Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The cigarette-smoking habit produces lung inflammation in everyone who smokes, and this inflammatory process underlies the pathogenesis of emphysema. However, because only 15 to 20% of heavy smokers develop emphysema, other host and environmental factors must be involved. Previous work from our laboratory (1) has suggested that latent adenovirus (Ad) infection amplifies the cigarette smoke-induced inflammatory process and that it may account for emphysema in a minority of smokers. Therefore, our primary objective was to determine whether the inflammation (1) and the emphysematous lung destruction produced by chronically exposing guinea pigs to cigarette smoke (6) is amplified by the presence of a latent Ad infection.

Our secondary objective was to determine whether the administration of all-trans-retinoic acid (RA) is capable of reversing cigarette smoke-induced emphysema in this model. This objective was based on Massaro and Massaro's report (7) that RA reverses the elastase-induced emphysema. RA and retinol are biologically active derivatives of vitamin A that are essential for normal lung development and growth and play a key regulatory role in the septation process that leads to the formation of new alveoli. Experiments in rats (8) have shown that RA stimulates the formation of new alveoli and blocks the inhibiting effect of dexamethasone on alveolar growth. The present study was designed to determine whether RA can either prevent or reverse the emphysematous destruction produced in guinea pig lungs by long-term exposure to cigarette smoke (6).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Study Design

The experiments were designed to determine the independent and combined effects of smoking, latent Ad infection, and RA treatment on emphysematous lung destruction, and the effects of smoking and latent Ad infection on the inflammatory cell response. Animals were randomized into Ad-infected and sham- infected groups. When viral replication stopped (9), each group was further randomized to smoking and nonsmoking groups, and these groups were again randomized to receive either RA or its vehicle. This design included a total of 76 guinea pigs because previous experiments suggested that eight or nine animals would be required in each of the eight experimental groups to have the power to detect differences in morphometric parameters (Figure 1).

View larger version (14K): [in this window] [in a new window]   Figure 1.   Experimental groups and the randomization of animals to Ad or sham infection and then to either cigarette smoke- exposed or nonexposed groups that either did or did not receive RA treatment. Bottom: the number of animals that died in each group. Vehicle: cottonseed oil.

Animals

Juvenile female guinea pigs (Cam Hartly), 250 to 300 g, were purchased from Charles River (St. Constant, PQ, Canada) and housed in polycarbonate cages fitted with high-efficiency particulate air filters. Animals were provided with food and water ad libitum and their body weights were measured weekly throughout the experimental period.

Ad Infection

Wild-type adenovirus 5 (Ad5) was obtained from the American Type Culture Collection (Rockville, MD). The virus was propagated in monolayer cultures of A549 cells (obtained from the same source) grown in Minimal Essential Medium (GIBCO BRL Life Technologies, Inc. Gaithersburg, MD) supplemented with 10% fetal bovine serum. At 2 to 5 d after inoculation the supernatant was collected and, after CsCl centrifugation viral bands were recovered, supernatant was dialysed against 10% glycerol in phosphate-buffered saline (PBS) (pH 7.4) at 4°C and then stored at 70°C (9). The viral titer was determined by plaque assay. Animals were anesthetized with Halothane and then given 10⁹ plaque-forming units of purified Ad5 in 200 µl PBS intranasally. The infected guinea pigs were kept in containment for 22 d to allow latent infection to develop (9). Animals given sterile PBS (pH 7.4) alone served as a control for the Ad5 infection. After the containment period, both Ad5- and sham-infected animals weighing 400 to 500 g at this time were exposed to cigarette smoke and RA treatment as described later. The persistence of the latent Ad infection was confirmed at the end of the study by immunohistochemistry to demonstrate the Ad E1A protein in the lungs of infected animals.

Cigarette and Smoke Exposure

Nonfilter cigarettes, purchased through the Canadian Tobacco Manufacturer's Council, that yielded 1.1 mg nicotine, 16 mg tar, and 11 mg carbon monoxide under standard smoking regimen when smoked to a 23-mm butt length were used in this study. Animals were exposed while awake using a previously described apparatus (10). Each guinea pig was placed in an individual stall and, to protect the animal's eyes from smoke, the snout protruded through a rubber dental dam into a chamber filled with cigarette smoke. The smoke was sucked from the cigarette into a 20-ml syringe and exhausted into these chambers in a cyclic manner. The animals were exposed to smoke from five cigarettes (20 ml puffs) delivered over a 40-min period, for 5 d each week. During the initial exposure to cigarette smoke the animals were sedated with 0.2 ml Inovar (50 µg/ml Fentanyl and 1.5 µg/ml Droperidol) delivered subcutaneously before beginning the exposure. The controls were placed in the apparatus for the same period of time and exposed to the room air instead of cigarette smoke. These smoke and sham exposures took place over 13 to 16 wk. When bronchospastic reactions developed during cigarette-smoke exposure, animals were removed from the smoke and treated with intraperitoneal (i.p.) injections of adrenaline (0.2 ml, 1:10.000) and oxygen.

Treatments

Animals treated with RA received a daily i.p. dose of 500 µg/kg of RA (Sigma-Aldrich, ON, Canada) diluted in cottonseed oil. The control groups received an equal volume of cottonseed oil delivered in the same way. The dose and the route of administration of RA and cottonseed oil were based on previous reports in rats (7). Both RA and cottonseed oil were handled under sterile conditions and protected from light.

Measurement of RA in the Serum

The animals were killed at least 24 h after the last RA treatment. A cardiac puncture was performed and the serum obtained from the blood drawn was stored at 70°C. Serum RA was measured using high-performance liquid chromatography (HPLC) as previously described (11).

Lung Processing and Morphometry Analysis

The body weight of each guinea pig was recorded just before i.p. administration of an overdose of pentobarbital. The lungs were removed and weighed. The right middle lobe was inflated with and immersed in Trizol and stored at 70°C for future analysis of RNA.

The remaining right upper and lower lobes were inflated and fixed in 10% buffered formalin for at least 24 h. The volumes of the right upper and lower lobes were measured by water displacement (± 0.02 ml), normalized to body weight, and used as a reference volume for morphometric analysis. Each lobe was sectioned transversally into four slices. Each transverse slice was processed into paraffin and at least two were randomly selected from each lobe for quantitative histologic analysis of emphysematous process. Histologic sections cut onto glass slides were stained with hematoxylin and eosin (H&E) and coded to prevent the observer from identifying the experimental groups under study. At least five randomly selected fields were analyzed on each slide to provide a coefficient of error < 0.1 (12). All slides were analyzed at ×100 magnification using a computer program (Gridder) to select the fields and apply the test lines.

A multilevel cascade design (13) was used to determine the volume fraction (Vv) of lung tissue, air space, airway wall, and lung surface density (Sv). A grid of lines with a point at each end was superimposed on the histologic image examined under the light microscopy. The number of end points that fell on tissue, air space, or airway wall, and the number of times the line intersected with lung parenchyma, were recorded. The Vv of tissue, air space, and airway wall were derived by dividing the number of points that fell on each of these lung components by the total number of points in the field. Sv, which provides the ratio of surface area to volume (SA/Vol), was determined from the number of points that fell on tissue and the number of alveolar intersections with the test lines (14, 15). The total SA of lung parenchyma was then calculated by multiplying Sv by the total lung volume.

The left lung was inflated with a 1:1 mixture of PBS (pH7.4) and CryoMatrix, then frozen in liquid nitrogen and stored at 70°C. The frozen sections from this lung were used for immunodetection of Ad E1A protein and for the histologic analysis of the inflammatory process. Each lobe of the left lung was transversally cut into two slices. A frozen slice was then randomly selected and sections approximately 8 µm thick were cut onto silanazed slides, air-dried, and fixed in 10% buffered formalin for 10 min. One complete set of sections was immunostained using an antibody against the Ad E1A protein (3), and other sets were used for immunodetection of inflammatory cells that included macrophages, T-lymphocyte subsets, and B-lymphocytes. Neutrophils and eosinophils were identified on paraffin sections stained with H&E and Hansel's method (16), respectively.

Monoclonal mouse immunoglobulin (Ig) G antibodies specific for cytoplasmic component of all guinea-pig tissue macrophages and monocytes (clone MR-1), guinea-pig T-lymphocyte helper/inducer subset (clone CT7) analogous to human CD4+ T-cell subset, T-lymphocyte cytotoxic/suppresser subset (clone CT6) analogous to human CD8+ T cell subset, and pan-B cell (clone MsGp9) were purchased from a commercial source (Serotech Ltd., Oxford, UK). All of the immunostaining was done on an automatic immunostainer (Dako Autostainer) and the antibody binding was detected with the alkaline phosphatase anti- alkaline phosphatase method (17) with naphthol AS-BI and New Fuchsin as substrate. A nonspecific mouse IgG was used as a negative control.

The number of specifically stained cells located in the parenchyma (tissue and air space) or airway wall was determined by point counting at ×400 magnification. Thus, the number of points that fell on stained cells was recorded and divided by the total number of points that fell on parenchyma (or airway wall). Applying the multilevel cascade design (13), this number was then multiplied by the Vv of parenchyma or airway wall. By multiplying the Vv of stained cell for each lung compartment through magnification cascade we determined the total volume of the specifically stained cell in the lung parenchyma and airway wall.

Statistical Analysis

The results were analyzed using a three-way analysis of variance (ANOVA) with two crossed factors nested within a third. The response variables considered for determining the magnitude of the effect of smoking, RA treatment, and Ad infection on the emphysematous process, as well as their interactions (either as interference or synergy), were changes in body weight, lung weight, lung volume, tissue volume, air-space volume, lung SA, and SA/Vol ratio. A Bonferroni (18) adjustment of the independent main effects and interaction effects was used for the seven analyses and a final P < 0.05 indicated a significant difference.

The magnitude of the independent main effects as well as interaction effects of cigarette-smoke exposure and latent Ad infection on changes in the total volume of each inflammatory cell type (neutrophils, macrophages, CD4 cells, CD8 cells, B cells, and eosinophils) on both parenchyma and the airway wall was determined using a two-way ANOVA. A Bonferroni correction of P < 0.05 for a total of 12 comparisons, because there are six cell types on the parenchyma and six cell types on the airway wall, set the level of significance at a P < 0.004.

Fisher's exact test (two-tail) was used to determine differences and the effect on survival of individual treatments (smoking, Ad, and RA) as well as their combined effects. The level of significance was p < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Photomicrographs of histologic samples of the lung from control (Figure 2A) and cigarette smoke-exposed (Figure 2B) guinea pigs show enlargement of the air space and destruction of the centrilobular regions of alveolar tissue in the smoke-exposed animal compared with the normal lung architecture of the nonsmoking control.

View larger version (65K): [in this window] [in a new window]   Figure 2.   Photomicrographs of histologic sections of lungs from (A) air-exposed and (B) smoke-exposed guinea pigs show enlargement of the air space and destruction of the alveolar tissue in the smoke-exposed animal compared with the normal lung architecture in the nonsmoking control. (A) Air exposure for 13 wk. (B) Cigarette-smoke exposure for 13 wk. Bar, 500 µm. RB, respiratory bronchiole; CLE, centrilobular emphysema.

Cigarette-smoke exposure for 13 to 16 wk had no statistically significant effect on body weight or lung weight. The results (Tables 1 and 2) show that smoking increased lung volume (P = 0.04) and air-space volume (P < 0.001) and reduced tissue volume (P = 0.03) and lung SA/Vol ratio (P < 0.001) without significant change in total SA.

Latent Ad infection had no main effect on any of the seven parameters analyzed, but acted synergistically with smoke exposure (Tables 1 and 2) to increase lung weight (P = 0.01) and amplify the effect of smoke exposure on lung volume (P < 0.001), air-space volume (P < 0.001), and the SA/Vol ratio of the lung (P < 0.001).

RA treatment did not reverse the cigarette smoke- induced changes in lung volume, air-space volume, or tissue volume, or the decrease in SA/Vol ratio, in either infected or uninfected groups (Table 1), and showed no independent or combined effect in any of the groups treated with RA (Table 2).

Eleven of the 76 guinea pigs died as a result of bronchospastic reactions during smoke exposure. Seven of these died before 13 wk of smoke exposure and were excluded from the final analysis. The groups in which death occurred are indicated at the bottom of Figure 1, which shows that nine of the 11 deaths occurred in animals that received the combination of cigarette-smoke exposure and RA treatment (P < 0.007).

Results from the HPLC analysis showed that the serum levels of RA were in the range of 2 to 6 nmol/liter and were not different between the control and the experimental animals 24 h after the last RA treatment.

Figure 3 shows results from the quantitative histology of four inflammatory cell types (neutrophils, macrophages, CD4 T-lymphocytes, and CD8 T-lymphocytes) from the lungs of animals with either sham infection or infection with Ad who were exposed to either room air or cigarette smoke. Data are shown as a total volume of a cell type and are expressed as milliliters per kilogram of body mass.

View larger version (24K): [in this window] [in a new window]   Figure 3.   The response of parenchymal and airway wall polymorphonuclear leukocytes, macrophages, CD4 cells, and CD8 cells to cigarette-smoke exposure and latent Ad infection. Bars represent the mean value in each group and error bars represent the standard deviation. Solid bars: parenchyma; open bars: airway wall. * Main effect; ** additive effect.

The results show that cigarette-smoke exposure caused an increase of neutrophils (P < 0.001), macrophages (P < 0.0001), and CD4 T-lymphocytes (P = 0.0009) on both parenchyma and airway wall, yielding an independent main effect of smoking (Table 3) on each of these cell types. B lymphocytes were also increased (values not shown) due to cigarette-smoke exposure but reached significant value (P = 0.004) only in the airway wall (Table 3). Smoking did not cause any increase of CD8 T-lymphocytes (Figure 3 and Table 3) in either the parenchyma or airway wall and had no effect on parenchymal or airway eosinophils (data not shown).

The data also show (Figure 3 and Table 3) that latent Ad infection independently increased parenchymal neutrophils (P < 0.001) and parenchymal and airway wall macrophages (P < 0.002 and 0.003, respectively) and CD8 T-lymphocytes (P < 0.001 and 0.0001, respectively). On the other hand, latent Ad infection had no effect on CD4 T-lymphocytes and B cells. Similar to cigarette-smoke exposure, the presence of latent Ad infection had no effect on parenchymal or airway eosinophils.

The ANOVA revealed that the effects of Ad and smoking were independent and that there was neither interference nor synergy between the effects of smoke exposure and latent Ad infection on any of the inflammatory cell types.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

These data confirm the previous report from Wright and Churg (6) that cigarette-smoke exposure produces emphysematous lung destruction in guinea pigs that is similar to the centrilobular form of emphysema in humans. The procedure used to expose the animals (10) produces mean carboxyhemoglobin levels (19) that are similar to those observed in human cigarette-smokers (20). The animals that completed the smoke exposure were not different from controls with respect to their body weight or eating habits. All of the 11 animals that died during the study were in the smoke-exposed groups, and nine of the 11 deaths occurred when smoke exposure was combined with RA treatment.

Although the changes in lung volume, the decrease in SA/Vol ratio, and preservation of total surface area that resulted from smoke exposure are similar to those caused by the elastase installation (7), RA treatment failed to either prevent or reverse these changes in any of the smoke-exposed groups. The severity of the emphysema produced by elastase treatment in rats (7) and by cigarette-smoke exposure in guinea pigs is consistent with mild human disease (14). Whether the difference between the study by Massaro and Massaro (7) and the present study is due to the stimulus used to cause emphysema or to species differences remains to be determined. One major difference is that the Massaros administered the RA treatment after the elastase installation, whereas treatment was begun at the same time as the smoke exposure and continued through the entire 13 to 16 wk of the present study. If RA had been added after the smoke exposure stopped, fewer deaths may have occurred.

The failure to measure differences in serum RA between treated and control groups by HPLC in the present study is attributed to the length of the time (24 h) between the last exposure and the serum collection. Inasmuch as the dose (500 µg/kg) and the route of administration (i.p.) were exactly the same as in the Massaros' report (7), we assume that the guinea pigs received as much RA in this study as in the earlier study on rats (7).

The presence of latent Ad infection amplified the effect of cigarette-smoke exposure by producing a further increase in lung volume and air-space volume as well as a greater reduction in the SA/Vol ratio. The combined effect of smoking and infection also caused an increase in lung weight without a decrease in tissue volume, which we attribute to greater amounts of fluid and cells in the inflammatory exudate.

The analysis of the inflammatory cells supports the hypothesis that the presence of a latent Ad infection increases the inflammatory response to cigarette smoke. The combination of cigarette-smoke exposure and latent Ad infection produced an additive effect on neutrophils and macrophages, but the CD4 cells increased exclusively due to cigarette-smoke exposure and the CD8 cells increased exclusively in the presence of the latent Ad infection. The increased number of both CD4 and CD8 lymphocytes that has been well documented in chronic obstructive pulmonary disease (21) could well be due to the independent effects of the cigarette-smoke exposure and the presence of an intracellular pathogen such as a latent Ad infection.