The immunolocalization of cadherins and beta-catenin in the cervix and vagina of cycling cows

The adherens junctions (AJs) maintain the epithelial cell layers' structural integrity and barrier function. AJs also play a vital role in various biological and pathological processes. AJs perform these functions through the cadherin-catenin adhesion complex. This study investigated the presence, cell-specific localization, and temporal distribution of AJ components such as classical type I cadherins and beta-catenin in the cow cervix and vagina during the estrous cycle. Immunohistochemistry and Western blot analysis results demonstrated that beta-catenin and epithelial (E)-, neural (N)-, and placental (P)-cadherins are expressed in the cow cervix and vagina during the estrous cycle. These adhesion molecules were localized in the membrane and cytoplasm of the ciliated and non-ciliated cervical cells and the stratified vaginal epithelial cells. Positive immunostaining for P-, N-cadherin, and beta-catenin was also observed in the vascular endothelial cells of the cervical and vaginal stroma. Quantitative immunohistochemistry examinations revealed that in the cervical and vaginal epithelia, P-cadherin's optical density values (ODv) were the highest; in contrast, the N-cadherin ODv were the lowest. The ODv of P-cadherin and beta-catenin in the cervical epithelium and E-cadherin in the vagina were significantly higher in the luteal phase versus the follicular phase of the estrous cycle. Furthermore, the ODv of P-cadherin, N-cadherin, and beta-catenin in the cervix's central and peripheral epithelial regions were different during the estrous cycle. These findings indicate that classical cadherins and beta-catenin in the cervix and vagina exhibit cell- and tissue-specific expression patterns under the influence of estrogen and progesterone hormones during the estrous cycle.


Introduction
The epithelial cells of the female reproductive tract (FRT) undergo a marked increase in growth, proliferation, and differentiation under the influence of estrogen (E2) and progesterone (P4) hormones secreted by the ovary during the estrous cycle. In this process, the maintenance of appropriate cell-cell adhesion and tissue integrity is essential for the physiology of FRT epithelial cells (Rowlands et al. 2000). Adherens junctions (zonula adherens, AJs) mediate cell-cell adhesion between neighboring epithelial cells and are crucial for epithelial tissue integrity and barrier function in reproductive tract organs (Blaskewicz et al. 2011). The core components of AJs are clusters of cadherin molecules and a group of intracellular anchor proteins, referred to as catenins (Gumbiner 2005).
The cadherins superfamily comprises four major subfamilies: classical, desmosomal, proto-, and atypical cadherins (Harris and Tepass 2010). Classical cadherins were initially identified as Ca 2+ -dependent, homophilic adhesion molecules in vertebrates. Based on phylogenetic relationships, classical cadherins are subdivided into two families, namely, type I [epithelial (E)-, placental (P), neural (N)-, and retinal (R)-cadherins] and type II [vascular endothelial (VE)-, kidney (K)-, and osteoblast (OB)-cadherins] (Nollet et al. 2000). E-cadherin is expressed in all epithelial tissues and plays an essential role in the development, cell differentiation, tissue homeostasis, and the establishment and maintenance of apicobasal polarity and structural integrity 1 3 of epithelial tissues (Nollet et al. 2000;van Roy and Berx 2008). E-cadherin is also a tumor suppressor protein frequently lost in epithelial malignancies (Pećina-Slaus 2003). N-cadherin is highly expressed in neural, endothelial, mesenchymal, and muscle cells (Radice 2013). In contrast, it is absent in normal epithelia and is upregulated in many invasive tumors (Derycke and Bracke 2004). P-cadherin is present in placental structures and undifferentiated cells in normal adult epithelial tissues, such as the basal layer of the epidermis, breast, prostate, ovary, cervix, and lung. P-cadherin plays a crucial role in cell-cell adhesion and is involved in cell growth, differentiation, motility, and survival (Vieira and Paredes 2015).
Classical and desmosomal cadherins feature an aminoterminal extracellular region (ectodomain) composed of five extracellular cadherins (EC) repeats and a carboxy-terminal intracellular region (cytoplasmic domain). Interactions between the ectodomains of classical cadherins on opposed cells mediate specific cell-cell contacts. In contrast, the cadherin cytoplasmic domain functionally links to cytoskeletal actin filaments through catenins [alpha (α)-, beta (β)-, and gamma (γ)-catenin]. Among adaptor proteins, betacatenin is a dual-function protein involved in the regulation and coordination of cell-cell adhesion and in the activation of transcription of Wnt target genes that regulate apoptosis and cell cycle control (Nelson and Nusse 2004). The cadherin cytoplasmic domain establishes a high affinity, 1:1 complex with beta-catenin, and beta-catenin binds to alphacatenin with a lower affinity (Huber and Weis 2001). The cadherin-beta-catenin complex coordinates all cells' form, polarity, and function in the epithelium and plays a vital role in maintaining epithelial integrity. Furthermore, this complex checks Wnt/beta-catenin signals (Tian et al. 2011).
Like the uterine and oviduct epithelial tissues, the epithelium lining the cervix and vagina undergoes extensive organspecific morphological changes associated with circulating estrogen (E2) and progesterone (P4) hormones in cycling animals. In the cow cervix and vagina, cyclical variations in circulating concentrations of E2 and P4 regulate several physiological processes, such as epithelial proliferation, the production and rheological properties of cervical mucus (Pessina et al. 2006;Larsen and Hwang 2011;Tsiligianni et al. 2011), and vaginal pH and secretion viscosity (Rizvi et al. 2009). Throughout pregnancy, progesterone promotes the production of highly viscous mucus resulting in the formation of a plug that temporarily seals the cervix so that pathogens do not harm the fetus (Sheldon et al. 2014). Thus, the cervical and vaginal epithelia are physical and biochemical barriers preventing bacterial pathogens and foreign material from ascending to cows' uterus, especially during pregnancy. In particular, the enlarged cervix allows bacteria to migrate from the external environment into the uterine lumen after parturition. When microorganisms break the epithelial barriers in the cervix and vagina, the reproductive tract becomes contaminated and inflamed (Sheldon et al. 2014;De Tomasi et al. 2019). Continuation of this condition may cause clinical diseases that lead to subfertility or infertility (Sheldon et al. 2014).
In cows, in contrast to the upper genital tract epithelium, the cervical and vaginal epithelia are permanently colonized by diverse microorganisms. The commensal cervical and vaginal microbiota play a vital role in the protection of the genital tract from pathogenic microbes by competition effect, and the microbial population diversity of cervicovaginal mucus is different depending on the estrous cycle phase (Wang et al. 2018;Quereda et al. 2020;Srinivasan et al. 2021). Recent studies have demonstrated pathogenic bacterial infections disrupt the cadherin-mediated cell-cell adhesion and the epithelial barrier in the cervix and vagina (Prozialeck et al. 2002;Politi et al. 2008;Devaux et al. 2019). In light of this information, it seems essential to maintain the integrity of the epithelial barrier in the cow cervix and vagina against pathogen invasion because the inability to reconstitute epithelial barrier function during the estrous cycle may cause significant pathophysiological consequences. Therefore, it is important to know the adhesion molecule expression in the cow cervix and vagina during the normal estrous cycle.
Many studies have been published on cadherin and betacatenin expression in the epithelia of the upper reproductive tract organs of various mammalian species, including 1 3 humans (Inoue et al. 1992;Fujimoto et al. 1996;Shih et al. 2004;Tsuchiya et al. 2006), mice (MacCalman et al. 1994Potter et al. 1996), monkeys (Allan et al. 2003), pigs (Ryan et al. 2001;Kiewisz et al. 2011), dogs (Yue et al. 2009Guo et al. 2010;Payan-Carreira et al. 2016), and cattle (Caballero et al. 2014;Tienthai 2018) during pregnancy and the estrous cycle. However, few studies have addressed cadherin and beta-catenin expression in normal cervical and vaginal epithelial cells (Inoue et al. 1992;Blaskewicz et al. 2011;Crasta et al. 2016). In addition, there are no data on whether altered levels of E2 or P4 cause significant changes in the expression of epithelial adhesion molecules and epithelial barrier function in the cervix and vagina. Moreover, differences in cadherins and catenin expressions in the cervical and vaginal epithelial cells between the phases of estrous cycles of mammalian species have not been studied in detail until now.
Given the increasing importance of AJs in regulating various functions in reproductive system organs, including maintaining epithelial tissue integrity and barrier function, and there is no information regarding the expression and localization of AJ proteins in cow cervix and vagina, this study intends: to investigate the cell-specific expression and localization patterns of classical E-, P-, N-cadherins, and beta-catenin proteins in normal cow cervix and vagina during the estrous cycle. Elucidation of the cell-and organspecific manner of AJ protein expression in the cow cervix and vagina may provide new insights regarding the potential roles of these proteins in hormone-dependent organs.

Animals' ethics statement and experimental conditions
In the present study, tissue samples of animals were taken under the Regulation rules on the Working Procedures and Principles of Animal Experiments Ethics Committees of the Republic of Turkey Ministry of Forestry and Water Affairs (dated 15 February 2014, 28,914). In this study, the reproductive tissue samples of healthy 2-8 years old Holstein cows (n = 30), which we used in our previous study, were used (Sağsöz et al. 2019). These tissue samples were obtained from local abattoirs in Diyarbakır Province, Turkey. Before slaughter, cows were examined for evidence of estrous behaviors, including mounting or attempting to other cattle, smelling and trailing other females, vulvar swelling and reddening, clear vaginal mucus discharge, and mucus smeared on the rump (Peralta et al. 2005). Immediately after the death of the cows, the entire reproductive tract was removed and macroscopically examined for the presence of disease. This study included tissue samples without pathologic signs, namely necrotic or hemorrhagic uterine mucosa, hyperaemic and oedematous cervical or vaginal mucosa, malodorous or non-odorous, purulent or mucopurulent vaginal discharge (Millward et al. 2019).

Collection of blood samples and measurement of hormone concentrations
The blood of pre-selected cows was collected into a tube at post-mortem bleeding and transported to the laboratory immediately after collection to measure serum concentrations of E2 and P4 hormones. Serum samples were obtained by centrifugation at 4 °C, 3969Xg for 5 min, and stored at − 20 °C. Serum E2 and P4 measurements were performed in a clinical laboratory (PRO-LABORATORY Laboratory Technologies, Istanbul) using commercially available kits [DRG Aurica Elisa Oestradiol Kit (Cat. No. EIA-2693) and DRG Aurica Elisa Progesterone Kit (Cat. No. EIA-1561)] by enzyme immunoassay (EIA) method, according to the manufacturer's protocol.

Tissue samples collection and processing
In the present study, ovaries and uteri were used for histological analysis to determine the estrous cycle. After the macroscopic examination of the entire genital tract of the killed cows (n = 30), each cow's right and left ovaries were longitudinally cut into two halves. One-half each of the right and left ovaries were used for histological analysis. The uterine horns were also opened by cutting the uterine wall along the anti-mesometrial side. In the cow cervix, the cervical mucosa forms three to four annular folds or rings that project into the lumen and numerous smaller longitudinal folds (Breeveld-Dwarkasing 2002). Therefore, the cervical tissue samples used in this study were harvested from all three rings of the cervix and the vaginal area adjacent to the vulva. One tissue fragment of each animal's cervix and vagina was used for histological and immunohistochemical analysis. After the dissection, all reproductive tissue samples were fixed in a 10% buffered formalin solution and routinely processed for inclusion in paraffin. Furthermore, for western blot analysis, cervical and vaginal tissue samples taken from one follicular and one luteal phase were frozen in liquid nitrogen.

Determination of estrous cycle phase
The phase of the estrous cycle of each cow was determined using macroscopic, histological, and biochemical parameters as described by Benbia et al. (2017), with slight modification. The presence of a preovulatory follicle and fully developed corpus luteum (CL) in the ovaries were assumed as the characteristic features of the follicular and luteal phases of the estrous cycle, respectively. For histological determination of the estrous cycle phase, ovarian, uterine, cervical, and vaginal tissue Sects. 5 μm-thick were stained with a modified Mallory's connective tissue stain (Crossmon 1937).
The estrous cycle of the cows is 21 days long (range 18-24 days) and divided into the follicular and luteal phases based on the dominant structures of the ovary. The follicular phase is the period from the regression of corpus luteum (CL) until ovulation occurs, and it comprises proestrus and oestrus. The luteal phase is from ovulation until CL regression, including metestrus and diestrus (Crowe 2011;Forde et al. 2011). Based on literature information (Crowe 2011;Forde et al. 2011;Benbia et al. 2017;Kim 2018) and the findings mentioned above in this study, we divided the cows into two groups: the follicular phase group (n = 13) and the luteal phase group (n = 17).
In order to control the specificity of the immunostaining of the cadherin and beta-catenin, negative and positive control tissues were used. Archived blocks of the bovine uterus and placenta were positive controls for E-, P-, N-cadherin, and beta-catenin. Archived paraffin blocks of the bovine ovary and liver were also stained for N-cadherin. Normal rabbit IgG (sc-2027, Santa Cruz Biotechnology, Santa Cruz, CA, USA) instead of E-and P-cadherin and normal mouse IgG (sc-2025, Santa Cruz Biotechnology, Santa Cruz, CA, USA) instead of anti-N-cadherin and anti-beta-catenin antibodies were used as negative controls. Both IgG had the same concentration as the primary antibodies.

Microscopic evaluation and imaging
In the present study, to define the location of the epithelial cells in the cervical mucosa, it was used that the terminology described by Mullins and Saacke (1989). Firstly, the epithelium surrounding the cervical lumen and lining the primary longitudinal folds was described as the central epithelium. The epithelium covering the secondary folds was called the peripheral region epithelium. Secondly, the epithelium was defined according to its location in the grooves. The term "basal area" was applied to the areas within grooves, while "apical area" referred to the areas between the grooves. The histological architecture of the cervical epithelium is given in the supplementary figure.
After IHC, slides were examined by conventional light microscopy using an Olympus BX51 microscope (Olympus, Tokyo, Japan). The immunostainings for E-, P-, N-cadherin, and beta-catenin were evaluated in two different ways in the study. First, immunostainings for each adhesion molecule were assessed using a four-point intensity score (IS) (Detre et al. 1995). Positive immunostainings for all cadherins and beta-catenin were determined in high-expression areas by scanning the cervical and vaginal sections at magnifications of X40, X100, X200, and X400. The staining was scored as negative, weak, moderate, and strong (or intense). The subcellular, cellular, and tissue localization of E-, P-, N-cadherin, and beta-catenin were evaluated independently for three tissue layers (epithelium, stroma, and smooth muscle layer) and 1 3 blood vessels in the cervix and vagina. The serosa was present in only some cervical sections as it was lost during the fixation and embedding procedures and is, therefore, not included in the results. Secondly, the immunostaining intensities of betacatenin and cadherin proteins in the cow cervical and vaginal tissue sections were evaluated with a semi-quantitative IHC method, as Crowe and Yue (2019) reported. One researcher (NL) assessed the staining results as an observer in a blind fashion. The three animals' cervical and vaginal tissue sections for each estrous cycle phase were quantitatively evaluated. Epithelial immunostaining was performed in each animal's cervix's central and peripheral region epithelium. A minimum of three representative areas of each region of the cervical epithelium and three different fields of the vaginal epithelium per cow per estrous cycle phase were randomly selected and photographed with an Olympus DP72 digital color microscope camera (Olympus, Tokyo, Japan). All images were captured using a 40X (NA 0.85) objective lens and were at an ultrahigh-resolution mode (4140 × 3096 pixels).

Digital image analysis for beta-catenin and cadherins
Free Fiji-ImageJ software with a color deconvolution plugin (version 1.2; https:// imagej. net/ Fiji/ Downl oads) was used to determine the semi-quantitative expressions of beta-catenin and cadherin proteins. After importing the TIFF image into Fiji-ImageJ software, the actual color was deconvoluted into three colors (green, gray, and blue) using the H DAB vector. The threshold tool converted the second color (gray) image into a black-and-white contrast for further processing. The threshold values of each tissue and antibody were set and held constant in analyzing all images. The staining intensity was measured as the "mean gray value" parameter. The mean gray value was measured by the "measure" function and then converted to optical density (OD) values using the formula OD = log 10 (Max intensity/Mean intensity) as described in the ImageJ User Guide (https:// imagej. nih. gov/ ij/ docs/ user-guide-USboo klet. pdf). Max Intensity is equal to 255 for the eight-bit image. Mean intensity was quantified as (255-"Mean gray value"). The immunostaining intensity was expressed as an OD value (scale range 0-2.4) (Irwin et al. 2016).

Western Blot (WB) analysis
To confirm the expression of the P-cadherin, N-cadherin, E-cadherin, and beta-catenin proteins in the cow cervix and vagina, Western blot analysis was used. For total protein extraction, cervical and vaginal tissue samples taken from one follicular and one luteal phase of normal cervix and vagina were lyzed with an ultrasonic homogenizer in a cold RIPA lysis buffer (sc-24948, Santa Cruz Biotechnology, Santa Cruz, CA) containing a protease inhibitor cocktail, phenylmethylsulfonyl fluoride (PMSF) and sodium orthovanadate and centrifuged (15 000Xg) for 15 min at 4 °C. The supernatants containing proteins were removed, and the protein concentrations were determined by Lowry assay. Equal amounts of protein samples (40 µg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a 7% running gel and then electrically transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Bedford, MA, USA) which were pretreated with methanol (1 min) and transfer buffer (48 mM Tris, 39 mM glycine, 10% (v/v) methanol) at a constant mA of 100 mA for 16 h. After the transfer, the membranes were blocked with 5% non-fat dry milk in TBS-T (Tris-buffered saline with 0.1% Tween-20) for one h at room temperature and then incubated overnight at 4 °C with the primary antibodies diluted in TBS-T containing 5% non-fat dry milk. The antibodies were identical to those used for immunohistochemistry, diluted at 1:200 (N-cadherin) and 1:500 (E-cadherin, P-cadherin, and beta-catenin). Afterward, the membranes were washed in TBS-T three times for 10 min and incubated with the horseradish peroxidaseconjugated secondary antibodies [anti-rabbit IgG (sc-2004) or anti-mouse IgG (sc-516102), Santa Cruz Biotechnology, CA, USA] for one h at room temperature. Membranes were washed four times in TBS-T for 5 min and two times in TBS for 5 min. Then, the membranes were incubated in enhanced chemiluminescence (ECL) substrate (Cat: 170-5061, Bio-Rad Laboratories, Hercules, CA) for 5 min, and the chemiluminescence signal was visualized using ChemiDoc™ XRS + System (Bio-Rad Laboratories, Hercules, CA). The membranes were stripped and reprobed with a beta-actin mouse monoclonal antibody (sc-69879, Santa Cruz Biotechnology, CA) at a 1:500 dilution to confirm the gel's equal loading of protein samples. Furthermore, to test the specificity of cadherin and beta-catenin antibodies against the bovine tissues, the lysates of the cow placenta and ovary were used as a positive control in Western blot analysis.

Statistical analysis
The results of the immunohistochemical analysis were presented as mean ± standard deviation (SD), and SPSS Statistics performed the statistical analysis for Windows, Version 20.0 (Armonk, NY: IBM Corp). The Shapiro-Wilk test assessed data normality. Student's t-test was used to analyze whether there was any significant difference in the OD values of the cervical and vaginal epithelium for beta-catenin, E-cadherin, N-cadherin, and P-cadherin between different phases of the estrous cycle or between the different regions of the cervical epithelium during each phase. Differences in the values were considered statistically significant when the P-value was < 0.05.

Immunolocalization of cadherins and beta-catenin in the cervix and vagina during the oestrus cycle
The streptavidin-peroxidase technique demonstrated that the positive control tissue slides were immunopositive for cadherins and beta-catenin (data not shown). In contrast, negative control slides of the cervix and vagina treated with a normal rabbit or mouse IgG instead of primary antibodies showed no immunostaining (Figs. 1,2,3,4,5,6,7,and 8).
Our results showed that P-cadherin, N-cadherin, E-cadherin, and beta-catenin proteins were localized in the cervical and vaginal tissue sections throughout the estrous cycle. The immunolocalization patterns of cadherins and beta-catenin in the cervical and vaginal cells are presented in Figs. 1, 2, 3, 4, 5, 6, 7, and 8. The OD and mean gray intensity values of cadherins and β-catenin levels in the cervix and vagina during the follicular and luteal phases of the estrous cycle as determined by immunohistochemistry are given in Fig. 9. Specifically, cadherins and beta-catenin were immunolocalized to the cervical and vaginal epithelial cells throughout the estrous cycle.

Cervical epithelium
During both phases of the estrous cycle, P-cadherin (Fig. 1A, B, E, and F) displayed strong membranous and moderate cytoplasmic expression patterns in the ciliated cells and moderate membranous immunolocalization in the non-ciliated cells. N-cadherin showed only moderate cytoplasmic immunostaining in the ciliated cells of the central and peripheral regions of the cervical epithelium during both phases of the estrous cycle ( Fig. 2A, B, E, and F). However, a relatively strong localization of E-cadherin (Fig. 3A, B, E, and F) and beta-catenin (Fig. 4A, B, E, and F) was observed in the lateral membrane of both ciliated and non-ciliated cells in the epithelium of the central and peripheral regions. Moreover, the cytoplasm of the ciliated epithelial cells exhibited weak to moderate immunostaining for E-cadherin and beta-catenin. Also, the basal compartment of these cells showed intense immunostaining for all proteins.

Vaginal epithelium
We observed that the vaginal epithelium stained positively for P-cadherin ( Fig. 5A and C), N-cadherin ( Fig. 6A and C), E-cadherin ( Fig. 7A and C), and beta-catenin ( Fig. 8A and C) throughout the estrous cycle. During both the follicular and luteal phases, the epithelium's basal and parabasal cell layers showed moderate cytoplasmic and strong membrane staining for P-and E-cadherin (Figs. 5A and C, and 7A and C, respectively) and beta-catenin ( Fig. 8A and C). During the follicular phase, the superficially located, tall, columnar, highly active, and mucus-secreting cells displayed strong membrane and weak cytoplasmic staining for P- (Fig. 5A) and E-cadherin (Fig. 7A), and beta-catenin (Fig. 8A). During the luteal phase, strong membrane and weak to moderate cytoplasmic immunostaining for P-cadherin (Fig. 5C), E-cadherin (Fig. 7C), and beta-catenin (Fig. 8C) were observed in the superficial squamous cells. However, during both the follicular and luteal phases, the vaginal epithelium showed weak cytoplasmic N-cadherin immunostaining throughout its layers (Fig. 6A, C).

Quantitative analysis of cadherins and beta-catenin in the cervical and vaginal epithelium
In the cervical epithelium, OD values of P-cadherin, N-cadherin, and beta-catenin among both phases of the estrous cycle were statistically different (P < 0.05), but not OD values of E-cadherin (P > 0.05). When P-cadherin OD values in the cervix's central and peripheral epithelial regions were compared, it was observed that the differences between the two regions were nonsignificant in both phases of the estrous cycle (P > 0.05). The OD values of N-cadherin in the central and peripheral epithelial regions were statistically different during the follicular phase (P < 0.001) but not during the luteal phase (P > 0.05). In contrast, OD values of beta-catenin among the central and peripheral epithelial regions were significantly different during the follicular (P < 0.05) and luteal (P < 0.01) phases of the estrous cycle (P < 0.05). When the follicular and luteal phases were compared, statistical differences were observed in the OD values of P-cadherin and N-cadherin for each region of the cervical epithelium (P < 0.001). In contrast, there were statistically important differences in the OD values of beta-catenin for only peripheral region epithelium (P < 0.001) (Fig. 9A).
In the vaginal epithelium, there were no statistical differences (P > 0.05) in OD values of P-cadherin, N-cadherin, and beta-catenin between the cycle phases. However, OD values of E-cadherin among the follicular and luteal phases of the estrous cycle were significantly different (P < 0.05) (Fig. 9B).

Stroma and muscle layer of the cervix and vagina
The cervical and vaginal smooth muscle cells exhibited weak to moderate cytoplasmic immunolabelling for P-cadherin (Fig. 5E) but did not stain positive for N-cadherin, E-cadherin, and betacatenin (data not shown). Furthermore, a moderate cytoplasmic expression pattern for P-cadherin was observed in the endothelial cells and the vascular smooth muscle cells in the cervical ( Fig. 1A and E) and vaginal (Fig. 5G) stroma. In contrast, strong N-cadherin (Figs. 2E, and 6C and E) and beta-catenin (Figs. 4E and F, and 8C and E) immunoreactivities were only observed in the lateral plasma membrane of endothelial cells.

Western blot analysis of cadherins and beta-catenin in the cervix and vagina
Western blot analysis confirmed that the antibodies used in the present study are specific to the cow cervix and vagina. As shown in Fig. 10a and e, the specific bands for P-cadherin were observed in the cow cervix, vagina, and placenta. P-cadherin was detected as a single band with a molecular weight of 130 kDa in the cervix (Fig. 10a), while it was observed as two bands with molecular weights of 91 and 130 kDa in the vagina and placenta (Fig. 10e). The N-cadherin protein was detected as a single band with a molecular weight of 130 kDa in the cow cervix, ovary (Fig. 10b), and vagina (Fig. 10f). However, the N-cadherin protein did not express in the cow placenta (Fig. 10b). As seen in Fig. 10b, N-cadherin expression in the cervix was higher in the follicular phase than in the luteal phase. E-cadherin was detected as a single band with a molecular weight of 80 kDa in the cervical (Fig. 10c) and vaginal tissue samples (Fig. 10g). In contrast, it was determined as three bands with molecular weights of 80, 100, and 120 kDa in the cow placenta ( Fig. 10c and f). Beta-catenin was observed as three bands with 70, 85, and 92 kDa in the cervix, vagina, and placenta ( Fig. 10d and h).

Discussion
This is the first study in which the expression, localization, and temporal distribution of the cadherin-mediated pathway in intercellular adherens junction in the cow cervix and vagina during the follicular and luteal phases of the estrous cycle have been examined by Western blot analysis and immunohistochemistry.
There is very little data on cadherins and beta-catenin expression in mammals' normal cervix and vagina (Inoue et al. 1992;Blaskewicz et al. 2011;Crasta et al. 2016). Therefore, the present study investigated whether several Fig. 4 Beta-catenin immunoreactivity in the cow cervix during the follicular and luteal phases of the estrous cycle; immunohistochemical stain, diaminobenzidine as the chromogen. In the cervix, beta-catenin (A, B, E, and F) exhibited membrane staining (arrowheads) in the ciliated (Cc) and non-ciliated cells (Nc) of the central and peripheral region epithelium of the cervix. Beta-catenin immunoreactivity was also detected in the lateral plasma membrane (arrowhead) of the endothelial cells (ve) of both the cervical blood vessels (E and F). Negative control, produced using mouse IgG, resulted in no immunostaining for beta-catenin in the cow cervix (C, D, G, and H). a, the apical area of the grooves; b, basal area of the groves; Ce, central epithelium; g, grooves in the central and peripheral regions of the cervix; i, intraepithelial lymphocyte; S, stoma; v, blood vessel. Scale bars: 10 µm (A, B, E, and F) and 20 µm (C, D, G, and H) During the follicular phase, the mucus-secreting superficial cells displayed membranous and cytoplasmic staining for P-(A), while during the luteal phase, the superficial squamous cells showed membranous and cytoplasmic immunostaining patterns for P-cadherin (C). P-cadherin immunoreactivity was observed in the nuclei and cytoplasm of some stromal cells (sc) (C), smooth muscle cells (smc) (E), vascular endothelial (ve) and smooth muscle cells (vsmc) (G). The negative control was produced with rabbit IgG, and no P-cadherin immunoreactivity was observed in any structure (B, D, F, and H). S, stroma; v, blood vessel. Scale bars: 10 µm (A, C, E, and G) and 20 µm (B, D, F, and H) adhesion proteins are expressed in the cow cervix and vagina using polyclonal and monoclonal antibodies produced against P-, N-and E-cadherin and beta-catenin proteins. Our study confirmed the specificity of antibodies for bovine tissues by western blot analysis. The results of the western blot analysis indicated that P-, N-and E-cadherin and beta-catenin proteins are expressed in cervical and vaginal tissue samples, one from the follicular and the other from the luteal periods. In addition, the immunohistochemistry results showed that P-, N-and E-cadherin and beta-catenin were mainly localized to the cervical and vaginal epithelium. This finding coincides with the knowledge that P-, N-and E-cadherin and beta-catenin are expressed in the epithelia of the reproductive tract organs of various mammalian species, including humans (Inoue et al. 1992;Fujimoto et al. 1996;Shih et al. 2004;Tsuchiya et al. 2006;Blaskewicz et al. Fig. 6 N-cadherin immunoreactivity in the cow vagina during the follicular and luteal phases of the estrous cycle; immunohistochemical stain, diaminobenzidine as the chromogen. During both the follicular and luteal phases, the vaginal epithelium (e) showed weak cytoplasmic N-cadherin immunostaining throughout its layer (A and C). Strong N-cadherin immunoreactivity was localized in the lateral plasma membrane of vascular endothelial cells (ve) in the vaginal stroma (E). The negative control was produced with mouse IgG, and no N-cadherin immunoreactivity was observed in any structure (B, D, and F). S, stroma; v, blood vessel. Scale bars: 10 µm (A, C, and E) and 20 µm (B, D, and F) 2011; Crasta et al. 2016), mice (MacCalman et al. 1994Potter et al. 1996), monkeys (Allan et al. 2003), pigs (Ryan et al. 2001;Kiewisz et al. 2011), dogs (Yue et al. 2009;Guo et al. 2010;Payan-Carreira et al. 2016), and cattle (Caballero et al. 2014;Tienthai 2018) during pregnancy and the estrous cycle.
In the present study, we determined that the optical densities (OD) of P-and N-cadherin and beta-catenin immunostainings in the cervix were statistically different between both phases of the estrous cycle of cows (P < 0.05). In addition, we determined that the OD values of E-cadherin in the vagina were statistically different between the two phases of the cow's estrous cycle. In contrast, the OD values of P-cadherin, N-cadherin, and beta-catenin in the vaginal epithelium did not significantly change during the estrous cycle. Unlike our findings in the cow vagina but similar to our results in the cow cervix, previous studies have shown that the expression of E-cadherin and catenin in the endometrium does not change during the selected estrus phase (Caballero et al. 2014;Tienthai 2018) or the menstrual cycle (Tabibzadeh et al. 1995;Tsuchiya et al. 2006;Carico et al. 2010). However, some studies in the uterus of humans (Fujimoto et al. 1996;Shih et al. 2004) and animals (MacCalman et al. 1994;Payan-Carreira et al. 2016) reported that both E2 and P4 were able to induce E-cadherin transcription. Together with this knowledge, our current findings suggest that changes in the adhesion molecule expression can be associated with the hormonal status of the cow cervix and vagina.
The cow cervical mucosal surface consists of a complex system of longitudinally arranged primary folds, secondary folds produced by longitudinal division within primary folds, and numerous uniformly spaced shallow grooves along all cervical structure surfaces (Mullins and Saacke 1989). In the cow cervix, the central and peripheral region epithelia and the epithelium lining grooves contain two distinct cell types; (1) non-ciliated secretory columnar and (2) ciliated secretory columnar epithelial cells. The structure of the epithelium and the amount and chemical composition of the mucin secreted by the non-ciliated columnar cells vary considerably depending on the levels of E2 and P4 hormones at different phases of the cycle (Wordinger et al. 1973;Mullins and Saacke 1989). The epithelial cells are tall columnar due to an accumulation of mucus in nonciliated cells during the follicular phase, while their size reduces during the luteal phase (Wordinger et al. 1973). In this study, we investigated whether the OD values of cadherins and beta-catenin immunostainings change due to structural changes in the cervical epithelium according  C). During the follicular phase, the mucus-secreting superficial cells displayed membranous and cytoplasmic staining for E-cadherin (A), while during the luteal phase, the superficial squamous cells showed membranous and cytoplasmic immunostaining for E-cadherin (C). The negative control was produced with rabbit IgG, and no E-cadherin immunoreactivity was observed in any structure (B and D). S, stroma; v, blood vessel. Scale bars: 10 µm (A and C) and 20 µm (B and D) to the phases of the cow's estrous cycle. Firstly, we examined whether the OD values of cadherins and beta-catenin were different between the central and peripheral regions of the cervix for each phase of the estrous cycle of cows. We observed that the differences between the OD values of P-cadherin in the cervix's central and peripheral epithelial regions were nonsignificant during both phases of the estrous cycle (P > 0.05). The OD values of N-cadherin in the central and peripheral epithelial regions were statistically different only during the follicular phase (P < 0.001).
In contrast, we detected that the OD values of beta-catenin among the central and peripheral epithelial regions were significantly different during the follicular (P < 0.05) and luteal (P < 0.01) phases of the estrous cycle (P < 0.05). These findings may indicate that the expression of cadherins and beta-catenin in the cervical epithelium is specific to the cervical mucosa region.
Secondly, we compared the follicular and luteal phases and observed statistical differences in the OD values of P-cadherin and N-cadherin for each region of the cervical Fig. 8 Beta-catenin immunoreactivity in the cow vagina during the follicular and luteal phases of the estrous cycle; immunohistochemical stain, diaminobenzidine as the chromogen. Throughout the estrous cycle, in the vagina, the basal (bc) and intermediate cell layers of the vaginal epithelium showed cytoplasmic and membranous staining (arrowheads) for beta-catenin (A and C). The membranous and cytoplasmic staining for beta-catenin was also observed in the mucussecreting superficial cells (A) and the superficial squamous cells (C) of the vaginal epithelium were found during the follicular and luteal phase, respectively. Beta-catenin immunoreactivity was also detected in the lateral plasma membrane (arrowheads) of the endothelial cells (ve) of both the cervical and vaginal blood vessels (C and E). Negative control, produced using mouse IgG, resulted in no immunostaining for beta-catenin in the cow vagina (B, D, and F). L, lumen; S, stroma; v, blood vessel. Scale bars: 10 µm (A, C, and E) and 20 µm (B, D, and F) epithelium (P < 0.001). In contrast, we detected significant differences in the OD values of beta-catenin for only peripheral region epithelium (P < 0.001). This expression of betacatenin and cadherin types that differ by location within the defined cervical structures would suggest that a complex pattern of adhesion molecules exists within the cervical epithelium and is determined by the structural features of the epithelium during the estrous cycle.
The present study's immunohistochemical results indicate that in the cervix, P-cadherin exhibited cytoplasmic and membranous expression patterns in the ciliated cells and lateral membrane localization in the non-ciliated cells during the follicular and luteal phases. This finding contrasts with previous studies showing that P-cadherin was absent in the normal columnar epithelium of the human cervix (de Boer et al. 1999;Han et al. 2000). E-cadherin and beta-catenin showed strong membranous and weak cytoplasmic expression patterns in the ciliated cells and moderate membranous localization in the non-ciliated cells of the cow cervix throughout the estrous cycle. This basic finding is consistent with previous studies that have reported E-cadherin (Vessey et al. 1995;de Boer et al. 1999;Ryan et al. 2001;Fadare et al. 2005;Blaskewicz et al. 2011;Auvinen et al. 2013) and beta-catenin (Fadare et al. 2005) expression in the lateral membrane of normal columnar epithelial cells lining the cervix, but not on the apical and basal cellular surfaces of these cells. However, this contradicts previous reports indicating no detectable intracellular E-cadherin and beta-catenin (de Boer et al. 1999;Fadare et al. 2005). N-cadherin exhibited a weak to moderate cytoplasmic immunostaining pattern in the ciliated cells of the cervical epithelium. This result is consistent with the findings of Jiang et al. (2019). However, it is inconsistent with Li et al. (2016) and Vornhagen et al.'s (2018) studies Fig. 9 The optical densities (OD) of cadherins and beta-catenin in the cow cervix (A) and vagina (B) during the follicular and luteal phases of the estrous cycle following IHC staining. Optical density was obtained by color deconvolution analysis. The OD values are displayed as mean ± SD. The results were evaluated using Student's t-test. Statistically significant results are marked by asterisks (*) directly in the graphs. * P < 0.05, ** P < 0.01, *** P < 0.001 of the human cervix. The presence of P-, N-and E-cadherin and beta-catenin in the cervical epithelium of cycling cows could confirm that these adhesion proteins are involved in maintaining the epithelial integrity (Paredes et al. 2012) and regulation of several cellular processes in the cervical epithelium throughout the estrous cycle.
Previous studies demonstrated that in the squamous epithelium of the human ectocervix and vagina, E-cadherin and beta-catenin are predominantly found along the cellto-cell borders in the basal and parabasal cell layers (Inoue et al. 1992;Vessey et al. 1995;Carico et al. 2001;Shinohara et al. 2001;Fadare et al. 2005;Blaskewicz et al. 2011;Auvinen et al. 2013;Zhang et al. 2014;Jiang et al. 2019;Donmez 2020) and P-cadherin is confined to the basal cell layer (Li et al. 2016). In contrast, the immunohistochemical findings indicate that in the cow vagina, P and E-cadherin and beta-catenin were localized to all of the cell layers of the stratified epithelium during the estrous cycle. The reason for this difference may be that the vaginal epithelium of the cow is different from the vaginal epithelium of most animals during the estrous cycle. In humans and most species, the superficial layers of the vaginal epithelium consist of dead squamous cells that have undergone a terminal cell differentiation program called cornification, which occurs , ovary, and vagina (f) but was absent from cow placentome (b). E-cadherin detected an intense band at 80 kDa in the cervix (c) and vagina (g), but it labeled three bands at ∼80, 100, and ∼120 kDa molecular weights in the placentome (c and g). Betacatenin antibody marked three bands with molecular weights ranging from 70, 85, and 92 kDa in the cervical (d), vaginal (h), and placental samples (d and h) under the influence of estrogen (Anderson et al. 2014). Consequently, terminally differentiated superficial cells do not have robust intercellular junctions (Anderson et al. 2014). In contrast, the luminal surface epithelium of the cow vagina is composed of mucus-secreting columnar cells during the follicular phase (Miroud and Noakes 1991). The desquamation is not observed in the vagina (Roark and Herman 1950).
The current study results indicate that the cellular localization patterns of these adhesion proteins varied with the structural changes that occur in the vaginal epithelium during the estrous cycle. During the follicular phase, the luminal surface columnar cells of the vaginal epithelium displayed intense membranous and weak cytoplasmic staining for Pand E-cadherins and beta-catenin. However, during the luteal phase, the luminal surface squamous cells of the vaginal epithelium showed moderate cytoplasmic, sometimes membranous expression for P-cadherin and strong membranous and moderate cytoplasmic immunostaining for E-cadherin and beta-catenin. Furthermore, the vaginal epithelial cells also exhibited nuclear immunostaining for P-cadherin throughout the estrous cycle. Fadare et al. (2005) demonstrated that in the normal human ectocervix, E-cadherin and beta-catenin decorated the epithelium in a circumferentially membranous fashion cytoplasmic or nuclear staining was present. However, Zhang et al. (2014) and Donmez (2020) showed that the normal epithelial cells of the human ectocervix displayed membranous and cytoplasmic beta-catenin expression in the basal and suprabasal layers, similar to the case in the cow vagina. These reports and the present study's findings prove that the epithelial localization patterns of P-and E-cadherin and beta-catenin are species-specific.
The OD values of N-cadherin in the cervical and vaginal epithelia were lower than those of the E-cadherin, P-cadherin, and beta-catenin OD levels. This finding is consistent with the results of Jiang et al. (2019), who demonstrated that the expression level of N-cadherin was very low in normal human cervical tissues. These differences indicate that the expression of junctional adhesion molecules is species-and tissue-specific.
E-and N-cadherin exhibit opposite effects, where E-cadherin mediates the adhesion between epithelial cells (van Roy and Berx 2008), and N-cadherin promotes cell movement (Hazan et al. 2004). Previous studies have demonstrated that the E-cadherin/catenin complex plays an essential role in maintaining the normal phenotype of epithelial cells, and E-cadherin is an important tumor suppressor (Jeanes et al. 2008;Tian et al. 2011;Jiang et al. 2019). Furthermore, studies have shown that E-cadherin and betacatenin expression gradually decreases with the progression of lesions in cervical cancers while N-cadherin expression increases (Islam et al. 1996;Rodríguez-Sastre et al. 2005;Mrozik et al. 2018;Jiang et al. 2019). Based on these reports (Islam et al. 1996;Orsulic et al. 1999;Jeanes et al. 2008;Jiang et al. 2019) and the present results, the low N-cadherin expression and strong membranous E-cadherin and beta-catenin expressions may be suggested to strengthened intercellular adhesion, prevent the formation of abnormal cells during the morphological changes that occur in the cervical and vaginal epithelium throughout the cow oestrus cycle, and inhibit tumorigenesis. However, further studies are needed to confirm these hypotheses.
Early studies have reported that cervical stromal cells do not exhibit any staining for E-, P-and N-cadherin (Ryan et al. 2001;Fadare et al. 2005;Li et al. 2016) and betacatenin (Fadare et al. 2005). Thus, we are not surprised to see no immunostaining for E-cadherin and beta-catenin in the cervical and vaginal stroma during the estrous cycle. Contrary, we observed nuclear and cytoplasmic immunostaining patterns with anti-P-cadherin antibodies in some cervical and vaginal stromal cells.
P-cadherin has a crucial role in conserving the structural integrity of epithelial tissues and regulates several cellular homeostatic processes important for cell differentiation, shape, growth, and migration that participate in embryonic development and maintain adult tissue architecture (Cavallaro and Dejana 20111). Based on this information, we could suggest that P-cadherin in stromal cells may be associated with the maintenance of adult tissue architecture during the estrous cycle by participating in molecular mechanisms that regulate cell differentiation, cell shape, growth, and migration. However, based on immunohistochemical findings, it is difficult to explain the function of P-cadherin and why it exhibits cytoplasmic and nuclear expression patterns in stromal cells. Therefore, further studies are required to clarify this issue.
To the best of our knowledge, no detailed information is available on the expression of cadherins and beta-catenin in the smooth muscle cells of the cervix and vagina of humans and other mammals. However, Taylor et al. (1996) demonstrated that while the human myometrium expressed numerous cadherins in a cell-specific manner, differing among smooth muscle cells, stromal cells, and endothelial cells, the expression of cadherins in the myometrium remained constant throughout the menstrual cycle. While Khan-Dawood et al. (1997) detected E-cadherin and its mRNA in both normal myometrium and leiomyoma, Tai et al. (2003) reported the absence of E-cadherin in the normal myometrium and uterine leiomyomas, a similar expression of P-cadherin in these two tissues, and significantly higher expression of N-cadherin and its mRNA in uterine leiomyomas, compared to the normal myometrium. Furthermore, these researchers (Tai et al. 2003) reported no difference in catenin expression between the normal myometrium and uterine leiomyomas. Our immunohistochemical findings corroborate the intense nuclear and moderate cytoplasmic P-cadherin expression in the smooth muscle 1 3 cells of the bovine cervix and vagina, which remained constant throughout the estrous cycle. It is known that many functions of smooth muscle cells, such as adhesion, migration, proliferation, contraction, differentiation, and apoptosis, are regulated by a broad spectrum of cell-cell and cell-matrix adhesion molecules (Frismantiene et al. 2018). The findings of this study were limited to immunohistochemistry. Therefore, we could not determine the function of P-cadherin in the smooth muscle cells in the cervix and vagina. Further studies are required to clarify this issue.
Limited data are available on the presence and localization of cadherins and beta-catenin in the blood vessels of mammalian reproductive organs. Unlike results for the endothelial cells of the human endometrium (Tabibzadeh et al. 1995), P-cadherin staining was localized to the cytoplasm and nuclei of the vascular endothelial and smooth muscle cells, and N-cadherin was found in the endothelial cells. In addition, similar to what was reported by Tabibzadeh et al. (1995), the immunohistochemical findings in the present study of the cow cervix and vagina indicate that beta-catenin was located at the junctions between the vascular endothelial cells. This finding corroborates that P-cadherin, N-cadherin, and beta-catenin are essential to endothelial cells in normal vascular patterning in the bovine cervix and vagina, as reported by previous studies (George and Beeching 2006;Clifford et al. 2008).
In conclusion, this report reveals that 1) classical E-, N-, P-cadherins, and beta-catenin are constitutively present in the cervix and vagina of cycling cows, 2) their distribution profiles are cell/tissue/organ type-specific manner, and 3) the OD values and immunolocalization patterns of these adhesion molecules in the cervical and vaginal epithelium are altered by the structural changes that occur under the influence of E2 and P4 during the estrous cycle. These outcomes suggest the possible role of E-, N-, P-cadherins, and betacatenin in maintaining the normal architecture, epithelial integrity, and barrier function in the cow cervix and vagina during the estrous cycle as reported in the other mammalian reproductive organs (for a review, see Rowlands et al. 2000;Poncelet et al. 2002;Shih et al. 2004;Blaskewicz et al. 2011;Tienthai 2018).
CRediT authorship contribution statement Narin Liman: Conceptualization, formal analysis, experimental design, knowledge transfer, analyzing and interpreting the data, and writing and revising the manuscript. Hakan Sağsöz: Experimental design, tissue collection and processing, interpretation of data, and drafting the manuscript.

Data availability
The datasets in this study are available from the corresponding author upon reasonable request.

Declarations
Ethical approval All applicable international, national, and/or institutional guidelines for the care and use of animals were followed.

Consent to participate All authors participated voluntarily in the research.
Consent for publication All authors read and approved the final manuscript.

Conflict of interest
All authors note no conflicts of interest relevant to this study.