INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In utero cigarette smoke exposure has been shown to produce abnormalities in lung function in newborn infants. Tager and coworkers showed that infants born to mothers who smoked during pregnancy had approximately 10% reduced expiratory flow parameters when compared with infants whose mothers did not smoke during pregnancy (1). Young and coworkers showed that infants whose mothers smoked during pregnancy had increased airway responsiveness to inhaled histamine 4 wk after delivery (2). In epidemiologic studies, exclusion of the effects of any postnatal smoke exposure is difficult; however, the use of statistical methods, in which these confounding variables are adjusted for, suggests that increased responsiveness in these infants is primarily associated with in utero cigarette smoke exposure (3). Despite these conclusions, the actual mechanisms resulting in abnormal postnatal lung function after in utero cigarette smoke exposure are unknown. Saetta and coworkers have previously shown that the number of alveolar attachments to the surrounding airway adventitia is reduced in active smokers and that this reduction is associated with a reduction in lung elastic recoil (4). Several studies have reported reduced respiratory function at birth in infants whose mothers smoked during pregnancy and propose altered lung/airway development in utero as a likely mechanism although no attempts were made to test these hypotheses (1). Therefore, we hypothesized that alterations in postnatal lung function observed after in utero cigarette smoke exposure are caused by altered lung/airway structure. The aims of this study were to examine the effects of in utero smoke exposure on postnatal lung function, specifically airway responsiveness, and lung (alveolar attachments) and airway (airway dimensions) structure using a small animal model of in utero smoke exposure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In Utero Model of Smoke Exposure

Nine pregnant Cam Hartley guinea pigs were obtained from a local breeding facility (Monash University breeding facilities, Clayton, Victoria, Australia) between 15 and 22 d postconception. Smoke exposure commenced at Day 28 postconception and occurred for 15 min a day for 4 d each week. Smoke exposure commenced at Day 28 to minimize the risk of failed implantation or spontaneous abortion resulting from the acute effects of cigarette smoke inhalation. The guinea pigs were placed in a 16-L Perspex chamber developed for this project. A cigarette was connected to the access port in the wall of the chamber. A bidirectional syringe pump was also connected to the chamber to enable cigarette smoke to be drawn in. After the lighting of the cigarette, smoke was drawn into the syringe by the pump and the smoke was then returned into the chamber through a further porthole. An oxygen monitor continuously monitored the oxygen concentrations in the exposure chamber and supplemental oxygen was bled into the chamber, as the combustion process continued, to maintain normoxic conditions. At the end of the 15-min exposure, animals were removed and returned to their standard cages. To limit any effects related to the stress of handling, control animals were placed in an identical exposure chamber at the same frequency but received no smoke exposure.

Serum Cotinine Measurements

Blood samples were drawn from each pregnant animal under light halothane anesthesia twice a week for measurement of serum cotinine. Blood samples were centrifuged and the serum removed. Analysis of serum cotinine was performed using a cotinine assay kit (DPC nicotine metabolite assay; Bio-Mediq DPC Pty Ltd., CA). Blood samples were drawn immediately before the first smoke exposure of each week and 24 h after the last smoke exposure each week. To control for any possible effect of the halothane anesthesia on fetal development, control non-smoke-exposed animals were also anesthetized and bled at an identical frequency.

Neonatal Animals

Pregnant animals were allowed to proceed to normal delivery, usually at 68 d gestation. To limit any effects of smoke metabolites that may cross through breast milk, newborn infants were removed from mothers as soon as possible and placed with lactating animals who had not received smoke exposure during pregnancy. Newborn animals received no smoke exposure postnatally and were studied at Day 21 of postnatal life.

Assessment of Airway Responsiveness

Animals were anesthetized with 0.3 ml ketamine and xylazine by intraperitoneal injection. Under supplemental halothane anesthesia a tracheostomy was performed and the animals were then placed in a pressure-sensitive plethysmograph, designed for the guinea pig and ventilated with a Harvard small animal ventilator (model 608; Harvard Apparatus, South Natick, MA) delivering a tidal volume of 3 ml at a frequency of 60 breaths/min. Animals were paralyzed with succinylcholine (0.5 mg/kg, intramuscularly) immediately after attachment to the ventilator. Volume signals were obtained from a pressure-sensitive transducer (MP 45, 2 ± 2 cm H₂O; Validyne, Northridge, CA) which measured changes in box pressure. The plethysmograph was calibrated with a 10-ml syringe, and a 3-ml volume change resulted in a 0.3 cm H₂O pressure change within the box. The pressure response was linear to a volume change of 20 ml. The volume signal was electronically differentiated to determine flow. Pulmonary pressure was determined using a differential pressure transducer (model 267BC; Hewlett-Packard, Waltham, MA) by comparing pressure at the tracheal opening with esophageal pressure measured from a saline-filled PE-90 tube in the distal esophagus. The pressure transducer was calibrated using a water-filled manometer. The response was linear over the range ± 50 cm H₂O. The volume, flow, and pressure signals were recorded (model RS4-5P recorder; General Scanning) and pulmonary resistance (RL) was calculated at 50% of tidal volume from these signals using the method of von Neergard and Wirz (5). After stable baseline RL values were present, guinea pigs were given isotonic saline aerosol followed by increasing doses of aerosolized acetylcholine solution (ACh), and the changes in RL were measured.

Acetylcholine chloride (Sigma Chemical, St. Louis, MO) was dissolved in normal saline to produce a stock solution of 50 mg/ml. Serial dilutions were made to produce solutions of 15, 5, 1.5, and 0.5 mg/ml. Aerosol was generated by placing 5 ml of saline or ACh solution into a Hudson nebulizer (Hudson Oxygen Therapy Sales Co., Temecula, CA) driven by compressed air at 7 L/min and connected by a T piece to the attachment tube for the tracheostomy. The ventilator was disconnected from the circuit during the administration of the ACh. Six 3-ml tidal breaths of nebulized ACh were delivered for each concentration of ACh by intermittent digital pressure on the sidearm of the T piece as we have previously described (6). After the administration of each dose, the ventilator was immediately reconnected to the circuit. The computer analyzes the first 20 breaths after recommencement of ventilation. Peak responses after ACh challenges generally occur within the first 10 breaths after reconnection to the ventilator. The average of the three highest RL values at 50% tidal volume was taken as the response to each ACh concentration. The next dose was delivered after RL had returned to baseline.

Specimen Preparation

At the completion of the assessment of airway responsiveness, the animal was killed and removed from the plethysmograph. The anterior chest wall was then removed and the trachea cannulated though the tracheostomy site. The lungs were then inflated over 60 min with 10% buffered formalin at 20 cm H₂O. After inflation the lungs and heart were removed from the chest cavity en bloc and immersed in formalin. Multiple tissue blocks were cut in a sagittal plane through the perihilar regions of both lungs and embedded in paraffin for histologic processing. Three micron sections were cut and stained with hematoxylin-eosin for morphometric analysis. Slides were examined using a video-linked microscope Leica Laborlux D (Leica, Germany), with the image being projected onto the monitor screen of a 486 DX computer. Images were assessed using the color image analysis program Quantimet 500+ (Leica Cambridge Ltd, Cambridge, UK).

Airway Morphometry

On all airways cut in transverse section (defined as an even thickness of epithelium and an even thickness from the basement membrane to the inner smooth muscle layer and a min/max diameter ratio 1:3), the following areas (A) and perimeters (P) were measured; the internal area and perimeter (Ai and Pi), defined by the luminal surface of the epithelial border; the basement membrane area and perimeter (Abm and Pbm), defined by the basement membrane; the outer muscle area and perimeter (Amo and Pmo), defined by the outer border of the airway smooth muscle; and the outer airway wall area and perimeter (Ao and Po), defined by the outer edge of the adventitia surrounding the airway (10). The number of alveolar attachments to the adventitial surface of the airway was counted for each airway (Figure 1). Where a large pulmonary vessel was adjacent to the airway and no alveolar attachments were possible, the length of the outer perimeter where the airway and vessel were adjacent was subtracted from the total outer perimeter to give the total length over which alveolar attachments could be counted to allow the mean distance between alveolar attachments to be calculated. The area of airway smooth muscle in the airway wall was measured by direct tracing. Airways in which more than 50% of the surface epithelium was missing or which were cut at a bifurcation were not measured. In airways with more than 50% of the epithelium intact, the luminal border of the epithelium was interpolated between two intact areas.

View larger version (108K): [in this window] [in a new window]   Figure 1.   Photomicrograph of an airway from a 3-wk-old guinea pig born after smoke exposure in utero from Day 28 of gestational life to term. Arrows highlight attachment points of the airway.

Calculations

The inner wall area (WA_i) was calculated by subtracting Ai from Amo. The outer wall area (WA_o) was calculated by subtracting Amo from Ao. The area of smooth muscle in the airway wall was normalized by dividing the measured area by the square of the basement membrane perimeter, which has been shown to be independent of muscle contraction and lung volume (8). The length of the outer perimeter was divided by the number of alveolar attachments to obtain the mean distance between attachment points in millimeters. The areas of smooth muscle and airway wall area were divided by the square of the basement membrane perimeter (Pbm).

Data Analysis

The Pbm was used to define airway size. As there were so few large cartilaginous airways available for examination, only airways with a Pbm of less than 4 mm were analyzed. These airways were predominantly membranous and small intraparenchymal cartilaginous airways. The results were expressed as the mean ± SE. Where data were normally distributed, differences between the two groups were tested using an unpaired t test, and where data were not normally distributed, differences were tested using a Mann-Whitney U test and expressed as median and 25-75 percentiles. A probability of < 5% was considered significant.

Intraobserver error was expressed as the coefficient of variation, and was calculated for measurements made on 6 airways 6 times, as previously described by Carroll and coworkers (9). All measurements were made by the one observer who was blinded to the case classification.

The study was reviewed by and approved by the Royal Children's Hospital Animal Experimentation Ethics committee.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Results were obtained from 17 animals (7 males, 10 female) born to six guinea pigs exposed to cigarette smoke during pregnancy. These results were compared with findings from eight animals (5 males, 3 female) born to three guinea pigs not exposed to cigarette smoke during pregnancy. In utero smoke-exposed animals had a significantly lower weight at birth compared with nonexposed animals (p < 0.05); however, the in utero smoke-exposed animals displayed accelerated postnatal growth. By Day 21 when airway responsiveness was measured there was no significant difference between the two groups in mean weight (271 g ± 10 g [smoke-exposed] versus 293 g ± 8 g [controls]) (p = 0.11).

Smoke Exposure during Pregnancy

Serum cotinine levels during pregnancy in the smoke-exposed group showed significant variation through the week after the 4-d exposure pattern. Serum cotinine levels peaked at Day 4 of exposure and had reached nonrecordable concentrations by the commencement of the following week's exposures. Peak serum cotinine levels ranged from 22 ± 12 ng/ml to 76 ± 12 ng/ml at Day 35 of gestation whereas all animals reached nonrecordable levels by the end of each week. No recordable cotinine concentrations were obtained from control animals.

Airway Responsiveness

There was no difference in baseline pulmonary resistance between the smoke-exposed animals and control animals. There was increased airway responsiveness (p < 0.05) to inhaled acetylcholine at higher doses in the animals that received smoke exposure during pregnancy compared with control non-smoke-exposed animals (Figure 2).

View larger version (12K): [in this window] [in a new window]   Figure 2.   Pulmonary resistance values after administration of 6 breaths of increasing concentrations of nebulized acetylcholine to 21-d-old guinea pigs who were exposed to cigarette smoke in utero (n = 17) or non-smoke-exposed controls (n = 8). *p < 0.05, **p < 0.01.

Airway Morphometry

A mean number of 8 ± 2 airways were measured from each animal. Pbm was the same (1.32 mm ± 0.11 mm versus 1.35 mm ± 0.08 mm) in the nonexposed and exposed animals, respectively, showing that similar sized airways were compared. Po adjusted for any lengths where adjacent pulmonary vessels prevented alveolar attachments, was also similar in the two groups (0.94 mm ± 0.06 mm nonexposed versus 1.0 mm ± 0.05 mm exposed), validating the expression of attachment points in this way. The inner airway wall area, the outer airway wall area and the area of airway smooth muscle normalized by dividing by Pbm² are shown in Table 1. Compared with nonexposed animals, the outer airway wall area was increased by 20% and the inner airway wall and smooth muscle areas by 10% in the smoke-exposed animals although these differences were not significant. The mean distance between alveolar attachment points was increased (p = 0.001) in the smoke-exposed animals (0.052 ± 0.009 mm¹) compared with nonexposed animals (0.046 ± 0.009 mm¹) (Table 2). There was no difference between the two groups for the amount of smooth muscle shortening observed. The coefficient of variation for airway measurements ranged from 0.68% to 3.2% with a mean value of 1.8% ± 0.5%.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, in utero exposure to cigarette smoke during pregnancy, in the absence of postnatal cigarette smoke exposure resulted in increased airway responsiveness to inhaled acetylcholine at Day 21 of life in the neonatal guinea pig. This increase in airway responsiveness was associated with a significant increase in the mean distance between alveolar attachment points in membranous and small cartilaginous airways, which was due to a combination of fewer alveolar attachments and an increase in the outer perimeter of the airway wall. The inner airway wall area and the airway smooth muscle area were increased by 10% and the outer airway wall was increased by 20% in the smoke-exposed animals compared with the nonexposed animals.

When interpreting the findings in this study, one needs to carefully consider a number of technical and methodological factors. First, we are confident that the level of smoke exposure to pregnant animals in this study was similar to that seen in studies of maternal smoking. To this extent, the pregnant animals exposed to cigarette smoke during this study had peak serum cotinine levels in the range of 50 to 80 ng/ml. This level approximates the concentrations seen in humans who smoke between 5 and 10 cigarettes a day (10). We believe therefore that the degree of fetal smoke exposure produced in this study is similar to the degree of exposure experienced by human fetuses whose mothers smoke to a moderate degree.

Second, for valid comparisons of morphologic data between different animals it is important that airways of similar size and dimensions are analyzed. When the frequency distribution of internal airway and outer airway perimeters was compared there was no difference between the groups, suggesting that similar sized airways were examined in each group. These data are also supported by the means and standard errors for both of these measurements, within any given airway size group, which were virtually identical for the two groups. Thus, we feel confident that the methods used to normalize the data in this study are valid. Similar methods have been used previously and the matching of airways by order or generation is not possible when lung tissue is collected by cutting lung lobes through the sagittal plane to sample airways (8, 9).

There are a number of studies in which lung function has been measured in infants close to birth whose mothers smoked during pregnancy (1, 11). These studies report reduced respiratory function at birth in such infants and propose altered lung/airway development in utero as a likely cause although no attempts were made to test these hypotheses. Several previous animal models have documented abnormalities in lung development after smoke exposure during pregnancy. Our animal study examined the effect of smoke exposure during pregnancy on postnatal lung function and on postnatal lung structure several weeks after birth in an attempt to elucidate possible mechanisms for this altered lung function. Collins and coworkers studied rats and found that in pregnant animals exposed to cigarette smoke from Day 5 to term at Day 21, their offspring were not only smaller but had reduced lung elastic tissue, and increased size but reduced number of saccules (fetal alveoli) (12). They did not examine the effects of these structural changes on postnatal lung function nor did they examine the effect of postnatal lung growth in a smoke-free environment.

Changes in lung or airway structure, or both, as a result of in utero cigarette smoke exposure might give rise to increased responsiveness to an inhaled smooth muscle agonist in a number of ways. In vivo, when airway smooth muscle is stimulated, it must overcome a number of forces that act to oppose muscle shortening. One of the major loads that smooth muscle must overcome when shortening in vivo is the load imparted on the muscle owing to the elastic recoil pressure of the lung parenchyma (13). This elastic load imparted on the muscle is translated to the airway wall through alveolar attachments to the airway. Thus, any reduction in the number of alveolar attachments to the airway may reduce the loads opposing smooth muscle shortening and allow increased shortening for the same stimulus (14). Similarly, an increase in the thickness of the airway wall outside the smooth muscle layer will have the same effect and reduce the effect of lung elastic recoil pressure when smooth muscle shortens. On the other hand, an increase in the amount of smooth muscle may enable the muscle to generate more force when stimulated to shorten, whereas an increase in the thickness of the airway wall inside the smooth muscle layer, will result in greater luminal narrowing, for the same degree of smooth muscle shortening. Therefore, it is plausible that the net effect of the modest increases in airway wall thickness and smooth muscle area seen in this study, coupled to the significant reduction in the mean distance between alveolar attachments (and therefore the number per unit area), might result in airway hyperresponsiveness such as that observed in the smoke-exposed animals.

Saetta and coworkers have previously shown a reduction in the number of alveolar attachment points, an increase in the distance between attachments, and an increase in the percentage of abnormal attachments in cigarette smokers compared with nonsmokers. This reduction was associated with an increased score for airway inflammation and with reduced elastic recoil pressure in smokers (10). Airway responsiveness was not assessed in the study of Saetta and coworkers. Petty and coworkers have also shown that in patients with mild emphysema (i.e., destruction and loss of the alveolar walls) elastic recoil pressure is reduced but that this is not associated with airflow limitation (15). The destruction of alveolar walls observed in patients with moderate to severe emphysema associated with cigarette smoking is associated with airflow obstruction and increased airway hyperresponsiveness. These studies have generally been performed in adults whereas our study examined the effects of passive cigarette smoke exposure in utero in a developing lung system. We did not examine airway or alveolar inflammation in this study, so whether the decreased number of alveolar attachment points in smoke- exposed animals in this study was the result of cellular infiltration and destruction of existing alveoli or abnormal growth in utero is unknown. Unlike the study by Saetta and coworkers we made no attempt to assess the degree of damage to existing alveolar attachments. Structural changes to the alveoli such as a reduction in elastic fibers or collagen deposition (fibrosis) of alveolar walls, although not measured in this study, might also alter lung function or response to inhaled agonists.

We did not observe any difference between the smoke- exposed and nonexposed animals in the amount of smooth muscle shortening observed, despite a significant increased response to inhaled acetylcholine at the higher doses in the smoke-exposed animals. We can think of two possible explanations for this observation. First, the lungs were inflated to a pressure of 20 cm H₂O with fixative before histologic examination. This distending pressure may reduce the amount of muscle shortening observed after fixation and may have been preserved if the lungs had been fixed in immersion without inflation. Carroll and coworkers compared smooth muscle shortening in patients with fatal asthma with nonasthmatic control patients dying of nonrespiratory causes in whom the lungs were inflated to a pressure of 25 cm H₂O and found no difference between the groups (16). We made a conscious decision to inflate the lungs in this study to increase our ability to clearly delineate alveolar attachments that are more accurately measured in the inflated lung. A second possible explanation for our findings is that the relatively modest increase in the inner airway wall thickness in the smoke-exposed group resulted in exaggerated luminal narrowing at high doses of acetylcholine. This is supported by studies that have modeled the effects of airway wall thickening in patients with asthma and showed that the most pronounced effects of increased thickness of the inner airway wall and luminal narrowing are seen at higher degrees of muscle shortening (14). This might explain why the changes were seen at the highest doses where presumably the smooth muscle is highly stimulated.

Young and coworkers examined postnatal lung function in 63 infants at 4.5 wk of age and found a strong association between increased airway reactivity and exposure to cigarette smoke in utero (2). Other investigators have also documented abnormalities in neonatal respiratory function, which are associated with maternal smoking during pregnancy (1, 3). Similarly, in the present study we observed increased airway responsiveness to acetylcholine at Day 21 postnatally in guinea pigs whose mothers had been directly exposed to cigarette smoke during pregnancy but in the absence of postnatal smoke exposure. This suggests that the mechanisms resulting in increased airway responsiveness may develop in utero in response to passive smoke exposure. To what extent the findings in guinea pigs can be extrapolated to humans is not clear; however, we feel confident that the findings from the present study provide important new information about possible mechanisms to explain the deleterious effect of in utero smoke exposure on postnatal lung function.

In utero smoke exposure could be considered as a distinct form of "passive" cigarette smoking in that the fetus is not directly exposed to cigarette smoke and thus any effects on fetal development are a result of indirect exposure to cigarette smoke. A variety of hypotheses to explain such changes have been postulated, including factors such as alterations in placental blood flow, altered cortisol levels, and changes in fetal breathing patterns, all of which may influence fetal lung development (17), but these were not assessed in this study.

In conclusion, neonatal guinea pigs born to animals exposed to cigarette smoke during pregnancy show increased airway responsiveness at Day 21 of postnatal life. Examination of lung structure at this stage shows a significantly increased mean distance between alveolar attachment points (a combination of fewer attachments and an increased outer airway wall perimeter) and modest changes in airway dimensions, suggesting that the observed increase in airway responsiveness may be related to altered airway and lung structure secondary to in utero cigarette smoke exposure.