Galectins: Double-edged Swords in the Cross-roads of Pregnancy Complications and Female Reproductive Tract Inflammation and Neoplasia.

Pregnancy poses a substantial challenge to the maternal immune system. The semi-allogeneic fetus, placenta and chorioamniotic membranes continuously interact with maternal immune cells in the uterus, which is an immune privileged site [ 169 ], and those in the maternal circulation [ 170 ]. During implantation and placentation, there is a continuous immune recognition and modulation of the maternal immune system by trophoblasts at the maternalfetal interface [ 171 - 174 ]. Moreover, there is a continuous deportation of fetal cells and trophoblastic debris into the maternal circulation, which leads to microchimerism and an increase in systemic inflammation in the mother during pregnancy [ 175 - 179 ]. Therefore, normal pregnancy is associated with a mild inflammatory state, especially by neutrophils of the innate immune system [ 180 , 181 ]. This is significantly pronounced in preeclampsia, where the activation state of neutrophils is higher than in sepsis [ 182 , 183 ]. Overtly activated neutrophils are also implicated in recurrent fetal loss or bacterially induced abortions [ 181 ]. It was also revealed that other great obstetrical syndromes (e.g. intrauterine growth restriction [IUGR] and preterm labor) are also associated with various changes in the phenotypes as well as the behavior of maternal peripheral blood leukocytes and systemic inflammation [ 183 - 186 ]. Since several galectins are expressed at the maternal-fetal interface, the site of contact between maternal and fetal cells that varies among different species [ 171 , 187 - 190 ], they are proposed to promote maternal-fetal immune tolerance and regulate local and systemic inflammation and infection[ 10 , 16 , 17 , 26 , 27 ]. Indeed, changes in the expression of galectins [ 23 , 32 - 38 ] have been reported in the great obstetrical syndromes (e.g. preterm labor, preeclampsia) [ 191 ], which are related to local and/or systemic inflammation and infection, and are responsible for most perinatal mortality and morbidity [ 192 - 205 ]. The human maternal-fetal interfaces dynamically change during gestation [ 190 ]. First, the syncytiotrophoblast is in direct contact with maternal cells in the decidua for a few days post-implantation and then with cells in the intervillous space. The latter is also the site of the interaction between the syncytiotrophoblast and maternal blood cells by the end of the first trimester, while invasive extravillous cytotrophoblasts in the placental bed and trophoblasts in the chorion laeve come into contact with maternal cells in the decidua [ 190 ]. In this dynamic context, the expression of several galectins is also spatio-temporally regulated during development ( Fig. 3 ) [ 8 ]. Galectin-1, -3, and -9 are broadly expressed during human and mouse embryogenesis, suggesting that they may play a role in embryo development in mammals [ 8 ]. Despite that, galectin-1 or galectin-3 knock out (KO) mice are viable [ 206 ], possibly due to the redundancy in galectin functions [ 8 ]. In addition, galectin-1, -3, -8, -9, -13, -14, and -16 are also strongly expressed at the maternal-fetal interface in various mammals, some in a developmentally regulated fashion [ 16 , 18 , 19 , 21 , 23 , 27 , 32 - 34 , 39 ]. For example, galectin-1 expression is strong in the differentiated syncytiotrophoblast but not in the cytotrophoblast during first and third trimesters [ 19 , 207 , 208 ], and its expression in the extravillous trophoblast is developmentally regulated during the first-trimester [ 19 , 209 ]. This latter phenomenon is also true for galectin-3, which also localizes to villous cytotrophoblasts [ 19 , 208 ]. Galectin-4 has weaker placental expression [ 27 ], which is down-regulated during trophoblast differentiation in rats [ 210 ]. Galectin-8 has expression in villous and extravillous trophoblasts [ 39 ], while galectin-9 is mainly located in the decidua [ 16 ]. RNA and protein evidence have shown that galectins in the chromosome 19 cluster (-13, -14, -16, and -17) are predominantly expressed by the syncytiotrophoblast but not by the underlying cytotrophoblasts [ 27 , 28 , 31 , 33 , 46 , 58 ]. This is supported by galectin-13 immunolocalization to the multinucleated luminal trophoblasts within converted decidual spiral arterioles in the first trimester [ 28 ]. A recent study demonstrated that the expression of galectin-13, -14, and -16 is related to the differentiation and syncytialization of the villous trophoblast [ 72 ], which is important in the production of placental hormones and immune proteins to control fetal development and immune tolerance [ 189 , 211 , 212 ]. In vitro assays demonstrated that the expression of these galectins is related to syncytium formation induced by cAMP [ 72 ]. Interestingly, the promoter evolution and the insertion of a primate-specific transposable element into the 5’ untranslated region of an ancestral galectin gene introduced several binding sites for transcription factors fundamental in syncytiotrophoblastic gene expression, leading to the gain of placental expression of these chromosome 19 cluster galectins [ 72 , 213 ]. Of note, DNA methylation also regulates the developmental expression of these genes [ 72 ] similar to other galectins [ 214 ]. Of interest, galectin-1, -7, -9, -13, -14, -16, and -17 are also expressed in the chorioamniotic membranes, but the developmental aspects of their regulation at this site have not yet been revealed [ 16 , 27 , 34 , 37 , 38 ]. Embryonic implantation can be considered a pro-inflammatory response in the decidua, which involves the chemotaxis of leukocytes and their active participation in the regulation of implantation via secreted immune and angiogenic factors [ 215 - 218 ]. Decidual cell-derived factors also have a key role in implantation [ 219 ]. Of importance, several galectins are expressed by the uterine endometrium and decidua in mammals and are strictly regulated by sex steroids [ 18 , 22 , 220 - 222 ]. The peak expression of these galectins coincides with the implantation time window; therefore, their possible roles in blastocyst attachment and in the regulation of immune cell functions during implantation have been implicated ( Fig. 3 ) [ 18 , 21 , 22 ]. For example, a temporal expression change of galectin-1, dependent on estrogen and progesterone, has been observed during the estrus cycle in mice [ 10 , 18 ]. In humans, the expression of galectin-1, -2, -3, -4, -8, -9, and -12 is described in the endometrium [ 21 , 22 , 98 , 113 , 223 - 225 ] where galectin-1 and galectin-3 are highly expressed during the implantation time window [ 22 , 221 ]. Galectin-3 expression is increased in glandular epithelial cells in the secretory phase, while galectin-1 expression is increased in stromal cells in the late secretory phase and further increased in the decidua [ 22 ]. Interestingly, galectin-1 is also expressed in the trophectoderm and inner cell mass of human pre-implantation stage embryos, where it may be involved in the attachment to the uterine epithelium [ 226 ]. In spite of the identification of galectin-3 in trophoblasts, its role in implantation has not been well defined. Data in humans and mice support that galectin-9 is also involved in implantation. In mouse models, galectin-9 is associated with cell-to-cell interactions and the establishment of an immuno-privileged local environment for implantation and early fetal development as well as the mediation of decidual cell migration and chemotaxis [ 223 ]. In humans, galectin-9 is expressed by the endometrial glandular epithelial cells during the implantation time window as well as by the human decidua during early pregnancy [ 21 ]. Electron microscopy clarified its localization on the apical projections of the human endometrial epithelium called uterodomes [ 223 ], which are membrane projections that exclusively feature the receptive endometrium during the implantation time window. The contribution of galectin-9 to the development of pregnancy is supported by the observation that normal pregnancy and cases of spontaneous abortions differ significantly in terms of endometrial galectin-9 splice variant profiles in both mice and humans [ 227 ]. A growing body of evidence suggests that human galectins play key roles in placentation events beyond implantation. For example, galectin-1, -3, and -8 are expressed in the extravillous trophoblast in the first trimester [ 19 , 39 ] throughout the invasive pathway of trophoblast differentiation [ 212 , 217 , 228 ]. These galectins are expressed in extravillous trophoblast cell columns, where they actively deposit extracellular matrix and can bind to major structural glycans of the placental bed (e.g. fibronectin and laminin) [ 8 , 19 , 39 , 229 ] Thus, galectin-1, -3, and -8 may play a role in the organization of the extracellular matrix and the modulation of cell adhesion in the cell columns [ 19 , 39 ]. In addition, galectin-1 and -3 may have a role in the regulation of the extravillous trophoblast cell cycle since they are absent from the differentiated, nonproliferating, interstitially migrating, highly invasive cytotrophoblasts ( Fig. 3 ) [ 19 ]. Not only the expression pattern of galectin-1 in the first trimester placenta but also the findings that blocking galectin-1 substantially abrogates migration of primary trophoblasts and HTR8/SVneo cells cultured in matrige [l19,209] suggest that galectin-1 modulates the invasive pathway of trophoblast differentiation and enhances trophoblast invasiveness. Extravillous trophoblastic galectin-3 [ 19 , 208 ] may interact between cell and extracellular matrix components, modulating adhesive interactions and immune reactions as observed in a murine model [ 230 ]. In the case of galectin-13 (PP13), a different mechanism is proposed to promote trophoblast invasion [ 28 ]. Galectin-13 is secreted by the syncytiotrophoblast to the maternal circulation, from where it is transferred into the decidua in the first trimester, coinciding with the time of early trophoblast invasion. Interestingly, galectin-13 forms crystal-like aggregates in the decidua, where it attracts, activates and kills maternal immune cells, diverting them from spiral arterioles and invading trophoblasts [ 28 ]. In this manner, PP13 may serve to establish a decoy inflammatory response, sequestering maternal immune cells away from the site of extravillous trophoblast spiral artery modification. In eutherian mammals multiple immune mechanisms exist which support the establishment and maintenance of immunological privilege in the pregnant uterus, as well as antigen-specific, local and systemic maternal-fetal tolerance [ 10 , 26 , 27 , 171 - 174 , 192 ]. These mechanisms are strongly affected by the type of placentation and the interactions between fetal trophoblasts and maternal immune cells at the maternal-fetal interfaces [ 171 , 187 , 189 ]. In this regard, it is important to note that galectins are also expressed by maternal immune cells, which infiltrate the decidua and play key roles in mammalian pregnancies ( Fig. 3 ) [ 20 , 25 , 51 , 70 , 231 ]. For example, galectin-1 is strongly expressed by uterine natural killer (uNK) cells compared to peripheral blood NK cells [ 20 ]. These CD56+galectin-1+uNK cells comprise ~70% of maternal leukocytes at the implantation site, promote angiogenesis and trophoblast invasion [ 20 , 171 ] and are pivotal for the maternal adaptation to pregnancy [ 232 ]. Galectin-1, secreted by human uNK cells, induces apoptosis of activated decidual T cells [ 25 ], which is supported by data indicating that galectin-1 can selectively induce apoptosis of Th1 and Th17 cells [ 25 , 63 ] and contribute to maternal immune-tolerance to the semi-allogeneic fetus [ 10 , 25 , 26 ]. In addition, galectin-1 is among the immunosuppressive molecules secreted by villous trophoblasts, which were identified by a proteomics study and found to inhibit T lymphocyte proliferation and adaptive immune responses [ 69 ]. The villous trophoblast secretes other galectins, expressed from the chromosome 19 galectin cluster (-13, -14, and -16), which induce the apoptosis of activated T cells, and thus, are assumed to exert special homeostatic and immunobiological functions at the maternalfetal interface [ 16 , 27 ]. As in vivo evidence for the pivotal functions of human galectin-1, a proteomics study identified it to be down-regulated in villous placenta in early pregnancy loss, reflecting abnormalities in the support for the maintenance of pregnancy [ 23 ]. Other in vivo evidence comes from a mouse model of stress-induced fetal loss in which the decidual expression of galectin-1 decreased, and these mice, similar to galectin-1 KO mice, had a higher rate of fetal loss in allogeneic pregnancies [ 10 ]. This effect was reversed by the administration of recombinant galectin-1 and also by progesterone treatment, supporting the progesterone-dependent regulation of decidual galectin-1 expression. Galectin-1 treatment also prevents the drop in progesterone and progesteroneinduced blocking factor serum concentrations in stressed animals, suggesting a synergistic effect of galectin-1 and progesterone in pregnancy maintenance [ 10 ]. It was also elucidated that galectin-1 exerts its immune modulatory effect through the induction of tolerogenic DCs, which in turn trigger the expansion of interleukin-10 expressing CD4+CD25+Treg cells in vivo [ 10 ]. Subsequently, it was determined that Treg cells, which normally expand during pregnancy and suppress the maternal allogeneic response directed against the fetus [ 187 ], also overexpress galectin-10, which has an important role in suppressive functions [ 70 , 231 ]. The galectin-9/TIM-3 (T-cell immunoglobulin domain and the mucin domain 3) pathway has been recognized as central in the regulation of Th1 immunity and tolerance induction [ 233 , 234 ]. Very recently, galectin-9 was also implicated in the regulation of uNK cell function and the maintenance of normal pregnancy [ 235 ] as galectin-9, secreted by human trophoblast cells, induces the transformation of peripheral NK cells into uNK-like cells via the interaction with TIM-3. In addition, a decreased number of TIM-3+uNK cells was detected in human miscarriages and abortion-prone murine models, and a Th2/Th1 imbalance was detected in TIM-3+uNK cells in human and mouse miscarriages, suggesting the importance of the galectin-9/TIM-3 pathway [ 235 ]. Moreover, Treg cells increase their galectin-9 expression with advancing gestational age in accord with the increasing galectin-9 concentrations in maternal blood, suggesting that galectin-9 expressing Treg cells may have important roles in the maintenance of pregnancy [ 236 ]. Aside from modulating the immune system and trophoblast invasion, human galectins have been implicated in key roles in angiogenesis ( Fig. 3 ). This is not surprising in light of the pivotal role of galectin-glycan interactions in angiogenesis [ 237 ] and the angiostimulatory roles of several galectins reviewed elsewhere [ 14 ]. The most studied galectin, with respect to placental angiogenesis, is galectin-1. When this lectin is added exogeneously in a rodent model of reduced angiogenesis, it enhances the production of pro-angiogenic factors (e.g. angiogenin, heparin-binding epidermal growth factor, and fibroblast growth factor-basic) and matrix metallopeptidases (MMP-3, MMP-8, and MMP-9) to promote normal vascular development, to rescue implantation and to support healthy placentation [ 238 ]. Galectin-1 acts via the NRP-1–VEGF–VEGF-R2 signaling pathway [ 239 , 240 ], which is important in promoting angiogenesis during implantation, decidualization and placentation [ 241 , 242 ]. Galectin-1 binding to neuropilin-1 promotes VEGF–VEGF-R2 interactions, and consequently, endothelial cell migration and adhesion [ 239 , 241 , 243 ], and these effects can be blocked by an NRP-1 neutralizing antibody, which inhibits VEGF–VEGF-R2 signaling [ 238 , 240 ]. Although several other galectins (-3, -8, and -9) have been implicated in angiogenesis and endothelial cell biology [ 14 ], their involvement in placental angiogenesis has not yet been elucidated. The effect of galectin-13 has recently been tested on rat vasculature, and it was found that recombinant galectin-13 reduces blood pressure and increases utero-placental perfusion in vivo , while it promotes vasodilation in isolated arteries in vitro [ 244 , 245 ]. Term parturition is characterized by local pro-inflammatory changes in the decidua and chorioamnion, which play fundamental roles in the initiation of labor and myometrial contractions [ 37 , 246 - 249 ]. Evidence from microarray studies have shown that galectins may also play a role in pathways leading to term labor as galectin-7 is up-regulated in the amnion in oxytocin-induced labor, and galectin-9 is down-regulated in the chorion at the site of rupture ( Fig. 4 ) [ 37 ]. Preterm parturition is a syndrome that has many etiologies, predominantly those associated with intra-amniotic infection and inflammation [ 193 , 205 , 250 ]. The pathways initiated in preterm parturition are different from those in term labor, whereas the terminal pathway of cervical effacement and dilatation, choriodecidual, as well as myometrial activation, are shared between the two [ 193 , 205 , 246 , 250 ]. Interestingly, proteomics studies show that galectin-1 is upregulated in the fetal membranes in preterm parturition [ 38 ], reflecting heightened local inflammation. Preterm premature rupture of the membranes (PPROM) is a syndrome in which approximately 32%–75% of the cases are associated with microbial invasion of the amniotic cavity [ 193 , 195 , 196 ]. To date, only galectin-1 expression has been studied in PPROM using detailed gene and protein expression profiling [ 34 ]; it is increased in the chorioamniotic membranes in patients with histologic chorioamnionitis, but not in those without this condition. Galectin-1 expression is increased [ 34 ] in a temporal and spatial fashion in amnion epithelial cells, maternal neutrophils and chorioamniotic macrophages and myofibroblasts [ 251 ] with advancing inflammation. Since galectin-1 is associated with the up-regulation of genes encoding for MMPs in DCs [ 252 ], it has been proposed that the overexpression of galectin-1 in the chorioamniotic membranes may be the link between inflammation, tissue remodeling, and membrane weakening, which may contribute to the membrane rupture [ 34 ]. Moreover, the increased expression of galectin-1 by chorioamniotic macrophages upon inflammation suggests a role for galectin-1 in the active barrier functions of the membranes, protecting the fetus from bacterial infection and promoting the recognition and phagocytic removal of invading maternal neutrophils [ 34 ]. This hypothesis is supported by findings that (1) activated macrophages are present in the fetal membranes in association with fetal inflammatory response upon infection [ 253 - 255 ], (2) the chorioamniotic membranes have antimicrobial properties [ 256 ], (3) galectin-1 expression is up-regulated in activated macrophages [ 257 ] where it regulates macrophage effector functions [ 258 ], (4) galectin-1 decreases macrophage inducible nitric oxide synthase expression and inhibits lipopolysaccharide-induced NO metabolism [ 259 ], and (5) it regulates the cell surface expression of FcγRI [ 258 ]. Due to galectins’ roles in immune responses, their relevant roles in term and preterm parturition in the neonate have also been investigated, mainly regarding galectin-1 and galectin-3 ( Fig. 4 ) [ 260 - 262 ]. In term parturition, in spite of the physiological systemic inflammation in the mother at the time of normal delivery, cord blood plasma contains more galectin-3 than maternal plasma, regardless of the delivery mode [ 262 ]. In addition, cord blood neutrophils show priming in comparison to maternal neutrophils by responding to galectin-3 with reactive oxygen species (ROS) production, suggesting that inflammatory stimuli associated with labor promotes neutrophils to develop a reactive phenotype with extensive priming features [ 262 ]. Indeed, when cord blood leukocytes are stimulated by invasive bacteria, there is an induction of galectin-3 expression, suggesting its importance for innate immunity in the neonate [ 260 ]. Although galectin-1 is also expressed in cord blood, lymphocytes expressing galectin-1 were not determined to have a major role in immune reactivity in cord blood [ 263 ]. In preterm parturition, the earlier preterm birth occurs, the higher the rate of intra-amniotic infection and inflammation [ 193 ]. Since 5%–13% of pregnancies are affected by preterm parturition [ 194 ], the resulting severe complications (i.e. intraventricular hemorrhage, cystic periventricular leukomalacia, bronchopulmonary dysplasia [BPD], and cerebral palsy) have disastrous short-term and life-long impacts on the neonate, and the healthcare and social impacts are immense [ 193 , 205 ]. In regard to these, galectin-3 concentrations are elevated in the cerebrospinal fluid of infants suffering from birth asphyxia, and even higher in those with abnormal outcomes [ 261 ]. Since galectin-3 is produced by activated microglia/macrophages and activates NADPH oxidase, leading to neurotoxic production of ROS and contributing to hypoxic brain injury in an animal model [ 264 ], it has been proposed to serve as a marker for abnormal outcomes [ 261 ]. In addition, in a small preliminary study, galectin-3 concentrations in tracheal aspirates of premature infants tended to be elevated in the first week of life in those who later developed BPD (Staretz et al ., personal communication). IUGR is one of the most heterogeneous syndromes in obstetrics; it is associated with fetal malformations and chromosomal abnormalities, as well as maternal autoimmune disorders and placental dysfunction resulting from poor implantation, making the understanding of an IUGR fetus a challenge. In addition, neonates may be small-for-gestational age (SGA) due to a normal condition in short-stature couples [ 265 ]. Of interest, a recent report showed that galectin-3 concentrations in cord blood have a positive correlation with gestational age, and SGA neonates have higher concentrations of galectin-3 than those that are appropriate for gestational age [ 260 ], which may be a sign of an inflammatory condition. Based on the above data, it is not surprising that galectins have been implicated in the development of preeclampsia, a syndrome with impaired trophoblast invasion, an anti-angiogenic state and an exaggerated maternal systemic immune response [ 190 , 266 ]. Preeclampsia is a severe complication of pregnancy, which affects 5%–7% of pregnant women and is a leading cause of maternal and perinatal morbidity and mortality [ 267 , 268 ]. It also confers a high risk to the mother and fetus for metabolic and cardiovascular diseases later in life [ 269 - 272 ]. Preeclampsia is a syndrome with a spectrum of phenotypes, which may present at various gestational ages, with different degrees of severity at clinical onset, and also with or without the involvement of the fetus [ 272 - 274 ]. It is a multi-stage disease that has placental origins [ 190 , 275 - 277 ] due to the failure of extravillous trophoblast invasion into the uterine tissues [ 278 , 279 ] and impaired villous trophoblastic syncytialization [ 72 , 280 , 281 ]. Subsequent rheological changes in uterine blood flow, metabolic changes, and ischemic stress of the villous placenta lead to the liberation of anti-angiogenic molecules, highly inflammatory placental debris, and cell-free fetal DNA that may also be pro-inflammatory and cause an exaggerated maternal systemic inflammatory response, anti-angiogenic conditions and end-organ damage [ 179 , 181 , 190 , 192 , 271 , 275 - 277 , 282 - 291 ]. Other, less severe pathologies are also implicated that result in the terminal pathway of systemic inflammation and an anti-angiogenic state [ 292 ]. Importantly, several members of the galectin family have been implicated in the development of various stages of this syndrome ( Fig. 4 ). Indirect evidence of galectin involvement is the up-regulation of galectin-1 and -3 in the extravillous trophoblasts in the placental bed during preeclampsia and HELLP syndrome [ 38 , 158 ], which is associated with the failure of extravillous trophoblast invasion [ 32 ]. It was also observed that low galectin-13 expression is associated with deficient trophoblast invasion, failure of spiral arteriole conversion, and the development of preeclampsia [ 28 ]. Galectin-13 and galectin-14 mRNA expression is decreased in the syncytiotrophoblast in preeclampsia associated with or without HELLP syndrome at the time of clinical onset, predominantly in the early-onset forms [ 28 , 33 , 72 ]. Importantly, decreased galectin-13 mRNA expression can be detected as early as the first trimester in laser captured specimens of chorionic villous trophoblasts as well as decreased galectin-13 protein and mRNA concentrations in first trimester maternal serum sampled from patients destined to develop preeclampsia [ 36 ]. This phenomenon possibly reflects abnormal villous trophoblast syncytialization starting from early pregnancy and may be one of the earliest placental indicators for the subsequent development of preeclampsia. A recent study [ 72 ] revealed that GCM1 and ESRRG, two transcription factors that regulate villous trophoblastic syncytialization and metabolic functions, are down-regulated in the placenta in preeclampsia. Functional and evolutionary evidence also implicates these two factors in regulating trophoblastic expression of chromosome 19 galectin cluster genes. This is supported by the observation of decreased GCM1-mediated trophoblast fusion in impaired galectin gene expression in preeclampsia [ 72 ]. Furthermore, the differential methylation of LGALS13 and LGALS14 is also found in the villous trophoblast in preterm preeclampsia, suggesting that potential additional disease-mechanisms may account for the trophoblastic pathology in preterm preeclampsia [ 72 ]. Galectin-1 and -8 are overexpressed in the villous trophoblast in preeclampsia and HELLP syndrome [ 32 , 35 ], where increased placental stress occurs preceding exaggerated maternal systemic inflammation [ 275 , 276 , 290 , 293 , 294 ]. It is possible that galectins may function as “alarmins” in this condition [ 12 , 35 ]. Alarmins are endogenous danger signals secreted by activated cells via non-classical pathways or released from necrotic cells, which signal tissue damage and contribute to the activation and/or resolution of immune responses [ 66 ]. Galectin-13 may also be considered a placental alarmin since it is excessively secreted or shed from the syncytiotrophoblast at the time of the clinical onset of preeclampsia and HELLP syndrome [ 33 , 64 ]. Interestingly, the syncytiotrophoblast microvillous membrane and microvesicles, which are shed from the syncytiotrophoblast, stain strongly for galectin-13, suggesting that the increased release of galectin-13–positive microvesicles from the syncytiotrophoblast may lead to elevated maternal serum galectin-13 concentrations when the clinical symptoms appear [ 33 , 46 ]. Placental and maternal blood galectin-1 expression is downregulated in patients with early-onset preeclampsia, and Lgals1 KO mice exhibit preeclampsia-like symptoms, probably due to the inhibition of pro-angiogenic effects of galectin-1 [ 238 ]. Moreover, blocking galectin-1–mediated angiogenesis with anginex, a synthetic peptide, also promotes preeclampsia-like symptoms in mice and inhibits human extravillous trophoblast functions in vitro [ 238 , 295 ]. The number of galectin-1–expressing NK cells and Treg cells is decreased in preeclampsia [ 51 , 296 , 297 ], which may reflect a failure of immune tolerance in this syndrome [ 298 ]. Recently, the involvement of galectin-9 and its TIM-3 ligand has been implicated in maternal systemic inflammation in preeclampsia [ 299 ]. In this regard, decreased TIM-3 expression by T cells, cytotoxic T cells, NK cells, and CD56dim NK cells, as well as increased frequency of galectin-9+peripheral lymphocytes, is detected in women with early-onset preeclampsia, suggesting that the impairment of the galectin-9/TIM-3 pathway can result in an enhanced systemic inflammatory response including the activation of Th1 lymphocytes in preeclampsia [ 299 ]. Due to the dysregulation of some galectins at the maternal-fetal interface and in maternal blood in various obstetrical syndromes, investigations have been expanded on their possible value as diagnostic, predictive and prognostic biomarkers of these pregnancy complications. Most data is available for galectin-13, also known as PP13, which has been widely investigated by international collaborative studies ( Fig. 4 ) [ 41 - 50 , 58 ]. The changes in the expression patterns of galectin-13 in the placenta during gestation in normal and preeclamptic pregnancies, the fact that galectin-13 is expressed only in the placenta [ 27 ], and it is not detected in non-pregnant patients (Madar-Shapiro et al ., personal communication), make this galectin a suitable and promising first trimester maternal blood biomarker for the prediction of preterm preeclampsia. In addition, genetic studies found certain single nucleotide polymorphisms, including an exonic variant(221delT) in the LGALS13 gene, which may increase the risk for preterm labor and preeclampsia [ 300 ]. Recent advancement in the field has also facilitated the study of the potential use of this galectin as a therapeutic drug for preeclampsia [ 244 , 245 ]. The utilization of other galectins as biomarkers has recently been started. In the first trimester of pregnancy, there is a lower PP13 mRNA content in maternal blood in preeclampsia compared to controls [ 301 , 302 ]; however, the predictive value of the detected maternal blood PP13 mRNA species is currently limited due to the varying and low amounts of trophoblastic mRNA reaching the maternal circulation. Much more promising results were derived from studies on maternal blood PP13 concentrations in the first trimester for the prediction of preeclampsia, which were analyzed by a recent meta-analysis [ 303 ]. The results were pooled from 19 studies on singleton pregnancies, which were included in prospective or nested case-control studies or fully prospective studies in which a total of 16,153 pregnant women were tested for PP13 between 6 and 14 weeks of gestation [ 42 - 48 , 50 , 58 , 304 - 313 ]. For all cases of preeclampsia, the mean detection rate (DR) for predicting preeclampsia was 47% (95% confidence interval [CI], 43 to 65) at a 10% false-positive rate (FPR). For preterm preeclampsia, the DR was 66% (95% CI, 48 to 78); for early-onset preeclampsia, the DR was 83% (95% CI, 25 to 100). For all cases of preeclampsia, the positive likelihood ratio (LR) [sensitivity/(1-specificity)] was 5.82, while the negative LR [(1-sensitivity)/specificity] was 0.46. For preterm preeclampsia, both of these indices were better (positive LR, 6.94; negative LR, 0.34). Of interest, the introduction of maternal ABO blood groups into the prediction model could improve the DRs for preeclampsia, which can be explained by the differential binding of PP13 onto ABO blood group antigen-containing cell surfaces and the varying bioavailability of PP13 in maternal blood depending on the ABO blood type [ 58 ]. Moreover, the performance of the first trimester PP13 test could further be improved by the inclusion of PP13 into panels of multiple biomarkers (e.g. ADAM metallopeptidase domain 12 [ADAM12], pregnancy associated plasma protein A [PAPP-A], placenta growth factor [PlGF]) [ 50 , 314 ], which is necessitated in light of the syndromic nature of preeclampsia [ 48 , 314 ]. In addition, risk predictions based on combining PP13 and uterine artery Doppler pulsatility index (PI) also showed increased prediction accuracy [ 42 , 44 , 304 , 306 , 314 - 316 ]. Moreover, the combination of PP13, Doppler PI, and maternal artery stiffness (MAP) increased the DR of preeclampsia to 93% for early-onset preeclampsia and to 86% for all cases of preeclampsia at 10% FPR [ 49 ]. This is in line with comprehensive risk algorithms based on combined multi-marker analysis of background risks, MAP, Doppler PI, and a panel of blood biomarkers that can yield much higher predictive value and accuracy than individual markers [ 306 ], especially for early-onset (<34 weeks) and preterm (<37 weeks) preeclampsia. Therefore, the introduction of a broad biomarker panel for the evaluation of preeclampsia and other obstetrical syndromes in the first trimester is suggested in order to change antenatal care as formulated by the inverted pyramid model of perinatal evaluation in pregnancy [ 317 ]. In the second trimester of pregnancy, galectin-13 does not have much diagnostic or predictive value due to the sharp increase in PP13 maternal blood concentrations in preeclampsia between the first and third trimesters compared to the moderate change in women with normal pregnancy [ 318 ]. Interestingly, galectin-1 has recently emerged as a potential preclinical biomarker for preeclampsia since a prospective study detected decreased galectin-1 maternal blood concentrations and placental expression in early-onset preeclampsia compared to normal pregnancy in mid pregnancy [ 238 ]. Of note, placental galectin-1 expression is increased in preterm and severe preeclampsia compared to normal pregnancy [ 35 , 238 ]. In the third trimester of pregnancy, galectin-13 may have diagnostic significance for the clinical development of preeclampsia according to a recent meta-analysis [ 318 ]. This included eight clinical studies that contained third trimester maternal blood PP13 data from 2750 pregnant women [ 33 , 45 , 46 , 58 , 319 , 320 ]. Maternal blood PP13 was higher in women who subsequently developed preeclampsia compared to unaffected women. The mean DR at 10% FPR for all preeclampsia cases was 59.4% (95% CI, 49.7 to 64.5), and for preterm preeclampsia was 71.7% (95% CI, 60.3 to 75.3). Interestingly, the DR appeared to be related to the severity of the cases in a given study, showing that the higher the hypertension and proteinuria, the higher the third trimester PP13 in maternal blood. A combined algorithm of PP13, MAP and proteinuria yielded a 95% DR for preterm preeclampsia and 85% for all preeclampsia at 5% FPR. The positive LR for all cases of preeclampsia was 5.94 and the negative LR was 0.45, providing an overall LR of 26.24. The positive LR for preterm preeclampsia was 7.17 and the negative LR was 0.31, providing an overall LR of 37.99. Therefore, the meta-analysis indicates that higher third trimester maternal blood PP13, among women who subsequently developed preeclampsia, reached clinical diagnostic levels [ 318 ].

Galectins are an evolutionarily ancient family of lectins that have pleiotropic functions in the regulation of key biological processes. Galectins are pivotal in immune responses, angiogenesis, cell migration and invasion, and due to these functions, they have double-edged functions in shared and unique pathways of embryonic and tumor development. Recent advances facilitate the use of galectins as biomarkers in obstetrical syndromes and in various malignancies, and their therapeutic applications are also under investigation.

More than half of all human proteins are glycosylated [ 1 ], and glycans are attached to various additional glycoconjugates (e.g. glycolipids) besides glycoproteins. Because of the abundance of glycans intra- and extracellularly and also their high complexity, glycans can store orders of magnitude larger biological information than other biomolecules (e.g. nucleic acids and proteins) [ 2 , 3 ]. Lectins are sugar-binding proteins, which are not an antibody or an enzyme, and can specifically bind glycans without catalyzing their modification [ 3 , 4 ]. The interactions of lectins with glycans are pivotal in the regulation of a wide variety of interactions of cells with other cells, the extracellular matrix or pathogens [ 2 - 4 ]. Galectins belong to a subfamily of lectins based on their unique structural and sugar-binding characteristics, since their carbohydrate-recognition domains (CRDs) contain consensus amino acid sequences and they specifically bind beta-galactoside–containing glycoconjugates [ 5 - 8 ]. Galectins are the most widely expressed animal lectins; they have been found in species ranging from sponges to humans [ 7 - 9 ]. They regulate a wide variety of key biological processes, such as cell growth, proliferation and differentiation, apoptosis, signal transduction, pre-mRNA splicing, as well as cell-cell and cell-extracellular matrix interactions [ 2 , 5 - 9 ]. Galectins are also pivotal in immune responses since they regulate host-pathogen interactions, acute and chronic inflammation, and immune tolerance ( Fig. 1 ) [ 8 , 10 - 13 ]. Moreover, some galectins are central to the regulation of angiogenesis in the placenta and in tumors [ 14 , 15 ]. Interestingly, galectins can have opposing functions, and the same galectin can also have varying or contrasting effects based on the biological context and the microenvironment since their functions depend on the differentiation or activation status of the cell, the dynamic changes of their glycan partners on the cell surfaces, the redox and oligomerization status of the galectin, or its intra- or extracellular localization [ 8 , 11 , 16 , 17 ]. Thus, galectins’ double-edged action may sometimes be beneficial or harmful to the organism. The fundamental functions of galectins indicate that they are strongly associated with reproductive functions as well as the establishment and maintenance of pregnancy [ 10 , 18 - 28 ]. Indeed, some galectins are highly expressed at the maternal-fetal interface [ 10 , 18 - 39 ], and these are evolutionarily linked to placental evolution in eutherian mammals [ 5 , 7 , 9 , 26 , 27 , 40 ]. Moreover, the dysregulated expression of these galectins in pregnancy complications has been increasingly documented [ 10 , 23 , 32 - 36 , 38 , 41 - 51 ]. Galectins have also been implicated in inflammatory, infectious and malignant diseases of the reproductive tracts. Of importance, the same galectins may be functional in pathways commonly shared by physiological and pathological, placental, and tumor developmental processes (e.g. cell invasion, angiogenesis, and immune tolerance). This review aims to give a general overview of galectins and also a focused review of them in the context of inflammation, infection and tumors in the female reproductive tract as well as in normal and complicated pregnancies. Galectins were originally termed “S-type lectins,” where “S” refers to their free cysteine residues [ 6 , 8 ]. Galectins or galectin-like proteins were also discovered in fungi, viruses, and even plants [ 8 , 9 ]. Because of the diversity between mammalian and non-mammalian galectins, their nomenclature has diverged as mammalian galectins have been named using sequential numbering, while non-mammalian galectins have retained specific names ( Table 1 ) [ 6 ]. Nineteen galectins have been identified in mammals to date, 13 of which were found in humans [ 9 , 27 ]. These galectins can be divided into three structural groups [ 5 - 8 ]: (1) “proto-type” galectins (-1, -2, -5, -7, -10, -13, -14, -15, -16, -17, -19, -20) contain a single CRD of ~130 amino acids, which homodimerize [ 5 - 8 , 52 ]; (2) “tandem-repeat-type” galectins (-4, -6, -8, -9, -12) contain two homologous CRDs connected by a short linker sequence. These may differ in their sugar-binding affinities and enable multivalent binding activity [ 5 - 7 , 52 ]; and (3) “chimera-type” galectin-3, which contains a C-terminal CRD and an N-terminal non-lectin domain important for multimerization and cross-linking as well as functional regulation [ 5 - 7 , 52 ]. Although the amino acid sequences of galectins have diverged during evolution, the topologies of their CRDs are very similar, often described as “jelly-roll;” these are β-sandwiches consisting of five- and six-stranded anti-parallel β-sheets ( Fig. 1 ) [ 6 - 8 , 53 - 55 ]. Highly conserved in galectin CRDs are eight residues, which are involved in glycan-binding by hydrogen-bonds as well as electrostatic and van der Waals interactions [ 53 , 55 ]. All galectins specifically bind beta-galactosides [ 5 - 8 ] except galectin-10, which has more affinity to beta-mannosides [ 53 ]. Of interest, some galectins have high affinity for poly-N-acetyllactosamine or ABO blood-group containing glycans, and the latter is responsible for their hemagglutinin activity [ 52 , 56 - 58 ]. Galectins have multiple functions both inside and outside the cell ( Table 1 ) [ 8 , 59 , 60 ]. Intracellularly, certain galectins can modulate cell growth, differentiation, apoptosis, and migration [ 8 , 59 , 60 ] via protein-protein interactions [ 8 , 59 , 60 ]. Some galectins (-1 and -3) shuttle into the nucleus where they function in pre-mRNA splicing [ 8 , 59 ]. In spite of the fact that they do not have a secretory signal sequence, galectins can be secreted from cells via a non-classical pathway, avoiding the endoplasmic reticulum and Golgi apparatus, which is characteristic of only a small set of proteins (e.g. high-mobility group box 1 protein, interleukin-1β) [ 61 ]. Extracellularly, galectins predominantly localize to lipid rafts on cell surfaces [ 8 , 52 , 62 ] where they exert their functions through binding to cell-surface or extracellular matrix molecules, which carry their glycan ligands [ 2 , 7 , 8 , 11 , 13 , 52 , 63 , 64 ]. They can form multivalent galectinglycan arrays, so-called lattices, by cross-linking their ligands on cell surfaces, and these lattices can organize lipid raft domains and modulate cell signaling for cell growth, metabolic functions, cytokine secretion, and survival, as well as many other intracellular and extracellular interactions [ 8 , 11 , 13 , 17 , 52 , 65 ]. Some galectins can also affect cell adhesion and apoptosis, and activate or inhibit immune responses [ 8 , 11 , 13 , 63 ]. An interesting trait of galectins is that their secretion is heightened upon response to stress conditions (e.g. inflammation and infection) and cellular damage (e.g. necrosis); therefore, galectins have been implicated as “alarmins” which signal tissue damage and elicit effector responses from immune cells, thereby promoting the activation and/or resolution of immune responses [ 12 , 35 , 64 , 66 ]. Accumulating evidence in various species shows that galectins have distinct but overlapping tissue expression patterns in mammals including humans ( Table 1 ) [ 6 , 8 , 27 , 67 , 68 ]. Among prototype galectins, galectin-1 and galectin-3 have a wide expression pattern in humans, galectin-1 being the most abundant in the endometrium/decidua [ 8 , 26 , 67 ]. Among tandem-repeat-type galectins, galectin-8 and galectin-9 have a broad and complex expression pattern. Alternatively spliced isoforms of galectin-8 are differentially expressed in various tissues [ 8 , 39 , 67 ] similar to galectin-9, which is encoded by three genes [ 67 ]. These galectins are highly expressed in the female reproductive tract and at the maternal-fetal interface [ 8 , 10 , 18 - 23 , 25 , 27 , 32 , 34 , 35 , 37 - 39 , 69 ]. Some galectins (-2, -4, -5, -6, -7, -12) have more restricted tissue distribution [ 8 , 67 ]. Of note, the expression of galectins in the chromosome 19 cluster is very restricted. Among these, galectin-10 is expressed in T regulatory (Treg) cells, as well as eosinophil and basophil lineages, and forms the so-called Charcot-Leyden crystals at sites of eosinophil-associated inflammation [ 53 , 70 ]. The expression of galectin-13, -14 and -16 is predominant in the placenta, while galectin-17 expression is low in any tissues [ 27 , 29 , 67 , 71 ]. Interestingly, these galectins (-10, -13, -14, -16, -17), which are expressed from the chromosome 19 cluster, emerged via birth-and-death evolution in anthropoid primates and may regulate unique aspects of pregnancies, including maternal-fetal immune regulation and tolerance in these species [ 16 , 27 , 72 ].