Main

Adverse effects of active maternal smoking during pregnancy on fetal development have been well established in both clinical and animal studies. Fetal growth retardation, increased rates of spontaneous abortion, reduced postnatal growth, and behavioral and cognitive abnormalities have been documented(1–10). Depending upon the population in question, up to 40% of pregnant women are active smokers(1, 8). In contrast, one-third to one-half of pregnant women are exposed to ETS(11–13). ETS is made up of two components, SS, the smoke which is emitted from the end of a burning cigarette, and exhaled MS, the smoke which is inhaled through a burning cigarette. Two major components of ETS, nicotine and carbon monoxide, have been shown to be 2.7 and 2.5 times higher in concentration, respectively, in SS, the major component of ETS, when compared with MS(12, 13). For these reasons, concern has also been raised about the effects of maternal passive exposure to tobacco smoke during pregnancy on the fetus(8–11, 14), as well as the effects of ETS on infants and young children(12–15). A variety of studies have examined the effects of ETS on birth weight. Although some studies conclude that maternal passive smoking during gestation increases the risk of fetal growth retardation, other investigations do not support this conclusion. Chen and Petitti(11) have recently studied this question, and they have reviewed this literature. Maternal passive exposure to smoking during gestation has been shown to increase the risk of cognitive impairment of the offspring(8–10). Increased risks of otitis media, asthma, and other respiratory illnesses have been demonstrated in children exposed to ETS(12, 13, 15), and an increased risk of sudden infant death syndrome in ETS-exposed infants has been suggested(16, 17).

Although a number of animal studies of the effects of MS on pregnancy outcome have been reported, animal research addressing the effects of ETS on fetal development has been limited(18–20). With respect to the neuro-developmental consequences of exposure to tobacco components, the effects of maternal nicotine exposure on brain development and function have been investigated. In particular, offspring of rats exposed to nicotine demonstrated subtle alterations in behavior(21–23). In addition, alterations in catecholaminergic systems and nicotinic cholinergic receptor development have been observed in offspring of nicotine-exposed dams(24–26).

An established method of determining the teratogenic effects of a substance on the CNS involves quantifying DNA, protein, and cholesterol in brain regions and then using these measurements as indicators of brain growth and development(27–30). In particular, prenatal maternal nicotine exposure has been shown to cause a reduction in fetal cerebellar DNA content, indicating a reduction in the number of cerebellar cells(24, 31–33).

As human maternal ETS exposure has been shown to cause fetal growth retardation as well as cognitive impairment in offspring(8–11), it is important to assess the effects of ETS on fetal brain development. Because maternal gestational exposure to nicotine resulted in alterations in the biochemical composition of brain, similar alterations may occur after maternal gestational exposure to ETS. We report the results of a study of both prenatal and postnatal exposure to SS on the biochemical composition of the brain.

METHODS

SS exposure system. ETS consists of a mixture of exhaled MS and SS(12, 13). We used a SS exposure system to expose animals to ETS(34). Dilute SS was generated by a modified ADL/II smoke exposure system (Oakridge National Laboratory) using conditioned 1R4F cigarettes from the Tobacco and Health Research Institute of the University of Kentucky. To ensure consistency in burning, cigarettes were conditioned by maintaining them in a desiccator jar at a temperature of 23°C and relative humidity of 60% for a minimum of 48 h before use. Two cigarettes at a time were smoked in a staggered fashion under Federal Trade Commission conditions of 1 puff (35 mL, 2-s duration) per min. The MS was collected on a filter and discarded. The SS was diluted with FA in a mixing chamber and was then passed into the exposure chamber. The exposure chamber was characterized by a relative humidity of 41.8 ± 7.2%, a temperature of 24.4 ± 0.8°C, a total suspended particulate concentration of 1.00 ± 0.07 mg/m3, a carbon monoxide concentration of 4.9± 0.7 ppm, and a nicotine concentration of 344 ± 85μg/m3 (mean ± SD). This concentration of carbon monoxide would be typical of a restaurant, bar, or other public gathering place where smoking is permitted, whereas the total suspended particulate and nicotine concentrations are roughly 10 and 30 times higher, respectively(12, 13).

Animal exposures to SS. The following protocol was approved by the University of California, Davis Animal Care and Use Administrative Advisory Committee. The study was designed to yield four exposure conditions:in utero FA followed by postnatal FA (FA/FA), in utero SS followed by postnatal FA (SS/FA), in utero FA followed by postnatal SS (FA/SS), and in utero SS followed by postnatal SS (SS/SS). Pregnant Sprague-Dawley rats (Charles River, Wilmington, MA) were randomly assigned to be exposed to either FA or SS for 4 h/d, 7 d/wk from d 3 of gestation until delivery. After birth, the female pups were randomly assigned to litters of 13-14 pups/dam which were then exposed to either FA or SS postnatally. At 21 d of life, the pups were weaned, and 8-13 in each exposure condition were continued on their same exposure regimen until 9 wk of life. The animals were weighed every 6 to 9 d during the postnatal exposure period. At 9 wk of life, the animals were killed, and the brains were removed and divided at the pontomesencephalic junction into forebrain and hindbrain portions. Each portion was weighed and then frozen at -70°C for later biochemical analyses of DNA, protein, and cholesterol.

Biochemical assays of tissue DNA, protein, and cholesterol. Tissue specimens were diluted 1:20 in 50 mM sodium phosphate, 5 mM EDTA, pH 7.4 (buffer 1), and homogenized. For the DNA assay, a 100-μL aliquot was diluted with 900 μL of 50 mM sodium phosphate, 5 mM EDTA, 2.2 M NaCl, pH 7.4 (buffer 2), and stored at -70°C. For the protein assay, a 100-μL aliquot of the homogenate was diluted with 400 μL of buffer 1 and stored at-70°C. For the cholesterol assay, the remainder of the homogenate was centrifuged at 13 500 rpm, 4°C, for 40 min. The pellet was then dissolved in 1 to 3 mL of buffer 1, and tissue lipids were extracted with 2:1 chloroform-:methanol. The lipid extracts were frozen at -70°C for subsequent analysis of cholesterol.

Standard techniques for measuring DNA, protein, and cholesterol were used. Tissue DNA content was assayed with the Hoechst 33258 (bis-benzimide, Sigma Chemical Co., St. Louis, MO) method(35) using a fluorescence spectrophotometer (SLM Instruments, Urbana, IL), with excitation and emission wavelength settings at 356 and 458 nm, respectively. Tissue protein concentration was assayed via the method developed by Bradford(36) using a Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA), and a UV-visible spectrophotometer (Beckman, Fullerton, CA) with a wavelength setting of 595 nm. Tissue cholesterol was measured via the method of Searcy and Bergquist(37) using a UV-visible spectrophotometer (Beckman, Fullerton, CA) with a wavelength setting of 490 nm.

These biochemical measurements can serve as indices of brain development. Specifically, DNA concentration (mg/g tissue weight) represents cellular density, cholesterol concentration (mg/g tissue weight) is a measurement of myelin density, and protein concentration (mg/g tissue weight) represents protein density, an indicator of general growth. Ratios between certain measurements provide other information about brain growth. The protein/DNA ratio represents cell size, whereas the cholesterol/DNA ratio represents myelinization and arborization per cell(27–30).

Statistical analysis. Results are presented as mean ± SD. Data were analyzed with Minitab Statistical Software for Windows (Minitab, Inc., State College, PA). A general linear model for an unbalanced 2-way analysis of variance was used for all analyses. Two factors, in utero (prenatal) exposure to FA or SS and postnatal exposure to FA or SS together with an in utero exposure by postnatal exposure(interaction) term, were used for this model. For all analyses, results were considered to be significant at the p < 0.05 level.

RESULTS

For all analyses performed, there were no significant effects of in utero SS exposure, and there were no significant in utero exposure by postnatal exposure interactions. Postnatal exposure to SS significantly increased the mortality of the pups during the first 18 d of life with 43% of the pups exposed to postnatal SS dying compared with 14% of the pups exposed to postnatal FA (χ2 = 12.4, p < 0.001).

Postnatal exposure to SS resulted in a significant reduction in animal weight at 9 wk of age (p = 0.012, F1,44 = 6.91). The weights of the animals in the SS/SS group were the lowest, whereas those in the FA/SS group were reduced to a lesser extent (Table 1). Because there was no in utero exposure by postnatal exposure interaction, these data were pooled into two groups (animals with and without postnatal SS exposure). By comparing these two groups, postnatal exposure to SS resulted in a 6.4% reduction in body weight (p = 0.016,F1,44 = 6.23). The difference in body weight was not present at birth and became apparent only after 6 wk of age (data not shown). SS exposure had no effect on total brain weight, forebrain weight, and hindbrain weight, whereas the total brain weight/body weight ratio was increased in animals exposed to SS postnatally (p = 0.001, F1,44= 11.99).

Table 1 Effects of in utero and/or postnatal SS exposure on body weight and brain weight
Full size table

The effect of SS exposure on the biochemical composition of the forebrain and hindbrain was limited to alterations of DNA concentration and the protein/DNA ratio. In the forebrain (Table 2), there was a significant reduction of DNA concentration in the animals exposed to SS during the postnatal period (p = 0.035, F1,44 = 4.78). This was most evident in the SS/SS group. By pooling the animals into two groups, postnatal SS exposure reduced forebrain DNA by 2.2% (p = 0.034, F1,44 = 4.82)

Table 2 Effects of in utero and/or postnatal SS exposure on forebrain weight and biochemical composition
Full size table

In the hindbrain (Table 3), there was a greater reduction of DNA concentration by postnatal SS exposure (p < 0.001, F1,44 = 14.87). As in the forebrain, the SS/SS group had the lowest DNA concentration which was 6.1% less than the FA/FA group. Because there was no in utero exposure by postnatal exposure interaction, these data were also pooled into two groups (animals with and without postnatal SS exposure). By comparing these two groups, postnatal SS exposure reduced hindbrain DNA concentration by 4.4% (p = 0.001,F1,44 = 13.24). In addition to the reduction in hindbrain DNA concentration, postnatal SS exposure also resulted in a significant increase in the protein/DNA ratio (p = 0.001, F1,44 = 13.74). For example, the protein/DNA ratio of the SS/SS group was 12.1% greater than the value for the FA/FA group. By pooling the data into two groups (as above), postnatal SS exposure increased the hindbrain protein/DNA ratio by 8.4% (p = 0.001, F1,44 = 13.13).

Table 3 Effects of in utero and/or postnatal SS exposure on hindbrain weight and biochemical composition
Full size table

DISCUSSION

This study demonstrates that perinatal exposure to SS significantly affects the biochemical composition of the brain. Although in utero exposure to SS did not affect the biochemical indicators of brain development, postnatal SS exposure did cause a reduction in DNA concentration. This change was more significant in the hindbrain, which contains the cerebellum, a brain region which undergoes significant postnatal development in the rat(27, 32, 38). The decrease in hindbrain DNA concentration suggests that cellular density in this brain region was reduced. In addition to this reduction in hindbrain cellular density, the total DNA content (an indicator of cell number) of the hindbrain was also decreased by postnatal SS exposure (p = 0.008,F1,44 = 7.65, data not shown). The reduction in hindbrain cellular density was accompanied by a significant increase in the protein/DNA ratio, which is an indicator of cell size. These two changes may explain why hindbrain weight was not affected by SS exposure (brain sparing), with cellular hypertrophy helping to counteract the reduction in cell number. Brain sparing, a situation where the adverse effects of a substance on brain growth are less than the effects on the growth of other organs, has been reported in studies of both maternal malnutrition and maternal exposure to teratogens(31, 32, 39, 40). Although cellular hypertrophy may have prevented a postnatal SS-induced reduction in brain weight, the reduction in cell number may still result in abnormal neurologic function. Therefore, it is likely erroneous to conclude that physiologic brain sparing occurred in this study.

Similar changes in the biochemical composition of brain regions have been described in pups of dams exposed to nicotine during gestation(24, 31–33). In one particular study, dams were exposed to nicotine by constant infusion for 16 d starting on d 4 of gestation(33). At 2 d of postnatal age, cerebellar DNA concentration was reduced, whereas the cerebellar protein/DNA ratio was increased and cerebellar weight was unchanged. Subsequently, postnatal cerebellar growth and cellular acquisition were impaired by the prenatal nicotine exposure in these animals, and therefore brain sparing was not present.

Although the majority of clinical studies of the effects of prenatal maternal exposure to ETS on fetal growth have focused on birth weight(11), one study did examine the effects of maternal passive smoking on neonatal head circumference. In that investigation, the offspring of nonsmoking, passive smoking, and direct smoking women were evaluated, and no difference in neonatal head circumference was detected(41). With only this small amount of clinical data regarding the effects of ETS on fetal head growth, it is difficult to correlate the results of our animal study with the observations made in newborn humans.

The absence of an effect of in utero SS exposure is difficult to explain directly from this study. It is possible that prenatal SS exposure did affect certain aspects of brain development which, in the SS/FA animals, normalized after 9 wk of postnatal FA exposure. Given the design of this study, where all of the pups were evaluated at 9 wk of age, any effects ofin utero SS exposure, which would be present only at birth, could not be assessed. Previous work on the effects of maternal nicotine exposure on fetal brain development suggests that this is a possible explanation. In a study of the effects of twice daily maternal nicotine injections from d 3-20 of gestation, postnatal growth of brain, other organs, and body growth were initially delayed, but normalized by the end of the fifth postnatal week(24).

The effects of ETS on fetal growth and pregnancy outcome have been assessed in a limited number of studies. In the present study, at 9 wk of postnatal age, there was a significant reduction in body weight in the animals exposed to SS postnatally, without a change in total brain weight. This reduction in body weight was not present at birth, but began to appear by 6 wk of age. Two previous studies of the effects of in utero SS exposure using the same exposure system described in this study have been reported(19, 20). Rajini et al.(19) demonstrated a small reduction in average fetal weight per litter at 20 d of gestation in the offspring of dams exposed to SS for 6 h per day on d 3, 6-10, and 13-17 of gestation without a change in the number of live pups, whereas Witschi et al.(20) noted that SS exposure for 6 h/d on d 3-11 of pregnancy reduced both the number of implantation sites per litter as well as the number of live pups per litter. The average pup weight per litter on d 21 of gestation was not affected by in utero SS exposure. With respect to the present investigation, those two studies used different SS exposure protocols, and the animal weights were expressed as mean pup weight per litter for each exposure group, as opposed to mean pup weight per exposure group. Therefore the results of the three studies are not directly comparable. In a study from another laboratory, where the various components of the SS exposure were not quantified, dams were exposed to SS for 2 h per day on d 1-20 of gestation. A 9% reduction in fetal weight on d 20 of gestation was noted in the SS-exposed pups(18). Therefore, under certain exposure conditions, in utero maternal SS exposure has been shown to adversely affect fetal growth.

The mechanism by which tobacco smoke produces teratogenic effects has not been established. In utero, the fetus may be exposed to nicotine by several routes. It has been suggested that transplacental passage of nicotine to the fetus may be small(42). However, amniotic fluid contains higher concentrations of nicotine than maternal blood(43), and fetal exposure to nicotine due to both swallowed amniotic fluid and dermal absorption may be substantial(42). Both nicotine and carbon monoxide have been implicated as probable teratogenic agents, with placental and fetal hypoxia resulting from exposure to these substances(22). However, nicotine likely has direct teratogenic effects on developing CNS and peripheral nervous system tissues, as well. A reduction in cerebellar DNA, together with an increase in brain [3H]nicotine binding sites and a reduction in kidney norepinephrine levels have been shown in the offspring of dams exposed to a low dose of nicotine(31). The low dose of nicotine used in these experiments did not adversely affect maternal weight, pup body weight, or pup brain weight. Therefore, nicotine likely has a direct teratogenic effect on developing neurotransmitter systems apart from its effect on systemic growth.

The observations made in a limited number of clinical studies suggest that maternal exposure to ETS during pregnancy has an adverse effect on cognitive function of the offspring(8–10). The present study provides biologic evidence of an alteration of brain development due to SS exposure. Although the changes in the biochemical composition of the hindbrain suggest both a reduction in cell number and an increase in cell size due to SS, morphologic studies will be necessary both to further quantify and to fully characterize these developmental effects. Although only postnatal SS exposure resulted in statistically significant changes in brain development, the observations from this study have important clinical implications. Brain development in the rat during the first 2 wk of postnatal life is substantial(brain growth spurt) and corresponds to human fetal brain growth during the third trimester(44). Therefore, maternal passive smoking during the latter months of pregnancy may adversely affect certain aspects of neuronal development.