Abstract
The direct effect of environmental tobacco smoke (ETS) exposure in utero on the development of the lung parenchyma is not known. We used design-based stereologic methods to evaluate in utero and postnatal ETS exposure on alveolar and respiratory bronchiole (RB) development in the rhesus macaque.
Methods
Timed-pregnant rhesus macaques and their offspring were exposed to filtered air or various amounts of ETS during the prenatal and postnatal period. The left cranial lobe from necropsied infants was evaluated by design-based stereological methods and general pathological review.
Results
Infants in the in utero and six-month ETS groups had an 18% and 17% relative decrease, respectively, in alveolar number and a 57% and 33% increase, respectively, in alveolar size compared to filtered air (FA) monkeys. Lung volume positively correlated with alveolar number in the FA and six-month ETS group and negatively correlated in the in utero ETS group. The distribution of alveolar size was much more variable in the in utero group. Overall, RB volume was significantly increased in the six-month ETS group (p < .04).
Conclusions
Taken together, these results indicate that in utero and postnatal ETS exposure is associated with altered parenchymal lung development.
Introduction
Although the overall prevalence of smoking has decreased in the past several years, exposure to secondhand environmental tobacco smoke (ETS) continues to be a problem. Recent studies of cotinine and cotinine metabolites in the general population have confirmed that ETS continues to be a factor for a majority of the population (Kandel et al. 2007). Although several studies have demonstrated a deleterious effect of maternal smoking on fetal development, the effect of ETS on the development of the lung parenchyma is not known. The development of the lung parenchyma, including alveoli and the small airways, begins to mature during the late fetal period. Interruptions of this stage could lead to chronic respiratory problems during infancy and early childhood. Recently, studies of alveolar development in nonhuman primates have confirmed that alveolar development continues after birth and well into early adulthood (Hyde et al. 2007). These same studies have revealed that overall lung volume is the best predictor of alveolar number. Fetal exposure to ETS during the intrauterine period is by a systemic effect through the mother’s circulation, whereas exposure to ETS after birth is by direct inhalation of unfiltered secondhand smoke. The relative contributions and consequences of these two separate exposures on lung parenchymal development are not clear. The purpose of this study was to evaluate the effect of ETS on alveolar and small airway development. The study was designed to address the additional influence of in utero ETS exposure to postnatal ETS. Using design-based stereologic methods, we examined archived tissue from fifteen infant rhesus macaques exposed to varied amounts of ETS. The aim of the study was to quantify alveolar and respiratory bronchiolar size and number to establish the effect of ETS on parenchymal lung development.
Methods
Animal Protocol
The animal protocol timeline is outlined in Figure 1. Briefly, fifteen timed-pregnant rhesus monkeys and their offspring were exposed to filtered air (FA) or to aged and diluted side-stream cigarette smoke as a surrogate to environmental tobacco smoke (a total suspended particulate concentration of 0.99 mg/m3 for six hours/day, five days/week). Three separate treatment groups were defined (n = 5/group). The dams used in this study ranged in age from five to nine years and had had prior successful pregnancies. Research cigarettes (1R4F), obtained from the Tobacco Research Institute at the University of Kentucky, were used. The first group was kept in an FA environment from forty-five to fifty days gestational age (DGA) to thirteen months postnatal age (PNA). The second group was exposed to ETS from forty-five to fifty days DGA to thirteen months PNA. The third group was kept in a FA environment from forty-five to fifty days DGA to six months PNA, at which point the monkeys were transferred to an ETS exposure environment to thirteen months PNA. All animal care was delivered by licensed veterinarians at the California National Primate Research Center (CNPRC). For necropsy the monkeys were deeply sedated with intravenous sodium pentobarbital dosed at the discretion of the attending veterinarian. The research protocol was approved by the institutional animal care committee. A similar experiment protocol has recently been published (Zhong et al. 2006).
Necropsy and Lung Preparation
Section on Lung Necropsy
The whole lung was separated into lobes, and the left cranial lobe was fixed by infusion of 1% glutaraldehyde–1% paraformaldehyde in cacodylate buffer (440 mosm, pH 7.4) at 30 cm fluid pressure for four hours.
Lobe Volume
The fixed left cranial lobe volume (VL) was determined by estimating its buoyant weight in PBS (Scherle 1970).
Sampling
After fixation and lobe volume determination, the lung was sampled so as to obtain isotropic, uniform, random (IUR) oriented samples (Hyde et al. 2007). The lungs were first embedded in agar and oriented using the Orientator (Mattfeldt et al. 1990). The lung was sliced and sampled using the Smooth Fractionator (Gundersen 2002; Hyde et al. 2007; Hyde et al. 2004). The sampled blocks were embedded in paraffin; serial sectioned (four to six sections), and stained using hematoxylin and eosin. Eight to ten blocks were sampled for each animal.
Estimation of the Volume of Parenchymal Components
We used Slidebook 4.1 software (Intelligent Imaging Innovations, Inc., Santa Monica, CA, USA) to sample 5 μm sections at a magnification of 185X. Images were collected using a spacing of 2200 μm and converted to tiff files. The tiff images were counted using a double lattice test system of 25/100 points. The volume densities of parenchyma (P) and nonparenchyma were determined, and these compartments were further divided into alveoli, terminal bronchioles, and respiratory bronchioles (RB) in the parenchymal compartment and into large airways, arteries, and veins in the non-parenchymal compartment. Absolute volumes were obtained by multiplying the volume density of a component by the volume density of parenchyma and the lobe volume. For the case of the respiratory bronchiolar volume, the calculation is:
Estimation of Number of Alveoli in the Left Cranial Lobe
Alveolar number was estimated by counting the number of alveolar entrance rings that are represented as the free edges of interalveolar septa in a physical dissector using the method of Hyde et al. (2004). We used fractionator sampling, where Δχ is the number of alveolar openings per sampled fraction and SF is the total sampling fraction, the product of the sampling fractions of tissue bars, blocks, dissector heights, and areas at subsequent sampling levels. The number of alveolar openings was estimated by counting the presence of bridges connecting the free edges of interalveolar septa or islands (newly appearing interalveolar septal tissue) in the sampling section, but not the look-up section, of the dissector. Bridges (B) and islands (I) were counted using the Computer-Assisted Stereology Toolbox software system (CAST, Visiopharm, Horsholm, Denmark).
The number of alveoli was calculated as follows:
Sampling fraction (SF) is the product of the sampling fractions (sf) of bars, blocks, section heights per block, and filed areas per section.
Calculation of Number-weighted Volume
The “number-weighted” mean alveolar volume, v̄ n alv, is estimated indirectly by dividing total volume of alveolar airspaces (excluding that of alveolar ducts), V(alv,lung), obtained by point counting, by total alveolar number, Nalv,lung, obtained by the dissector:
The units were converted to μm3.
Estimation of Volume-weighted Volume
“Volume-weighted” mean alveolar volume (v̄ v alv) was calculated using the point-sampled intercept method that estimates the volume of structures provided the tissue is sampled under IUR conditions (Gundersen and Jensen 1985; Hyde et al. 2007). The CAST software system was used to perform the measurements and calculations:
where Fsgv is a correction for shrinkage (μm3) as a result of processing and l̄ − 3 IAS is directly measured on the CAST software system.
Airway Tethering
Longitudinal cross-sections of respiratory bronchioles were captured using an Olympus BX61 light microscope and Slide-book 4.1 software. The perimeter of each RB was measured, and alveolar attachments to the RB wall (the proportion of the abluminal airway wall anchored by interalveolar septal tissue) were quantitated using the NIH Image J stereology toolbox. The tethering index is expressed as number of interalveolar septal attachments per mm length of RB wall.
Tissue Pathology
Tissue sections that were used for stereological analysis were also stained with H&E for review by a pathologist. The analysis was performed in a blinded fashion, where each tissue section was semiquantitatively examined for degree, type, and distribution of inflammation. A range of five to ten fields from each of nine to twelve parenchymal sections per animal were analyzed. Inflammation was graded as present if greater than 50% of the fields within a section had inflammatory cells in the form of macrophages or leukocytes. The inflammation was categorized as Bronchus Associated Lymphoid Tissue (BALT), bronchial inflammation, parenchymal inflammation, or predominant alveolar macrophages. The pathology scores were then sorted by treatment group.
Results
Pathology
Representative images from each group are shown in Figure 2. Inflammatory changes were graded and classified as described in the Methods section. In general, the changes in all of the animals were very mild. The majority of sections that were reviewed demonstrated very little inflammation. The lungs from animals exposed to FA showed essentially normal histological features. One animal from this group showed very focal acute pneumonia of less than forty-eight hours duration in two of nine sections, and one animal showed focal chronic inflammation (eight of eight sections). Three of the five lungs from animals receiving both prenatal and postnatal ETS showed increased alveolar macrophages, and two of the five showed very focal, nonspecific chronic inflammation (present only in one of nine and two of six sections). Lung microanatomy was otherwise normal. All of the animals in the group receiving postnatal ETS showed prominent collections of alveolar macrophages. There was multifocal alveolar, interstitial, or perivascular chronic inflammation present in three animals (nine of nine, ten of ten, and nine of ten sections, respectively). One animal showed focal chronic inflammation (one of ten sections), and one was normal except for the increased numbers of alveolar macrophages. A semiquantitative graph of the pathological findings is represented in Figure 2D, which demonstrates that the most statistical difference was seen in the abundance of alveolar macrophages in the postnatal group.
Changes in Alveolar Number and Size
Alveolar number and volume were quantitated as previously described (Gundersen 2002; Hyde et al. 2004). The correlation between alveolar number and lung volume proved to be significantly different between the groups. The trend was toward a smaller number of alveoli in the two ETS treatment groups (Figure 3A). Accordingly, the tethering of noncartilaginous airways was decreased in the same two treatment groups (Figure 3B). Alveolar volume was determined by two different methods termed volume weighted (v̄ n alv) and number weighted volume (v̄ v alv). The overall trend was an increase in alveolar size in the group exposed to ETS in the in utero and postnatal period (in utero group). As shown in Figure 4A, the group exposed to ETS in the in utero and postnatal period has a steeper increase in size when calculated by the volume weighted method as compared to the number weighted method. This finding is representative of the greater distribution of alveolar sizes in this treatment group. There is a positive correlation between alveolar number and overall lung volume in the FA group and postnatal ETS group, but a negative correlation between alveolar number and lung volume in the in utero ETS group (Figure 4B). This trend was confirmed when the number of alveoli was plotted against total surface area (Figure 4C). This finding represents a change in the normal relationship between alveolar number and overall lung volume and alveolar surface area and is statistically different (p < .001). Taken together, these results demonstrate a larger distribution of alveolar volumes with a relatively smaller number of alveoli in the group that was exposed to ETS in utero.
Changes in Terminal and Respiratory Bronchioles
The relative differences in overall volume of nonparenchyma (conducting airway) and parenchymal are shown in Figure 5A. There were no statistical differences between treatment groups. The relative volumes of respiratory and terminal bronchioles are represented in Figure 5B. There was a statistically significant increase in respiratory bronchiole volume in the group that received ETS postnatally.
Discussion
This study was designed to examine the effects of ETS during the prenatal and postnatal periods. The focus of the study was the quantitative assessment of parenchymal lung growth including alveoli and small airways. Fifteen pregnant rhesus macaques were identified and divided into three separate exposure groups. The exposures began early during the gestational period and continued until the thirteenth postnatal month. The three groups consisted of a FA group; an in utero group, which was exposed to ETS during the entire prenatal and postnatal periods; and the six-month group, which was exposed to ETS in the last six months of the study. This study was designed to address the differential influence of ETS exposure during intrauterine development and the early postnatal period versus later in the postnatal period. The in utero group received three separate types of ETS exposure. One type was the inhalation of the ETS during the postnatal period. The other types consisted of systemic exposure from the mother during the prenatal period as well as exposure from the mother’s milk during the early postnatal period. The stages of development covered by this exposure timeline are as follows: fetus, 45–165 days gestation; newborn, twenty-four hours postnatal; neonate, up to one month; infant, one to twelve months; and juvenile, twelve to twenty-four months. This study was designed to match the real-life exposures that an infant may encounter. Infants born to smoking mothers have exposure to ETS in utero and in the postnatal period. Infants born to nonsmoking mothers will have their exposure occur later in the postnatal period.
Lungs from the offspring of the fifteen female monkeys obtained at time of necropsy were examined by a pathologist in blinded fashion. These lungs were semiquantitatively scored for degree and type of inflammation; the lungs from the six-month ETS exposure group had the highest degree of parenchymal inflammation. Macrophages were the dominant cell type described. One explanation for these findings may be the effect of systemic nicotine stimulation of the nicotinic acetylcholine receptor (alpha7 nAChR). Matthay et al. recently reported that stimulation of alpha7 nAChR with nicotine resulted in a decrease in excess lung water and protein concentration in the Bronchoalveolar Lavage (BAL) (Su et al. 2007). Similar findings have been reported with overall amount of surfactant protein after nicotine stimulation. Since the in utero monkeys received ETS exposure by systemic and inhalation means, their overall nicotine exposure may have been higher, which may have led to a decreased inflammatory response to postnatal ETS exposure. This finding has been demonstrated in other studies, where stimulation of alpha7 nAChR with nicotine has resulted in a decrease of Tumor Necrosis Factor (TNF) alpha release in response to LPS as well as a decrease in toll receptor 4 expression on monocytes (Giebelen et al. 2007; Hamano et al. 2006). Moreover, a similar cohort of infant rhesus macaques exposed to ETS in utero and in the postnatal period had a decrease in the TH1 cytokine gene expression in the lungs (Yu et al. 2008).
Recent work from our lab has established that alveolar development in the rhesus macaque continues into the adult period. This study demonstrated that alveolar number continues to grow, whereas the size of alveoli remains relatively unchanged (Hyde et al. 2007). Using the same stereological approach, alveolar number and size were estimated for each of the three exposure groups. Compared to the FA group, the overall number of alveoli was less for the two ETS groups. In the six-month ETS group and the FA group, the number of alveoli correlated with the overall surface area and volume of each lung. The in utero group had a reciprocal relationship between alveolar number and lung volume. Not only was this finding opposite of the other two groups, but it was also opposite to the normal relationship that exists during this stage of development (Hyde et al. 2007). Moreover, the distribution of alveolar size was greater in the in utero groups, further supporting the idea that alveolar development has been disrupted. These results are slightly different from a previous study of pregnant rhesus monkeys treated with nicotine during the prenatal period (Sekhon et al. 1999). In that study, lungs from the offspring of nicotine-treated monkeys demonstrated larger alveoli that were fewer in number, with an overall decrease in lung volume. These findings would suggest that in utero exposure to ETS leads to an interruption of programmed alveolar development, which may lead to an inability to support adequate wound healing of postnatal lung injury. This hypothesis is supported by a study of rhesus macaque during the neonatal period, which demonstrated increased apoptosis with ETS exposure in lung parenchyma (Zhong et al. 2006).
The most statistically significant morphometric finding in this study was the increase in RB volume in the group exposed to ETS during the last six months of the postnatal period. This is a similar finding to a previous study from our group, which involved ozone exposure of bonnet macaques (Fujinaka et al. 1985). This study showed an increase in respiratory bronchiole volume correlated with the inflammatory response to ozone exposure. The six-month ETS exposure had the highest degree of parenchyma inflammation, which correlates with the measured increase in respiratory bronchiole volume. The in utero ETS exposure group had the same amount of postnatal ETS exposure as the six-month group but did not have an increase in the volume of respiratory bronchioles. This finding may be owing to the decreased level of inflammatory response to ETS. Other possibilities include the possibility of key growth factors being affected differentially by in utero versus postnatal ETS exposure. The current study was not designed to directly investigate causal relationships, but rather to identify a key time point in alveolar development. The causal relationships will be explored in future studies.
This study supports the findings of previous work that has focused on the effect of ETS on lung development. Our study demonstrates that ETS exposure in utero disrupts normal alveolar development and leads to an altered repair phenotype of postnatal injury to the lungs. As a result, a pre-emphysematous phenotype is formed. The molecular basis for these findings is not established but may involve the trigger of apoptotic gene expression leading to increased cell death, alteration of the vascular development, or a primary inflammatory process. Most importantly, this work highlights the potential consequence of secondhand tobacco exposure in pregnant mammals and further supports the efforts to limit such exposure.