1
2
3 4 Junko Nio-Kobayashi a, b , Hazirah B. Z. Abidin b , Jeremy K. Brown b , Toshihiko Iwanaga a , 5 Andrew W. Horne b , W. Colin Duncan b 6 7 a Laboratory of Histology and Cytology, Hokkaido University Graduate School of 8 Medicine, Sapporo, JAPAN, b MRC Centre for Reproductive Health, The Queen’s 9 Medical Research Institute, The University of Edinburgh, Edinburgh, UK 10 11
12
13 Laboratory of Histology and Cytology, Hokkaido University Graduate School of 14 Medicine, Kita 15-Nishi 7, Kita-ku, Sapporo 060-8638, JAPAN. Tel. & Fax: +81-11 15 7067151, E-mail: niojun@med.hokudai.ac.jp 16 17
18
1

19 Galectin-1 and galectin-3 are abundantly expressed at implantation sites in the uterus, 20 suggesting their involvement in the establishment of pregnancy. In this study, we 21 examined the expression and localization of galectin-1 and galectin-3 in the Fallopian 22 tubes from non-pregnant women, and in those presenting with tubal ectopic pregnancy. 23 There was no significant difference in the expression of both galectin-1 (LGALS1) and 24 galectin-3 (LGALS3) transcripts in Fallopian tube across the menstrual cycle. Their 25 expression in the Fallopian tube were inversely correlated to each other (r=−0.5134, 26 P<0.0001) and differentially localized. Galectin-1 protein was abundant in the stroma of 27 non-pregnant Fallopian tubes whereas galectin-3 was mainly localized to the epithelium, 28 notably to the cilia of ciliated cells and the apical cytoplasm of secretory cells. In ectopic 29 pregnancies, LGALS3 expression was significantly reduced (P<0.0001) but LGALS1 30 expression did not alter when compared to non-pregnant Fallopian tube collected during 31 the mid-secretory phase. The percentage of Fallopian tube epithelial cells expressing 32 galectin-3 in cilia tended to be reduced (P=0.0685) with an accompanying loss of normal 33 ciliary structure, while nuclear galectin-3 increased (P<0.05) in ectopic pregnancies. 34 Epithelial immunostaining for galectin-1 tended to be elevated in the Fallopian tube from 35 women with ectopic pregnancy. Co-culture of human trophoblast-origin SW71 cells 36 significantly increased LGALS1 expression in human Fallopian tube epithelial OE-E6/E7 2

37 cells, suggesting that trophoblast-derived products regulate LGALS1 expression in the 38 oviductal epithelium. These findings imply a differential contribution of galectin-1 and 39 galectin-3 in the homeostasis of human Fallopian tube and in pathophysiology of ectopic 40 pregnancy. 41 42 43 44 45 46 47 48 49 50 51 52 53 54
3

55 Successful embryonic implantation at the correct site is crucial for human reproduction. 56 Ectopic pregnancy is defined as a pregnancy where a blastocyst implants outside the 57 normal uterine cavity, and over 95% of ectopic pregnancies are located in the Fallopian 58 tube [Walker, 2007; Varma and Gupta, 2012]. Tubal ectopic pregnancy is a life-threating 59 condition and is one of the leading causes of maternal death in the first trimester in both 60 developed and developing countries [Farquhar, 2005; Varma and Gupta, 2012]. The 61 etiology of ectopic pregnancy is still unknown, and early diagnosis and treatment prior to 62 the rupture of the Fallopian tube are essential to reduce associated morbidity and mortality. 63 Major risk factors for the tubal ectopic pregnancy include tubal damage as a result of 64 surgery or infection, cigarette smoking, and in vitro fertilization. These lead to impaired 65 tubal transport that delays passage of the embryo along the Fallopian tube and/or to 66 alterations in the tubal environment that results in early implantation [Shaw et al., 2010b]. 67 Galectin is a  -galactoside-binding animal lectin which has high affinity for N 68 acetyllactosamine residue (Gal  1-3/4GlcNAc) of glycoconjugates [Barondes et al., 1994; 69 Hirabayashi et al., 2002]. In mammals, sixteen members of galectin have been identified 70 so far and eleven galectin subtypes exist in human tissues (galectin-1 to 4, 7 to 10, 12, 13, 71 and 16). Galectins are differentially distributed throughout the mammalian body, and 72 involved in various biological functions including cell differentiation, migration, and 4

73 apoptosis, or in pathological events such as inflammation and cancer metastasis [Yang et 74 al., 2008]. Galectin-1 and galectin-3 are the major subtypes expressed in the female 75 reproductive tracts. They are abundantly expressed in the uterus and at the utero-placental 76 interface, and have been implicated in the process of implantation [Powell, 1980; 77 Hirabayashi et al., 1984; Poirier et al., 1992; von Wolff et al., 2005)]. As there is 78 convincing evidence that galectin-1 and galectin-3 contribute to the establishment of 79 successful pregnancy [Blois et al., 2007; Than et al., 2008; Blidner and Rabinovich, 2013; 80 Barrientos et al., 2014], we hypothesized that galectins expressed in the Fallopian tube 81 may have a role in the pathophysiology of ectopic pregnancy. 82 Although the expression and localization of galectins in the uterus and developing 83 placenta have been examined in various animals [Phillips et al., 1996; Choe et al., 1997; 84 Kim et al., 2008; Yang et al., 2012b; Orazizadeh et al., 2013], only three studies have 85 dealt with the expression of galectin-3 in the oviduct/Fallopian tube. Galectin-3 has been 86 localized to the epithelium in the Fallopian tube of women [John et al., 2002; Roldán and 87 Mrini, 2014], and in the oviduct of cows [Kim et al., 2008] and pigs [Roldán and Mrini, 88 2014]. However, there is no information about the detailed expression and localization of 89 galectins, particularly galectin-1, in human Fallopian tube across the menstrual cycle, and 90 after tubal implantation in ectopic pregnancy. In this study, we investigated the expression 5

91 and cellular localization of galectin-1 and galectin-3 in the Fallopian tube of non-pregnant 92 women throughout the menstrual cycle and in the Fallopian tube from women with 93 ectopic pregnancy. 94 95 96 97 98 99 100 101 102 103 104 105 106
107
Human Fallopian tube and serum collection
108 Ethical approval for this study was obtained from the Lothian Research Ethics Committee, 6

109 and informed written consent was obtained from all of women participating in this study 110 (LREC 10/S1102/40). Serum samples (10 mL) and the biopsies (2‒3 cm) at ampullary 111 region of the Fallopian tube were collected from study participants at the time of 112 hysterectomy for benign gynecological conditions or during surgical management of 113 tubal ectopic pregnancy. Women were between 18 and 45 years of age. The stage of the 114 menstrual cycle of each patient at the time of hysterectomy was determined by histologic 115 examination and staging of an endometrial biopsy taken with the Fallopian tube, and 116 measurement of serum estradiol and progesterone levels as described previously [Duncan 117 et al., 2011]. Biopsies of non-pregnant Fallopian tubes (n=32) and those from women 118 with tubal ectopic pregnancy (n=25) free from trophoblast contamination [Duncan et al., 119 2011] were divided into equivalent portions and either immersed in RNAlater (Ambion, 120 TX, USA) at 4°C overnight, and then flash frozen and stored at −80°C for subsequent 121 RNA extraction, or fixed in 10% neutral-buffered formalin overnight at 4°C followed by 122 storage in 70% ethanol, and subsequent embedding in paraffin wax for 123 immunohistochemical staining. Information about the samples used in this study is 124 summarized in Table 1 as RNA or paraffin embedded tissue were not available in some 125 patients. Serum concentrations of female sex steroids and human chorionic 126 gonadotrophin (hCG) were available in only 7 and 15 samples respectively out of the 28 7

127 samples of the Fallopian tube from women with ectopic pregnancy. 128 129
Cell culture
130 Human Fallopian tube epithelial cells (OE-E6/E7) [Lee et al., 2001] and human first 131 trimester trophoblast cells (SW71) [Straszewski-Chavez et al., 2009] were maintained in 132 either DMEM or RPMI 1640 medium containing 10% fetal bovine serum, 2 mM L 133 glutamine, 10 unit/mL penicillin, and 0.1 mg/mL streptomycin in 5% CO 2 at 37°C. OE 134 E6/E7 cells in DMEM medium were seeded at 1x10 5 cells per well in 12-well dishes. A 135 Transwell ® (Corning, Corning, NY, USA) was placed onto each well, and 1x10 5 SW71 136 cells in RPMI 1640 medium were seeded into the insert. Control cells were cultured with 137 PRMI 1640 medium without cells in the insert. They were cultured in 5% CO 2 at 37°C 138 for three days, and OE-E6/E7 cells were collected for RNA extraction. 139 140
Quantitative RT-PCR (qRT-PCR)
141 Fallopian tube tissues used for a quantitative gene expression analysis were classified as 142 proliferative phase (n=8), mid-secretory phase (n=15), late-secretory to menstrual phase 143 (n=8), or ectopic pregnancy (n=25). Total RNA was extracted from frozen human 144 Fallopian tube or cultured OE-E6/E7 cells using RNeasy Mini Kit (Qiagen Ltd., Crawley, 8

145 UK) according to the manufacturer’s protocol. RNA (200 ng) was used to prepare cDNA 146 by TaqMan Reverse Transcription regents (Applied Biosystems, Foster City, CA, USA). 147 The sequences of the primer sets used for this study are described previously [Nio 148 Kobayashi et al., 2014]. Primers were pre-validated by standard PCR and by generating 149 standard curves using qRT-PCR. Each reaction buffer contained 5.0 µL 2×PowerSYBR ® 150 Green PCR Master Mix (Applied Biosystems), 0.5 µL primer pair (5 µM), 3.5 µL of 151 nuclease free H 2 O, and 1.0 µL cDNA, and each reaction was conducted in duplicate. The 152 qRT-PCR cycling program consisted of a denaturing step (95°C for 10 min), annealing 153 and extension step (95°C for 15 sec and 60°C for 1 min repeated for 40 cycles), and a 154 dissociation step (95°C, 60°C, and 95°C for 15 sec each) using a 7900 Sequence Detection 155 System (Applied Biosystems). The relative expression levels of each target to the 156 housekeeping gene (glucose-6-phosphate dehydrogenase: G6PDH), previously validated 157 using geNorm analysis (Primerdesign Ltd, Southampton, UK), were quantified using the 158 Δ Ct or ΔΔ Ct methods. After testing for normality, all statistical analyses were performed 159 using unpaired t-tests or one-way ANOVA, with pairwise comparison, using GraphPad 160 Prism 6 software (GraphPad Software Inc., San Diego, CA, USA), and P<0.05 was 161 regarded as significant. 162 9

163
Immunohistochemistry
164 Fixed human Fallopian tube tissues collected during the proliferative phase (n=5), the 165 mid-secretory phase (n=9), late-secretory to menstrual phase (n=4), and from women with 166 ectopic pregnancy (n=26) were available for immunohistochemical analysis. The sections, 167 at 5  m thickness, were de-waxed and washed in phosphate-buffered saline (PBS). 168 Subsequently the sections were incubated with 3% hydrogen peroxide for 20 min and 169 Avidin/Biotin blocking solution (Vector Laboratories Inc., Burlingame, CA) for 15 min 170 for each reagent. Then the sections were incubated with normal rabbit or goat serum for 171 60 min at room temperature. They were incubated with goat anti-human galectin-1 172 antibody (1:1000; AF1152, R&D systems Inc., Minneapolis, MN) or rabbit anti-human 173 galectin-3 antibody (1:200; sc-20157, Santa Cruz Biotechnology Inc., Dallas, TX) in 174 rabbit or goat serum at 4°C overnight. Control sections were incubated with non-immune 175 serum. After washing twice in PBS, the sections were incubated with biotinylated anti 176 goat or anti-rabbit IgG (1:500; Vector laboratories Inc.) for 60 min at room temperature. 177 The reaction sites were visualized using Vectastain ABC Elite kit (Vector Laboratories 178 Inc.) for 60 min followed by ImmPACT TM DAB Peroxidase Substrate Kit (Vector 179 Laboratories Inc.) for 5 min. The sections were counterstained with haematoxylin and 180 observed under a light microscope (BX51; Olympus corporation, Tokyo, Japan). 10

181 To quantify the number of cells with positive galectin-3 immunoreaction in either 182 the cilia, nucleus, or cytoplasm of the Fallopian tube epithelium collected in the mid 183 secretory phase or during surgery for ectopic pregnancy, at least five images were taken 184 from the different parts of the representative sections (n=8 for the mid-secretory phase 185 and n=10 for the ectopic pregnancy) using stratified random sampling, and the ratio of 186 positive cells in total epithelial cells was calculated by observers blinded to sample source. 187 188
Dual immunohistochemistry
189 Some sections after the reaction with anti-galectin-3 antibody overnight were washed 190 with PBS and subsequently incubated with AlexaFluor 594-labeled anti-rabbit IgG 191 (1:200; Life technologies Japan, Tokyo, Japan) for 2 hours at room temperature. The 192 sections were blocked with 10% normal goat serum for 60 min at room temperature and 193 then incubated with mouse anti  -tubulin antibody (1:2,000; T6793) at 4°C overnight. 194 Sections were washed with PBS and incubated with AlexaFluor 488-labeled anti-mouse 195 IgG (1:200; Life technologies Japan) for 2 hours at room temperature, and observed under 196 a confocal laser scanning microscope (FV300; Olympus, Tokyo, Japan). 197 198 11

199 200 201 202 203 204 205 206 207 208 209 210
211
Expression and localization of galectin-1 and galectin-3 in the Fallopian tube of non-
212
pregnant women across the menstrual cycle
213 We examined the mRNA expression of galectin-1 (LGALS1) and galectin-3 (LGALS3) by 214 qRT-PCR in the Fallopian tube of non-pregnant women during the proliferative, mid 215 secretory, and late-secretory to menstrual phases. The expression of both LGALS1 and 216 LGALS3 transcripts did not significantly alter between these phases (Table 1). 12

217 To identify cells expressing galectin-1 and galectin-3 in the Fallopian tube of non 218 pregnant women, we carried out immunohistochemistry using specific antibodies against 219 human galectin-1 and galectin-3. There were no clear differences in the staining pattern 220 of galectins between the phases. Galectin-1 was predominantly localized to the stromal 221 cells, largely fibroblasts and extracellular matrix of the lamina propria, while the 222 epithelium showed limited immunoreactivity for galectin-1 (fig. 1A, B). At higher 223 magnification, epithelial cells were slightly immunoreactive for galectin-1. However, 224 numerous round cells within the epithelium were significantly immunostained for 225 galectin-1 (arrowheads in fig. 1B, C). 226 In contrast, an intense immunoreactivity for galectin-3 was found in the apical region 227 of epithelium while the stroma was weakly immunoreactive for galectin-3 (fig. 1D). 228 Galectin-3 in the epithelium was mainly localized to the cilia of ciliated cells (fig. 1E, 1F) 229 and the apical cytoplasm of secretory cells (arrows in fig. 1E). Occasionally, the galectin 230 3 immunoreactivity was found in the nucleus or whole cytoplasm of both types of 231 epithelial cells in the human Fallopian tube (asterisk in fig. 1F). Dual immunostaining for 232 galectin-3 and  -tubulin, a maker for cilia, confirmed the localization of galectin-3 both 233 in the  -tubulin-positive cilia of ciliated cells (fig. 1G-I) and in the apical cytoplasm of 234  -tubulin-negative non-ciliated secretory cells (arrows in fig. 1G-I). 13

235 236
Changes in mRNA expression and localization of galectins in the Fallopian tube from
237
women with ectopic pregnancy
238 In the Fallopian tube collected after ectopic implantation, the mRNA expression of 239 LGALS3 was significantly decreased in the Fallopian tube from women with ectopic 240 pregnancy when compared to the Fallopian tube of non-pregnant women during the mid 241 secretory phase (P<0.0001; Table 1). In contrast, the LGALS1 expression did not alter 242 between two groups (Table 1). 243 Immunohistochemical analysis of the Fallopian tubes from women with ectopic 244 implantation revealed that galectin-1 immunoreactivity was abundant in the stroma, as 245 was seen in the Fallopian tube from non-pregnant women (fig. 2A). However, the 246 epithelium consistently showed elevated immunoreactivity for galectin-1 (fig. 2B) with 247 intensified immunolabeling of the round cells within the epithelium (arrowheads in fig. 248 2B). Similar to non-pregnant women, galectin-3 immunoreactivity was mainly localized 249 to the epithelium (fig. 2C). At higher magnification, epithelial galectin-3 immunoreaction 250 was found in the cilia of ciliated cells (fig. 2D) and the apical region of non-ciliated 251 secretory cells (arrows in fig. 2D) like the Fallopian tube from non-pregnant women. The 252 nuclear immunostaining for galectin-3 was more frequently observed in the Fallopian 14

253 tube from women with ectopic pregnancy (asterisks in fig. 2D). The immunoreactivity 254 for galectin-3 in  -tubulin-negative secretory cells tended to gather at the plasma 255 membrane rather than the apical cytoplasm (arrows in fig. 2D, E). 256 Cytoplasmic staining for galectin-3 was abundant in both types of cells in the 257 Fallopian tube epithelium with ectopic pregnancy, and ciliated cells containing abundant 258 galectin-3 in the cytoplasm appeared to display scattered or irregular ciliary structure 259 (arrows in fig. 2F, G). Dual immunostaining for galectin-3 and  -tubulin clearly 260 demonstrated that cells with abundant immunoreactivity for galectin-3 in the cytoplasm 261 had almost lost the  -tubulin-positive ciliary structure (arrowheads in fig. 2H-J) while 262 cells devoid of the cytoplasmic galectin-3 displayed intact  -tubulin-positive ciliary 263 structure (arrows in fig. 2H-J). The percentage of cells with galectin-3-positive intact cilia 264 tended to decrease in the Fallopian tube from women with ectopic pregnancy (fig. 2K, 265 P=0.0685). On the other hand, the number of epithelial cells with nuclear galectin-3 266 immunostaining was increased in the Fallopian tube from women with ectopic pregnancy 267 (fig. 2L, P<0.01). The number of epithelial cells with cytoplasmic galectin-3 268 immunoreactivity in the Fallopian tube did not differ between non-pregnancy and ectopic 269 pregnancy (fig. 2M). 270 These findings suggest that LGALS3 mRNA expression was significantly decreased 15

271 and alterations in subcellular localization of galectin-3 are remarkable in the epithelial 272 cells in the Fallopian tube from women with ectopic pregnancy. 273 274 Galectin-1 and galectin-3 are inversely expressed in the Fallopian tube 275 To evaluate the influence of the serum sex steroids on the expression of galectin mRNAs 276 in the Fallopian tube, we performed the correlation analysis. Although there was no 277 statistical significance, the expression of LGALS1 and LGALS3 seemed to be 278 differentially correlated to the concentration of serum sex steroids: LGALS1 expression 279 seemed to be positively whereas LGALS3 to be negatively correlated to the concentration 280 of both estradiol and progesterone (fig. 3A-D). Interestingly, the mRNA expression of 281 LGALS1 and LGALS3 were inversely correlated to each other (r=−0.5134, P<0.0001; fig. 282 3E). 283 284
Changes in galectin mRNA expression in human Fallopian tube epithelial OE-E6/E7 cells
285
induced by co-culture with trophoblast-origin SW71 cells
286 To examine whether the presence of the implanting embryo affects the mRNA expression 287 of galectins in human Fallopian tube epithelium, we used two types of cells derived from 288 human Fallopian tube epithelium (OE-E6/E7) and human trophoblasts (SW71). OE 16

289 E6/E7 cells were co-cultured with SW71 cells using transwell system for 3 days, and the 290 mRNA expression of LGALS1 and LGALS3 in OE-E6/E7 cells were analyzed by qRT 291 PCR. LGALS1 expression was increased in OE-E6/E7 cells when co-cultured with SW71 292 cells (fig. 5A; P<0.0001) while the mRNA abundance of LGALS3 did not alter (fig. 5B). 293 This finding suggests that certain molecules secreted from SW71 cells may stimulate the 294 mRNA expression of LGALS1 in epithelial OE-E6/E7 cells. 295 296 297 298
299 In this study, we demonstrated the subtype-specific expression and localization of 300 galectin-1 and galectin-3 in the Fallopian tube of women. Galectin-1 was mainly localized 301 to the stroma but galectin-3 was expressed apical sites of the epithelium in the Fallopian 302 tube of non-pregnant women. Although the expression of both galectins did not alter 303 during menstrual cycle, the expression of LGALS1 and LGALS3 was inversely correlated 304 each other. In the ectopic pregnancies, the expression of LGALS3 was dramatically 305 decreased and subcellular localization of galectin-3 changed in association of ciliary loss 306 in the Fallopian tube epithelium. On the other hand, the epithelial immunostaining of 17

307 galectin-1 slightly increased in the ectopic pregnancies and co-culture with trophoblast 308 origin SW71 cells significantly enhanced the expression of LGALS1 in the Fallopian tube 309 epithelium-derived OE-E6/E7 cells. Although the detailed function of galectins is still 310 unclear, these results suggest that galectin-1 and galectin-3 may differentially contributed 311 to the pathophysiology of tubal ectopic pregnancy. 312 We have for the first time described here the expression and localization of galectin 313 1 in human Fallopian tube. Galectin-1 was predominantly localized to cells in the stroma 314 while minimal immunoreactivity was detected in the epithelium. Interestingly, 315 intraepithelial round cells were immunostained for galectin-1. We believe that these cells 316 are leukocytes as we and other research group have previously identified leukocytes with 317 similar morphology in human Fallopian tube epithelium by immunohistochemistry 318 [Ulziibat et al., 2006; Shaw et al., 2011]. According to the previous studies, most of 319 intraepithelial leukocytes are identified as CD8-positive suppressor T lymphocytes 320 [Ulziibat et al., 2006; Shaw et al., 2011]. This suggests a possible contribution of galectin 321 1 in mucosal immunity in human Fallopian tube. Galectin-1 is well-known to regulate 322 inflammation [Liu, 2000; Cedeno-Laurent and Dimitroff, 2012] and to be a pivotal 323 regulator of the fetomaternal immune tolerance during pregnancy [Blois et al., 2007]. 324 Previous studies have reported that galectin-3 was localized to the apical surface of 18

325 non-ciliated cells in the oviduct/Fallopian tube of pigs and women [John et al., 2002; 326 Roldán and Mrini, 2014], in agreement with our data showing the apical staining of 327 galectin-3 in non-ciliated secretory cells of the human Fallopian tube. We further 328 demonstrated the unique localization of galectin-3 in the cilia of ciliated cells of human 329 Fallopian tube epithelium. Thus it is reasonable to consider differential localization of 330 galectin-1 and galectin-3 is established in human Fallopian tube and is related to 331 pathophysiology of this organ including ectopic implantation. 332 Although the mRNA expression of galectins in the Fallopian tube did not alter during 333 menstrual phase, there seems to be weak correlation between the expression of galectins 334 in the Fallopian tube and serum concentration of estradiol and progesterone. As shown in 335 fig. 3, the expression of LGALS1 tended to be positively correlated to the concentration 336 of serum steroids, especially estradiol. On the other hand, the expression of LGALS3 337 seemed to be negatively correlated to that. Regulation of the expression of galectin-1 and 338 galectin-3 by estradiol and progesterone has been reported in the murine uterus [Choe et 339 al., 1997; Hirota et al., 2012], human endometrial epithelial cells [Yang et al., 2012a], and 340 human trophoblast cell lines [Yang et al., 2011; Ramhorst et al., 2012]. Although Than et 341 al. [2008] noted the existence of estrogen responsive element in the 5’ promoter of 342 LGALS1, it remains unclear whether the expression of galectin-3 in the female 19

343 reproductive organ is directly regulated by sex steroid hormones. 344 We examined, for the first time, the expression and localization of galectins in the 345 Fallopian tube from women with ectopic pregnancy. Unfortunately it is not possible to 346 obtain human Fallopian tubes for study during normal early pregnancy, and the effect of 347 prolonged exposure to pregnancy hormones cannot be assessed. Thus, we had to compare 348 the expression to the Fallopian tube during mid-secretory stage when progesterone levels 349 are highest, mimicking the early stages of pregnancy. There was no significant difference 350 in the expression of LGALS1 in the Fallopian tube of women with ectopic pregnancy. 351 Similar to that in the non-pregnant Fallopian tube, galectin-1 protein was mainly localized 352 to the stroma and intraepithelial leukocytes were intensely immunostained for galectin-1, 353 while the epithelial galectin-1 immunoreaction was slightly increased in ectopic 354 pregnancies. Although we have previously shown that immune cell populations are 355 increased in the Fallopian tube from women with ectopic pregnancy [Shaw et al. 2011], 356 the number of galectin-1-positive intraepithelial cells did not alter between non 357 pregnancy and ectopic pregnancy (supplementary figure 1), suggesting that there is no 358 significant contribution of galectin-1-positive intraepithelial cells to the development of 359 ectopic pregnancy. In contrast, there seemed to be an increased epithelial immunolabeling 360 more generally in the Fallopian tube from women with ectopic pregnancy. As LGALS1 20

361 mRNA expression in OE-E6/E7 cells was enhanced by co-culture with trophoblast SW71 362 cells, the increased galectin-1 immunoreaction in the Fallopian tube epithelium with 363 ectopic pregnancy may be a result of ectopic blastocyst implantation and regulated by 364 trophoblast products. As sex steroids also can up-regulate galectin-1 in the uterus [Choe 365 et al., 1997] and LGALS1 expression tended to be positively correlated to the serum 366 concentration of female sex steroids in the Fallopian tube as revealed by this study, the 367 prolonged exposure to steroids in early pregnancy is also a potential mechanism for this 368 change. 369 On the other hand, the mRNA expression of LGALS3 was significantly decreased in 370 the Fallopian tube collected from women with ectopic pregnancy compared to that of 371 non-pregnant women during mid-secretory phase. This may be partially due to prolonged 372 progesterone exposure during ectopic pregnancy. It remains possible that altered galectin 373 3 expression is a cause rather than a consequence of tubal ectopic pregnancy. Down 374 regulation of galectin-3 expression in the mouse endometrium was observed at the 375 beginning of pregnancy [Orazizadeh et al., 2013], suggesting that a decreased epithelial 376 galectin-3 is associated with successful embryo implantation. However many researchers 377 have noted an increased expression of galectin-3 in the uterus during pregnancy, and the 378 numbers of implanted embryos decreased when galectin-3 was knocked down selectively 21

379 in mouse endometrium [Yang et al., 2012b]. Therefore the exact role of galectin-3 during 380 normal blastocyst implantation and in the Fallopian tube epithelium with ectopic 381 pregnancy remains uncertain. 382 One interesting observation in the present study was that the subcellular localization 383 of galectin-3 changed in the Fallopian tube epithelium from women with ectopic 384 pregnancy. Cells with galectin-3-positive cilia significantly decreased in number and 385 there was intimate relationship between loss of normal ciliary structure and the nuclear 386 or cytoplasmic translocation of galectin-3. Koch et al. [2010] have reported that galectin 387 3 is a novel centrosome-associated protein and knockout of this gene resulted in the 388 abnormal morphology of primary cilia in the renal epithelial cells. Recently, the same 389 research group demonstrated that galectin-3 at the base of the motile cilia in tracheal 390 ciliated cells plays a crucial role in the maintenance of the coordinated orientation and 391 stabilization of motile cilia [Clare et al., 2014]. Although the function of galectin-3 in the 392 pathogenesis of ectopic pregnancy is still unclear, translocation of galectin-3 from the 393 cilia to cell body could be related to ciliary loss and a decrease of ciliary motility, which 394 is a major cause of the ectopic implantation [Vasquez et al., 1983]. It remains possible 395 that these findings are specific to the Fallopian tubes prone to ectopic implantation, and 396 investigation of the Fallopian tubes damaged by past inflammation would be of interest. 22

397 As galectins are sugar-binding animal lectins, an identification of the ligand 398 glycoconjugates is also important to elucidate the exact role of galectins in the human 399 Fallopian tube. A previous study reported that the binding of biotinylated 400 neoglycoproteins, which contains galectin-recognizing  -galactose and lactose, in the 401 epithelium of the oviduct of rabbits [Biermann et al., 1997]. Because sialylation and 402 fucosylation on terminal galactose of glycoconjugates are important modifications on 403 subtype-specific sugar binding affinity of galectins, to investigate the changes in the 404 glycan structures in the human Fallopian tube would be of interest. 405 In conclusion, herein we describe the detailed expression of galectin-1 and galectin 406 3 in the Fallopian tube from non-pregnant women, and their changes in the Fallopian tube 407 from women with ectopic pregnancy. The differential expression and localization of 408 galectin-1 and galectin-3 suggest the subtype-specific contribution to the homeostasis of 409 human Fallopian tube and to the pathogenesis of ectopic pregnancy. 410 411 412 413 414 23

415 416 417 418 419 420 421 422
423 We are grateful to Ms. Lyndsey Boswell, Dr. Fiona Connolly, Ms. Zety Adin, and Ms. 424 Linda Nicol, The University of Edinburgh, for their kind advice and excellent technical 425 support. We thank the research nurses at Royal Infirmary of Edinburgh for help in tissue 426 collection and for all patients participated in this study. We thank Prof. Yeung, The 427 University of Hong Kong, for the use of the human tubal epithelial cell line. 428 429
430 This study was supported by the Cunningham Trust to WCD. WCD was supported by a 431 Scottish Senior Fellowship from the Scottish Funding Council, AWH by a Clinician 432 Scientist Fellowship from the Medical Research Council, and JN-K by a Postdoctoral 24

433 Fellowship for Research Abroad from Japan Society for the Promotion of Science. 434 435
436 There is nothing to be declared. 437 438 439 440
441 Barondes, S.H., Castronovo, V., Cooper, D.N., Cummings, R.D., Drickamer, K., Feizi, T., 442 443 Gitt, M.A., Hirabayashi, J., Hughes, C., Kasai, K. et al. (1994) Galectins: a family of animal beta-galactoside-binding lectins. Cell 76: 597-598. 444 Barrientos, G., Freitaq, N., Tirado-González, I., Unverdorben, L., Jeschke, U., Thijssen, 445 446 V.L., S.M. Blois (2014) Involvement of galectin-1 in reproduction: past, present and future. Hum Reprod Update 20: 175-193. 447 Biermann, L., Gabius, H.J., H.W. Denker (1997) Neoglycoprotein-binding sites 448 449 (endogenous lectins) in the Fallopian tube, uterus and blastocyst of the rabbit during the preimplantation phase and implantation. Acta Anat 160: 159-171. 450 Blidner, A.G., G.A. Rabinovich (2013) ‘Sweetening’ pregnancy: galectins at the 25

451 fetomaternal interface. Am J Reprod Immunol 69: 369-382. 452 Blois, S.M., Ilarregui, J.M., Tometten, M., Garcia, M., Orsal, A.S., Cordo-Russo, R., 453 454 455 Toscano, M.A., Bianco, G.A., Kobelt, P., Handjiski, B., Tirado, I., Markert, U.R., Klapp, B.F., Poirier, F., Szekeres-Bartho, J., Rabinovich, G.A., P.C. Arck (2007) A pivotal role for galectin-1 in fetomaternal tolerance. Nat Med 13: 1450-1457. 456 Cedeno-Laurent, F., C.J. Dimitroff (2012) Galectin-1 research in T cell immunity: past, 457 present and future. Clin Immunol 142: 107-116. 458 Choe, Y.S., Shim, C., Choi, D., Lee, C.S., Lee, K.K., K. Kim (1997) Expression of 459 460 galectin-1 mRNA in the mouse uterus is under the control of ovarian steroids during blastocyst implantation. Mol Reprod Dev 48: 261-266. 461 Clare, D.K., Magescas, J., Piolot, T., Dumoux, M., Vasque, C., Pichard, E., Dang, T., 462 463 Duvauchelle, B., Poirier, F., D. Delacour (2014) Basal foot MTOC organizes pillar MTs required for coordination of beating cilia. Nat Commun 5: 4888. 464 Duncan, W.C., Shaw, J.L., Burgess, S., McDonald, S.E., Critchley, H.O., A.W. Horne 465 466 467 (2011) Ectopic pregnancy as a model to identify endometrial genes and signaling pathways important in decidulization and regulated by local trophoblast. PLos One 6: e23595. 468 C.M. Farquhar (2005) Ectopic pregnancy. Lancet 366: 583-591. 26

469 Hirabayashi, J., Hashidate, T., Arata, Y., Nishi, N., Nakamura, T., Hirashima, M., 470 471 472 Urashima, T., Oka, T., Futai, M., Muller, W.E., Yagi, F, K. Kasai (2002) Oligosaccharide specificity of galectins: a search by frontal affinity chromatography. Biochim Biophys Acta 1572: 232-254. 473 Hirabayashi, J., K. Kasai (1984) Human placenta beta-galactoside-binding lectin. 474 Purification and some properties. Biochem Biophys Res Commun 122: 938-944. 475 Hirota, Y., Bumum, K.E., Acar, N., Rabinovich, G.A., Daikoku, T., S.K. Dey (2012) 476 477 Galectin-1 markedly reduces the incidence of resorptions in mice missing immunophilin FKBP52. Endocrinology 153: 2486-2493. 478 John, C.M., Jarvis, G.A., Swanson, K.V., Leffler, H., Cooper, M.D., Huflejt, M.E., J.M. 479 480 481 Griffiss (2002) Galectin-3 binds lactosaminylated lipooligosaccharides from Neisseria gonorrhoeae and is selectively expressed by mucosal epithelial cells that are infected. Cell Microbiol 4: 649-662. 482 Kim, M., Kim, S., Kim, H., Kim, H., Joo, H.G., T. Shin (2008) Immunohistochemical 483 484 485 Koch, A., Poirier, F., Jacob, R., D. Delacour (2010) Galectin-3, a novel centrosome 486 localization of galectin-3 in the reproductive organs of the cow. Acta Histochem 110: 473-480. associated protein, required for epithelial morphogenesis. Mol Biol Cell 21: 219-231. 27

487 Lee, Y.L., Lee, K.F., Xu, J.S., Wang, Y.L., Tsao, S.W. W.S. Yeung (2001) Establishment 488 489 490 F.T. Liu (2000) Galectins: a new family of regulators of inflammation. Clin Immunol 97: 491 and characterization of an immortalized human oviductal cell line. Mol Reprod Dev 59: 400-409. 79-88. 492 Nio-Kobayashi, J., Boswell, L., Amano, M., Iwanaga, T., W.C. Duncan (2014) The loss 493 494 of luteal progesterone production in women is associated with a galectin switch via  2,6-sialylation of glycoconjugates. J Clin Endocrinol Metab 99: 4616-4624. 495 Orazizadeh, M., Khorsandi, L., G. Saki (2013) Immunohistochemical assessment of 496 497 galectin-3 during pre-implantation in mouse endometrium. Iran J Reprod Med 11: 119-126. 498 Phillips, B., Knisley, K., Weitlauf, K.D., Dorsett, J., Lee, V., H. Weitlauf (1996) 499 500 Differential expression of two beta-galactoside-binding lectins in the reproductive tracts of pregnant mice. Biol Reprod 55: 548-558. 501 Poirier, F., Timmons, P.M., Chan, C.T., Guénet, J.L., P.W. Rigby (1992) Expression of the 502 503 L14 lectin during mouse embryogenesis suggests multiple roles during pre- and post implantation development. Development 115: 143-155. 504 J.T. Powell (1980) Purification and properties of lung lectin. Rat lung and human lung 28

505 beta-galactoside-binding proteins. Biochem J 187: 123-129. 506 Ramhorst, R.E., Giribaldi, L., Fraccaroli, L., Toscano, M.A., Stupirski, J.C., Romero, 507 508 509 M.D., Durand, E.S., Rubinstein, N., Blaschitz, A., Sedlmayr, P., Genti-Raimondi, S., Fainboim, L., G.A. Rabinovich (2012) Galectin-1 confers immune privilege to human trophoblast: implications in recurrent fetal loss. Glycobiology 22: 1374-1386. 510 Roldán, M.L., P.E. Marini (2014) First evidence of the interaction between delated in 511 512 513 Shaw, J.L., Dey, S.K., Critchley, H.O, A.W. Horne (2010b) Current knowledge of the 514 malignant brain tumor 1 and galectin-3 in the mammalian oviduct. Histochem Cell Biol 141: 181-190. aetiology of human tubal ectopic pregnancy. Hum Reprod Update 16: 432-444. 515 Shaw, J.L., Fitch, P., Cartwright, J., Entrican, G., Schwarze, J., Critchley, H.O., A.W. 516 517 Horne (2011) Lymphoid and myeloid cell populations in the non-pregnant human Fallopian tube and in ectopic pregnancy. J Reprod Immunol 89: 84-91. 518 Straszewski-Chavez, S.L., Abrahams, V.M., Alvero, A.B., Aldo, P.B., Ma, Y., Guller, S., 519 520 521 Than, N.G., Romero, R., Erez, O., Weckle, A., Tarca, A.L., Hotra, J., Abbas, A., Han, 522 Romero, R., G. Mor (2009) The isolation and characterization of a novel telomerase immortalized first trimester trophoblast cell line, Swan 71. Placenta 30: 939-948. Y.M., Kim, S.S., Kusanovic, J.P., Gotsch, F., Hou, Z., Santolaya-Forgas, J., 29

523 524 525 526 Benirschke, K., Papp, Z., Grossman, L.I., Goodman, M., D.E. Wildman (2008) Emergence of hormonal and redox regulation of galectin-1 in placental mammals: implication in maternal-fetal immune tolerance. Proc Natl Acad Sci U S A 105: 15819-15824. 527 Ulziibat, S., Ejima, K., Shibata, Y., Hishikawa, Y., Kitajima, M., ujishita, A., Ishimaru, T., 528 T. Koji (2006) Identification of estrogen receptor  -positive intraepithelial 529 lymphocytes and their possible roles in normal and tubal pregnancy oviducts. Human 530 Reprod. 21: 2281-2289. 531 Varma, R., J. Gupta (2012) Tubal ectopic pregnancy. Clin Evid (Online) pii:1406. 532 Vasquez, G., Winston, R.M., I.A. Brosens (1983) Tubal mucosa and ectopic pregnancy. 533 Br J Obstet Gynaecol 90: 468-474. 534 J.J. Walker (2007) Ectopic pregnancy. Clin Obstet Gynecol 50: 89-99. 535 Yang, H., Lei, C., Cheng, C., Feng, Y., Zhang, W., Petracco, R.G., S. Sak (2012a) The 536 537 538 Yang, H., Lei, C., W. Zhang (2012b) Expression of galectin-3 in mouse endometrium and 539 antiapoptotic effect of galectin-3 in human endometrial cells under the regulation of estrogen and progesterone. Biol Reprod 87: 1-7 its effect during embryo implantation. Reprod Biomed Online 24: 116-122. 540 Yang, H., Taylor, H.S., Lei, C., Cheng, C., W. Zhang (2011) Hormonal regulation of 30

541 542 Yang, R.Y., Rabinovich, G.A., F.T. Liu (2008) Galectins: structure, function and 543 galectin-3 in trophoblasts and its effects on endometrium. Reprod Sci 18: 1118-1127. therapeutic potential. Expert Rev Mol Med 10: e17 544 von Wolff, M., Wang, X., Gabius, H.J., T. Strowitzki (2005) Galectin fingerprinting in 545 546 human endometrium and decidua during the menstrual cycle and in early gestation. Mol Hum Reprod 11: 189-194. 547 548 549 550 551 552 553 554 555 556 557 558 31

559 560 561 562 563
564
565
566 mainly found in the stroma, and the epithelium shows a limited immunoreactivity for 567 galectin-1 (A, B). Round cells within the epithelium are positive in galectin-1 568 immunoreaction (arrowheads in B, C). Galectin-3 is mainly localized in the apical region 569 of the epithelium, and the stroma is weakly immunoreactive for galectin-3 (D). At higher 570 magnification, galectin-3 immunoreactivity is found in the cilia of ciliated cells (E, F) 571 and the apical cytoplasm of secretory cells (arrows in E). Occasionally, the nuclear and 572 cytoplasmic staining for galectin-3 is observed in both types of epithelial cells (asterisk 573 in F shows an example in a ciliated cell). Dual immunostaining for galectin-3 (red) and 574  -tubulin (green) as a maker of cilia clearly demonstrates the positive immunoreaction 575 for galectin-3 in the  -tubulin-positive cilia of ciliated cells (G-I) and the apical 576 cytoplasm of  -tubulin-negative non-ciliated secretory cells (arrows in G-I). The 32

577 localization of the immunoreactivities did not change between the menstrual phases, and 578 the representative tissue sections at the mid-secretory phase are used for illustration. Insert 579 in A is a control section where the primary antibody was omitted.  -tub:  -tubulin, G1: 580 galectin-1, G3: galectin-3. 581
582
583 stroma of the Fallopian tube from women with ectopic pregnancy (EP) (A) as it is in the 584 Fallopian tube from non-pregnant women. At higher magnification, the immunoreactivity 585 for galectin-1 slightly increased with entire length of epithelium (B). Intraepithelial round 586 cells are also positive in galectin-1 immunroeaction (arrowheads in B). Galectin-3 is 587 localized in both the epithelium and the stroma (C). At higher magnification, the galectin 588 3 immunoreactivity is found in the cilia of ciliated cells as it is in the non-pregnant women 589 (D) but the nuclear galectin-3 staining in both types of epithelial cells is more frequently 590 observed in the Fallopian tube from women with ectopic implantation (asterisks in D). 591 Galectin-3 immunoreactivity tends to gather at the apical plasma membrane of  -tubulin 592 negative non-ciliated secretory cells (arrows in D and E). Cells with abundant 593 cytoplasmic galectin-3 lose the normal ciliary structure (arrows in F and G). Dual 594 immunostaining for galectin-3 and  -tubulin clearly shows the decreased  -tubulin 33

595 immunoreactive ciliary structures in cells with abundant cytoplasmic galectin-3 (arrows 596 in H-J) whereas an intact  -tubulin-positive ciliary structure is found in cells without 597 cytoplasmic galectin-3 (arrowheads in H-J). The percentage of epithelial cells with 598 galectin-3-immunoreactive intact cilia tends to decrease in the Fallopian tube with EP (K). 599 On the other hand, the number of cells with nuclear staining for galectin-3 is significantly 600 increased in EP (L). The number of cells with cytoplasmic staining for galectin-3 does 601 not alter between two groups (M).  -tub:  -tubulin, G1: galectin-1, G3: galectin-3, n.s.: 602 not significant. **P<0.01. 603 604
605
606 positively correlated to the serum concentration of estradiol (n=35) and progesterone 607 (n=42) (A, B). On the other hand, the expression of LGALS3 seems to be negatively 608 correlated to these female steroids (C, D). The expression of LGALS1 and LGALS3 in the 609 Fallopian tube is negatively correlated (n=56) (E). n.s.: not significant. 610 611
612
34

613 expression of LGALS1 is significantly increased in OE-E6/E7 cells co-cultured with 614 SW71 cells (A) while LGLAS3 mRNA expression does not alter (B). Cont: control culture 615 of OE-E6/E7 cells only. ****P<0.0001. 35