mitoseslmm of guinea pig t[ac~eal epithelium for the 24-h examinalio~ perio~ after exposure to 100 pu~fs whole cigarette smoke. ~la~t~ error bars are indicateU, and experimental values szgnihcantly ~ifferenl (p < 0.05) from those of the ¢ontroI group ar~ asl~isked. D. Basement brahe are~ mm~/mm guinea pig tracheal epilhelium for the 24-h examination period alte~ exposure to 100 pulls whole mg~relte smoke. ~tandard error bars are indicated, and ~xperimenlal values significantly ~ifle~enl (p < 0.05} from those of the control 9~oup are asterisked. (PMN = polymorphonuclear leuko- cyzes). gested that the increased nonspecific reactivity is related to "mucosal dam- age." Although the nature of this mu- cosal damage has not been well de- fined, Boucher"and co-workers (2) have presented data suggesting that the hyperreactivity and increased mucosa[ permeability may be linked. The pres- ent study was relevant to this question because it showed that increased muco- saI permeability is a feature of the exu- dative phase of the acute inflammatory reaction. This raised the possibility that the common denominator between viral infections, ozone, NO~, cigarette smoke, etc., may well be linked to the increased mucosal permeability that is associated with mucosal inflammation. ~ s , °il[i'l r I Hours Post Cholle~e w~th 100 Pulls Whole C go:elle ~moke F~g. 7 A. Goblet cellslmm guine~ pig tracheal epilhelium for the 24.h examination period afler exposure to 1~ p~lfs whole mgarette smoke. B. Mast cell~mm of guinea pig Vacheal eoithelium for the 24-h ex- ammahon period after exposure to 1~ puffs whole mgarette smoke. C. Eosinophilslmm ot guinea pig Jrac~eal epithehum fo~ the 24.h examination per~ afte~ exposure to 1~ puffs w~ole cigarette smoke. D. Plasma cell~mm guinea pig ffacheal epithelium 1or the 24-h examination peri~ alt~ exposure to 1~ puffs whole cigarelte smoke. Standard error bars are indicate, and experimenlsI ~lues significantly diif~ent (p < 0.05] fr~ Jhose of the contr~ group a~e asterisked. The precise reason for the increased re- activity could not be pinpointed in these experiments, but increased exposure of nerve endings (figure 2A and B), that occurred as a result of the epithelial damage is one possibility. As shown in figure 6C the epithelial repair begins about 12 h after the in- jury when the number of basal cell mi- toses increase. Previous investigators (17) have demonstrated that exposure to tobacco smoke results in a dose-re- lated increase in mitotic activity that peaks 24 to 36 h after exposure, with delays as long as 72 h in previously in- jured epithelia (18). Our data provided additional information that showed that permeability remains high after the injury until the repair phase begins. This means that toxic chemicals, in- cluding carcinogens present in cigarette smoke, could have prolonged access to the dividing basal cell layer. Our work was consistent with others (!!) who have demonstrated that the new epithe- lium is composed of squamous noncil- iated cells that previously comprised the basal cell layer. It is of interest that the tracheal wet wt/dry wt ratios follow a pattern al- most directly opposite to that observed for basement membrane thickness. Comparison of Figure 6A with 6D shows that the basement membrane thickness decreased when the wet wt/ dry wt ratio increased in the first 6 h after the injury. By 24 h, however, the wet wt/dry wt ratio has decreased below control values (p < 0.05), whereas the basement membrane thickness was sig- nificantly increased (p < 0.05) above that of the control. Because of this in- verse relationship, we felt that the in- crease in basement membrane thickness at 24 h after smoking represented the synthesis of new material and could not be attributed to swelling of the base- ment membrane by an inflammatory exudate. This suggested that increased production of collagen and glycopro- rein moities comprising the basement membrane occurs in concerl with the increase in basal cell mitoses, all within a 24-h period. This observation was es- pecially interesting in light of the recen-" work by Vracko 119) and Vracko ant' Benditt (20) relating basal membran~ thickness in diabetics to endothelia turnover. They found that each genera tion of endolhelium produces a laye of basement membrane, and, mor.:- over, that diabetics have a higher tha: normal rate of endothelial turnove~ T!04231005

TABLE 1 INTEROBSERVER REPRODUCIBILITY Polymorpho- nuclear Mast Plasma Milalic Goblet Leukocyles Cells Cells Figures Cells Number of observations 368 83 113 83 145 Correlation 0.998 0.996 0.988 0.987 0.987 Slope of regression 0.907 0.990 0.900 0.985 0.810 Intercept of regression 1.27 1.65 0.73 - 0.88 0.67 Thus, Vracko argued that the excessive accumulation of basal membrane found in diabetics resulted directly from in- creased turnover of the vascular en- dothelium. Our data suggested that a similar relationship may exist for the airway epithelial turnover and its base- ment membrane and raised the possi- bility that the thickened basement membrane seen in chronic inflamma- tory states such as asthma may well be related to an increased epithelial cell turnover. The cellular component of the in- flammatory reaction was indicated by a peak number of PMNs at 6 h after the exposure (figure 60). Kilburn and co- workers (21) have reported that in the hamster, peak recruitment of PMNs in response to tobacco smoke exposure occurred at 24 h after insult, whereas cotton dust and .tobacco extract caused a peak recruitment between 6 and 12 h after exposure. The difference between our results and those of Kilburn and co-workers (21) could represent either exposure or species differences. The data on eosinophils, mast cells, and plasma cells (figure 7) reported here were included for descriptive pur- poses. It is possible that the reduction in eosinophils could be related to their emigration to the epithelial surface be- cause we found the mucous layer rich in eosinophils in favorable sections where it was well fixed. However, this observation cannot be quantified since not all the tissues demonstrated an in- tact mucous layer. The increase in epi- thelial mast cells is interesting and sug- gests that tobacco smoke might cause the release of positive chemotactic fac- tors that attract mast cells to the in- jured region. The possibility that in- creased numbers of mast cells might contribute to airway hyperreactivity deserves further investigation. The sharp reduction in the number of goblet cells observed in this study (figure 7A) at 30 rain after exposure is coincident with the increase in epitheli- al permeability. This is interesting with respect to previous studies (22) of guin- ea pig tracheal epithelia using freeze cleave techniques where a high degree of disarray was found in the junctional complexes between ciliated and dis- charged goblet cells. A similar observa- tion of junctional disarray was also re- ported by Wade and Karnovsky (23) in osmotically stressed epithelia. These data suggested that there may be an in- terrelation between goblet cell dis- charge and increased mucosal perme- ability. Although previous studies have reported an increase in goblet cells after exposure to tobacco smoke, they have not concentrated on the changes that occur immediately after injury. An exception to this is a study by Reid (24) in rats showing a decreased num- ber of goblet cells 2 wk after injury and then an increase to values that ex- ceeded those of control values 4 wk af- ter exposure. In summary, the acute exposure of guinea pigs to 100 puffs of whole to- bacco smoke caused an inflammatory reaction with a well-defined exudative and proliferative phase. Increased mu- cosal permeability was associated with the exudative phase of the inflamma- tory reaction and appeared to begin with goblet cell discharge. The mucosal permeability returned to control values 12 h after the exposure when the epi- thelial repair phase began. Acknowledgment The writers express their gratitude to Joan Dixon and Linda Hare in preparing the manuscript, to Irene Nicols for help with the graphics, and to Wendy Laird for re- sourceful technical assistance. References 1. Simani IAS, Inoue S, Hogg JC. Penetration of the respiratory epithelium of guinea pigs fol- lowing exposure to cigarette smoke. Lab Invest 1974; 3 ! :75-87. 2. Boucher RC, Johnston J, inoue S, Hulbert W, Hogg JC. The effect of cigarette smoke on the permeability of guinea pig airways. Lab In- vest 1980; 43:94-100. 3. Widdlcome JG. Reflex control of trachea- bronchial smooth muscle in experimental and human asthma. In: Lichtcnstein LM, Austcn cds, Asthma--Physiology, immunopharmacol- ogy and treatmenh New York: Academic Pres.g 1977:225-31. 4. Guerzon GM, Pare PD. Michood MC, Hogg JC. The number and distribution of mast cells in monkey lungs. Am Rev Rcspir Dis 1979; 119: 59-66. 5. Moroz LA, Joubert JR, Hogg JC. A potent non-peridoxasc glycoprotein tmmunogen in com- mercial horseradish peroxidase. J Immunol 1974; i 12:1094-9. 6. Saunders GL. Art of solid phase im- munoassay including selected protocols. In: Nakamura RM, ed, immunoassay in the clinical lab. New York: Alan Liss Co., 1979:173-89. 7. Karnovsky MJ The uhrastructural basis of capillary permeability studied with peroxidase as a tracer. J Cell Biol 1967; 35:215-30. 8. Reynolds ES The use of lea~l citrate at high pH as an electron opaque slain in electron micro- scopy. J Biophys Riochem Cytol 1963; 10:308-12. 9. Dunnett CW. A multiple comparison proce- dure for comparing several treatments with a control. J Am Statist Assoc 1955; 50:1090-121. 10. Marin ML, Gordon RE, LaneBP. Develop- ment of tight junctions in rat tracheal epithelium during the early hours after mechanical injury. Am Rev Respir Dis 1979; i 19:101-6. i 1. Gordon RE, Lane BP. Regeneration of rat tracheal epithelium after mechanical injury. II. Restoration of surface integrity during the early hours after injury. Am Rev Respir Dis 1976; 113:799-807. 12. Jones JG, Lawler P, Crawley JCW, Minty BD, Hulands G, Veall N. Increased alveolar epi- thelial permeability in cigarette smoke. Lancet 1980; 1(8159):66-8. 13. Laitmen LA, Ekin RB, Empey DW. et aL Changes in bronchial reactivity after administra- tion of llve attenuated virus. Am Rev Respir Dis 1976; 113:194-201. 14. Orehek J, Massari JP0 Gagrard P, Grinoud G, Charpin J. Effect of short term, low level ni- trogen dioxide exposure on bronchial sensitivity in asthmatic patients. J Clin Invest 1976; 57: 301-7. 15. Golden JA, Nadel J, Boushey HA. Bron- chial hyperirritability in healthy" subjects after ex- posure to ozone. An1 Rev Respir Dis 1978; 118: 29:287-94. 16. Empey DW, Laitmen LA. Jacobs J, Gold WM, Nadel J. Mechanisms of bronchial hyper- reactivity in normal subjects after upper re- spiratory infection. Am Rev Respir Dis 1976; 113: 131-9. 17. Wells AB, Lamerton LF. Regenerative re- sponse of the rat tracheal epithelium after acute exposure to tobacco smoke: a quantitative study. J Nail Cancer lnst 1975; 55:887-94. I$. Wilhelm D. Regeneration of tracheal epi- thelium. J Pathol 1953; 65:543-57. 19. Vracko RV. Basal lamina layering in diabe- tes mellitus. Evidence for accelerated rate of cell death and cell regeneration. Diabetes 1974: 23: 94-104. 20. Vracko RV. Benditt EP. Manifestations of diabetes mellitus: their possible relationships to an underlying cell defect. Am J Pathol 1974; 75:204-21. 21. Kilburn KH, McKenzie WN, Thurston RJ. Cellular effects of cigarette smoke on hamster Ti04231006

airways. C'~t 1975: 67:545-9. 22. lnoue S, Hogg JC. Freeze-etch study of Ih¢ lracheal epithelium of normal guinea pigs with particular relrerenc¢ to intercellular junctions. J Uhraslruc! R~ 1977: 61:89-97. 7.3. Wade J'B, Karnovsky M.,]. Fracture faces of osmotically' disrupted zonulae occludentes..I Cell Biol 1974; 62:344-53. HULBERT, WALKER. JACKSOH, Arid HOGG 24. Rcid L Bronchial mucus production in health and disease, in: Liebow AA, Smith DE, ecb,. "[he lung. International Academy of Path- ology monograph no. 8. Ne'~-" York: The Williams and Wilkins Co., 1978: 8:87-108. T104231007