Misoprostol-Induced Modification of the Notch Signaling Pathway in the Human Cervix

Introduction

Preterm labor can be described as the onset of labor before the 37th week of gestation. ¹ Contemporary theories of uterovaginal embryology suggest that the cervix should be regarded not only as a part of the uterus but also as a transition organ between 2 structures with embryologically different roots. ²

Cervical ripening, the clinical features of which are softening, thinning, and dilatation, is a process that usually occurs prior to labor and independently of uterine contractions. ³ This process resembles an inflammatory reaction involving the rearrangement of collagen fibers in the cervical stroma, infiltration of white blood cells, and generation of edema. ^4,5 One of the events leading to cervical ripening is thought to be a local increase in prostaglandins (PGs) with autocrine and paracrine consequences. Prostaglandins have diverse functions in parturition, including regulation of uterine contractility and tissue remodeling. ⁶ Misoprostol, a synthetic prostaglandin E1 analog and is widely used as a uterotonic agent and in the prevention of postpartum hemorrhage due to its potent ability to induce uterine contractions ^6,7 by selectively binding to EP-2/EP-3 prostanoid receptors in the myometrium. ⁸ The administration of PGs increases the amount of submucosal water in the cervix content and allows the dissolution of the colloidal fibers. These changes are similar when compared to a pregnant woman who performs normal birth and premature birth. Prostaglandins are endogenous substances found in fetal membranes, myometrium, and decidua in pregnancy. ⁹

The Notch signaling pathway is conserved in many species. ¹⁰ Notch signaling links the fate of one cell with that of a cellular neighbor through physical interactions between the Notch receptor and the membrane-bound ligands that are expressed in an apposing cell. ¹¹ Notch itself is a cell-surface receptor that transduces short-range signals by interacting with transmembrane ligands such as δ (termed δ-like in humans) and serrate (termed jagged in humans) on neighboring cells. ¹² In Notch signaling, the cells that initiate the signal express Notch ligands belonging to the δ-like (DLL1, DDL3, DLL4) or the jagged (J1, J2) family, whereas the cells that receive Notch signals express Notch receptors. The Notch signal is activated as a result of interactions between Notch receptors and their ligands: δ or Jagged. ¹³ Ligand binding leads to cleavage and release of the Notch intracellular domain, which then travels to the nucleus to regulate transcriptional complexes containing the DNA-binding protein CBF1/RBPjk/Su(H)/Lag1 (CSL). Following their synthesis, Notch receptors are cleaved by protein convertases during exocytosis at site 1 (S1), which regulates their trafficking and signaling activity. ¹⁴ However, the responses of the cells to the Notch signal are very large and vary. For example, Notch in some cases initiates cell proliferation, in some cases provide apoptosis. Different responses that depend on Notch’s interaction with other pathways are thought to have emerged. ^13,15,16 Notch signals can influence differentiation, proliferation, and apoptotic cell fates ¹¹ and are necessary for decidualization in pregnancy. ¹⁵

Notch1 signaling is reported to modulate multiple signaling mechanisms crucial for decidualization in the artificial decidualization model in mice ¹⁷ and in primates, ¹⁸ which is essential for the establishment of a successful pregnancy. It is also supposed that Notch signaling is an important factor in the regulation of pregnancy and may be involved, in part, in inflammation-induced preterm labor. ¹⁹ Leukocyte infiltration and upregulation of cytokines/chemokines occur in cervix, fetal membranes, and myometrium during labor at term these inflammatory processes interact with stretch and endocrine signals to initiate and amplify the labor cascade. ²⁰ The detection of Notch in human cervical cancers more than a decade ago prompted an investigation into the possible role of this pathway in driving these cancers. ²¹ However, the number of studies related to cervix-notch and especially preterm labor is very limited. Our objective was to identify changes in Notch signaling associated with misoprostol-induced cervical ripening. We still need to learn more about the finer details of the Notch pathway and how it specifically operates in different spatial and temporal cellular contexts. ²²

Materials and Method

This study was conducted by the Medical Faculty Department of Obstetrics and Gynecology at Akdeniz University with 20 pregnant women between 25 and 35 years of age and at 12 to 20 weeks’ gestational age undergoing medical abortion for fetal or maternal indications and who had approved the provision of biopsy material. The preferred method of biopsy in our study has been used successfully in different studies to obtain cervical material during the first trimester of pregnancy, ²³ and in intrapelvic histologic examinations, no complication has been reported. ²⁴ In addition, in 12 to 20 weeks of pregnancy, after 4 hours of the oral intake of misoprostol, cervical ripening success was found high. ²⁵

Despite its widespread use, there are different applications for dose optimization. It has been reported that cervical ripening after 200 to 400 μg sublingual misoprostol application at 3 hours can be achieved faster than vaginal and other oral method. Longer practice is thought to increase the risk of bleeding. ²⁶ In addition, in the 12 to 20 week pregnant group, when misoprostol different ways of use were compared, abortion success after sublingual application at 4-hour intervals was found as 86%. This method is more successful than other treatment groups, and the complication rate is low. ²⁵

Cervical biopsies were performed 4 hours after the oral intake of 200 μg misoprostol (n = 10). The control group (n = 10) underwent cervical biopsy before ingestion of misoprostol.

The tissues include biopsy specimens of the same patients before and 4 hours after misoprostol administration. A small amount of sample was used effectively. Every patient’s tissue participated in all experiments. It was evaluated before and after the same patient to exclude the effects of individual differences.

The Tru-Cut needle (22 gauge) was used to obtain a 1.5 mm-sized biopsy from the anterior cervix. The biopsy material was first divided into 2 parts. One of the pieces was put into liquid nitrogen for Western blot (WB) analysis, and the other was fixed with 10% neutral-buffered formalin during 24 hours for immunohistochemical [IHC] evaluations. After fixing, the tissues were passed through a 70%, 80%, 90%, and 100% ethanol series, respectively, to remove excess water and afterward were treated with xylol to make them clearer, followed by embedding in paraffin.

Results

Our study results were evaluated separately for cervical epithelium, vascular structures, and muscle and cellular components (Figure 1/H-score). This was because the expression patterns for the proteins of interest differed based upon cell type within tissues.

OPEN IN VIEWER

Figure 1.

Possible expression scheme for Notch receptors and ligands in cervical structures (H-Score). a indicates artery; Bc, basal cells; ic, intermediate cells; mac, macrophage; (sq), squamous epithelium; pc, parabasal cells; sc, superficial cells; smc: stromal smooth muscle.

Epithelium

The nonkeratinized, multilamellar, squamous epithelium (squamous-sq) found in the ectocervical area of the cervix in the control group showed differentiated expression of Notch receptors from the parabasal layer to the epithelial surface, especially in the areas of the anterior cervix, where tissue integrity was preserved. N1 is expressed in the basal cells but decreases toward the superficial area in the control group (Figure 2A). Although staining intensity of N2 is high in the basal part, it is minimal in the surface epithelium (Figure 2C). Notch 3 (Figure 2E) was expressed on all epithelial layers, while Notch 4 expression was less than N1 and N3 (Figure 2G). DLL1, DLL3, and J2 ligands are expressed in the superficial region, but there is no expression of DLL4 and J1 (Figure 2I-R). In the misoprostol groups, no fully conserved anterior epithelial areas were seen in the biopsy material, therefore no evaluations could be performed (it is thought to cause the separation of the stroma by affecting the delicate epithelial structure with the cause of bleeding in the vessels).

OPEN IN VIEWER

Figure 2.

The number of leukocytes counted in the control and misoprostol group area (40× T) and the immunohistochemical staining intensity of Notch receptors and ligands in leukocyte stroma (U). The difference in expression of Notch proteins between the layers in the anterior squamous epithelium of the ectocervix (Figure 2-small square). lc indicates leukocyte; sq, squamous epithelium.

Stroma

When the biopsy specimen was examined for expression levels in cells and blood vessels in the stroma:

In the Control Group

N1 were expressed in the vascular smooth muscle structures, arteries, and venule endothelium (Figure 3A). In addition, some cells in the stroma were expressing strong N1. Immunohistochemical staining showed that these cells were macrophages (Figure 4). N2 expression in the arterial wall (Figure 3C), N3 in the stromal smooth muscle, in arterioles and venules (Figure 3E), while N4 was exclusively expressed in the blood vessels (Figure 3G). Notch ligands showed that DLL1 was not expressed to the arterial wall when there was weak expression in the muscles and endothelium (Figure 5A). When DLL3 was expressed in the muscles and endothelium (Figure 5C), the staining intensity of DLL4 was quite high in the arterial wall (Figure 5E). J1 is only expressed on the vessel wall in a weak manner (Figure 5G), and J2 is intensively expressed on the vessel walls with the endothelium (Figure 5I).

OPEN IN VIEWER

Figure 3.

The results of the evaluations of the immunohistochemical staining intensity (40×) (J) and Western blot analysis (I) of endocervical tissue in the control and misoprostol groups of N1–N4 receptors. Thick arrow, arterial wall; arrow head, smooth muscle structure; thin arrow, macrophage; asterisk, arterioles–venules; neg, negative.

OPEN IN VIEWER

Figure 4.

Strong Notch1 (N1) expression in macrophages in the control group (on the left-C), and the same cells are confirmed with the antimacrophage antibody (serial histological section, on the right-Mac); a indicates artery; circle-mac, macrophage; neg, negative; v, vein.

OPEN IN VIEWER

Figure 5.

The results of the evaluations of the immunohistochemical staining intensity (40×) (L) and Western blot analysis (K) of endocervical tissue in the control and misoprostol groups of DLL1–J2 ligands. Thick arrow indicates arterial wall; arrow head, smooth muscle structure; thin arrow, macrophage; asterisk, arterioles–venules; CD34, endothelial marker; PCK, pancytokeratin; marker of the glandular epithelium.

In the Misoprostol Group

The intensity of staining for N1 was very strong in the muscles (Figure 3B). Although N2 is expressed in the endothelium and stromal muscle structures of blood vessels (Figure 3D), expression of N3 was small enough to be absent in the muscles (Figure 3F). N4 was not expressed in the blood vessels but instead was expressed in the muscles and macrophages (Figure 3H). Expression of DLL1 in arterial wall structure and endothelial and stromal muscle structures was strong (Figure 5B). DLL3 was intensely stained in the endothelium and few muscle cells (Figure 5D), while DLL4 was intensively stained in the muscles (Figure 5F). When J1 is weakly expressed on the vessel walls (Figure 5H), J2 is strongly expressed (Figure 5J).

In addition, CD34+ (Figure 5E/CD34) and PCK (Figure 5I) staining were used for the confirmation of vascular structures and the monitoring of gland structures. Due to the biopsy specimen taken from the distal endocervical region, vessels and smooth muscle structures were observed in the examined areas.

The difference between the groups was found to be significant for IHC (P < .05; Figures 3J and 5L: except for N2, DLL3) and WB (P < .05; Figures 3I and 5K: except for N2, DLL3, and J1).

Conclusions

In pregnancy, the mother’s immune system adapts to a fetal/placental unit that is genetically distinct from hers. During normal quiescent pregnancies, Th2 is the dominant immunological condition. ¹⁶ Suppression of Th1 responses in the uterus and enhancement of Th2 are necessary to reduce the risk of spontaneous abortion because Th1-type cytokines produce proinflammatory and prolabor responses. ²⁷ The Notch ligand δ stimulates Th1, while Jagged evokes Th2 differentiation. ²⁸ There is evidence that preterm labor following intrauterine lipopolysaccharide begins by the δ ligand, which affects angiogenesis in the uterus and placenta. Lipopolysaccharide treatment suppressed Notch-dependent angiogenesis production of Jagged 1 and 2, DLL-4, and VEGF. ²⁹

Examining the results of both groups: the control group in which the pregnancy continues and the misoprostol group in which the labor is induced, the cells and structures expressing the Notch signal receptor and ligand were altered. However, when cervical tissue is evaluated in terms of increased expressions, the N1 receptor, and DLL1 and J2 ligands appear to increase with the onset of preterm labor (Figure 6). This suggests that DLL1 and J2 ligands tend to work together to preserve a balance in preterm labor rather than the expression of a single ligand.

OPEN IN VIEWER

Figure 6.

Preterm labor with misoprostol induction in cervix (cross talk). a indicates artery; Leu, leukocyte; mac, macrophage; Pg ER, prostoglandin E Receptor, smc, smooth muscle.

It is thought that PGs may perform their regulatory role on the uterus and cervix during pregnancy by using autocrine and paracrine pathways. ⁶ In this process, Notch1 is thought to mediate the stromal differentiation of the uterus and is important for complete decidualization. ¹⁷ Leukocyte invasion is also considered necessary in remodeling the extracellular matrix during the normal labor process. ²⁰ However, in mice, incompatible macrophage polarization in the maternal and fetal compartments was found to be associated with spontaneous abortion ³⁰ and intrauterine infections. ³¹

Macrophages are generally thought to originate from circulating monocytes that migrate from blood vessels and then transform into macrophages in the peripheral tissue. ^32,33

It was found that cells present in the stroma, especially in the control group, and which expressed strong N1, were macrophages. With the onset of labor, it was observed that the expression strength of these cells disappeared in the stroma, but the intensity of N1 staining in the muscles increased considerably. It was also found that the increase in leucocyte counts observed in the bleeding foci was higher in the misoprostol group. This led us to assume that the monocyte–macrophage transformation process, which is particularly likely to occur with possible diapedesis, can end with the onset of preterm labor, and with the contraction–relaxation control of the prostaglandin receptors, ³⁴ the expression started to increase with the induction of these receptors in the muscle.

There is evidence in published literature that the Notch signal in the uterus of a mouse participates in the polarization process in macrophages, DLL1 and N1 affect the cytokines involved in M1/M2 conversion. ²⁹ There are studies in literature showing that Notch 1,4 and DLL1,4, J1 are expressed in the endothelium, and Notch 1,3 and DLL1, J1,2 are expressed in the smooth muscle cells in the uterus and blood vessels. ^35,36 In inflammatory-induced preterm labor, it is known that in the uterus and placenta, N1 and DLL1 increase, whereas angiogenesis-specific ligands J1,2 and DLL4 decrease. ¹⁹ It is thought that leukocytes perform DLL1 expression, which facilitates trophoblast invasion by providing IFN-γ release through interaction with Notch receptors, leading to vascular smooth muscle cell damage. ³⁷ Notch ligands DLL4, J1,2 are expressed by decidual cells and trophoblasts. ^38,39 Some studies suggest that proinflammatory factors may be associated with M1 macrophage polarization as well as upregulation in Notch-signaling molecules. ⁴⁰ However, the expression status in the cervix and macrophages is unknown.

Our study results showed that Notch receptors and ligands were expressed by different cells before and after the onset of preterm labor. Initially, it was seen that blood vessels performed DLL4 expression, muscle cells expression N3, and macrophages N1 expression; and with the onset of preterm labor, DLL1 and J2 were expressed in blood vessels, N1 was expressed in muscles, and N4 was expressed in macrophages.

We think that leukocyte/macrophages may be effective in modifying the Notch signal and that muscles can produce a contractile response with N1 receptor expression. When the process is evaluated together with protein expression levels, reciprocal cross talk between the cervical structures is important (Figure 1), and it may be thought that with the change in balance between the DLL1 and J2 ligands, preterm labor may start with N1 activation. However, we also think that the blood vessels in the cervix may contain prostanoid receptors, and sensitive to the levels of oxytocin in the body, they may regulate the hormonal step that controls the preterm labor by Notch signaling. A detailed determination of the role of molecular mechanisms in the preterm labor process may help clarify unknown points as well as contribute to the assessment of treatment potencies of topically used agents in the clinic.