Progesterone-induced blocking factor (PIBF) and galectins modulate the maternal immune response during pregnancy. We hypothesized that the relative transcript abundance of the above genes would be different during the luteal phase/early pregnancy and would be affected by progesterone supplementation. To further test this, hypothesis protein expression analyses were carried out to evaluate the abundance and localization of LGALS9 and PIBF. Following estrus synchronization, heifers were inseminated (n = 140) or not (n = 70). Half the heifers in each status (cyclic or potentially pregnant) were randomly assigned to receive a progesterone-releasing intravaginal device (PRID) on day 3 after estrus, which elevated progesterone concentrations from day 3.5 to 8 (P < 0.05), resulting in four treatment groups: cyclic and pregnant heifers, each with normal and high progesterone. After confirmation of pregnancy status in inseminated animals, uterine tissue was collected on days 5, 7, 13, or 16 of the luteal phase of the cycle/pregnancy. Gene and protein expression was determined using Q-RT-PCR and IHC, respectively, on 5 heifers per treatment per time point (i.e., 80 in total). Progesterone concentrations did not affect expression of any of the genes (P > 0.05). LGALS9 and LGALS3BP were expressed at low levels in both cyclic and pregnant endometria until day 13. On day 16, expression increased only in the pregnant heifers (P < 0.0001). LGALS1 and LGALS3 decreased on day 7 (P < 0.0001) and remained low until day 16. Pregnancy had no effect on the expression of LGALS1, LGALS3, and PIBF. Additionally, LGALS9 and PIBF proteins were expressed in distinct uterine cell types. These results indicate that the galectins may be involved in uterine receptivity and/or implantation in heifers.
- early pregnancy
- estrous cycle
during the estrous cycle, the uterus undergoes remodeling under the influence of estrogen and progesterone, to create an optimum environment for embryo survival and successful implantation. Numerous studies have identified a linear and quadratic relationship between progesterone concentrations during early pregnancy and probability of embryo survival/pregnancy outcome in both lactating dairy cows and nonlactating heifers (12, 35, 52). Elevated concentrations of circulating progesterone in the immediate postconception period have been associated with an advancement of conceptus elongation (8, 20, 46), an increase in interferon-tau (IFNT) production (32, 33) and higher pregnancy rates in cattle and sheep (35, 52). This advancement in conceptus elongation, due to elevated systemic progesterone concentrations, is thought to be mediated by the indirect action of progesterone on the uterine endometrium (10). Elevated progesterone, acting on the endometrium, actively downregulates the nuclear progesterone receptor (38) and alters the timing and expression of genes thought to be important for establishing uterine receptivity to implantation (15, 17). In a high progesterone environment, the advanced conceptus, producing more IFNT, increases the expression of genes involved in the maternal immune response that are required for successful pregnancy (16).
Progesterone-induced blocking factor (PIBF) and members of the galectin family of proteins assert their actions on the maternal immune system by promoting a biased T-helper (TH) type 2 cytokine production, which favors pregnancy (6, 30, 43, 44). Progesterone, PIBF, and galectin 1 (lgals1) have been reported to act synergistically in mice to enhance fetal survival (6). Previous studies on the expression of some of the members of the galectin family in cattle relate to a single time point around the initiation of implantation, e.g., day 18 (2) or day 20 (34). In addition, the expression profile of LGALS1 has been reported during the estrous cycle (36).
Galectins are members of the lectin family, characterized by a conserved carbohydrate recognition domain and high affinity for β-galactosidases (6, 30, 41, 56), and display a wide tissue distribution in several species, where they mediate cell growth, adhesion, apoptosis, and immune modulation (22, 42) by recognizing carbohydrate groups in intracellular ligands, cell signaling receptors, and glycoproteins (24, 42). LGALS1 in humans (56), mice (6), and cattle (34, 36), galectin-3 (LGALS3) in humans (56), galectin-3 binding protein (LGALS3BP) in cattle (2), and galectin-9 (LGALS9) in humans (41, 48) and cattle (2) are expressed in the uterine endometrium. An additional member of the galectin family, galectin 15 (LGALS15), was identified in the bovine, ovine, and caprine genome (31). However, it is only expressed in sheep and goat endometrium (22, 31). Galectins lack a signal peptide for release and possess features that are characteristic of cytosolic proteins; hence, they are secreted from the cell via nonclassical pathways, enabling them to mediate both intra- and extracellular processes (30). They are implicated in the mediation of distinct functions in cell growth, adhesion, chemotaxis (41), apoptosis, immune function, and inflammation (21, 45), all of which are important processes for endometrial function. Indeed, LGALS15 has been shown to stimulate migration and attachment of the conceptus trophectoderm in sheep trophoblast cells cultured in vitro (14).
Despite this available evidence, a comparative analysis of the expression profiles of galectins and PIBF during the estrous cycle and early pregnancy has not been carried out. A large-scale microarray study by our group identified that LGALS1, LGALS3, LGALS3BP, and LGALS9 were expressed in the endometrium of cattle throughout the luteal phase of the estrous cycle and early pregnancy (16). Therefore, in this study, the expression of these galectins as well as PIBF was characterized in the bovine endometrium on day 5, 7, 13, and 16 of the estrous cycle or early pregnancy. Gene expression and localization of selected proteins were analyzed by quantitative real-time PCR (Q-RT-PCR) and immunohistochemical (IHC) procedures, respectively. We hypothesized that the relative abundance of PIBF, LGALS1, LGALS3, LGALS3BP, and LGALS9 mRNA would display a differential expression pattern in the endometrium of cyclic and pregnant heifers at different stages of the luteal phase of the estrous cycle/early pregnancy and that elevated circulating progesterone concentrations would influence this expression pattern. To further test this hypothesis protein localization was performed to evaluate the abundance of LGALS9 and PIBF in the different cell types of the uterus.
MATERIALS AND METHODS
Animal model and tissue collection.
Unless otherwise stated all chemicals and reagents were sourced from Sigma (Dublin, Ireland). All experimental procedures involving animals were licensed by the Department of Health and Children, Ireland, in accordance with the Cruelty to Animals Act (Ireland 1897) and European Community Directive 86/609/EC and sanctioned by the Animal Research Ethics Committee of University College Dublin. The experimental design was as previously described (8). In summary, to ensure sufficient numbers of pregnant and cyclic animals on the selected days, the estrous cycles of 263 cross-bred beef heifers (predominantly Charolais and Limousin cross) were synchronized using an 8-day controlled internal drug-releasing device (CIDR, 1.94 g progesterone; InterAg, Hamilton, New Zealand). Three days prior to CIDR removal, all heifers received an intramuscular injection of 2 ml of a prostaglandin F2α analog (Estrumate, Shering-Plough Animal Health, Hertfordshire, UK; equivalent to 0.5 mg cloprostenol). Of the 210 heifers that displayed standing estrus within a narrow time window (day of estrus = day 0), 140 were artificially inseminated to generate a pregnant group, while the remaining heifers were left as noninseminated cyclic controls. On day 3 of the estrous cycle/early pregnancy, half of each group were randomly assigned to receive a progesterone-releasing intravaginal device (PRID 1.55 g P4; CEVA, Animal Health, Chesham, UK), to elevate circulating concentrations of progesterone (8). This resulted in four treatment groups: 1) pregnant, high progesterone; 2) pregnant, normal progesterone; 3) cyclic, high progesterone; and 4) cyclic, normal progesterone. All heifers were randomly assigned for slaughter on either days 5, 7, 13, or 16 of the estrous cycle or potential early pregnancy. These days, in pregnant heifers, correspond to the 16-cell/early morula stage, blastocyst stage, initiation of conceptus elongation, and maternal recognition of pregnancy, respectively. Within 30 min of slaughter, the reproductive tracts of all heifers were retrieved and flushed with 20 ml of PBS containing 10% fetal calf serum. In the inseminated group, only tissues from those heifers with an appropriately developed embryo/conceptus for the day of pregnancy, as seen under a stereomicroscope and classified according to the stage of development using the criteria of the International Embryo Transfer Society (51) were further processed; i.e., for day 5, the embryos were at 16-cell/early morula stage, blastocyst stage at day 7, ovoid conceptus at day 13, and elongated conceptus at day 16. For IHC analysis, a 25 mm whole uterine cross section was obtained from the horn ipsilateral to the corpus luteum from each animal and fixed for 24 h in 10% buffered formalin. The samples for IHC were then processed by dehydration through a series of ascending concentrations of alcohol, cleared in xylene, and finally impregnated with paraffin wax prior to sectioning for IHC analysis. Strips of endometrial tissue (∼2 cm long and 300 mg in weight) were removed from the midsection of the ipsilateral horn caudal to the samples collected for IHC, immersed in 1:5 wt/vol RNAlater, and transported back to the laboratory on ice. These samples were stored at 4°C for 24 h removed from the RNAlater, placed into a new tube, and stored at −80°C prior to RNA extraction for Q-RT-PCR analysis. For both gene expression and protein localization analysis, five heifers per treatment per time point were analyzed (i.e., a total of 80 heifers). To ensure the greatest diversity between groups (while still remaining within physiological ranges) heifers were selected for gene expression and protein localization analysis based on having the highest (PRID groups) or lowest (control groups) progesterone profiles.
Changes in gene expression were analyzed by Q-RT-PCR as previously described (16). Briefly, total RNA was extracted from ∼100 mg strips of endometrium using TRIzol reagent (Invitrogen, Carlsbad, CA), as per manufacturer's instructions. On column, DNase digestion and RNA cleanup was performed using a Qiagen mini kit (Qiagen, Crawley, West Sussex, UK). Both quality and quantity of the RNA were determined using the Agilent bioanalyser, and all samples used for further analysis had an RNA integrity number of >8.5 (Agilent Technologies, Santa Clara, CA). Complementary DNA was synthesized from 5 μg of total RNA using Superscript III (Invitrogen) and random hexamers as per manufacturer's instructions. Following reverse transcription, all cDNA was diluted 1:500 to generate working stocks of cDNA with a concentration of 10 ng/μl. All primers were designed using Primer Express Software (Applied Biosystems, Foster City, CA) and manufactured by Eurofins MWG (Ebersberg, Germany; Supplementary Table S1).1 All reactions were performed using 50 ng of cDNA, 10 μl of SYBR Green master mix (Applied Biosystems), and primers at a concentration of 300 nM each. Two transcript variants exist for LGALS9, due to the deletion of a coding exon, resulting in transcript variant 1 (long isoform), which codes for the entire length of 355 amino acids, and transcript variant 2 (short isoform), which lacks 32 amino acids (aa149–aa180), both of which were analyzed by Q-RT-PCR. Final reaction volumes were made up to 20 μl with RNase/DNase-free H2O. All Q-RT-PCR reactions were carried out in duplicate on the 7500 Fast Real-Time PCR System (Applied Biosystems) under the following cycling conditions: 50°C for 2 min, 95°C for 10 min, 40 cycles at 95°C for 15 s and at 60°C for 1 min. A dissociation curve was included in each Q-RT-PCR run to ensure specificity of the amplicons. Transcript identity for each of the genes of interest was verified by routine TA cloning (Invitrogen) and sequencing (MWG Eurofins; Supplementary File S1). Analysis of the most appropriate Q-RT-PCR normalization was carried out using the geNorm application in qbaseplus software (Biogazelle, Zwijnaarde, Belgium). The optimal number of reference targets in this experimental situation was determined as 3 (geNorm V < 0.15 when comparing a normalization factor based on the 3 or 4 most stable targets). As such, the optimal normalization factor was calculated as the geometric mean of reference targets ACTB, RPL19, and ERK1. All expression data for genes of interest were calibrated normalized and the expression values for each gene were determined in arbitrary units.
Based on the mRNA profiles of LGALS9 and PIBF, which increased as the day of the luteal phase of the estrous cycle/early pregnancy progressed, protein localization was carried out using IHC. The procedure for IHC was carried out as previously described (38). In summary, one paraffin wax-embedded block per animal was processed, and there were five animals per treatment per time point as outlined above. From each block (i.e., animal), 15 4-μm sections were taken. From among these 15 sections, three were randomly chosen for optimization of antibody concentrations. In addition, sections from the block used for optimization that gave the most discrete and strong staining were used as positive control slides for each IHC run. After optimization of the antibodies, one 4-μm section of embedded tissue was chosen per heifer in each of the four treatment groups above (n = 5 heifers per treatment per time point), for each of the two antibodies. Two negative control slides (where the primary antibody was omitted) and a positive control slide were also included in each assay. The sections were dewaxed in xylene and rehydrated through a series of graded alcohol steps. The sections were then blocked for endogenous peroxidase activity using 1% hydrogen peroxide solution in methanol (Sigma, Dublin, Ireland), and for nonspecific binding by using 2% normal goat serum for anti-PIBF and normal mouse serum for anti-LGALS9 (Dako Diagnostics, Cambridgeshire, UK). Primary and secondary antibodies were added to the slides and incubated at predetermined optimized temperatures for specific periods (Table 2). The bound antibody was visualized using Elite Vectastain ABC kit (Vector Labs, Peterborough, UK) and 3,3′-diaminobenzidine tetrahydrochloride chromogen substrate (Sigma), which was prepared according to manufacturer's instructions and the color allowed to develop for 10 min. The slides were then washed, dehydrated through ascending concentrations of alcohol, and cleared in xylene. Slides were mounted with DPX (AGB Scientific, Dublin, Ireland) and observed under ×10 magnification. Using a digital camera, we captured four images per tissue section: two images showing the luminal epithelium (LE), superficial glands (SG), and stroma (STR) and two images showing the deep glands (DG) and myometrium (MYO). Intensity of staining for all regions was determined using Image-Pro Plus software (version 6.2; MediaCybernetics, Bethesda, MD). To ensure lighting was consistent for each of the slides, all pictures were taken in one sitting at ×10 magnification and the same light intensity. Before saving the images of each of individual slide, we calculated the auto-exposure and used that for all subsequent image analysis. For each slide the white balance was set to “auto.” Four images were saved for each of the slides, two displaying different areas of the lumen and superficial glands and two displaying different areas of the myometrium and deeper glands. For each image, the color channel was processed to display the best representation of the original image. This gray-scale image was used to save the outlines that represented areas that had been stained i.e., protein localized. The count was measured, the background of the image flattened, and the outlines were saved. Once the outlines had been saved analysis returned to the original color image. To determine the intensity, the desired area(s) (e.g., the LE) were selected and the previously saved outlines were then loaded. The measurements to be recorded were selected (blue, green, red, mean density/intensity) and exported to an excel file. For this analysis the mean density/intensity was used and to obtain working values, this number was subtracted from 255, which is the brightest point on the intensity scale.
Differences between treatment groups for mRNA expression and protein localization for each endometrial region were determined using SAS (SAS Institute, Cary, NC). Parameters were checked for the assumptions underlying the analysis of variance i.e., examined for normality and homogeneity using histograms, qqplots, and formal statistical tests as part of the PROC UNIVARIATE procedure in SAS. Parameters that violated these assumptions were transformed using the appropriate lambda value obtained from PROC TRANSREG. Analysis was performed using PROC GLM function with day, pregnancy status, progesterone concentration, and their interactions as the main effects. Tukey's test was used to separate the treatment differences. The graphs show calibrated, normalized, relative expression values in arbitrary units and SE.
Progesterone profiles for heifers in each group used in this study have been published (8). In brief, insertion of a PRID on day 3 of either pregnancy or the estrous cycle increased concentrations of progesterone in serum from day 3.5, and values remained statistically different to day 7 of pregnancy and the estrous cycle. There was no effect on embryo development on day 5 (8- to 16-cell stage) or day 7 (morula/early blastocyst stage); however, conceptuses recovered from heifers with elevated progesterone were significantly larger on day 13 (ovoid conceptus) and day 16 (elongated filamentous conceptus) compared with those recovered from heifers with normal serum concentrations of progesterone (8).
LGALS3BP and LGALS9 transcript variants 1 (_tv1) and 2 (_tv2) displayed similar expression patterns to each other with a significant effect of day, pregnancy, and day*pregnancy interaction identified (P < 0.0001). The expression of LGALS3BP (Fig. 1A) and LGALS9 transcript variants 1 and 2 (Fig. 1B, Table 1, Supplementary Table S2) was similar in pregnant and cyclic endometria on days 5, 7, and 13 (P > 0.05). On day 16, however, the expression of these three transcripts was significantly higher in pregnant compared with cyclic endometria (P < 0.0001). There was abundant LGALS9 protein expression in the bovine uterus (Fig. 2, Supplementary Table S2). It was expressed in the LE and glandular epithelia (GE), as well as in the endothelial cells. There was limited staining in the STR, which varied widely across the treatments. Analysis of the intensity of localized protein in all cell types showed no difference between the cyclic and pregnant heifers across the days studied (P > 0.05).
Overall, a significant day effect on LGALS1 and LGALS3 expression was identified (P < 0.0001); however, there was no overall effect of pregnancy or progesterone status. The expression of LGALS1 decreased significantly from day 5 to day 7 and from day 7 to day 13 (P < 0.0001). Thereafter, levels plateaued to day 16 (Fig. 1C). Similarly, LGALS3 expression decreased on day 7 (P = 0.0002), and levels did not change on day 13 but decreased further on day 16 (P < 0.0001, Fig. 1D).
There was no effect of day, pregnancy, progesterone, or any interaction on the expression of PIBF (P > 0.05, Table 1). PIBF protein was localized to the nucleus and cytoplasm and was highly expressed in the LE, superficial GE, and to a lesser extent in the STR (Fig. 3). There was also a high expression in the MYO, deep GE, and endothelial cells. This expression across all uterine regions and in the endothelial cells was similar among all treatments and was not affected by progesterone concentration, pregnancy status, or day of the estrous cycle/early pregnancy (P > 0.05).
Pregnancy recognition is a complex event that requires synchrony between the developing conceptus and the maternal uterine environment. Furthermore, since the embryo contains both maternal and paternal factors, the maternal immune system must be tolerant to the semiallogeneic fetus for successful implantation. Galectins and PIBF have been associated with modulating the maternal immune system during pregnancy by promoting a TH2 cytokine profile (28, 54). In addition, PIBF also controls the activity of the NK cells (53). Therefore in this study, we focused on selected galectins and PIBF in the bovine uterus. The main findings are 1) LGALS3BP and LGALS9 mRNA expression is significantly higher in day 16 pregnant endometria, corresponding to the day of maternal recognition of pregnancy in cattle; 2) PIBF and LGALS9 protein are expressed in all uterine cell types and endometrial epithelia and endothelial cells, respectively, and are not affected by progesterone concentrations, pregnancy status, or day of the early luteal phase of the estrous cycle/early pregnancy; 3) Elevated concentrations of progesterone in circulation do not affect the expression of PIBF LGALS1, -3, -3BP, and -9 mRNA in the bovine endometrium throughout the early luteal phase of the estrous cycle and early pregnancy; 4) LGALS1 and -3 mRNA expression declines as the luteal phase progresses and are not affected by pregnancy status.
Of interest in this study was the result that elevated progesterone in the early luteal phase did not affect the expression of any of the genes studied. Data in the literature indicate that both beef heifers and dairy cows (with the confounding effects associated with these animal models) display a linear and quadratic relationship between concentrations of progesterone in the early luteal phase (up to day 7) and the subsequent probability of embryo survival/calving rate. In addition, data in both beef heifers (using this animal model) and sheep have shown that supplementation of exogenous progesterone concentrations significantly alters the expression of genes in the endometrium resulting in enhanced conceptus elongation (8, 47). Available evidence indicates that progesterone or IFNT alone, or working in concert, can alter the expression of genes in the endometrium; however, a number of genes that are modulated in the uterine endometrium during early pregnancy occur due to the actions of IFNT alone (16, 50). Given the process of immune modulation that genes in this study are involved in, it is unsurprising that some, e.g., LGALS3BP and -9, appear to be regulated by pregnancy, most likely conceptus derived IFNT, alone.
LGALS1 suppresses autoimmune inflammation and apoptosis by signaling through distinct cell surface glycans expressed by TH1 and TH17 cells, while the TH2 cytokines are shielded from this effect through sialylation of cell surface glycoproteins (55). It also regulates homoeostasis (43) and inhibits the activity of CD4+ and CD25+ regulatory T cells (19). Expression of LGALS1 has been reported to be higher in in vivo-derived day 7 bovine embryos compared with their less viable in vitro-derived counterparts and may contribute to lower pregnancy rates following embryo transfer as LGALS1 promotes maternal fetal tolerance (37). Extensive studies in mice showed control of Lgals1 to be under the influence of estrogen and progesterone (9). Treatment with estrogen, which acts through its nuclear receptors, increased the expression of Lgals1, while an antiprogestin reduced Lgals1 expression by antagonizing the progesterone effect. Delayed implantation also led to a decrease in Lgals1 expression (9). In cattle, LGALS1 mRNA expression in the endometrium was highest at estrus (36). A strong expression of LGALS1 was reported in the bovine superficial GE at estrus (36), while during implantation, expression in pregnant heifers was reported to be mainly stromal and concentrated in the caruncular regions (34), indicating discrepancies in the localization of the mRNA or perhaps a shift in expression due to pregnancy and a potential role for LGALS1 during implantation. The declining levels of LGALS1 in the current study and the lack of an effect of progesterone on expression give us reason to speculate that in the bovine endometrium, the levels may be under the influence of estrogen.
LGALS3 is the only galectin with both pro- and antiapoptotic properties (18, 58). After its secretion from epithelial cells, it modulates cell adhesion by binding a number of ligands such as fibronectin, laminin, and integrins (26). While LGALS1 reduces the levels of the TH1 cytokines by promoting TH2 cytokine secretion (43), LGALS3 treatment was found to increase TH1 cytokine levels (11). Mice exhibiting induced abortion due to stress showed higher levels of Lgals3 and low levels of Lgals1, which was reversed by treatment with recombinant LGALS1 and an anti-LGALS3 (5). The current study revealed similar expression patterns between LGALS1 and LGALS3 in the bovine endometrium. In the human endometrium, expression between the two galectins was opposite with low expression of LGALS3 during the proliferative phase and increased expression during the secretory phase and in the deciduas (41). These studies show that there is a strict regulation of LGALS1 and LGALS3 expression during the estrous/menstrual cycle and early pregnancy.
Mice stress models showed synergy between expression of LGALS1, PIBF, and progesterone to maintain pregnancy. Treatment of pregnant mice under stressed conditions with a progesterone derivative prevented abortion by increasing PIBF levels and decreasing TH1 cytokines (7). A similar study showed that the progesterone derivative increased LGALS1 levels, which significantly increased progesterone and PIBF concentrations in the stress-challenged pregnancies to levels that were much higher than were found in normal pregnancy (6). Furthermore, in humans, the level of PIBF present in the urine could be related to the outcome of pregnancy, with higher levels being favorable (40). In the current study PIBF transcript levels were unaffected by progesterone concentrations or pregnancy status. Our results pose questions on the significance of expression of PIBF in the bovine endometrium and its regulation since its expression is quite different to that described in humans and mice above, where it has been associated with immunomodulation during pregnancy. In addition, the expression profiles of LGALS1 and PIBF mRNA in the bovine endometrium are opposite to each other, therefore making synergism unlikely. In view of the conflicting results in expression of LGALS1, LGALS3, and PIBF between the current findings in the bovine and previous findings in mice (5–7) and humans (13, 56), it is apparent that a species difference in the expression of LGALS1, LGALS3, and PIBF may exist. This may be influenced by the types of placentation where humans and rodents have a highly invasive hemochorial placentation, while ruminants have a relatively noninvasive synepitheliochoral placentation (3, 4).
LGALS3BP and LGALS9 are associated with cell adhesion and aggregation (2, 25, 27). LGALS3BP was reported to interact with LGALS3 to form cell aggregates (27). In the current study, the expression of LGALS3BP and LGALS9 in the bovine endometrium was similar among all treatment groups until day 16, where there was a significant increase in their expression in pregnant heifers. Previous studies on day 16 (16) and on day 18 (2) bovine endometrium reported increased expression in LGALS3BP and LGALS9 in the pregnant group (2), further emphasizing that they are pregnancy specific. We postulate that in the bovine endometrium, LGALS9 expression may be stimulated by IFNT from the conceptus trophectoderm and that it may be one of the new class of “nonclassical” interferon-stimulated genes, which include LGALS15, GRP, CSTL, CST3, and IGFBP1 in sheep (4). We further propose that it may serve to stimulate proliferation of the trophectodem cells.
The localization of LGALS9 protein in the bovine uterus in the current study revealed that it was expressed in the LE and GE, with no expression in MYO and a low dispersed expression in the STR. This is similar to observations in humans (41), and the distinct expression of LGALS9 during the window of implantation has been suggested to play a potential role during embryo implantation (48). One of the characteristic processes of implantation is immunomodulation, which is characterized by an influx of immune cells at the implantation site (49). LGALS9 mediates cell adhesion and chemotaxis (42, 57) and is thought to be one of the processes by which LGALS9 regulates leukocytes within the endometrium (48). A study in TK-1 cells showed that LGALS9 contributes to a TH2 cytokine bias by binding to Tim-3, a membrane protein expressed on the surface of fully differentiated TH1 cells, which negatively regulates TH1 responses (59), hence promoting a pregnancy protective effect. In this study, LGALS9 protein was localized in the cytoplasm and cellular membranes of luminal and glandular epithelial cells, which are responsible for uterine secretions into the lumen (1, 23) and may therefore be secreted at the feto-maternal interface where, as in humans (48), they may mediate the regulation of the immune system during implantation, (49). The reason for the apparent discrepancy between transcript abundance and protein expression data is unclear but suggests caution in interpretation given that it is ultimately the protein that exerts the biological function.
The LE and superficial LE offer the primary site of superficial attachment of the fetal trophectoderm cells during placentation (3). Galectins take part in cell adhesion by binding distinct oligosaccharide structures (39) and cell adhesion molecules (29). They have also been suggested to bind integrins, thereby strengthening their otherwise weak bonds and eliciting biological activity (50), e.g., expression of LGALS15 in the ovine endometrium enables attachment of the ovine trophectoderm cells to maternal epithelia by binding and signaling through integrins (14, 31). In the current study, we showed that LGALS9 was only expressed in the epithelial cells, which is similar to the localization pattern of LGALS15 in the ovine endometrium (31). Although an integrin ligand for LGALS9 has yet to be identified (48), signaling through integrins may be one of the ways in which LGALS9 may function in cattle to mediate the process of implantation.
In conclusion, we have shown that there is differential regulation of some galectin genes in bovine endometrium, specific to the stage of luteal phase of the estrous cycle/early pregnancy, but independent of progesterone concentration. We have also described for the first time in the bovine uterus, the expression of the above galectins as the luteal phase of the estrous cycle/early pregnancy progresses and further described localization of LGALS9 protein in the bovine endometrium. We have also described expression of PIBF mRNA and protein in the bovine endometrium for the first time, and our results raise questions about its exact role in the bovine endometrium. These results have shown that the expression of LGALS1 and LGALS3 is tightly regulated, with decreasing expression as the cycle progresses, which is in contrast to what is reported in humans and mice. This study has also shown that both LGALS3BP and LGALS9 expression is increased in the endometrium of pregnant heifers at the time of pregnancy recognition, much earlier than in previous reports. The marked increase of LGALS3BP and LGALS9 in pregnant heifers only on day 16 and the exclusive expression of LGALS9 in the LE and GE suggest a possible key role in maternal immune tolerance of the conceptus during maternal recognition of pregnancy and a potential role during implantation. The tight regulation of these galectins in the bovine endometrium during pregnancy recognition indicates that they may also mediate uterine receptivity in cattle. (Table 2).
This work was funded by Science Foundation Ireland PI Grant 06/INI/B62.
No conflicts of interest, financial or otherwise, are declared by the author(s).
The authors express gratitude to all the principal investigators, postdoctoral scientists, technical staff, and all graduate students involved in sample collection and assays as well as advice on various aspects of the study.
↵1 The online version of this article contains supplemental material.
- Copyright © 2011 the American Physiological Society