The association between smoking and ectopic pregnancy: why nicotine is BAD for your fallopian tube

Abstract

Epidemiological studies have shown that cigarette smoking is a major risk factor for tubal ectopic pregnancy but the reason for this remains unclear. Here, we set out to determine the effect of smoking on Fallopian tube gene expression. An oviductal epithelial cell line (OE-E6/E7) and explants of human Fallopian tubes from non-pregnant women (n = 6) were exposed to physiologically relevant concentrations of cotinine, the principle metabolite of nicotine, and changes in gene expression analyzed using the Illumina Human HT-12 array. Cotinine sensitive genes identified through this process were then localized and quantified in Fallopian tube biopsies from non-pregnant smokers (n = 10) and non-smokers (n = 11) using immunohistochemistry and TaqMan RT-PCR. The principle cotinine induced change in gene expression detected by the array analysis in both explants and the cell line was significant down regulation (P<0.05) of the pro-apoptotic gene BAD. We therefore assessed the effect of smoking on cell turnover in retrospectively collected human samples. Consistent with the array data, smoking was associated with decreased levels of BAD transcript (P<0.01) and increased levels of BCL2 transcript (P<0.05) in Fallopian tube biopsies. BAD and BCL2 specific immunolabelling was localized to Fallopian tube epithelium. Although no other significant differences in levels of apoptosis or cell cycle associated proteins were observed, smoking was associated with significant changes in the morphology of the Fallopian tube epithelium (P<0.05). These results suggest that smoking may alter tubal epithelial cell turnover and is associated with structural, as well as functional, changes that may contribute to the development of ectopic pregnancy.

Figure 1. Diagram comparing array data in each of the four analytical groups.

In the OE-E6/7 cell line the addition of 40 ng/ml cotinine (‘Low Cell’) altered the expression of 946 genes (increasing 585 genes and reducing the expression of 361 genes). Of the 676 genes (361 increased and 315 reduced) altered by the addition of 400 ng/ml cotinine (‘High Cell’), 135 were shared with the ‘Low Cell’ treatment. In FT explants the addition of 40 ng/ml cotinine (‘Low Tissue’) altered the expression of 596 genes (137 increased and 459 reduced). Explant treatment with 400 ng/ml cotinine (‘High Tissue’) changed 2419 genes (increasing 313 and reducing 2106), 265 of which were shared with the ‘Low Tissue’ treatment. The effect of cotinine in Fallopian tube therefore is mainly to inhibit gene expression. When comparing the effect of cotinine in the OE-E6/7 cells when compared to FT explants common were genes identified. There were 16 altered genes the ‘Low Tissue’ group shared with the ‘Low Cell’ group and 27 genes shared with the ‘High Cell’ group. There were 111 genes that changed expression in common between the ‘High Tissue’ and ‘Low Cell’ group and 54 with the ‘High Cell’ group. There was only one up-regulated gene and one down regulated gene in common to all groups.

Figure 2. TaqMan RT-PCR analysis of BAD and BCL2 transcript abundance in FT from smokers and non-smokers.

Relative expression of BAD (A) and BCL2 (B) in the FT of non-smokers (clear bars: n = 11) and smokers (filled bars: n = 10). Observed differences are significant at * P<0.05, **P<0.01. Gene expression was related to a G6PDH internal control.

Figure 3. Immunohistochemistry.

A) BAD (brown) is expressed in the epithelium (E), most prominently towards the lumen (L), of the FT and not in the stroma (S). B) BCL2 (brown) immunolabelling showing predominant epithelial staining with occasional cell staining in the stroma. C) Higher power BCL2 immunolabelling showing no staining in the ciliated epithelial cells (black arrow) with staining in interspaced cells with no obvious cilia (red arrow). D) Scanning EM highlighting the two populations of epithelial cells with and without cilia. E) Apoptotic cells (brown) identified by immunolabelling for cleaved caspase 3. The arrow shows an apoptotic cell in the tubal epithelium. F) Higher power view showing a cell stained for cleaved caspase 3 (arrow) at the epithelial and stromal junction. G) Cells stained by the proliferation marker Ki-67 (brown). H) Higher power view showing a cell stained for Ki-67 (arrow) in the tubal epithelium. I) Representative image of a section of FT from a smoker immunostained for BAD showing the apical smooth protuberances or ‘epithelial bledding’ (arrow). J) Section of FT immunostained for BAD with the ciliated epithelial cells with no surface ‘epithelial blebbing’.

Figure 4. Assessment of cell death in the FT.

A) Frequency of immunopositive cells in the FT of non-smokers (clear bars: n = 11) and smokers (filled bars: n = 10). Relative expression of of CASP3 (B) and CASP9 (C) transcripts within the FT of smokers and non smokers and (D) their correlation to each other in a single FT sample. Observed differences are significant at * P<0.05. Gene expression was related to a G6PDH internal control.

Figure 5. Assessment of changes in cell proliferation and morphology in the FT of smokers.

A) Frequency of Ki67 positive cells in the FT of non-smokers (clear bars: n = 11) and smokers (filled bars: n = 10). B) Relative expression of Cyclin D1 transcripts within the FT of smokers and non smokers Expression of CCND1 in the FT of smokers and non smokers. C) Histoscore analysis of the degree of ‘epithelial blebbing’ in non-smokers compared to smokers. Observed differences are significant at * P<0.05. Gene expression was related to a G6PDH internal control.