Smoking is a well-established risk factor for cardiovascular disease. Therefore, it is not surprising that smoking during pregnancy is associated with multiple poor pregnancy outcomes and may also significantly affect maternal health in later life. The prevalence of smoking during pregnancy in the United States is as high as 30–50%,1 and therefore represents a major, modifiable risk factor for poor pregnancy outcomes. Interestingly, smoking is negatively associated with preeclampsia, and given the many similarities between preeclampsia and atherosclerosis, the mechanism of this association remains puzzling.2
Cigarette smoke exposure causes cardiovascular disease through vascular compromise secondary to endothelial dysfunction and oxidative stress. Smoking is associated with alterations in endothelial function as measured by in vivo studies of endothelium-dependent vasodilation and in vitro studies of circulating biochemical markers.3–8 Reactive oxygen species damage endothelial cells directly through cellular injury, and indirectly through effects on lipid peroxidation and nitric oxide scavenging. Evidence to support these claims include the reversal of endothelial dysfunction in smokers after smoking cessation9 or inhibition of xanthine oxidase, the primary source of the reduced O2 species.10,11 Additionally, xanthine oxidase transcription and activity are upregulated by tobacco smoke condensate exposure of endothelial cells in vitro.12
In the past two decades, several different laboratory measures of endothelial dysfunction have been validated, including biochemical markers, genetic markers, and assessment of in vivo flow-mediated vasodilation. Measurement of biochemical markers is an accepted method of assessing endothelial dysfunction and generally includes a panel of cell adhesion molecules (intracellular adhesion molecule 1 [ICAM-1], vascular cell adhesion molecule 1 [VCAM-1], E-selectin, P-selectin, sCD40L), cytokines (interleukin-6, interleukin-18, tumor necrosis factor-α, high-sensitivity C-reactive protein [hs-CRP], endothelin-1 [ET-1], and metalloproteinases. 13
Pregnancy represents a naturally occurring period of significant physiologic adaptation. This adaptation affects many cells and organ systems, including the vascular system and endothelial function. In vivo and in vitro studies indicate augmented endothelial-mediated relaxation in normal pregnancy.14 In addition, markers of endothelial activation, including ICAM-1 and VCAM-1 that are known to be responsive to cytokines, are increased during pregnancy, a finding consistent with the generalized increased inflammatory response of normal pregnancy.15 Smoke exposure during pregnancy has been associated with further changes in endothelial cell function beyond what naturally occurs. Compared with uncomplicated pregnancy, we found that women who smoked during pregnancy had elevated levels of circulating ICAM-1 and uric acid.16,17 Elevated circulating ICAM-1 is indicative of endothelial dysfunction, while an increase in uric acid may represent an upregulation of xanthine oxidase activity.
Although intriguing, our prior studies were limited in both sample size and biomarkers assayed. The present study is a different patient population than previously studied. The objectives of this study were to measure markers of endothelial dysfunction, particularly cell adhesion molecules (CAMs), in a large prospective study of primigravid women who did or did not smoke during pregnancy. In addition, we sought to determine the source of any soluble markers whose expression was altered by cigarette smoke exposure.
MATERIALS AND METHODS
Samples for this study were obtained randomly from healthy primiparous patients enrolled in an ongoing longitudinal study at Magee-Womens Hospital (Pittsburgh, PA) between January 2003 and September 2004. Information regarding smoking status was not available at sample collection. Samples were collected at the time of a routine prenatal visit. The institutional review board at Magee-Womens Hospital approved this study, and written informed consent was obtained from each participant at enrollment. Demographic data and outcome data were obtained by interview and medical record abstraction. Abstracted charts were examined by a jury of clinicians to establish whether pregnancies were uncomplicated or had medical or obstetric complications.
Three hundred twenty-five blood samples were collected from a cohort of 198 women at time points throughout pregnancy when patients were otherwise having routine blood work. Plasma was prepared from blood anticoagulated with ethylenediaminetetra-acetic acid and divided into aliquots under sterile conditions and stored at -80°C until assayed. Smoking status was determined by plasma cotinine concentrations measured by a direct enzyme-linked immunosorbent assay (ELISA) kit (Immunalysis, Pomona, CA). The sensitivity of the cotinine assay is 1 ng/mL. A cotinine concentration of more than 5 ng/mL was considered positive for smoking and a concentration of more than 100 ng/mL was considered positive for heavy smoking.18,19
All plasma samples for CAMs were run in duplicate. A quality–control sample was assayed with each microtiter plate. Plasma ICAM-1, VCAM-1, and E-selectin, were determined by ELISA using kits (R&D Systems, Minneapolis, MN). Standard curves for ICAM-1 were determined from 0 to 100 ng/mL with interassay variability of 9.3% and sensitivity of less than 0.5 ng/mL. Standard curves for VCAM-1 were calculated from 0 to 75 ng/mL with interassay variability of 11.2% and sensitivity of 0.5 ng/mL. Standard curves for E-selectin were generated from 0 to 10.6 ng/mL with a minimum detectable concentration of 0.1 ng/ml and interassay variability of 7.3%.
Freshly delivered placentas were rinsed in sterile phosphate-buffered saline. A small sample was fixed in 10% buffered formalin and embedded in paraffin. Sections were cut at 5-micron thickness and analyzed for immunoreactivity to primary antibodies to ICAM-1 (Santa Cruz Biotechnology, Santa Cruz, CA) and proliferating cell nuclear antigen (Dako, Carpinteria, CA). After rehydration, slides were processed for proliferating cell nuclear antigen and ICAM-1 staining. Endogenous peroxidases were quenched by incubating the slides in 3% H2O2. After a citrate antigen retrieval step, the slides were placed into Protein Blocking Agent (Shandon, Pittsburgh, PA). The primary antibodies were applied, and slides were placed into a humidified chamber for 1 hour. Intracellular adhesion molecule 1 was added at a 1:500 dilution and proliferating cell nuclear antigen was at a 1:200 dilution. Nonimmune mouse serum (Biogenics, San Raman, CA) was used as a negative control for both antibodies, and a 16-week placenta was used as a positive control for both antigens. Antibodies were detected using the Super Sensitive Detection Kit (Biogenics). A biotinylated secondary antibody to mouse immunoglobin was applied followed by a horseradish peroxidase-conjunjucated strepavidin label. Diaminobenzide (Biogenex, San Roman, CA) was applied and slides were counterstained with Gill’s hematoxlin (Fisher Pittsburg, PA). The slides were dehydrated and mounted with cytoseal 60 (Stephen Scientific, Kalamazoo, MI) for microscopic viewing. For analysis, we determined the number of proliferating cell nuclear antigen + nuclei/microscope field (40× objective) in at least 10 random fields per slide. Intracellular adhesion molecule 1 was scored as positive or negative and as an estimated percentage of positive cells per field.
Total RNA was isolated from flash-frozen placenta using Trizol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol. After isolation, double- and single-stranded DNA were removed by treatment with DNaseI (Invitrogen). Quality and quantity of RNA were determined by measuring absorption at 280 and 260 nm. The cDNA was synthesized from 2 μg of total RNA using the Superscript III First Strand Synthesis System (Invitrogen). Relative levels of gene expression for ICAM-1 (Hs00164932, Applied Biosystems, Foster City, CA) were determined by real-time polymerase chain reaction (PCR) on an ABI Prism 7700 Sequence Detection System (Applied Biosystems). Polymerase chain reaction was performed using TaqMan Gene Expression Assays (Applied Biosystems) which contain TaqMan unlabeled primers and a FAM (fluorescein) dye-labeled MGB (minor groove binder) probe for all target genes. The GAPDH gene (Assay ID) was used as an endogenous reference for all samples. All amplification cycles were performed in a singleplex 50-μL reaction with cDNA equivalent to 100 ng of total RNA. A typical 50-μL reaction sample also contained 25 μL of TaqMan Universal PCR Master Mix (containing 1× TaqMan buffer, 200 μM dNTPS, 5 mM MgCl2, 1.25 U AmpliTaqGold, and 0.5 U of Amperase uracil-N-glycolsylase [UNG] along with the TaqMan primers and MGB probes). Thermal cycling conditions were 2 minutes at 50°C and 10 minutes at 95°C followed by 45 cycles at 95°C for 30 seconds and 60°C for 1 minute. Relative quantification of gene expression was performed using the comparative CT (Cycle-Threshold) method (Applied Biosystems), which consists of the normalization of the number of target gene copies (ICAM1) to an endogenous reference gene (GAPDH). All experiments were conducted in triplicate and an average CT value calculated for the replicates. The CT value refers to the cycle number wherein the fluorescent intensity crosses the threshold line, which is set in the exponential phase of the amplification plot above background levels.
Human umbilical vein endothelial cells were purchased from Cambrex (Walkersville, MD) along with the corresponding EGM endothelial growth media and EGM Bullet Kit media. The cells were grown to confluence and exposed to cotinine at concentrations of 75, 600, and 1,200 ng/mL for 4 hours at 37°C. After cotinine exposure, the supernatant was saved for determination of ICAM-1 by ELISA (R&D Systems, Minneapolis, MN). Cells were harvested and fixed with 0.5% paraformaldehyde. Cellular surface molecule expression of CD31, CD106, CD62P, CD62E, and CD54 purchased from PharMingen (San Diego, CA) were performed by flow cytometry. In this experiment IL-1β was used as a positive stimulation control, and media without cotinine was the negative control. There was no significant cell death in response to cotinine exposure.
Demographic variables were compared among smokers and nonsmokers using the Student t test and χ2. Cellular adhesion molecule concentrations were compared between the two groups using Mann-Whitney and, when stratified by severity of smoking, Kruskal-Wallis. Only 62% of patients had more than 1 sample obtained during pregnancy, so repeated-measures analysis of variance (ANOVA) was not used. To maintain independence, samples were divided into gestational age categories of 6 weeks (< 20, 20–26, 26–32, and > 32 weeks of gestational age) with no patient having more than one sample in any category. Analysis was performed for each gestational age interval. The associations of CAMs with smoking and cotinine concentrations were evaluated with univariate and multivariate regression models controlling for gestational age of sample, race, maternal age, and body mass index. Demographic data are presented as mean ± standard deviation, and experimental data are presented as medians. Polymerase chain reaction data were compared by using the Student t test. Human umbilical vein endothelial cell supernatant ICAM-1 concentrations were compared across concentrations using ANOVA, and post hoc analysis was done with Dunnett’s test comparing increasing concentrations with no treatment. Statistical analysis was done using SPSS 13.0 for Windows (SPSS, Inc, Chicago, IL) statistical software. Power calculations were done using Sample Power 1.2. We estimated that 300 samples taken throughout pregnancy with a smoking rate of 30–40% would have greater than 80% power to detect a 40-ng/mL mean difference in endothelial markers (α = 0.05) between women who smoked and those who did not smoke.
Demographic characteristics of the 198 primiparous subjects are listed in Table 1. Overall, the mean maternal age of the cohort was 23.5 ± 4.8 years with a mean body mass index of 25.1 ± 6.2. The racial distribution was 33% African American, 65% white, and 2% other. Smokers were slightly younger compared with nonsmokers (22.6 ± 3.5 vs 23.9 ± 5.1 years, P = .04). Body mass index and racial distribution were not statistically different by smoking status.
Cotinine concentrations were evaluated on all samples. Using 5 ng/mL as a cutoff for smoking, 32% of women were smokers according to their first sample cotinine concentration. The mean cotinine concentration for the smokers was 137.1 ± 16.1 ng/mL. When stratified by severity, 17% of the cohort were light smokers and 15 % were heavy smokers with mean concentrations of 44.6 ± 5.4 and 245.6 ± 20.7 ng/mL, respectively.
Cellular adhesion molecule comparisons between heavy smokers and all other subjects in each of the gestational age intervals were performed with Mann-Whitney test. Medians are displayed in Fig. 1. The concentrations of ICAM-1 were higher in smokers for all time points and significantly higher at time points 1 and 2 (Fig. 1A). When stratified by severity of smoking, ICAM was greatest in heavy smokers with significance at time points 1, 2, and 4 (Fig. 2A). Concentrations of VCAM-1 were not different by smoking status (Fig. 1B and Fig. 2B). E-selectin concentrations were significantly lower in smokers in mid and late pregnancy between smokers and nonsmokers (Fig. 1C), but significance did not remain when stratified by severity (Fig. 2C). Concentrations of ICAM-1 were correlated with cotinine concentrations (r = 0.15, P < .01) and VCAM-1 concentrations (r = 0.25; P < .001). Using logistic regression, smokers have an increased odds ratio (OR) of ICAM-1 concentrations in the upper quartile (OR 2.4, 95% confidence interval [CI] 1.2–2.5) and heavy smoking during pregnancy is associated with a further increase risk (OR 6.9, 95% CI 2.7–17.9). This association for both smoking and heavy smoking remained significant controlling for confounders (gestational age of sample, race, maternal age, and body mass index) (OR 2.4, 95% 1.2–4.9; OR 7.7, 95% CI 2.9–20.8).
To determine whether the placenta might be the source of elevated ICAM-1 production, we determined the relative expression of both mRNA and protein levels of ICAM-1 in placentas from smokers and nonsmokers. Figure 3 represents the relative expression of ICAM-1 as determined by real-time reverse-transcription PCR. Relative placental expression for smokers was significantly less compared with expression for nonsmokers (1.2 ± 0.2 vs 2.2 ± 0.3; P = .02). Immunohistochemistry was used to determine the relative level and distribution of ICAM-1 protein in placentas. Although not quantitative, we did not observe any differences between the two groups.
Ex vivo exposure of HUVECs to cotinine was used to determine whether endothelial cells could upregulate ICAM-1 expression in response to cigarette smoke exposure. After 4 hours exposure to cotinine, the major metabolite of nicotine, the media on the HUVECs showed a significant increase in soluble ICAM-1, but only at the highest concentration of cotinine tested. As shown in Figure 4, supernatant concentrations of ICAM-1 increased with increasing concentrations of cotinine, as determined by ANOVA (P < .01). Concentrations of ICAM-1 were increased to 0.13 ± 0.02 ng/mL after exposure to 1,200 ng/mL cotinine compared with no cotinine exposure (0.07 ± 0.02 ng/mL) and was different from no exposure by post hoc analysis using Dunnett’s test (P = .003). There were no differences in any of the surface markers analyzed on HUVECs exposed to cotinine.
We performed a prospective, large-cohort study to determine the effects of cigarette smoke exposure on endothelial function during pregnancy. Women who smoked heavily during pregnancy had increased circulating concentrations of ICAM-1. Differences in ICAM-1 but not VCAM-1 may be secondary to differences in regulation,20 function, or sources of these two molecules. Intracellular adhesion molecule 1 is expressed on endothelial cells and leukocytes, binds to CD18/CD11a or CD18/CD11b, and mediates the adhesion of monocytes, lymphocytes, and neutrophils to endothelial cells.7 Vascular cell adhesion molecule 1 is expressed on endothelial cells and vascular smooth muscle cells, binds to integrin a4b1, and mediates the adhesion of lymphocytes and monocytes to endothelial cells.7 Alterations in circulating ICAM-1 may be secondary therefore to either maternal endothelial cell dysfunction, leukocyte shedding, or abnormalities of placentation.
In contrast, E-selectin was decreased in smokers. E-selectin is a surface-bound adhesion molecule specific to activated endothelial cells, and it mediates interaction of leukocytes, platelets, and endothelial cells. The association of decreased E-selectin among pregnant smokers may be an important piece of the negative association of smoking and preeclampsia. A decrease in E-selectin would suggest a decrease in endothelial-leukocyte adhesion among women who smoke and thus be consistent with a decreased rate of preeclampsia.
Intracellular adhesion molecule 1 is expressed in the placenta in both cytotrophoblasts and syncytiotrophoblasts, and therefore alterations in placental shedding or production with smoking may contribute to elevated maternal circulating concentrations. Intracellular adhesion molecule 1 is increased with abnormal implantation such as with preeclampsia or growth restriction.21 Our data demonstrate the relative expression of protein levels of ICAM-1 by immunohistochemistry in placentas from smokers and nonsmokers is not different. In addition, the relative expression of mRNA was decreased with any amount of smoking. It is therefore unlikely that the placenta is the source of increased ICAM-1 circulating concentrations with smoking. In normal pregnancies, endovascular cytotrophoblasts express VCAM-1 but in abnormal implantation, as occurs in intrauterine growth restriction and preeclampsia, VCAM-1 is not produced.22 We did not evaluate placental production of VCAM-1 given the lack of difference in circulating maternal concentrations.
To evaluate the endothelial contribution of circulation ICAM-1 concentrations, we measured increased concentrations of ICAM-1 in HUVEC supernatant after exposure to physiologic doses of cotinine. The expression of ICAM-1 is increased in pulmonary endothelial cells of nonpregnant smokers and in HUVECs exposed to serum from smokers.23,24 Concentrations of ICAM-1 were increased with cotinine concentrations equivalent to in vivo very heavy exposure only. This suggests that the endothelium is a possible source for the increased maternal concentrations of ICAM-1 but perhaps only under conditions of very heavy exposure.
In conclusion, the increase in plasma ICAM-1 in the maternal circulation of heavy smokers may primarily be endothelial in nature although shedding from leukocytes cannot be ruled out. The placenta as a primary source of the increase is less likely. If shedding from either the endothelium or leukocytes decreased surface expression then an overall decreased endothelial-leukocyte adhesion may result. This decrease in endothelial activation may contribute to the decrease in preeclampsia among smokers. Further studies and analyses are underway to evaluate the contribution of circulating leukocytes.
1. American College of Obstetricians and Gynecologists. Smoking and Women’s Health. ACOG Educational Bulletin 240. Washington, DC: ACOG; 1997.
2. Conde-Agudelo A, Althabe F, Belizan JM, Kafury-Goeta AC. Cigarette smoking during pregnancy and risk of preeclampsia: a systematic review. Am J Obstet Gynecol 1999;181:1026–35.
3. Lekakis J, Papamichael C, Vemmos C, Nanas J, Kontoyannis D, Stamatelopoulos S, et al. Effect of acute cigarette smoking on endothelium-dependent brachial artery dilatation in healthy individuals. Am J Cardiol 1997;79:529–31.
4. Celermajer DS, Sorensen KE, Bull C, Robinson J, Deanfield JE. Endothelium-dependent dilation in the systemic arteries of asymptomatic subjects relates to coronary risk factors and their interaction. J Am Coll Cardiol 1994;24:1468–74.
5. Bergmann S, Siekmeier R, Mix C, Jaross W. Even moderate cigarette smoking influences the pattern of circulating monocytes and the concentration of sICAM-1. Respir Physiol 1998;114:269–75.
6. Takeuchi N, Kawamura T, Kanai A, Nakamura N, Uno T, Hara T, et al. The effect of cigarette smoking on soluble adhesion molecules in middle-aged patients with Type 2 diabetes mellitus. Diabet Med 2002;19:57–64.
7. Ridker PM, Hennekens CH, Roitman-Johnson B, Stampfer MJ, Allen J. Plasma concentration of soluble intercellular adhesion molecule 1 and risks of future myocardial infarction in apparently healthy men.[see comment]. Lancet 1998;351:88–92.
8. Blann AD, Steele C, McCollum CN. The influence of smoking on soluble adhesion molecules and endothelial cell markers. Thromb Res 1997;85:433–8.
9. Celermajer DS, Sorensen KE, Georgakopoulos D, Bull C, Thomas O, Robinson J, et al. Cigarette smoking is associated with dose-related and potentially reversible impairment of endothelium-dependent dilation in healthy young adults. Circulation 1993;88:2149–55.
10. Guthikonda S, Sinkey C, Barenz T, Haynes WG. Xanthine oxidase inhibition reverses endothelial dysfunction in heavy smokers. Circulation 2003;107:416–21.
11. Guthikonda S, Woods K, Sinkey CA, Haynes WG. Role of xanthine oxidase in conduit artery endothelial dysfunction in cigarette smokers. Am J Cardiol 2004;93:664–8.
12. Kayyali US, Budhiraja R, Pennella CM, Cooray S, Lanzillo JJ, Chalkley R, et al. Upregulation of xanthine oxidase by tobacco smoke condensate in pulmonary endothelial cells. Toxicol Applied Pharmacol 2003;188:59–68.
13. Deanfield J, Donald A, Ferri C, Giannattasio C, Halcox J, Halligan S, et al. Endothelial function and dysfunction. Part I: Methodological issues for assessment in the different vascular beds: a statement by the Working Group on Endothelin and Endothelial Factors of the European Society of Hypertension. J Hypertens 2005;23:7–17.
14. Cockell AP, Poston L. Flow-mediated vasodilatation is enhanced in normal pregnancy but reduced in preeclampsia [see comments]. Hypertension 1997;30:247–51.
15. Redman CW, Sacks GP, Sargent IL. Preeclampsia: an excessive maternal inflammatory response to pregnancy. Am J Obstet Gynecol 1999;180:499–506.
16. Lain KY, Wilson JW, Crombleholme WR, Ness RB, Roberts JM. Smoking during pregnancy is associated with alterations in markers of endothelial function. Am J Obstet Gynecol 2003;189:1196–201.
17. Lain KY, Markovic N, Ness RB, Roberts JM. Effect of smoking on uric acid and other metabolic markers throughout normal pregnancy. J Clin Endocrinol Metab 2005;90:5743–6.
18. Lindqvist R, Lendahls L, Tollbom O, Aberg H, Hakansson A. Smoking during pregnancy: comparison of self-reports and cotinine levels in 496 women. Acta Obstet Gynecol Scand 2002;81:240–4.
19. de Weerd S, Thomas CM, Kuster JE, Cikot RJ, Steegers EA. Variation of serum and urine cotinine in passive and active smokers and applicability in preconceptional smoking cessation counseling. Environ Res 2002;90:119–24.
20. Collins T, Read MA, Neish AS, Whitley MZ, Thanos D, Maniatis T. Transcriptional regulation of endothelial cell adhesion molecules: NF-kappa B and cytokine-inducible enhancers. FASEB J 1995;9:899–909.
21. Labarrere CA, Faulk WP. Intercellular adhesion molecule-1 (ICAM-1) and HLA-DR antigens are expressed on endovascular cytotrophoblasts in abnormal pregnancies. Am J Reprod Immunol 1995;33:47–53.
22. Zhou Y, Fisher SJ, Janatpour M, Genbacev O, Dejana E, Wheelock M, et al. Human cytotrophoblasts adopt a vascular phenotype as they differentiate. A strategy for successful endovascular invasion? [see comments.]. J Clin Invest 1997;99:2139–51.
23. Schaberg T, Rau M, Oerter R, Liebers U, Rahn W, Kaiser D, et al. Expression of adhesion molecules in peripheral pulmonary vessels from smokers and nonsmokers. Lung 1996;174:71–81.
© 2006 The American College of Obstetricians and Gynecologists
24. Adams MR, Jessup W, Celermajer DS. Cigarette smoking is associated with increased human monocyte adhesion to endothelial cells: reversibility with oral L-arginine but not vitamin C. J Am Coll Cardiol 1997;29:491–7.